water supply needs and sources assessment...

Special Publication SJ97-SP7

WATER SUPPLY NEEDS AND SOURCES ASSESSMENTALTERNATIVE WATER SUPPLY STRATEGIES INVESTIGATION

SURFACE WATER AVAILABILITY AND YIELD ANALYSIS

by

CH2M Hill

St. Johns River Water Management DistrictPalatka, Florida

1997

Executive Summary

EXECUTIVE SUMMARYThis technical memorandum (TM) is the third in a series concernedwith the feasibility of developing selected surface water sources tohelp meet municipal water supply needs within the St. Johns RiverWater Management District. The first surface water supply TMaddressed data availability and development of the methodology to beused in the feasibility evaluation. The second TM addressed selectionof six candidate surface water withdrawal sites for quantitativeanalysis. This TM presents the results of the quantitative water supplyavailability and yield analysis.

The six candidate withdrawal sites include Lake Griffin on HainesCreek, located in Lake County, and five sites located on the main stemof the St. Johns River extending from Cocoa downstream toJacksonville. For each of the six candidate withdrawal sites, a similarseries of analyses was conducted.

WATER SUPPLY YIELDThe estimated maximum reliable municipal water supply yield foreach of the six candidate withdrawal sites is summarized below:

• Lake Griffin (Haines Creek) 28 mgd• St. Johns River near Cocoa 108 mgd• St. Johns River near Titusville 143 mgd• St. Johns River at Sanford (Lake Monroe) 279 mgd• St. Johns River near De Land 351 mgd• St. Johns River above Jacksonville 419 mgd

The maximum water supply yield estimates are based on applicationof the previously established surface water evaluation methodology.However, because planned SJRWMD minimum flows and levelsanalysis for Lake Griffin may result in different, and possibly morerestrictive, withdrawal criteria, only 50 percent of the calculatedmaximum yield, or 14 mgd, will be considered in subsequent areawidealternative water supply evaluations.

Maximum reliable yields for the Lake Griffin and St. Johns River sitesare independent as they are relatively independent hydrologicsystems. Thus, water supply development on Lake Griffin will notaffect the potential for water supply development on the St. JohnsRiver.

Surface Water: Availability and Yield Analysis

Hi

Executive Summary

However, the maximum yield values for the individual St. Johns Riversites are not independent. These values represent the cumulativeamount for each individual site and all upstream sites. For example, ifa 100-mgd reliable water supply were developed near Titusville, thenthe maximum reliable yield at De Land (or another downstream site)would be reduced by 100 mgd.

TREATMENT REQUIREMENTSFacilities required to develop the surface water sources include a riverdiversion structure, an off-line raw water reservoir, a water treatmentplant, and an aquifer storage recovery system. Treatmentrequirements vary considerably by location. For example, the LakeGriffin site is the only true freshwater site, while the St. Johns Riversites will require some desalting facilities, likely reverse osmosismembrane treatment, to provide the required product water quality.

The four upstream St. Johns River sites provide raw water that can beclassified as either slightly or moderately brackish under certain flowconditions. The most downstream site, the St. Johns River aboveJacksonville, is tidal and has poor water quality characteristics. Rawwater at this site is classified as saline, which would require extensivedesalting facilities and would generate large quantities of wasteconcentrate.

In conclusion, five of the six candidate water supply withdrawal sitesare technically viable municipal water supply sources and should beconsidered in subsequent phases of the alternative water supplyinvestigations. The viable sites include Lake Griffin and the fourupstream sites located on the main stem of the St. Johns River, fromnear Cocoa to near De Land. The most downstream site, the St. JohnsRiver above Jacksonville, does not provide a viable municipal watersupply source and should not be considered further.


iv

Table of Contents

EXECUTIVE SUMMARY iiiWATER SUPPLY YIELD iiiTREATMENT REQUIREMENTS ivFIGURES viiTABLES ix

INTRODUCTION 1

METHODS 3FACTORS AFFECTING SURFACE WATER SUPPLY DEVELOPMENT 3

Streamflow Characteristics 3Minimum Streamflow Requirements 4Demand Characteristics 5Required System Reliability 5

GENERAL FACILITIES REQUIREMENTS 6River Diversion Structure 6Raw Water Storage Reservoir 6Water Treatment Plant 8ASR Systems 9

SIMULATION OF CONTINUOUS WATER SUPPLY SYSTEMS 10Overview and Application 10Simulation Logic 11

SELECTED CANDIDATE WITHDRAWAL SITES 12

STREAMFLOW DURATION ANALYSIS 16FLOW DURATION CURVES AND MINIMUM STREAMFLOW REQUIREMENTS 16

Haines Creek 16St. Johns River near Cocoa 18St. Johns River near Tirusville 18St. Johns River at Sanford 18St. Johns River near De Land 21St. Johns River above Jacksonville 21

EFFECT OF WATER SUPPLY WITHDRAWAL ON FLOW DURATION RELATIONSHIPS 25

POTENTIAL YIELD ANALYSIS 33

SURFACE WATER SUPPLY FACILITY REQUIREMENTS 42TRIAL FACILITY COMPONENTS 42

Raw Water Diversion and Pumping Station 42Off-line Raw Water Storage Reservoir 42Water Treatment Plant 44Treated Water ASR System 45

FACILITY REQUIREMENTS SIMULATION RESULTS 45WATER QUALITY AND TREATMENT REQUIREMENTS 46RELIABLE NET YIELD AND FACILITY REQUIREMENTS 52


V

Table of Contents

PROPOSED COST ESTIMATION PROCEDURE 60SITES CONSIDERED 60COST PARAMETERS AND CRITERIA 60CONSTRUCTION AND O&M COST COMPONENTS 61LAND COSTS 62OTHER COST ESTIMATES 62COST FUNCTIONS 62TECHNICAL MEMORANDUM 63

SUMMARY AND RECOMMENDATIONS 64SUMMARY 64

Flow Duration Analysis 64Potential Yield Analysis 65Surface Water Supply Facility Requirements 66

RECOMMENDATIONS 68

REFERENCES 70


vi

Table of Contents

FIGURES1 Facilities Required to Develop Reliable Surface Water Supply

to Meet Urban Demands 7

2 Location of the Candidate Surface Water Withdrawal Sitesin Lake County 13

3 Locations of Candidate Surface Water Withdrawal Siteson the Main Stem of the St. Johns River 14

4 Flow Duration Curve for Haines Creek near Lisbon, Florida... 17

5 Flow Duration Curve for the St. Johns River near Cocoa,Florida 19

6 Flow Duration Curve for the St. Johns River near Titusville(Christmas), Florida 20

7 Approximate Flow Duration Curve for the St. Johns Rivernear Sanford, Florida 22

8 Flow Duration Curve for the St. Johns River near De Land,Florida 23

9 Flow Duration Curve for the St. Johns River aboveJacksonville, Florida 24

10 Maximum Effect of Water Supply Withdrawal on FlowDuration Curve for Haines Creek at Lisbon, Florida 26

11 Maximum Effect of Water Supply Withdrawal on FlowDuration Curve for the St. Johns River near Cocoa, Florida 27

12 Maximum Effect of Water Supply Withdrawal on FlowDuration Curve for the St. Johns River near Titusville(Christmas), Florida 28

