CECW-EH-Y

Engineer Manual

1110-2-1420

Department of the ArmyU.S. Army Corps of Engineers

Washington, DC 20314-1000

EM 1110-2-1420

31 October 1997

Engineering and Design

HYDROLOGIC ENGINEERING REQUIREMENTS FOR RESERVOIRS

Distribution Restriction StatementApproved for public release; distribution is

unlimited.

DEPARTMENT OF THE ARMY EM 1110-2-1420U.S. Army Corps of Engineers

CECW-EH-Y Washington, DC 20314-1000

ManualNo. 1110-2-1420 31 October 1997

Engineering and DesignHYDROLOGIC ENGINEERING REQUIREMENTS

FOR RESERVOIRS

1. Purpose. This manual provides guidance to field office personnel for hydrologic engineeringinvestigations for planning and design of reservoir projects. The manual presents typical studymethods; however, the details of procedures are only presented if there are no convenient referencesdescribing the methods. Also, publications that contain the theoretical basis for the methods arereferenced. Many of the computational procedures have been automated, and appropriate referencesare provided.

2. Applicability. This manual applies to all HQUSACE elements and USACE commands having civilworks responsibilities.

FOR THE COMMANDER:

OTIS WILLIAMSColonel, Corps of EngineersChief of Staff

i

DEPARTMENT OF THE ARMY EM 1110-2-1420U.S. Army Corps of Engineers

CECW-EH-Y Washington, DC 20314-1000

ManualNo. 1110-2-1420 31 October 1997

Engineering and DesignHYDROLOGIC ENGINEERING REQUIREMENTS

FOR RESERVOIRS

Table of Contents

Subject Paragraph Page Subject Paragraph Page

Chapter 1 Managing Competitive andIntroduction Complementary Functions . . . . . . . . 3-3 3-2Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1 1-1Applicability . . . . . . . . . . . . . . . . . . . . . . 1-2 1-1References . . . . . . . . . . . . . . . . . . . . . . . . 1-3 1-1Related H&H Guidance . . . . . . . . . . . . . . 1-4 1-1

Part 1 Hydrologic Engineering Concepts for Reservoirs

Chapter 2Reservoir PurposesCongressional Authorizations . . . . . . . . . 2-1 2-1Reservoir Purposes . . . . . . . . . . . . . . . . . 2-2 2-2Flood Control . . . . . . . . . . . . . . . . . . . . . 2-3 2-2Navigation . . . . . . . . . . . . . . . . . . . . . . . . 2-4 2-3Hydroelectric Power . . . . . . . . . . . . . . . . 2-5 2-3Irrigation . . . . . . . . . . . . . . . . . . . . . . . . . 2-6 2-4Municipal and Industrial Water Supply . . . . . . . . . . . . . . . . . . . . . 2-7 2-4Water Quality . . . . . . . . . . . . . . . . . . . . . 2-8 2-4Fish and Wildlife . . . . . . . . . . . . . . . . . . . 2-9 2-5Recreation . . . . . . . . . . . . . . . . . . . . . . . . 2-10 2-5Water Management Goals and Objectives . . . . . . . . . . . . . . . . . . . . 2-11 2-6

Chapter 3 Chapter 5Multiple-Purpose Reservoirs Hydrologic EngineeringHydrologic Studies for DataMultipurpose Projects . . . . . . . . . . . . . . . 3-1 3-1Relative Priorities of Topographic Data . . . . . . . . . . . . . . . 5-2 5-3 Project Functions . . . . . . . . . . . . . . . . . . 3-2 3-1

Operating Concepts . . . . . . . . . . . . . . 3-4 3-2Construction and Physical Operation . . . . . . . . . . . . . . . . . . . . 3-5 3-3General Study Procedure . . . . . . . . . . 3-6 3-4

Chapter 4Reservoir SystemsIntroduction . . . . . . . . . . . . . . . . . . . . 4-1 4-1System Description . . . . . . . . . . . . . . 4-2 4-1Operating Objectives and Criteria . . . . . . . . . . . . . . . . . . . 4-3 4-2System Simulation . . . . . . . . . . . . . . . 4-4 4-2Flood-Control Simulation . . . . . . . . . 4-5 4-2Conservation Simulation . . . . . . . . . . 4-6 4-3System Power Simulation . . . . . . . . . 4-7 4-3Determination of Firm Yield . . . . . . . 4-8 4-4Derivation of Operating Criteria . . . . . . . . . . . . . . . . . . . . . . 4-9 4-4System Formulation Strategies . . . . . 4-10 4-5General Study Procedure . . . . . . . . . . 4-11 4-7

Part 2 Hydrologic Analysis

Meteorological Data . . . . . . . . . . . . . 5-1 5-1

Streamflow Data . . . . . . . . . . . . . . . . 5-3 5-4

EM 1110-2-142031 Oct 97

ii

Subject Paragraph Page Subject Paragraph Page

Adjustment of Streamflow Flood Volume Frequencies . . . . . . . . . 10-3 10-2 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-4 5-5Simulation of Operation Constraints Streamflow Data . . . . . . . . . . . . . . . . . . 5-5 5-8

Chapter 6 Determinations . . . . . . . . . . . . . . . . . 10-6 10-4Hydrologic Frequency Spillways . . . . . . . . . . . . . . . . . . . . . . . 10-7 10-4Determinations Flood-Control SystemIntroduction . . . . . . . . . . . . . . . . . . . . . . . 6-1 6-1Duration Curves . . . . . . . . . . . . . . . . . . . 6-2 6-1Flood-Frequency Determinations . . . . . . . 6-3 6-1Estimating Frequency Curves . . . . . . . . . 6-4 6-3Effect of Basin Developments Conservation Storage on Frequency Relations . . . . . . . . . . . . . 6-5 6-4Selection of Frequency Data . . . . . . . . . . 6-6 6-4Climatic Variations . . . . . . . . . . . . . . . . . 6-7 6-5Frequency Reliability Analyses . . . . . . . . 6-8 6-6Presentation of Frequency Fish and Wildlife . . . . . . . . . . . . . . . . . 11-4 11-4 Analysis Results . . . . . . . . . . . . . . . . . . 6-9 6-6

Chapter 7 Water Quality Requirements . . . . . . . . 11-7 11-11Flood-Runoff Analysis Reservoir Water QualityIntroduction . . . . . . . . . . . . . . . . . . . . . . . 7-1 7-1Flood Hydrograph Analysis . . . . . . . . . . . 7-2 7-1Hypothetical Floods . . . . . . . . . . . . . . . . . 7-3 7-1

Chapter 8 Introduction . . . . . . . . . . . . . . . . . . . . . 12-1 12-1Water Surface Profiles Problem Description . . . . . . . . . . . . . . 12-2 12-1Introduction . . . . . . . . . . . . . . . . . . . . . . . 8-1 8-1Steady Flow Analysis . . . . . . . . . . . . . . . 8-2 8-1Unsteady Flow Analysis . . . . . . . . . . . . . 8-3 8-2Multidimensional Analysis . . . . . . . . . . . 8-4 8-2Movable-Boundary Profile Analysis . . . . . 8-5 8-2

Chapter 9 Storage Limitations . . . . . . . . . . . . . . . 12-9 12-4Reservoir Sediment Analysis Effects of ConservationIntroduction . . . . . . . . . . . . . . . . . . . . . . . 9-1 9-1Sediment Yield Studies . . . . . . . . . . . . . . 9-2 9-1Reservoir Sedimentation Problems . . . . . 9-3 9-1Downstream Sediment Problems . . . . . . . 9-4 9-2Sediment Water Quality . . . . . . . . . . . . . 9-5 9-3Sediment Investigations . . . . . . . . . . . . . . 9-6 9-3

Part 3 Reservoir Storage Requirements

Chapter 10Flood-Control StorageGeneral Considerations . . . . . . . . . . . . . . 10-1 10-1Regulated Release Rates . . . . . . . . . . . . . 10-2 10-1

Hypothetical Floods . . . . . . . . . . . . . . . 10-4 10-3

and Criteria . . . . . . . . . . . . . . . . . . . . 10-5 10-3Storage Capacity

Formulation . . . . . . . . . . . . . . . . . . . . 10-8 10-5General Study Procedure . . . . . . . . . . . 10-9 10-6

Chapter 11

General Considerations . . . . . . . . . . . . 11-1 11-1Water Supply . . . . . . . . . . . . . . . . . . . . 11-2 11-2Navigation and Low-Flow Augmentation . . . . . . . . . . . . . . . . . . 11-3 11-3

Hydroelectric Power . . . . . . . . . . . . . . 11-5 11-5Water Quality Considerations . . . . . . . 11-6 11-11

Management . . . . . . . . . . . . . . . . . . . 11-8 11-12

Chapter 12Conservation Storage Yield

Study Objectives . . . . . . . . . . . . . . . . . 12-3 12-1Types of Procedures . . . . . . . . . . . . . . . 12-4 12-2Factors Affecting Selection . . . . . . . . . 12-5 12-3Time Interval . . . . . . . . . . . . . . . . . . . 12-6 12-3Physical Constraints . . . . . . . . . . . . . . 12-7 12-4Priorities . . . . . . . . . . . . . . . . . . . . . . . 12-8 12-4

and Other Purposes . . . . . . . . . . . . . . 12-10 12-6Simplified Methods . . . . . . . . . . . . . . . 12-11 12-8Detailed Sequential Analysis . . . . . . . . 12-12 12-9Effects of Water Deficiencies . . . . . . . . 12-13 12-11Shortage Index . . . . . . . . . . . . . . . . . . 12-14 12-12General Study Procedures . . . . . . . . . . 12-15 12-12

Chapter 13Reservoir SedimentationIntroduction . . . . . . . . . . . . . . . . . . . . . 13-1 13-1Reservoir Deposition . . . . . . . . . . . . . . 13-2 13-1Distribution of Sediment Deposits in the Reservoir . . . . . . . . . . 13-3 13-2

EM 1110-2-142031 Oct 97

iii

Subject Paragraph Page Subject Paragraph Page

Part 4 Hydrologic Engineering Studies for Reservoirs

Chapter 14Spillways and Outlet WorksFunction of Spillways and Outlet Works . . . . . . . . . . . . . . . . . 14-1 14-1Spillway Design Flood . . . . . . . . . . . . . . . 14-2 14-1Area and Capacity of the Reservoir . . . . . 14-3 14-2Routing the Spillway Design Flood . . . . . 14-4 14-2Sizing the Spillway . . . . . . . . . . . . . . . . . 14-5 14-3Outlet Works . . . . . . . . . . . . . . . . . . . . . . 14-6 14-4

Chapter 15Dam Freeboard RequirementsBasic Considerations . . . . . . . . . . . . . . . . 15-1 15-1Wind Characteristics over Reservoirs . . . . . . . . . . . . . . . . . . . 15-2 15-1Computation of Wave and Wind Tide Characteristics . . . . . . . . . . . . . . . . 15-3 15-2Wave Runup on Sloping Embankment . . . . . . . . . . . . . . 15-4 15-3Freeboard Allowances for Wave Action . . . . . . . . . . . . . . . . . . 15-5 15-4

Chapter 16Dam Break AnalysisIntroduction . . . . . . . . . . . . . . . . . . . . . . . 16-1 16-1Dam Breach Analysis . . . . . . . . . . . . . . . 16-2 16-1Dam Failure Hydrograph . . . . . . . . . . . . . 16-3 16-2Dam Break Routing . . . . . . . . . . . . . . . . . 16-4 16-3Inundation Mapping . . . . . . . . . . . . . . . . 16-5 16-4

Chapter 17Channel Capacity StudiesIntroduction . . . . . . . . . . . . . . . . . . . . 17-1 17-1Downstream Channel Capacity . . . . . . . . . . . . . . . . . . . . . 17-2 17-1Stream Rating Curve . . . . . . . . . . . . . 17-3 17-2Water Surface Profiles . . . . . . . . . . . . 17-4 17-2

Chapter 18Real Estate and Right-of-WayStudiesIntroduction . . . . . . . . . . . . . . . . . . . . 18-1 18-1Definition of Terms . . . . . . . . . . . . . . 18-2 18-1Real Estate Acquisition Policies for Reservoirs . . . . . . . . . . . 18-3 18-1Hydrologic Evaluations . . . . . . . . . . . 18-4 18-2Water Surface Profile Computations . . . . . . . . . . . . 18-5 18-2Development of an Envelope Curve . . . . . . . . . . . . . . . . 18-6 18-2Evaluations to Determine Guide Taking Lines (GTL) . . . . . . . 18-7 18-2Acquisitions of Lands for Reservoir Projects . . . . . . . . . . . . . . 18-8 18-3Acquisition of Lands Downstream from Spillways for Hydrologic Safety Purposes . . . . . . . . . . . . . . . . . . . . . 18-9 18-4

Appendix AReferences

EM 1110-2-142031 Oct 97

1-1

Chapter 1Introduction

1-1. Purpose

This manual provides guidance to field office personnel forhydrologic engineering investigations for planning anddesign of reservoir projects. The manual presents typicalstudy methods; however, the details of procedures are onlypresented if there are no convenient references describingthe methods. Also, publications that contain the theoreticalbasis for the methods are referenced. Many of the com-putat ional procedures have been automated, and appropri-ate references are provided.

1-2. Applicability

T his manual applies to all HQUSACE elements andUSACE commands having civil works responsibilities.

a. Scope. This manual provides information onhydrologic engineering studies for reservoir projects.These studies can utilize many of the hydrologic engineer-ing methods described in the manuals listed in para-graph 1-4. Hydraulic design of project features are notincluded here; they are presented in a series of hydraulicdesign manuals.

b. Organization. This manual is divided into fourparts. Part 1 provides basic hydrologic concepts for reser-voi rs . Reservoir purposes and basic hydrologic concernsand methods are presented. Part 2 describes hydrologicdata and analytical methods. Part 3 covers storagerequirements for various project purposes, and the last,Part 4, covers hydrologic engineering studies.

1-3. References

Required and related publications are listed in Appendix A.

1-4. Related H&H Guidance

a. Engineer manuals. This engineer manual (EM)relies on, and references, technical information presented inother guidance documents. Some of the key EM's forreservoir studies are listed below. Additionally, there arerelated documents on hydraulic design for project featuresassociated with reservoir projects. This document does notpresent hydraulic design concepts.

EM 1110-2-1201 Reservoir Water Quality Analysis

EM 1110-2-1415 Hydrologic Frequency Analysis

EM 1110-2-1416 River Hydraulics

EM 1110-2-1417 Flood-Runoff Analysis

EM 1110-2-1602 Hydraulic Design of ReservoirOutlet Works

EM 1110-2-1603 Hydraulic Design of Spillways

EM 1110-2-1701 Hydropower

EM 1110-2-3600 Management of Water ControlSystems

EM 1110-2-4000 Sedimentation Investigation ofRivers and Reservoirs

These manuals provide the technical background for studyprocedures that are frequently required for reservoir analy-sis. Specific references to these EM's are made throughoutthis document.

b. Engineer regulations. There are several engineerregulations (ER) which prescribe necessary studies asso-ciated with reservoir projects. The most relevant ER's arelisted below.

ER 1110-2-1460 Hydrologic EngineeringManagement

ER 1110-2-7004 Hydrologic Analysis for WatershedRunoff

ER 1110-2-7005 Hydrologic Engineering Require-ments for Flood Damage ReductionStudies

ER 1110-2-7008 Hydrologic Engineering in DamSafety

ER 1110-2-7009 Hydrologic Data Collection andManagement

ER 1110-2-7010 Local Protection - Safety/Workability

ER 1110-8-2(FR) Inflow Design Floods for Damsand Reservoirs

T hese and other regulations should be consulted prior toperforming any hydrologic engineering study for reservoirs.A current index of regulations should be consulted for newand updated regulations.

EM 1110-2-142031 Oct 97

PART 1

HYDROLOGIC ENGINEERING CONCEPTSFOR RESERVOIRS

EM 1110-2-142031 Oct 97

2-1

Figure 2-1. Purposes and programs authorized by Congress

Chapter 2Reservoir Purposes

2-1. Congressional Authorizations

a. Authorization of purposes. The United StatesCongress authorizes the purposes served by U.S. ArmyCorps of Engineers reservoirs at the time the authorizinglegislation is passed. Congress commonly authorizes aproject Asubstantially in accordance with the recommen-dations of the Chief of Engineers,@ as detailed in a separatecongressional document. Later, additional purposes are c. Additional authorization. Figure 2-1 illustratessometimes added, deleted, or original purposes modified, how additional authorizations have increased the number ofby subsequent congressional action. When the original purposes for which the Corps is responsible both inpurposes are not seriously affected, or structural or planning and managing water resource developmentoperational changes are not major, modifications may projects. The first authorizations were principally forbe made by the Chief of Engineers (Water Supply Act navigation, hydroelectric power, and flood control. Later1958). authorizations covered a variety of conservation purposes

and programs. During drought when there is a water

b. General legislation. Congress also passes generallegislation that applies to many projects. The 1944

Flood Control Act, for example, authorizes recreationalfacilities at water resource development projects. Thisauthority has made recreation a significant purpose at manyCorps reservoirs. Similar general legislation has beenpassed to enhance and promote fish and wildlife (1958) andwetlands (1976). The Water Resource Development Act of1976 authorizes the Chief of Engineers, under certainconditions, to plan and establish wetland areas as part of anauthorized water resource development project. Achronology of the congressional legislation authorizingvarious purposes and programs is shown in Figure 2-1(USACE 1989).

shortage, all purposes compete for available water and areaffected by the shortage. The more purposes and programs

EM 1110-2-142031 Oct 97

2-2

Figure 2-2. Typical storage allocation in reservoirs

there are to serve, the greater the potential for conflict, and included in this manual, are nonetheless important in waterthe more complex the task of managing existing supplies. control management.AAuthorized and Operating Purposes of Corps of Engi-neers' Reservoirs@ (USACE 1992) lists the purposes for d. Inactive capacity. The inactive space is com-which Corps operated reservoirs were authorized and are monly used to maintain a minimum pool and for sedimentoperated. storage. Sediment storage may affect all levels of the

2-2. Reservoir Purposes

a. Storage capacity. A cross section of a typical ation, fish and wildlife, and water quality.reservoir is shown in Figure 2-2. The storage capacity isdivided into three zones: exclusive, multiple-purpose, and e. Storage space allocation. Reservoir storageinactive. While each Corps reservoir is unique both in its space may not be allocated to specific conservation pur-allocation of storage space and in its operation, the divi- poses. Rather, reservoir releases can serve several pur-sion of storage illustrated by Figure 2-2 is common. poses. However, the amount of water needed to serve

b. Exclusive capacity. The exclusive space is events. Each reservoir's water control plan defines thereserved for use by a single purpose. Usually this is flood goals of regulation. Usually, a compromise is achieved tocontrol, although navigation and hydroelectric power have best utilize the storage space to reduce flooding from bothexclusive space in some reservoirs. The exclusive capac- major and minor flood events. In special circumstancesity reserved for flood control is normally empty. Some where reservoir inflows can be forecast several days orreservoirs with exclusive flood control space have no weeks in advance (for example, when the runoff occursmultiple-purpose pool but have a nominal inactive pool from snowmelt), for the best utilization of storage space,that attracts recreational use. Recreational use is also the degree of control for a particular flood event may becommon on pools originally established exclusively for determined on the basis of forecasts. When runoff isnavigation. seasonal, the amount of designated flood control storage

c. Multiple-purpose capacity. Multiple-purpose voirs for multiple-purpose regulation.storage serves a variety of purposes. These purposesinclude both seasonal flood control storage, often in addi- b. Releases. Flood control releases are based upontion to exclusive storage, and conservation. Conservation the overall objectives to limit the discharges at the down-purposes include: navigation, hydroelectric power, water stream control points to predetermined damage levels.supply, irrigation, fish and wildlife, recreation, and water The regulation must consider the travel times caused byquality. Other conservation purposes such as wetlands, storage effects in the river system and the local inflowsgroundwater supply and endangered species, while not between the reservoir and downstream control points.

reservoir storage. Also, the inactive capacity may some-times be used during drought when it can provide limitedbut important storage for water supply, irrigation, recre-

each purpose varies. During drought, with limited multi-ple-purpose storage available, the purposes requiringgreater releases begin to compete with purposes requiringless. For example, if the greater releases are not made, thestorage would last longer for the purposes served by thelesser releases.

f. General information. A brief description ofproject purposes is presented below. Additional detail anda discussion of reservoir operating procedures may befound in EM 1110-2-3600, from which the followingsections are excerpts.

2-3. Flood Control

a. Utilizing storage space. Reservoirs are designedto minimize downstream flooding by storing a portion orsometimes the entire runoff from minor or moderate flood

space may be varied seasonally to better utilize the reser-

EM 1110-2-142031 Oct 97

2-3

c. Intervening tributary and downstream damage b. Waterflow requirements. Navigation locks locatedareas. A multiple-reservoir system is generally regulated at dams on major rivers generally have sufficient waterfor flood control to provide flood protection both in inter- from instream flows to supply lockage water flowvening tributary areas and at downstream main stem dam- requirements. Navigation requirements for downstream useage areas. The extent of reservoir regulation required for in open river channels may require larger quantities ofprotecting these areas depends on local conditions of flood water over a long period of time (from several months to adamage, uncontrolled tributary drainage, reservoir storage year), to maintain water levels for boat or barge trans-capacity, and the volume and time distribution of reservoir portation. Usually, water released from reservoirs forinflows. Either the upstream or downstream requirements navigation is also used for other purposes, such as hydro-may govern the reservoir regulation, and usually the electric power, low-flow augmentation, water quality,optimum regulation is based on the combination of the two. enhancement of fish and wildlife, and recreation. Seasonal

d. Coordinated reservoir regulation. Water control define the use of water for navigation. The amount ofwith a system of reservoirs can incorporate the concept of a stored water to be released depends on the conditions ofbalanced reservoir regulation, with regard to filling the water storage in the reservoir system and downstreamreservoirs in proportion to each reservoir's flood control requirements or goals for low-flow augmentation, as well ascapability, while also considering expected inflows and factors related to all uses of the water in storage.downstream channel capacities. Evacuation of flood waterstored in a reservoir system must also be accomplished on acoordinated basis. Each reservoir in the system is drawndown as quickly as possible, considering conditions atcontrol points, to provide space for controlling futurefloods. The objectives for withdrawal of water in thevarious zones of reservoir storage are determined tominimize the risk of encroaching into the flood controlstorage and to meet other project requirements. Sometimesthe lower portion of the flood control pool must beevacuated slower to transition to a lower flow to minimizebank caving and allow channel recovery.

2-4. Nav igation

a. Navigational requirements. Problems related tothe management of water for navigation use vary widelyamong river basins and types of developments. Controls t ructures at dams, or other facilities where navigation isone of the project purposes, must be regulated to providerequired water flows and/or to maintain project navigationdepths. Navigational requirements must be integrated withother water uses in multiple-purpose water resourcesystems. In the regulation of dams and reservoirs, thenavigational requirements involve controlling water levelsin the reservoirs and at downstream locations, and provid-ing the quantity of water necessary for the operation oflocks. There also may be navigational constraints in theregulation of dams and reservoirs with regard to rates ofchange of water surface elevations and outflows. There arenumerous special navigational considerations that mayinvolve water control, such as ice, undesirable currents andwater flow patterns, emergency precautions, boating events,and launchings.

or annual water management plans are prepared which

c. Using water for lockage. Navigational constraintsare also important for short-term regulation of projects tomeet all requirements. In some rivers, supply of water forlockage is a significant problem, particularly during periodsof low flow or droughts. The use of water for lockage isgenerally given priority over hydropower or irrigationusages. However, this is dependent on the storage allocatedto each purpose. In critical low-water periods, acurtailment of water use for lockage may be instituted byrestricting the number of locks used, thereby conserving theutilization of water through a more efficient use of thenavigation system. Water requirements for navigationcanals are sometimes based on lockage and instream flowsas necessary to preserve water quality in the canal.

2-5. Hydroelectric Power

a. Reservoir project categories. Reservoir projectswhich incorporate hydropower generally fall into twodistinct categories: storage reservoirs which have sufficientcapacity to regulate streamflow on a seasonal basis and run-of-river projects where storage capacity is minor relative tothe volume of flow. Most storage projects are multiple-purpose. Normally, the upstream reservoirs includeprovisions for power production at the site, as well as forrelease of water for downstream control. Run-of-riverhydropower plants are usually developed in connectionwith navigation projects.

b. Integration and control of a power system.Integration and control of a major power system involvinghydropower resources is generally accomplished by acent ralized power dispatching facility. This facility

EM 1110-2-142031 Oct 97

2-4

contains the equipment to monitor the entire power system c. Meeting irrigation demands. The general modeoperation, including individual plant generation, substation for regulation of reservoirs to meet irrigation demands is tooperation, transmission line operation, power loads and capture all runoff in excess of minimum flow demands andrequirements by individual utilities and other bulk power water rights during the spring and early summer. Thisusers, and all factors related to the electrical system control usually results in refilling the reservoirs prior to thefor real-time operation. The dispatching center is manned irrigation demand season. The water is held in storage untilon a continuous basis, and operations monitor and control the natural flow recedes to the point where it is no longer ofthe flow of power through the system, rectify outages, and sufficient quantity to meet all demands for downstreamperform all the necessary steps to ensure the continuity of i rrigat ion. At that time, the release of stored water frompower system operation in meeting system loads. reservoirs is begun and continued on a demand basis until

c. Regulation of a hydropower system. Regulation of October). During the winter, projects release water ashydropower systems involves two levels of control: required for instream flows, stock water, or other projectscheduling and dispatching. The scheduling function is purposes.performed by schedulers who analyze daily requirementsfor meeting power loads and resources and all other projectrequirements. Schedules are prepared and thoroughlycoordinated to meet water and power requirements of thesystem as a whole. Projections of system regulation, whichindicate the expected physical operation of individualplants and the system as a whole, are prepared for one tofi ve days in advance. These projections are updated on adai l y or more frequent basis to reflect the continuouslychanging power and water requirements.

2-6. Irrigation the right to withdraw water from the lake or to order

a. Irrigation diversion requirements. Irrigation water subject to Federal restrictions with regard to overalldiverted from reservoirs, diversion dams, or natural river regulation of the project and to the extent of availablechannels is controlled to meet the water duty requirements. storage space.T he requirements vary seasonally, and in most irrigatedareas in the western United States, the agricultural growing b. T emporary withdrawal. In times of drought,season begins in the spring months. The diversion special considerations may guide the regulation of projectsrequirements gradually increase as the summer progresses, with regard to water supply. Adequate authority to permitreaching their maximum amounts in July or August. They temporary withdrawal of water from Corps projects isthen recede to relatively low amounts by late summer. By contained in 31 U.S.C. 483a (HEC 1990e). Such with-the end of the growing season, irrigation diversions are drawal requires a fee that is sufficient to recapture lostterminated, except for minor amounts of water that may be project revenues, and a proportionate share of operation,necessary for domestic use, stock water, or other purposes. maintenance, and major replacement expenses.

b. Irrigation as project purpose. Corps of Engineers'reservoir projects have been authorized and operatedprimarily for flood control, navigation, and hydroelectricpower. However, several major Corps of Engineersmultiple-purpose reservoir projects include irrigation as aproject purpose. Usually, water for irrigation is suppliedfrom reservoir storage to augment the natural streamflow asrequired to meet irrigation demands in downstream areas.In some cases, water is diverted from the reservoir bygravity through outlet facilities at the dam which feeddirectly into irrigation canals. At some of the run-of-riverpower or navigation projects, water is pumped directly fromthe reservoir for irrigation purposes.

the end of the growing season (usually September or

2-7. Municipal and Industrial Water Supply

a. Municipal and industrial use. Regulation ofreservoirs for municipal and industrial (M&I) water supplyis performed in accordance with contractual arrangements.S torage rights of the user are defined in terms of acre-feetof s tored water and/or the use of storage space betweenfixed limits of reservoir levels. The amount of storagespace is adjusted to account for change in the total reservoircapacity that is caused by sediment deposits. The user has

releases to be made through the outlet works. This is

2-8. Water Quality

a. Goal and objective. Water quality encompassesthe physical, chemical, and biological characteristics ofwater and the abiotic and biotic interrelationships. Thequality of the water and the aquatic environment is signi-ficantly affected by management practices employed by thewater control manager. Water quality control is anauthorized purpose at many Corps of Engineers reservoirs.However, even if not an authorized project purpose, waterqual i ty is an integral consideration during all phases of aproject's life, from planning through operation. Theminimum goal is to meet State and Federal water quality

EM 1110-2-142031 Oct 97

2-5

standards in effect for the lakes and tailwaters. The oper- levels present both problems and opportunities to the waterat ing objective is to maximize beneficial uses of the control manager with regard to fishery management. Theresources through enhancement and nondegradation of seasonal fluctuation that occurs at many flood controlwater quality. reservoirs, and the daily fluctuations that occur with

b. Release requirements. Water quality releases for shoreline vegetation and subsequent shoreline erosion,downstream control have both qualitative and quantitative water quality degradation and loss of habitat. Adverserequirements. The quality aspects relate to Corps' policy impacts of water level fluctuations also include loss ofand objectives to meet state water quality standards, main- shoreline shelter and physical disruption of spawning andtain present water quality where standards are exceeded, nests.and maintain an acceptable tailwater habitat for aquatic life.T he Corps has responsibility for the quality of waterdischarged from its projects. One of the most importantmeasures of quality is quantity. At many projects autho-ri zed for water quality control, a minimum flow at somedownstream control point is the primary water qualityobjective. Other common objectives include temperature,dissolved oxygen, and turbidity targets at downstreamlocations.

c. Coordinated regulation. Coordinated regulation ofmultiple reservoirs in a river basin is required to maximizebenefits beyond those achievable with individual projectregulation. System regulation for quantitative aspects, suchas flood control and hydropower generation, is a widelyaccepted and established practice, and the same principleappl ies to water quality concerns. Water qualitymaintenance and enhancements may be possible throughcoordinated system regulation. This applies to all facets ofquality from the readily visible quantity aspect to traditionalconcerns such as water temperature and dissolved oxygencontent.

d. System regulation. System regulation for waterqual i t y is of most value during low-flow periods whenavai lable water must be used with greatest efficiency toavoid degrading lake or river quality. Seasonal watercontrol plans are formulated based on current and fore-casted basin hydrologic, meteorologic and quality condi-tions, reservoir status, quality objectives and knowledge ofwater quality characteristics of component parts of thesystem. Required flows and qualities are then apportionedto the individual projects, resulting in a quantitatively andqualitatively balanced system. Computer programs capableof simulating reservoir system regulation for water qualityprovide useful tools for deriving and evaluating watercontrol alternatives.

2-9. Fish and Wildlife control plan. The goals may be minimum project outflows

P roject regulation can influence fisheries both in the reser- boat ing or fishing. Of special importance is minimizingvoi r pool and downstream. One of the most readily any danger that might result from changing conditions ofobservable influences of reservoir regulation is reservoir outflows which would cause unexpected rise or fall in riverpool fluctuation. Periodic fluctuations in reservoir water levels. Also, river drifting is becoming an important

hydropower operation often result in the elimination of

2-10. Recreation

a. Reservoir level. Recreational use of the reservoirsmay extend throughout the entire year. Under mostcircumstances, the optimum recreational use of reservoirswould require that the reservoir levels be at or near fullconservation pool during the recreation season. The degreeto which this objective can be met varies widely, dependingupon the regional characteristics of water supply, runoff,and the basic objectives of water regulation for the variousproject purposes. Facilities constructed to enhance therecreational use of reservoirs may be designed to beoperable under the planned reservoir regulation guidecurves on water control diagrams, which reflect the rangesof reservoir levels that are to be expected during therecreational season.

b. Downstream river levels. In addition to theseasonal regulation of reservoir levels for recreation,regulation of project outflows may encompass requirementsfor specific regulation criteria to enhance the use of therivers downstream from the projects, as well as to ensurethe safety of the general public. The Corps has theresponsibility to regulate projects in a manner to maintainor enhance the recreational use of the rivers below projectsto the extent possible (i.e., without significantly affectingthe project function for authorized purposes). During thepeak recreation season, streamflows are regulated to ensurethe safety of the public who may be engaged in waterrelated activities, including boating, swimming, fishing,rafting, and river drifting. Also, the aesthetics of the riversmay be enhanced by augmenting streamflows during thelow-water period. Water requirements for maintaining orenhancing the recreational use of rivers are usually muchsmaller than other major project functional uses.Nevertheless, it is desirable to include specific goals toenhance recreation in downstream rivers in the water

or augmented streamflows at times of special need for

EM 1110-2-142031 Oct 97

2-6

recreational use of rivers, and in some cases it may be b. Water control systems management. EM 1110-2-poss ible to enhance the conditions of stream flow for 3600 provides guidance on water control plans and projectrelatively short periods of time for this purpose. management. A general prime requirement in project

regulation is the safety of users of the facilities and the

2-11. Water Management Goals and Objectiv es

a. Water management. ER 1110-2-240 paragraph 6,defines the goals and objectives for water regulation by theCorps. Basically, the objective is to conform with specificprovisions of the project authorizing legislation and watermanagement criteria defined in Corps of Engineers reportsprepared in the planning and design of the project orsystem. Beyond this, the goals for water management willinclude the provisions, as set forth in any applicableauthorities, established project construction, and allapplicable Congressional Acts and Executive Ordersrelating to operations of Federal facilities.

general public, both at the project and at downstreamlocations. The development of water control plans and thescheduling of reservoir releases must be coordinated withappropriate agencies, or entities, as necessary to meetcommitments made during the planning and design of theproject. Additionally, water control plans must be reviewedand adjusted, when possible, to meet changing localconditions.

c. Regional management. Regional water manage-ment should consider the interaction of surface-ground-water resources. HEC Research Document 32 providesexamples for several regions in the United States (HEC1991c).

EM 1110-2-142031 Oct 97

3-1

Chapter 3Multiple-Purpose R eservoirs

3-1. Hydrologic Studies for MultipurposeProjects

a. Conception. Multipurpose reservoirs were origi-nal l y conceived as projects that served more than onepurpose independently and would effect savings throughthe construction of a single large project instead of two ormore smaller projects. As the concept developed, the jointuse of water and reservoir space were added asmul t ipurpose concepts. Even such competitive uses asflood control and water supply could use the same reservoirspace at different times during the year.

b. Feasibility. The feasibility of multiple-purposedevelopment is almost wholly dependent upon the demon-st rated ability of a proposed project to serve several pur-poses simultaneously without creating conditions thatwould be undesirable or intolerable for the other purposes.In order to demonstrate that multipurpose operation isfeas ible, detailed analyses of the effects of various combi-nations of streamflows, storage levels, and water require-ments are required. Detailed analyses of these factors maybe overlooked during the planning phase because theanalyses are complex and simplifying methods or assump-tions may not consider some details that may be important.However, ignoring the details of multipurpose operation inthe planning phase is risky because the operation criteriaare critical in determining the feasibility of serving severalpurposes simultaneously.

c. Def in ing the multipurpose project. One of thefactors that make detailed sequential analyses of multipur-pose operation difficult during planning studies is thatsufficient data on various water demands are either notavailable or not of comparable quality for all purposes. Toadequately define the multipurpose operation, the analysesmust include information on the magnitude and seasonalvariations of each demand, long-term changes in demands,relative priority of each use, and shortage tolerances.Information on magnitude and seasonal variation indemands and on long-term variations in demands is usuallymore readily available than information on relativepriorities among uses and on shortage tolerances. Ifinformation on priorities and shortages is not available fromthe various users, one can make several assumptionsconcerning the priorities and perform sequential routingstudies for each set of assumptions. The results of thesestudies can determine the consequences of various prioritiesto potential water users. It may be possible for the potential

users to adopt a priority arrangement based on the value ofthe water for the various demands.

d. Success of multipurpose projects. The success ofmultipurpose operation also depends on the formulation ofoperational rules that ensure that water in the properquantities and qualities is available for each of the purposesat the proper time and place. Techniques for formulatingoperational rules are not fixed, but the logical approachinvolves determining the seasonal variation of theflood-control space requirement, and the seasonal variationof conservation requirements, formulation of generaloperational rules that satisfy these requirements, anddetai led testing of the operational rules to ascertain theadequacy of the plan for each specific purpose.

e. Multipurpose project rules. The judgment of anexperienced hydrologic engineer is invaluable in the initialformulation and subsequent development and testing ofoperational rules. Although the necessary rules cannot becompletely developed until most of the physical dimensionsof the project are known, any tendency to discount theimportance of operational rules as a planning variableshould be resisted because of the important role they oftenassume in the feasibility of multipurpose projects. As aminimum, the operational rules used in a planning studyshould be sufficiently refined to assist the engineer inevaluating the suitability of alternative projects to satisfywater demands for specified purposes.

3-2. Relativ e Priorities of Project Functions

a. Developing project rules. As indicated above, theuse of operational rules based on the relative prioritiesamong the project purposes appears to offer the bestapproach to multipurpose operational problems. Thedegree of success that can be realized depends on a realisticpriority system that accurately reflects the relative value ofwater from the project for a given purpose at a given time.Unless a realistic priority system is used to develop theoperational rules, it will not be possible to follow the rulesduring the project life because the true priorities maycontrol the operational decisions and prevent the projectfrom supplying the services it was designed to provide.

b. Typical system priorities. Priorities among thevarious water resource purposes vary with locale, waterrights, the need for various types of water use, the legal andpolitical considerations, and with social, cultural, andenvironmental conditions. Although these variations makei t impossible to specify a general priority system, it isuseful to identify a set of priorities that would be typicalunder average conditions. In such a situation, operation forthe safety of the structure has the highest priority unless the

EM 1110-2-142031 Oct 97

3-2

consequences of failure of the structure are minor (which is s torage space (although the storage reservation can varyseldom the case). Of the functional purposes, flood control seasonally) because of the basic conflict between reservingmust have a high priority, particularly where downstream empty storage space for regulating potential floods andlevees, bridges, or other vital structures are threatened. It is filling that space to meet future water supply requirements.not unusual for conservation operations to cease entirely However, applying specific storage allocations orduring periods of flood activity if a significant reduction in reservations for competing conservation purposes should beflooding can be realized thereby. Among the conservation kept to a minimum because it reduces operationalpurposes, municipal and industrial water supply and flexibility.hydroelectric power generation are often given a highpriority, particularly where alternative supplies are not c. Operational conflicts. Allocation of specific stor-readily available. After those purposes, other project age space to several purposes within the conservation poolpurposes usually have a somewhat lower priority because can result in operational conflicts that might make it impos-temporary shortages are usually not disastrous. It should be s ible or very costly to provide water for the various pur-emphasized again that there can be marked exceptions to poses in the quantities and at the time they are needed. Thethese relative priorities. There are regional differences in concept of commingled or joint-use conservation storagerelative needs and, legal and institutional factors may for all conservation purposes with operational criteria togreatly affect priorities. maximize the complementary effects and minimize the

c. Complex system priorities. In complex reservoir designed, will provide better service for all purposes.systems, with competing demands and several alternative W here the concept of joint-use storage is used, theprojects to meet the demand, the relative priority among operational criteria should be studied in the planningprojects and purposes may not be obvious. The operation process in such a way that the relative priorities of therules, which can be evaluated with detailed simulation, may various purposes are taken into account. This allowsnot be known or may be subject to criticism. In these careful evaluation of a number of priority systems andsituations, it may be useful to apply a system analysis based operational plans. The operational decisions that resulton consistent values for the various project purposes. The from such disputes are frequently not studied in enoughresults of the analysis could suggest an operational strategy detail (from the engineering point of view), and as a result,which can be tested with more detailed analysis. Chapter 4 the abi lity of the project to serve some purposes may bepresents information and approaches for system analysis. seriously affected.

3 -3. Managing Competitiv e and Complementary 3-4. Operating ConceptsFunctions

a. Identifying interactions between purposes. Beforeoperation rules can be formulated, the adverse (competi-t i ve) and the beneficial (complementary) interactionsbetween purposes must be identified. The time of occur-rence of the interactions is often as important as the degreeof interaction, particularly if one or more of the water useshas s ignificant variations in water demand. In supplyingwater from a single reservoir for several purposes withseasonally varying demands, it is possible for normallycomplementary purposes to become competitive at timesdue to differences in their seasonal requirements.

b. Allocating storage space. When several purposesare to be served from a single reservoir, it is possible toallocate space within certain regions of the reservoir storagefor each of the purposes. This practice evolved fromprojects that served only flood control and one conservationpurpose because it was necessary to reserve a portion of thereservoir storage for storing floodwater. It is still necessaryto have a specific allocation of flood-control

competitive effects is far easier to manage and, if carefully

a. Operating goals. Reservoir operating goals varywith the storage in the reservoir. The highest zone in thereservoir is that space reserved at any particular time for thecontrol of floods. This zone includes the operational flood-cont rol space and the surcharge space required for thepassage of spillway flows. Whenever water is in this zoneit must be released in accordance with flood-controlrequi rements. The remaining space can be designated asconservation space. The top zone of conservation spacemay include storage that is not required to satisfy the firmconservation demands, including recreational use of thereservoir. Water in this space can be released as surplus toserve needs or uses that exceed basic requirements. Themiddle zone of conservation space is that needed to storewater to supply firm water needs. The bottom zone ofconservation space can be termed buffer space, and whenoperation is in this zone the firm services are curtailed inorder to prevent a more severe shortage later. The bottomzone of space in the reservoir is designated as the minimumpool reserved for recreation, fish, minimum power head,sediment reserve, and other storage functions.

EM 1110-2-142031 Oct 97

3-3

b. Storage zone boundaries. The boundaries between purposes, and to change from one level of service for astorage zones may be fixed at a constant level or they may given purpose to a lower level of service for that samevary seasonally. In general, the seasonally varying purpose when storage levels are too low to ensure theboundaries offer the potential for a more flexible operating continuation of firm supplies for all purposes. As with theplan that can result in higher yields for all purposes. other techniques for implementing a multipurpose opera-However, the proper location of the seasonal boundaries t ion, the amount of buffer storage and the location of therequi res more study than the location of a constant boundaries cannot be determined accurately except byboundary. This is discussed in more detail in Chapter 11. successive approximations and testing by sequential routingFurthermore, an additional element of chance is introduced studies.when the boundaries are allowed to vary, because the jointuse of storage might endanger firm supplies for one or morespecific purposes. The location of the seasonally varyingboundaries is determined by a process of formulating a setof boundaries and attendant operational rules, testing thescheme by a detailed sequential routing study, evaluatingthe outcome of the study, changing the rules or boundariesif necessary, and repeating the procedure until a satisfactoryoperation results.

c. Demand schedules. Expressing demand schedulesas a function of the relative availability of water is anothermeans of incorporating flexibility and relative priority inoperational rules. For example, the balance between hydroand thermal power generation might well be a continuousfunction of available storage. As another example, it mightbe possible to have two or more levels of navigation serviceor lengths of navigation season with the actual level ofservice or length of season being dependent upon theavailability of water in the reservoir. By regulating thelevel of supply to the available water in the reservoir, userscan plan emergency measures that will enable them towithstand partial reductions in service and thereby avoidcomplete cessation of service, which might be disastrous.Terms such as desired flow and minimum required flow fornavigation can be used to describe two levels of service.

d. Levels of service. There can be as many levels ofservice as a user desires, but each level requires criteria fordetermining when the level is to be initiated and when it isto be terminated. The testing and development of thecri teria for operating a multipurpose project with severalpurposes and several levels of service are accomplished bydetailed sequential routing studies. Because the devel-opment and testing of these criteria are relatively difficult,the number of levels of service should be limited to theminimum number needed to achieve a satisfactory plan ofoperation.

e. Buf fer storage. Buffer storage or buffer zones areregions within the conservation storage where operationalrules effect a temporary reduction in firm services. Thetwo primary reasons for temporarily reducing services areto ensure service for a high-priority purpose whileel iminating or curtailing services for lower-priority

3-5. Construction and Physical Operation

a. General. In addition to hydrologic determinationsdiscussed above, a number of important hydrologicdeterminations are required during project construction andduring project operation for ensuring the integrity of theproject and its operation.

b. Cof ferdams. From a hydrologic standpoint,during construction the provisions for streamflow diversionare a primary concern. If a cofferdam used for dewateringthe work area is overtopped, serious delays and additionalconstruction costs can result. In the case of highcofferdams where substantial poundage occurs, it ispossible that failure could cause major damage indownstream areas. Cofferdams should be designed on thesame principles as are permanent dams, generally on thebasis of balancing incremental costs against incrementalbenefits of all types. This will require flood frequency andhypothetical flood studies, as described in Chapters 6 and 7of th is manual. Where major damage might result fromcofferdam failure, a standard project flood (SPF) or even aprobable maximum flood (PMF) may be used as a primarybasis for design.

c. Over topping. Where a major dam embankmentmay be subject to overtopping during construction, thediversion conduit capacity must be sufficient to regulatefloods that might occur with substantial probability duringthe critical construction period. It is not necessary that theregulated releases be nondamaging downstream, but it isvital that the structure remain intact. An explicit evaluationof risk of embankment failure and downstream impactsduring construction should be presented in the designdocument.

d. Conduits, spillways, and gates. Conduits, spill-ways, and all regulating gates must be functionally ade-quate to accomplish project objectives. Their sizes,dependability, and speed of operation should be testedus ing recorded and hypothetical hydrographs and antici-pated hydraulic heads to ensure that they will performproperly. The nature of stilling facilities might be dictatedby hydrologic considerations if frequency and duration of

EM 1110-2-142031 Oct 97

3-4

high outflows substantially influence their design. effectiveness by enabling the operating agency, throughT he necessity for multilevel intakes to control the quality rel iable forecasts of hydrologic conditions, to increaseof reservoir releases can be assessed by detailed operation efficiency. Hydrologic aspects of monitoringreservoir stratification studies under all combinations of faci l i ties and forecasts will be presented in a new EM onhydrologic and reservoir conditions. Techniques for hydrologic forecasting.conducting reservoir stratification studies are discussed inEM 1110-2-1201. i. St ream gauges. Because gauged data are most

e. Design. The design of power facilities can be in locating the gauge. Stream gauges should not be locatedgreatly influenced by hydrologic considerations, as dis- on bridges or other structures that are subject to beingcussed in Chapter 11 of this manual and EM 1110-2-1701. washed out. To the extent possible, the gauges should beGeneral considerations in the hydrologic design of capable of working up through extreme flood events, andspillways are discussed in Chapter 10 and more detailed stage-discharge relationships should be developed up toinformation is presented in Chapter 14 herein. that level. The gauge should have reasonable access for

f. Extreme floods. Regardless of the reservoir pur- local flooding, and backwater effects from downstreamposes, it is imperative that spillway facilities provided will tributaries should all be considered when finding a suitableensure the integrity of the project in the event of extreme location. More detailed information on stream gauges canfloods. Whenever the operation rules of a reservoir are be found in many USGS publications, such as Carter andsubstantially changed, spillway facilities should be Davidian (1968), Buchanan and Somers (1968 and 1969),reviewed to ensure that the change in project operation does or Smoot and Novak (1969).not adversely alter the capability to pass extreme floodswi thout endangering the structure. The capability of aspillway to pass extreme floods can be adversely affectedby changes in operation rules that actually affect the floodoperation itself or by changes that result in higher poolstages during periods of high flood potential.

g. Special operating rules. A number of situationsmight require special operating rules. For example, oper-at ing rules are needed for the period during which a reser-voi r is initially filling, for emergency dewatering of areservoir, for interim operation of one or more componentsin a system during the period while other components areunder construction, and for unanticipated conditions thatseem to require deviation from established operating rules.T he need for operation rules during the filling period isespecially important because many decisions must be basedon the filling plan. Among the important factors that aredependent upon the filling schedule are the on-line date forpower generating units, the in-service dates for variouspurposes such as water supply and navigation, and theeffective date for legal obligations such as recreationconcessions.

h. Specification of monitoring facilities. One of themore important considerations in the hydrologic analysis ofany reservoir is the specification of monitoring facilities,including streamflow, rainfall, reservoir stage, and otherhydrologic measurements. These facilities serve two basicpurposes: to record all operations and to provideinformation for operation decisions. The former purposesatisfies legal requirements and provides data for futurestudies. The latter purpose may greatly increase the project

important during flood events, special care should be taken

checking and repair during the flood. Reservoir spilling,

3-6. General Study Procedure

As indicated earlier, there is no fixed procedure for devel-oping reservoir operational plans for multipurpose projects;however, the general approach that should be common toall cases would include the following steps:

a. S urvey the potential water uses to be served bythe project in order to determine the magnitude of eachdemand and the seasonal and long-term variations in thedemand schedule.

b. Develop a relative priority for each purpose anddetermine the levels of service and required priority thatwill be necessary to serve each purpose. If necessary, makesequential studies illustrating the consequences of variousalternative priority systems.

c. Establish the seasonal variation of flood-controlspace required, using procedures discussed in Chapter 10.

d. Establish the total power, water supply, andlow-flow regulation requirements for competitive purposesduring each season of the year.

e. Establish preliminary feasibility of the projectbased on physical constraints.

f. Establish the seasonal variation of the storagerequirement to satisfy these needs, using proceduresdescribed in Chapter 11.