13 Maximum Effect of Water Supply Withdrawal on Flow DurationCurve for the St. Johns River at Sanford, Florida 29

14 Maximum Effect of Water Supply Withdrawal on Flow DurationCurve for the St. Johns River at De Land, Florida 30

15 Maximum Effect of Water Supply Withdrawal on FlowDuration Curve for the St. Johns River above Jacksonville,Florida 31


vii

Table of Contents

16 Potential Yield Curve for Haines Creek at Lisbon, Florida 35

17 Potential Yield Curve for the St. Johns River near Cocoa,Florida 36

18 Potential Yield Curve for the St. Johns River near Titusville(Christmas), Florida 37

19 Potential Yield Curve for the St. Johns River at Sanford,Florida 38

20 Potential Yield Curve for the St. Johns River nearDe Land, Florida 39

21 Potential Yield Curve for the St. Johns River aboveJacksonville, Florida 40

22 Surface Water Facility Requirements for Lake Griffin(Haines Creek) 54

23 Surface Water Facility Requirements for the St. Johns Rivernear Cocoa 55

24 Surface Water Facility Requirements for the St. Johns Rivernear Titusville 56

25 Surface Water Facility Requirements for the St. Johns Riverat Sanford 57

26 Surface Water Facility Requirements for the St. Johns Rivernear De Land 58

27 Surface Water Facility Requirements for the St. Johns Riverabove Jacksonville 59


Viii

Table of Contents

TABLES1 Maximum Potential Yield for the Six Candidate Surface

Water Withdrawal Sites 41

2 Water Quality Characteristics at Candidate WithdrawalPoints 47

3 Major Water Treatment Processes Required for EachCandidate Surface Water Withdrawal Site 49

4 Surface Water Facility Requirements Summary for Rangeof Total Reliable Net Yield 53

5 Water Supply Availability and Yield Analysis ResultsSummary 67

Surface Water: Availability and "Yield Analysis

ix

Introduction

INTRODUCTIONThis technical memorandum (TM) is the third in a series concernedwith the feasibility of developing selected surface water sources tohelp meet municipal water supply needs within the St. Johns RiverWater Management District (SJRWMD). The first surface water supplyTM addressed data availability and development of the methodologyto be used in the feasibility evaluation (CH2M HILL, 1996a). Thesecond TM addressed selection of six candidate surface waterwithdrawal sites for quantitative evaluation (CH2M HILL, 1996b).This TM presents the results of the quantitative evaluation of the sixpreviously selected candidate withdrawal sites using the establishedevaluation procedure.

The six candidate withdrawal sites include Lake Griffin on HainesCreek, located in Lake County, and five sites located on the main stemof the St. Johns River extending from Cocoa downstream toJacksonville. For each of the six candidate withdrawal sites, thefollowing series of analyses was performed:

• Development of flow duration curves

• Determination of minimum streamflow requirements and themaximum water supply withdrawal rate

• Evaluation of the effect of water supply withdrawal on the flowduration relationship

• Evaluation of the potential water supply yield as a function ofmaximum installed withdrawal capacity

• Determination of streamflow water quality characteristics andwater treatment requirements

• Estimation of the type and size of water supply facilities requiredas a function of the total reliable water supply yield developed

The analysis is primarily based on available streamflow and waterquality records. First, flow duration curves for each site weredeveloped. Then, minimum streamflow requirements and themaximum withdrawal rate were determined for each site according toflow duration curve characteristics. Once these parameters weredetermined, the maximum effect of water supply withdrawal on theexisting streamflow duration was determined and reported for eachcandidate withdrawal site.


Introduction

The remainder of the feasibility analysis focused on determiningsurface water supply facilities requirements. The first step in thisprocess was developing the potential yield curve for each candidatewithdrawal site. The potential yield is defined as the water supplyyield that could be developed if adequate storage and treatmentfacilities are provided and no waste is produced by the treatmentprocess. This curve defines the relationship between the installedstreamflow diversion capacity and the long-term average flow ratethat can be diverted for water supply purposes. Also, this curveestablishes the maximum water supply that could be developed as afunction of the maximum streamflow diversion rate.

Once the potential yield curve was developed, trial water supplyfacilities were selected. Long-term water supply systems were thensimulated to evaluate the reliability of the trial water supply systemsto meet a variety of target yields. By doing so, the facilities required tomeet a projected long-term demand at the desired reliability wereestablished.

Preliminary water treatment process requirements were determinedfrom the water quality characteristics observed at each candidatewithdrawal site. Once conceptual treatment requirements wereestablished, the net reliable yield for each trial water supply facilitywas estimated. The final results are relationships between net reliablewater supply yield and facility requirements for each candidatewithdrawal site.

This TM concludes with a proposed procedure for developingplanning-level cost estimates for the required surface water supplyfacilities. Planning-level water supply facility cost estimates andplanning-level cost functions will be the subject of the fourth and finalsurface water supply TM. These cost estimates will then be availablefor use in the University of Florida Decision Model, which willcompare the surface water supply alternative with other water supplystrategies.


Methods

METHODSThe methods applied to the surface water supply availability and yieldanalysis are fully documented in TM B.l.f, Surface Water DataAcquisition and Evaluation Methodology (CH2M HILL, 1996a). Thissection presents a summary of the methodology established in thatTM.

FACTORS AFFECTING SURFACE WATER SUPPLYDEVELOPMENT

Several factors affect the technical feasibility of developing a surfacewater supply source, such as streamflow characteristics, minimumstreamflow requirements, water supply demand characteristics, andrequired system reliability.

Streamflow Characteristics

Important streamflow characteristics include streamflow magnitude,streamflow variability, and water quality. The magnitude of thestreamflow, which is defined as the long-term average discharge rateor watershed yield, establishes the total water resource available. Ingeneral, the larger the streamflow magnitude, the greater the potentialfor significant water supply development.

Streamflow variability defines the day-to-day, month-to-month, andyear-to-year variation of the total streamflow. Variability can be asimportant as magnitude in determining the economic feasibility ofdeveloping a surface water supply source. This is because a highlyvariable source is more difficult to develop than a source with littlevariability. In general, variable sources require larger storage facilitiesand are therefore more expensive to develop than sources with lessvariable streamflow regimes.

The streamflow characteristics of magnitude and variability aredefined by observed streamflow records. Fortunately for this analysis,long-term streamflow records are available at, or near, each of the sixcandidate withdrawal sites. Daily streamflow records are used todevelop flow duration curves, which illustrate the inherent variabilityof the source.

The long-term monthly streamflow array for each candidatewithdrawal site is used directly in the system simulation analysis.


Methods

Therefore, streamflow magnitude and variability are explicitlyaccounted for throughout this surface water supply feasibility analysis.

Water quality records are used to define the characteristics of the rawwater available for treatment. These data are used to establish theappropriate treatment processes required to deliver high-qualityproduct water. Treatment requirements greatly influence the cost ofdeveloping each individual candidate water supply site. For example,if desalting is required, treatment facilities will be relatively moreexpensive and the reliable yield will be reduced because part of thediverted flow will become waste concentrate.

Useful water quality data were available at or near each of the sixcandidate withdrawal sites.

Minimum Streamflow Requirements

Minimum streamflow requirements for streams and rivers withinSJRWMD will be defined, in the future, by the ongoing minimumflows and levels program. However, at this time, minimum flows andlevels have not been established for any of the water bodies orwithdrawal sites considered in this evaluation. Thus, approximateminimum streamflow requirements were established for thispreliminary evaluation of surface water supply feasibility. Onceminimum flows and levels are established for the candidatewithdrawal sites considered in this TM, the feasibility of developing aviable surface water supply should be re-evaluated using the site-specific minimum flows and levels criteria established by SJRWMD.

For this effort, minimum streamflow requirements are defined as thepositive streamflow rate observed 95 percent of the time. This meansthat at least 5 percent of the time (during low-flow periods) watersupply withdrawal will not be allowed. In addition, the maximumwithdrawal rate considered is equal to 25 percent of the long- termaverage flow rate. Lesser withdrawal rates are also considered.

Consider, for example, a stream with positive flow 97 percent of thetime and a mean flow rate of 500 cubic feet per second (cfs). Theallowable diversion frequency would be 92.2 percent of the time(0.95 x 0.97 = 0.922), and the maximum diversion rate would be 125 cfs(0.25x500 = 125).


Methods

Demand Characteristics

The characteristics of the projected demand to be met by the watersupply system also influences the water supply facilities required tomeet the demand. For this effort, only municipal demands areconsidered. Demand characteristics of interest include the long-termaverage demand to be met, or average daily demand (ADD), themaximum daily demand to be met (MDD), and the seasonaldistribution of demands. The seasonal distribution of demands isdefined by the ratio of monthly demands to the ADD.

In this feasibility analysis, the targeted yields are varied over a rangecompatible with the expected water supply yield for a given candidatewithdrawal site. The objective is to define facilities requirements as afunction of demand met. The monthly demand ratios previouslyestablished for this analysis are based on the City of Cocoa operationalexperience reported below:

Month Demand RatioJan 0.868Feb 0.919Mar 1.059Apr 1.127May 1.149Jun 1.070Jul 1.084

Aug 1.067Sep 1.002Oct 0.944Nov 0.892Dec 0.879

In addition, the MDD is assumed to be equal to 1.5*ADD.

Required System Reliability

Domestic water supply systems must be highly reliable, in that theymust be able to supply the desired quantity and quality of water for ahigh percentage of the time. In most cases, however, an inability tomeet the desired target yield, or target yield deficiency, meansproviding only a portion of the desired yield or providing water thatdoes not fully meet all desired quality criteria. When a yielddeficiency occurs, implementation of water use restrictions is morelikely to occur than a complete lack of supply.


Methods

For the purpose of this preliminary feasibility analysis, the previouslyestablished water supply system reliability target is 98.3 percent. Thisvalue is based on the allowance of an average of one monthly targetyield deficiency once every 5 years.

GENERAL FACILITIES REQUIREMENTSThe facilities required to develop a safe, reliable surface water supplyinclude some combination of the following components (Figure 1):

• River diversion structure• Raw water storage• Water treatment plant• Aquifer storage recovery (ASR)

Under favorable conditions, including a high-volume, low-variabilitysource and limited water supply needs, the required water supplysystem may be developed with only a river diversion structure and awater treatment plant. However, in most situations, some type ofstorage will be required to provide the system reliability.

Raw water will likely be available for diversion, in quantities adequateto meet the desired yield, for only a portion of the time. Storagefacilities, including either raw water storage reservoirs or ASRsystems, can be used to store water when it is available for later usewhen it is needed. Storage provides the flow attenuation necessary tomatch a variable water supply source to a variable water supplydemand.

River Diversion Structure

A river diversion structure consists of a raw water intake and apumping station. In most cases, some type of coarse screen or barscreen is provided to prevent damage to the pumps or otherdownstream treatment equipment. The diversion pumping stationcapacity (Qd) must be sized to allow diversion of the necessary volumeof water, which is subject to withdrawal constraints defined byminimum streamflow requirements and maximum allowablediversion rates.

Raw Water Storage Reservoir

There are two types of raw water storage reservoirs: on-streamreservoirs and off-line reservoirs. Development of on-streamreservoirs requires construction of a dam across the stream, flooding a


Raw Water Diversion andPumping Station

\XJ\

Finished WaterASR System

Off-Line RawWater Storage

Reservoir

Distribution to Meet Demand

Figure 1. Facilities Required to Develop Reliable Surface Water Supply to Meet Urban Demands.

Methods

portion of the upstream valley and subsequently providing therequired water supply storage. Off-line reservoirs, which areconstructed adjacent to the free-flowing stream, are filled by pumpingdivertable streamflow into the reservoir. The off-line reservoir isusually built by constructing a levee around the perimeter of thereservoir site. The storage volume provided is then a function of thearea enclosed and the depth to which water can be impounded.

Both on-stream and off-line reservoirs receive additional inflow fromdirect rainfall and incur water losses through lake evaporation. Undercertain conditions, additional water could also be lost by seepage.

In Florida, construction of on-line reservoirs is difficult because streamvalleys are wide and favorable dam sites are rare. In addition, theconstruction of on-line reservoirs greatly impacts the natural flowregime of a stream, often flooding productive adjacent wetlands. Incontrast, off-line reservoirs do not interfere with the naturalstreamflow regime and, therefore, are less environmentally disruptivethan on-line reservoirs. However, off-line reservoirs are usuallylocated on or near floodplains, where they may impact wetlands.

For this preliminary feasibility analysis, only off-line reservoirs will beconsidered and their use will be minimized. Priority will be given tousing ASR as the primary storage method; however, off-line raw waterstorage will be used to provide a buffer between the river diversionstructure and the water treatment plant, providing some flexibility inoperation. The presence of a raw water storage reservoir will allowthe plant to operate during short periods when streamflow diversion isnot allowed because of minimum streamflow requirements.

A minimum off-line raw water storage reservoir that provides 5 daysof operational storage is included in this analysis.

Water Treatment Plant

The purpose of the water treatment plant is to provide a safe, potablefinish water that meets all necessary drinking water standards. If theraw water is of reasonably high quality, then conventional treatment isusually all that is necessary. For surface water sources, conventionaltreatment usually consists of some type of clarification and filtrationwith disinfection. If the raw water is of poor quality, including a highdissolved minerals content, membrane treatment may also be required,which produces a waste concentrate. Therefore, addition of membrane


8

Methods

ASR Systems

treatment would result in reduction of the net water supply yield, aswell as additional requirements for a concentrate disposal system.