EM 1110-2-142031 Oct 97

3-5

g. Determine the amount of storage needed as a test ing, evaluating, and changing until satisfactoryminimum pool for power head, recreation, sedimentation operation is obtained.reserve, and other purposes.

h. Using the above information, estimate the size of t ial routing studies with stochastic hydrologic data toreservoir and seasonal distribution of space for the various evaluate the possibility of historical bias in the proposedpurposes that would satisfy the needs. Determine the rules.reservoir characteristics, including flowage, spillway,power plant, and outlet requirements. k. Determine the needs for operating and monitoring

i. T est and evaluate the operation of the project the project.through the use of recorded hydrologic data in a sequentialrouting study to determine the adequacy of the storage l. As detailed construction plans progress, evaluateestimates and proposed rules with respect to the operational cofferdam needs and protective measures needed for theobjectives for each purpose. If the record is short, supple- integrity of project construction, particularly diversionment it with synthetic floods to evaluate flood storage capacity as a function of dam construction stage and floodreserves. If necessary, make necessary changes and repeat threat for each season.

j. Test proposed rules of operation by using sequen-

equipment required to ensure proper functional operation of

EM 1110-2-142031 Oct 97

4-1

Chapter 4Reservoir Systems

4-1. Introduction

Water resource systems should be designed and operatedfor the most effective and efficient accomplishment ofoveral l objectives. The system usually consists of reser-voirs, power plants, diversions, and canals that are eachconstructed for specific objectives and operated based onexisting agreements and customs. Nevertheless, there isconsiderable latitude in developing an operational plan forany water resource system, but the problem is greatlycomplicated by the legal and social restrictions thatordinarily exist.

a. Mathematical modeling. Water resource systemoperation is usually modeled mathematically, rather thanwith physical models. The mathematical representation ofa water resource system can be extremely complex. Oper-ations research techniques such as linear programming anddynamic programming can be applied to a water resourcesystem; however, they usually are not capable ofincorporating all the details that affect system outputs. It isusually necessary to simulate the detailed sequentialoperation of a system, representing the manner in whicheach element in the system will function under realisticconditions of inputs and requirements on the system. Thes imulation can be based on the results from the optimiza-tion of system outputs or repeated simulations. Succes-s ively refining the physical characteristics and operationalrules can be applied to find the optimum output.

b. Inputs and requirements. A factor that greatlycomplicates the simulation and evaluation of reservoirsystem outputs is the stochastic nature of the inputs and ofthe requirements on the system. In the past, it has beencustomary to evaluate system accomplishments on theassumption that a repetition of historical inputs andrequirements (adjusted to future conditions) would ade-quately represent system values. However, this assumptionhas been demonstrated to be somewhat deficient. It isdes irable to test any proposed system operation under agreat many sequences of inputs and requirements. Thisrequires a mathematical model that will define the fre-quency and correlation characteristics of inputs andrequirements and that is capable of generating a number oflong sequences of these quantities. Concepts foraccomplishing this are discussed in paragraph 5-5.

4-2. System Description

a. Simulating system operation. Water resourcesystems consist of reservoirs, power plants, diversionstructures, channels, and conveyance facilities. In order tosimulate system operation, the system must be completelydescribed in terms of the location and functionalcharacteristics of each facility. The system should includeall components that affect the project operation and providethe required outputs for analysis.

(1) Reservoirs. For reservoirs, the relation of surfacearea and release capacity to storage content must bedescribed. Characteristics of the control gates on theoutlets and spillway must be known in order to determineconstraints on operation. The top-of-dam elevation must bespeci fied and the ability of the structure to withstandovertopping must be assessed.

(2) Downstream channels. The downstream channelsmust be defined. Maximum and minimum flow targets arerequired. For short-interval simulation, the translation offlow through the channel system is modeled by routingcriteria. The travel time for flood flow is important indetermining reservoir releases and potential limits for floodcontrol operation to distant downstream locations.

(3) Power plants. For power plants at storage reser-voi rs , the relation of turbine and generation capacity tohead must be determined. To compute the head on theplant, the relation of tailwater elevation to outflow must beknown. Also, the relation of overall power plant efficiencyto head is required. Other characteristics such as turbineleakage and operating efficiency under partial load are alsoimportant.

(4) Diversion structures. For diversion structures,maximum diversion and delivery capacity must be estab-lished. The demand schedule is required, and the consump-tive use and potential return flow to the system may beimportant for the simulation.

b. Preparing data. While reservoir system data mustbe defined in sufficient detail to simulate the essence of thephysical system, preparing the required hydrologic datamay require far more time and effort. The essential flowdata are required for the period of record, for major floodevents, and in a consistent physical state of the system.Flow records are usually incomplete, new reservoirs in thesystem change the flow distribution, and water usage in thewatershed alters the basin yield over

EM 1110-2-142031 Oct 97

4-2

time. Developing a consistent hydrologic data series, mak- an upstream-to-downstream direction. At each pertinenting maximum use of the available information, is discussed locat ion, requirements for each service are noted, and thein Chapter 5. reservoirs at and above that location are operated in such a

4-3. Operating Objectiv es and Criteria

a. User services. Usually, there is a fixed objectivefor each function in a water resource system. Projects areconst ructed and operated to provide services that arecounted on by the users. In the case of power generationand water supply, the services are usually contracted, and iti s essential to provide contracted amounts insofar aspossible. Services above the contracted amounts areordinarily of significantly less value. Some services, suchas flood control and recreation, are not ordinarily coveredby contracts. For these, service areas are developed toprovide the degree of service for which the project was4-5. Flood-Control Simulationconstructed.

b. Rules for services. Shortages in many of theservices can be very costly, whereas surpluses are usuallyof m inor value. Accordingly, the objectives of waterresource system operational are usually fixed for any partic-ular plan of development. These are expressed in terms ofoperational rules that specify quantities of water to bereleased and diverted, quantities of power to be generated,reservoir storage to be maintained, and flood releases to bemade. These quantities will normally vary seasonally andwi th the amount of storage water in the system. Rulecurves for the operation of the system for each function aredeveloped by successive approximations on the basis ofperformance during a repetition of historical streamflows,adjusted to future conditions, or on the basis of syntheticstream flows that would represent future runoff potential.

4-4. System Simulation

T he evaluation of system operation under specified opera-t ion rules and a set of input quantities is complex andrequires detailed simulation of the operation for longperiods of time. This is accomplished by assuming thats teady-state conditions prevail for successive intervals oftime. The time interval must be short enough to capture thedetails that affect system outputs. For example, averagemonthly flows may be used for most conservationpurposes; however, for small reservoirs, the flow variationwithin a month may be important. For hydropower reser-voirs, the average monthly pool level or tailwater elevationmay not give an accurate estimate of energy production.

To simulate the operation during each interval, the simula-tion solves the continuity equation with the reservoirrelease as the decision variable. The system is analyzed in

way as to serve those requirements, subject to systemconstraints such as outlet capacity, and channel capacity,and reservoir storage capacity. As the computationprocedure progresses to downstream locations, the tentativerelease decisions made for upstream locations becomeincreasingly constraining. It often becomes necessary toassign priorities among services that conflict. Where powergeneration causes flows downstream to exceed channelcapacity, for example, a determination must be made as towhether to curtail power generation. If there is inadequatewater at a diversion to serve both the canal and riverrequirements, a decision must be made.

Flood discharge can change rapidly with time. Therefore,s teady-state conditions cannot be assumed to prevail forlong periods of time (such as one month). Also, physicalconstraints such as outlet capacity and the ability to changegate settings are more important. The time translation forflow and channel storage effects cannot ordinarily beignored. Consequently, the problem of simulating theflood-control operation of a system can be more complexthan for conservation.

a. Computational interval. The computation intervalnecessary for satisfactory simulation of flood operations isusually on the order of a few hours to one day at the most.Sometimes intervals as short as 15 or 30 min are necessary.It is usually not feasible to simulate for long periods oftime, such as the entire period of record, using such a shortcomputation interval. However, period-of-record may beunnecessary because most of the flows are of noconsequence from a flood-control standpoint. Accordingly,simulation of flood-control operation is usually made onlyfor important flood periods.

b. Starting conditions. The starting conditions forsimulating the flood-control operation for an historic floodperiod would depend on the operation of the system forconservation purposes prior to that time. Accordingly, theconservation operation could be simulated first to establishthe state of the system at the beginning of the month duringwhich the flood occurred as the initial conditions for theflood simulation. However, the starting storage for floodoperation should be based on a realistic assessment oflikely future conditions. If it is likely that the conservationpool is full when a flood occurs, then that would be a betterstarting condition to test the flood-pool capacity. It may bepossible that the starting pool would be higher if there wereseveral storms in sequence, or if the flood operation does

EM 1110-2-142031 Oct 97

4-3

not start the instant excessive inflows raise the pool levelinto flood-control space.

c. His toric sequences. While simulating historicsequences are important, future floods will be different andoccur in different sequences. Therefore, the analysis offlood operations should utilize both historic and syntheticfloods. The possibility of multiple storms, changes in theupstream catchment, and realistic flood operation should beincluded in the analysis. Chapter 7 presents flood-runoffanalysis and Chapter 10 presents flood-control storagerequirements.

d. Upstream-to-downstream solution. If the opera-t ion of each reservoir in a system can be based onconditions at or above that reservoir, an upstream-to-downstream solution approach can establish reservoirreleases, and these releases can be routed through channelreaches as necessary in order to obtain a realistic simula-tion. Under such conditions, a simple simulation model iscapable of simulating the system operation with a highdegree of accuracy. However, as the number of reservoirsand downstream damage centers increase, the solutionbecomes far more complex. A priority criteria must beestablished among the reservoirs to establish which shouldrelease water, when there is a choice among them.

e. Combination releases. The HEC-5 Simulation ofFlood Control and Conservation Systems (HEC 1982c)computer program can solve for the combination of releasesat upstream reservoirs that will satisfy channel capacityconstraints at a downstream control point, taking intoaccount the time translation and channel storage effects,and that will provide continuity in successive time intervals.T he time translation effects can be modeled with a choiceof hydrologic routing methods. Reservoir releases aredetermined for all designated downstream locations, subjectto operation constraints. The simulation is usuallyperformed with a limited foresight of inflows and a con-t ingency factor to reflect uncertainty in future flow values.T he concept of pool levels is used to establish prioritiesamong projects in multiple-reservoir systems. Standardoutput includes an indicator for the basis of reservoirrelease determination, along with standard simulation out-put of reservoir storage, releases, and downstream flows.

f. Period-of-record flows. Alternatively, a singletime interval, such as daily, can be used to simulate period-of-record flows for all project purposes. This approach isroutinely used in the Southwestern Division with the com-puter program “ Super” (USACE 1972), and in the NorthPacific Division with the SSARR program (USACE 1991).The SSARR program is capable of simulation on variabletime intervals.

4-6. Conserv ation Simulation

W hi le the flood-control operation of a reservoir system issensitive to short time variations in system input, theoperation of a system for most conservation purposes isusually sensitive only to long-period streamflow variations.His torically, simulation of the conservation operation of awater resource system has been based on a relatively longcomputation interval such as a month. With the ease ofcomputer simulation and available data, shortercomputational intervals (e.g., daily) can provide a moreaccurate accounting of flow and storage. Some aspects ofthe conservation operation, such as diurnal variations inpower generation in a peaking project, might require evenshorter computational intervals for selected typical orcritical periods to define important short-term variations.

a. Hydropower simulation. Hydropower simulationrequires a realistic estimate of power head, which dependson reservoir pool level, tailwater elevation, and hydraulicenergy losses. Depending on the size and type of reservoir,there can be considerable variation in these variables.General ly, the shorter time intervals will provide a moreaccurate estimate of power capacity and energyproductions.

b. Evaporation and channel losses. In simulatingthe operation of a reservoir system for conservation, thet ime of travel of water between points in the system isusually ignored, because it is small in relation to the typicalcomputation interval (e.g., monthly or weekly). On theother hand, evaporation and channel losses might be quiteimportant; and it is sometimes necessary to account forsuch losses in natural river channels and diversion canals.

c. Rule curves. Rule curves for the operation of areservoir system for conservation usually consist of stan-dard power generation and water supply requirements thatwill be served under normal conditions, a set of storagelevels that will provide a target for balancing storage amongthe various system reservoirs, and maximum and minimumpermissible pool levels for each season based on floodcont rol, recreation, and other project requirements. Oftensome criteria for decreasing services when the systemreservoir storage is critically low will be desirable.

4-7. System Power Simulation

W here a number of power plants in the water resourcesystem serve the same system load, there is usually con-s iderable flexibility in the selection of plants for powergeneration at any particular time. In order to simulate theoperat ion of the system for power generation, it is neces-sary to specify the overall system requirement and the

EM 1110-2-142031 Oct 97

4-4

minimum amount of energy that must be generated at each water resource system is probably more difficult than theplant during each month or other interval of time. Because derivation of optimum configuration and unit sizes becausethe entire system power requirement might possibly be any small change in operation rules can affect manysupplied by incidental generation due to releases made for functions in the system for long periods of time and in veryother purposes, it is first necessary to search the entire subtle ways.system to determine generation that would occur with onlyminimum power requirements at each plant and with all a. Simulation. Operation criteria generally consistrequirements throughout the system for other purposes. If of release schedules at reservoirs, diversion schedules atinsufficient power is generated to meet the entire system control points, and minimum flows in the river at controlload in this manner, a search will be made for those power points, in conjunction with reservoir balancing levels thatreservoirs where storage is at a higher level, in relation to define the target storage contribution among the variousthe rule curves, than at other power reservoirs. The reservoirs in the system. All of these can vary seasonally,additional power load requirement will then be assigned to and target flows can vary stochastically. Once the unitthose reservoirs in such a manner as to maintain the s izes and target flows are established for a particular planreservoir storage as nearly as possible in conformance with of development, a system of balancing levels must bethe rule curves that balance storage among the reservoirs in developed. The system response to a change in thesethe most desirable way. This must be done without balancing levels is a complicated function of many system,assigning more power to any plant than it can generate at input, and requirement characteristics. For this reason, theoverload capacity and at the system load factor for that development of a set of balancing levels is an iterationinterval. EM 1110-2-1701 paragraph 5-14, describes process, and a complete system simulation must be done forhydropower system analysis. each iteration.

4-8. Determination of Firm Yield

If the yield is defined as the supply that can be maintainedthroughout the simulation period without shortages, thenthe process of computing the maximum yield can beexpedited. This is done by maintaining a record of theminimum reserve storage (if no shortage has yet occurred)or of the amount of shortage (if one does occur) in relationto the total requirement since the last time that all reservoirswere full. The surplus or shortage that existed at the end ofany computation interval would be expressed as a ratio ofthe supply since the reservoirs were last full, and theminimum surplus ratio (if no shortage occurs) or maximumshortage ratio (if a shortage does occur) that occurs duringthe ent ire simulation period would be used to adjust thetarget yield for the next iteration. This basic procedure forcomputing firm yield is included in the HEC-5 computerprogram. Additionally, the program has a routine to makean initial estimate of the critical period and expected yield.After the yield is determined using the critical period, theprogram will evaluate the yield by performing a simulationwi th the entire input flow record. Chapter 12 describesstorage-yield procedures.

4-9. Deriv ation of Operating Criteria

A plan of development for a water resource system consistsnot only of the physical structures and their functionalcharacteristics but also of the criteria by which the systemwi ll be operated. In order to compare alternative plans ofdevelopment, it is necessary that each plan be operatedoptimally. The derivation of optimal operation criteria for a

(1) When first establishing balancing levels in thereservoir system, it usually is best to simulate systemoperation only for the most critical periods of historicalstreamflows. The final solution should be checked bysimulation for long periods of time. The balancing levelsdefin ing the flood-control space are first tentatively estab-l i shed on the basis of minimum requirements for floodcont rol storage that will provide the desired degree ofprotection. Preliminary estimates of other levels can beestablished on the basis of reserving the most storage in thesmaller reservoirs, in those reservoirs with the least amountof runoff, and in those reservoirs that supply operation ser-vices not producible by other reservoirs.

(2) After a preliminary set of balancing levels isestablished, they should be defined approximately in termsof a minimum number of variables. The general shape andspacing of levels at a typical reservoir might be defined bythe use of four or five variables, along with rules forcomputing the levels from those variables. Variations inlevels among reservoirs should be defined by one or twovariables, if possible, in order to reduce the amount of workrequired for optimization to an acceptable quantity.

(3) Optimization of a set of balancing levels for oper-at ional rule curves can be accomplished by successiveapproximations using a complete system simulation com-putation for critical drought periods. However, the proce-dures are limited to the input specifications of demands andstorage allocation. While one can compare simulationresults and conclude one is better than another based on

EM 1110-2-142031 Oct 97

4-5

performance criteria, there is no way of knowing that an (3) The development of the penalty functions requiresoptimum solution has been achieved. an economic evaluation of the values to be placed on flow

b. Optimization. While water resource agencies have are disagreements on the values, due to the difficulty ofgenerally focused on simulation models for system analysis, defining values for some purposes. However, the processthe academic community and research literature have does provide a method for defining and reviewing theemphasized optimization and stochastic analysis purposes and their relative values.techniques. Research performed at HEC (HEC 1991b) hasfound a proliferation of papers on optimization of reservoir (4) The primary disadvantage of the HEC-PRM is thatsystem operations written during the past 25 years, the monthly flow data and lack of channel routing limit itsprimarily by university researchers. There still remains a application for short interval simulation, such as floodcons iderable gap between the innovative applications cont rol and peaking hydropower. Additionally, thereported in the literature and the practices followed by the optimized solution is provided in terms of period-of-recordagencies responsible for water resource development. One flows and storage; however, the basis for the systembasic problem is that many of the reported applications are operation are not explicitly defined. The post-processing ofuniquely formulated to solve a specific problem for a given the results requires interpretation of the results in order tosystem. There is a general view that the models develop an operation plan that could be used in basicperformance, or the methods assumptions, would not simulation and applied operation. More experience withsufficiently evaluate a different problem and system. this analysis of results is still required to define these

c. Prescriptive reservoir model. HEC has developeda system analysis tool based on a network flow model(HEC 1991a). The Prescriptive Reservoir Model (HEC-PRM) will identify the water allocation that minimizes poorperformance for all defined system purposes. Performanceis measured with analyst-provided functions of flow ors torage or both. The physical system is represented as anetwork, and the allocation problem is formulated as aminimum-cost network flow problem. The objectivefunctions for this network problem are convex, piecewise-linear approximations of the summed penalty functions foreach project purpose (HEC 1991d).

(1) S ystems have been analyzed in studies on theMissouri River (HEC 1991d) and the Columbia River(HEC 1991f). A preliminary analysis of the Phase IMissouri River study has developed initial methodologiesfor developing operation plans based on PRM results(HEC 1992b). Continued application experience isrequired to define generalized procedures for theseanalyses.

(2) The primary advantages for the HEC-PRMapproach are the open state of the system and the requiredpenal ty functions for each system purpose. There are norule curves or details of storage allocation, only basicphysical constraints are defined. The reservoir systeminformation defines maximum and minimum storage in thereservoirs and the linking of the system through thenetwork of channels and diversions. The other primaryreservoir data is traditional period-of-record monthly flowsfor the system.

and storage in the system. The process is difficult and there

procedures.

4-10. System Formulation Strategies

a. Determining the best system. A system is best forthe national income criteria if it results in a value for systemnet benefits that exceeds that of any other feasible system.Except where noted, the following discussion wasdeveloped in a paper presented at the International Com-missions on Large Dams Congress (Eichert andDavis 1976). For a few components, analysis of the num-ber of alternative systems that are feasible is generallymanageable, and exhaustive evaluation provides thestrategy for determining the best system. When the numberof components is more than just a few, then the exhaustiveevaluation of all feasible alternative systems cannotpractically be accomplished. In this instance, a strategy isneeded that reduces the number of system alternatives to beevaluated to a manageable number while providing a goodchance of identifying the best system. System analysis doesnot permit (maximum net benefit system) for reasonablycomplex systems even with all hydrologic-economic dataknown. An acceptable strategy need not make the absoluteguarantee of economic optimum because seldom will theoptimum economic system be selected as best.

b. Incremental test. The incremental test of the valueof an individual system component is definitive for theeconomic efficiency criteria and provides the basis forseveral alternative formulation strategies. If existingreservoir components are present in the system, then theydefine the base conditions. If no reservoirs exist, the basecondition would be for natural conditions. The strategies

EM 1110-2-142031 Oct 97

4-6

described below are extensions of currently used techni- (1) The analysis is then repeated for the next stage byques and are based upon the concept of examining in detail computing the first added value of each component to thethe performance of a selected few alternative systems. The system again, the base now including the first componentperformance is assumed to be evaluated generally by added. The strategy is continued to completion by suc-traditional simulation methods, like the use of HEC-5. cessive application of the first added analysis until no more

c. Reasoned thought strategy. This strategy is predi-cated upon the idea that it is possible to reason out using (2) T he strategy does have a great deal of practicaljudgment and other criteria, reasonable alternative systems. appeal and probably would accomplish the important taskThe strategy consists of devising through rational thought, of identifying the components that are clearly good addi-sampling, public opinion, literature search, and tions to the system and that should be implemented at anbrainstorming, a manageable number of system alternatives early stage. The strategy, however, ignores any systemthat will be evaluated. No more than 15 to 20 alternative value that could be generated by the addition of more thansystems could be evaluated by detailed simulation in a one component to the system at a time, and this could omitpractical sense. potentially useful additions to the system. For example, the

(1) The total performance of each system in terms of tributaries above a damage center are justified, but eithereconomic (net benefit) and performance criteria is evalu- one analyzed separately is not, i.e., the system effect isated by a system simulation. A system (or systems if more great enough to justify both. The number of systemthan one have very similar performance) is selected that analyses required to formulate a system based on thismaximizes the contribution towards the formulation st rategy could range upwards to 120 for a moderatelyobjectives (those that exhibit the highest value of net complex (15 component) system, which is probably close tobenefits while satisfying the minimum performance crite- being an unmanageably large number of evaluations.ria). To confirm the incremental justification of eachcomponent, the contribution of each system component in e. Last added strategy. This strategy, similar to firstthe last added position is evaluated. The last added value is added strategy, is designed such that successive applicationthe difference between the value (net benefits) of the yields the formulated system. Beginning with all proposedsystem with all components in operation and the value (net components to the system, the value of each component inbenefits) of the system with the last added component the last added position is computed. The project whoseremoved. If each component is incrementally justified, as delet ion causes the value (net benefit) of the system toindicated by the test, the system is economically justified, increase the most is dropped out. The net benefits wouldand formulation is complete. If any components are not increase if the component is not incrementally justified.incrementally justified, they should be dropped and the last T he strategy is continued through successive stagedadded analysis repeated. appl i cations until the deletion of a component causes the

(2) The system selected by this strategy will be afeas ible system that is economically justified. Assuming (1) The last added strategy will also yield a system inthe method of devising the alternative systems is rational, which all components are incrementally justified and inthe chances are good that the major worthwhile projects which the total system will be justified. This strategywill have been identified. On the other hand, the chances would probably identify the obviously desirable projects, asthat this system provides the absolute maximum net bene- would the others. However, its weakness is that it isfits is relatively small. This strategy would require between sl ightly possible, though not too likely, that groups of30 and 60 system evaluations for a moderately complex projects that would not be justified are carried along(15 component) system. because of their complex linkage with the total system. For

d. First added strategy. This strategy is designed two tributaries above a damage center are not justifiedsuch that its successive application will yield the formu- together, but deletion of each from a system that includeslated system. The performance of the systems, including both results in such a great loss in system value that indi-the base components (if any), are evaluated with each vidual analysis indicates neither should be droppedpotential addition to the system in the “ first added” posi- individually.tion. The component that contributes the greatest value (netbenefit) to the system is selected and added to the base (2) The number of systems analyses required for thissystem. st rategy would be similar to the first added strategy

component additions to the system are justified.

situation sometimes exists where reservoirs on, say, two

total system value (net benefits) to decrease.

example, the situation sometimes exists where reservoirs on

EM 1110-2-142031 Oct 97

4-7

requiring perhaps 10-20 percent more evaluations. Twenty- needed to provide a reasonable degree of protection, usingtwo last added analyses were made in the four stages procedures described in Chapter 10. Distribute this storagerequired to select four new projects out of seven in a reasonable way among contemplated reservoirs in orderal ternatives. This strategy is more efficient than the first to obtain a first approximation of a plan for flood control.added if the majority of the potential system additions are Include approximate rule curves for releasing some or all ofgood ones. this storage for other uses during the nonflood season

f. Branch-and-bound enumeration. “ Branch-and-bound enumeration is a general-purpose technique for c. Determine approximately for each tributary,identifying the optimal solution to an optimization problem where appropriate, the total water needed each month forwithout explicitly enumerating all solutions,” (HEC 1985a). al l conservation purposes and attendant losses, and, usingT he technique provides a framework to evaluate procedures described in Chapter 11, estimate the storageindependent alternatives by dividing the entire set into needed on each principle tributary for conservationsubsets for evaluation. The method has been applied in services. Formulate a basic plan of development includingresource planning to problems of sizing, selecting, detailed specification of all reservoir, canal, channel, andsequencing, and scheduling projects. HEC has developed a powerplant features and operation rules; all flowt raining document illustrating the application to flood- requirements; benefit functions for all conservationdamage-mitigation plan selection (HEC 1985). Addition- services; and stage-damage functions for all flood damageal l y, HEC Research Document No. 35 (Bowen 1987) index locations. Although this part of plan formulation isi l l us trates an application for reservoir flood control plan not entirely a hydrologic engineering function, a satisfac-selection using computer program HEC-5 for reservoir tory first approximation requires good knowledge of runoffs imulation. The procedure can use any criteria for evalu- characteristics, hydraulic structure characteristics andat ion and supports detailed simulation in the analysis l imitations, overall hydroelectric power characteristics,process. engineering feasibility, and costs of various types of struc-

4-11. General Study Procedure

There is no single approach to developing an optimum planof improvement for a complex reservoir system. Ordinarilymany services are fixed and act as constraints on systemoperation for other services. In many cases, all but oneservice is fixed, and the system is planned to optimize theoutput for one remaining service, such as power generation.It should also be recognized that most systems have beendeveloped over a long period of time and that manyservices are in fact fixed, as are many system features.Nevertheless, an idealized general study procedure ispresented below:

a. Prepare regional and river-system topographicmaps showing locations of hydrologic stations, existing andcontemplated projects, service and damage areas, andpertinent drainage boundaries. Obtain all precipitation,evaporation, snowpack, hydrograph timing and runoff datapertinent to the project studies. Obtain physical and oper-ational data on existing projects. Construct a normal sea-sonal isohyetal map for the river basin concerned.

b. F or each location where flood protection is to beprovided, estimate approximately the nondamaging flowcapacity that exists or could be ensured with minor channeland levee improvements. Estimate also the amount ofstorage (in addition to existing storage) that would be

where appropriate.

tures, and relocations.

d. Using the general procedures outlined in Part 2,develop flood frequencies, hypothetical flood hydrographs,and stage-discharge relations for unregulated conditionsand for the preliminary plan of development for floodcontrol. It may be desirable to do this for various seasonsof the year in order to evaluate seasonal variation of flood-cont rol space. Evaluate the flood-control adequacy of theplan of development, using procedures described inparagraph 4-5 and Chapter 10, modify the plan, as neces-sary, to improve the overall net benefits for flood controlwhi le preserving basic protection where essential. Eachmodification must be followed by a new evaluation of netbenefits for flood control. Each iteration is costly and time-consuming; consequently, only a few iterations are feasible,and considerable thought must be given to each planmodification.

e. For system analysis to determine the best alloca-t ion of flow and storage for conservation purposes,consider optimization using a tool HEC-PRM (para-graph 4-9c). The program outputs would then be analyzedto infer an operation policy that could be defined forsimulation and more detailed analysis. The alternative is torepeatedly simulate with critical low-flow periods todevelop a policy to meet system goals and then perform aperiod-of-record simulation to evaluate total systemperformance.

EM 1110-2-142031 Oct 97

4-8

f. Consider generating synthetic sequences of flow to future conditions are estimated at several stages into theevaluate the system's performance with different flow future. The system analysis should be performed for eachsequences (see paragraph 5-5). Future system flows repli- stage. While these analyses will take additional time andcate the period-of-record. Also, projected changes in the effort , they will also provide some indication of howbasin should be factored into the analysis. Typically, responsive the system results are to changing conditions.

EM 1110-2-142031 Oct 97

PART 2

HYDROLOGIC ANALYSIS

EM 1110-2-142031 Oct 97

5-1

Chapter 5Hydrologic Engin eering Data

5-1. Meteorological Data Reports” and “ Technical Memorandum” prepared by the

The extent of meteorological observations is determined bythe data needs and use and the availability of personnel andequipment. Data usually recorded at weather stationsinclude air (sometimes water) temperature, precipitation,wind, and evaporation. As indicated below, more extensiverecording of various types of data is often made for specialpurposes. The primary source of meteorological data forthe United States is the National Oceanic and AtmosphericAdministration (NOAA) National Climatic Data Center inAsheville, North Carolina. Data are available from NOAAin computer-readable form as well as published reports.NOAA publication “ Selective Guide to Climatic DataS ources” is an excellent reference for data availability.P rivate vendor sources employing compact disc (CD)technology using NOAA records are also available.

a. Storm meteorology. The determination of runoffpotential, particularly flood potential, in areas wherehydrologic data are scarce can be based on a knowledge ofs torm meteorology. This includes sources of moisture inthe paths over which the storm has traveled, as well as aknowledge of the mechanics of storm activity. Derivationof hydrologic quantities associated with various stormsmust take into consideration the type of storm, its path,potential moisture capacity and stability of the atmosphere,i sobar, wind and isotherm patterns, and the nature andintensity of fronts separating air masses. These are usuallydescribed adequately in the synoptic charts that areprepared at regular (usually 6-hr) intervals for weatherforecasting purposes and associated upper-air soundings.Where such charts are available, it is important that they beretained as a permanent record of meteorological activityfor use in supplementing information contained in theregularly prepared hemispheric charts. These latter chartssummarize the daily synoptic situation throughout thehemisphere but do not contain all of the data that are ofinterest or that would have direct bearing on the derivationof design criteria.

(1) NOAA monthly publication “ Storm Data” (1959-present) documents the time, location, and the meteorologiccharacteristics of all reported severe storms or unusualweather phenomena. Synoptic maps are published byNOAA on a weekly basis in “ Daily Weather Maps, WeeklySeries” and on a monthly basis, “ Synoptic Weather Maps,

Daily Series, Northern Hemisphere Sea-Level and500-Millibar Charts and Data Tabulations.”

(2) Storm data including synoptic charts for selectedhis toric storms are included in the “ Hydrometeorological

National Weather Service (NWS) Office of Hydrology inS i l ver Spring, MD. Other sources of meteorological datainclude the National Hurricane Center in Coral Gables, FL,and state climatologists as well as U.S. Geological Surveyand Corps flood reports.

b. Precipitation. Monthly summaries of observedhourly and daily precipitation data are published by NOAAin “ Climatical Data” and “ Hourly Precipitation Data.”P recipitation data are also available from NOAA incomputer-readable media. Precipitation data for significanthistoric storms (1870's-1960's) are tabulated in “ StormRainfall in the United States, Depth, Area, Duration Data.”

(1) There are usually local, or state agencies, collect-ing precipitation data for their own use. These data couldprovide additional storm information. However, precipita-tion measurements at remote unattended locations may notbe consistently and accurately recorded, particularly wheresnow and hail frequently occur. For this reason, recordsobtained at unattended locations must be interpreted withcare. When an observer is regularly on-site, the times ofoccurrence of snowfall and hail should be noted to makeaccurate use of the data. The exact location and elevationof the gauge are important considerations in precipitationmeasurement and evaluation. For uniform use, this is bestexpressed in terms of latitude and longitude and in metersor feet of elevation above sea level. Of primary importancein processing the data is tabulating precipitation at regularintervals. This should be done daily for non-recordinggauges with the time of observation stated. Continuouslyrecording gauges should be tabulated hourly. The originalrecording charts should be preserved in order to permits tudy of high-intensity precipitation during short intervalsfor certain applications.

(2) Procedures to develop standard project and prob-able maximum precipitation estimates are presented inNW S hydrometeorological reports and technical memo-randum. Chapter 7 of this manual provides an overview ofhypothetical storms and their application to flood-runoffanalysis.

c. Snowpack. Where snowmelt contributes signifi-cant ly to runoff, observations of the snowpack characteris-t i cs can be of considerable value in the development ofhydrologic design criteria. The observation of water

EM 1110-2-142031 Oct 97

5-2

equivalent (weight) of a vertical column sampled from the maximum and minimum temperatures should be maintainedsnowpack at specified locations and observation times is of and, where feasible, published. The NOAA report seriesprimary importance. As the observations will ordinarily be on “ Climatography of the United States” by city, state, orused as an index for surrounding regions, the elevation and region also provides information on daily and monthlyexposure of the location must be known. The depth of normal temperatures.snowpack is of secondary importance, but some obser-vations of the areal extent of the snowpack are often useful. e. Moisture. Atmospheric moisture is a major factor

(1) An important element in processing snowpack can be measured by atmospheric soundings which recordobservation data is the adjustment of observations at all temperature, pressure, relative humidity, and other items.locations to a common date, such as the first of each month T otal moisture in the atmosphere can be integrated andduring the snowpack accumulation season. Since these expressed as a depth of water. During storms, the verticalobservations are often made by traveling survey teams, they distribution of moisture in the atmosphere ordinarilyare not made simultaneously. Also, they cannot always be follows a rather definite pattern. Total moisture canmade at a specified time because it is impossible to obtain therefore be related to the moisture at the surface, which isaccurate or representative measurements during a function of the dew point at the surface. Accordingly, asnowstorms. record of daily dew points is of considerable value. Here,

(2) Where continuous recordings of snowpack water s tat ion must be known. NOAA publication “ Localequivalent by means of radioactive gauges or snow pillows Climatological Data” is a primary source for observed deware available, these can be used on a basis for adjusting point, pressure, and temperature data.manual observations at nearby locations. Otherwise, somejudgment or correlation technique based on precipitation f. Winds. Probably the most difficult meteorologicalmeasurements is required to adjust the observations data to element to evaluate is wind speed and direction. Quitea uniform date at all locations. It is important to preserve commonly, the direction of surface winds reversesthe original records whenever such adjustments are made. diurnally, and wind speeds fluctuate greatly from hour toHowever, data that are disseminated for use in design hour and minute to minute. There is also a radical changeshould be the adjusted systematic quantities. of wind speed and direction with altitude. The speed and

(3) T he primary agency for the collection and distri- obst ructions such as mountains, and locally by smallbut ion of snowpack data is the Soil Conservation Service obstructions such as buildings and trees. Accordingly, it is(S CS ) (Department of Agriculture, Washington, D.C.). important that great care is exercised in selecting a locationFrom January through May each year, the SCS publishes a and altitude for wind measurement. For most hydrologicmonthly report titled, “ Water Supply Outlook.” These applications, wind measurements at elevations of 5 to 15 mreports provide snowpack and streamflow forecast data for above the ground surface are satisfactory. It is important toeach state and region. The SCS also issues “ Basin Outlook preserve all basic records of winds, including data on theReports,” a monthly regional summary of snow depth and locat ion, ground elevation, and the height of thewater content. Additionally, NOAA distributes an annual anemometer above the ground. An anemometer is antabulation of snowpack data in their “ Snow Cover instrument for measuring and indicating the force or speedS urveys.” Climatoligcal precipitation data published of the wind. Where continuous records are available,monthly by NOAA also include information on snowfall hourly tabulations of speed and direction are highlyand snow on the ground. desirable. Total wind movement and the prevailing direc-

d. Temperatures. In most hydrologic applications of each state are published in NOAA publications “ Localtemperature data, maximum and minimum temperatures for Climatological Data” and “ Climatological Data.”each day at ground level are very useful. Continuousrecords of diurnal temperature variations at selected loca- g. Evaporation. Evaporation data is usually requiredtions can be used to determine the daily temperature pattern for reservoir studies, particularly for low-flow analysis.fai rly accurately at nearby locations where only the Reservoir evaporation is typically estimated by measuringmaximum and minimum temperatures are known. In pan evaporation or computing potential evaporation. Thereapplying temperature data to large areas, it must be recog- are several methods of estimating potential evaporation,nized that temperatures normally decrease with increasing based on meteorological information. Pruitt (1990)elevation and latitude. It is also important to preserve all of reviewed various approaches in an evaluation of thethe original temperature records. Summaries of daily methodology and results published in “ A Preliminary

influencing the occurrence of precipitation. This moisture

again, the elevation, latitude and longitude of the measuring

direction at lower levels is greatly influenced by

tion for each day are also useful data. Daily wind data for

EM 1110-2-142031 Oct 97

5-3

Assessment of Corps of Engineer Reservoirs, Their Pur- hydrologic studies. The USGS 7.5 Minute Series (Scaleposes and Susceptibility to Drought,” (HEC 1990e). 1:24,000), with a typical contour interval of 5 or 10 ft (1.5

(1) Evaporation is usually measured by using a pan The USGS publications “ Catalog of Published Maps” andabout 4 ft (1.2 m) in diameter filled with water to a depth of “ Index to Maps” are excellent guides to readily availableabout 8 in. (0.2 m). Daily evaporation can be calculated by topographic data for each state. Reports by the USGS aresubtracting the previous day's reading from today's reading avai lable through Books and Open File Reports Section,and adding the precipitation for the intervening period. The US GS, Federal Center, Box 25425, Denver, CO, or bypan should be occasionally refilled and this fact noted in contacting the National Technical Information Servicethe record. This volume of added water, divided by the (NTIS), 5285 Port Royal Road, Springfield, VA 22161.area of the pan, is equal to the daily evaporation amountexpressed in inches or millimeters. A tabulation of daily b. Digital mapping. Increasingly, topographic dataevaporation amounts should be maintained and, if possible, are avai lable in digital form. One form of computerpublished. It is essential that a rain gauge be maintained at description of topography is a digital elevation modelthe evaporation pan site, and it is usually desirable that (DEM). Geographic information systems (GIS) can linktemperature, dew point (or wet-bulb temperature) and low- land attributes to topographic data and other informationlevel wind measurements also be made at the site for future concerning processes and properties related to geographicstudy purposes. locations. DEM and GIS representations of topologic data

(2) NOAA Technical Report “ NWS 33, Evaporation S ome of the earliest applications in hydrologic modelingAt las for the Contiguous United States” (Farnsworth, used grid cell (raster) storage of information. An exampleThompson, and Peck 1982) provides maps showing annual of raster-based GIS is the Corps' Geographic Resourceand May-October evaporation in addition to pan Analysis Support System (GRASS). An alternate approachcoefficients for the contiguous United States. Companion ut i lizes a collection of irregularly spaced points connectedreport “ NWS 34, Mean Monthly, Seasonal, and Annual Pan by lines to produce triangles, known as triangular irregularEvaporation for the United States” documents monthly network (TIN). The use of DEM and TIN data andevaporation data which was used in the development of the processing software is rapidly changing and may soonevaporation atlas. Daily observed evaporation data are become the standard operation for developing terrain andpublished for each state in NOAA publications “ Local related hydrologic models. A review of GIS applications inClimatological Data” and in “ Climatological Data.” hydrology is provided by DeVantier et al. (HEC 1993).

h. Upper air soundings. Upper air soundings are c. Stream patterns and profiles. Where detailedavai lable from NOAA National Climatic Data Center in studies of floodplains are required, computation of water-Asheville, NC. The soundings provide atmospheric pres- surface profiles is necessary. Basic data needed for thissure, temperature, dew point temperature, wind speed, and computation include detailed cross sections of the river anddirection data from which lapse rate, atmospheric stability, overbank areas at frequent intervals. These are usuallyand jet stream strength can be determined. These meteo- obtained by special field surveys and/or aerial photography.rological parameters are necessary to a comprehensive When these surveys are made, it is important to documentstorm study. and date the data and resulting models, then permanently

5-2. Topographic Data

a. Mapping. For most hydrologic studies, it isessent ial that good topographic maps be used. It isimportant that the maps contain contours of ground-surfaceelevation, so that drainage basins can be delineated andimportant features such as slopes, exposure, and streampatterns can be measured. United States OperationalNavigation Charts, with a scale of 1:1,000,000 and contourintervals of 1,000 ft, are available for most parts of theworld. However, mapping to a much larger scale(1:25,000) and smaller contour intervals in the range of5-20 ft (1.5 to 6 m) are usually necessary for satisfactory

or 3.0 m), provides a good basic map for watershed studies.

are part of a general group of digital terrain models (DTM).

preserve the information so it is readily available for futurereference. Observations of actual water-surface elevationsduring maximum flood stages (high-water marks) areinvaluable for calibrating and validating models for profilecomputations.

d. L akes and swamps. The rate of runoff from anywatershed is greatly influenced by the existence of lakes,swamps, and similar storage areas. It is therefore importantto indicate these areas on available maps. Data on theoutlet characteristics of lakes are important because, in theabsence of outflow measurements, the outflow can often becomputed using the relationship between the amount ofwater stored in the lake and its outlet characteristics.

EM 1110-2-142031 Oct 97

5-4

e. Soi l and geology. Certain maps of soils and documented, and the effect of errors in the estimates on thegeology can be very useful in surface-water studies if they technical procedure and results should be considered.show characteristics that relate to perviousness of the basin.These can be used for estimating loss rates during storms. a. Measurement. Streamflow data are usually bestOf particular interest are areas of extensive sandy soils that obtained by means of a continuous record of river stage,do not contribute to runoff and areas of limestone and supplemented by frequent meter measurements of flowsvolcanic formations that are highly pervious and can store that can be related to corresponding river stages. It islarge amounts of water beneath the surface in a short time. important that stage measurements be made at a goodAddi tionally, watershed sediment yield estimates will cont rol section, even if a weir or other control structuresdepend on similar information. The SCS soil survey must be constructed. Each meter measurement shouldreports are the primary source of soil and permeability data. cons is t of velocity measurements within each of severalState geologic survey or mineral resource agencies are also (6-20, where practical) subdivisions within the channela useful source of geologic data. cross section. Velocity for a subdivision is usually taken as

f. Vegetal cover. Often the type of vegetation more from the surface to the streambed or as the average ofaccurately reflects variation in hydrologic phenomena than velocities taken at 20 and 80 percent (0.2 and 0.8) of thedoes the type of soil or the geology. In transposing infor- depth at the middle of the subdivision. River stage readingsmation to areas of little or no hydrologic data, generalized should be made immediately before and after the crossmaps of vegetal cover are very helpful. As with soil and section is metered. The average of these two stages is thegeology, vegetation has a significant impact on sediment stage associated with the measurement. The measurementyield. In the arid southwest, time since the watershed last i s computed by integrating the rates of flow (m /s) in allburned is a significant parameter in estimating total sedi- subdivisions of the cross section.ment yield for a storm event. The U.S. Forest Service, theAgricultural Stabilization and Conservation Service (1) Measurements of stream velocity and computed(AS CS ) and, in western states, the Bureau of Land Man- streamflow are usually recorded on standard forms. Whenagement are sources of vegetal cover maps. State forest, measurements have been made for a sufficient range ofagricultural agencies, or USGS topographic maps also flows, the rating curve of flow versus stage can beprovide information on vegetal cover. developed. The rating curve can be used to convert the

g. Existing improvements. Streamflow at any partic- T he flows should be averaged for each day in order toular location can be greatly affected by hydraulic structures construct a tabulation of mean daily flows. This constituteslocated upstream. It is important, therefore, that essential the most commonly published record of runoff. data be obtained on all significant hydraulic structureslocated in and upstream from a study area. For diversion (2) For flood studies, it is particularly important tostructures, detailed data are required on the size of the obtain accurate records of short-period variations duringdivers ion dam, capacity of the diversion canal, and the high river stages and to obtain meter measurements at orprobable size of flood required to wash out the diversion near the maximum stage during as many floods as possible.dam. In the case of storage reservoirs, detailed data on the W here the river profile is very flat, as in estuaries andrelation of storage capacity to elevation, location, and size major rivers, it is advisable to obtain measurementsof outlets and spillways, types, sizes, and operation of frequently on the rising and on the falling stage to deter-cont rol gates, and sizes of power plant and penstocks mine if a looped, or hysteresis, effect exists in the rating.should be known. Bridges can produce backwater effects T he reason for this is that the hydraulic slope can changewhich will cause upstream flooding. This flooding may be greatly, resulting in different rating curves for rising andproduced by the approach roads, constriction of the channel falling stages.and floodplain, pier shapes, the angle between the piers andthe streamflow, or the pier length-width ratios.