The quality characteristics of the raw water define treatmentrequirements and, in part, treatment costs. However, the demands tobe met and the amount and type of storage provided will define thetreatment capacity (Qt) requirements. If no finish water storage isprovided, by an ASR system for example, then the treatment plantmust be sized to meet the maximum day demands (Qt = MDD). Or, iffinish water storage is provided, the treatment plant can be somewhatsmaller because maximum day demands can be met from finish waterstorage.

In general, ASR systems can be used to store both raw water andtreated finish water (Pyne, 1995). In raw water applications, the ASRsystem could replace the off-line raw water storage reservoir discussedpreviously. However, in most water supply applications implementedto date, ASR has been used to provide treated water storage, so onlytreated water ASR is being considered for SJRWMD. Water processedby the water treatment plant and not needed at the time of treatment isinjected into a suitable storage aquifer for later recovery anddistribution. In general, the recovered water is re-disinfected, but noadditional treatment is required.

ASR involves injecting water to be stored into a suitable aquifer. Thenative ground water is displaced by the injected water, which is thenavailable for recovery when needed. However, some inefficiencies andlosses occur, which prevent all of the water injected from ultimatelybeing recovered and used. As water is injected, some of it mixes withthe native ground water. Depending on the mixing characteristics ofthe aquifer and the quality of both the injected water and nativeground water, only a portion of this mixture can be recovered beforethe water quality is unacceptable for the intended purpose.

The mixing characteristics of the storage aquifer and water quality ofthe native ground water are not usually as restrictive in ASRapplications as they might first appear, if the ASR system is developedand operated properly. Even if the native ground water quality ispoor and considerable initial mixing occurs, a viable ASR system canstill usually be developed by injecting an initial volume of treatedwater to develop a buffer between the native ground water and treatedinjected water. Once developed, the buffer will allow good recovery


Methods

efficiencies if the injected water is not recovered, but is allowed toremain in the buffer.

SIMULATION OF CONTINUOUS WATER SUPPLY SYSTEMSDirect calculation of requirements for surface water supply facilitiesfor a given set of conditions is not possible because of the complexityof the system and the large number of interacting factors. Facilityrequirements must be determined on a trial-and-error basis using astructured, continuous simulation approach.

Overview and Application

The water supply systems simulation was designed to simulate thelong-term operation of a trial water supply system subject to a givenset of monthly demands and to track the performance of the system, asmeasured in terms of its reliability or ability to meet demands. Thebasic approach that was used defines a number of trial water supplysystems using appropriate components, as defined in Figure 1. Severalsets of monthly target yield arrays (small-to-large) were alsoestablished.

Each trial water supply system was then evaluated by the simulationrelative to its ability to deliver the desired yields. The reliability of thetrial system is tracked for each target yield array simulated. In thismanner, relationships between facility size and water supply yield forthe given system reliability, were developed. This is the basicapproach used previously by CH2M HILL to evaluate surface watersupply facilities requirements for the Peace River Water SupplySystem (CH2M HILL, 1985,1987,1993; Wycoff, 1985) and for theFlorida Lower East Coast water supply planning project(CH2M HILL, 1994).

The procedure involves multiple, long-term simulations. In a givenapplication, for each candidate withdrawal site, six trial water supplysystems were identified and ten target yield arrays were defined. Sixtysimulation runs were used to fully define facility requirements, yield,and reliability relationship. Some applications involved more than onecomplete iteration because the initially defined target yield levelsproved to be inappropriate once initial results from the simulationwere available.

The simulation uses a monthly time step, which is the appropriatelevel of detail for preliminary surface water supply planning purposes


10

Methods

(McMahon, 1992). The length of simulation varied, depending on thestreamflow records used. In general, whole calendar years of monthlystreamflow records were used.

Simulation Logic

The simulation was constructed around a flow distribution logic thatdefines how the water supply system will operate and providescriteria defining how a given monthly demand will be met, based onthe monthly divertable streamflow, available facilities, and previouslystored water. The flow distribution logic is defined as follows:

• Condition A. Monthly divertable river flow is greater than or equalto monthly target yield

- Treat diverted flow and distribute to meet desired yield.

- Treat and inject into the ASR system remaining divertable flowup to the available treatment capacity or ASR injection capacity,whichever is less.

- Remaining divertable flow, if any, goes to the surface reservoir.

- If the surface reservoir is full, potential divertable flow is lostfrom the water supply system (i.e., not diverted).

• Condition B. Monthly divertable flow is less than monthly targetyield

- Treat divertable flow, if any, and distribute.

- Obtain remaining desired yield from the ASR system up to themaximum recovery rate or recoverable ASR volume, or both.

- Obtain remaining desired yield, if any, from the surfacereservoir, treat, and distribute.

- If total desired yield cannot be met, a yield deficit occurs.

The above logic is applied to each time step in the simulation and thenumber of yield deficits is tracked. The total number of deficitsdivided by the total number of simulation time steps is equal to thewater supply system deficit rate. One minus the deficit rate thenequals system reliability.


11

Methods

SELECTED CANDIDATE WITHDRAWAL SITESInitially, many sites were considered on the St. Johns River and theHaines Creek/Palatlakaha Chain of Lakes. The site selection processconsidered location of streamflow gauging stations, municipal watersupply demand growth areas, and potential maximum surface watersupply availability. Based on these criteria, the six candidate siteswere chosen for this analysis.

Only one site from the Haines Creek/Palatlakaha Chain of Lakeswatershed was selected. The chosen site is Lake Griffin in LakeCounty (Figure 2), which could potentially supply a portion of thefuture projected needs of northern Lake County. Lake Griffin receivesthe majority of its inflow from Haines Creek, and the water availabilityanalysis is based on the Haines Creek gauging station, located nearLisbon Florida (U.S. Geologic Survey [USGS] Gauge No. 2238000). Forthe purpose of this feasibility analysis, withdrawals from Lake Griffinwill only be allowed when Haines Creek flow rates are above theminimum streamflow requirements established for the Haines Creekgauge.

The following remaining five candidate withdrawal sites are alllocated on the main stem of the St. Johns River, as shown in Figure 3:

• St. Johns River near Cocoa• St. Johns River near Titusville• St. Johns River at Sanford(Lake Monroe)• St. Johns River near De Land• St. Johns River above Jacksonville

USGS stream gauging records, including daily flow observations andperiodic water quality data, are available at or near each of these sites.Streamflow and water quality characteristics for the Cocoa site aredefined by USGS Station No. 2232400, located near Cocoa. Thegauging station near Christmas Florida (2232500) is used for theTitusville site evaluation.