5-3. Streamflow Data

The availability of streamflow data is a significant factor inthe selection of an appropriate technical method forreservoir studies. It is important to be cognizant of thenature, source, reliability, and adequacy of available data.If estimates are needed, the assumptions used should be

the velocity at a depth of 60 percent (0.6) of the distance

3

continuous record of stage into a continuous record of flow.

b. Streamflow data sources. The USGS is the pri-mary agency for documenting and publishing flow data inthe United States. Daily flow data for each state arepublished in the USGS annual “ Water Data Report.” TheUSGS National Water Data Exchange (NAWDEX) com-puterized database identifies sources of water data. TheNational Water Data Storage and Retrieval System(W ATSTORE) provides processing, storage, and retrievalof surface water, groundwater, and water quality data.

EM 1110-2-142031 Oct 97

5-5

NAWDEX is only an index of the contents ofWATSTORE. These programs will eventually be inte-grated into a National Water Information System, whichwi l l also combine the National Water-Use InformationProgram and Water Resources Scientific InformationCenter (Mosley and McKerchar 1993). Commercial dataservices have also provided convenient access to USGSdaily and peak flow files on CD.

c. Flow conditions. Reservoirs substantially alter thedistribution of flow in time. Many other developments,such as urbanization, diversions, or cultivation andirrigation of large areas can also have a significant effect onwatershed yield and the distribution of flow in time. Thedegree that flows are modified depends on the scale andmanner of the development, as well as the magnitude, time,and areal distribution of rainfall (and snowmelt, ifpertinent). Most reservoir evaluations require an assess-ment based on a consistent flow data set. Various terms areused to define what condition the data represent:

(1) Natural conditions in the drainage basin are definedas the hydrologic conditions that would prevail if no regula-tory works or other development affecting basin runoff andst reamflow were constructed. The effects of natural lakesand swamp areas are included.

(2) Present conditions are defined as the conditions thatexis t at, or near, the time of study. If there are upstreamreservoirs in the basin, the observed flow record wouldrepresent “ regulated flow.” Flow records, precedingcurrent reservoir projects, would be adjusted to reflectthose project operations in order to have consistent “ presentconditions” flow.

(3) Unregulated conditions reflect the present (orrecent) basin development, but without the effect of reser-voir regulation. Unlike natural conditions which aredi fficult to determine, only the effect of reservoir operationand major diversions are removed from the historic data.

(4) Without-project conditions are defined as thecondi tions that would prevail if the project under consid-eration were not constructed but with all existing and futureprojects under construction assumed to exist.

(5) W ith-project conditions are defined as the condi-tions that will exist after the project is completed and aftercompletion of all projects having an equal or higher priorityof construction.

5-4. Adjustment of Streamflow Data

The adjustment of recorded streamflows is often requiredbefore the data can be used in water resources developmentstudies. This is because flow information usually isrequired at locations other than gauging stations and forconditions of upstream development other than those underwhich flows occurred historically. In correlating flowsbetween locations, it is important to use “ natural” flows(unaffected by artificial storage and diversion) in order thatcorrelation procedures will apply logically and efficiently.In generating flows, natural flows should be used becausegeneral frequency functions, characteristic of natural flows,are employed in this process.

a. Natural conditions. When feasible, flow datashould be converted to natural conditions. The conversionis made by adding historical storage changes (plus netevaporation) and upstream diversions (less return flows) tohistorical flows at the gauging stations for each timeinterval in turn. Under some conditions, it may be neces-sary to account for differences in channel and overbankinfiltration losses, distributary flow diversions, travel times,and other factors.

b. Unregulated conditions. It is not always feasibleto convert flows to natural conditions. Often, required dataare not available. Also, the hydrologic effects and timingof some basin developments are not known to sufficientlydefine the required adjustments. An alternative is to adjustthe data to a uniform basin condition, usually near currenttime. The primary adjustments should remove specialinfluences, such as major reservoirs and diversions, thatwould cause unnatural variations of flow.

c. Reservoir holdouts. The primary effect of reser-voi r operation is the storage of excess river flow duringhigh-flow periods, and the release of stored water duringlow-flow periods. The flow adjustment process requiresthe addition of the change in water stored (hold-outs) ineach time step to the observed regulated flow. Holdouts,both positive and negative, are routed down the channel toeach gauge and algebraically added to the observed flow.Hydrologic routing methods, typically used for theseadjustments, are described in Chapter 9 of EM 1110-2-1417. The HEC Data Storage System (HEC-DSS) software(HEC 1995) provides a convenient data managementsystem and utilities to route flows and add, subtract, oradjust long time-series flow data.

EM 1110-2-142031 Oct 97

5-6

d. Reservoir losses. The nonproject inflow represents ratio of precipitation falling on the reservoir. This is oftenthe flow at the project site without the reservoir and estimated to be 0.7 for semi-arid regions. This increase inincludes runoff from the entire effective drainage area runoff is subtracted from gross reservoir evaporation, oftenabove the dam, including the reservoir area. Under non- estimated as 0.7 of pan evaporation, to obtain a net loss.project conditions, runoff from the area to be inundated bythe reservoir is ordinarily only a fraction of the total e. Other losses. In final project studies it is oftenprecipitation which falls on that area. However, under necessary to consider other types of project losses whichproject conditions infiltration losses over the reservoir area may be of minor importance in preliminary studies. Often,are usually minimal during a rainfall event. Thus, these losses cannot be estimated until a project design haspractically all the precipitation falling on the reservoir area been adopted. The importance of these losses is dependentwill appear as runoff. Therefore, the inflow will be greater upon their relative magnitude. That is, losses of 5 m /sunder project conditions than under nonproject conditions, might be considered unimportant for a stream which has ai f inflow is defined as total contribution to the reservoir minimum average annual flow of 1,500 m /s. Such losses,before evaporation losses are considered. In order to though, would be significant from a stream with a minimumdetermine the amount of water available for use at the average annual flow of 25 m /s. Various types of losses arereservoir, evaporation must be subtracted from project discussed in the following paragraphs.inflow. In operation studies, nonproject inflow is ordinarilyconverted to available water in one operation without com- (1) T he term “ losses” may not actually denote aputing project inflow as defined above. This is done in one physical loss of water from the system as a whole. Usually,of two ways: by means of a constant annual loss each year water unavailable for a specific project purpose is called awith seasonal variation or with a different loss each period, “ loss” for that purpose although it may be used at someexpressed as a function of observed precipitation and other point or for some other purpose. For example, waterevaporation. These two methods are described in the which leaves the reservoir through a pipeline for municipalfollowing paragraphs. water supply or fish hatchery requirements might be called

(1) Constant annual loss procedure consists of esti- conduits, and spillway gates are considered losses tomating the evapotranspiration and infiltration losses overhydropower generation, but they are ordinarily not losses tothe reservoir area for conditions without the project, and the flow requirements at a downstream station. Furthermore,evaporation and infiltration losses over the reservoir area such losses that become available for use below the damwith the project. Nonproject losses are usually estimated as should be added to inflow at points downstream from thethe difference between average annual precipitation and project.average annual runoff at the location, distributed seasonallyin accordance with precipitation and temperature variations. (2) Leakage at a dam or in a reservoir area is consid-These are expressed in millimeters of depth. Under project ered a loss for purposes which are dependent upon avail-conditions, infiltration losses are usually ignored, and ability of water at the dam or in the reservoir itself. Theselosses are considered to consist of only direct evaporation purposes include power generation, pipelines from thefrom the lake area, expressed in millimeters for each period. reservoir, and any purpose which utilizes pump intakesT he difference between these losses is the net loss due to which are located at or above the dam. As a rule, leakagethe project. Figure 5-1 illustrates the differences between through, around, or under a dam is relatively small and isnonproject and project losses. difficult to quantify before a project is actually constructed.

(2) The variable loss approach uses historical records geologic area may be used as a basis for estimating losses atof long-term average monthly precipitation and evaporation a proposed project. The amount of leakage is a function ofdata to account for the change in losses due to a reservoir the type and size of dam, the geologic conditions, and theproject. This is accomplished by estimating the average hydrostatic pressure against the dam.runoff coefficient, the ratio of runoff to rainfall, for thereservoir area under preproject conditions and subtracting (3) Leakage from conduits and spillway gates is athis from the runoff coefficient for the reservoir area under function of gate perimeter, type of seal, and head on theproject conditions. The runoff coefficient for project gate, and it varies with the square root of the head. Thecondi t ions is usually 1.0, but a lower coefficient may be amount of leakage may again be measured at existingused if substantial infiltration or leakage from the reservoir projects with various types of seals, and a leakage rateis anticipated. The difference between preproject and computed per meter of perimeter for a given head. Thisproject runoff coefficients is the net gain expressed as a rate may then be used to compute estimated leakage for a

3

3

3

a loss to power. Likewise, leakages through turbines, dams,

In some cases, the measured leakage at a similar dam or

EM 1110-2-142031 Oct 97

5-7

Figure 5-1. Project and nonproject reservoir losses

proposed project by using the proposed size and number of rate is then reduced by multiplying by the proportion ofgates and the proposed head on the gates. time the unit will be closed. For example, suppose that

(4) If a proposed project will include power, and if the and the operation records indicate that the unit is closedarea demand is such that the turbines will sometimes be 60 percent of the time. The average leakage rate would beidle, it is advisable to estimate leakage through the estimated at 0.1 × 0.6 or 0.06 m /s.turbines when closed. This leakage is a function of thetype of penstock gate, type of turbine wicket gate, number (5) The inclusion of a navigation lock at a damof turbines, and head on the turbine. The actual leakage requires that locking operations and leakages through thethrough a turbine may be measured at the time of accep- lock be considered. The leakage is dependent upon the lifttance and during annual maintenance inspections, or the or head, the type and size of lock, and the type of gates andmeasurements of similar existing projects may be used to seals. Again, estimates can be made from observed leakageestimate leakage for a proposed project. An estimate of at similar structures. Water required for locking operationsthe percent of time that a unit will be closed may be should also be deducted from water available at the damobtained from actual operational records for existing units site. These demands can be computed by multiplying thein the same demand area. The measured or estimated leakage volume of water required for a single locking operation

leakage through a turbine has been measured at 0.1 m /s,3

3

EM 1110-2-142031 Oct 97

5-8

times the number of operations anticipated in a given time characteristics. Correlation procedures and suggestedperiod and converting the product to a flow rate over the bas in characteristics are described in Chapter 9 ofgiven period. EM 1110-2-1415.

(6) Water use for purposes related to project operations g. Preproject conditions. After project flows for ais often treated as a loss. Station use for sanitary and drink- specified condition of upstream development are obtaineding purposes, cooling water for generators, and water for for all pertinent locations and periods, they must be con-condensing operations have been estimated to be about 0.06 verted to preproject (nonproject) conditions. Nonprojectm /s per turbine at some stations in the southwestern United conditions are those that would prevail during the lifetime3

S tates. Examining operation records for comparable of the proposed project if the project was not constructed.projects in a given study area may also be useful in This conversion is made by subtracting projected upstreamest imat ing these losses. If house units are included in a diversions and storage changes and by accounting forproject to supply the project's power requirements, data evaporation, return flows, differences in channelshould be obtained from the designer in order to estimate infiltration, and timing. Where nonproject conditions willwater used by the units. vary during the project lifetime, it is necessary to convert to

(7) The competitive use of water should also be con- and end of the proposed project life. Separate operations idered when evaluating reservoir losses. When initially studies would then be made for each condition. Thisestimating yield rates for various project purposes at a conversion to future conditions can be made simultaneouslymul t iple-purpose project, competitive uses of water are with project operation studies, but a separate evaluation ofoften treated as losses. For example, consider a pro- nonproject flows is usually required for economicposed project on a stream with an average minimum evaluation of the project.usable flow of 16 m /s. The reservoir of this project is3

to supply 1.5 m /s by pipeline for downstream water3

supply and 2.0 m /s for a fish hatchery in addition to3

providing for hydroelectric power production. The mini-mum average flow available for power generation is thus,16 - (1.5 + 2.0) = 12.5 m /s. Care should be exercised in3

accounting for all such competitive uses when makingpreliminary yield estimates.

f. Missing data.

(1) After recorded flows are converted to uniformconditions, flows for missing periods of record at eachpertinent location should be estimated by correlation withrecorded flows at other locations in the region. Usually,only one other location is used, and linear correlation offlow logarithms is used. It is more satisfactory, however, touse all other locations in the region that can contributeindependent information on the missing data. Althoughthis would require a large amount of computation, thecomputer program HEC-4 Monthly Streamflow Simulationaccomplishes this for monthly streamflow (HEC 1971).

(2) Flow estimates for ungauged locations can beestimated satisfactorily on a flow per basin area basis insome cases, particularly where a gauge exists on the samestream. In most cases, however, it is necessary to correlatemean flow logarithms (and sometimes standard deviation offlow logarithms) with logarithms of drainage area size,logarithm of normal seasonal precipitation, and other basin

two or more sets of conditions, such as those at the start

5-5. Simulation of Streamflow Data

a. Introduction. The term “ simulation” has beenused to refer to both the estimation of historic sequencesand the assessment of probable future sequences ofs t reamflow. The former reference concerns the applicationof continuous precipitation-runoff models to simulatest reamflow based on meteorologic input such as rainfalland temperature. The latter reference concerns the appli-cat ion of stochastic (probabilistic) models that employMonte Carlo simulation methods to estimate the probableoccurrence of future streamflow sequences. Assessment ofthe probable reliability of water resource systems can bemade given the assessment of probable future sequences offlow. Statistical methods used in stochastic models canalso be employed to augment observed historic data byfilling in or extending observed streamflow records.

b. Historic sequences from continuous precipitation-runoff models. Many different types of continuous simula-tion runoff models have been used to estimate the historicsequence of streamflow that would occur from observedprecipitation and other meteorologic variables. Among themost prominent are the various forms of the StanfordWatershed Model (Mays and Tung 1992) and the SSARRModel used by the North Pacific Division (USACE 1991).F or a further description of the application of the modelssee EM 1110-2-1417 Section 8.

EM 1110-2-142031 Oct 97

5-9

c. Stochastic streamflow models. Stochastic stream- s ignificantly with the number of simulations. For furtherflow models are used to assess the probable sequence of explanation of this point, see “ Stochastic Analysis offuture flows. As with any model, a model structure is Drought Phenomena” (HEC 1985b).assumed, parameters are estimated from observed data, andthe model is used for prediction (Salas et al. 1980). d. Assessment of reliability with stochastic stream-Typically, stochastic streamflow models are used to simu-flow models. The advantage of using a stochastic stream-late annual and/or monthly streamflow volumes. Stochastic flow model over that of employing only historic records iss t reamflow models have not been successfully developed that it can be used to provide a probabilistic estimate of afor daily streamflow. water resource system's reliability. For example, the

(1) Although many different structures have been certain goals can be estimated by simulating the stochasticproposed in the research literature, regression is most flow sequences with a reservoir simulation model. Oncecommonly used as the basis for stochastic streamflow again, the number of flow sequences used are sufficientmodels. The regressions involve both correlation between when the estimate of the probabilities stabilize.flows at different sites and the correlation between currentand past flow periods, termed serial correlation. The e. Available software for stochastic streamflowcorrelation between sites is useful in improving parameterss imulation. HEC-4, “ Monthly Streamflow Simulation”estimates from regional information. The serial correlation (HEC 1971), and LAST (Lane 1990) are public domainbetween periods is important in modeling the persistence, software for performing stochastic streamflow simulation.or the tendency for high flow or drought periods. A HEC-4 performs monthly stochastic streamflow simulation.random error component is added to the regression to LAST utilizes a more modern approach where annual andprovide a probabilistic component to the model. shorter time period (seasonal, monthly, etc.) stochastic

(2) The model parameters are estimated to preserve thecorrelation structure observed in the observed data. If the f. Extending and filling in historic records. Statis-appropriate correlation structure is preserved, then the t i cal techniques can be used to augment existing historicregression residuals should closely approximate the records by either “ filling in” missing flow values orbehavior that would be expected from a random error extending the observed record at a gauge based on obser-component. vations at other gauges. The statistical techniques used are

(3) Model prediction is performed via the application and are a modification of regression based techniquesof Monte Carlo simulation. Monte Carlo simulation is a (Alley and Burns 1983, and Salas 1992). MOVEnumerical integration technique. This numerical technique algorithms have been instituted because the variance ofi s necessary because the stochastic model effectively repre- series augmented by regression alone is underestimated.sents a complex joint probability distribution of T he MOVE technology is only generally applicable whenst reamflows in time and space that cannot be evaluated serial correlation does not exist in the streamflow records.analytically. The simulation is performed by producing However, monthly or annual sequences of streamflowrandom sequences of flows via a computer algorithm that volumes usually do exhibit a degree of serial correlation.employs random number generators. These sequences of In these circumstances, the information provided by theflows are analyzed to assess supply characteristics, for longer record station may not be useful in extending aexample the probability for a certain magnitude or duration shorter record station. For a discussion of the impact ofof drought. The number of flow sequences generated is serial correlation see Matalas and Langbein (1962) andsufficient when the estimated probabilities do not change Tasker (1983).

probability that a particular reservoir will be able to meet

streamflow can be co-simulated.

referred to as MOVE, maintenance of variance extension,

EM 1110-2-142031 Oct 97

6-1

Chapter 6Hydrologic Frequency Det erminations

6-1. Introduction

F requency curves are most commonly used in Corps ofEngineers studies to determine the economic value of floodreduction projects. Reservoir applications also include thedetermination of reservoir stage for real estate acquisitionand reservoir-use purposes, the selection of rainfallmagnitude for synthetic floods, and the selection of runoffmagnitude for sizing flood-control storage.

a. Annual and partial-duration frequency. There aretwo bas ic types of frequency curves used in hydrologicwork. A curve of annual maximum events is ordinarilyused when the primary interest lies in the very large eventsor when the second largest event in any year is of minorconcern in the analysis. The partial-duration curverepresents the frequency of all events above a given basevalue, regardless of whether two or more occurred in thesame year. This type of curve is ordinarily used in eco-nomic analysis when there are substantial damages result-ing from the second largest and third largest floods inextremely wet years. Damage from floods occurring morefrequently than the annual event can occur in agriculturalareas, when there is sufficient time between events forrecovery and new investment. When both the frequencycurve of annual floods and the partial-duration curve areused, care must be exercised to assure that the two areconsistent.

b. Seasonal frequency curves. In most locations,there are seasons when storms or floods do not occur or arenot severe, and other seasons when they are more severe.Also, damage associated with a flood often varies with theseason of the year. In studies where the seasonal variationis of primary importance, it becomes necessary to establishfrequency curves for each month or other subdivision of theyear. For example, one frequency curve might representthe largest floods that occur each January; a second onewould represent the largest floods that occur each February,etc. In another case, one frequency curve might representfloods during the snowmelt season, while a second mightrepresent floods during the rainy season. Occasionally,when seasons are studied separately, an annual-event curvecovering all seasons is also prepared. Care should beexercised to assure that the various seasonal curves areconsistent with the annual curve.

6-2. Duration Curv es

a. Flow duration curve. In power studies, for run-of-river plants particularly and in low-flow studies, theflow-duration curve serves a useful purpose. It simplyrepresents the percent of time during which specified flowrates are exceeded at a given location. Ordinarily, varia-t ions within periods less than 1 day are inconsequential,and the curves are therefore based on observed mean-dailyflows. For the purposes served by flow duration curves, theext reme rates of flow are not important, and consequentlythere i s no need for refining the curve in regions of highflow.

b. Preparing flow-duration curve. The procedureordinarily used to prepare a flow duration curve consists ofcounting the number of mean-daily flows that occur withingiven ranges of magnitude. The lower limit of magnitudein each range is then plotted against the percentage of daysof record that mean-daily flows exceed that magnitude. Aflow duration curve example is shown in Figure 6-1.

6-3. Flood-Frequency Determinations

At many locations, flood stages are a unique function offlood discharges for most practical purposes. Accordingly,i t i s usual practice to establish a frequency curve of riverdischarges as the basic hydrologic determination for flooddamage reduction project studies. In special cases, factorsother than river discharge, such as tidal action oraccumulated run-off volume, may greatly influence rivers tages. In such cases, a direct study of stage frequencybased on recorded stages is often warranted.

a. Determination made with available data. Whererunoff data at or near the site are available, flood-frequencydeterminations are most reliably made by direct study ofthese data. Before frequency studies of recorded flows aremade, the flows must be converted to a uniform condition,usually to conditions without major regulation or diversion.Developing unregulated flow requires detailed routingstudies to remove the effect of reservoir hold-outs anddiversions. As damaging flows occur during a very smallfraction of the total time, only a small percentage of dailyflows are used for flood-frequency studies. These consistof the largest flow that occurs each year and the secondarypeak flows that cause damage. However, for mostreservoirs studies, the period-of-record flow will berequired for analysis of nonflood purposes and impacts.

EM 1110-2-142031 Oct 97

6-2

Figure 6-1. Example flow duration curve

b. Historical data. Flood frequency estimates are damaging events are ordinarily made separately. Thesubject to considerable uncertainty, even when fairly long addition of historical information can be very important inrecords are available. In order to increase the reliability of verifying the frequency of large recorded events.frequency estimates, empirical theoretical frequency Historical information on large damaging floods can berelations are used in specific frequency studies. These obtained through standard sources such as USGS waterstudies require that a complete set of data be used. In supply series or from newspapers and local museums. Theorder to comply with this requirement, the basic frequency latter sources often are more qualitative but give importantstudy ordinarily is based on the maximum flow for each insight into the relative frequency of recorded events.year of record. Supplementary studies that include other

EM 1110-2-142031 Oct 97

6-3

c. Selecting and computing frequency curves. The accordingly to make inferences regarding their “ parentunderlying general assumption made in all frequencypopulation” (i.e., the distribution from which they werestudies is that each observed event represents an approxi- derived). The procedure is applied to annual maxima ofmately equal proportion of the future events that will occur unregulated flow, which are assumed to be independentat the location, if controlling conditions do not change. random events. The fact that the set of observations couldDetailed procedures for selecting data and computing resul t from any of many sets of physical conditions orflood-frequency curves are presented in EM 1110-2-1415 distributions leaves considerable uncertainty in the derivedand HEC-FFA program (HEC 1992c). curve. However, using statistical processes, the most

d. Regional correlation of data. Where runoff data at were derived can be estimated. Because this in allor near the site do not exist or are too fragmentary to probabi lity is not the true parent population, the relativesupport direct frequency calculations, regional correlation chance that variations from this distribution might be trueof frequency statistics may be used for estimating fre- must be evaluated. Each range of possible parentquencies. These correlations generally relate the mean and populations is then weighted in proportion to its likelihoodstandard deviation of flows to drainage basin characteristics in order to obtain a weighted average. A probabilityand location. Techniques of regional correlation are obtained from this weighted average is herein referred to aspresented in Chapter 9 of EM 1110-2-1415. the expected probability P . Chapter 3 of EM 1110-2-1415

e. Ext reme floods. In the analysis of reservoir proj-ects, the project's performance during floods larger than the (3) Regional. Because of the shortness of hydrologicmaximum recorded events is usually required. Extra- records, frequency determinations for rare events arepolating derived frequency relations is uncertain, so special relatively unreliable when based on a single record. Also, itstudies of the potential magnitudes of extreme flood events is often necessary to estimate frequencies for locationsare usually required. The most practical approach is where no record exists. For these reasons, regionalizedthrough examination of rainstorms that have occurred in the frequency studies, in which frequency characteristics areregion and determination of the runoff that would result at related to drainage-basin features and precipitation charac-the project location if these storms should occur in the teri s tics, are desirable. Regionalized frequency studiest ributary area. This subject is discussed in the following usually develop relationships for analytical frequencychapter, “ Flood-Runoff Analysis.” statistics. An alternative approach is to develop predictive

6-4. Estimating Frequency Curv es

a. Approaches. There are two basic approaches toestimating frequency curves--graphical and analytical.Each approach has several variations, but the discussionherein will be limited to recommended methods. Theprimary Corps reference for computing frequency curves isEM 1110-2-1415.

(1) Graphical. Frequencies are evaluated graphicallyby arranging observed values in the order of magnitude andrepresenting frequencies by a smooth curve through thearray of values. Each observed value represents a fractionof the future possibilities and, when plotting the frequencycurve, it is given a plotting position that is calculated togive it the proper weight. Every derived frequency relationshould be plotted graphically, even though the results canbe obtained analytically. Paragraph 2-4 of EM 1110-2-1415 presents “ Graphical Frequency Analysis.”

(2) Analytical. In the application of analytical (sta-t i s t i cal) procedures, the concept of theoretical populationsor distributions is employed. The events that have occurredare presumed to constitute a random sample and are used

probable nature of the distribution from which the data

Ncovers analytical flood-frequency analysis.

equations for the flow for specific recurrence intervals.Chapter 9 of EM 1110-2-1415 presents regression analysisand its application to regional studies.

b. Flood volume-duration frequencies. A compre-hensive flood volume-duration frequency series consists ofa set of: 1, 3, 7, 15, 30, 60, 120, and 183-day average flowsfor each water year. These durations are normally availablefrom the USGS WATSTORE files, and they are the defaultdurations in the computer program STATS (HEC 1987a).Runoff volumes are expressed as average flows in orderthat peak flows and volumes can be readily compared andcoordinated. Paragraph 3-8 of EM 1110-2-1415 coversflood volumes.

c. Low flow frequencies. Reservoir analysis oftenrequires the evaluation of the frequencies of low flowsforvarious durations. The same fundamental procedurescan be used, except that minimum instead of maximumrunoff values are selected from the basic data. For lowflows, the effects of basin development are usually mores igni fi cant than for high flow. A relatively moderatedivers ion may not be very significant during a flood;however, it may greatly modify or even eliminate lowflows. Accordingly, one of the most important aspects of

EM 1110-2-142031 Oct 97

6-4

low flows concerns the evaluation of past and future effects b. Regulated runoff frequency curves. If it is prac-of bas in developments. Chapter 4 of EM 1110-2-1415 tical, the most complete approach to determining frequencydescribes low-flow frequency analysis. curves of regulated runoff consists of routing flows for the

d. Reservoir level frequency. A reservoir frequency works, arranging the annual peak regulated flows in ordercurve of annual maximum storage is ordinarily constructed of magnitude, assigning a plotting position to the peakgraphically, using the procedures for flood-flow frequency. values, plotting the peak flow values at the assignedObserved storage should be used to the extent available, but plotting position, and drawing the frequency curve based ononly if the reservoir has been operated in the past in the plotted data. A less involved method consists of routingaccordance with future plans. If historical data are not the largest floods of record, or multiples of a largeavai lable, or if it is not appropriate for future use, then hypothetical flood, to estimate the regulated frequencyreservoir routings should be used to develop data for curve. This approach requires the assumption that theexpected reservoir operations. Stage-duration curves can frequency of the regulated peak flow is the same as thebe constructed from historical operation data or from unregulated peak flow, which is probably true for thes imulations. These curves are usually constructed for the largest floods. Paragraph 3-9d of EM 1110-2-1415entire period-of-record, or for a selected wet or dry period. describes these methods.F or some purposes, particularly recreational use, theseasonal variation of reservoir stages is of importance, and c. Erratic stage-discharge frequency curves. Ina set of frequency or duration curves for each month of the general, cumulative frequency curves of river stages areyear may be required. Reservoir stage (or elevation) curves determined from frequency curves of flow. In cases whereshould indicate significant reservoir levels such as: the stage-discharge relation is erratic, a frequency curve ofminimum pool, top of conservation pool, top of flood- stages can be derived directly from stage data. Chapter 6 ofcontrol pool, spillway crest elevation, and top of dam. EM 1110-2-1415 presents stage-frequency analysis.

6-5. Effect of Basin Dev elopments on FrequencyRelations

a. Ef fects of flood-control works. Most hydrologicfrequency estimates serve some purpose relating to theplanning, design, or operation of water resources manage-ment projects. The anticipated effects of a project onflooding can be assessed by comparing the peak dischargeand volume frequency curves with and without the project.Also, projects that have existed in the past have affected therates and volumes of floods, and recorded values must beadjusted to reflect uniform conditions in order for thefrequency analysis to conform to the basic assumptions ofrandomness and common population. For a frequencycurve to conform reasonably with a generalizedmathematical or probability law, the flows must beessentially unregulated by man-made storage or diversionst ructures. Consequently, wherever practicable, recordedrunoff values should be adjusted to unregulated conditionsbefore a frequency analysis is made. However, in caseswhere the regulation results from a multitude of relativelysmall hydraulic structures that have not changed appre-ciably during the period of record, it is likely that thegeneral mathematical laws will apply as in the case ofnatural flows, and that adjustment to natural conditionswould be unnecessary. The effects of flood control worksare presented in paragraph 3-9 of EM 1110-2-1415, andeffects of urbanization in paragraph 3-10.

ent i re period of record through the proposed management

d. Reservoir or channel modifications. Projectconstruction or natural changes in streambed elevation maychange the relationship between stage and flow at alocation. By forming constrictions, levees may raise riverstages half a meter for some distance upstream. Reservoiror channel modifications may cause changes in degradationor aggradation of streambeds, and thereby change ratingcurves. Thus, the effect of projects on river stages ofteninvolves the effects on channel hydraulics as well as theeffects on streamflow. Consult EM 1110-2-1416 forinformation on modeling these potential changes.

6-6. Selection of Frequency Data

a. Primary considerations. The primary consider-at ion in selecting an array of data for a frequency study isthe objective of the frequency analysis. If the frequencycurve that is developed is to be used for estimating damagesthat are related to instantaneous peak flows in a stream,peak flows should be selected from the record. If thedamages are related to maximum mean-daily flows or tomaximum 3-day flows, these items should be selected. Ifthe behavior of a reservoir under investigation is related tothe 3-day or 10-day rain-flood volume, or to the seasonalsnowmelt volume, that pertinent item should be selected.Normally, several durations are analyzed along with peakflows to develop a consistent relationship.

EM 1110-2-142031 Oct 97

6-5

b. Selecting a related variable. Occasionally, it is flows are better adapted to analytical methods and are morenecessary to select a related variable in lieu of the one easily compared within a region. Frequency curves ofdes i red. For example, where mean daily flow records are present-regulated conditions (those prevailing under currentmore complete than the records of peak flows, it may be practices of regulation and diversion) or of future-regulateddesirable to derive a frequency curve of mean-daily flows condi tions can be constructed from the frequency curve ofand then, from the computed curve, derive a peak-flow natural flow by means of an empirical or logicalcurve by means of an empirical relation between mean daily relationship between natural and regulated flows. Whereflows and peak flows. All reasonably independent values data recorded at two different locations are to be combinedshould be selected, but the annual maximum events should for construction of a single frequency curve, the dataordinarily be segregated when the application of analytical should be adjusted as necessary to a single location, usuallyprocedures is contemplated. the location of the longer record, accounting for differences

c. Data selected. Data selected for a frequency study channel characteristics between the locations. Where themust measure the same aspect of each event (such as peak stream-gauge location is somewhat different from theflow, mean-daily flow, or flood volume for a specified project location, the frequency curve should be constructedduration), and each event must be controlled by a uniform for the stream-gauge location and subsequently adjusted toset of hydrologic and operational factors. For example, it the project location.would be improper to combine items from old records thatare reported as peak flows but are in fact only daily f. Runoff record interruptions. Occasionally, areadings, with newer records where the peak was actually runoff record may be interrupted by a period of one or moremeasured. Similarly, care should be exercised when there years. If the interruption is caused by the destruction of thehas been significant change in upstream storage regulation gauging station by a large flood, failure to fill in the recordduring the period of record so as not to inadvertently for that flood would have a biasing effect, which should becombine unlike events into a single series. In such a case, avoided. However, if the cause of the interruption isthe entire flow record should be adjusted to a consistentknown to be independent of flow magnitude, the entirecondition, preferably the unregulated flow condition. period of interruption should be eliminated from the

d. Hydrologic factors. Hydrologic factors and rela- no runoff records are available on the stream concerned, itt i onships operating during a winter rain flood are usually is possible to estimate the frequency curve as a whole usingqui te different from those operating during a spring snow- regional generalizations. An alternative method is tomelt flood or during a local summer cloudburst flood. estimate a complete series of individual floods fromWhere two or more types of floods are distinct and do not recorded precipitation by continuous hydrologic simulationoccur predominantly in mutual combinations, they should and perform conventional frequency analysis on thenot be combined into a single series for frequency analysis. simulated record.T hey should be considered as events from different parentpopulations. It is usually more reliable in such cases tosegregate the data in accordance with type and to combineonly the final curves, if necessary. For example, in themountainous region of eastern California, frequency studiesare made separately for rain floods, which occur principallyfrom November through March, and for snowmelt floods,which occur from April through July. Flows for each ofthese two seasons are segregated strictly by cause, thosepredominantly caused by snowmelt and thosepredominantly caused by rain. In desert regions, summerthunderstorms should be excluded from frequency studiesof winter rain flood or spring snowmelt floods and shouldbe considered separately. Similarly, in coastal regions itwould be desirable to separate floods induced by hurricanesor typhoons from other general flood events.

e. Data adjustments. When practicable, all runoffdata should be adjusted to unregulated hydrologic condi-t ions before making the frequency study because natural

of drainage area and precipitation and, where appropriate,

frequency array, since no bias would result. In cases where

6-7. Climatic Variations

S ome hydrologic records suggest regular cyclic variationsin precipitation and runoff potential. Many attempts havebeen made to demonstrate that precipitation or stream flowsdisplay variations that are in phase with various cycles, par-ticularly the well-established 11-year sunspot cycle. Thereis no doubt that long duration cycles or irregular climaticchanges are associated with general changes of land massesand seas and with local changes in lakes and swamps.Also, large areas that have been known to be fertile in thepast are now arid deserts, and large temperate regions havebeen covered with glaciers one or more times. Althoughthe existence of climatic changes is not questioned, theireffect is ordinarily neglected because long-term climaticchanges generally have insignificant effects during theperiod concerned in water development projects, and short-term climatic changes tend to be self-compensating. Forthese reasons, and because of the difficulty in

EM 1110-2-142031 Oct 97

6-6

di fferentiating between fortuitous and systematic changes, reliability of frequency estimates is greater than that ofit is considered that, except for the annual cycle, the effect measurement errors. For this reason, it is usually better toof natural cycles or trends during the period of useful include an estimated magnitude for a major flood than toproject life can ordinarily be neglected in hydrologic ignore it. For example, a flood event that was not recordedfrequency studies. because of gauge failure should be estimated, rather than to

6-8. Frequency Reliability Analyses

a. Inf luences. The reliability of frequency estimatesis influenced by the amount of information available, thevariability of the events, and the accuracy with which thedata were measured.

(1) In general with regard to the amount of informationavailable, errors of estimate are inversely proportional tothe square root of the number of independent itemscontained in the frequency array. Therefore, errors ofest imates based on 40 years of record would normally behalf as large as errors of estimates based on 10 years ofrecord, other conditions being the same.

(2) The variability of events in a record is usually themost important factor affecting the reliability of frequencyest imates. For example, the ratio of the largest to the6-9. Presentation of Frequency Analysis Resultssmallest annual flood of record on the Mississippi River atRed River Landing, LA, is about 2.7; whereas the ratio of Information provided with frequency curves should clearlythe largest to the smallest annual flood of record on the indicate the scope of the studies and include a briefKings R iver at Piedra, CA, is about 100 or 35 times as description of the procedure used, including appropriategreat. Statistical studies show that as a consequence of this references. When rough estimates are adequate or neces-di fference in variability, a flow corresponding to a given sary, the frequency data should be properly qualified infrequency that can be estimated within 10 percent on the order to avoid misleading conclusions that might seriouslyMississippi River, can be estimated only within 40 percent affect the project plan. A summary of the basic dataon the Kings River. consisting of a chronological tabulation of values used and

(3) The accuracy of data measurement normally has helpful. The frequency data can also advantageously berelatively little influence on the reliability of a frequency presented in graphical form, ordinarily on probability paper,estimate, because such errors ordinarily are not systematic along with the adopted frequency curves.and tend to cancel. The influence of extreme events on

omit it from the frequency array. However, it is advisableto always use the most reliable sources of data and to guardagainst systematic errors.

b. Errors in estimating flood frequencies. It shouldbe remembered that possible errors in estimating floodfrequencies are very large, principally because of thechance of having a nonrepresentative sample. Sometimesthe occurrence of one or two rare flood events can changethe apparent exceedance frequency of a given magnitudefrom once in 1,000 years to once in 200 years. Neverthe-less, the frequency-curve technique is considerably betterthan any other tool available for certain purposes andrepresents a substantial improvement over using an arrayrestricted to observed flows only. Reliability criteria usefulfor illustrating the accuracy of frequency determinations aredescribed in Chapter 8 of EM 1110-2-1415.

indicating sources of data and adjustments made would be

EM 1110-2-142031 Oct 97

7-1

Chapter 7Flood-Runoff Analysis

7-1. Introduction

Flood-runoff analysis is usually required for any reservoirproject. Even without flood control as a purpose, a reser-voir must be designed to safely pass flood flows. Rarelyare there sufficient flow records at a reservoir site to meetall analysis requirements for the evaluation of a reservoirproject. This chapter describes the methods used to ana-lyze the flood hydrographs and the application of hypo-thetical floods in reservoir projects. Most of the details onmethods are presented in EM 1110-2-1417. The damsafety standards are dependent on the type and location ofthe dam. ER 1110-8-2 defines the requirements for designfloods to evaluate dam and spillway adequacy. Require-ments for flood development and application are alsoprovided.

7-2. Flood Hydrograph Analysis

a. Unit hydrograph method. The standard Corpsprocedure for computing flood hydrographs from catch-ments is the unit hydrograph method. The fundamentalcomponents are listed below:

(1) Analysis of rainfall and/or snowmelt to determinethe time-distributed average precipitation input to eachcatchment area.

(2) Infiltration, or loss, analysis to determine theprecipitation excess available for surface runoff.

(3) Unit hydrograph transforms to estimate the surfaceflow hydrograph at the catchment outflow location.

(4) Baseflow estimation to determine the subsurfacecontribution to the total runoff hydrograph.

(5) Hydrograph routing and combining to movecatchment hydrographs through the basin and determinetotal runoff at desired locations.

For urban catchments, the kinematic-wave approach isoften used to compute the surface flow hydrograph, insteadof unit hydrograph transforms. Each of the standard floodrunoff and routing procedures is described in Part 2 ofEM 1110-2-1417. HEC-1 Flood Hydrograph Package(HEC 1990c) is a generalized computer program providingthe standard methods for performing the requiredcomponents for basin modeling.

b. Rainfall-runoff parameters. Whenever possibleunit hydrographs and loss rate characteristics should bederived from the reconstitution of observed storm andflood events on the study watershed, or nearby watershedswith similar characteristics. The HEC-1 program hasoptimization routines to facilitate the determination of best-fit rainfall-runoff parameters for each event. When runoffrecords are not available at or near the location of interest,unit hydrograph and loss characteristics must bedetermined from regional studies of such characteristicsobserved at gauged locations. Runoff and loss coefficientscan be related to drainage basin characteristics by multiplecorrelation analysis and mapping procedures, asdescribed in Chapter 16, “Ungauged Basin Analysis” ofEM 1110-2-1417.

c. Developing basin models. Flood hydrographsmay be developed for a number of purposes. Basin modelsare developed to provide hydrographs for historical eventsat required locations where gauged data are not available.Even in large basins, there will be limited gauged data andmany locations where data are desired. With some gaugeddata, a basin model can be developed and calibrated forobserved flood events. Chapter 13 of EM 1110-2-1417provides information on model development andcalibration.

d. Estimating runoff. Basin models can estimate therunoff response under changing conditions. Even withhistorical flow records, many reservoir studies will requireestimates of flood runoff under future, changed conditions.The future runoff with developments in the catchment andmodifications in the channel system can be modeled with abasin runoff model.

e. Application. For reservoir studies, the mostfrequent application of flood hydrograph analysis is todevelop hypothetical (or synthetic) floods. The threecommon applications are frequency-storms, SPF and PMF.Frequency-based design floods are used to develop flood-frequency information, like that required to computeexpected annual flood damage. SPF and PMF are used asdesign standards to evaluate project performance under themore rare flood events.

7-3. Hypothetical Floods

a. General. Hypothetical floods are usually used inthe planning and design of reservoir projects as a primarybasis of design for some project features and to substanti-ate the estimates of extreme flood-peak frequency. Whererunoff data are not available for computing frequencycurves of peak discharge, hypothetical floods can be usedto establish flood magnitudes for a specified frequency

EM 1110-2-142031 Oct 97

7-2

from rainstorm events of that frequency. This approach is values for a flood of this magnitude. Part 2 ofnot accurate where variations in soil-moisture conditions EM 1110-2-1417 provides detailed information on the unitand rainfall distribution characteristics greatly influence hydrograph procedure and the simulation of hypotheticalflood magnitudes. In general, measured data should be floods is described in Chapter 13. The computer programused to the maximum extent possible, and when approxi- HEC-1 Flood Hydrograph Package provides the SPS andmate methods are used, several approaches should be taken SPF computation procedures, as described in the SPFto compute flood magnitudes. determination manual.

b. Frequency-based design floods. In areas where (3) While the frequency of the standard project floodinfiltr ation losses are small, it may be feasible to compute cannot be specified, it can be used as a guide in extrapo-hypothetical floods from rainfall amounts of a specified lating frequency curves because it is considered to liefrequency and to assign that frequency to the flood event. within a reasonable range of rare recurrence intervals, suchNOAA publishes generalized rainfall criteria for the United as between once in 200 years and once in 1,000 years.States. They contain maps with isopluvial lines of pointprecipitation for various frequencies and durations. These d. Probable maximum flood. The PMF is the floodpoint values are then adjusted for application to areas that may be expected from the most severe combination ofgreater that 10 square miles, based on precipitation critical meteorologic and hydrologic conditions that areduration and catchment area. Section 13-4 of EM 1110-2- reasonably possible in the region. The PMF is calculated1417 provides information on simulation with frequency- from the Probable Maximum Precipitation (PMP). Thebased design storms. PMP values encompass the maximized intensity-duration

c. Standard project flood. The SPF is the flood that and variations of precipitation are considered with respectcan be expected from the most severe combination of to location, areal coverage of a watershed, and stormmeteorologic and hydrologic conditions that are considered duration. The probable maximum storm amounts arereasonably characteristic of the region in which the study determined in much the same way as are SPS amounts,basin is located. The SPF, which provides a performance except that precipitation amounts are first increased tostandard for potential major floods, is based on the Stan- correspond to maximum meteorologic factors such as winddard Project Storm (SPS). speed and maximum moisture content of the atmosphere.

(1) The SPS is usually an envelope of all or almost all (1) Estimates of PMP are based generally on theof the storms that have occurred in a given region. The results of the analyses of observed storms. More than 600size of this storm is derived by drawing isohyetal maps of storms throughout the United States have been analyzed inthe largest historical storms and developing a depth-area a uniform manner, and summary sheets have beencurve for the area of maximum precipitation for each distributed to government agencies and the engineeringstorm. Depth-area curves for storm rainfall of specified profession. These summary sheets include depth-area-durations are derived from this storm-total curve by a study duration data for each storm analyzed along with broadof the average time distribution of precipitation at stations outlines of storm magnitudes and their seasonal and geo-representing various area sizes at the storm center. When graphical variations. NWS (1977) Hydrometeorlogicalsuch depth-area curves are obtained for all large storms in Report No. 51 (HMR 51) contains generalized all-seasonthe region, the maximum values for each area size and estimates for the United States, east of the 105 longitude.duration are used to form a single set of depth-area- The PMP is distributed in space and time to develop theduration curves representing standard project storm PMS, which is a hypothetical storm that produces the PMFhyetographs for selected area sizes, using a typical time for a particular drainage basin.distribution observed in major storms. EM 1110-2-1411provides generalized SPS estimates for small and large (2) NWS (1981) HMR No. 52 provides criteria anddrainage basins, and projects for which SPF estimates are instructions for configuring the storm to produce the PMF.required. The generalized rainfall criteria and recom- The precipitation on a basin is affected by the stormmended procedures for SPS computations for U.S. drainage placement, storm-area size, and storm orientation. Thebasins located east of the 105 longitude are presented. HMR52 PMS (HEC 1984) computer program uses ath

(2) The SPF is ordinarily computed using the unit However, several trials are suggested to ensure that thehydrograph approach with the SPS precipitation. The unit maximum storm is produced. The PMS is then input to ahydrograph and basin losses should be based on reasonable rainfall-runoff model to determine the flood runoff.

values obtained from storms of a single type. Storm type

th

procedure to produce maximum precipitation on the basin.