Data from the USGS De Land gauging station (2236000) are used toestablish streamflow characteristics for both the De Land site and theSanford site. For the Sanford site, which is located upstream from theDe Land gauge, the observed streamflow values are multiplied by thepreviously established adjustment factor of 0.797 to account for thereduced streamflow expected at the upstream location (CH2M HILL,1996b).


12

rM A R I O N C O U N T Y

Lady Lakea

_J

2.96 mgdEwtis

1 UtsburgD17.93 mgd

m

D Mt. Dora_ 1.64 mgd

Figure 2. Location of the CandidateSurface Water Withdrawal Site inLake County.

Stream gauging site

Gauging stationnumber/approximate maximumdevelopable surface water supply

7.93 mad Projected public water supplyincrease

Withdrawal site

ApproximateScale in Miles

10

13

ST- «<St. Augustine

> •»• • • ̂ W •* ̂ » ^M * * ̂ ^ • ' __

29.92 mqdTrtUSVffle

•Orlando

Melbourne

Figure 3. Locations of CandidateSurface Water Withdrawal Sites on theMain Stem of the St. Johns River. 2244450

768 mgd

Stream gauging site

Gauging stationnumber/approximate maximumdevelopable surface water supply

73 95 mad Projected public water supply** » Increase

Withdrawal site

ApproximateScale in Miles

10

14

Methods

The final St. Johns River candidate withdrawal site is located innorthern St. Johns County, just upstream of the City of Jacksonville.Streamflow and water quality records from the USGS station nearSwitzerland (2246500) are used to establish the characteristics of thissite.

Surface 'Water: Availability and Yield Analysis

15

Streamflow Duration Analysis

STREAMFLOW DURATION ANALYSISThe Streamflow duration analysis includes three major parts. First,individual flow duration curves were developed for each candidatewithdrawal site. These curves, which are based on analysis of long-term USGS daily flow records, define streamflow magnitude andvariability.

The previously developed minimum streamflow criteria were thenapplied to the flow duration curves to establish approximate minimumstreamflow requirements and the maximum streamflow diversion rateconsidered in this feasibility analysis. Once these parameters wereestablished, the effect of water supply withdrawal on each streamflowduration relationship was evaluated.

FLOW DURATION CURVES AND MINIMUMSTREAMFLOW REQUIREMENTSHaines Creek

Figure 4 presents the flow duration curve for Haines Creek nearLisbon Florida. During the 52-year period of record, Haines Creekflow has range from no flow to a daily maximum of 1,470 cubic feetper second (cfs), with a mean of 248 cfs. Periods of no streamflow arerare, occurring only about 0.7 percent of the time, or 2 to 3 days peryear on the average.

Based on previously established minimum flow criteria, the minimumstreamflow requirement for Haines Creek is 14 cfs. This flow rate isexceeded 94.3 percent of the time (0.95 * 0.993 = 0.943), which is alsothe maximum allowable diversion frequency. Again, using thepreviously established criteria, the maximum diversion rateconsidered in this analysis is 62 cfs (40 mgd), which is equal to25 percent of the long- term observed mean flow rate.

SJRWMD plans to conduct a minimum flows and levels analysis forLake Griffin. Once this analysis is complete, it is likely that a site-specific surface water diversion rule for Lake Griffin, based on lakestage, will be established. It is also likely that this rule will differ fromthe Haines Creek criteria used in this preliminary water supplyfeasibility analysis.


16

1,600

1,000

800

600

400

200

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Percent of Time Flow Rate is Exceeded

Mean Flow = 248 cfsMinimum Flow = 14 cfsMaximum Diversion Rate = 62 cfsMaximum Diversion Frequency = 94.3%

Figure 4. Flow Duration Curve for Haines Creek near Lisbon, Florida.

17


Because of the uncertainty associated with the outcome of the plannedminimum flows and levels analysis for Lake Griffin, it is possible thatthe estimated maximum water supply yield resulting from thispreliminary water supply evaluation could be overstated. For thisreason, an additional constraint could be placed on the Lake Griffincandidate withdrawal site when evaluated by the University of Floridadecision model. The additional constraint would limit the maximumwater supply considered to 50 percent of the theoretical maximumobtained by application of the Streamflow diversion criteriaestablished for this investigation.

The feasibility of developing a viable water supply from Lake Griffinshould be re-evaluated once the planned Lake Griffin minimum flowsand levels analysis is complete.

St. Johns River near Cocoa

Figure 5 presents the flow duration curve for the St. Johns River nearCocoa. Streamflow has ranged from 5.6 cfs to 10,700 cfs at this station.The long-term mean flow is 958 cfs. The maximum diversionfrequency, based on these Streamflow characteristics, is 95 percent.This flow frequency corresponds to a minimum Streamflowrequirement of 57 cfs, as determined from the flow frequency curve.Therefore, for the purpose of this analysis, water supply diversionswill not be allowed when Streamflow is less than 57 cfs and themaximum diversion rate to be considered is 239 cfs (154 mgd).

St. Johns River near Titusville

The overall Streamflow characteristics of the St. Johns River nearTitusville, as illustrated in Figure 6, are similar to the characteristicsnear Cocoa. However, during the 61-year period of record, there were5 days with no Streamflow. The mean flow is 1,273 cfs and themaximum observed daily flow rate is 11,600 cfs. Based on ourplanning criteria, the minimum Streamflow requirement is 70 cfs andflow can be withdrawn for water supply 95 percent of the time. Themaximum diversion rate investigated is 318 cfs (206 mgd).

St. Johns River at Sanford

The flow duration curve for the Sanford (Lake Monroe) candidatewithdrawal site is based on analysis of the flow records from theDe Land gauge adjusted for the smaller tributary area. As previouslydiscussed, the De Land gauging station daily flow data were


18

tr

.2

12,000

10,000

8,000

6,000

4,000

2,000

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%


Mean Flow = 958 cfsMinimum Flow = 57 cfsMaximum Diversion Rate = 239 cfsMaximum Diversion Frequency = 95.0%

Figure 5. Flow Duration Curve for the St. Johns River near Cocoa, Florida.

19

IILL

12,000

10,000

8,000

6,000

4,000

2,000

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%


100%

Mean Flow = 1,273 cfsMinimum Flow = 70 cfsMaximum Diversion Rate = 318 cfsMaximum Diversion Frequency = 95.0%

Figure 6. Flow Duration Curve for the St. Johns River near Titusville (Christmas), Florida.

20


multiplied by an adjustment factor equal to 0.797. The characteristicsof these adjusted data are illustrated in Figure 7.

As Figure 7 shows, the estimated streamflow characteristics for theSanford site are somewhat different than those observed at the twoprevious upstream sites. At the Sanford site, significant negative orupstream flows are exhibited. Daily flow rates range from -2,966 cfs to13,629 cfs, and the mean flow rate is 2,425 cfs. Although significantnegative flow rates can occur, they only occur about 2.3 percent of thetime.