EM 1110-2-142031 Oct 97

7-3

(3) The HMR52 User's Manual shows an example f. Snowmelt contribution. Satisfactory criteria andapplication with the HEC-1 Flood Hydrograph Package. procedures have not yet been developed for the computa-The storm hyetographs can be written to an output file, in tion of standard project and probable maximum snowmeltHEC-1 input format, or to an HEC-DSS file. HEC-1 can floods. The problem is complicated in that deep snowpackread the DSS file to obtain the basin precipitation. tends to inhibit rapid rates of runoff, and consequently,

(4) Hydrometeorlogical criteria are being updated for necessarily correspond to maximum snowpack depth orvarious areas of the country. A check should be made for water equivalent. Snowpack and snowmelt differ at variousthe most recent criteria. Figure 13-3 in EM 1110-2-1417 elevations, thus adding to the complexity of the problem.shows the regional reports available in 1993. The HMR52computer program does not apply to U.S. regions west of (1) Where critical durations for project design arethe 105 meridian. short, high temperatures occurring with moderate snow-th

(5) In the determination of both the SPF and the PMF, produce the most critical runoff. Where critical durationsselection of rainfall loss rates and the starting storage of are long, as is the more usual case in the control of snow-upstream reservoirs should be based on appropriate melt floods, prolonged periods of high temperature orassumptions for antecedent precipitation and runoff for the warm rainfall occurring with heavy snowpack amounts willseason of the storm. Also, PMF studies should consider produce critical conditions.the capability of upstream reservoir projects to safelyhandle the PMF contribution from that portion of the (2) The general procedure for the computation ofwatershed. There could be deficiencies in an upstream hypothetical snowmelt floods is to specify an initial snow-project spillway that significantly affects the downstream pack for the season that would be critical. In the case ofproject's performance. SPF's a maximum observed snowpack should be assumed.

e. Storm duration. Hypothetical storms to be used that which produces the most critical runoff conditions andfor any particular category of hypothetical flood computa- should be selected from an observed historical sequence.tion must be based on data observed within a region. For In the case of PMF computation, the most criticalapplication in the design of local flood protection projects, snowpack possible should be used and it should beonly peak flows and runoff volumes for short durations are considerably larger or more critical than the standardusually important. Accordingly, the maximum pertinent project snowpack. The temperature pattern should beduration of storm rainfall is only on the order of the time of selected from historical temperature sequences augmentedtravel for flows from the headwaters to the location to represent probable maximum temperature for the season.concerned. After a reasonable maximum duration of Where simultaneous contribution from rainfall is possible,interest is established, rainfall amounts for this duration a maximum rainfall for the season should be added duringand for all important shorter durations must be established. the time of maximum snowmelt. This would require someFor standard project storm determinations, this would moderation of temperatures to ensure that they areconsist of the amounts of observed rainfall in the most consistent with precipitation conditions. EM 1110-2-1406severe storms within the region that correspond to area covers snowmelt for design floods, standard project andsizes equal to the drainage area above the project. In the maximum probable snowmelt flood derivation.case of hypothetical storms and floods of a specifiedfrequency, these rainfall amounts would correspond to (3) Snowmelt computations can be made in accor-amounts observed to occur with the specified frequency at dance with an energy budget computation, accounting forstations spread over an area the size of the project drainage radiation, evaporation, conductivity, and other factors, orarea. Larger rates and smaller amounts of precipitation by a simple relation with air temperature, which reflectswould occur for shorter durations, as compared with the most of these other influences. The latter procedure islonger durations of interest. Once a depth-duration curve is usually more satisfactory in practical situations. Snowmelt,established that represents the desired hypothetical storm loss rate, and unit hydrograph computations can be maderainfall, a time pattern must be selected that is reasonably by using a computer program like Flood Hydrographrepresentative of observed storm sequences. The HEC-1 Analysis, HEC-1. EM 1110-2-1417 has detailed descrip-computer program has the capability of accepting any tions of each computational component.depth-duration relation and selecting a reasonable timesequence. It is also capable of accepting specified timesequences for hypothetical storms.

probable maximum snowmelt flood potential does not

pack depths after some melting has occurred will probably

The temperature sequence for SPF computation would be

EM 1110-2-142031 Oct 97

8-1

Chapter 8Water Surface Profiles

8-1. Introduction

a. General. Water surface profiles are required formost reservoir projects, both upstream and downstreamfrom the project. Profile computations upstream from theproject define the “backwater” effect due to high reservoirpool levels. The determination of real estate requirementsare based on these backwater profiles. Water surfaceprofiles are required downstream to determine channelcapacity, flow depths and velocities, and other hydraulicinformation for evaluation of pre- and post-projectconditions.

b. Choosing a method. The choice of an appropriatemethod for computing profiles depends upon the followingcharacteristics: the river reach, the type of flow hydro-graph, and the study objectives. The gradually varied,steady flow profile computation (e.g., HEC-2), is used formany studies. However, the selection of the appropriatemethod is part of the engineering analysis. EM 1110-2-1416 provides information on formulating a hydraulicstudy and a discussion of the analytical methods in generaluse. The following sections provide general guidance onthe methods and the potential application in reservoirrelated studies.

8-2. Steady Flow Analysis

a. Method assumptions. A primary consideration inone-dimensional, gradually varied, steady flow analysis isthat flow is assumed to be constant, in time, for the profilecomputation. Additionally, all the one-dimensionalmethods require the modeler to define the flow path whendefining the cross-sectional data perpendicular to the flow.The basic assumptions of the method are as follows:

(1) Steady flow - depth and velocity at a given locationdo not vary with time.

(2) Gradually varied flow - depth and velocity changegradually along the length of the water course.

(3) One-dimensional flow - variation of flow charac-teristics, other than in the direction of the main axis of flowmay be neglected, and a single elevation represents thewater surface of a cross section perpendicular to the flow.

(4) Channel slope less than 0.1 m/m - because thehydrostatic pressure distribution is computed from thedepth of water measure vertically.

(5) Averaged friction slope - the friction loss betweencross sections can be estimated by the product of therepresentative slope and reach length.

(6) Rigid boundary - the flow cross section does notchange shape during the flood.

b. Gradually varied steady flow. The assumption ofgradually varied steady flow for general rainfall andsnowmelt floods is generally acceptable. Dischargechanges slowly with time and the use of the peak dischargefor the steady flow computations can provide a reasonableestimate for the flood profile. Backwater profiles,upstream from a reservoir, are routinely modeled usingsteady flow profile calculations. However, inflow hydro-graphs from short duration, high intensity storms, e.g.,thunderstorms, may not be adequately modeled assumingsteady flow.

c. Downstream profile. Obviously, the downstreamprofile for a constant reservoir release meets the steadyflow condition. Again, the consideration is how rapidlyflow changes with time. Hydropower releases for a peak-ing operation may not be reasonably modeled using steadyflow because releases can change from near zero to turbinecapacity, and back, in a short time (e.g., minutes) relative tothe travel time of the resulting disturbances. Dam-breakflood routing is another example of rapidly changing flowwhich is better modeled with an unsteady flow method.

d. Flat stream profiles. Another consideration iscalculating profiles for very flat streams. When the streamslope is less than 0.0004 m/m (2 ft/mile), there can be asignificant loop in the downstream stage-dischargerelationship. Also, the backwater effects from downstreamtributaries, or storage, or flow dynamics may stronglyattenuate flow. For slopes greater than 0.0009 m/m(5 ft/mile), steady flow analysis is usually adequate.

e. Further information. Chapter 6 of EM 1110-2-1416 River Hydraulics provides a detailed review of theassumptions of the steady flow method, data requirements,and model calibration and application. Appendix D pro-vides information on the definition of river geometry andenergy loss coefficients, which is applicable to all the one-dimensional methods.

EM 1110-2-142031 Oct 97

8-2

8-3. Unsteady Flow Analysis

a. Unsteady flow methods. One-dimensionalunsteady flow methods require the same assumptions listedin 8-2(a), herein, except flow, depth, and velocity can varywith time. Therefore, the primary reason for usingunsteady flow methods is to consider the time varyingnature of the problem. Examples of previously mentionedrapidly changing flow are thunderstorm floods,hydroelectric peaking operations, and dam-break floods.The second application of unsteady flow analysis consid-eration, mentioned above, is streams with very flat slopes.

b. Predicting downstream stages. Another applica-tion of unsteady flow is in the prediction of downstream a. Reservoirs. Reservoirs disrupt the flow of sedi-stages in river-reservoir systems with tributaries, or lock- ment when they store or slow down water. At the upperand-dam operations where the downstream operations limit of the reservoir, the velocity of inflowing wateraffect the upstream stage. Flow may not be changing rap- decreases and the ability to transport sediment decreasesidly with time, but the downstream changes cause a time and deposition occurs. Chapter 9 herein presents reservoirvarying downstream boundary condition that can affect the sediment analysis. Reservoir releases may be sedimentupstream stage. Steady flow assumes a unique stage- deficient, which can lead to channel degradationdischarge boundary condition that is stable in time. downstream from the project because the sediment is

c. Further information. Chapter 5, “Unsteady Flow,”in EM 1110-2-1416 provides a detailed review of model b. River and reservoir sedimentation. EM 1110-2-application including selection of method, data require- 4000 is the primary Corps reference on reservoir sedimen-ments, boundary conditions, calibration, and application. tation. Chapter 3 covers sediment yield and includes

8-4. Multidimensional Analysis

a. Two- and three-dimensional modeling. Multi-dimensional analysis includes both two- andthree-dimensional modeling. In river applications, two-dimensional modeling is usually depth-averaged. That is,variables like velocity do not vary with depth, so an averagevalue is computed. For deep reservoirs, the variation ofparameters with depth is often important (see Chapter 12,EM 1110-2-1201). Two-dimensional models, for deepreservoirs, are usually laterally-averaged. Three-dimen-sional models are available; however, their applicationshave mostly been in estuaries where both the lateral andvertical variation are important.

b. Two-dimensional analysis. Two-dimensional,depth-averaged analysis is usually performed in limited

portions of a study area at the design stage of a project.The typical river-reservoir application requires both thedirection and magnitude of velocities. Potential modelapplications include areas upstream and downstream fromreservoir outlets. Additionally, flow around islands, andother obstructions, may require two-dimensional modelingfor more detailed design data.

c. Further information. Chapter 4 of EM 1110-2-1416 provides a review of model assumptions and typicalapplications.

8-5. Movable-Boundary Profile Analysis

removed from the channel.

methods based on measurement and mathematical models.Chapter 4 covers river sedimentation, and Chapter 5 pre-sents reservoir sedimentation. Section III, of Chapter 5,provides an overview of points of caution, sedimentationproblems associated with reservoirs, and the impact ofreservoirs on the stream system. Section IV providesinformation on levels of studies and study methods.

b. Further information. Chapter 7 of EM 1110-2-1416 presents water surface profile computation withmovable boundaries. The theory, data requirements andsources, plus model development and application are allcovered. The primary math models, HEC-6 Scour andDeposition in Rivers and Reservoirs (HEC 1993) andOpen-Channel Flow and Sedimentation TABS-2 (Thomasand McAnally 1985) two-dimensional modeling package,are also described. The focus for the material is riverine.

EM 1110-2-142031 Oct 97

9-1

Chapter 9Reservoir Sediment Analysis

9-1. Introduction

a. Parameters of a natural river. Nature maintains avery delicate balance between the water flowing in anatural river, the sediment load moving with the water, andthe stream's boundary. Any activity which changes any oneof the following parameters:

� water yield from the watershed.

� sediment yield from the watershed.

� water discharge duration curve.

� depth, velocity, slope or width of the flow.

� size of sediment particles.

or which tends to fix the location of a river channel on itsfloodplain and thus constrains the natural tendency willupset the natural trend and initiate the formation of a newone. The objective of most sediment studies is to evaluatethe impact on the flow system resulting from changing anyof these parameters.

b. Changes caused by reservoirs. Reservoirs inter-rupt the flow of water and, therefore, sediment. In terms ofthe above parameters, the reservoir causes a change in theupstream hydraulics of flow depth, velocity, etc. by forcingthe energy gradient to approach zero. This results in a lossof transport capacity with the resulting sediment depositionin the reservoir. The reservoir also alters the downstreamwater discharge-duration relation and reduces the sedimentsupply which may lead to the degradation of thedownstream channel.

c. Areas of analysis. Sedimentation investigationsusually involve the evaluation of the existing condition aswell as the modified condition. The primary areas ofreservoir sediment analysis are the estimation of volumeand location of sediment deposits in the reservoir and theevaluation of reservoir releases' impact on the downstreamchannel system. Sediment deposits start in the backwaterarea of the reservoir, which increase the elevation of thebed profile and the resulting water surface profile.However, reservoirs may also cause sediment depositsupstream from the project, which affect the upstream watersurface profiles.

d. Further information. The primary Corps refer-ence for sediment analysis is EM 1110-2-4000. Majortopics include developing a study work plan, sedimentyield, river sedimentation, reservoir sedimentation, andmodel studies.

9-2. Sediment Yield Studies

a. General. Sediment yield studies determine theamount of sediment that leaves a basin for an event or overa period of time. Sediment yield, therefore, involveserosion processes as well as sediment deposition anddelivery to the study area. The yield provides the necessaryinput to determine sedimentation impacts on a reservoir.

b. Required analysis. Each reservoir projectrequires a sediment yield analysis to determine the storagedepletion resulting from the deposition of sediment duringthe life of the project. For most storage projects, asopposed to sediment detention structures, the majority ofthe delivered sediment is suspended. However, the datarequired for the headwater reaches of the reservoir shouldinclude total sediment yield by particle size because that iswhere the sands and gravels will deposit.

c. Further information. Corps of Engineer methodsfor predicting sediment yields are presented in Appendix Cof EM 1110-2-4000. A literature review, conducted by theHydrologic Engineering Center under the Land SurfaceErosion research work unit, showed numerous mathe-matical models are available to estimate sediment dischargerates from a watershed and the redistribution of soil withina watershed. An ETL on the methods will be issued soon.

9-3. Reservoir Sedimentation Problems

a. Sediment deposition. As mentioned above, theprimary reservoir sediment problem is the deposition ofsediment in the reservoir. The determination of the sedi-ment accumulation over the life of the project is the basisfor the sediment reserve. Typical storage diagrams ofreservoirs, showing sediment (or dead) storage at thebottom of the pool can be misleading. While the reservoirstorage capacity may ultimately fill with sediment, thedistribution of the deposits can be a significant concernduring the life of the project. The reservoir sedimentationstudy should forecast sediment accumulation anddistribution over the life of the project. Sediment depositsin the backwater area of the reservoir may form deltas,particularly in shallow reservoirs. A number of problemsassociated with delta formations are discussed below.

(1) Deposits forming the delta may raise the watersurface elevation during flood flows, thus requiring special

EM 1110-2-142031 Oct 97

9-2

consideration for land acquisition. In deep reservoirs, thisis usually not a problem with the reservoir area becauseproject purposes dictate land acquisitions or easements.Deltas tend to develop in the upstream direction. Inshallow reservoirs, the increase in water surface elevationis a problem even within the reservoir area. That is, floodsof equal frequency may have higher water surfaceelevations after a project begins to develop a delta depositthan was experienced before the project was constructed.Land acquisition studies must consider such a possibility.

(2) Aggradation problems are often more severe ontributaries than on the main stem. Analysis is complicatedby the amount of hydrologic data available on thetributaries, which is usually less than on the main stemitself. Land use along the tributary often includes recre-ation sites, where aggradation problems are particularlyundesirable.

(3) Reservoir deltas often attract phreatophytes due tothe high moisture level. This may cause water-use prob-lems due to their high transpiration rate.

(4) Reservoir delta deposits are often aestheticallyundesirable.

(5) Reservoir sediment deposits may increase the watersurface elevation sufficiently to impact on the groundwatertable, particularly in shallow impoundments.

(6) In many existing reservoirs, the delta and back-water-swamp areas support wildlife. Because the charac-teristics of the area are closely controlled by the operationpolicy of the reservoir, any reallocation of storage wouldneed to consider the impact on the present delta and swampareas.

b. Upstream projects. It is important to identify andlocate all existing reservoirs in a basin where a sedimentstudy is to be made. The projects upstream from the pointof analysis potentially modify both the sediment yield andthe water discharge duration curve. The date ofimpoundment is important so that observed inflowingsediment loads may be coordinated with whatever condi-tions existed in the basin during the periods selected forcalibration and verification. Also, useful information onthe density of sediment deposits and the gradation ofsediment deposits along with sediment yield are oftenavailable from other reservoirs in the basin. Information onthe rate of sediment deposition that has occurred at otherreservoir sites in the region is the most valuableinformation when estimating sediment deposition for a newreservoir.

9-4. Downstream Sediment Problems

a. Channel degradation. Channel degradationusually occurs downstream from the dam. Initially, afterreservoir construction, the hydraulics of flow (velocity,slope, depth, and width) remain unchanged from pre-project conditions. However, the reservoir acts as a sinkand traps sediment, especially the bed material load. Thisreduction in sediment delivery to the downstream channelcauses the energy in the flow to be out of balance with theboundary material for the downstream channel. Because ofthe available energy, the water attempts to re-establish theformer balance with sediment load from material in thestream bed, and this results in a degradation trend.Initially, degradation may persist for only a short distancedownstream from the dam because the equilibriumsediment load is soon re-established by removing materialfrom the stream bed.

b. Downstream migratory degradation. As timepasses, degradation tends to migrate downstream. How-ever, several factors are working together to establish anew equilibrium condition in this movable-boundary flowsystem. The potential energy gradient is decreasingbecause the degradation migrates in an upstream-to-downstream direction. As a result, the bed material isbecoming coarser and, consequently, more resistant tobeing moved. This tendency in the main channel has theopposite effect on tributaries. Their potential energygradient is increasing which results in an increase intransport capacity. This will usually increase sedimentpassing into the main stem which tends to stabilize themain channel resulting in less degradation than might beanticipated. Finally, a new balance will tend to be estab-lished between the flowing water-sediment mixture and theboundary.

c. Extent of degradation. The extent of degradationis complicated by the fact that the reservoir also changesthe discharge duration curve. This will impact for aconsiderable distance downstream from the project becausethe existing river channel reflects the historical phasingbetween flood flows on the main stem and those fromtributaries. That phasing will be changed by the operationof the reservoir. Also, the reduced flow will probablypromote vegetation growth at a lower elevation in thechannel. The result is a condition conducive to depositionin the vegetation. Detailed simulation studies should beperformed to determine future channel capacities and toidentify problem areas of excessive aggradation ordegradation. All major tributaries should be included.

EM 1110-2-142031 Oct 97

9-3

9-5. Sediment Water Quality

a. Sediments and pollutants. When a river carryingsediments and associated pollutants enters a reservoir, theflow velocity decreases and the suspended and bed loadsediments start settling down. Reservoirs generally act asdepositories for the sediments because of their high sedi-ment trap efficiency. Due to a high adsorption capacity,sediments act as sinks for contaminants in the reservoirsand, in agricultural and industrial areas, may contain PCB's,chlorinated hydrocarbon pesticides, oil and grease, heavymetals, coliform bacteria, or mutagenic substances. Burialof these contaminants by sedimentation may be animportant factor and an effective process in isolatingpotentially toxic substances from surface waters andimportant biological populations. Toxic inorganic andorganic contaminants associated with the sediments canalso be bioconcentrated by the aquatic organisms present inreservoirs.

b. Monitoring chemical contaminants. These incom-ing sediments and associated pollutants significantly affectthe water quality of the reservoir pool and downstreamreleases. Therefore, it is essential that these sedimentreservoir interactions be characterized by their depositionalbehavior, particle size distribution, and pollutantconcentrations to successfully plan a management strategyto quantify contaminant movement within reservoirs.Analytical and predictive methods to assess the influenceof contaminated sediments in reservoirs have not beendeveloped enough to be used in Corps field offices, butWES Instruction Report E-86-1, “General Guidelines forMonitoring Contaminants in Reservoirs” (Waide 1986),does provide general guidance on the design and conduct ofprograms for monitoring chemical contaminants inreservoir waters, sediments, and biota.

c. Sedimentation patterns. Sedimentation patternscan often be associated with water quality characteristics.There seems to be a relationship between longitudinalgradients in water quality (a characteristic of many reser-voirs) and sediment transport and deposition. High con-centrations of inorganic particulates can reduce light avail-ability near inflows and thus influence algal production anddecrease dissolved oxygen. The association of dissolvedsubstances, such as phosphorus, with suspended solids mayact to reduce or buffer dissolved concentrations, thusinfluencing nutrient availability.

9-6. Sediment Investigations

a. General. The level of detail required for theanalysis of any sediment problem depends on the objective

of the study. Chapter 1 of EM 1110-2-4000 describesstaged sedimentation studies in Section I. Section II, ofthat chapter, provides reporting requirements. Problemidentification and the development of a study work plan arecovered in Chapter 2.

b. Sediment deposits. Considering a dam site as animportant natural resource, it is essential to provide enoughvolume in the reservoir to contain anticipated depositsduring the project life. If the objective of a sediment studyis just to know the volume of deposits for use in screeningstudies, then trap efficiency techniques can provide asatisfactory solution. The important information that mustbe available is the water and sediment yields from thewatershed and the capacity of the reservoir. Chapter 3 ofEM 1110-2-4000 covers sediment yield. Section 3-7provides information on reservoir sedimentation, includingtrap efficiency.

c. Land acquisition. If the sediment study mustaddress land acquisition for the reservoir, then knowingonly the volume of deposits is not sufficient. The locationof deposits must also be known, and the study must takeinto account sediment movement. This generally requiressimulation of flow in a mobile boundary channel. Sortingof grain sizes must be considered because the coarsermaterial will deposit first, and armoring must be consideredbecause scour is involved. Movable-bed modeling is usefulto predict erosion or scour trends downstream from thedam, general aggradation or degradation trends in riverchannels, and the ability of a stream to transport the bed-material load. The computer program, HEC-6 Scour andDeposition in Rivers and Reservoirs (HEC 1993), isdesigned to provide long-term trends associated withchanges in the frequency and duration of the waterdischarge and/or stage or from modifying the channelgeometry.

d. Details of investigations. The details of reservoirsedimentation investigations are covered in Chapter 5 ofEM 1110-2-4000. The primary emphasis is on theevaluation of the modified condition, which includesconsideration of quality and environmental issues. Thelevels of sedimentation studies and methods of analysis arepresented in Section IV of Chapter 5. Model studies and ashort review of HEC-6 and the two-dimensional TABS-2modeling system are covered in Chapter 6.

EM 1110-2-142031 Oct 97

PART 3

RESERVOIR STORAGE REQUIREMENTS

EM 1110-2-142031 Oct 97

10-1

Chapter 10Flood-Control Storage

10-1. General Considerations

a. Reservoir flood storage. Where flood damage at anumber of locations on a river can be significantly reducedby construction of one or more reservoirs, or where areservoir site immediately upstream from one damagecenter provides more economical protection than localprotection works, reservoir flood storage should beconsidered. Whenever such reservoirs can serve needsother than flood control, the integrated design and opera-tion of the project for multipurpose use should beconsidered.

b. Flood-control features. In planning and designingthe flood-control features of a reservoir, it is important thatthe degree and extent of continuous ensured protection beno less than that provided by a local protection project, ifthe alternatives of reservoir construction or channel andlevee improvement are to be evaluated fairly. This meansthat the storage space and release schedule for floodcontrol must be provided at all times when the floodingpotential exists. In some regions this may be for the entireyear, but more commonly there are dry seasons when theflood potential is greatly reduced and storage reservationfor flood control can be reduced correspondingly. Exceptwhere spring snowmelt floods can be forecasted reliably orwhere safe release rates are sufficient to empty flood spacein a very short time, it is not ordinarily feasible to provideflood-control space only after a flood is forecasted. Spacemust be provided at all times during the flood season unlessit can be demonstrated that the necessary space can beevacuated on a realistic forecast basis. Also, space may bereduced if less storage is needed due to low snowpack, orthere is some other reliable basis for long range floodforecasting.

c. Runoff volume durations. Whereas the peak ratesof runoff are critical in the design of local protectionprojects, runoff volumes for pertinent durations are criticalin the design of reservoirs for flood control. The criticaldurations will be a function of the degree of flood protec-tion selected and of the release rate or maximum rate offlow at the key downstream control point. As the proposeddegree of protection is increased and as the proposed ratesof controlled flows at key damage centers are reduced, thecritical duration is increased. If this critical durationcorresponds to the duration of a single rainstorm period ora single snowmelt event, the computation of hypotheticalfloods from rainfall and snowmelt can constitute the

principle hydrologic design element. On the other hand, ifthe critical duration is much longer, hypothetical floods andsequences of hypothetical floods computed from rainfall orsnowmelt become less dependable as guides to design. Itthen is necessary to base the design primarily on thefrequency of observed runoff volumes for long durations.Even when this is done, it will be advisable to construct atypical hydrograph that corresponds to runoff volumes forthe critical duration and that reasonably characterizeshydrographs at the location, in order to examine theoperation of the proposed project under realisticconditions.

d. Hypothetical flood simulations. When hypotheti-cal floods are selected, they must be routed through theproposed reservoir under the operation rules that would bespecified for that particular design. In effect, a simulationstudy of the proposed project and operation scheme wouldbe conducted for each flood. It is also wise to simulate theoperation for major floods of historical record in order toensure that some peculiar feature of a particular flood doesnot upset the plan of operation. With present software, it isrelatively inexpensive to perform a complete period ofrecord simulation once the flood-control storage is set.

10-2. Regulated Release Rates

a. Flood reduction purposes. For flood reductionpurposes reservoirs must store only the water that cannotbe released without causing major damage downstream. Ifmore water can be released during a flood, less water needsto be stored. Thus, less storage space needs to be plannedfor flood control. Because reservoir space is costly andusually in high demand for other purposes, good flood-control practice consists of releasing water whenevernecessary at the highest practical rates so that a minimumamount of space need be reserved for flood control. Asthese rates increase, it becomes costly also to improvedownstream channels and to provide adequate reservoiroutlets, so there is an economic balance between releaserates and storage capacity for flood control. In general, it iseconomical to utilize the full nondamage capacity ofdownstream channels, and it may pay to provide someadditional channel or levee improvements downstream.However, as described in paragraph f, full channel capacitymay not be available, so analyses should consider theimpact of reduced capacity.

b. Channel capacities. Channel capacities should beevaluated by examing water-surface profile data fromactual flood events whenever possible. Under naturalchannel conditions, it will ordinarily be found that floodswhich occur more frequently than once in two years are notseriously damaging, while larger floods are.

EM 1110-2-142031 Oct 97

10-2

c. Minor versus major damage releases. In some efficiency. It may be necessary to delay flood releases tocases, it is most economical to sustain minor damage by permit removal of equipment, cattle, etc., from areas thatreleasing flows above nondamaging stages in order to would be flooded. Releases might be curtailed temporarilyaccommodate major floods and thereby protect the more in order to permit emergency repairs to canals, bridges, andimportant potential damage areas from flooding. In such other structures downstream. If levees fail, releases mightsituations, a stepped-release schedule designed to protect be reduced in order to hasten the drainage of flooded areas.all areas against frequent minor floods, with provision to Release can be reduced in order to facilitate rescueincrease releases after a specified reservoir stage is operations. These and various other conditions result inreached, might be considered. However, such a plan has reduced operation efficiency during floods. To account forserious drawbacks in practice because protection of the this, less nondamage flow capacity than actually existsminor damage areas would result in greater improvements (often about 80 percent) is assumed for design studies. It isin those areas; and it soon becomes highly objectionable, if important, however, that every effort be made in actualnot almost impossible, to make the larger releases when operation to effect the full non-damage releases in order tothey are required for protection of major damage areas. In attain maximum flood-control benefits.any case, it is necessary to make sure that the minordamage areas are not flooded more frequently or severely g. Gradually increasing and decreasing releases.with the project than they would have been without it. During flood operations, reservoir releases must be

d. Maintenance and zoning. It is important on all damage and undue hardship downstream. Graduallystreams in developed areas to provide for proper mainte- increasing releases will usually permit an orderly evacua-nance of channel capacity and zoning of the floodplain tion of people, livestock, and equipment from the riverwhere appropriate. This is vital where upstream reservoirs areas downstream. If releases are curtailed too rapidly,are operated for flood control because proper reservoir there is some danger that the saturated riverbanks willregulation depends as much on the ability to release slough and result in the loss of valuable land or damage towithout damage as it does on the ability to store. Minor levees.inadequacies in channel capacity can lead to the loss ofcontrol and result in major flooding. This situation isaggravated because the reduced frequency of floodingbelow reservoirs and the ability to reduce reservoir releaseswhen necessary often increase the incentive to develop thefloodplain and sometimes even remove the incentive formaintaining channel capacity.

e. Forecasted runoff. When a reservoir is locatedsome distance upstream from a damage center, allowancemust be made for any runoff that will occur in the inter-mediate area. This runoff must be forecasted, a possibleforecast error added, and the resulting quantities subtractedfrom project channel capacity to determine per-missiblerelease rates considering attenuation when routing therelease from the dam to the damage center and the con-tribution of flow from the intermediate drainage area.Also, with high intensity rainfall, the added rainfall depth tothe total downstream channel flow should be considered.

f. Delaying flood releases. Experience in the flood-control operation of reservoirs has demonstrated that theactual operation does not make 100 percent use ofdownstream channel capacities. Due to many contributingfactors average outflows during floods are less than maxi-mum permissible values. It is usually wise to approachmaximum release rates with caution, in order to ascertainany changes in channel capacity that have taken place sincethe last flood, and this practice reduces operational

increased and decreased gradually in order to prevent

10-3. Flood Volume Frequencies

a. Critical durations. Flood volume frequencystudies usually consist of deriving frequency curves ofannual maximum volumes for each of various specifieddurations that might be critical in project design. Criticaldurations range from a few hours in the case of regulating“cloudburst” floods to a few months where large storageand very low release rates prevail. The annual maximumvolumes for a specific duration are usually expressed asaverage rates of flow for that duration. It is essential thatthese flows represent a uniform condition of developmentfor the entire period of observation, preferably unregulatedconditions. Procedures for computing the individualfrequency curves are discussed briefly in Chapter 6 hereinand are described in detail in EM 1110-2-1415.

b. Flood-control space requirement. Determinationof the flood-control space needed to provide a selecteddegree of protection is based on detailed hydrographanalysis, but a general evaluation can be made as illustratedin Figure 10-1. The curve of runoff versus duration isobtained from frequency studies of runoff volumes or fromSPF studies at the location. The tangent line represents auniform flow equal to the project release capacity (reducedby an appropriate contingency factor). The interceptrepresents the space required for control of the flood. Thechart demonstrates that a reservoir capable of storing

EM 1110-2-142031 Oct 97

10-3

Figure 10-1. Flood-control space requirement

155,000 units of water and releasing 30,000 units per day in the event of project failure. The SPF, which representscan control 100-year runoff for any duration, and that the the largest flood for that location that is reasonablycritical duration (period of increasing storage) is about characteristic for the region, is a flood of considerably5 days. The volume-duration curve would be made for lesser magnitude and represents a high degree of designeach damage area and should include more than 100 per- for projects protecting major urban and industrial areas.cent of the local uncontrolled runoff downstream from the These floods can result from heavy rainfall or fromreservoir and above the control point in order to allow for snowmelt in combination with some rainfall.errors of forecast which would be reflected in reducedproject releases. If this local runoff appreciably exceeds e. Computing hydrographs. SPF and PMF hydro-nondamage flow capacity at the damage centers, the vol- graphs are computed from the storm hyetographs by unitume over and above the flow capacity is damaging water hydrograph procedures. In the case of the SPF, groundthat cannot be stored in the project reservoir. conditions that are reasonably conducive to heavy runoff

10-4. Hypothetical Floods

a. Two classes. Two classes of hypothetical floods is provided in Chapter 7 of this manual. Detailed methodsare important in the design of reservoirs for flood control. for performing these computations are described in EMOne is a balanced flood that corresponds to a specified 1110-2-1417. The computer program HEC-1 Floodfrequency of occurrence; the other is a flood that repre- Hydrograph Package contains routines for computingsents a maximum potential for the location, such as the floods from rainfall and snowmelt and also containsSPF or PMF. ER 1110-8-2(FR) sets forth hydrologic standard project criteria for the eastern United States.engineering requirements for selecting and accom-modating inflow design floods for dams and reservoirs.

b. Specified frequencies. A hypothetical flood a. General. As stated earlier, whenever floodcorresponding to a specified frequency should contain releases are required, it is imperative that they be made atrunoff volumes for all pertinent durations corresponding to maximum rates consistent with the conditions down-that specified frequency. The derivation of frequency stream. This means that the outlets should be designed tocurves is as discussed in the preceding section. A bal- permit releases at maximum rates at all reservoir levelsanced flood hydrograph is constructed by selecting a within the flood-control space. In some cases wheretypical hydrograph pattern and adjusting the ordinates so controlled releases are very high, such an outlet design isthat the maximum volumes for each selected duration not economical, and releases at lower stages might becorrespond to the volumes for that duration at the specified restricted because of limited outlet capacity. This con-frequency. straint, of course, should be taken into account during the

c. Longer duration floods. Where flood durationslonger than the typical single-flood duration are importantin the design, a sequence of flood hydrographs spacedreasonably in time should be used as a pattern flood. Inorder to represent average natural sequences of floodevents, the largest portions of the pattern flood shouldordinarily occur at or somewhat later than the midpoint ofthe entire pattern, because rainfall sequences are fairlyrandom but ground conditions become increasingly wetand conducive to larger runoff as any flood sequencecontinues.

d. Maximum flood potential. Two types of hypo-thetical floods that represent maximum flood potential areimportant in the design of reservoirs. The PMF, which isthe largest flood that is reasonably possible at the location,is ordinarily the design flood for the spillway of a structurewhere loss of life or major property damage would occur

are used. In the case of the PMF, the most severe groundconditions that are reasonably consistent with storm mag-nitudes are used. A general description of these analyses

10-5. Operation Constraints and Criteria

design studies.

EM 1110-2-142031 Oct 97

10-4

b. Downstream damage centers. Where damage c. Detailed operational study. Although there arecenters are at some distance downstream from the reser- approximate methods for estimating storage capacity, it isvoir, local runoff below the reservoir and above the damage essential that the final project design be tested by a detailedcenter must be considered when determining releases to be operational study. The analyses are based on actual outletmade. This will ordinarily require some forecasting of the capacities and realistic assumptions for limiting rates oflocal runoff and, consequently, some estimate of the release change, forecast errors, and operationalforecast uncertainty. The permissible release at any time is contingencies, and include various combinations ofdetermined by adding a safe error allowance to the reservoir inflow and local flow that can produce a specificforecasted local inflow and subtracting this sum from the downstream flood event. It is also important to route thenondamaging flow capacity. largest floods of record and synthetic floods through the

c. Rate-of-change of release. The rate-of-change of that the project provides the degree of protection for whichrelease must be restricted to the maximum changes that it was designed.will not cause critical conditions downstream. As a prac-tical matter, these rates-of-change of release should be less d. Seasonal distribution of storage requirements.than the rates-of-change of flow that occurred before the Where some of the flood-control space will be madereservoir was built. After the main flood has passed, water available for other uses during the dry season, a seasonalstored in the flood-control space must be released and distribution of flood-control storage requirement should bemaximum rates of release will continue until the desired developed. The most direct approach to this entails theamount of water is released, except that the rate of release construction of runoff frequency curves for each month ofshould be decreased gradually toward the end of the release the year. The average frequency of the design flood duringperiod. This reduction in release must be started while the rainy-season months can be used to select floodconsiderable flood waters remain in the reservoir in order magnitudes for other months. These could then serve as athat water retained for other purposes is not inadvertently basis for determining the amount of space that must bereleased. Schedules for this operation are discussed in made available during the other months.Part 3.

10-6. Storage Capacity Determinations

a. Determining required storage capacity. Thestorage capacity required to regulate a specific flood (rep-resented by a flood hydrograph at the dam) to a specifiedcontrol discharge immediately downstream of the dam is10-7. Spillwaysdetermined simply by routing the hydrograph through ahypothetical reservoir with unlimited storage capacity and Spillways are provided to release floodwater which nor-noting the maximum storage. However, there are many mally cannot be passed by other outlet works. The spillwayspecial practical considerations that complicate this is sized to ensure the passage of major floods withoutprocess. Release rates should not be changed suddenly; overtopping the dam. A general discussion of spillways istherefore, the routing should conform to criteria that provided in Section 4-2 of EM 1110-2-3600. EM 1110-2-specify the maximum rate of change of release. Also, 1603 describes the technical aspects of design for theoutlet capacities might not be adequate to supply full hydraulic features of spillways and ER 1110-8-2(FR) setsregulated releases with low reservoir stages. If this is the forth requirements for selecting and accommodating inflowcase, a preliminary reservoir design is required in order to design floods.define the relation of storage capacity to outlet capacity.

b. Specified flood. In the more common cases, where is usually selected as a large hypothetical flood deriveddamage centers exist at some distance downstream of the from rainfall and snowmelt. Other methods of estimatingreservoir, the storage requirement for a specified flood is extreme flood magnitudes, such as flood-frequency analy-determined by successive approximations, operating the sis, are not reliable due to limited observations. Thehypothetical reservoir to regulate flows at each damage selection of a spillway design flood depends on the policiescenter to nondamaging capacity, and allowing for local of the construction agency and regulations governing daminflow and for some forecasting error. construction. Usually, the spillways for major dams, whose

project to determine that the project design is adequate and

e. Further information. Sequential routing in plan-ning, design, and operation of flood-control reservoirs canbe accomplished with the computer program HEC-5 Simu-lation of Flood Control and Conservation Systems(HEC 1982c).

a. Spillway design flood. The spillway design flood

EM 1110-2-142031 Oct 97

10-5

failure might constitute a major disaster, are designed to e. Spillway types. While the spillway is primarilypass the PMF without a major failure; however, the intended to protect the structure from failure, the fact that itspillways for many small dams are designed for smaller can cause some water to be stored above ordinary full poolfloods such as the SPF. level (surcharge storage) is of some consequence in

b. Hydrologic design. The hydrologic design of a require higher dams and can, therefore, be highly effectivespillway is accomplished by first estimating a design and in partially regulating floods that exceed project designthen testing it by routing the spillway design flood. In magnitude, whereas wide spillways and gated spillways arerouting the spillway design flood, the initial reservoir stage less effective for regulating floods exceeding designshould be as high as reasonably expected at the start of magnitude. Where rare floods can cause great damagesuch a major flood, considering the manner in which the downstream, the selection of spillway type andreservoir is planned to operate or how in the future the characteristics can appreciably influence the benefits thatreservoir might operate differently from the planned are obtained for flood control. Accordingly, it is notoperation. In the case of ungated spillways, it is possible necessarily the least costly spillway that yields the mostthat the outlets of the dam will be closed gradually as the economical plan of development. In evaluating flood-spillway goes into operation, in order to delay damaging control benefits, computing frequency curves for regulatedreleases as long as possible and possibly to prevent them. conditions should be based on spillway characteristics andHowever, if spillway flows continue to increase, it may be operation criteria as well as on other project features.necessary to reopen the outlets. In doing so, care should beexercised to prevent releases from exceeding maximuminflow quantities. The exact manner in which outlets willbe operated should be specified so that the spillway designwill be adequate under conditions that will actually prevailafter project construction. Consideration should be givento the possibility that some outlets or turbines might be outof service during flood periods.

c. Large spillway gates. The operation of largespillway gates can be extremely hazardous, since openingthem inadvertently might cause major flooding at down-stream areas. Their operation should be controlled by rigidregulations. In particular, the opening of the gates duringfloods should be scheduled on the basis of inflows andreservoir storage so that the lake level will continue to riseas the gates are opened. This will ensure that inflowexceeds outflow as outflows are increased. The adequacyof a spillway to pass the spillway design flood is tested forgated spillways in the same manner as for ungatedspillways described above. Methods for developingspillway-gate operation regulations are described inChapter 14.

d. Preventing overtopping. To ensure that thespillway is adequate to protect the structure from overtop-ping, some amount of freeboard is added to the dam abovethe maximum pool water-surface elevation. This can varyfrom zero for structures that can withstand overtopping to2 m or more for structures where overtopping wouldconstitute a major hazard. The freeboard allowanceaccounts for wind set and wave action. Methods forestimating these quantities are discussed in Chapter 15.Risk analysis should be performed to determine the appro-priate top-of-dam elevation.

reducing downstream flooding. Narrow, ungated spillways

10-8. Flood-Control System Formulation

a. Objectives. The objectives of system formulationare to identify the individual components, determine thesize of each, determine the order in which the systemcomponents should be implemented, and develop anddisplay the information required to justify the decisions andthus secure system implementation. Section 4-10 describesseveral formulation strategies.

b. Criteria. Criteria for system formulation areneeded to distinguish the best system from amongcompeting alternative systems. The definition of “best” iscrucial. A reasonable viewpoint would seem to recognizethat simply aggregating the most attractive individual com-ponents into a system, while assuring physical com-patibility, could result in the inefficient use of resourcesbecause of system effects, data uncertainty, and thepossibility that all components may not be implemented. Itis proposed that the best system be considered to be asfollows:

(1) The system that includes the obviously goodcomponents while preserving flexibility for modification ofcomponents at future dates.

(2) The system which could be implemented at anumber of stages, if staging is possible, such that eachstage could stand on its own merits (be of social value) ifno more components were to be added.

c. General guidance. General guidance for formula-tion criteria are contained in the Principles and Standards(Water Resources Council 1973). The criterion of

EM 1110-2-142031 Oct 97

10-6

economic efficiency from the national viewpoint has been combinations of historic and synthetic floods are typicallyinterpreted to require that each component in a system used to evaluate reservoir flood-reduction performanceshould be incrementally justified, that is, each component (i.e., to develop regulated conditions frequency relations ataddition to a system should add to the value (net benefits) damage index stations), particular attention must be paid toof the total system. The environmental quality criteria can the selection or development of the system hydrology. Thebe viewed as favoring alternatives that can be structured to problem arises when evaluating complex reservoir systemsminimize adverse environmental impacts and provide with many reservoirs above common damage centers. Theopportunities for mitigation measures. Additional criteria problem increases with the size and complexity of the basinthat are not as formally stated as U.S. national policy are because the storm magnitudes and locations can favor oneimportant in decisions among alternatives. A formulated reservoir location over another. There are a large numberflood-control system must draw sufficient support from of storm centerings that could yield similar flows at aresponsible authorities in order to be implemented. In particular control point. Because of this, the contributionaddition, flood-control systems should be formulated so of a specific system component to reduced flooding at athat a minimum standard of performance (degree of risk) is downstream location is uncertain and dependent uponprovided so that public safety and welfare are adequately storm centering. This makes the selection or developmentprotected. of representative centerings crucial if all upstream

d. Environmental and other assessments. Of thesecriteria, only the national economic efficiency and mini- g. Desired evaluation. The desired evaluation formum performance standard have generally accepted regulated conditions is the expected or average condition somethods available for their rigorous inclusion in formula- that economic calculations are valid. The representativetion studies. Environmental quality analysis and social/ hydrograph procedure is where several proportions (ratiospolitical/institutional analyses related to implementation of one or more historic or synthetic events used tohave not developed technology applicable on a broad scale. represent system hydrology) are compatible with the sim-As a consequence, these criteria must guide the formulation ulation technique used, but care must be taken to reason-studies but, as yet, probably cannot directly contribute in a ably accommodate the storm centering uncertainty.structured formulation strategy. In discussions that follow, Testing the sensitivity of the expected annual damage tofocus is of necessity upon the economic criteria with the system hydrology (event centering) is appropriate andacceptable performance as a constraint, with the assump- necessary. Even if all historical floods of record are used,tion that the remaining criteria will be incorporated when there still may be some bias in computing expected annualthe formulation strategy has narrowed the range of alterna- damages if most historical floods were, by chance, centeredtives to a limited number for which the environmental and over a certain part of the basin and not over others. Forother assessments can be performed. instance, one reservoir site may have experienced several

e. Degrees of uncertainty. There will be varying adjacent to the area may, due to chance, not have had anydegrees of uncertainty in the information used in system severe floods.formulation. The hydrology will be better defined neargauging stations than it is in remote areas, and certainpotential reservoirs will have been more thoroughly inves-tigated than others. In addition, the accuracy of economicdata, both costs and value, existing or projected, is gener-ally lower than the more physically based data. Also, sinceconditions change over time, the data must be continuouslyupdated at each decision point. The practical accommoda-tion of information uncertainty is by limited sensitivityanalysis and continuing reappraisal as each component of asystem is studied for implementation.

f. Sensitivity analysis. Sensitivity analysis has, as itsobjective, the identification of either critical elements ofdata, or particularly sensitive system components, so thatfurther studies can be directed toward firming up theuncertain elements or that adjustments in system formula-tion can be made to reduce the uncertainty. Because

components are to be evaluated on a comparable basis.

severe historical floods, while another site immediately

10-9. General Study Procedure

After various alternative locations are selected for a reser-voir site to protect one or more damage centers, the fol-lowing steps are suggested for conducting the requiredhydrologic engineering studies:

a. Obtain a detailed topographic map of the regionshowing the locations of the damage areas, of proposedreservoir sites, and of all pertinent precipitation, snowpack,and stream-gauging stations. Prepare a larger scaletopographic map of the drainage basin tributary to the mostdownstream damage location. Locate damage centers,project sites, pertinent hydrologic measurement stations,and drainage boundaries above each damage center, projectsite, and stream-gauging station. Measure all pertinenttributary areas.

EM 1110-2-142031 Oct 97

10-7

throughout the basin and use several proportions of thoseb. Establish stage-discharge relations for each dam-age reach, relating the stages for each reach to a selectedindex location in that reach; procedures for doing this aredescribed in Flood-Damage Analysis Package User'sManual (HEC 1990b). Where local protection works areconsidered part of an overall plan of improvement, estab-lish the stage-discharge relation for each plan of localprotection.

c. Obtain area- and storage-elevation curves for eachreservoir site; select alternative reservoir capacities asappropriate for each site; select outlet and spillway ratingcurves for each reservoir, and develop a plan offlood-control operation for each reservoir. Determinemaximum regulated flows for each damage center.

d. Estimate the maximum critical duration of runofffor any of the plans of improvement, considering therelation of regulated flows at damage centers tounregulated flood hydrographs of design magnitude atthose damage centers. Prepare frequency curves ofunregulated peak flows and volumes of each of variousrepresentative durations, as described for peak flows inChapter 6, for each damage center index location, and foreach reservoir site. If seasonal variation of flood-controlspace is to be considered, these curves should be developedfor each season.

e. The two basic approaches for flood-control sim-ulation are complete period-of-record analysis and repre-sentative floods analysis. If flooding can occur during anytime of the year, the complete sequential analysis might befavored. However, if there is a separable flood season, e.g.,in the western states, then the representative stormapproach may be sufficient. For the storm approach,develop data for historical floods with storm centerings

floods to obtain flows at the damage centers representingthe full range of the flow-frequency-damage relationshipfor base conditions and for regulated conditions. Also,develop synthetic events that have consistency in volumesof runoff and peak flows and are reasonably representativeregarding upstream contributions to downstream flows.

f. Perform sequential analysis with the developedhydrology. The period-of-record simulation providessimulated regulated flow which can be analyzed directly todevelop flow-frequency relations. The representative floodapproach requires an assumption that the regulated-flowfrequency is the same as the natural-flow frequency.Frequency curves of regulated conditions at each damagecenter can then be derived from frequency curves ofunregulated flows simply by assuming that a given ratio ofthe base flood will have the same recurrence frequencywhether it is modified by regulatory structures or not. Thisassumption is valid as long as larger unregulated floodsalways correspond to the larger regulated flows.

g. Derive a flow-frequency and stage-dischargecurve for the index station at each damage center asdescribed in Chapters 6 and 8, for unregulated conditionsfor each plan of improvement. These can be used fordetermining average annual damage for unregulated con-ditions and for each plan of development and would thusform the primary basis for project selection.

h. Develop a PMF for each reservoir site, usingprocedures described in Chapter 7. These will be used as apossible basis for spillway design. Route the PMF througheach reservoir, assuming reasonably adverse conditions forinitial storage and available outlet capacity.

EM 1110-2-142031 Oct 97

11-1

Chapter 11Conservation Storage

11-1. General Considerations

a. Purposes. Water stored in the conservation poolcan serve many purposes. The primary purposes forconservation storage are water supply, navigation, low-flowaugmentation, fish and wildlife, and hydroelectric power.The water requirements for these purposes are discussed inthis chapter along with water quality considerations.Methods for estimating the conservation storage, or yield,are presented in Chapter 12.

b. Operational policy. In general, the operationalpolicy is to conserve available supplies and to release onlywhen supplemental flow is needed to meet downstreamrequirements. Water stored in the conservation pool alsoprovides benefits within the pool, such as lake recreationand fish and wildlife habitat.

c. Changing hydrology. When a reservoir is filled,the hydrology of the inundated area and its immediatesurroundings is changed in a number of respects. Theeffects of inflows at the perimeter of the reservoir aretranslated rapidly to the reservoir outlet, thus, effectivelyspeeding the flow of water through the reservoir. Also,large amounts of energy are stored and must be dissipatedor utilized at the outlet. The reservoir loses water byevaporation, and this usually exceeds preproject evapo-transpiration losses from the lake area. Siltation usuallyseals the reservoir bottom, but rising and falling waterlevels may alter the pattern of groundwater storage due tomovement into and out of the surrounding reservoir banks.At high stages, water may seep from the reservoir throughpermeable soils into neighboring catchment areas and so belost to the area of origin. Finally, sedimentation takes placein the reservoir and scour occurs downstream.

d. Storage allocation. The joint use of storage formore than one purpose creates problems of storage alloca-tion for the various purposes. While retained in reservoirstorage, water may provide benefits to recreation, fish,wildlife, hydropower, and aesthetics. Properly dischargedfrom the reservoir, similar benefits are achieved down-stream. Other benefits that can be derived from the reser-voir are those covered in this chapter, including municipaland industrial water supply, agricultural water supply,navigation, and low-flow augmentation.

e. Supplemental storage capacity. In most areas,supplemental storage capacity is required for sediment

deposition; otherwise, the yield capability of the reservoirmay be seriously diminished during the project's economiclife. Sediment storage is determined by estimating theaverage annual sediment yield per square mile of drainagearea from observations in the region and multiplying by thedrainage area and the economic life of the project. Trapefficiency of the reservoir is evaluated and the distributionof this estimated volume of sediment is determined, usingmethods described in EM 1110-2-4000 SedimentationInvestigations of Rivers and Reservoirs. Sediment surveyswithin the reservoir during actual operation will establishthe reliability of these estimates. Storage allocation levelsmay then be revised if the sediment surveys show asignificant difference between what was projected and whatwas measured. More complete descriptions of thetechniques used to determine reservoir sedimentation arepresented in EM 1110-2-4000.

f. Minimum pool. A minimum pool at the bottom ofactive conservation storage is usually established toidentify the lower limit of normal reservoir drawdown. Theinactive storage below the minimum pool level can be usedfor recreation, fish and wildlife, hydropower head, sedi-ment deposition reserve, and other purposes. In rareinstances, it might be used to relieve water supplyemergencies.

g. Reservoir outlets. Reservoir outlets must belocated low enough to withdraw water at desired rates withthe reservoir stage at minimum pool. These outlets candischarge directly into an aqueduct or into the river. In thelatter case, a diversion dam may be required downstream atthe main canal intake.

h. Computing storage capacity. Because the pri-mary function of reservoirs is to provide storage, their mostimportant physical characteristic is storage capacity.Capacities of reservoirs on natural sites must usually bedetermined from topographic surveys. The storage capac-ity can be computed by planimetering the area enclosedwithin each elevation contour throughout the full range ofelevations within the reservoir site. The increment ofstorage between any two elevation contours is usuallycomputed by multiplying the average of the areas at the twoelevations by the elevation difference. The summation ofthese increments below any elevation is the storage volumebelow that level. An alternative to the average-end-areamethod is the determination of the storage capacity by theconic method, which assumes that the volumes are morenearly represented by portions of a cone. This method isavailable in the HEC-1 Flood Hydrograph Packagecomputer program and is described in the program user'smanual. In the absence of adequate topographic maps,cross sections of the reservoir area are sometimes

EM 1110-2-142031 Oct 97

11-2

surveyed, and the capacity is computed from these vertical provide a specified yield. Chapter 12 describes procedurescross sections by using the formula for the volume of a for yield determination.prism.

11-2. Water Supply

a. Introduction. Water supply for any purpose isusually obtained from groundwater or from surface waters.Groundwater yields and the methods currently in use arecovered in Physical and Chemical Hydrogeology(Domenico and Schwartz 1990). This discussion is limitedto surface water supplies for low-flow regulation or fordiversion to demand areas.

(1) In some cases, water supply from surface watersinvolves only the withdrawal of water as needed from anearby stream. However, this source can be unreliablebecause streamflows can be highly variable, and the desiredamount might not always be available. An essentialrequirement of most water supply projects is that thesupply be available on a dependable basis. Reservoirs playa major role in fulfilling this requirement. Whatever theultimate use of water, the main function of a reservoir is tostabilize the flow of water, either by regulating a varyingsupply in a natural stream or by satisfying a varyingdemand by the ultimate consumer. Usually, some overallloss of water occurs in this process.

(2) In determining the location of a proposed reservoirto satisfy water needs, a number of factors should beconsidered. The dam should be located so that adequatecapacity can be obtained, social and environmental effectsof the project will be satisfactory, sediment deposition inthe reservoir and scour below the dam will be tolerable, thequality of water in the reservoir will be commensurate withthe ultimate use, and the cost of storing and transportingthe water to the desired location is acceptable. It isvirtually impossible to locate a reservoir site having com-pletely ideal characteristics, and many of these factors willbe competitive. However, these factors can be used asgeneral guidelines for evaluating prospective reservoirsites.

(3) In the planning and design of reservoirs for watersupply, the basic hydrologic problem is to determine howmuch water a specified reservoir capacity will yield. Yieldis the amount of water that can be supplied from thereservoir to a specified location and in a specified timepattern. Firm yield is usually defined as the maximumquantity of water that can be guaranteed with somespecified degree of confidence during a specific criticalperiod. The critical period is that period in a sequentialrecord that requires the largest volume from storage to

b. Municipal and industrial water use. The waterrequirement of a modern city is so great that a communitysystem capable of supplying a sufficient quantity of potablewater is a necessity. The first step in the design of awaterworks system is a determination of the quantity ofwater that will be required, with provision for the estimatedrequirements of the future. Next, a reliable source of watermust be located and, finally, a distribution system must beprovided. Water at the source may not be potable, sowater-purification facilities are ordinarily included as anintegral part of the system. Water use varies from city tocity, depending on the population, climatic conditions,industrialization, and other factors. In a given city, usevaries from season to season and from hour to hour.Planning of a water supply system requires that theprobable water use and its variations be estimated asaccurately as possible.

(1) Municipal uses of water may be divided intovarious classes. Domestic use is water used in homes,apartment houses, etc., for drinking, bathing, lawn andgarden sprinkling, and sanitary purposes. Commercial andindustrial use is water used by commercial establishmentsand industries. Public use is water required in parks, civicbuildings, schools, hospitals, churches, street washing, etc.Water that leaks from the system, unauthorizedconnections, and other unaccounted-for water is classifiedas loss and waste.

(2) The average daily use of water for municipal andindustrial purposes is influenced by many factors. Morewater is used in warm, dry climates than in humid climatesfor bathing, lawn watering, air conditioning, etc. Inextremely cold climates water may be wasted at faucets toprevent freezing of pipes. Water use is also influenced bythe economic status of the users. The per capita use ofwater in slum areas is much less than that in high-costresidential districts. Manufacturing plants often requirelarge amounts of water; however, some industries developtheir own water supply and place little or no demand on amunicipal system. The actual amount depends on theextent of the manufacturing and the type of industry.Zoning of the city affects the location of industries and mayhelp in estimating future industrial demands.

(3) About 80 percent of industrial water may be usedfor cooling and need not be of high quality, but water usedfor process purposes must be of good quality. In somecases, industrial water must have a lower content ofdissolved salts than can be permitted in drinking water.

EM 1110-2-142031 Oct 97

11-3

The location of industry is often much influenced by the the growing season. Other factors that affect wateravailability of water supply. If water costs are high, less requirements are the quality of the water, the amount ofwater is used, and industries will often develop their own land to be irrigated, and, of course, the cost of the water tosupply to obtain cheaper water. In this respect, the instal- the irrigator.lation of water meters in some communities has reducedwater use by as much as 40 percent. The size of the city (3) In estimating the amount of storage that will bebeing served is a factor affecting water use. Per capita use required in a reservoir for irrigation, the losses and wastetends to be higher in large cities than in small towns. The that occur in the irrigation system must be considered.difference results from greater industrial use, more parks, Losses and waste are usually divided into conveyance andgreater commercial use, and, perhaps, more loss and waste irrigation losses and waste. Conveyance losses and wastein the larger cities. All of these factors, plus estimated are those that occur in the conveyance and distributionpopulation projections, should be considered in designing a system prior to the application of water to crops. These arewaterworks system. dependent on the design and construction of the system and

(4) The use of water in a community varies almost Irrigation losses and waste are those that occur due to thecontinually. In midwinter the average daily use is usually slope of the irrigated land, the preparation of the land, soilabout 20 percent lower than the daily average for the year, condition, the method of irrigation, and the practices of thewhile in summer it may be 20 to 30 percent above the daily irrigator.average for the year. Seasonal industries such as canneriesmay cause wide variation in water demand during the year. (4) Usually, most of the irrigation losses and waste, asIt has been observed that for most communities, the well as a portion of the water applied to the irrigated lands,maximum daily use will be about 180 percent of the return to the stream. If there are requirements for flowaverage daily use throughout the year. Within any day, downstream of the reservoir, these return flows can belarge variations can be as low as 25 percent to as high as important in determining the amount of water that must be200 percent of the average for portions of the day. The released to meet such requirements.daily and hourly variations in water use are not usuallyconsidered in reservoir design, because most communities (5) In most areas, the need for irrigation water isuse distribution reservoirs (standpipes, etc.) to regulate for seasonal and depends on the growing season, the number ofthese variations. crops per year, and the amount of precipitation. For these

c. Agricultural water use. The need for agricultural from no water for some months up to 20 to 30 percent ofwater supply is primarily for irrigation. Irrigation can be the annual total for other months. This variation can have adefined as the application of water to soil to supplement very large effect on the amount of storage required and thedeficient rainfall in order to provide moisture for plant time of year when it is available.growth. In the United States, about 46 percent of all thewater used is for irrigation. Irrigation is a consumptive use;that is, most of the water is transpired or evaporated and isessentially lost to further use.

(1) In planning an irrigation project a number offactors must be considered. The first step would be toestablish the capability of the land to produce crops thatprovide adequate returns on the investment in irrigationworks. This involves determining whether the land isarable (land which, when properly prepared for agriculture,will have a sufficient yield to justify its development) andirrigable.

(2) The amount of water required to raise a cropdepends on the kind of crop and the climate. The plantsthat are the most important sources of food and fiber needrelatively large amounts of water. The most importantclimatic characteristic governing water need is the length of

also on how the system is operated and maintained.

reasons the variation of the demand is often high, ranging

11-3. Navigation and Low-Flow Augmentation

a. Objective. In designing a reservoir to supplywater for navigation and low-flow augmentation, theobjective is significantly different from objectives for theother purposes that have been discussed previously in thischapter. The objective is to supplement flows at one ormore points downstream from the reservoir. For naviga-tion, these flows aid in maintaining the necessary depth ofwater and alleviate silting problems in the navigable chan-nel. Low-flow augmentation serves a number of purposesincluding recreation, fish and wildlife, ice control, pollutionabatement, and run-of-river power projects. Under certainconditions, low-flow augmentation provides water for theother purposes discussed in this chapter. For instance, ifthe intake for a municipal and industrial water supply is atsome point downstream of the reservoir, the objective maybe to supplement low flows at that point.

EM 1110-2-142031 Oct 97

11-4

b. Criteria for navigability. There are no absolutecriteria for navigability and, in the final analysis, economiccriteria control. The physical factors that affect the cost ofwaterborne transport are depth of channel, width andalignment of channel, locking time, current velocity, andterminal facilities. Commercial inland water transport is,for the most part, accomplished by barge tows consisting of1 to 10 barges pushed by a shallow-draft tug. The cost of atrip between any two terminals is the sum of the fuel costsand wages, fixed charges, and other operating expensesdepending on the time of transit. Reservoirs aid inreducing these costs by providing the proper depth of waterin the navigation channel, or by providing a slack-waterpool in lock and dam projects. Storage reservoirs canrarely be justified economically for navigation purposesalone and are usually planned as multipurpose projects.Improving navigation by using reservoirs is possible whenflood flows can be stored for release during low-flowseasons.

c. Supplying deficiencies without waste. The idealreservoir operation for navigation or low-flow augmenta-tion would provide releases so timed as to supply thedeficiencies in natural flow without waste. This is possibleonly if the reservoir is at the head of a relatively shortcontrol reach. As the distance from the reservoir to thereach is increased, releases must be increased to allow foruncertainties in estimating intermediate runoff and forevaporation and seepage enroute to the reach to be served.Moreover, the releases must be made sufficiently far inadvance of the need to allow for travel time to the reach,and in sufficient quantity so that after reduction by channelstorage, the delivered flows are adequate. The waterrequirement for these releases is considerably greater thanthe difference between actual and required flows.

d. Climate. Climate can also affect reservoir opera-tion for low-flow regulation. Depending on the purpose tobe served, the releases may be required only at certaintimes of the year or may vary from month to month. Forpollution abatement, the important factors are the quality ofthe water to be supplemented, the quality of the water inthe reservoir, and the quality standard to be attained. Also,the level of the intakes from which releases will be madecan be a very sufficient factor in pollution abatement, sincethe quality can vary from one level to another in thereservoir. Long-term variations can occur due to increasedcontamination downstream of a reservoir. This should beconsidered in determining the required storage in thereservoir.

11-4. Fish and Wildlife

a. Added authorized purposes. As shown in Fig-ure 2-1, fish and wildlife and subsequent environmentalpurposes have been added as authorized purposes since1960. Because many of the reservoirs were built prior tothat time, their authorized purposes and regulation plansmay not adequately reflect the more recent environmentalobjectives. Therefore, there is an increasing demand andneed for the evaluation of environmental impacts for theseprojects.

b. Water level fluctuations. The seasonal fluctua-tion that occurs at many flood control reservoirs and thedaily fluctuations that occur with hydropower operationoften result in the elimination of shoreline vegetation andsubsequent shoreline erosion, water quality degradation,and loss of habitat for fish and wildlife. Adverse impactsof water level fluctuations also include loss of shorelineshelter and physical disruption of spawning and nests.

c. Water level management. Water-level manage-ment in fluctuating warm-water and cool-water reservoirsgenerally involves raising water levels during the spring toenhance spawning and the survival of young predators.Pool levels are lowered during the summer to permitregrowth of vegetation in the fluctuation zone. Fluctua-tions may be timed to benefit one or more target species;therefore, several variations in operation may be desirable.In the central United States, managers frequently recom-mend small increases in pool levels during the autumn forwaterfowl management.

d. Fishery management. Guidelines to meetdownstream fishery management potentials are developedbased on project water quality characteristics and watercontrol capabilities. To do so, an understanding of thereservoir water quality regimes is critical for developing thewater control criteria to meet the objectives. For example,temperature is often one of the major constraints of fisherymanagement in the downstream reach, and water controlmanagers must understand the temperature regime in thepool and downstream temperature requirements, as well asthe capability of the project to achieve the balance requiredbetween the inflows and the releases. Releasing cold waterdownstream where fishery management objectives requirewarm water will be detrimental to the downstream fishery.Conversely, releasing warm water creates difficulty inmaintaining a cold-water fishery downstream.

EM 1110-2-142031 Oct 97

11-5

e. Water temperature management. Water control reach. In addition, storage for hydropower reduces theactivities can also impact water temperatures within the quantity of spill, and as a result, juvenile fish must passpool by changing the volume of water available for a through the turbines. The delay in travel time subjects theparticular layer. In some instances, cold-water reserves juvenile fish to greater exposure to birds and predator fish,may be necessary to maintain a downstream temperature and passage through the powerhouse turbines increasesobjective in the late summer months; therefore, the avail- mortality. To improve juvenile survival, storage has beenability of cold water must be maintained to meet this made available at some projects to augment river flows,objective. For some projects, particularly in the southern and flows are diverted away from the turbine intakes andUnited States, water control objectives include the mainte- through tailraces where the fish are collected fornance of warm-water fisheries in the tailwaters. In other transportation or released back into the river. Barges orinstances, fishery management objectives may include the tank trucks can be used to transport juveniles from themaintenance of a two-story fishery in a reservoir, with a collector dams to release sites below the projects. Otherwarm-water fishery in the surface water, and a cold-water Corps projects have been modified so the ice and trashfishery in the bottom waters. Such an objective challenges spillways can be operated to provide juvenile fish passage.water control managers to regulate the project to maintainthe desired temperature stratification while maintaining i. Wildlife habitat. Project regulation can influencesufficient dissolved oxygen in the bottom waters for the wildlife habitat and management principally through watercold-water fishery. Regulation to meet this objective level fluctuations. The beneficial aspects of periodicrequires an understanding of operational affects on drawdowns on wildlife habitat are well documented inseasonal patterns of thermal stratification, and the ability to wildlife literature. Drawdowns as a wildlife managementanticipate thermal characteristics. technique can, as examples, allow the natural and artificial

f. Minimum releases. Minimum instantaneous flows tion and maintenance of artificial nesting structures, allowcan be beneficial for maintaining gravel beds downstream the control of vegetation species composition, and ensurefor species that require this habitat. However, dramatic mast tree survival in greentree reservoirs. Wildlife benefitschanges in release volumes, such as those that result from of periodic flooding include inhibiting the growth offlood-control regulation, as well as hydropower, can be undesirable and perennial plants, creating access anddetrimental to downstream fisheries. Peaking hydropower foraging opportunities for waterfowl in areas such asoperations can result in releases from near zero to very greentree reservoirs, and ensuring certain water levels inhigh magnitudes during operations at full capacity. Main- stands of vegetation to encourage waterfowl nesting andtaining minimum releases and incorporating reregulation reproduction.structures are two of the options available to mitigate thisproblem.

g. Fishing versus peak power. In some instances,tailwater fishing is at a maximum during summer weekendsand holidays, and this is a time when power generation maybe at a minimum and release near zero. Maintainingminimum releases during weekend daylight hours mayimprove recreational fishing, but may reduce the capabilityto meet peak power loads during the week because of lowerwater level (head) in the reservoir. In these instances,water control managers will be challenged to regulate theproject with consideration of these two objectives.

h. Anadromas fish. Regulation for anadromous fishis particularly important during certain periods of the year.Generally, upstream migration of adult anadromous fishbegins in the spring of each year and continues throughearly fall, and downstream migration of juvenile fish occurspredominantly during the spring and summer months. Thereduced water velocities through reservoirs, in comparisonwith preproject conditions, may greatly lengthen the traveltime for juvenile fish downstream through the impounded

revegetation of shallows for waterfowl, permit the installa-

11-5. Hydroelectric Power

a. General. The feasibility of hydroelectric devel-opment is dependent upon the need for electric power, theavailability of a transmission system to take the power fromthe point of generation to the points of demand, and theavailability of water from streamflow and storage to pro-duce power in accordance with the capacity and energydemands in the power market area. Also, the project'spower operations must be coordinated with the operationsfor other project purposes to ensure that all purposes areproperly served. Each of these factors must be investigatedto ensure that the project is both feasible and desirable andto minimize the possibility that unforeseen conflicts willdevelop between power and other water uses during theproject life.

(1) The ability of a project to supply power is meas-ured in terms of two parameters: capacity and energy.Capacity, commonly measured in kilowatts (kw), is themaximum amount of power that a generating plant candeliver. Energy, measured in kilowatt-hours (kwh), is the

EM 1110-2-142031 Oct 97

11-6

Figure 11-1. Typical annual load duration curve

amount of actual work done. Both parameters are impor-tant, and the operation of a hydroelectric project is sensi-tive to changes in the demand for either capacity orenergy.

(2) Experience has indicated that it is very unlikelythat power demands will remain unchanged during theproject life. Furthermore, the relative priority of variousother water uses can change during the project life, andthere are often legal, institutional, social, or environmentalfactors that might affect the future use of water at a par-ticular project. Consequently, the feasibility studies for aproposed project must not be limited to conditions that areonly representative of the current time or the relativelynear future. Instead, the studies must include consider-ations of future conditions that might create irreconcilableconflicts unless appropriate remedial measures are pro-vided for during project formulation.

(3) This section presents general concepts for thehydrologic analyses associated with the planning, design,and operation of hydroelectric projects and systems. Moredetailed information is provided in EM 1110-2-1701.Other investigations that influence or affect the hydrologicstudies will be discussed to the extent that their outcomemust be understood by the hydrologic engineer.

b. Types of hydroelectric load. Power developments,for purposes of this discussion, can be classified withrespect to the type of load served or the type of sitedevelopment proposed. The two categories related to thetype of load served are baseload and peaking plants.

(1) Base load. Baseload plants are projects that gen- generate hydroelectric power to supplement baseloaderate hydroelectric power to meet the baseload demand. generation during periods of peak power demands. TheThe baseload demand is the demand that exists 100 per- peak power demands are the loads that exist primarilycent of the time. The baseload can readily be seen in Fig- during the daylight hours. The time of occurrence andure 11-1 as the horizontal dashed line on a typical annual magnitude of peak power demands are shown on a loadload duration curve. This curve displays the percent of curve in Figure 11-2. This curve shows the time variationtime during a given year that a given capacity demand is in power demands for a typical week. Depending upon theequaled or exceeded. The area under this curve represents quantity of water available and the demand, a peakingthe total energy required to meet the load during the year. plant may generate from as much as 18 hr a day to asUsually, the baseload demand is met by thermal little as no generation at all, but it is usually 8 hr a day orgenerating facilities. However, in cases where there is a less. Peaking plants must supply sufficient capacity torelatively abundant supply of water that is available with a satisfy the peak capacity demands of a system andhigh degree of reliability and where fuel is relatively sufficient energy to make the capacity usable on the load.scarce, hydroelectric projects may be developed to meet This means that energy or water should be sufficient tothe baseload demands. These projects would then operate supply peaking support for as long and as oftenat or near full capacity 24 hr per day for long periods of as the capacity is needed. In general, a peaking hydroelec-time. This type of development is not feasible where there tric plant is desirable in a system that has thermalis a large seasonal variation in streamflow unless the generation facilities to meet the baseload demands. Thebaseflow is relatively high or unless there is a provision hydroelectric generating facilities are par t icular lyfor a large volume of power storage in the project. adaptable to the peaking operat ion because

(2) Peaking load. Peaking plants are projects that

EM 1110-2-142031 Oct 97

11-7

Figure 11-2. Weekly load curve for a large electric system

their loading can be changed rapidly. Also, the factors and sometimes rapid fluctuations in releases of waterthat make seasonal variations in streamflow a major through the generating units. It is often necessary toproblem in baseload operation are usually quite easily provide facilities to re-regulate the power releases if fluc-overcome in a peaking plant if some storage can be tuations of water levels below the project are not tolerable.provided. Because storage projects are conducive to multipurpose

c. Project types. There are three major classifica- function of the guaranteed output during a multi-year drytions of hydroelectric projects: storage, run-of-river, and period, it is usually necessary to make detailed routingpumped storage. There are also combinations of projects studies to determine the storage requirements, installedthat might be considered as separate classifications, but for capacity, firm energy, and an operating plan.purposes of discussing hydrologic analysis it is necessaryto define only these three types.

(1) Storage projects. Storage plants are projects thatusually have heads in the medium to high range (™ 25 m)and have provisions for storing relatively large volumes ofwater during periods of high streamflow in order to pro-vide water for power generation during periods of defi-cient streamflow. Considerable storage capacity may berequired because the period of deficient flow is quitefrequently more than a year long and, in some instances,may be several years long. Because use of the storedwater entails drawdown of the power storage, it is desir-able that other water uses associated with the developmentof a storage plant permit frequent and severe drawdownsduring dry periods. Peaking operation, which is quitefrequently associated with storage projects requires large

use and because the power output from a storage plant is a

(2) Run-of-river projects. Run-of-river plants havelittle or no power storage and, therefore, must generatepower from streamflow as it occurs. The projects gener-ally have productive heads in the low to medium range (5-30 m) and are quite frequently associated with naviga-tional developments or other multipurpose developmentswith limitations on reservoir drawdowns. Because of theabsence or near-absence of storage in run-of-river projects,there is usually very little operational flexibility in theseprojects, and it is necessary that all water uses becompatible. The existence of one or more storage projectsin the upstream portion of a river basin may make arun-of-river project in the lower portion of the basin feasi-ble where it would not otherwise be feasible. In thissituation, the storage projects provide a regulated outflow

EM 1110-2-142031 Oct 97

11-8

that is predictable and usable, while the natural streamflow turbines and generator motors located along a tunnel ormight be neither. penstock connecting the forebay and afterbay. The water is

pumped from the afterbay to the forebay when the normal(a) Run-of-river projects may have provisions for a

small amount of storage, often called pondage. Thispondage detains the streamflow during off-peak periods indaily or weekly cycles for use in generating power duringpeak demand periods. If the cycle of peaking operation is asingle day, the pondage requirements are based on the flowvolume needed to sustain generation at or near installedcapacity for 12 hr. If more storage capacity is availableand large fluctuations in the reservoir surface arepermissible, a weekly cycle of peaking operation may beconsidered. Because industrial and commercial consump-tion of power is significantly lower on weekends than onweek days, an “off-peak” period is created from Fridayevening until Monday morning. If generation from thehydroelectric peaking plants is not required during thisperiod, water can be stored in the pondage for use duringthe 5-day peak-load period.

(b) Because of the relatively low heads associated withrun-of-river projects, the tailwater fluctuations are usuallyquite important, particularly in peaking operations. Also,flood flows may curtail power generation due to hightailwater. While flow-duration analysis can be used toestimate average annual energy production, sequentialanalysis may be required for more detailed analysis ofextreme conditions.

(3) Pumped-storage projects. Pumped-storage plantsare projects that depend on pumped water as a partial ortotal source of water for generating electric energy. Thistype of project derives its usefulness from the fact that thedemand for power is generally low at night and on week-ends; therefore, pumping energy at a very low cost will beavailable from idle thermal generating facilities or run-of-river projects. If there is a need for peaking capacity and ifthe value of peaking power generation sufficiently exceedsthe cost of pumping energy (at least a ratio of 1.5 to 1.0because roughly 3 kwh of pumping energy are necessary todeliver enough water to provide 2 kwh of energygeneration), pumped storage might be feasible. There arethree types of pumped-storage development: diversion,off-channel, and in-channel, which are detailed in Chapter7 of EM 1110-2-1701.

(a) In general, pumped storage projects consist of ahigh-level forebay where pumped water is stored until it isneeded for generation and a low-level afterbay where thepower releases are regulated, if necessary, and from whichthe water is pumped. The pumping and generating aredone by generating units composed of reversible pump

power demand is low and least expensive and releasedfrom the forebay to the afterbay to generate power whenthe demand is high and most costly. The feasibility ofpumped-storage developments is dependent upon the needfor relatively large amounts of peaking capacity, theavailability of pumping energy at a guaranteed favorablecost, and a load with an off-peak period long enough topermit the required amount of pumping.

(b) A unique feature of pumped-storage systems isthat very little water is required for their operation. Oncethe headwater and tailwater pools have been filled, onlyenough water is needed to take care of evaporation andseepage. For heads up to 300 m, reversible pump turbineshave been devised to operate at relatively high efficiency aseither a pump or turbine. The same electrical unit serves asa generator and motor by reversing poles. Such a machinemay reduce the cost of a pumped-storage project byeliminating the extra pumping equipment and pump house.The reversible pump turbine is a compromise in designbetween a Francis turbine and a centrifugal pump. Itsfunction is reversed by changing the direction of rotation.

d. Need for hydroelectric power. The need forpower is established by a power market study or survey.The feasibility of a particular hydroelectric project orsystem is determined by considering the needs as estab-lished by the survey, availability of transmission facilities,and the economics of the proposed project or projects.Although forecasts of potential power requirements withina region to be served by a project are not hydrologicdeterminations, they are essential to the development ofplans for power facilities and to the determination ofproject feasibility and justification. The power marketsurvey is a means of evaluating the present and potentialmarket for electrical power in a region.

(1) The survey must provide a realistic estimate of thepower requirements to be met by the project and mustshow the anticipated rate of load growth from initial oper-ation of the project to the end of its economic life. Thesurvey also provides information regarding the character-istics of the anticipated demands for power. Thesecharacteristics, which must be considered in hydrologicevaluations of hydroelectric potential, include the seasonalvariation of energy requirements (preferably on a monthlybasis), the seasonal variation of capacity requirements (alsopreferably on a monthly basis), and the range of usableplant factors for hydroelectric projects under both adverseand average or normal flow conditions.

kW =1

11.81QHe

kWh =1

11.81 P

t

0

QtHtedt

EM 1110-2-142031 Oct 97

11-9

(11-1)

(11-2)

(2) The results of a power market survey might befurnished to the hydrologic engineer in the form of loadduration or load curves (Figures 11-1 and 11-2) showingthe projected load growth, the portion of the load that canbe supplied by existing generating facilities, and the portion In order to convert a project's power output to energy,that must be supplied by future additions to the generating Equation 11-1 must be integrated over time:system. From these curves, the characteristics of plannedhydroelectric generating facilities can be determined.Because these data are developed from the needs alonewithout consideration of the potential for supplying theseneeds, the next step is to study the potential forhydroelectric development, given the constraints estab- wherelished in the study of needs.

e. Estimation of hydroelectric power potential.Traditionally, hydroelectric power potential has been kWh = energy generated during a time period, kWhdetermined on the basis of the critical hydro-period asindicated by the historical record. The critical hydro- Q = average streamflow during the time period,period is defined as the period when the limitations of m /sechydroelectric power supply due to hydrologic conditionsare most critical with respect to power demands. Thus, the H = average head during the time period, m (Head =critical period is a function of the power demand, the headwater elevation - tailwater elevation -streamflow, and the available storage. In preliminary hydraulic losses)project planning, the estimates of power potential are oftenbased on a number of simplifying assumptions because of t = number of hours in the time periodthe lack of specific information for use in more detailedanalyses. Although these estimates and the assumptions e = overall efficiency expressed as the product ofupon which they are based are satisfactory for preliminary the generator efficiency and the turbineinvestigations, they are not suitable for every level of efficiencyengineering work. Many factors affecting the design andoperation of a project are ignored in these computations. In practice, the summation of energy production over theTherefore, detailed sequential analyses of at least the critical period is performed with a sufficiently small timecritical hydro-period should be initiated as early as step to provide reasonable estimates of head and, therefore,possible, usually when detailed hydrologic data and some energy. Two basic approaches are available: flow-durationapproximate physical data concerning the proposed project and sequential analysis.become available. Because of the availability of computerprograms for accomplishing these sequential routings, they (2) For run-of-river projects, where the headwatercan be done rapidly and at a relatively low cost. elevation does not vary significantly, the flow-duration

(1) The manner in which the streamflow at a given site production. The duration curve can be truncated at theis used to generate power depends upon the storage minimum flow rate for power production. The curve canavailable at the site, the hydraulic and electrical capacities also be truncated for high-flows if the tailwater elevation isof the plant, streamflow requirements downstream from the too high for generation. The remaining curve is convertedplant, and characteristics of the load to be served. In to capacity-duration and integrated to obtain averagetheory, the hydroelectric power potential at a particular annual energy. Hydropower Analysis Using Flow-Durationsite, based on repetition of historical runoff, can be esti- Procedures HYDUR (HEC 1982d) was developed tomated by identifying the critical hydro-period and obtaining perform energy computations based on flow-duration data.estimates of the average head and average streamflow EM 1110-2-1701 describes HYDUR in paragraph C-2b andduring this critical period. The data can then be used in the the flow-duration method in Section 5-7.equation below to calculate the power available from theproject:

kW = power available from the project, kW

3

approach can be used to estimate average annual energy

EM 1110-2-142031 Oct 97

11-10

(3) Sequential streamflow analysis will be applied to requirements and also reduce the frequency of use ofmost reservoir studies. The procedure allows detailed flood-control facilities.computations of the major parameters affecting hydro-power (e.g., headwater and tailwater elevation, efficiency, (2) Water for municipal, industrial, or agricultural useand flow release). By performing the analysis in suffi- can be passed through the generating units with no harmfulciently small time steps, an accurate simulation of the effects if the point of withdrawal for the other use is belowreservoir operation, power capacity and energy production the point where the power discharge enters the river.can be obtained. Chapter 5 of EM 1110-2-1701, Sec- Re-regulation may be required for hydropower peakingtions 5-8 through 5-10, provides a discussion of sequential operations to “smooth out” the power releases. Conflictsrouting studies. Appendix C provides information con- between power and these consumptive uses more likelycerning computer programs that are available for use in occur when the withdrawal for other uses is directly fromthese studies. the reservoir. When the withdrawal is from the reservoir of

f. Hydropower effect on other project purposes. purpose may require that special attention be given toUsually, power generation must have a high priority rela- intake facilities for the other purposes because of thetive to other conservation uses. Consequently, thorough relatively large drawdown associated with storage projects.investigations of all aspects of the power operation must beconducted to ensure that the power operations do not create (3) Low-flow augmentation for navigation, recreation,intolerable situations for other authorized or approved or fish and wildlife can be accomplished by releaseswater uses. Likewise, the power operations must be through power generating units. In the case of baseloadcoordinated with other high priority purposes such as flood projects, the power release is ideally suited for this type ofcontrol and municipal water supply to ensure that the use. With peaking projects, however, a re-regulationplanned power operation will not interfere with the structure may be necessary to provide the relativelyoperations for these purposes. The operation rules that are uniform releases that might be required for navigation ornecessary to effect the coordination are usually developed for in-stream recreation. Release of water for qualityand tested using engineering judgment and detailed enhancement can sometimes be accomplished through thesequential routing studies. However, it is necessary to generating units. Although the intakes for the turbines aredefine the interactions between power and other project usually located at a relatively low elevation in the reservoirpurposes before initiating operation studies. where dissolved oxygen content might be low, the

(1) Power generation is generally compatible with most below the project may produce water with an acceptablepurposes that require releases of water from a reservoir for dissolved oxygen content. The water released from thedownstream needs. However, power generation usually lower levels of the reservoir is normally at a relatively lowcompetes with purposes that require withdrawal of the temperature and, thus, ideal for support or enhancement ofwater directly from the reservoir or that restrict fluctuations a cold-water fishery downstream. If warm waters arein the reservoir level. Flood-control requirements needed for in-stream recreation, for fishery requirements,frequently conflict with power operations because or for any other purpose, a special multilevel intake may beflood-control needs may dictate that storage space in a required to obtain water of the desired temperature.reservoir be evacuated at a time when it would be benefi-cial to store water for use in meeting future power (4) Recreation values at a reservoir project may bedemands. Furthermore, when extensive flooding is antici- enhanced, somewhat, by the inclusion of power because apated downstream from a reservoir project, it may be much larger reservoir is frequently required, and that maynecessary to curtail power releases to accomplish increase opportunities for extensive recreational activities.flood-control objectives. It is often possible to pass part or Unfortunately, however, the large drawdowns associatedall of the flood-control releases through the generating with the big storage projects create special problems withunits, thereby reducing the number of additional outlets respect to the location of permanent recreational facilitiesneeded and significantly increasing the energy production and may create mudflats that are undesirable from theover what would be possible if the flood-control releases standpoint of aesthetics and public health requirements.were made through conduits or over the spillway. Also, The drawdown may also expose boaters, swimmers, andmany of the smaller floods can be completely regulated other users to hazardous underwater obstacles unlesswithin the power drawdown storage, an operation that is provisions are made to remove these obstacles to a pointbeneficial to power because it provides water for power well below the maximum anticipated drawdown.generation that might otherwise have been spilled. This Obviously the time of occurrence of extreme drawdownjoint use can reduce the exclusive flood-control storage

a storage project, the inclusion of power as a project

oxygenation that occurs in the tailrace and in the stream

EM 1110-2-142031 Oct 97

11-11

conditions is an important factor in determining the degree source waters throughout the world to decline as a result ofof conflict with recreation activities. increased urbanization and industrialization and as a result

11-6. Water Quality Considerations

a. Water quality definition. Water quality deals withthe kinds and amounts of matter dissolved and suspendedin natural water, the physical characteristics of the water,and the ecological relationships between aquatic organismsand their environment. It is a term used to describe thechemical, physical, and biological characteristics of waterin respect to its suitability for a particular purpose. Thesame water may be of good quality for one purpose or use,and bad for another, depending on its characteristics andthe requirements for the particular use.

b. Water quality parameters. In general, physicalparameters define the water quality characteristics thataffect our senses while chemical and biological parametersindex the chemical and biological constituents present inthe water resource system. However, these are notindependent, but are actually highly related. For example,chemical waste discharges may affect such physical factorsas density and color, may alter chemical parameters suchan pH and alkalinity, and may affect the biologicalcommunity in the water. Even so, the physical, chemical,and biological subdivision is a useful way to discuss waterquality conditions. The following sections describe thewater quality parameters frequently associated withreservoirs. EM 1110-2-1201 provides details on parame-ters, assessment techniques, plus data collection andanalysis.

11-7. Water Quality Requirements

A wide variety of demands are made for the use of waterresources. Water of a quality that is unsatisfactory for oneuse may be perfectly acceptable for another. The level ofacceptable quality is often governed by the scarcity of theresource or the availability of water of better quality.

a. Domestic use. The use of water for domesticpurposes such as drinking, culinary use, and bathing isgenerally considered to be the most essential use of ourwater resources. The regulations for the quality of thiswater are likewise higher than for most (but not all) otherbeneficial uses of water. In early times, the quality of thewater supply source and the quality at the delivery pointwere synonymous; but the general degradation of bothsurface water quality and shallow ground water quality hasmade it necessary, in most cases, for some degree of watertreatment to be used to produce acceptable water fordomestic use. In recent decades, there has been a strongtrend (which is likely to continue) for the quality of the

of changes in agricultural practices. At the same time,populations are coming to expect a higher standard ofhealth and well-being; and as a result, the regulations foracceptable domestic water continue to rise and enlarge therole of water treatment.

b. Drinking water standards. Drinking water stan-dards for the world as a whole have been set by the WorldHealth Organization (WHO). One should keep in mind thatthese standards do not describe an ideal or necessarilydesirable water, but are merely the maximum values ofcontaminant concentration which may be permitted. It ishighly desirable to have water of much better quality. Inthe United States, the Environmental Protection Agency(EPA) sets regulations that legally apply to public drinkingwater and water supply systems. The regulations aredivided into three categories: bacterial, physical, andchemical characteristics. They are defined in terms ofmaximum contaminant levels (MCL's). Bacterial quality isdefined by establishing the sampling sequence, the methodof analysis, and the interpretation of test results for thecoliform organisms which serve as presumptive evidence ofbacterial contamination from intestinal sources. Analysis isgenerally made for total coliform, fecal coliform, andstreptococci coliform. The limits on biological andphysical parameters, and on chemical elements orcompounds in water are documented in Water Supply andSewerage (McGhee 1991).

c. Quality of source waters. The drinking waterstandards are the end product of a production line whichbegins with the source water as a raw material and pro-ceeds through the various unit processes of water treatmentand finally water distribution. The quality of source watersfor other uses such as agricultural and industrial watersupply, fish and other aquatic life, and recreation are set bystate regulations of receiving waters. Other specific usesmay include regulations for navigation, wild and scenicrivers, and other state-specific needs. The state regulationsare subject to EPA approval. The regulations of a givenstate may take a variety of forms but are often specified bystream reaches including associated natural or constructedimpoundments. Each reach may be classified for itsvarious water uses and water quality standards defined foreach use.

(1) Industry uses water as a buoyant transportingmedium, cleansing agent, coolant, and as a source of steamfor heating and power production. Often the qualityrequired for these purposes is significantly higher than thatrequired for human consumption. The availability of waterof high quality is often an important parameter in site

EM 1110-2-142031 Oct 97

11-12

selection by an industry. The needs of a particular industry cyanides, oil, solids, turbidity, or insecticides are known toas to both quantity and quality of water varies with the exist.competition for water, the efficiency of the plant processwith regard to water utilization, the recycling of water, the (5) For an aquatic system to be acceptable for swim-location of the plant site and the ratio of the cost of the ming and bathing, it must be aesthetically enjoyable (i.e.,water to the cost of the product. For economic reasons and free from obnoxious floating or suspended substances,for reasons of quality control and operation responsibility, objectionable color and foul odors), it must contain noindustries with high water requirements usually develop substances that are toxic upon ingestion or irritating to thetheir own supply and treatment facilities. skin or sense organs, and it must be reasonably free of

(2) Farmstead water is that water used by the human first two terms as related to swimming and bathing exceptfarm population for drinking, food preparation, bathing, in qualitative terms. In the United States, numerous stan-and laundry. It also includes water used for the washing dards exist for bacteriological quality based on the coliformand hydrocooling of fruits and vegetables, and water used count in the water. Generally, the standards rangein the production of milk. The quality of water desired for downward from 1,000 coliform bacteria per 100 ml to asfarmstead use is generally that required for public water low as 50 per 100 ml. Such standards are not based onsupplies. It is not feasible to set rigid quality standards for demonstrated transmittal of disease, but have been estab-irrigation waters because of such varied and complex lished because in these ranges the standards arefactors as soil porosity, soil chemistry, climatic conditions, economically reasonable and no problems appear. The usethe ratio of rainwater to irrigation water, artificial and of an aquatic system for boating and aesthetic enjoyment isnatural drainage, relative tolerance of different plants, and generally not so demanding as the requirements for swim-interferences between and among constituents in the water. ming or the propagation of fish and aquatic life althoughExamples of the latter are the antagonist influence of these three water uses are usually closely linked.calcium-sodium, boron-nitrates, and selenium-sulfates.

(3) Water quality parameters of importance forirrigation are sodium, alkalinity, acidity, chlorides, bicar-bonates, pesticides, temperature, suspended solids, radio-nuclids, and biodegradable organics. All of these factorsneed to be weighed carefully in evaluating the suitability ofwater for irrigating a particular crop. There is surprisinglylittle data on the effect of water quality on livestock, butgenerally they thrive best on water meeting human drinkingwater standards. The intake of highly mineralized water byanimals can cause physiological disturbances of varyingdegrees of severity. In some cases, particular ions such asnitrates, fluorides, selenium salts, and molybdenum may beharmful. Certain algae and protozoa have also been proventoxic to livestock.

(4) The basic purpose of water quality criteria foraquatic life is to restore or maintain environmental condi-tions that are essential to the survival, growth, reproduc-tion, and general well-being of the important aquaticorganisms. These criteria are ordinarily determined with-out the aid of economic considerations. Generally, anumber of major problems arise in establishing waterquality criteria for an aquatic community because of theinability to quantify the effects of the pertinent parametersand reduce them to a conceptual model that describes thenature of the biological community which will developunder a given set of conditions. But extensive researchshould be considered when unusually high concentrationsof such parameters as alkalinity, acidity, heavy metals,

pathogenic bacteria. Standards generally do not cover the

11-8. Reservoir Water Quality Management

Reservoirs may serve several purposes in the managementof water quality. If used properly, substantial benefits canbe achieved. On the other hand, unwise use of reservoirsmay cause increased quality degradation. Benefits mayaccrue as a result of detention mixing or selective with-drawal of water in a reservoir or the blending of watersfrom several reservoirs. The effects of improper manage-ment are often far-reaching and long-term. They mayrange from minor to catastrophic, and may be as obvious asa fish kill or subtle and unnoticed. It is essential that allwater control management activity and especially real-timeactions include valid water quality evaluation as a part ofthe daily water control decision process. It must beunderstood that water quality benefits accumulate slowly,build on each other, and can become quite substantial overtime. This is in contrast to the sudden benefits that comefrom a successful flood-control operation. Water qualitymanagement requirements, objectives, and standards arepresented in EM 1110-2-3600.

a. Reservoirs in streams. The presence of a reser-voir in a stream affects the quality of the outflow as com-pared to the inflow by virtue of the storage and mixingwhich takes place in the reservoir. The effect of such animpoundment may be easily evaluated for conservativeparameters if the waters of the reservoir are sufficientlymixed that an assumption of complete mixing within an

EM 1110-2-142031 Oct 97

11-13

analysis time period does not lead to appreciable error. reservoirs that receive substantial sediment will have aHowever, this assumption is limited to relatively small, short useful life. Planning should include evaluation of theshallow reservoirs. ultimate fate and possible replacement of such reservoirs.