Applying our planning criteria, the maximum diversion frequency forthe Sanford site will be about 92.8 percent of the time and thecorresponding minimum streamflow requirement is 571 cfs. Themaximum stream diversion rate to be considered is 606 cfs (392 mgd).

St. Johns River near De Land

Figure 8 presents the flow duration curve for the De Land candidatewithdrawal site. Since both the De Land and Sanford flow durationcurves were derived from the same observed streamflow sequence, theshapes of these curves are the same. However, the magnitude of flowat De Land is larger than that at Sanford.

Daily flow rates range from -3,030 cfs to 17,100 cfs and the mean flowrate is 3,043 cfs. Applying the streamflow diversion criteria for thisinvestigation, the maximum diversion frequency for the De Land sitewill be about 92.8 percent of the time and the corresponding minimumstreamflow requirement is 717 cfs. The maximum stream diversionrate to be considered is 761 cfs (492 mgd).

St. Johns River above Jacksonville

The final withdrawal site considered is the St. Johns River in northernSt. Johns County, just upstream from Jacksonville. As can be seen inFigure 9, the streamflow variability at this site is large. Daily flowrates range from -202,000 cfs to 185,000 cfs, and negative or upstreamflow occurs nearly 30 percent of the time. However, the net freshwateroutflow is substantial. The mean flow equals 9,129 cfs.

Applying the streamflow diversion criteria for this investigation, theminimum streamflow requirement is 1,663 cfs and water supplywithdrawal would be allowed 67.2 percent of the time. The maximumwithdrawal rate considered here is 2,282 cfs (1,475 mgd).


21

-4,00010% 20% 30% 40% 50% 60% 70% 80% 90% 100%


Mean Flow = 2,425 cfsMinimum Flow = 571 cfsMaximum Diversion Rate = 606 cfsMaximum Diversion Frequency = 92.8%

Figure 7. Approximate Flow Duration Curve for the St. Johns River near Sanford, Florida.

22

18,000

16,000

14,000

12,000

10,000

8,000DC5E 6,000

4,000

2,000

-2,000

-4,00010% 20% 30% 40% 50% 60% 70%


80% 90% 100%

Mean Flow = 3,043 cfsMinimum Flow = 717 cfsMaximum Djversion Rate = 761 cfsMaximum Diversion Frequency = 92.8%

Figure 8. Flow Duration Curve for the St. Johns River near De Land, Florida.

23

o -50,000IE

-250,00010% 20% 30% 40% 50% 60% 70%


80% 90% 100%

Mean Flow = 9,129cfsMinimum Flow = 1,663 cfsMaximum Diversion Rate = 2,282 cfsMaximum Diversion Frequency = 67.2%

Figure 9. Flow Duration Curve for the St. Johns River above Jacksonville, Florida.

24


The extremely large daily flow range and frequent upstream flow arecaused by strong tidal influences at this downstream location. Thesetidal influences would make development of a municipal water supplydifficult because of the large volume and significant duration ofseawater inflow. Because water supply withdrawals can only occurduring a certain time period, the facilities required for water supplydevelopment would be relatively large compared with the amount ofwater that would be gained. In addition, the raw water quality will beextremely variable because of tidal mixing, resulting in difficult andexpensive treatment requirements.

Also, the available Streamflow records at this site are of poor quality.Again, this is because of the tidal influences, which make flowmeasurement difficult. Large incoming and outgoing tidal flows occurdaily, and the differences in these flow volumes is the net dailyoutflow or river flow.

EFFECT OF WATER SUPPLY WITHDRAWAL ON FLOWDURATION RELATIONSHIPS

If a significant surface water supply were developed at any of thecandidate withdrawal sites, the Streamflow duration relationshipwould be modified. An analysis was conducted to quantify themaximum effects that could result from these withdrawals at each ofthe six sites.

In each case, the effect of withdrawal at the maximum rate considered,as well as withdrawals at 1/3 and 2/3 of the maximum, wereevaluated. Modified flow duration curves, one for each withdrawalrate, were developed and are shown in Figures 10 through 15. Thesefigures also demonstrate the effect of withdrawal on the entireStreamflow regime and on the normal to low-flow regime.

The effect of water supply withdrawal using the previouslyestablished withdrawal criteria are similar for each site. In the upperportions of Figures 10 through 15, the effects relative to the entireStreamflow regime are shown. In most cases, when the entireStreamflow regime is considered, the effects on the flow duration curveappear to be relatively minor.

However, the effects on the flow duration relationship are moreapparent in the bottom portions of Figures 10 through 15, whichillustrate only the low and moderate flow regime. Under theseconditions, water supply development tends to flatten the flowduration curve. At flow rates less than the minimum withdrawal rate,


25

1,600

Existing StreamflowDiversion Rate = 21 cfsDiversion Rate = 41 cfsDiversion Rate = 62 cfs

10% 20% 30% 40% 50% 60% 70%


80% 90% 100%

a) Entire Streamflow Regime

300

200

*cc

100


30% 40% 50% 60% 70% 80%


b) Moderate and Low Streamflow Regime

90% 100%

Figure 10. Maximum Effect of Water Supply Withdrawal on Flow Duration Curve for HainesCreek at Lisbon, Florida.

26

12,000

10,000Existing StreamflowDiversion Rate = 80 cfsDiversion Rate = 159 cfsDiversion Rate = 239 cfs

o% 10% 20%


30% 40% 50% 60% 70%


80% 90% 100%

1,500

1,000

DC

o

500


t t t—| 1 1 1 1 1 1 1 1 t 1 1 1 1-

30% 40% 50% 60% 70% 80% 90% 100%



Figure 11. Maximum Effect of Water Supply Withdrawal on Flow Duration Curve for the St.Jonns River near Cocoa, Florida.

27

12,000

10,000

Existing StreamflowDiversion Rate = 106 cfs_Diversion Rate = 212 cfsDiversion Rate = 318 cfs

30% 40% 50% 60% 70% 80%


90%


0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%



2,000

Existing StreamflowDiversion Rate = 106 cfsDiversion Rate = 212 cfs

— Diversion Rate = 318 cfs

100%

Figure 12. Maximum Effect of Water Supply Withdrawal on Flow Duration Curve for theSt. Johns River near Titusville (Christmas), Florida.

28


-2,000

-4,000o% 10% 20%


30% 40% 50% 60% 70%


80% 90% 100%

4,000

2,000

£u

Icc

o

-2,000


30% 40% 90%50% 60% 70% 80%

Percent of Time Flow Rate is Exceededb) Moderate and Low Streamflow Regime

Figure 13. Maximum Effect of Water Supply Withdrawal on Flow Duration Curve for the St.Jonns River at Sanford, Florida.

29

Existing StreamflowDiversion Rate = 254 cfsDiversion Rate = 507 cfsDiversion Rate = 761 cfs I

-4,000o% 10% 20%


30% 40% 50% 60% 70%


80% 90% 100%

4,000


2,000

Iir

-2,00030% 40% 50% 60% 70% 80%



90%

Figure 14. Maximum Effect of Water Supply Withdrawal on Flow Duration Curve for the St.Johns River at De Land, Florida.