Reservoir sedimentation is covered in EM 1110-2-4000.b. Reservoir outflow and inflow. The simplest tech-

nique requires the assumption that the reservoir outflowduring a given time period is of constant quality and equalto the quality of the reservoir storage at the end of thecomputation time period. It is then assumed that the inflowfor the time period occurs independently of the outflow,and reservoir quality is determined by a quality massbalance at the end of the time period. This approach isequivalent to the mass balance of water in reservoirrouting.

c. Reservoir water quality. Simple mass balanceprocedures may be applicable is some situations; however,usually more comprehensive methods should beconsidered. Chapter 4, “Water Quality Assessment Tech-niques,” in EM 1110-2-1201 describes various techniquesavailable for assessing reservoir water quality conditions.There is a hierarchy of available techniques that reflectsincreasing requirements of time, cost, and technical exper-tise. The increasing efforts should provide accompanyingincreases in the degree of understanding and resolution ofthe problem and causes. This hierarchy includes screeningdiagnostic and predictive techniques, which are described.

d. Reservoirs as detention basins. Reservoir mixingis a continual process where low inflows of poor quality arestored and mixed with higher inflows of better quality.Generally, this is accomplished in large reservoirs whereannual or even multiple-year flows are retained, but theconcept extends to small reservoirs in which weekly oreven daily quality changes occur due to variability ofloading associated with the inflow.

(1) The use of a reservoir as a mixing devise should beconsidered whenever the inorganic water quality isunacceptable during some periods but where the averagequality falls within the acceptance level. Lake Texoma onthe Red River is an example of a reservoir which modifiesthe quality pattern. Although monthly inflow quality hasequalled 1,950 mg/l chloride concentration, the outflow hasnot exceeded 520 mg/l.

(2) Many materials which enter a reservoir areremoved by settling. This applies not only to incomingsettleable solids, but also to colloidal and dissolvedmaterials which become of settleable size by chemicalprecipitation or by synthesis into biological organisms.Reservoirs are often used to prevent such settleable mate-rial from entering navigable rivers where settleablematerials would interfere with desired uses. However,

e. Reservoirs as stratified systems. Reservoirsbecome stratified if density variations caused by tempera-ture or dissolved solids are sufficiently pronounced toprevent complete mixing. This stratification may behelpful or harmful depending on the outlet works, inflowwater quality, and the operating procedure of the reservoir.

(1) Temperature stratification can be beneficial forcold-water fisheries if the water which enters the reservoirduring the cooler months can also be stored and releasedduring the warmer months. The cooler water releasedduring the warm months can also be valuable as a coolingwater source, can provide for higher oxygen transfer(re-aeration) or slower organic waste oxidation (deoxygen-ation), and can make the water more aesthetically accept-able for water supply and recreational purposes.

(2) Dissolved oxygen stratification usually occurs indensity stratified lakes, particularly during the warmermonths. The phenomenon occurs because oxygen whichhas been introduced into the epilimnion by surfacere-aeration does not transfer through the metalimnion intothe hypolimnion at a rate high enough to satisfy the oxygendemand by dissolved and suspended materials and by thebenthal organisms. Thus, the cool bottom waters which aresometimes desirable may be undesirable from a dissolvedoxygen standpoint unless energy dissipation structures areconstructed to transfer substantial oxygen into the reservoirpool or the reservoir discharge. Mechanical reservoirmixing to equalize temperature and transfer oxygen tolower reservoir levels is one possible tool for managingreservoir water quality.

f. Reservoirs as flow management devices. Reser-voirs may improve water quality by merely permitting themanagement of flow. This management may includemaintenance of minimum flows, blending selective releasesfrom one or more reservoirs to maintain a given streamquality, and the exclusion of a flow from a system bydiversion.

(1) Minimum flow is often maintained in a stream fornavigation, recreation, fish and wildlife, and water rightspurposes. Such flows may also aid in maintaining accept-able water quality.

(2) There is general agreement that water may bestored and selectively released to help reduce natural waterquality problems where source control is not possible, and

EM 1110-2-142031 Oct 97

11-14

also that water should not be stored and released solely to will improve the river quality or poor quality water that isimprove water quality where similar improvement may be to be discharged when it will do a minimum of harm (e.g.,achieved by treatment at the source. The use of a water during high flow). The water quality version of the HEC-5resource to dilute treatable waste materials is regarded as reservoir simulation program (HEC-5Q) is designed tothe misuse of a valuable resource in most cases. perform quality analysis based on a reservoir simulation for

(3) Selective release of water from one or more reser- releases to meet quality objectives (HEC 1986). Sec-voirs may help improve quality at one or more downstream tion III, Chapter 4 of EM 1110-2-1201 describes variouslocations. Such releases may be one of the governing predictive techniques, including numerical and physicalfactors in establishing reservoir management rules. The models.water to be released may either be good quality water that

quantity demands and subsequently determine additional

EM 1110-2-142031 Oct 97

12-1

Chapter 12Conservation Storage Yield

12-1. Introduction

a. Purpose. There are three purposes of this chapter:(1) to provide a descriptive summary of the technicalprocedures used in the hydrologic studies to analyze reser-voirs for conservation purposes; (2) to furnish backgroundinformation concerning the data requirements, advantages,and limitations of the various procedures; and (3) toestablish guidelines which will be helpful in selecting aprocedure, conducting the studies, and evaluating results.

b. Procedures. The procedures presented are gener-ally used to determine the relationship between reservoirstorage capacity and reservoir yield (supply) for a singlereservoir. The procedures may be used to determinestorage requirements for water supply, water quality con-trol, hydroelectric power, navigation, irrigation, and otherconservation purposes. Although the discussions arelimited to single reservoir analysis, many of the principlesare generally applicable to multi-reservoir systems. Chap-ter 4, “Reservoir Systems,” presents concepts regarding theanalysis of a multi-reservoir system.

12-2. Problem Description

a. Determining storage yield relationships. Thedetermination of storage-yield relationships for a reservoirproject is one of the basic hydrologic analyses for reser-voirs. The basic objective can be to determine the reservoiryield given a storage allocation, or find the storage requiredto obtain a desired yield. The determination follows atraditional engineering approach:

(1) Determine the study objectives. This includes theproject purposes, operation goals, and the evaluationcriteria.

(2) Determine the physical and hydrologic constraintsfor the site. This includes the reservoir storage and outflowcapability, as well as the downstream channel system.

(3) Compile the basic data. The basic data includedemands, flow, and losses. Also, the appropriate timeinterval for analysis, which depends on the data and itsvariations in time.

(4) Select the appropriate method, one that meets thestudy objectives and provides reliable information toevaluate results based on accepted criteria.

(5) Perform the analysis, evaluate the results, andpresent the information.

b. Evaluating hydrologic aspects of planning.Many of the methods described in other chapters of thismanual are necessary to develop and provide data toevaluate the hydrologic aspects of reservoir planning,design, and operation. In many cases, the methodsrequired to provide data for a reservoir analysis are morecomplex than the method for the reservoir study itself.However, because the usefulness and validity of the reser-voir analysis are directly dependent upon the accuracy andsoundness of basic data, complex methods can often bejustified to develop the data.

c. General information. This chapter containsinformation on types of procedures, considerations of timeinterval, storage allocation, project purposes, several typesof studies, and a summary of methods to analyze the resultsof reservoir studies. The methods can be characterized assimplified, including sequential and nonsequential analysis,and detailed sequential analysis. Emphasis is given to thesequential routing because:

(1) It is adaptable to study single or multiple reservoirsystems.

(2) With an appropriate time interval, the variations insupply and demand can be directly analyzed.

(3) It gives results that are easily understood andexplained by engineers familiar with basic hydrologicprinciples.

(4) The accuracy and results of the study can beclosely controlled by the engineers performing and super-vising the studies.

(5) It can be used with sparse basic data for prelimi-nary analyses, as well as with detailed data and analyses.

12-3. Study Objectives

a. Establish and consider objectives. Before anymeaningful storage-yield analysis can be made, it is nec-essary to establish and consider the objectives of thehydrologic study. The objectives could range from apreliminary study to a detailed analysis for coordinatingreservoir operation for several purposes. The objectives,together with the available data, will control the degree ofaccuracy required for the study.

b. Determination of storage required. Basically,there are two ways of viewing the storage-yield

EM 1110-2-142031 Oct 97

12-2

relationship. The most common viewpoint involves the c. Simplified method. A simplified method can bedetermination of the storage required at a given site to used if demands for water are relatively simple (constant)supply a given yield. This type of problem is usually or if approximate results are sufficient, as in the case ofencountered in the planning and early design phases of a many preliminary studies. However, it should be empha-water resources development study. sized that the objective of the simplified methods is to

c. Determination of yield. The second viewpoint achieved by detailed sequential analysis. Simplifiedrequires the determination of yield from a given amount of methods consist generally of mass curve and depth durationstorage. This often occurs in the final design phases or in analyses, which are discussed later.re-evaluation of an existing project for a more compre-hensive analysis. Because a higher degree of accuracy is d. Computer models. Computer models havedesirable in such studies, detailed sequential routings are changed the role of the simplified methods because of theusually used. relatively low cost of a detailed sequential routing. Com-

d. Other objectives. Other objectives of a and Conservation Systems (HEC 1982c) provide efficientstorage-yield analysis include the following: determination models of reservoirs, based on the level of data availability.of complementary or competitive aspects of multiple proj- For preliminary studies, minimum reservoir and demandect development, determination of complementary or information are sufficient. The critical and more complexcompetitive aspects of multiple purpose development in a problem is the development of a consistent flow sequence,single project, and analysis of alternative operation rules which is required by all methods of analysis. Simplifiedfor a project or group of projects. Each objective and the methods still have a role in screening studies or as tools tobasis for evaluation dictates implicitly the method which obtain good estimates of input data for the sequentialshould be used in the analysis. routings.

12-4. Types of Procedures

a. Selecting. The procedures used to determine thestorage-yield relationship for a potential dam site may bedivided into either simplified or detailed sequential analy-sis. The selection of the appropriate technical proceduremay be governed by the availability of data, study objec-tives, or budgetary considerations. In general, the simpli-fied techniques are only satisfactory when the studyobjectives are limited to preliminary or screening studies.Detailed methods are usually required when the studyobjectives advance to the feasibility and design phases.

b. Simulation and mathematical programming analy-ses. The detailed sequential methods may be furthersubdivided into simulation analyses and mathematicalprogramming analyses. In simulation analysis, the physicalsystem is simulated by performing a sequential reservoirrouting with specified demands and supply. In this type ofstudy, attempts are made to accurately reproduce thetemporal and spatial variation in streamflow and reservoirstorage in a reservoir-river system. This is accomplishedby accounting for as many significant accretions anddepletions as possible. In mathematical programminganalysis, the objective is to develop a mathematical modelwhich can be used to analyze the physical system withoutnecessarily reproducing detailed factors. The modelusually provides a simulation that will provide a maximum(or minimum) value of the objective function, subject tosystem constraints.

obtain a good estimate of the results which could be

puter programs like HEC-5 Simulation of Flood Control

e. Detailed sequential routing. In the past, detailedsequential routings have been used almost exclusively forthe development of operating plans for existing reservoirsand reservoir systems. However, the advent of the com-prehensive basin planning concept, the growing demand formore efficient utilization of water resources, and theincreasing competition for water among various projectpurposes indicate a need for detailed sequential routings inplanning studies. Also, these complex system studiesprovide an opportunity to use optimization to suggestsystem operations or allocations (ETL 1110-2-336).

f. Mathematical programming. Mathematicalprogramming (optimization) has generally been applied inwater management studies of existing systems. The ques-tions addressed usually deal with obtaining maximum gainfrom available resources, e.g., energy production from ahydropower system. Recent Corps applications include thereview of operation plans for reservoir systems, e.g.,Columbia and Missouri River Systems (HEC 1991d 1991f,and 1992a). These studies utilized the HEC PrescriptiveReservoir Model HEC-PRM (HEC 1991a).

g. Further information. Wurbs (1991) provides areview of modeling and analysis approaches includingsimulation models, yield analysis, stochastic streamflowmodels, impacts of basinwide management on yield, andoptimization techniques.

EM 1110-2-142031 Oct 97

12-3

12-5. Factors Affecting Selection

a. Examining objectives and data availability.Before initiating a storage-yield study, the study objectivesand data availability should be examined in order toascertain: (1) the method best suited for the studyrequirements; (2) the degree of accuracy required to pro-duce results consistent with the study objectives; and(3) the basic data required to obtain the desired accuracyusing the selected method. In preliminary studies, limita-tions in time and scope might dictate the data and methodto be used and the accuracy. More detailed analyses areneeded when a higher degree of accuracy is desired. Atechnical study work plan is very useful in organizing studyobjectives, inventory of available data, and the selection ofgeneral procedures.

b. Availability of data. The availability of basicphysical and hydrologic data will quite frequently be acontrolling factor in determining which of the severaltechnical methods can be used. Obviously, the detailedmethods require more data which may not be available.However, detailed simulation can be performed withlimited system data if the historic flow data are available.The simplified methods require less data, but the reliabilityof the results decreases rapidly as the length of hydrologicrecord decreases. Therefore, it is often desirable tosimulate additional hydrologic data for use with simplifiedmethods. Hydrologic data and data simulation are dis-cussed in Chapter 5.

c. Significant aspects. The study level and availabledata are not the only deciding factors. The study methodsmust capture the significant aspects of the prototype sys-tem. If simplified methods do not utilize data which havemajor influences upon the results, it would be necessary toutilize a more detailed method which accounts forvariations in these data. For example, the NationalHydropower Study (USACE 1979) used flow-durationtechniques for most reservoirs, but used sequential routingfor reservoirs with significant storage.

12-6. Time Interval

a. Selection. The selection of an appropriate timeinterval depends primarily on the type of analysis and thesignificance of the data variation over time. Time intervalsof one month are usually adequate for nonsequential andpreliminary sequential analyses. For more detailed studies,shorter routing intervals will ordinarily be required.Average daily flow data are increasingly used because theyare readily available, and computer speed is sufficient toprocess the data in a reasonable time. Only in exceptionalcases will routing intervals of less than one day be required

for conservation studies. Considerable work is involvedwith shorter intervals, and the effects of time translation,which are usually ignored in conservation routing studies,become important with shorter intervals. Shorter intervalsare necessary and should be used during flood periods orduring periods when daily power fluctuations occur.

b. Using short time intervals. Ordinarily, whenusing short time intervals of one day or less, it is necessaryto obtain adequate definition of the conditions under study.Periods selected for analysis should exhibit criticalcombinations of hydrologic conditions and demandcharacteristics. For example, analysis of hourly powergeneration at a hydroelectric plant under peaking condi-tions might be studied for a one-month period whereextremely low flows could be assumed to coincide withextremely high power demands. As a rule, studies involv-ing short time intervals are supplementary to one or morestudies of longer periods using longer time intervals. Theresults of the long period study are often used to establishinitial conditions such as initial reservoir storage for theselected periods of short-interval analysis.

c. Selecting a routing interval. In sequential con-servation routing studies, the selection of a routing intervalis dependent upon four major factors: (1) the demandschedule that will be utilized in determining the yield;(2) the accuracy required by the study objectives; (3) thedata available for use in the study; and (4) the phaserelationship between periods of high and low demands andhigh and low flow. If the water demand schedule isrelatively uniform, it is ordinarily possible to estimate theamount of storage required for a specified yield by gra-phical analysis using the Rippl diagram or the nonsequen-tial analysis discussed later herein. Demand scheduleswhich show marked seasonal variations usually precludethe use of graphical techniques alone in determining stor-age requirements. This is especially true when the demandis a function which cannot be described in terms of aspecific amount of water, as in the cases of hydroelectricpower and water quality. In order to obtain accurateestimates of storage requirements when the demandschedule is variable, it is necessary to make sequentialrouting studies with routing intervals short enough todelineate important variations in the demand schedule.Simplified techniques may be utilized in obtaining a firstestimate of storage requirements for the detailed sequentialrouting.

d. More accurate results. As a general rule, shorterrouting intervals will provide more accurate results. This isdue to many factors, such as better definition of rela-tionships between inflow and releases, and better estimatesof average reservoir levels for evaporation and power-head

EM 1110-2-142031 Oct 97

12-4

calculations. Average flow for longer routing intervals cities are of utmost importance, which makes floodtends to reduce the characteristic variations of streamflows, reduction the highest priority in a multiple-purpose projectthus producing a “dampened” storage requirement. This during actual operation. During periods of flood opera-will tend to overestimate yield, or underestimate required tions, conservation requirements might be reduced in orderstorage for reservoirs with small storage. Therefore, the to provide the best flood operation. Although this chaptervolume of conservation storage, compared to the average is not concerned directly with flood-control operation orflow in a time period, is an additional consideration in criteria, it is necessary to integrate flood-control constraintsselecting the time interval. For example, if a monthly with the conservation study to ensure that operatinginterval is used and there is no sufficient conservation conditions and reservoir levels for conservation purposesstorage to control the variation of flow within the month, do not interfere with flood-control operation. Prioritiesthe use of average monthly will conceal that fact. among the various conservation purposes vary with locale,

e. Effects on storage requirements. When fluctua- utilization. In multipurpose projects, every effort should betions in streamflows or demands have a significant effect made to develop operation criteria which maximize theon storage requirements, computations should be refined complementary uses for the various conservation purposes.for critical portions of the studies, or shorter routing inter-vals should be used. However, the routing interval shouldnot be shorter than the shortest period for which flow anddemand data are available. Attempts to “manufacture”flow or demand data are usually time consuming and maycreate errors or give a false impression of accuracy unlessreliable information is available for subdivision of basicdata.

f. Nonsequential methods. The selection of the flowinterval for analysis by nonsequential methods is usuallynot as critical as for a sequential analysis. Because thenonsequential analysis is restricted to uniform demands, itdoes not produce results as accurate as those obtained bysequential methods. Therefore, there is very little gain inaccuracy with short intervals. Flow intervals of one monthare usually suitable for nonsequential methods.

12-7. Physical Constraints

Physical constraints which should be considered instorage-yield studies include conservation storage avail-able, minimum pool, outlet capacities, and channel capaci-ties. The addition of hydropower as a purpose will requirethe inclusion of constraints to power generation, e.g.,maximum and minimum head, penstock capacity, andpower capacity. If flood control is to be included as aproject purpose, the maximum conservation storagefeasible at a given site will be affected by the flood-controlanalysis.

12-8. Priorities

In order to determine optimum yield in a multiple-purposeproject, some type of priority system for the various pur-poses must be established. This is necessary when thecompetitive aspects of water use require a firm basis for anoperating decision. Safety of downstream inhabitants and

water rights, and with the need for various types of water

12-9. Storage Limitations

One of the reasons for making sequential conservationrouting studies is to determine the effect of storage limita-tions on yield rates. Simplified yield methods cannotaccount for operational restrictions imposed by storagelimitations in a multiple-purpose project. As shown inFigure 2-2, three primary storage zones, any or all of whichmay exist in a given reservoir project, may generally bedescribed as follows:

• Exclusive capacity, generally for flood control, inthe uppermost storage space in the reservoir.

• Multiple-purpose capacity, typically conservationstorage, immediately below the flood controlstorage.

• Inactive capacity, or dead storage, the loweststorage space in the reservoir.

An additional space, called surcharge, exists between thetop of the flood-control space and the top of the dam.Surcharge storage is required to pass flood waters over thespillway. The boundaries between the storage zones andoperational boundaries within the zones may be fixedthroughout the year, or they may vary from season toseason as shown on Figure 12-1. The varying boundariesusually offer a more flexible operational plan which mayresult in higher yields for all purposes, although an addi-tional element of chance is often introduced when theboundaries are allowed to vary. The purpose of detailedsequential routing studies is to produce an operatingscheme and boundary arrangement which minimizes thechance of failure to satisfy any project purpose whileoptimizing the yield for each purpose. The three storagezones and the effect of varying their boundaries are dis-cussed in the following paragraphs.

EM 1110-2-142031 Oct 97

12-5

Figure 12-1. Seasonally varying storage boundaries

a. Flood-control storage. The inclusion of flood- flood-control storage and conservation storage may becontrol storage in a multiple-purpose project may used. The general procedure is to hold the top of theadversely affect conservation purposes in two ways. First, conservation pool at a low level when conservationstorage space which might otherwise be utilized for con- demands are not critical in order to reserve more storageservation purposes is reserved for flood-control usage. space for flood-control regulation. Then, as the likelihoodSecond, flood-control operations may conflict with conser- of flood occurrence decreases, the top of the conservationvation goals, with a resultant reduction or loss of pool is raised to increase the storage available for conser-conservation benefits. However, detailed planning and vation purposes. Water management criteria are thenanalysis of criteria for flood-control and conservation tested by detailed sequential routing for the period ofoperations can minimize such adverse effects. Even with- recorded streamflow. Several alternative patterns andout dedicated flood storage, conservation projects must be magnitudes of seasonal variations should be evaluated toable to perform during flood events. determine the response of the storage-yield relationship

(1) Where competition between flood-control and of the boundary. A properly designed seasonally-varyingconservation requirements exists, but does not coincide in storage boundary should not reduce the effectiveness oftime, the use of a seasonally varying boundary between flood-control storage to increase the conservation yield.

and the flood-control efficiency to the seasonal variation

EM 1110-2-142031 Oct 97

12-6

(2) Flood-control operation is generally simplified inconservation studies because the routing interval for suchstudies is frequently too long to adequately define theflood-control operation. Nevertheless, flood-control con-straints should be observed insofar as possible. Forexample, channel capacities below the reservoir are con-sidered for release purposes, and storage above the top offlood-control pool is not utilized.

b. Conservation storage. The conservation storagemay be used to regulate minor floods in some multipurposeprojects, as well as to supply water for conservationpurposes. In addition to seasonal variations in its upperboundary between flood control and conservation, thelower conservation storage boundary may also vary sea-sonally. If several conservation purposes of differentpriorities exist, there may be need for a buffer zone in theconservation storage. The seasonal variation in the boun-dary between conservation storage and buffer zone wouldbe determined by the relationship between seasonaldemands for the various purposes.

c. Buffer storage. Buffer storage may be required forone or two reasons. First, it may be used in multipurposeprojects to continue releases for a higher priority purposewhen normal conservation storage has been exhausted bysupplying water for both high and low priority purposes.Second, it may be used in a single-purpose project tocontinue releases at a reduced rate after normalconservation storage has been exhausted by supplyingwater at a higher rate. In either case, the boundary betweenthe normal conservation storage and buffer storage is usedto change the operational criteria. The location of thisboundary and its seasonal variation are important factors indetailed sequential routing because of this change in watermanagement criteria. The amount of buffer storage and,consequently, the location of the seasonally varyingboundary between the buffer zone and the remainder of theconservation storage is usually determined by successiveapproximations in sequential routing studies. However, asimplified procedure, which produces a satisfactoryestimate in cases without seasonally varying boundaries, isdescribed in Section 12-11.

d. Inactive or dead storage. The inactive storagezone is maintained in the reservoir for several purposes,such as a reserve for sedimentation or for fish and wildlifehabitat. As a rule, the reservoir may not be drawn belowthe top of the inactive storage. Although it may bepossible to vary the top of the reserve pool as shown inFigure 12-1, it is seldom practical to do so. This couldreflect the desire to maintain a higher recreation poolduring the summer.

12-10. Effects of Conservation and OtherPurposes

As previously indicated, the seasonal variation of demandschedules may assume an important role in the determina-tion of required yield. The effect of seasonal variation ismost pronounced when the varying demand is large withrespect to other demands, as is often the case whenhydroelectric power or irrigation is a large demand item.The quantity of yield from a specified storage may beoverestimated by as much as 30 percent when a uniformyield rate is used in lieu of a known variable yield rate.Also, variable demand schedules often complicate theanalysis of reservoir yield to the extent that it is impossibleto accurately estimate the maximum yield or the optimumoperation by approximate methods. Because detailedsequential routing is particularly adaptable to the use ofvariable demand schedules, every effort should be made toincorporate all known demand data into the criteria forrouting. Thus, successive trials using detailed sequentialanalysis must often be used to determine maximum yield.Computer programs such as HEC-5 provide yielddetermination for reservoirs by performing the successivesequential routing until a firm yield is determined. Firmyield is the amount of water available for a specific use ona dependable basis during the life of a project. The projectpurposes which often require analysis of seasonalvariations in demand are discussed in more detail in thefollowing subparagraphs.

a. Low-flow regulation. The operation of a reservoirfor low flow regulation at a downstream control point isdifficult to evaluate without a detailed sequential routing,because the operation is highly dependent upon the flowswhich occur between the reservoir and the control point,called intervening local flow. As these flows can varysignificantly, a yield based on long period average inter-vening flows can be subject to considerable error. Adetailed sequential routing, in which allowance is made forvariations of intervening flows within the routing interval,produces a more reliable estimate of storage requirementsfor a specified yield and reduces the chance ofoverestimating a firm yield. Ordinarily, the yield and thecorresponding operation of a reservoir for low-flowregulation are determined by detailed sequential routing ofthe critical period and several other periods of low flow.The entire period of recorded streamflow may not berequired, unless summary-type information is needed forfunctions such as power.

b. Diversion and return flows. The analysis of yieldfor diversions is complicated by the fact that diversionrequirements may vary from year to year as well as from

EM 1110-2-142031 Oct 97

12-7

season to season. Furthermore, the diversion requirements (2) Determination of firm power or firm energy ismay be stated as a function of the natural flow and water usually based on sequential routings over the criticalrights rather than as a fixed amount. In addition, diversion period. The critical period must consider the combinationamounts may often be reduced or eliminated when storage of power demand and critical hydrologic conditions.in the reservoir reaches a certain critical low value. When Various operational plans are used in an attempt to maxi-any one of these three items is important to a given mize power output while meeting necessary commitmentsreservoir analysis, detailed sequential analysis for the entire for other project purposes. When the optimum output isperiod of flow record should be made to determine achieved, a water management guide curve can be devel-accurately the yield and the water management criteria. oped. The curve is based on the power output itself and onCoordination of the water management criteria for other the plan of operation followed to obtain the maximumpurposes with the diversion requirements may also be output. Critical period analysis and curve development areachieved with the detailed sequential analysis results. described in Section 12-11. Additional sequential routings

c. Water quality control. Inclusion of water quality rule curve developed in the critical period studies. Thesecontrol and management as a project purpose almost routings are used to coordinate power production withalways dictates that sequential routing studies be used to flood-control operation and to determine the averageevaluate project performance. Practically every variable annual potential energy available from the project.under consideration in a water quality study will varyseasonally. Following are the variables which must be (3) In areas where hydroelectric power is used pri-considered in a water quality study: (1) variation in quality marily for peaking purposes, it is important that storageof the inflow, (2) subsequent change in quality of the requirements be defined as accurately as possible becausereservoir waters due to inflow quality and evaporation, the available head during a period of peak demand is(3) variation in quality of natural streamflow entering the required to determine the peaking capability of the project.stream between the reservoir and the control station, An error in storage requirements, on the other hand, can(4) variation in effluent from treatment plants and storm adversely affect the head with a resultant loss of peakingdrainage outflow between the reservoir and the control capability.station, and (5) variation in quality requirements at thecontrol station. Accurate evaluation of project perfor- (4) Tailwater elevations are also of considerablemance must consider all of these variations which pertain importance in power studies because of the effect of headto water quality control. Additionally, there are several on power output. Several factors which may adverselyquality parameters which may require study, and each affect the tailwater elevation at a reservoir are constructionparameter introduces additional variations which should be of a reregulation reservoir below the project underevaluated. For example, if temperature is an important consideration, high pool elevations at a project immediatelyparameter, the level of the reservoir from which water is downstream from the project under consideration, andreleased should be considered in addition to the above backwater effects from another stream if the project is nearvariables. Likewise, if oxygen content is important, the the confluence of two streams. If any of these conditionseffects of release through power units versus release exist, the resultant tailwater conditions should be carefullythrough conduits must be evaluated. evaluated. For projects in which peaking operation is

d. Hydroelectric power generation. be used to determine reservoir releases for the sequential

(1) If hydroelectric power is included as a project is defined as the tailwater elevation resulting frompurpose, detailed sequential routings are necessary to sustained generation at or near the plant's rated capacitydevelop water management criteria, to coordinate power which represents the condition under which the project isproduction with other project purposes, and to determine expected to operate most of the time. Although in realitythe project's power potential. As a rule, simplified methods the peaking operation tailwater would fluctuateare usable for power projects only for preliminary or considerably, the use of block-loading tailwater elevationscreening studies, reservoirs with very little power storage, ensures a conservative estimate of storage requirementsor when energy is a by-product to other operations. Flow- and available head.duration analysis, described in Chapter 11, is typicallyapplied in these situations. Power production is a function (5) Reversible pump-turbines have enhanced theof both head and flow, which requires a detailed sequential feasibility of the pumped-storage type of hydroelectricstudy when the conservation storage is relatively large and development. Pumped storage includes reversible pump-the head can be expected to fluctuate significantly. turbines in the powerhouse along with conventional

for the entire period of flow record are then made using the

anticipated, an assumed “block-loading” tailwater should

reservoir routing. The “block-loading” tailwater elevation

EM 1110-2-142031 Oct 97

12-8

Figure 12-2. Storage determination using a sequential mass curve

generating units, and an afterbay is constructed below the a. Sequential mass curve. The most commonlymain dam to retain water for pumping during nongenerat- used simplified sequential method is the sequential massing periods. Sequential routing studies are required for curve analysis, sometimes referred to as the Rippl Method.analyses of this type because of the need to define storage This method produces a graphical estimate of the storagerequirements in the afterbay, pumping requirements and required to produce a given yield, assuming that thecharacteristics, and the extent to which plan should be seasonal variations in demand are not significant enough todeveloped. Many of the existing and proposed pumped- prohibit the use of a uniform draft (demand) rate. Thestorage projects in the United States, however, are single sequential mass curve is constructed by accumulatingpurpose projects which do not have conventional units and inflows to a reservoir site throughout the period of recordoften utilize off-channel forebays. and plotting these accumulated inflows versus the sequential

12-11. Simplified Methods

If demands for water are relatively constant or if approxi- 38,000 m /year, is represented by the slope of a straight line.mate results are sufficient, as in the case of many prelimi- Straight lines are then constructed parallel to the desirednary studies, a simplified method can be used to save time yield rate and tangent to the mass curve at each low pointand effort. The use of simplified techniques which do not (line ABC) and at the preceding high point that gives theconsider sequential variations in streamflow or demand are highest tangent (line DEF). The vertical distance betweengenerally limited to screening studies or developing first these two lines (line BE) represents the storage required toestimates of storage or yield. The following procedures provide the desired yield during the time period between thewill generally produce satisfactory results and continue to two tangent points (points D and B). The maximum verticalhave a role in storage-yield determinations. difference in the period is the required

time period as illustrated in Figure 12-2.

(1) The desired yield rate, in this example3

I O = � S

EM 1110-2-142031 Oct 97

12-9

(12-1)

storage to meet the desired yield, during the given flow (4) The nonsequential mass curve (Figure 12-4) issequence. developed by selecting the desired volume-duration curve

(2) The critical period is the duration of time from The desired yield is then used to determine the storagepoint D, when conservation storage drawdown begins, topoint F, when the reservoir conservation storage fills. Thecritical drawdown is from point D to point B, while duringthe time from B to F the reservoir would be refilling.

(3) The sequential mass curve method does not indi-cate the relative frequency of a shortage. However, byusing nonsequential methods, a curve of yield versusshortage frequency can be determined.

b. Nonsequential mass curve. Several nonsequentialmethods can be used to develop a relation for storage yieldversus shortage frequency. The application of thisprocedure is limited, however, to water supply demandsthat are uniform in time. These methods involve thedevelopment of probability relations for varying durationsof streamflow. The historical flows, supplemented bysimulated flows where needed, are used to determinefrequency tables of independent low-flow events for sev-eral durations. A series of low-flow events for a particularduration is selected by computing and arranging in order ofmagnitude, the independent minimum-flow rates for thatduration for the period of record.

(1) After the frequency tables of independent low-flowevents are computed for various durations, low-flowfrequency curves are obtained by plotting the average flowon log-probability paper. Chapter 4 of EM 1110-2-1415describes the procedure and presents an example table andfrequency plot.

(2) Care must be exercised in the interpretation of thelow-flow curves because the abscissa is “nonexceedancefrequency per 100-years,” or the number of events within100-years that have a flow equal to or less than the indi-cated flow. Thus, when low-flow durations in excess ofone year are evaluated, the terminology cannot be usedinterchangeably with probability. For instance, during a100-year period, the maximum number of independentevents of 120 months (10-years) duration is 10. Therefore,the 120-month duration curve cannot cross the value of 10on the “nonexceedance frequency per 100-year” scale.

(3) Minimum runoff-duration curves for variousfrequencies, as shown in Figure 12-3 are obtained byplotting points from the low-flow frequency curves onlogarithmic paper. The flow rates are converted to volumes(millions of cubic meters in this example). The logarithmicscales simply permit more accurate interpolation betweendurations represented by the frequency curves.

from Figure 12-3 and plotting this curve on arithmetic grid.

requirement for the reservoir. The storage requirement isdetermined by drawing a straight line, with slope equivalentto the required gross yield, and by plotting this line tangentto the mass curve. The absolute value of the negativevertical intercept represents the storage requirement. Theapplication of this procedure is severely limited everywherein the case of seasonal variations in runoff and yieldrequirements because the nonsequential mass curve doesnot reflect the seasonal variation in streamflows, and thetangent line does not reflect seasonal variations in demand.However, the method does provide an estimate of yieldreliability.

c. Evaporation losses. Another disadvantage ofthese simplified types of storage-yield analysis is theinability to evaluate evaporation losses accurately. Thismay not be critical in humid areas where net evaporation(lake evaporation minus pre-project evapotranspiration) isrelatively small, but it can cause large errors in studies forarid regions. Also, these procedures do not permit con-sideration of seasonal variations in requirements, systemnonlinearities, conflicting and complementing servicerequirements, and several other factors.

12-12. Detailed Sequential Analysis

a. Sequential analysis. Sequential analysis is cur-rently the most accepted method of determining reservoirstorage requirements. Many simplified methods have givenway to the more detailed computer simulation approaches.In many instances, the computer solution provides moreaccurate answers at a lower cost than the simple handsolutions.

b. Accounting for reservoir water. Sequential analy-sis applies the principle of conservation of mass to accountfor the water in a reservoir. The fundamental relationshipused in the routing can be defined by:

where

I = total inflow during the time period, in volumeunits

O = total outflow during the time period, in volumeunits

�S = change in storage during the time period, involume units

St = St&1 % It & Ot

EM 1110-2-142031 Oct 97

12-10

Figure 12-3. Minimum runoff-duration curves

Figure 12-4. Nonsequential mass curve fromFigure 12-3

(12-2)

The inflow and outflow terms include all types of inflowand outflow. The inflows should include natural stream-flow, releases from upstream reservoirs, local inflow to thereservoir, precipitation falling on the reservoir surface(sometimes included in computation of net evaporation),and diversions into the reservoir. Outflows consist ofreservoir releases plus evaporation losses, leakage, anddiversions out of the reservoir. Sequential routing pro-vided the framework for accounting for all water in thesystem. The application can be as detailed as required.The development of the required data constitutes the majoreffort in most sequential routing studies.

c. Sequential routing. Sequential routing uses arepetitive solution of Equation 12-1 in the form of:

where

S = storage at the end of time t, volume unitst

S = storage at the end of time t-1, volume unitst-1

EM 1110-2-142031 Oct 97

12-11

I = average inflow during time step �t, converted to period. After the firm yield is estimated, the program willt

volume units perform a period of record simulation to ensure that the

O = average outflow during time step �t, converted period of record simulations can be performed in onet

to volume units computer run, based on user input specifications.

The primary input includes reservoir storage capacity andallocation, requirements, losses, flow at all model locationsfor the simulation period, and system connectivity andconstraints. The primary output for the reservoir is theaverage reservoir release for each time step and theresulting reservoir storage at the end of each time step.The releases are made to meet specified requirements,subject to all specified constraints such as storage alloca-tion and maximum release capability. Downstreamaccounting of flows adds reservoir releases and subtractsdiversions and losses to the local downstream flows tocompute regulated flow at desired locations. If a short timeinterval is used, the flow travel time must be considered.Channel routing is usually done with hydrologic routingmethods.

d. Multipurpose reservoir routings. The HEC-5Simulation of Flood Control and Conservation Systems(HEC 1982c) computer program performs multipurposereservoir routings for reservoir systems providing forservices at the reservoirs and downstream control points.Releases from a reservoir are determined by the specifiedrequirements for project purposes. Reservoir releases maybe controlled at the dam site by hydroelectric powerrequirements, downstream control for flow, diversion,water rights, or quality. Additional diversions may bemade directly from the reservoir.

(1) The program operates to meet the downstream flowrequirement, considering available supplies and supplementflow from the intervening area. The storage allocation andmost demands can be defined as constant, monthly varying,or seasonally varying. Historic simulations can beperformed with period-by-period demands for low flow andhydropower.

(2) Computer program HEC-5 has a firm yield routinecalled optimization of conservation storage in the programuser's manual. The routine can either determine therequired storage to meet a specified demand or themaximum reservoir yield that can be obtained from aspecified amount of storage. While designed for a singlereservoir, it can use up to six reservoirs in a single run,provided the reservoirs operate independently. Optimiza-tion can be accomplished on monthly firm energy require-ments, minimum monthly flow, monthly diversions, or allof the conservation purposes. The routine can estimate thecritical period and make a firm yield estimate based on that

firm yield can be met. Several cycles of critical period and

12-13. Effects of Water Deficiencies

a. Water storages. Absolute guarantees of wateryield are usually not practical, and the designer shouldtherefore provide estimates of shortages that couldreasonably develop in supplying the demands with avail-able storage. If nonsequential procedures have been used,information on future shortages is limited to the probabilityor frequency of occurrence, and the duration or severity ofshortages will not be known. In using the Rippl Method,the computations are based on just meeting the demand;therefore, no shortages are allowed during the period ofanalysis. The result gives no information on the shortagesthat might be expected in the future. Only in the detailedsequential analysis procedure is adequate information onexpected future shortages obtainable.

b. Amount and duration of water shortage. Theamount and duration of shortage that can be tolerated inserving various project purposes can greatly influence theamount of storage required to produce a firm yield. Thesetolerances vary a great deal for different project purposesand should be analyzed carefully in reservoir design. Also,changes in reservoir operation should be considered tomeet needs during drought (HEC 1990a).

c. Intolerable shortages. Shortages are generallyconsidered to be intolerable for purposes such as drinkingwater. However, some reduction in the quantity of muni-cipal and industrial water required can be tolerated withoutserious economic effects by reducing some of the lessimportant uses of water such as lawn watering, car wash-ing, etc. Shortages greater than 10 percent may causeserious hardship. Most designs of reservoir storage formunicipal and industrial water supply are based on sup-plying the firm yield during the most critical drought ofrecord. Typically, drought contingency plans are devel-oped to meet essential demands during drought conditionsthat may be more severe than the historic critical period.ETL 1110-2-335 provides guidance for developing andupdating plans.

d. Irrigation shortages. For irrigation water supply,shortages are usually acceptable under some conditions.Often the desired quantity can be reduced considerablyduring the less critical parts of the growing season withoutgreat crop loss. Also, if there is a reliable forecast of adrought, the irrigator may be able to switch to a crop

SI =

100 (

i=N

i=1

SA

DA

2

N

EM 1110-2-142031 Oct 97

12-12

(12-3)

having less water requirements or use groundwater to makeup the deficit. Shortages of 10 percent usually havenegligible economic effect, whereas shortages as large as50 percent are usually disastrous.

e. Water supply for navigation and low flow aug-mentation. In designing a reservoir to supply water fornavigation or low-flow augmentation, the amount andduration of shortages are usually much more importantthan the frequency of the shortages. Small shortages mightonly require rescheduling of fully-loaded vessels, whereas,large shortages might stop traffic altogether. The samething is true for such purposes as fish and wildlife whereone large shortage during the spawning season, forexample, could have serious economic effects for years tocome.

f. Effects of shortages. Each project purpose shouldbe analyzed carefully to determine what the effects ofshortages will be. In many cases, this will be the criterionthat determined the ultimate amount of reservoir storageneeded for water supply and low-flow regulation.

12-14. Shortage Index

a. Definition. A general approach to shortagedefinition is to use a shortage index. The shortage index isequal to the sum of the squares of the annual shortagesover a 100-year period when each annual shortage isexpressed as a ratio to the annual requirements, as shownbelow:

where

SI = shortage index

N = number of years in routing study

S = annual shortage (annual demand volume minusA

annual volume supplied)

D = annual demand volumeA

This shortage index reflects the observation that economicand social effects of shortages are roughly proportional tothe square of the degree of shortage. For example, ashortage of 40 percent is assumed to be four times assevere as a shortage of 20 percent. Similarly, as illustratedin Table 12-1, shortages of 50 percent during 4 out of

Table 12-1Illustration of Shortage Index

Shortage No. of Annual Short- Annual ShortageIndex ages per 100 Years (S /D ) In %A A

1.00 100 10

1.00 25 20

1.00 4 50

0.25 25 10

0.25 1 50

100 years are assumed four times as severe as shortages of10 percent during 25 out of 100 years.

The shortage index has considerable merit over shortagefrequency alone as a measure of severity because shortagefrequency considers neither magnitude nor duration. Theshortage index can be multiplied by a constant to obtain arough estimate of associated damages.

b. Additional criteria needed. There is a definiteneed for additional criteria delineating shortage acceptabil-ity for various services under different conditions. Thesecriteria should be based on social and economic costs ofshortages in each individual project study, or certain stan-dards could be established for the various services andconditions. Such criteria should account for degree ofshortage as well as expected frequency of shortage.

12-15. General Study Procedures

a. Water supply. After alternative plans for one ormore water supply reservoirs have been established, thefollowing steps can be followed in performing hydrologicstudies required for each plan:

(1) Obtain all available daily and monthly streamflowrecords that can be used to estimate historical flows at eachreservoir and diversion or control point. Compute monthlyflows and adjust as necessary for future conditions at eachpertinent location. A review of hydrologic data ispresented in Chapter 5.

(2) Obtain area-elevation data on each reservoir site tobe studied and compute storage capacity curves.Determine maximum practical reservoir stage from phys-iographic and cultural limitations.

(3) Estimate monthly evapotranspiration losses fromeach site and monthly lake evaporation that is likely tooccur if the reservoir is built.

(4) Determine seasonal patterns of demands and totalannual requirements for all project purposes, if applicable,

EM 1110-2-142031 Oct 97

12-13

as a function of future time. Synthesize stochastic varia- and knowing the type of project and the approximatetions in demands, if significant. storage usable for power production, determine the histor-

(5) Establish a tentative plan of operation, considering hydrologic data.flood control and reservoir sedimentation as well as con-servation requirements, and perform an operation study (4) An estimate of potential hydroelectric energy forbased on runoff during the critical period of record. The the assumed critical hydro-period is made using Equa-HEC-5 Simulation of Flood Control and Conservation tion 11-2. If the energy calculated from this equation is forSystems computer program can be used for this purpose. a period other than the basic marketing contract period

(6) Revise the plan of operation, including sizes of critical hydro-period should be converted to a firm orvarious facilities, as necessary to improve accomplishments minimum quantity for the contract period (minimum annualand perform a new operation study. Repeat this process or annual firm in the case of a calendar year).until a near-optimum plan of development is obtained.

(7) Depending on the degree of refinement justified in hydroelectric energy and peaking capacity is a complexthe particular study, test this plan of development using the function of the head, the streamflow, the storage, andentire period of estimated historical inflows and as many operation for all other purposes, the energy estimatesequences of synthetic streamflows and demands as might obtained in step (4) is only an approximation. Althoughbe appropriate. Methods for developing synthetic flow this approximation is useful for planning purposes it shouldsequences are presented in Chapter 12 of the Hydrologic be verified by simulating the operation of the project for allFrequency manual (EM 1110-2-1415). authorized purposes by means of a sequential routing

(8) Modify the plan of development to balance yields analyzing sequential routing studies.and shortages for the maximum overall accomplishment ofall project objectives. (6) From the results of detailed sequential routing

b. Hydroelectric power. The study procedure for units and power-related facilities of the project should beplanning, designing, and operating hydroelectric develop- developed. The design head and design output of thements can be summarized as follows: generating units, approximate powerhouse dimensions,

(1) From an assessment of the need for power gener- dimensions of the project depend on the power installation.ation facilities, obtain information concerning the feasibil-ity and utility of various types of hydroelectric projects.This assessment could be made as part of the overall studyfor a given project or system, or it could be available froma national, regional, or local power authority.

(2) From a review of the physical characteristics of aproposed site and a review of other project purposes, ifany, develop an estimate of the approximate amount ofspace that will be available for either sole- or joint-usepower storage. This determination and the needs devel-oped under step (1) will determine whether the project willbe a storage, run-of-river, or pumped-storage power projectand whether it will be operated to supply demands forpeaking or for baseload generation.

(3) Using information concerning seasonal variation inpower demands obtained from the assessment of needs,

ical critical hydro-period by review of the historical

(usually a calendar year), the potential energy during the

(5) Because the ability of a project to produce

study. Chapter 11 provides methods for performing and

studies, the data necessary for designing power-generating

approximate sizes of water passages, and other physical

(7) Operation rules for the project must be developedbefore construction is completed. These rules are devel-oped and verified through sequential routing studies thatincorporate all of the factors known to affect the project'soperation. For many multipurpose projects, these operationrules are relatively complex and require the use ofcomputerized simulation models to facilitate the computa-tions involved in the sequential routing studies.

(8) If the project is to be incorporated into an existingsystem or if the project is part of a planned system, systemoperation rules must be developed to define the role of theproject in supplying energy and water to satisfy the systemdemands. These rules are also developed and tested usingsequential routing studies. Sequential routing studies forplanning or operating hydroelectric power systems are bestaccomplished using a computer program such as HEC-5.

EM 1110-2-142031 Oct 97

13-1

Figure 13-1. Conceptual deposition in deep reservoirs

Chapter 13 Reservoir Sedimentation

13-1. Introduction

AThe ultimate destiny of all reservoirs is to be filled withsediment,@ (Linsley et al. 1992). The question is how longwill it take? Also, as the sediment accumulates with time,will it adversely affect water control goals?

a. Transport capacity. A reservoir changes thehydraulics of flow by forcing the energy gradient toapproach zero. This results in a loss of transport capacitywith the resulting deposition. The smaller the particles,the farther they will move into the reservoir before depos-iting. Some may even pass the dam. Deep reservoirs arenot fully mixed and are conducive to the formation ofdensity currents.

b. Sediment deposits. The obvious consequence ofsediment deposits is a depletion in reservoir storagecapacity. Figure 13-1 illustrates components of sedimentdeposition in a deep reservoir. The volume of sedimentmaterial in the delta and the main reservoir depends on theinflowing water and sediment, reservoir geometry, projectoperation and life among other things. The delta willcontinue to develop, with time, and the reservoir willeventually fill with sediment.

13-2. Reservoir Deposition

a. Total available sediment. The first step is toestimate the total sediment that will be available for depo-sition during the design life of the project. Required datainclude design life of the reservoir, reservoir capacity,water and sediment yield from the watershed, the compo-sition of the sediment material, and the unit weight ofsediment deposits. With this information, the trap effi-ciency can be determined.

b. Trap efficiency. Trap efficiency is the percent ofinflowing sediment that remains in the reservoir. Someproportion of the inflowing sediment leaves the reservoirthrough the outlet works. The proportion remaining in thereservoir is typically estimated based on the trap effi-ciency. Trap efficiency is described in Section 3-7(a) ofEM 1110-2-4000, and the calculations are described in anappendix therein. The efficiency is primarily dependenton the detention time, with the deposition increasing as thetime in storage increases.

c. Existing reservoirs. Existing reservoirs areroutinely surveyed to determine sediment deposition, andresulting loss of storage. Section 5-30 and Appendix K ofEM 1110-2-4000 describe the Corps program. Thishistoric deposition data can be useful for checking com-puted estimates. ASediment Deposition in U.S. Reservoirs(Summary of Data Reported 1981-85)@ provides

EM 1110-2-142031 Oct 97

13-2

data on reservoir locations, drainage areas, survey dates, primarily composed of suspended sediments brought fromreservoir storage capacities, ratios of reservoir capacities to upstream by density currents. The region is called theaverage annual inflows, specific weights (dry) of sediment reservoir dead storage and generally does not affect thedeposits, and average annual sediment-accumulation rates storage capacity. Some of the finest material may not settle(U.S. Geological Survey 1992). Reservoirs are grouped by out and will pass through the dam. In order to calculate thedrainage basins. volume of material which will deposit as a function of

13-3. Distribution of Sediment Deposits in theReservoir

The planning or design of a reservoir requires an analysisto determine how sediment deposits will be distributed inthe reservoir. This is a difficult aspect of reservoir sedi-mentation because of the complex interaction betweenhydraulics of flow, reservoir operating policy, inflowingsediment load, and changes in the reservoir bed elevation.The traditional approach to analyzing the distribution ofdeposits has relied on empirical methods, all of whichrequire a great deal of simplification from the actualphysical problem.

a. Main channel deposition. Conceptually, deposi-tion starts in the main channel. As flow enters a reservoir,the main channel fills at the upstream end until theelevation is at or above the former overbank elevations oneither side. Flow then shifts laterally to one side or theother, but present theory does not predict the exact loca-tion. During periods of high water elevation, depositionwill move upstream. As the reservoir is drawn down, achannel is cut into the delta deposits and subsequentdeposition moves material farther into the reservoir. Thelateral location of the channel may shift from year to year,but the hydraulic characteristics will be similar to those ofthe natural channel existing prior to impounding thereservoir. Vegetation will cover the exposed delta depositsand thus attract additional deposition until the delta takeson characteristics of a floodplain.

b. Sediment diameters. The diameter of sedimentparticles commonly transported by streams ranges over fivelog cycles. Generally, the coarse material will settle first inthe outer reaches of the reservoir followed by progressivelyfiner fractions farther down toward the reservoir dam.Based on this depositional pattern, the reservoir is dividedinto three distinct regions: top-set, fore-set, and bottom-setbeds. The top-set bed is located in the upper part of thereservoir and is largely composed of coarse material or bedload. While it may have a small effect on the reservoirstorage capacity, it could increase upstream stages. Thefore-set region represents the live storage capacity of thereservoir and comprises the wash load. The bottom-setregion is located immediately upstream of the dam and is

distance, grain size must be included as well as themagnitude of the water discharge and the operating policyof the reservoir.

c. Reservoir shape. Reservoir shape is an importantfactor in calculating the deposition profile. For example,flow entering a wide reservoir spreads out, thus reducingtransport capacity, but the path of expanding flow does notnecessarily follow the reservoir boundaries. It becomes a2-dimensional problem to calculate the flow distributionacross the reservoir in order to approximate transportcapacity and, therefore, the resulting deposition pattern.On the other hand, flow entering a narrow reservoir has amore uniform distribution across the section resulting inhydraulic conditions that are better approximated by 1-dimensional hydraulic theory.

d. Flood waves. Flood waves attenuate upon enter-ing a reservoir. Therefore, their sediment transport capac-ity decreases from two considerations: (1) a decrease invelocity due to the increase in flow area and (2) a decreasein velocity due to a decrease in water discharge resultingfrom reservoir storage. As reservoir storage is depleted bythe sediment deposits in the delta, the impact of attenuationon transport capacity diminishes. The resulting configura-tion, therefore, is assumed to depend upon the firstconsideration, whereas, the time for delta development isinfluenced somewhat by the second consideration.

e. Flood-pool index method. If flood control is aproject purpose, the next level of detail in reservoir sedi-mentation studies is to divide the total volume of predicteddeposits into that volume settling into the flood-controlpool and that volume settling in the remainder of thereservoir. The flood-pool index method requires the depthof flood-control pool, depth of reservoir, and the percent oftime the reservoir water level is at or above the bottom ofthe flood-control pool. Based on the index, the percent ofsediment trapped in the flood-control pool is estimated by ageneral empirical relationship. Appendix H of EM 1110-2-4000 describes the index method and provides severalother methods for estimating the distribution of sedimentdeposits in reservoirs. Chapter 5, Section IV, EM 1110-2-4000, provides an overview of levels of sedimentationstudies and methods of analysis.

EM 1110-2-142031 Oct 97

PART 4

HYDROLOGIC ENGINEERING STUDIESFOR RESERVOIRS

EM 1110-2-142031 Oct 97

14-1

Chapter 14Spillways and Outlet Works

14-1. Function of Spillways and Outlet Works

Spillways and outlet works are necessary to providecapability to release an adequate rate of water from thereservoir to satisfy dam safety and water control regulationof the project. Sections 4-2 and 4-3 in EM 1110-2-3600provide general descriptions of types and operationrequirements for spillways and outlet works, respectively.

a. Spillway adequacy. While the outflow capabilitymust be provided throughout the operational range of thereservoir, the focus of hydrologic studies is usually on thehigh flows and spillway adequacy. Dam failures have beencaused by improperly designed spillways or by insufficientspillway capacity. Ample capacity is of great importancefor earthfill and rockfill dams, which are likely to bedestroyed if overtopped; whereas concrete dams may beable to withstand moderate overtopping.

b. Spillway classification. Spillways are ordinarilyclassified according to their most prominent feature, eitheras it pertains to their shape, location, or discharge channel.Spillways are often referred to as controlled oruncontrolled, depending on whether they are gated orungated. EM 1110-2-1603 describes a variety of spillwaytypes and provides hydraulic principles, design criteria, andresults from laboratory and prototype tests.

c. Outlet works. Outlet works serve to regulate orrelease water impounded by a dam. It may releaseincoming flows at a reduced rate, as in the case of adetention dam; divert inflows into canals or pipelines, as inthe case of a diversion dam; or release stored-water at suchrates as may be dictated by downstream needs, evacuationconsiderations, or a combination of multiple-purposerequirements.

d. Outlet structure classification. Outlet structurescan be classified according to their purpose, their physicaland structural arrangement, or their hydraulic operation.EM 1110-2-1602 provides information on basic hydrau-lics, conduits for concrete dams, and conduits for earthdams with emphases on flood-control projects. Appen-dix IV of EM 1110-2-1602 provides an illustrative exampleof the computation of a discharge rating for outlet works.

e. Low level outlets. Low level outlets are providedto maintain downstream flows for all levels of the reservoir

operational pool. The outlets may also serve to empty thereservoir to permit inspection, to make needed repairs, or tomaintain the upstream face of the dam or other structuresnormally inundated.

f. Outlets as flood-control regulators. Outlet worksmay act as a flood-control regulator to release waterstemporarily stored in flood control storage space or toevacuate storage in anticipation of flood inflows. In thiscase, the outflow capacity should be able to release channelcapacity, or higher. The flood control storage must beevacuated as rapidly as safely possible, in order to maintainflood reduction capability.

14-2. Spillway Design Flood

a. Spillway design flood analyses. Spillway designflood (SDF) analyses are performed to evaluate the ade-quacy of an existing spillway or to size a spillway. For amajor project, the conservative practice in the UnitedStates is to base the spillway design flood on the probablemaximum precipitation (PMP). The PMP is based on themaximum conceivable combination of unfavorable meteo-rological events. While a frequency is not normallyassigned. a committee of ASCE has suggested that thePMP is perhaps equivalent to a return period of10,000 years.

b. Probable maximum flood. The PMF inflowhydrograph is developed by centering the PMP over thewatershed to produce a maximum flood response. The unithydrograph approach, described in Chapter 7 of thismanual, is usually applied. Section 13-5 of EM 1110-2-1417 contains information on PMP determination andcomputation of the PMF.

c. Flood hydrographs. The inflow design floodhydrographs are usually for rainfall floods. Normally, suchfloods will have the highest peak flows but not always thelargest volumes. When spillways of small capacities inrelation to these inflow design flood peaks are considered,precautions must be taken to ensure that the spillwaycapacity will be sufficient to evacuate storage so that thedam will not be overtopped by a recurrent storm, andprevent the flood storage from being kept partially full by aprolonged runoff whose peak, although less than the inflowdesign flood, exceeds the spillway capacity. To meet theserequirements, the minimum spillway capacity should be inaccord with the following general criteria (Hoffman 1977):

(1) In the case of snow-fed perennial streams, thespillway capacity should never be less than the peak dis-charge of record that has resulted from snowmelt runoff.

EM 1110-2-142031 Oct 97

14-2

(2) The spillway capacity should provide for the spillway crest. The peak outflow will occur at maximumevacuation of sufficient surcharge storage space so that in pool elevation, which should always be less, to somerouting a succeeding flood, the maximum water surface degree, than the peak inflow.does not exceed that obtained by routing the inflow designflood. In general, the recurrent storm is assumed to begin 4 b. Gated spillway. With a gated spillway, thedays after the time of peak outflow obtained in routing the normal operating level is usually near the top of the gates,inflow design flood. although at times it may be drawn below this level by other

(3) In regions having an annual rainfall of 40 in. or available storage and head, while at the same time limitingmore, the time interval to the beginning of the recurrent backwater damages by providing a high initial dischargestorm in criterion (2) should be reduced to 2 days. capacity. In routing the spillway design flood, an initial

(4) In regions having an annual rainfall of 20 in. or assumed. Operating rules for spillway gates must be basedless, the time interval to the beginning of the recurrent on careful study to avoid releasing discharges that would bestorm in criterion (2) may be increased to 7 days. greater than would occur under natural conditions before

14-3. Area and Capacity of the Reservoir

a. Reservoir capacity and operations. Dam designsand reservoir operating criteria are related to the reservoircapacity and anticipated reservoir operations. The reser-voir capacity and reservoir operations are used to properlysize the spillway and outlet works. The reservoir capacityis a major factor in flood routings and may determine thesize and crest elevation of the spillway. The reservoiroperation and reservoir capacity allocations will determinethe location and size of outlet works for the controlledrelease of water for downstream requirements and floodcontrol.

b. Area-capacity tables. Reservoir area-capacitytables should be prepared before the final designs andspecifications are completed. These area-capacity tablesshould be based on the best available topographic data andshould be the final design for administrative purposes untilsuperseded by a reservoir resurvey. To ensure uniformreporting of data for design and construction, standarddesignations of water surface elevations and reservoircapacity allocations should be used.

14-4. Routing the Spillway Design Flood

a. Discharge facilities. The facilities available fordischarging inflow from the spillway design flood dependon the type and design of the dam and its proposed use. Asingle dam installation may have two or more of thefollowing discharge facilities: uncontrolled overflowspillway, gated overflow spillway, regulating outlet, andpower plant. With a reservoir full to the spillway crest atthe beginning of the design flood, uncontrolled dischargewill begin at once. Surcharge storage is created when theoutflow capacity is less than the inflow and the excesswater goes into storage, causing the pool level to rise above

outlets. A gated spillway's main purpose is to maximize

reservoir elevation at the normal full pool operating level is

construction of the reservoir. By gate operation, releasescan be reduced and additional water will be held in storage,which is called “induced-surcharge storage.” The releaserates should be made in accordance with spillway gateregulation schedules developed for each gated reservoir.EM 1110-2-3600 Section 4-5 describes induced surchargestorage and the development and testing of the regulationschedules.

c. Surcharge storage. The important factor in therouting procedure is the evaluation of the effect of storagein the upper levels of the reservoir, surcharge storage, onthe required outflow capacity. In computing the availablestorage, the water surface is generally considered to belevel. There will be a sloping water surface at the head ofthe reservoir due to backwater effect, and this conditionwill create an additional “wedge storage.” However, inmost large and deep reservoirs this incremental storage canbe neglected.

d. Drawdown. If a reservoir is drawn down at thetime of occurrence of the spillway design flood, the initialincrements of inflow will be stored with the correspondingreduction in ultimate peak outflow. Therefore, formaximum safety in design it is generally assumed that areservoir will be full to the top of flood-control pool at thebeginning of the spillway design flood.

e. Large flood-control storage reservations. Theremay be exceptions to the above criteria in the case ofreservoirs with large reservations for flood-control storage.However, even in such cases, a substantial part(> 50 percent) of the flood-control storage should beconsidered as filled by runoff from antecedent floods. Theeffect on the economics and safety of the project should beanalyzed before adopting such assumptions. ER 1110-8-2(FR) contains guidance on inflow design flood develop-ment and application.

EM 1110-2-142031 Oct 97

14-3

f. Release rates. Assuming a reservoir can be sig- spillway can then be proportioned to conform to thenificantly drawn down in advance of the spillway design required capacity and the specific site conditions, and aflood by using a short-term flood warning system is gen- complete layout of the spillway can be established. Costerally not acceptable for several reasons. The volume that estimates of the spillway and dam can then be made.can be released is the product of the total rate of discharge Estimates of various combinations of spillway capacity andat the dam times the warning time. Because the warning dam height for an assumed spillway type, and of alternativetime is usually short, except on large rivers, the release rate types of spillways, will provide a basis for the selection ofmust be the greatest possible without flood damage the most economical spillway type and the optimumdownstream. Even under the most favorable conditions, it relation of spillway capacity to the height of the dam.is unlikely that the released volume will be significant,relative to the volume of the spillway design flood. d. Maximum reservoir level. The maximum reser-

14-5. Sizing the Spillway

a. Storage and spillway capacity. In determining thebest combination of storage and spillway capacity toaccommodate the selected inflow design flood, all pertinentfactors of hydrology, hydraulics, design, cost, and damageshould be considered. In this connection and when applica-ble, consideration should be given to the following factors:

(1) The characteristics of the flood hydrograph.

(2) The damage which would result if such a floodoccurred without the dam.

(3) The damage which would result if such a floodoccurred with the dam in-place.

(4) The damage which would occur if the dam orspillway were breached.

(5) Effects of various dam and spillway combinationson the probable increase or decrease of damages above orbelow the dam.

(6) Relative costs of increasing the capacity of thespillways.

(7) The use of combined outlet facilities to serve morethan one function.

b. Outflow characteristics. The outflow characteris-tics of a spillway depend on the particular device selectedto control the discharge. These control facilities may takethe form of an overflow weir, an orifice, a tube, or a pipe.Such devices can be unregulated, or they can be equippedwith gates or valves to regulate the outflow.

c. Flood routing. After a spillway control of certaindimensions has been selected, the maximum spillwaydischarge and the maximum reservoir water level can bedetermined by flood routing. Other components of the

voir level can be determined by routing the spillway designflood hydrograph using sequential routing procedures andthe proposed operation procedures. This is a basic step inthe selection of the elevation of the crest of the dam, thesize of the spillway, or both.

e. Peak rate of inflow. Where no flood storage isprovided, the spillway must be sufficiently large to pass thepeak of the flood. The peak rate of inflow is then ofprimary interest, and the total volume in the flood becomesless important. However, where a relatively large storagecapacity above normal reservoir level can be madeeconomically available by a higher dam, a portion of theflood volume can be retained temporarily in reservoirsurcharge space, and the spillway capacity can be reducedconsiderably. If a dam could be made sufficiently high toprovide storage space to impound the entire volume of theflood above normal storage level, theoretically, no spillwayother than an emergency type would be required, providedthe outlet capacity could evacuate the surcharge storage ina reasonable period of time in anticipation of a recurringflood. The maximum reservoir level would then dependentirely on the volume of the flood, and the rate of inflowwould be of no concern. From a practical standpoint,however, relatively few sites will permit complete storageof the inflow design flood by surcharge storage.

f. Overall cost. The spillway length and corre-sponding capacity may have an important effect on theoverall cost of a project because the selection of the spill-way characteristics is based on an economic analysis. Inmany reservoir projects, economic considerations willnecessitate a design utilizing surcharge. The mosteconomical combination of surcharge storage and spillwaycapacity requires flood routing studies and economicstudies of the costs of spillway-dam combinations. Amongthe many economic factors that may be considered aredamage due to backwater in the reservoir, cost-heightrelations for gates, and utilization in the dam of materialexcavated from the spillway channel. However,consideration must still be given to the minimum sizespillway which must be provided for safety.

EM 1110-2-142031 Oct 97

14-4

g. Comprehensive study. The study may require designed to close if the operating gates fail, or wheremany flood routings, spillway layouts, and spillway and dewatering is desired to inspect or repair the operatingdam estimates. Even then, the study is not necessarily gates.complete because many other spillway arrangements couldbe considered. A comprehensive study to determine d. Continuous low-flow releases. Continuous low-alternative optimum combinations and minimum costs may flow releases are usually required to satisfy the needs ofnot be warranted for the design of some dams. Judgment fish, wildlife and existing water rights downstream from theon the part of the designer would be required to select for dam. When the low-flow release is small, one or twostudy only those combinations which show definite separate small bypass pipes, with high-pressure regulatingadvantages, either in cost or adaptability. For example, valves, are provided to facilitate operations. Flood-although a gated spillway might be slightly cheaper than an regulating gates may be used for making low-flow releasesungated spillway, it may be desirable to adopt the latter when those low-flow releases require substantial gatebecause of its less complicated construction, its automatic openings (EM 1110-2-1602).and trouble-free operation, its ability to function without anattendant, and its less costly maintenance. e. Uses of an outlet works. An outlet works may be

14-6. Outlet Works

a. Definition. An outlet works consists of theequipment and structures which together release therequired water for a given purpose or combination ofpurposes. Flows through river outlets and canal or pipelineoutlets change throughout the year and may involve a widerange of discharges under varying heads. The accuracy andease of control are major considerations and a great amountof planning may be justified in determining the type ofcontrol devices that can be best utilized.

b. Description. Usually, the outlet works consist ofan intake structure, a conduit or series of conduits throughthe dam, discharge flow control devices, and an energydissipating device where required downstream of the dam.The intake structure includes a trash-rack, an entrancetransition, and stop-logs or an emergency gate. The controldevice can be placed at the intake on the upstream face, atsome point along the conduit and be regulated fromgalleries inside the dam, or at the downstream end of theconduit with the operating controls placed in a gate-houseon the downstream face of the dam. When there is a powerplant or other structure near the face of the dam, the outletconduits can be extended farther downstream to dischargeinto the river channel beyond these features. In this case, acontrol valve may be placed in a gate structure at the end ofthe conduit.

c. Discharge. Discharges from a reservoir outletworks fluctuate throughout the year depending upondownstream water needs and reservoir flood controlrequirements. Therefore, impounded water must bereleased at specific regulated rates. Operating gates andregulating valves are used to control and regulate the outletworks flow and are designed to operate in any positionfrom closed to fully open. Guard or emergency gates are

used for diverting the river flow or portion thereof during aphase of the construction period, thus avoiding thenecessity for supplementary installations for that purpose.The outlet structure size dictated by this use rather than thesize indicated for ordinary outlet requirements maydetermine the final outlet works capacity.

f. Intake level. The establishment of the intakelevel is influenced by several considerations such as main-taining the required discharge at the minimum reservoiroperating elevation, establishing a silt retention space, andallowing selective withdrawal to achieve suitable watertemperature and/or quality. Dams which will impoundwaters for irrigation, domestic use, or other conservationpurposes must have the outlet works intake low enough tobe able to draw the water down to the bottom of theallocated storage space. Further, if the outlets are to beused to evacuate the reservoir for inspection or repair ofthe dam, they should be placed as low as practicable.However, it is usual practice to make an allowance in areservoir for inactive storage for silt deposition, fish andwildlife conservation, and recreation.

g. Elevation of outlet intake. Reservoirs becomethermally stratified, and taste and odor vary between ele-vations. Therefore, the outlet intake should be establishedat the best elevation to achieve satisfactory water qualityfor the purpose intended. Downstream fish and wildliferequirements may determine the temperature at which theoutlet releases should be made. Municipal and industrialwater use increases the emphasis on water quality andrequires the water to be drawn from the reservoir at theelevation which produces the most satisfactory com-bination of odor, taste, and temperature. Water supplyreleases can be made through separate outlet works atdifferent elevations if requirements for the individual wateruses are not the same and the reservoir is stratified.

EM 1110-2-142031 Oct 97

14-5

h. Energy-dissipating devices. The two types of spillway to utilize the spillway-stilling device forenergy dissipating devices most commonly used in con- dissipating the energy of the water discharging from thejunction with outlet works on concrete dams are hydraulic river outlets. Energy-dissipating devices for free-flowjump stilling basins and plunge pools. On some dams, it is conduit outlet works are essentially the same as those forpossible to arrange the outlet works in conjunction with the spillways.

EM 1110-2-142031 Oct 97

15-1

Chapter 15Dam Freeboard Requirements

15-1. Basic Considerations

a. Freeboard. Freeboard protects dams andembankments from overflow caused by wind-induced tidesand waves. It is defined as the vertical distance betweenthe crest of a dam and some specified pool level, usuallythe normal operating level or the maximum flood level.Depending on the importance of the structure, the amountof freeboard will vary in order to maintain structuralintegrity and the estimated cost of repairing damagesresulting from overtopping. Riprap or other types of slopeprotection are provided within the freeboard to controlerosion that may occur even without overtopping.

b. Estimating freeboard. Freeboard is generallybased on maximum probable wind conditions when thereference elevation is the normal operating level. Whenestimating the freeboard to be used with the probablemaximum reservoir level, a lesser wind condition is usedbecause it is improbable that maximum wind conditionswill occur simultaneously with the maximum flood level.A first step in wave height determinations is a study ofavailable wind records to determine velocities and relateddurations and directions. Three basic considerations aregenerally used in establishing freeboard allowance. Theseare wave characteristics, wind setup, and wave runup.

c. Further information. The Corps of EngineersCoastal Engineering Research Center (CERC) has devel-oped criteria and procedures for evaluating each of theabove areas. The primary references are EM 1110-2-1412and EM 1110-2-1414. The procedures presented in thesemanuals have received general acceptance for use in esti-mating freeboard requirements for reservoirs.

d. Applications. In applications for inland reservoirs,it is necessary to give special consideration to theinfluences that reservoir surface configuration, surroundingtopography, and ground roughness may have on windvelocities and directions over the water surface. Theeffects of shoreline irregularities on wave refraction andinfluences of water depth variations on wave heights andlengths must be accounted for. Although allowances canonly be approximated, the estimates of wave and wind tidecharacteristics in inland reservoirs can be preparedsufficiently accurate for engineering purposes.

15-2. Wind Characteristics over Reservoirs

a. General. The more violent windstorms experi-enced in the United States are associated with tropicalstorms (hurricanes) and tornadoes. Hurricane wind char-acteristics may affect reservoir projects located near Atlan-tic and Gulf coastlines, but winds associated with tornadoesare not applicable to the determination of freeboardallowances for wave action. In mountainous regions, theflow of air is influenced by topography as well as meteo-rological factors. These “orographic” wind effects, whenaugmented by critical meteorological patterns, may pro-duce high wind velocities for relatively long periods oftime. Therefore, they should be given special considerationin estimating wave action in reservoirs located inmountainous regions. In areas not affected by majortopographic influences, air movement is generally the resultof horizontal differences in pressure which in turn are dueprimarily to large-scale temperature differences in airmasses. Wind velocities and durations associated withthese meteorological conditions, with or without majorinfluences of local topography, are of major importance inestimating wave characteristics in reservoirs.

b. Isovel patterns. Estimates of wind velocities anddirections near a water surface at successive intervals oftime, as a windstorm passes the area, may be established byderiving “isovel” patterns. Sequence relations canrepresent wind velocities at, say, one-half hour intervalsduring periods of maximum winds, and one-hour or longerintervals thereafter. The “isovel” lines connect points ofequal wind speeds, resembling elevation contour maps.Wind directions are indicated by arrows. EM 1110-2-1412,Sections 1.9 and 1.10 describe storms and the storm surgegeneration process. Figure 1-1a shows an example windisovel pattern and pertinent parameters.

c. Relation of wind duration to wave heights. Ifwind velocity over a particular fetch remains constant,wave heights will progressively increase until a limitingmaximum value is attained, corresponding approximatelyto relations dependent on fetch distance, wind velocity, andduration. Accordingly, wind velocity-duration relationsapplicable to effective reservoir fetch areas are needed foruse in computing wave characteristics in reservoirs.

d. Wind velocity-duration relations. In some casesit is desired to estimate wave characteristics in existingreservoirs in order to analyze causes of riprap damage orfor other reasons. Wind records, supplemented by meteo-rological studies are usually required. Data on actual

EM 1110-2-142031 Oct 97

15-2

windstorms of record have been maintained at many over reservoir areas, and vice versa. The relationships areU.S. Weather Bureau stations. Index values, such as the not constant, but vary with topographic and vegetativefastest mile, 1-min average or 5-min average velocities, cover of land areas involved, reservoir configurations, andwith direction indications, are usually presented in clima- other conditions affecting air flow. However, on the basistological data publications. Some data collected by other of research and field studies (Technical Memorandumagencies and private observers may be available in pub- No. 132, USACE 1962), the following ratios representlished or unpublished form. However, information averages that are usually suitable for computing waveregarding wind velocities sustained for several hours or characteristics in reservoirs that are surrounded by terraindays is not ordinarily published in detail. Accordingly, of moderate irregularities and surface roughness:special studies are usually required to determine windvelocity-duration relations applicable to specific effectivefetch areas involved in wave computations. Basic recordsfor such studies are usually available from the U.S. Wea-ther Bureau offices or other observer stations. Somesummaries of wind velocities over relatively long periodsof time have been published by various investigators, andothers may be available in project reports related to waterresources development.

e. Generalized wind velocity-duration relations.Studies show that maximum wind velocities in one generaldirection during major windstorms, in most regions of theUnited States, have averaged approximately 40 to 50 mphfor a period of 1 hr. Corresponding velocities in the samegeneral direction for periods of 2 hr and 6 hr have averaged95 percent and 88 percent, respectively, of the maximum1-hr average velocity. In EM 1110-2-1414, Figure 5-26provides the ratio of wind speed of any duration to the 1-hrwind speed. Extreme wind velocities for brief periods,normally referred to as “fastest mile” or 1-min average,have been recorded as high as 150 to 200 percent ofmaximum 1-hr averages in most regions. In EM 1110-2-1414, Figures 5-18 through 5-20 provide the annualextreme fastest-mile speed 30 ft (9.1 m) above ground forthe 25-year, 50-year, and 100-year recurrence intervals,respectively. However, these extreme values are seldom ofinterest in computing wave characteristics in reservoirs.Generalized wind velocity-duration relations are consideredto be fairly representative of maximum values that arelikely to prevail over a reservoir in generally a singledirection for periods up to 6 hr (excluding projects locatedin regions that are subject to severe hurricanes ororographic wind-flow effects). Special studies of windcharacteristics associated with individual project areasshould be made when determinations of unusualimportance, or problems requiring consideration of winddurations exceeding 6 hr, are involved.

f. Ratio of wind velocities over water and land areas.The wind velocities described in paragraph d are for overland. Under comparable meteorological conditions, windvelocities over water are higher than over land surfacesbecause of smoother and more uniform surface conditions.Winds blowing from land tend to increase with passage

Fetch (F ) in Miles Over Lande

Wind ratio Over Water

0.5 1.08

1 1.13

2 1.21

3 1.26

4 1.28

5 (or over) 1.30

15-3. Computation of Wave and Wind TideCharacteristics

a. Effective wind fetch (F ) for wave generation.e

The characteristics of wind-generated waves are influencedby the distance that wind moves over the water surface inthe “fetch” direction. The generally narrow irregularshoreline of inland reservoirs will have lower waves thanan open coast because there is less water surface for thewind to act on. The method to compensate for the reducedwater surface for an enclosed body of water is computationof an effective fetch. The effective fetch (F ) adjusts radiale

lines from the embankment to various points on thereservoir shore. The radials spanning 45 deg on each sideof the central radial are adjusted by the cosign of theirangle to the central radial to estimate an average effectivefetch. The computation procedure is shown in EM 1110-2-1414, Figure 5-33 and Example Problem 7-2. Generalizedrelations are based on effective fetch distances derived inthis manner.

b. Fetch distance for wind tide computations. Fetchdistances for use in estimating wind tide (set-up) effects areusually longer than effective fetch distances used inestimating wave heights. In as much as wind tide effects indeep inland reservoirs are relatively small, extensivestudies to refine estimates are seldom justified. For prac-tical purposes, it is usually satisfactory to assume that thewind tide fetch is equal to twice the effective fetch (F ). Ife

wind tide heights determined in this manner are relativelylarge in relation to overall freeboard requirements, moredetailed analyses are advisable using methods as generallydiscussed in Chapter 3 of EM 1110-2-1414.

S =U 2 F

1400 D

EM 1110-2-142031 Oct 97

15-3

(15-1)

c. Generalized diagrams for wave height and wavemany slopes, wave conditions, and embankment porosityperiod in deep inland reservoirs. EM 1110-2-1414, Fig- provide sufficient data to make estimates of wave runup onure 5-34, presents generalized relations between significant a prototype embankment.wave height, wave period, fetch (F ), and wind velocitiese

corresponding to critical durations. These diagrams were b. Relative runup relations. EM 1110-2-2904developed from research and field studies based on wind Plate 25 presents generalized relations on wave runup onspeed at 10 m (33 ft). If wind speeds are for a different rubble-mound breakwaters and smooth impervious slopes.level, Equation 5-12 can be used to adjust to the 10 m level Plate 26 provides similar curves for various embankment(EM 1110-2-1414). slopes for water depths greater than three-times wave

d. Wave characteristics in shallow inland lakes and large number of small and large-scale hydraulic model testreservoirs. In the analyses of wave characteristics, lakes results, and have been adjusted for model scale effects toand reservoirs are considered to be shallow when depths in represent prototype conditions. The relations were basedthe wave-generating area are generally less than about one- primarily on tests involving mechanically generated waveshalf the theoretical deep-water wave length (L ) and may differ somewhat from relations associated witho

corresponding to the same wave period (T). Curves pre- individual waves in natural wind-generated spectrums ofsented in Figures 5-35 through 5-44 represent relations waves. However, general field observations andbetween wave characteristics, fetch distances (in feet) and comparisons with wave experiences support the conclusionconstant water depths in the wave-generating area, ranging that relations presented.from 5 to 50 ft (1.5 to 15.2 m) (EM 1110-2-1414).

e. Wind tides (set-up) in inland waters. (1) If waves generated in deep water (i.e., depths

(1) When wind blows over a water surface, it exerts a exceeding about one-third to one-half the wave length),horizontal stress on the water, driving it in the direction of reach the toe of an embankment without breaking, thethe wind. In an enclosed body of water, this wind effect vertical height of runup may be computed by multiplyingresults in a piling up of water at the leeward end, and a the deep-water wave height (H ) by the relative runuplowering of water level at the windward end. This effect is ration (R/H) obtained from EM 1110-2-2904, Plate 26, forcalled “wind tide” or “wind set-up.” Wind set-up can be the appropriate slope and wave steepness (H /L ). In thisreasonably estimated for lakes and reservoirs, based on the case, deep-water values of H and L should be used asfollowing equation: indices, even though wave heights and lengths are modified

in which S is wind tide (set-up) in feet above the stillwater located in deep water or in shallow water, provided thelevel that would prevail with zero wind action; U is the wave does not break before reaching the toe of theaverage wind velocity in statute miles per hour over the structure.fetch distance (F) that influences wind tide; D is theaverage depth of water generally along the fetch line (2) If waves are generated primarily by winds over(EM 1110-2-1414). The fetch distance (F) used in the open-water areas where the relative depth (d/L) isabove formula is usually somewhat longer than the effec- appreciably less than 0.3, the wave heights and periodstive fetch (F ) used in wave computations, as indicated in should be computed by procedures applicable to shallowe

paragraph 15-3b. Refer to EM 1110-2-1414, Section 3-2 waters.for a discussion of prediction models.

15-4. Wave Runup on Sloping Embankment

a. Introduction. Most dam embankments are frontedby deep water, have slopes between 1 on 2 and 1 on 4, andare armored with riprap. Rock-fill dams are considered aspermeable rubble slopes and earth-fill dams with ripraparmor are considered impermeable. Laboratory tests of

height. The curves correspond to statistical averages of a

c. Runup of waves on sloping embankments.

o

o o

o o

by passing through areas in which water depths are lessthan L /3 (provided the depth is not small enough to causeo

the wave to break before reaching the embankment). Thatis, the height of runup may be computed by using the deep-water steepness H /L , whether the structure under study iso o

(3) Waves generated by wind over open-water areasof a particular depth change characteristics when theyreach areas where the constant depth is substantially less,the height (H) tending to increase while the length (L)decreases. The distribution of wave energy changes as awave enters the shallow water, the proportion of totalenergy which is transmitted forward with the wave towardthe shore increasing, although the actual amount of this

EM 1110-2-142031 Oct 97

15-4

translated energy remains constant except from minor establish allowances needed to provide for wave action thatfrictional effects. If the depth continues to decrease, the is likely to affect various project elements, as follows:steepness ratio (H/L) increases, until finally the wavebecomes unstable and breaks, resulting in appreciable (1) Main embankment of the dam, and supplementalenergy dissipation. Theoretically, the maximum wave dike sections.cannot exceed 0.78 D, where D is the depth of waterwithout wave action. After breaking, the waves will tend to (2) Levees that protect areas within potential flowagereform with lower heights within a distance equal to a few limits of the reservoir.wave lengths. For most engineering applications, it issatisfactory to assume that the wave height after breaking (3) Highway and railroad embankments that intersectwill equal approximately 0.78 D in the shallow area and the reservoir limits.that L will be the same as before the wave broke. Plate 26would then be entered with a wave-steepness ratio equal to (4) Structures located within the reservoir area.0.78/L to determine the relative runup ration (R/H), andthis ratio would be multiplied by 0.78 D to obtain the (5) Shoreline areas that are subject to adverse effectsestimated runup height (R). This procedure should provide of wave action.conservative results under circumstances in which thedistance between the point where waves reach breaking b. Freeboard on dams. The establishment of free-depths and location of the structure under study is long board allowances on dams includes not only the consider-enough to permit waves to reform, and short enough to ation of potential wave characteristics in a reservoir, butpreclude substantial build-up by winds prevailing over the several other factors of importance, including certain policyshallower area. More accurate values could be obtained by matters.using this breaking height (0.78 D) and period to obtaincomparable deep water values of H and L . c. Freeboard allowances for wave action ono o

d. Adjustments in wave runup estimates for varia- limits.tions in riprap.

(1) A rough riprap layer on an embankment tends to establishing design grades and slope protection measuresreduce the height of runup after a wave breaks. If the for highway, railroad, levee, and other embankments thatriprap layer thickness is small in comparison with wave intersect or boarder a reservoir. The design of operatingmagnitudes and the underlying surface is relatively imper- structure, boat docks, recreational beaches, and shorelinemeable, so that the void spaces in the riprap remain mostly protection measures at critical locations involves thefilled with water between successive waves during severe consideration of wave characteristics and frequencies understorm events, the height of runup may closely approach a range of conditions. Estimates of wave characteristicsheights attained on smooth embankments of comparable affecting the design of these facilities can have a majorslope. However, if the riprap layer is sufficiently rough, influence on the adequacy of design and costs ofthick, and free draining to quickly absorb the water that relocations required for reservoir projects, and in theimpinges on the embankment as each successive wave development of supplemental facilities.breaks, further wave runup will be almost completelyeliminated. (2) The freeboard reference level selected as a base

(2) The design of riprap to absorb most of the energy several types of facilities referred to above will be gov-of breaking waves is practicable if waves involved will be erned by considerations associated with the particularrelatively small or moderate, but costs and other practical facility. Otherwise, procedures generally as described withconsiderations usually preclude such design where large respect to the determination of freeboard allowances forwaves are encountered. Accordingly, the design dams should be followed, and stage hydrographs andcharacteristics of riprap layers are usually somewhere related wave runup elevations corresponding to the selectedbetween the two extremes described above. wind criteria should be prepared. However, the freeboard

15-5. Freeboard Allowances for Wave Action

a. Purpose. In connection with the design of damsand reservoirs, the estimate of freeboard is required to

embankments and structures within reservoir flowage

(1) Wave action effects must be taken into account in

for estimating wave effects associated with each of the

reference level and coincident wind velocity-durationrelations selected for these studies usually correspond toconditions that would be expected with moderatefrequency, instead of the rare combinations assumed inestimating the height of dam required for safety.

EM 1110-2-142031 Oct 97

15-5

(3) In estimating effects of wave action on embank- consideration is particularly important in estimating thements and structures, the influences of water depths near effects waves may have on bridge structures that arethe facility should be carefully considered. If the shallow partially submerged under certain reservoir conditions.depths prevail for substantial distances from the embank-ment or structure under study, wave effects may be greatly (4) Systematic analyses of wave effects associatedreduced from those prevailing in deep-water areas. On the with various key locations along embankments that cross orother hand, facilities located where sudden reductions in border reservoirs provide a practical basis for varyingwater depths cause waves to break are likely to be design grades and erosion protection measures to establishsubjected to greater dynamic forces than would be imposed the most economical plan to meet pertinent operational andon similar facilities located in deep water. This maintenance standards.

EM 1110-2-142031 Oct 97

16-1

Chapter 16Dam Break Analysis

16-1. Introduction

a. Corps policy. It is the policy of the Corps of Engi-neers to design, construct, and operate dams safely(ER 1110-8-2(FR)). When a dam is breached, catastrophicflash flooding occurs as the impounded water escapesthrough the gap into the downstream channel. Usually, theresponse time available for warning is much shorter thanthat for precipitation-runoff floods, so the potential for lossof life and property damage is much greater.

b. Hazard evaluation. A hazard evaluation is thebasis for selecting the performance standards to be used indam design or in evaluating existing dams. When floodingcould cause significant hazards to life or major propertydamage, the design flood selected should have virtually nochance of being exceeded. ER 1110-8-2(FR) provides damsafety standards with respect to the appropriate selection ofan inflow design flood. If human life is at risk, the generalrequirement is to compute the flood using PMP. If lesserhazards are involved, a smaller flood may be selected fordesign. However, all dams should be designed to withstanda relatively large flood without failure even when there isapparently no downstream hazard involved under presentconditions of development.

c. Safety design. Safety design includes studies toascertain areas that would be flooded during the designflood and in the event of dam failure. The areas down-stream from the project should be evaluated to determinethe need for land acquisition, flood plain management, orother methods to prevent major damage. Informationshould be developed and documented suitable for releasingto downstream interests regarding the remaining risks offlooding.

d. National Dam Safety Act. The potential for cata-strophic flooding due to dam failures in the 1960's and1970's brought about passage of the National Dam SafetyAct, Public Law 92-367. The Corps of Engineers becameresponsible for inspecting U.S. Federal and non-Federaldams, which met the size and storage limitations of the act,in order to evaluate their safety. The Corps inventorieddams; surveyed each State and Federal agency'scapabilities, practices, and regulations regarding the design,construction, operation, and maintenance of the dams;developed guidelines for the inspection and evaluation ofdam safety; and formulated recommendations for acomprehensive national program.

e. Flood emergency documents. In support of theNational Dam Safety Program, flood emergency planningfor dams was evaluated in the 1980's, and a series of docu-ments were published: Emergency Planning for Dams,Bibliography and Abstracts of Selected Publications (HEC1982a), Flood Emergency Plan, Guidelines for CorpsDams, Research Document 13 (HEC 1980), ExampleEmergency Plan for Blue Marsh Dam and Lake (HEC1983a), and Example Plan for Evacuation of Reading,Pennsylvania in the Event of Emergencies at Blue MarshDam and Lake (HEC 1983b). The development of anemergency plan requires the identification of the type ofemergencies to be considered, the gathering of needed data,performing the analyses and evaluations, and presenting theresults. HEC Research Document 13 (HEC 1980) providesguidelines for each step of the process.

16-2. Dam Breach Analysis

a. Causes of dam failures. Dam failures can becaused by overtopping a dam due to insufficient spillwaycapacity during large inflows to the reservoir, by seepage orpiping through the dam or along internal conduits, slopeembankment slides, earthquake damage and liquification ofearthen dams from earthquakes, or landslide-generatedwaves within the reservoir. Hydraulics, hydrodynamics,hydrology, sediment transport mechanics, and geotechnicalaspects are all involved in breach formation and eventualdam failure. HEC Research Document 13 lists theprominent causes as follows:

(1) Earthquake.

(2) Landslide.

(3) Extreme storm.

(4) Piping.

(5) Equipment malfunction.

(6) Structural damage.

(7) Foundation failure.

(8) Sabotage.

b. Dam breach characteristics. The breach is theopening formed in the dam when it fails. Despite the factthat the main modes of failure have been identified aspiping or overtopping, the actual failure mechanics are notwell understood for either earthen or concrete dams. Inprevious attempts to predict downstream flooding due todam failures, it was usually assumed that the dam failed

EM 1110-2-142031 Oct 97

16-2

completely and instantaneously. These assumptions of (1) The first approach uses statistically derivedinstantaneous and complete breaches were used for reasons regression equations, like those formulated by MacDonaldof convenience when applying certain mathematical and Langridge-Monopolis (1984) and by Froelich (1987).techniques for analyzing dam-break flood waves. The Both sets of equations are based on actual data frompresumptions are somewhat appropriate for concrete arch- dozens of historic dam failures. The MacDonald, andtype dams, but they are not suitable for earthen dams and Langridge-Monopolis study was based on data fromconcrete gravity-type dams. 42 constructed earth- and rock-fill dams. The Froelich

(1) Earthen dams, which exceedingly outnumber all earthen dams. Both studies resulted in a set of graphs andother types of dams, do not tend to completely fail, nor do equations that can be used to predict the approximate sizethey fail instantaneously. Once a developing breach has of the breach and the time it takes for the breach to reachbeen initiated, the discharging water will erode the breach its full size.until either the reservoir water is depleted or the breachresists further erosion. The fully formed breach in earthen (2) The second approach is a physically based com-dams tends to have an average width (b) in the range puter model called BREACH, developed by Dr. Danny(h yby3h ) where, h is the height of the dam. Breach Fread (1989) for the National Weather Service. The breachd d d

widths for earthen dams are therefore usually much less model uses sediment transport and hydraulic routingthan the total length of the dam as measured across the equations to simulate the formation of either a piping orvalley. Also, the breach requires a finite interval of time overtopping type of failure. The model requires infor-for its formation through erosion of the dam materials by mation about the physical dimensions of the dam, as wellthe escaping water. The total time of failure may range as a detailed description of the soil properties of the dam.from a few minutes to a few hours, depending on the height Soils information includes D50 (mm), porosity, unit weightof the dam, the type of materials used in construction, and (lb/ft ), internal friction angle, cohesive strength (lb/ft ),the magnitude and duration of the flow of escaping water. and D90/D30. These parameters can be specifiedPiping failures occur when initial breach formation takes separately for the inner-core and outside-bank materials ofplace at some point below the top of the dam due to erosion a dam.of an internal channel through the dam by escaping water.As the erosion proceeds, a larger and larger opening isformed. This is eventually hastened by caving-in of the topportion of the dam.

(2) Concrete gravity dams also tend to have a partialbreach as one or more monolith sections formed during thedam construction are forced apart by the escaping water.The time for breach formation is in the range of a fewminutes.