30

Existing Streamflow(Effect of WithdrawalCannot be Shown forEntire Streamflow Regime)

-200,000

-250,0000% 10% 20% 30% 40% 50% 60% 70%


80% 90% 100%


20,000

15,000

•5- 10,000•5

o 5,000

-5,000

Existing StreamflowDiversion Rate = 761 cfsDiversion Rate = 1,521 cfsDiversion Rate = 2,282 cfs

30% 35% 40% 45% 50% 55% 60% 65% 70% 75% 80%



Figure 15. Maximum Effect of Water Supply Withdrawal on Flow Duration Curve for theSt. Johns River above Jacksonville, Florida.

31


the flow duration relationship would be unchanged as diversionwould not be allowed. However, the duration of time at or near thelow-flow value may increase noticeably.

This flattening effect appears to be more pronounced for the candidatewithdrawal sites with the smallest total streamflow magnitude; forexample, at the Haines Creek site (Figure 10) and St. Johns River site nearCocoa (Figure 11). The apparent effect on the Haines Creek flow regimeis only theoretical if the water supply withdrawal is located in LakeGriffin, as envisioned. The creek itself would not be affected becausethe water would actually be withdrawn from the downstream lake.


32

Potential Yield Analysis

POTENTIAL YIELD ANALYSISThe purpose of the potential yield analysis is to develop a quantitativerelationship between installed river diversion capacity and themaximum long-term water supply that could be developed. Thepotential yield analysis establishes the maximum theoretical watersupply yield for each diversion capacity investigated. This is anecessary intermediate step between the flow duration analysis anddetermination of water supply facility requirements because itestablishes the upper limit of the developable water supply andprovides insight into the magnitude of the facilities required todevelop each site.

Potential yield is a function of the minimum streamflow criteria, themaximum installed water supply diversion capacity, and thestreamflow variability. For each withdrawal site, the minimumstreamflow has been determined by applying the establishedstreamflow diversion criteria, and streamflow variability is defined bythe observed streamflow records. Therefore, the only remainingvariable is the installed diversion capacity. For each diversion capacityanalyzed, the potential yield is calculated for each observed dailystreamflow. These individual daily potential yield values are thenaveraged over the period of record to establish the overall potentialyield for that diversion capacity. The process is repeated for a numberof diversion capacities to establish a relationship between installeddiversion capacity and total divertable flow or potential yield.

For example, consider the St. Johns River near Cocoa, which has aminimum streamflow requirement of 57 cfs (about 37 mgd), with aninstalled river diversion capacity of 50 mgd. If the streamflow on agiven day is 30 mgd, which is less than the minimum streamflowrequirement, then the allowed diversion for that day is zero. If,however, the daily streamflow is equal to 60 mgd, then the allowablediversion is 23 mgd (60 mgd minus 37 mgd). If the daily streamflow islarge, say 300 mgd, then the withdrawal that day is limited by thediversion capacity and would be equal to 50 mgd. The divertableamount of water is calculated on a daily basis for the entire period ofstreamflow record to establish the long-term potential yield for eachmaximum diversion capacity considered.

The potential yield curve for each candidate withdrawal site is afunction of streamflow magnitude and variability and the minimumstreamflow criteria applied. Potential yield curves are both site-


33


specific and applicable only for a given set of minimum flow criteria.Thus, the potential yield analysis presented here applies only to thepreviously established minimum streamflow requirements. If thesecriteria change, then the potential yield curves, subsequent facilityrequirements, and water supply development costs will also change.

The potential yield curves for each of the six candidate withdrawalsites are illustrated in Figures 16 through 21. In each case, the installedriver diversion capacity, in mgd, is plotted on the x-axis and thepotential yield, in mgd, is shown on the y-axis. An empirical equationwas fit to each potential yield curve, and each of these equations isshown on the figures. The equations may be used to obtain anestimate of the potential yield for any diversion rate up to themaximum.

The maximum diversion rate considered and the resulting maximumpotential yield are also shown on Figures 16 through 21. For example,on Figure 16, which shows the potential yield relationship for HainesCreek, the maximum diversion rate investigated is 40 mgd and theresulting maximum potential yield is 30 mgd, or about 75 percent ofthe diversion capacity. Because the maximum diversion rate is25 percent of the mean streamflow, the maximum potential yield inthis case is equal to about 18.8 percent (0.75 x 0.25 = 0.188) of the totalstreamflow. Table 1 summarizes the maximum potential yield of eachcandidate withdrawal site as a function of the mean annualstreamflow.

Table 1 also provides some insight into the relative water supplydevelopment potential of each candidate withdrawal site. The yieldratio, which is defined as the potential yield divided by the installeddiversion capacity (PY/Qd), illustrates the relative scale of facilitiesrequired at each site. In general, a site with a larger yield ratio willrequire smaller facilities relative to the water supply developed. Basedon a comparison of the yield ratio, the most attractive sites would bethe St. Johns river at Sanford or the St. Johns River near De Land. Theyield ratio at these locations is equal to 84 percent, which represents anefficient use of constructed water supply facilities.

The least desirable location would be the St. Johns River nearJacksonville, with a yield ratio of just 64 percent. This low-yield ratiois a direct result of the significant tidal influences that restrict the timeperiod during which flow could be withdrawn from the river.Therefore, facilities at this site would have to be proportionately largerto meet a given target yield.


34

Pot

entia

l Yie

ld (P

Y) [

mgd

]_i

_i

HO

|\j

W

COO

O1

OO

1O

O1

OO

1

x^//

«r

//

/

PY

\̂\

= Qd/(1.0S

^̂

3 + 0.0071:

^

2*Qd)

D 5 10 15 20 25 30 35 40 45

Diversion Capacity (Qd) [mgd]

Maximum Diversion Rate = 40 mgd

Maximum Potential Yield = 30 mgd

Figure 16. Potential Yield Curve for Haines Creek at Lisbon, Florida.

35

140

120

^ 100o>

a 80

1

3 60t&

S. 40

20

0 </_/

^

/^

/^

_^

<,X

PY = Qd/(1

^

s.

0555 + 0.00

^

124*Qd)

— '

) 20 40 60 80 100 120 140 160




Figure 17. Potential Yield Curve for St. Johns River near Cocoa, Florida.

36

180

160

140

Jb 100«^ 80

ia eoQ.

40

on

0

s/

/

S

/,^r

//

S

^^^

s\\

PY = Qd/(1.06I

.4

/

/

>37 + 0.000868*Qd)

'

0 50 100 150 200 250




Figure 18. Potential Yield Curve for St. Johns River near Titusville (Christmas), Florida.