(3) Poorly constructed earthen dams and coal-wasteslag piles which impound water tend to fail within a fewminutes and have average breach widths in the upper rangeor even greater than those for the earthen dams mentionedabove.

c. Dam breach parameters. The parameters of fail-ure depend on the dam and the mode of failure. For floodhydrograph estimation, the breach is modeled assumingweir conditions, and the breach size, shape, and timing arethe important parameters. The larger the breach openingand the shorter the time to total failure, the larger the peakoutflow. HEC Research Document 13, Table 1, listssuggested breach parameters for earth-fill, concrete-gravity, and concrete-arch dams. There are two basicapproaches used to determine possible breach sizes andtimes.

study included data from constructed and landslide-formed

3 2

16-3. Dam Failure Hydrograph

a. Flow hydrograph. The flow hydrograph from abreached dam may be computed using traditional methodsfor flow routing through a reservoir and downstreamchannel. The reservoir routing approach is the same asrouting for the spillway design flood, described in Chap-ter 14. Generally, a short time step is required because thebreach formation and resulting reservoir outflow changerapidly with time.

b. Routing methods. The choice between hydraulicand hydrologic routing depends on many factors, includingthe nature of available data and accuracy required. Thehydraulic method is the more accurate method of routingthe unsteady flow from a dam failure flood through thedownstream river. This technique simultaneouslycomputes the discharge, water surface elevation, andvelocity throughout the river reach. Chapter 9 ofEM 1110-2-1417 describes the routing methods andapplicability of routing techniques. Chapter 5 of EM 1110-2-1416 describes unsteady flow computations.

c. Geometry and surface area. The geometry andsurface area of the reservoir can also affect the choice ofmethod. For very narrow and long reservoirs where thedam is relatively large, the change of water level at the

Qmax =827

Wb gYo3/2

y =49

Yo

EM 1110-2-142031 Oct 97

16-3

(16-1)

(16-2)

failed dam is rapid, and the unsteady flow method is useful.However, for very large reservoirs where the dam is smallcompared to the area of the lake, the change in water levelis relatively slow and the storage routing method (ModifiedPuls) is economical in developing the failure hydrograph.Because of the rapid change in water level, small timeperiods are required for both methods.

d. Height of downstream water. The height of thewater downstream of a dam (tailwater) also affects theoutflow hydrograph in a failure analysis. It also affects theformation or nonformation of a bore in front of the wave.

e. Deriving the peak outflow. By assuming a rectan-gular cross section, zero bottom slope, and an instan-taneous failure of a dam, the peak outflow can be derivedby the mathematical expression originally developed by St.Venant, as follows:

where

Y is the initial depth, W is the width of the breach, g iso b

the gravity coefficient, and the water depth, y, justdownstream of the dam is

This equation is applicable only for relatively long andnarrow rectangular channels where the dam is completelyremoved. Guidelines for Calculating and routing a Dam-Break Flood, HEC Research Document No. 5 (HEC 1977)describes this approach.

f. Failed dam outflow hydrograph. The outflowhydrograph from a failed dam may also be approximatedby a triangle. For instantaneous failure, a right triangle isapplicable. The base represents the time to empty thereservoir volume, and the height represents theinstantaneous peak outflow. In erosion analysis, the Officeof Emergency Services, after consultation with other agen-cies, suggested an isosceles triangle. The rising side of theisosceles triangle is developed by assuming that half of thereservoir storage is required to erode the dam to naturalground level. The apex of the triangle represents the peakflow through the breach under the assumption that the flowoccurs at critical depth.

g. Potential for overtopping. The HydrologicEngineering Center's HEC-1 Flood Hydrograph Package(HEC 1990c) can be used to determine the potential forovertopping of dams by run off resulting from variousproportions of the PMF. This technique is most appro-priate for simulating breaches in earthen dams caused byovertopping. Other conditions may be approximated,however, such as instantaneous failure. This methodmakes six assumptions:

(1) Level-pool reservoir routing to determine time-history of pool elevation.

(2) Breach shape is a generalized trapezoid withbottom width and side slopes prespecified by the analyst.

(3) Bottom of the breach moves downward at aconstant rate.

(4) Breach formation begins where the water surfacein the reservoir reaches a prespecified elevation.

(5) Breach is fully developed when the bottomreaches a prespecified elevation.

(6) Discharge through the breach can be calculatedindependently of downstream hydraulics, i.e., criticaldepth occurs at or near the breach. A tailwater rating curveor a single cross section (assuming normal-depth for arating) can be used to simulate submergence effects.

The total discharge from the dam at any instant is cal-culated by summing the individual flows through the lowlevel outlet, over the spillway and top of the dam, andthrough the breach.

h. Peak flow values. With several calculations oftheoretical flood peaks from assumed breaches, peak flowvalues may seem either too low or too high. One way ofchecking the reasonableness of the assumption is to com-pare the calculated values with historical failures. Anenvelope of estimated flood peaks from actual dam failuresprepared by the Bureau of Reclamation is a good means ofcomparing such values. HEC Research Document No. 13,Figure 2, provides an envelope of experienced outflowrates from breached dams, as a function of hydraulic depth.

16-4. Dam Break Routing

a. Dam-break flood hydrographs. Dam-break floodhydrographs are dynamic, unsteady flow events.

EM 1110-2-142031 Oct 97

16-4

Therefore, the preferred routing approach is to utilize a full e. Special features. The following special featuresunsteady flow routing model. The HEC-1 Flood Hydro- and capacities are included in FLDWAV: variable �t andgraph package provides the capability to compute and route�x computational intervals; irregular cross-sectional geo-the inflow design flood and compute the breach and metry; off-channel storage; roughness coefficients that varyresulting hydrograph, but its channel routing is limited to with discharge or water surface elevation, and with distancehydrologic methods. The most appropriate HEC-1 along the waterway; capability to generate linearlyapproach is the Muskingum-Cunge option. The option uses interpolated cross sections and roughness coefficientsa simple cross section plus reach slope and length to define between input cross sections; automatic computation ofa routing reach. No downstream backwater effects are initial steady flow and water elevations at all cross sectionsconsidered. If simplified representations of the down- along the waterway; external boundaries of discharge orstream river reaches are acceptable, an adequate routing water surface elevation time series (hydrographs), a single-may be obtained. valued or looped depth-discharge relation (tabular or

b. St. Venant equations. The St. Venant equations internal boundaries enable treatment of time-dependentapply to gradually varied flow with a continuous profile. If dam failures, spillway flows, gate controls, or bridge flows,features which control or interrupt the water surface profile or bridge-embankment overtopping flow; short-circuitingexist along the main stem of the river or its tributaries, of floodplain flow in a valley with a meandering river;internal boundary conditions are required. These features levee failure and/or overtopping; a special computationalinclude dams, bridges, roadway embankments, etc. If the technique to provide numerical stability when treatingstructure is a dam, the total discharge is the sum of spillway flows that change from supercritical to subcritical , orflow, flow over the top of the dam, gated-spillway flow, conversely, with time and distance along the waterway; andflow through turbines, and flow through a breach, should a an automatic calibration technique for determining thebreach occur. The spillway flow and dam overtopping are variable roughness coefficient by using observedtreated as weir flow, with corrections for submergence. hydrographs along the waterway. The gated outlet can represent a fixed gate or one in whichthe gate opening can vary with time. These flows can also f. UNET. The unsteady flow program UNET (HECbe specified by rating curves which define discharge 1995) has a dam-break routing capability. However, therepassing through the dam as a function of upstream water has been limited application of this feature. UNET couldsurface elevation. be used to route the outflow hydrograph computed in an

c. Unsteady flow computer programs. There are an and write hydrographs using the HEC Data Storage System,increasing number of available unsteady flow computer HEC-DSS (HEC 1995a).programs. The FLDWAV program is a generalizedunsteady-flow simulation model for open channels. Itreplaces the DAMBRK, DWOPER, and NETWORKmodels, combining their capabilities and providing newhydraulic simulation procedures within a more user-friendly model structure (DeVries and Hromadka 1993).Given the long history of application by the NationalWeather Service, this program is likely the most capablefor this purpose.

d. FLDWAV. FLDWAV can simulate the failure ofdams caused by either overtopping or piping failure of thedam. The program can also represent the failure of two ormore dams located sequentially on a river. The program isbased on the complete equations for unsteady open-channelflow (St. Venant equations). Various types of external andinternal boundary conditions are programmed into themodel. At the upstream and downstream boundaries of themodel (external boundaries), either discharges or watersurface elevations, which vary with time, can be specified.

computed); time-dependent lateral inflows (or outflows);

HEC-1 runoff-dam break model. Both programs can read

16-5. Inundation Mapping

a. Preparation of maps. To evaluate the effects ofdam failure, maps should be prepared delineating the areawhich would be inundated in the event of failure. Landuses and significant development or improvements withinthe area of inundation should be indicated. The mapsshould be equivalent to or more detailed than the USGSquadrangle maps, 7.5-min series, or of sufficient scale anddetail to identify clearly the area that should be evacuated ifthere is evident danger of failure of the dam. Copies of themaps should be distributed to local government officials foruse in the development of an evacuation plan. The intentof the maps is to develop evacuation procedures in case ofcollapse of the dam, so the travel time of the flood waveshould be indicated on every significant habitation areaalong the river channel.

EM 1110-2-142031 Oct 97

16-5

b. Evaluation of hazard potential. To assist in the f. Evacuation plans. evaluation of hazard potential, areas delineated on inun-dation maps should be classified in accordance with the (1) Evacuation plans should be prepared and imple-degree of occupancy and hazard potential. The potential mented by the local jurisdiction controlling inundationfor loss of life is affected by many factors, including but areas. The assistance of local civil defense personnel, ifnot limited to the capacity and number of exit roads to available, should be requested in preparation of the eva-higher ground and available transportation. Hazard poten- cuation plan. State and local law enforcement agenciestial is greatest in urban areas. The evaluation of hazard usually will be responsible for the execution of much of thepotential should be conservative because the extent of plan and should be represented in the planning effort. Stateinundation is usually difficult to delineate precisely. and local laws and ordinances may require that other state,

c. Hazard potential for recreation areas. The hazard preparation, review, approval, or execution of the plan.potential for affected recreation areas varies greatly, Before finalization, a copy of the plan should be furnisheddepending on the type of recreation offered, intensity of to the dam agency or owner for information and comment.use, communications facilities, and available transpor-tation. The potential for loss of life may be increased (2) Evacuation plans will vary in complexity in accor-where recreationists are widely scattered over the area of dance with the type and degree of occupancy in thepotential inundation because they would be difficult to potentially affected area. The plans may includelocate on short notice. delineation of the area to be evacuated; routes to be used;

d. Industries and utilities. Many industries and utili- emergency transportation; special procedures for the evac-ties requiring substantial quantities of water are located on uation and care of people from institutions such as hospi-or near rivers or streams. Flooding of these areas and tals, nursing homes, and prisons; procedures for securingindustries, in addition to causing the potential for loss of the perimeter and for interior security of the area; proce-life, can damage machinery, manufactured products, raw dures for the lifting of the evacuation order and reentry tomaterials and materials in process of manufacture, plus the area; and details indicating which organizations areinterrupt essential community services. responsible for specific functions and for furnishing the

e. Least hazard potential. Rural areas usually have HEC Research Documents 19 and 20 provide examplethe least hazard potential. However, the potential for loss emergency plans and evacuation plans, respectively (HECof life exists, and damage to large areas of intensely cul- 1983a and b).tivated agricultural land can cause high economic loss.

county, and local government agencies have a role in the

traffic control measures; shelter; methods of providing

materials, equipment, and personnel resources required.

EM 1110-2-142031 Oct 97

17-1

Chapter 17Channel Capacity Studies

17-1. Introduction

a. General. Channel capacity studies tend to focuson high flows. Flood operations for a reservoir will requireoperational downstream targets for nondamaging flowswhen excess water must be released. Nondamagingchannel capacity may be defined at several locations, andthe target flow may be defined at several levels. There maybe lower targets for small flood events and, under extremeflood situations, the nondamaging target may cause someminor damage. Also, the nondamaging flow target mayvary seasonally and depend on floodplain land use.

b. Withstanding release rates. Channel capacity isalso concerned with the capability of the channel to with-stand reservoir release rates. Of particular concern is thereach immediately downstream from the reservoir. Highrelease rates for hydropower or flood control could damagechannel banks and cause local scour and channeldegradation.

c. Channel capacity. While flood operation mayfocus on maximum channel capacity, planning studiesusually require stage-discharge information over the entirerange of expected operations. Also, low-flow targets maybe concerned with maintaining minimum downstream flowdepth for navigation, recreation, or environmental goals.Channel capacity studies typically provide information onsafe channel capacity and stage-discharge (rating) curvesfor key locations.

17-2. Downstream Channel Capacity

a. Downstream channel erosion. Water flowing overa spillway or through a sluiceway is capable of causingsevere erosion of the stream bed and banks below the dam.Consequently, the dam and its appurtenant works must beso designed that harmful erosion is minimized. The outletworks for a dam usually require an energy-dissipatingstructure. The design may vary from an elaborate multiple-basin arrangement to a simple head wall design, dependingon the number of conduits involved, the erosion resistanceof the exit channel bed material, and the duration, intensity,and frequency of outlet flows. A stilling basin may be pro-vided for outlet works when such downstream uses asnavigation, irrigation, and water supply, require frequentoperation or when the channel immediately downstream iseasily eroded. Sections 4-2b and 4-3j of EM 1110-2-3600

provide a general discussion of energy dissipators forspillways and outlet works, respectively.

b. Adequate capacity. The channel downstreamshould have adequate capacity to carry most flows fromreservoir releases. After the water has lost most of itsenergy in the energy-dissipating devices, it is usuallytransported downstream through the natural channel to itsdestination points. With the expected release rates, thechannel should be able to resist excessive erosion andscour, and have a large enough capacity to prevent down-stream flooding except during large floods.

c. River surveys. River surveys of various typesprovide the basic physical information on which riverengineering planning and design are based. Survey datainclude information on the horizontal configuration (plan-form) of streams; characteristics of the cross sections(channel and overbank); stream slope; bed and bank mate-rials; water discharge; sediment characteristics and dis-charge; water quality; and natural and cultural resources.

d. Evaluating bank stability. To evaluate bankstability, it is essential to understand the complex historicalpattern of channel migration and bank recession of thestream and the relationship of channel changes to stream-flow. Studies of bank caving, based on survey data andaerial photographs, provide information on the progres-sively shifting alignment of a stream and are basic to layingout a rectified channel alignment. The concepts andevaluation procedures presented in “Stability of FloodControl Channels” (USACE 1990) are applicable to thechannel capacity evaluation.

e. Interrupted sediment flow. A dam and reservoirproject tends to interrupt the flow of sediment, which canhave a significant impact on the downstream channelcapacity. If the project is relatively new, the affect may notbe seen by evaluating historic information or currentchannel conditions. The future channel capacity willdepend on the long-term trends in aggradation and degra-dation along the river. General concepts on sedimentanalysis are presented in Chapter 9. Sediment Inves-tigations of Rivers and Reservoirs, EM 1110-2-4000, is theprimary reference for defining potential problems andanalyses procedures.

f. Downstream floodplain land use. Channel capa-city also depends on the long-term trends in downstreamfloodplain land use. While it is not a hydrologic problem,channel capacity studies should recognize the impact offloodplain encroachments on what is considered thenondamaging channel capacity. Anecdotal history has

EM 1110-2-142031 Oct 97

17-2

shown that many Corps’ projects are not able to make deviations which increase with time, but a flood can flushplanned channel-capacity releases due to development and away sediment and aquatic weed and cause a suddenencroachments downstream. reversal of the rating curve shift.

17-3. Stream Rating Curve

a. Stage-discharge relationship. The relationshipbetween stage and discharge, the “rating” at a gauging sta-tion, is based on field measurements with a curve fitted toplotted data of stage versus discharge. For subcritical flow,the stage-discharge relationship is controlled by the streamreach downstream of the gauge; for supercritical flow, thecontrol is upstream of the gauge. The stage-dischargerelationship is closely tied to the rate of change ofdischarge with time, and the rating curve for a rising stagecan be different from that for the falling stage in alluvialrivers.

b. Tailwater rating curve. The tailwater rating curve,which gives the stage-discharge relationship of the naturalstream below the dam, is dependent on the natural con-ditions along the stream and ordinarily cannot be altered bythe spillway design or by the release characteristics.Degradation or aggradation of the river below the dam,which will affect the ultimate stage-discharge conditions,must be recognized in selecting the tailwater rating curvesto be used for design. Usually, river flows which approachthe maximum design discharges have never occurred, andan estimate of the tailwater rating curve must either beextrapolated from known conditions or computed on thebasis of assumed or empirical criteria. Thus, the tailwaterrating curve at best is only approximate, and factors ofsafety to compensate for variations in tailwater must beincluded in dependent designs.

c. Extrapolation. Extrapolation of rating curves isnecessary when a water level is recorded below the lowestor above the highest gauged level. Where the cross sectionis stable, a simple method is to extend the stage-area andstage-velocity curve and, for given stage values, take theproduct of velocity and cross-section area to give dischargevalues beyond the stage values that have been gauged. Generally, water-surface profiles should be computed todevelop the rating beyond the range of observed data.

d. Rating curve shifts. The stage-discharge relation-ship can vary with time, in response to degradation,aggradation, or a change in channel shape at the controlsection, deposition of sediment causing increased approachvelocities in a weir pond, vegetation growth, or iceaccumulation. Shifts in rating curves are best detectedfrom regular gauging and become evident when several a. Appropriate methods. For most channel-capacitygaugings deviate from the established curve. Sediment studies, water surface profiles will be computed to developaccumulation or vegetation growth at the control will cause the required information. Given the technical concerns

e. Flow magnitude and bed material. Stream bedconfiguration and roughness in alluvial channels are afunction of the flow magnitude and bed material. Bedforms range from ripples and dunes in the lower regime(Froude number y 1.0) to a smooth plane bed, to antiduneswith standing waves (bed and water surface waves inphase) and with breaking waves and, finally, to a series ofalternative chutes and pools in the upper regime as theFroude number increases.

f. Upper and lower rating portions. The large chan-ges in resistance to flow that occur as a result of changingbed roughness affect the stage-discharge relationship. Theupper portion of the rating is relatively stable if it repre-sents the upper regime (plane-bed, transition, standingwave, or antidune regime) of bed form. The lower portionof the rating is usually in the dune regime, and the stage-discharge relationship varies almost randomly with time.Continuous definition of the stage-discharge relationship atlow flow is a very difficult problem, and a mean curve forthe lower regime is frequently used for gauges with shiftingcontrol.

g. Break up of surface material. In gravel-bedrivers, a flood may break up the armoring of the surfacegravel material, leading to general degradation until a newarmoring layer becomes established and ratings tend toshift between states of quasi-equilibrium. It may then bepossible to shift the rating curve up or down by the changein the mean-bed level, as indicated by plots of stage andbed level versus time.

h. Ice. Ice at the control section may also affect thenormal stage-discharge relationship. Ice effects vary withthe quantity and the type of ice (surface ice, frazil ice, oranchor ice). When ice forms a jam in the channel andsubmerges the control or collects in sufficient amountsbetween the control and the gauge to increase resistance toflow, the stage-discharge relationship is affected; however,ice may form so gradually that there is little indication ofits initial effects. Surface ice is the most common form andaffects station ratings more frequently than frazil ice oranchor ice. The major effect of ice on a rating curve is dueto backwater and may vary from day to day.

17-4. Water Surface Profiles

EM 1110-2-142031 Oct 97

17-3

described in the preceding section on rating curves, the systems. Appendix D, “River Modeling - Lessonsselection of the appropriate method requires some Learned” (EM 1110-2-1416), provides an overview ofevaluation of the physical system and the expected use of technical issues and modeling impacts that apply to profilethe information. The modeling methods are described in calculations. Stability of Flood Control Channels (USACEChapter 8 and are presented in EM 1110-2-1416. While 1990) provides case examples of stream stability problems,steady-flow water surface profiles are used in a majority of causes, and effects. While the focus is not on reservoirs,profile calculations, the unsteady flow aspects of reservoir the experience reflects the high flow conditions that are aoperation or the long-term effects of changes in sediment major concern with reservoir operation. And EM 1110-2-transport may require the application of methods that 4000 provides procedures for problem assessment andcapture those aspects. modeling. All of these documents should be reviewed prior

b. Further information. The Corps, and other agen-cies, have accumulated considerable experience with river

to formulating and performing technical studies.

EM 1110-2-142031 Oct 97

18-1

Chapter 18Real Estate and Right-of-Way Studies

18-1. Introduction

a. General. This chapter provides guidance on theapplication of hydrologic engineering principles to deter-mine real estate acquisition requirements for reservoirprojects. Topics include selection of the analysis method,potential problems, evaluation criteria, and referencesassociated with the acquisition of real estate for reservoirprojects developed by the Corps of Engineers.

b. Related documents. Real estate reporting require-ments associated with feasibility reports, General DesignMemoranda, and Real Estate Design Memoranda are setforth in ER 405-1-12. Real estate reporting requirementsassociated with the acquisition of lands downstream fromspillways are set forth in ER 1110-2-1451, paragraph 9.

18-2. Definition of Terms

A list of terms and definitions used in this chapter is asfollows:

a. Project design sediment. The volume and distri-bution of sediment deposited in a reservoir over the life ofthe project.

b. Land acquisition flood. A hypothetical orrecorded flood event used to determine requirements forreal estate acquisition.

c. Full pool. The maximum reservoir elevation forstoring water for allocated project purposes.

d. Induced surcharge. Storage created in a reservoirabove the top of flood control pool by regulating outflowsduring flood events.

e. Envelope curve. A curve which connects the highpoints of intersection of preproject and postproject water-surface profiles.

f. Guide taking line. A contour line used as a guidefor land acquisition in the reservoir area. (Also referred toas the guide contour line or guide acquisition line.)

18-3. Real Estate Acquisition Policies forReservoirs

a. Basic policies. Basic policies and proceduresrelated to the acquisition of lands for reservoirs are pre-sented in ER 405-1-12. Paragraph 2-12 of ER 405-1-12states that, “Under the Joint Policy the Corps will take anadequate interest in lands, including areas required forpublic access, to accomplish all the authorized purposes ofthe project and thereby obtain maximum public benefitstherefrom.” Paragraph 2-12.a(2) further states that land tobe acquired in fee shall include, “lands below a guidecontour line...established with a reasonable freeboardallowance above the top pool elevation for storing water forflood control, navigation, power, irrigation, and otherpurposes, referred to in this paragraph as “full pool” ele-vation. In nonurban areas generally, this freeboardallowance will be established to include allowances forinduced surcharge operations plus a reasonable additionalfreeboard to provide for adverse effects of saturation, waveaction and bank erosion.”

b. Considering factors. Factors such as estimatedfrequency of occurrence, probable accuracy of estimates,and relocation costs will be taken into consideration.Where freeboard does not provide a minimum of 300 fthorizontally from the conservation pool, defined as the topof all planned storage not devoted exclusively to flood-control storage, then the guide acquisition line will beincreased to that extent. In the vicinity of urban com-munities or other areas of highly concentrated develop-ments, the total freeboard allowance between the full poolelevation and the acquisition line may be greater thanprescribed for nonurban areas generally. Also, there shouldbe sufficient distance to assure that major hazards to life orunusually severe property damages would not result fromfloods up to the magnitude of the SPF. In suchcircumstances, however, consideration may be given toeasements rather than fee acquisition for select sections iffound to be in the public interest. However, when theproject design provides a high level spillway, the crest ofwhich for economy of construction is considerably higherthan the storage elevation required to regulate the reservoirdesign flood, the upper level of fee acquisition willnormally be at least equal to the top elevation of spillwaygates or crest elevation of ungated spillway, and mayexceed this elevation if necessary to conform with othercriteria prescribed.

EM 1110-2-142031 Oct 97

18-2

18-4. Hydrologic Evaluations

a. Development of land acquisition flood. Toestablish a reasonable surcharge allowance above the toppool elevation, a land acquisition flood, which includes theeffects of any upstream reservoirs, should be selected androuted through the project to determine the impact on theestablishment of the guide acquisition line.

b. Nonurban areas. In nonurban areas, the landacquisition flood should be selected from an evaluation of arange of floods with various frequencies of occurrence.The impact of induced surcharge operations on existingand future developments, hazards to life, land use, andrelocations must be evaluated. The land acquisition floodwill be chosen based on an evaluation of the risk anduncertainty associated with each of these frequency events.Basic considerations to be addressed during the landacquisition flood selection process should include thecredibility of the analysis, identification and significance ofrisk, costs and benefits, and legal, social, and politicalramifications.

c. Urban areas. In urban areas or other areas withhighly concentrated areas of development, the SPF will beused for the land acquisition flood.

d. Project design sedimentation volume. Projectcapacity data should be adjusted for projected sedimentvolumes when routing the land acquisition flood. Projectdesign sediment should be based on appropriate rates ofsedimentation for the project area for the life of the project.

18-5. Water Surface Profile Computations

a. Backwater model development. A basic backwatermodel should be developed for the project area from theproposed flat pool area through the headwater area whereimpacts of the proposed reservoir are expected to besignificant. The model should reflect appropriate cross-sectional data and include parameters based on historicalflood discharges and high water marks. EM 1110-2-1416presents the model requirements and calibrationprocedures.

b. Preproject profiles. A series of preproject watersurface profiles should be developed utilizing preprojectcross-section geometry, calibrated Manning's “n” values,and appropriate starting water surface elevations for theinitial cross section. Flow rates used in the water surfaceprofile computations should be selected from the peak andrecession side of the land acquisition flood hydrograph.

c. Postproject profiles. A series of water surfaceprofiles shall be developed utilizing the postproject crosssections which are adjusted to reflect project designsedimentation over the life of the project. Manning'sroughness coefficients are based on adjusted preprojectroughness coefficients to account for factors such as vege-tation and land use changes which decrease hydraulic con-veyance. Agricultural lands existing in the headwater areasprior to land purchases will likely revert to forested areassome years after the reservoir is filled. Preproject flowrates and coincident reservoir pool elevations from landacquisition flood routing should be used to computepostproject profiles.

d. Project design sedimentation distribution. Post-project cross-section geometry must be adjusted to reflectthe impacts of sedimentation over the life of the project.Sedimentation problems associated with reservoir projectsand methods of analysis to address sediment volumes anddistributions are given in Chapter 5 of EM 1110-2-4000.

18-6. Development of an Envelope Curve

The development of an envelope curve is based on pre-project and postproject water-surface profiles. A selecteddischarge from the land acquisition flood is used to com-pute a preproject and a postproject profile. A point ofintersection is established where the profiles are within 1 ftof each other. The point of intersection is placed at theelevation of the higher of the two profiles. A series ofpoints of intersection are derived from water-surfaceprofile computations utilizing a range of selected dis-charges from the land acquisition flood. A curve is drawnthrough the series of points of intersection to establish theenvelope curve.

18-7. Evaluations to Determine Guide TakingLines (GTL)

a. Land acquisition flat pool. The land acquisitionflat pool of a reservoir project is established by the maxi-mum pool elevation designated for storing water for allo-cated project purposes to include induced surcharge storageand is not impacted by the backwater effects of mainstream or tributary inflows. In flat pool areas, the eleva-tions of the GTL are based on the flat pool elevation and afreeboard allowance to account for adverse effects ofsaturation, bank erosion, and wave action.

b. Headwater areas. In headwater areas, the GTLmay be based on the envelope curve elevations andappropriate allowances to prevent damages associated withsaturation, bank erosion, and wave action.

EM 1110-2-142031 Oct 97

18-3

c. Flood-control projects. The selection of an Register, dated 22 February 1962, Volume 27, page 1734.appropriate land acquisition flood for flood-control projects On July 1966, the Joint Policy was again published in 31,located in rural areas should be based on an elevation of aF.R. 9108, as follows:range of frequency flood events. The land acquisitionflood selection for flood-control projects in rural locations JOINT POLICIES OF THEmust include regulation by upstream reservoirs and reflect DEPARTMENTS OF THE INTERIORpostproject conditions which minimize adverse impacts AND OF THE ARMY RELATIVE TOwithin the project area resulting from induced flood RESERVOIR PROJECT LANDSelevations and duration of flooding. In highly developedareas along the perimeter of flood-control projects, the SPF“A joint policy statement of the Department of the Interiorshould be used for land acquisition. An envelope curve canand the Department of the Army was inadvertently issuedbe developed from the land acquisition flood routings andas a notice in 27 F.R. 1734. Publication should have beenwater-surface profile computations for preproject and made as a final rule replacing regulations then appearingpostproject conditions. The land acquisition GTL may be in 43 CFR part 8. The policy as it appears in 27 F.R. 1734established from the envelope curve and appropriatehas been the policy of the Department of the Interior andallowances for reservoir disturbances. the Department of the Army since its publication as a

d. Nonflood-control projects. Nonflood controlprojects may be any combination of purposes such as water Sectionsupply, hydropower, recreation, navigation and irrigation.The land acquisition flood selection process for nonflood- 8.0 Acquisition of lands for reservoircontrol projects located in rural areas is based on an projectsevaluation of a range of frequency floods and is used todetermine postproject flood elevations and duration of 8.1 Lands for reservoir construction andflooding in the project area. As with flood-control projects, operationregulation of flows by upstream reservoirs must beincorporated in the development process. The land 8.2 Additional lands for correlative purposesacquisition flood used to evaluate real estate acquisitions inrural areas should reflect postproject conditions which 8.3 Easementsminimize adverse impacts. The land acquisition flood fordeveloped areas should be the SPF. The maximum pool 8.4 Blocking outelevation designed for storing water for allocated projectpurposes is used in the development of the land acquisition 8.5 Mineral rightsflood routing. An envelope curve based on preproject andpostproject water-surface profiles utilizing project design 8.6 Buildingsedimentation and distribution should be developed. Theenvelope curve and appropriate allowances for reservoir Authority: The provisions of this Part 8disturbances may be used to establish the land acquisition issued under Sec. 7, 32 Stat., 389, Sec. 14, 53GTL. Stat. 1197, 43 U.S.C. 421, 389.

18-8. Acquisitions of Lands for ReservoirProjects

Land acquisition policies of the Department of the Armygoverning acquisition of land for reservoir projects ispublished in ER 405-1-12, Change 6, dated 2 January 1979.Paragraph is as follows:

Joint Land Acquisition Policy for Reservoir Projects.The joint policies of the Department of the Interior andDepartment of the Army, governing the acquisition of landfor reservoir projects, are published in the Federal

Notice and is now codified as set forth below.

8.0 Acquisition of Lands for Reservoir Projects.Insofar as permitted by law, it is the policy of the Depart-ments of the Interior and of the Army to acquire, as a partof reservoir project construction, adequate interest inlands necessary for the realization of optimum values forall purposes including additional land areas to assure fullrealization of optimum present and future outdoor recrea-tional and fish and wildlife potentials of each reservoir.

8.1 Lands for Reservoir Construction and Operation.The fee title will be acquired to the following:

EM 1110-2-142031 Oct 97

18-4

a) Lands necessary for permanent structures. 8.6 Buildings. Buildings for human occupancy as

b) Lands below the maximum flowage line of the operation of the project for any project purpose will bereservoir including lands below a selected freeboard where prohibited on reservoir project lands."necessary to safeguard against the effects of saturation,wave action, and bank erosion and to permit inducedsurcharge operation.

c) Lands needed to provide for public access to themaximum flowage line, as described in Paragraph 1b, orfor operation and maintenance of the project.

8.2 Additional Lands for Correlative Purposes. Thefee title will be acquired for the following:

a) Such lands as are needed to meet present and futurerequirements for fish and wildlife as are determined pursuant to the Fish and Wildlife Coordination Act.

b) Such lands as are needed to meet present and futurepublic requirements for outdoor recreation, as may beauthorized by Congress.

8.3 Easements. Easements in lieu of fee title may betaken only for lands that meet all of the followingconditions:

a) Lands lying above the storage pool,

b) Lands in remote portions of the project area,

c) Lands determined to be of no substantial value forprotection or enhancement of fish and wildliferesources, or for public outdoor recreation,

d) It is to the financial advantage of the Government totake easements in lieu of fee title.

8.4 Blocking Out. Blocking out will be accomplishedin accordance with sound real estate practices, for exam-ple, on minor sectional subdivision lines: and normally,land will not be acquired to avoid severance damage if theowner will waive such damage.

8.5 Mineral Rights. Mineral, oil and gas rights willnot be acquired except where the development thereofwould interfere with project purposes, but mineral rightsnot acquired will be subordinated to the Government'sright to regulate their development in a manner that willnot interfere with the primary purposes of the project,including public access.

well as other structures which would interfere with the

18-9. Acquisition of Lands Downstream fromSpillways for Hydrologic Safety Purposes

a. General. A real estate interest will be acquireddownstream of dam and lake projects to assure adequatesecurity for the general public in areas downstream fromspillways. Real estate interests must be obtained fordownstream areas where spillway discharges create orsignificantly increase a hazardous condition.

b. Evaluation criteria. Combinations of floodevents and flood conditions which result in a hazardouscondition or increase the hazard from the preproject topostproject flood conditions are determined for areasdownstream from the spillway. These combinations offlood events and flood conditions are identified as criticalconditions.

c. Flood events and conditions. Flood events up tothe magnitude of the spillway design flood are evaluatedfor preproject and postproject conditions for areas down-stream from the spillway. Flood conditions to be analyzedinclude flooded area, depth of flooding, duration,velocities, debris, and erosion.

d. Hazardous and nonhazardous conditions. Theimposed critical conditions are analyzed to determine ifthese conditions are hazardous or nonhazardous. Non-hazardous areas are characterized by the following criteria:

(1) Flood depths do not exceed 2 ft in urban and ruralareas.

(2) Flood depths are essentially nondamaging tourban property.

(3) Flood durations do not exceed 3 hr in urban areasand 24 hr in agricultural areas.

(4) Velocities do not exceed 4 fps.

(5) Debris and erosion potential are minimal.

(6) Imposed flood conditions would be infrequent.The exceedance frequency should be less than 1 percent.

EM 1110-2-142031 Oct 97

A-1

Appendix AReferences

A-1. Required Publications

Flood Control Act of 1944.

National Dam Safety Act, Public Law 92-367.

Water Resources Development Act of 1976.

Water Supply Act of 1958.

ER 405-1-12Real Estate Handbook

ER 1110-2-240Water Control Management

ER 1110-2-1451Acquisition of Lands Downstream from Spillways forHydrologic Safety Purposes

ER 1110-8-2(FR)Inflow Design Floods for Dams and Reservoirs

EM 1110-2-1201Reservoir Water Quality Analyses

EM 1110-2-1406Runoff from Snowmelt

EM 1110-2-1411Standard Project Flood Determinations

EM 1110-2-1412Storm Surge Analysis and Design Water LevelDeterminations

EM 1110-2-1414Water Levels and Wave Heights for Coastal EngineeringDesign

EM 1110-2-1415Hydrologic Frequency Analysis

EM 1110-2-1416River Hydraulics

EM 1110-2-1417Flood Run-off Analysis

EM 1110-2-1419Hydrologic Engineering Requirements for Flood DamageReduction Studies

EM 1110-2-1602Hydraulic Design of Reservoir Outlet Works

EM 1110-2-1603Hydraulic Design of Spillways

EM 1110-2-1701Hydropower

EM 1110-2-2904Design of Breakwater and Jetties

EM 1110-2-3600Management of Water Control Systems

EM 1110-2-4000Sedimentation Investigations of Rivers and Reservoirs

ETL 1110-2-335Development of Drought Contingency Plans

ETL 1110-2-336Operation of Reservoir Systems

A-2. Computer Program/Document References

Fread 1989Fread, D. 1989. BREACH: An Erosion Model for EarthenDam Failures, User’s Manual, National Weather Service,Hydrologic Research Laboratory, Silver Spring, MD.

Hydrologic Engineering Center (HEC) 1971Hydrologic Engineering Center (HEC). 1971. HEC-4“Monthly streamflow simulation,” User's Manual,U.S. Army Corps of Engineers, Davis, CA.

HEC 1982cHEC. 1982c. HEC-5, “Simulation of flood control andconservation systems,” User's Manual, U.S. Army Corps ofEngineers, Davis, CA.

HEC 1982dHEC. 1982d. HYDUR, “Hydropower analysis using flow-duration procedures,” User's Manual, U.S. Army Corps ofEngineers, Davis, CA.

EM 1110-2-142031 Oct 97

A-2

HEC 1984 Lane 1990HEC. 1984. HMR52, “Probable maximum storm (eastern Lane. 1990. LAST, “Applied stochastic techniques,”United States),” User's Manual, U.S. Army Corps of User’s Manual, U.S. Bureau of Reclamation, Denver, CO.Engineers, Davis, CA.

HEC 1986HEC. 1986. HEC-5(Q), “Simulation of flood control and analysis of simulation for a multi-purpose reservoir sys-conservation systems,” Appendix on Water Quality Analy- tem,” Southwestern Division, Dallas, TX.sis, U.S. Army Corps of Engineers, Davis, CA.

HEC 1987aHEC. 1987a. STATS, “Statistical analysis of time series reservoir regulation,” User's Manual, North Pacificdata,” Input Description, U.S. Army Corps of Engineers, Division, Portland, OR.Davis, CA.

HEC 1990bHEC. 1990b. “Flood damage analysis package on themicrocomputer,” Installation and User's Guide, TrainingDocument 31, U.S. Army Corps of Engineers, Davis, CA.

HEC 1990c 1284.HEC. 1990c. HEC-1, “Flood hydrograph package,” User'sManual, U.S. Army Corps of Engineers, Davis, CA.

HEC 1990d reservoir flood control plan selection,” Research DocumentHEC. 1990d. HEC-2, “Water surface profiles,” User'sManual, U.S. Army Corps of Engineers, Davis, CA.

HEC 1991a Buchanan, T. J., and Somers, W. P. 1968. “Stage mea-HEC. 1991a. HEC-PRM, “Prescriptive reservoir model,program description,” User's Manual, U.S. Army Corps ofEngineers, Davis, CA.

HEC 1992c Buchanan, T. J., and Somers, W. P. 1969. “DischargeHEC. 1992c. HEC-FFA, “Flood flow frequency analysis,”User's Manual, U.S. Army Corps of Engineers, Davis, CA.

HEC 1993HEC. 1993. HEC-6, Scour and deposition in rivers andreservoirs,” User's Manual, U.S. Army Corps of Engineers,Davis, CA.

HEC 1995a DeVries, J. J., and Hromadka, T. V. 1993. “ComputerHEC. 1995a. HEC-DSS, “Users guide and utility programmanual,” U.S. Army Corps of Engineers, Davis, CA.

HEC 1995bHEC. 199b. UNET, “One-dimensional unsteady flowthrough a full network of open channels,” User's Manual,U.S. Army Corps of Engineers, Davis, CA.

USACE 1972USACE. 1972. SUPER, “Regulation simulation and

USACE 1991USACE. 1991. SSARR, “Model streamflow synthesis and

A-3. Related Publications

Alley and Burns 1983Alley, W. M., and Burns, A. W. 1983. “Mixed stationextension of monthly streamflow records,” Journal ofHydraulic Engineering, ASCE, HY109(10), October 1272-

Bowen 1987Bowen, T. H. 1987. “Branch-bound enumeration for

35, HEC, Davis, CA.

Buchanan and Somers 1968

surements at gauging stations,” U.S. Geological Survey,TWI 3-A7.

Buchanan and Somers 1969

measurements at gauging stations,” U.S. Geological Sur-vey, TWI 3-A8.

Carter and Davidian 1968Carter, R. W., and Davidian, J. 1968. “General procedurefor gauging streams,” U.S. Geological Survey, TWI 3-A6.

DeVries and Hromadka 1993

models for surface water,” Handbook of hydrology D. R.Maidment, ed., McGraw-Hill, New York.

Domenico and Schwartz 1990Domenico, P. A., and Schwartz, F. W. 1990. Physical andChemical Hydrogeology. John Wiley and Sons, New York.

EM 1110-2-142031 Oct 97

A-3

Eichert and Davis 1976 HEC 1985bEichert, B. S., and Davis, D. W. 1976. “Sizing flood HEC. 1985b. “Stochastic Analysis of Drought Phenom-control reservoir systems by systems analysis,” Technical ena,” Training Document 25, U.S. Army Corps ofPaper 44, Hydrologic Engineering Center, Davis, CA. Engineers, Davis, CA.

Farnsworth, Thompson, and Peck 1982 HEC 1990aFarnsworth, R. K., Thompson, E. S., and Peck, E. L. 1982. HEC. 1990a. “Modifying reservoir operations to improve“Evaporation atlas for the contiguous 48 United States,” capabilities for meeting water supply needs duringTechnical Report NWS 33, National Oceanic and droughts,” Research Document 31, U.S. Army Corps ofAtmospheric Administration (NOAA), Washington, DC. Engineers, Davis, CA.

Froelich 1987 HEC 1990eFroelich, D. C. 1987. “Embankment-dam breach para- HEC. 1990e. “A preliminary assessment of Corps ofmeters,” Proceedings of the 1987 National Conference onEngineers' reservoirs, their purposes and susceptibility toHydraulic Engineering at Williamsburg, VA. American drought,” Research Document 33, U.S. Army Corps ofSociety of Civil Engineers, New York. Engineers, Davis, CA.

Hoffman 1977 HEC 1991bHoffman, C. J. 1977. “Design of spillways and outlet HEC. 1991b. “Optimization of multiple-purpose reservoirworks,” Handbook of Dam Engineering A. R. Golzé, ed., system operations: A review of modeling and analysisVan Nostrand Reinhold, New York. approaches,” Research Document 34, U.S. Army Corps of

Hydrologic Engineering Center (HEC) 1977Hydrologic Engineering Center (HEC). 1977. “Guidelinesfor calculating and routing a dam break flood,” ResearchDocument 5, U.S. Army Corps of Engineers, Davis, CA.

HEC 1980HEC. 1980. “Flood emergency plan guidelines for Corpsdams,” Research Document 13, U.S. Army Corps ofEngineers, Davis, CA.

HEC 1982aHEC. 1982a. “Emergency planning for dams, biblio-graphy and abstracts of selected publications,” ResearchDocument 17, U.S. Army Corps of Engineers, Davis, CA.

HEC 1983aHEC. 1983a. “Example emergency plan for Blue MarshDam and Lake,” Research Document 19, U.S. Army Corpsof Engineers, Davis, CA.

HEC 1983bHEC. 1983b. “Example plan for evacuation of Reading,Pennsylvania, in the event of emergencies at Blue MarshDam and Lake,” Research Document 20, U.S. Army Corpsof Engineers, Davis, CA.

HEC 1985aHEC. 1985a. “Flood-damage-mitigation plan selectionwith branch-and-bound enumeration,” Training Document23, U.S. Army Corps of Engineers, Davis, CA.

Engineers, Davis, CA.

HEC 1991cHEC. 1991c. “Importance of surface-ground water inter-action to Corps total water management: Regional andnational examples,” Research Document 32, U.S. ArmyCorps of Engineers, Davis, CA.

HEC 1991dHEC. 1991d. “Missouri River System Analysis Model -Phase I,” Project Report 15, U.S. Army Corps ofEngineers, Davis, CA.

HEC 1991fHEC. 1991f. “Columbia River System Analysis Model -Phase I,” Project Report 16, U.S. Army Corps ofEngineers, Davis, CA.

HEC 1992aHEC. 1992a. “Missouri River System Analysis Model -Phase II,” Project Report 17, U.S. Army Corps ofEngineers, Davis, CA.

HEC 1992bHEC. 1992b. “Developing Operation Plans from HECPrescriptive Reservoir Model, Results for the MissouriSystem: Preliminary Results,” Project Report 18,U.S. Army Corps of Engineers, Davis, CA.

EM 1110-2-142031 Oct 97

A-4

HEC 1993 Pruitt 1990HEC. 1993. “Review of GIS Applications in Hydrologic Pruitt, W. O. 1990. Evaluation of Reservoir EvaporationModeling,” Technical Paper 144, U.S. Army Corps of Estimates, submitted to USACE, Sacramento District,Engineers, Davis, CA. Sacramento, CA.

Linsley et al. 1992 Salas 1992Linsley, R. K., Franzini, J. B., Freyberg, D. L., and Salas, J. D. 1992. “Analysis and modeling of hydrologicTchobanoglous, G. 1992. Water resources engineering, time series,” Handbook of Hydrology, Maidment, ed.4th ed., McGraw-Hill, New York.

MacDonald and Langridge-Monopolic 1984MacDonald, T. C., and Langridge-Monopolis, J. 1984. 1980. “Applied modeling of hydrologic time series,” Water“Breaching characteristics of dam failures,” Journal of Resources Publications, Littleton, CO, 484.Hydraulic Engineering, American Society of Civil Engi-neers, 110 (5), 567-586.

Matalas and Langbein 1962Matalas, N. C., and Langbein, W. 1962. “Information vey, TWI 3-A11.Content of the Mean,” in Journal of Geophysical Research,67(9), 3441-3448.

Mays and Tung 1992Mays, L. W., and Tung, Y-K. 1992. Hydrosystems engi-neering & management, McGraw-Hill, New York.

McGhee 1991McGhee, T. J. 1991. Water supply and sewerage, Report HL-85-1, U.S. Army Engineer Waterways Experi-6th Edition, McGraw-Hill, New York. ment Station, Vicksburg, MS.

Mosley and McKerchar 1993 USACE 1962Mosley, M. P., and McKerchar, A. I. 1993. “Streamflow,” USACE. 1962. “Waves in Inland Reservoirs: SummaryHandbook of Hydrology. D. R. Maidment ed., McGraw- Report on CWI Projects,” CW-164 and CW-165.Hill, New York. Washington, DC.

National Weather Service 1960 USACE 1979National Weather Service. 1960. “Generalized estimates USACE. 1979. National hydroelectric power study,of PMP for the U.S. west of the 105th Meridian for areas Institute of Water Resources, Washington, DC.less than 400 square miles and durations to 24 hr,” Tech-nical Paper 38, Silver Spring, MD.

National Weather Service 1977National Weather Service. 1977. “Probable maximumprecipitation estimates,” United States east of 105th Meri-dian, Hydrometeorological Report No. 51, Silver Spring,MD.

National Weather Service 1981National Weather Service. 1981. “Application of probablemaximum precipitation estimates - United States east of the105th meridian,” Hydrometeorological Report No. 52,Silver Spring, MD.

Salas, Delleur, Yevjevich, and Lane 1980Salas, J. D., Delleur, J. W., Yevjevich, V., and Lane, W. L.

Smoot and Novak 1969Smoot, G. F., and Novak, C. E. 1969. “Measurement ofdischarge by moving-boat method,” U.S. Geological Sur-

Tasker 1983Tasker, G. D. 1983. “Effective record length of T-yearevent,” Journal of Hydrology, 64, 39-47.

Thomas and McAnally 1985Thomas, W. A., and McAnally, W. H. 1985. “Open-Channel Flow and Seidmentation TABS2,” Instruction

USACE 1989USACE. 1989. Digest of water resources policies andauthorities, 1165-2-1, Washington, DC.

USACE 1990USACE. 1990. “Stability of flood control channels,” EC1110-8-1(FR), Washington, DC.

USACE 1992USACE. 1992. “Authorized and operating purposes ofCorps of Engineers' reservoirs,” Washington, DC.

EM 1110-2-142031 Oct 97

A-5

U.S. Geological Survey 1992 Water Resources Council 1973United States Geological Survey. 1992. “Sediment depo- Water Resources Council. 1973. “Principles and standardssition in U.S. Reservoirs - summary of data reported 1981- for planning water and related land resources,” The Federal85,” Interagency Advisory Committee on Water Data - Register, The National Archives of the United States,Subcommittee on Sedimentation. September 10, 1973.

Waide 1986 Wurbs 1991Waide, J. B. 1986. “General guidelines for monitoring Wurbs, R. A. 1991. “Optimization of multiple-purposecontaminants in reservoirs,” Instruction Report E-86-1, reservoir system operations: A review of modeling andU.S. Army Engineers Waterways Experiment Station, analysis approaches,” Research Document 34, HydrologicVicksburg, MS. Engineering Center, Davis, CA.