37

350

300

75*250E

a 200

2

*1

io- 100

50//

./

/r~

^/^

^^\^ \

\\

PY = Qd

_^

^^

^

/(1. 0635 + 0.

\^

00031 2*Qd

/

U0 50 100 150 200 250 300 350 400




Figure 1 9. Potential Yield Curve for St. Johns River at Sanford, Florida.

38

450

350

£300

P£.250

± 200(B•*s

{1500.

TOO

50

S/

//

^

S*S^

^

^_^~\^

<^

S*^^ \

^^S^S^

VX

.s*^^

^S^

PY = Qd/(1.0635 + 0.000249*Qd)

°0 50 100 150 200 250 300 350 400 450 500




Figure 20. Potential Yield Curve for St. Johns River near De Land, Florida.

39

1000

900

800

§700

P600

"" 500

400

300

200

100

X\

PY = Qd/(1.479 + 0.0000458*Qd)

200 400 600 800 1000 1200


1400 1600

Maximum Diversion Rate = 1,475 mgd


Figure 21. Potential Yield Curve for St. Johns River above Jacksonville, Florida.

40


Table 1. Maximum Potential Yield for the Six Candidate Surface WaterWithdrawal Sites

Withdrawal SiteHaines Creek(Lake Griffin)St. Johns River nearCocoaSt. Johns River nearTitusvilleSt. Johns River atSanford (LakeMonroe)St. Johns River nearDe LandSt. Johns Riverabove Jacksonville

MaximumDiversion

RateCQdHmgd)

40

154

206

392

492

1,475

MaximumPotential

Yield[PY] (mgd)

30

124

166

331

415

953

Yield RatioPY/Qdas

percentage75

81

81

84

84

64

MaximumPotentialYield as a

Percentage ofMean

Streamflow18.8

20.1

20.1

21.1

21.1

16.2


41

Surface Water Supply Facility Requirements

SURFACE WATER SUPPLY FACILITYREQUIREMENTS

Facility requirements for each candidate withdrawal site weredetermined by simulating the water supply systems and evaluatingthe treatment requirements, based on analysis of available waterquality data. The first step in this process is to establish the initialtarget yield array and the trial facility components array to beevaluated. Both of these arrays are based on the potential yield curveof the withdrawal site.

For example, consider the St. Johns River near Cocoa, with anestimated maximum potential yield of 124 mgd (Figure 17). Aspreviously discussed, the maximum reliable yield will probably besomewhat less than this value. Therefore, the trial average daily yieldinvestigated was limited to 120 mgd or less. The trial average dailyyield array was established on equal increments of from 12 to 120 mgd.Initial target yields for each of the six candidate withdrawal site wereselected in a similar manner.

TRIAL FACILITY COMPONENTSThe trial facility components were also selected on the basis of thepotential yield curve and selected interrelationships among thecomponents. Major surface water supply system facilities, which areidentified in Figure 1, include a raw water diversion structure, a rawwater off-line storage reservoir, a water treatment plant, and a treatedwater ASR system.

Raw Water Diversion and Pumping Station

Again, considering the St. Johns River near Cocoa, the maximumdiversion capacity considered is equal to 154 mgd, as illustrated inFigure 17. The maximum diversion rate was divided into six elements(ranging from 25.7 mgd to 154 mgd) to define the diversion capacity ofthe six trial water supply systems. The sizes of the remainingcomponents are then based on the diversion capacity to definereasonably balanced trial water supply systems.

Off-line Raw Water Storage Reservoir

The purpose of the off-line raw water storage reservoir is primarily toprovide operational flexibility. Because streamflow cannot be diverted


42


for at least 5 percent of the time, the raw water storage allows the plantto continue operation while withdrawals are not allowed. Becauseshutting down and starting up a water treatment plant is a significanttask, the frequency of shut-down and start-up should be minimized.For the purpose of this preliminary facilities analysis, the off-linestorage reservoir is sized to provide 5 days of storage, based on theinstalled diversion capacity. During periods when streamflowwithdrawals are not allowed, plant operations would probably bereduced to approximately 20 percent of plant capacity, but would notbe shut down completely. Thus, the off-line reservoir would allow theplant to operate at this reduced rate for at least 25 days before it wouldbe necessary to completely cease operations.

An off-line raw water storage reservoir would also provide otherfunctions and operational flexibility. For example, if a contaminant isspilled in the river, the raw water intake could be closed while thecontaminated flow moves downstream, without interrupting plantoperations. In addition, the raw water storage reservoir wouldprovided some sedimentation and mixing of the raw water prior toentering the treatment facilities.

The off-line storage reservoir may be difficult to site and permit, giventhat these reservoirs will probably be located near the riverwithdrawal site and, therefore, within or adjacent to floodplains orwetlands, or both. The need for the off-line reservoir is driven in partby the assumed minimum streamflow criteria, which would not allowwater supply diversion for at least 5 percent of the time during lowstreamflow periods. If the streamflow diversion criteria were relaxedto allow reduced diversion during low-flow periods (say 20 percent oftreatment capacity), then the need for the off-line reservoir would bereduced and, possibly, could be eliminated. To a large extent, theenvironmental trade-off is whether it is better to allow some reduceddiversion at all times and avoid floodplain or wetland encroachment,or fully protect the riverine system during low-flow periods. Thistrade-off will likely be answered on a case-by-case basis if surfacewater supply proves to be a technically and economically feasiblewater supply option.

An off-line storage reservoir may not be necessary at the Lake Griffinsite, which is itself a large reservoir. If some water supply withdrawalwere allowed on a continuous basis, it may be possible to develop aviable water supply without constructing a separate off-line raw waterstorage reservoir. The same general reasoning also applies along the


43


main stem of the St. Johns River. Because of its mild slopes, theSt. Johns River system is very much like a system of interconnectedlakes. Again, if some continuous diversion were allowed, the off-linereservoir component may be unnecessary.

The only currently operational municipal surface water supply on theSt. Johns River is located on Lake Washington and serves the City ofMelbourne, Florida. This system does not include an off-line rawwater reservoir. However, the City of Melbourne also uses groundwater as a source of supply. Therefore, the overall reliability of thewater supply system does not depend on surface water alone. Astand-alone surface water supply system located on the Peace River insouthwest Florida does incorporate an off-line raw water reservoir.Therefore, examples of each application exist. The final decision for agiven site should be based on detailed engineering alternatives andenvironmental systems analysis.

For the purpose of this preliminary feasibility analysis, the off-line rawwater reservoir is included because a reservoir will be necessary tomeet the minimum streamflow criteria established for thisinvestigation. However, the cost of the off-line reservoirs will betracked separately so that alternatives that do not include thiscomponent can be evaluated by SJRWMD if desired.

Water Treatment Plant

The conventional water treatment plant capacity is set equal to95 percent of the raw water diversion capacity. The small differencebetween the diversion capacity and treatment capacity allows somecapability to refill the off-line reservoir, while at the same timeoperating the water treatment plant at full capacity.

If desalting is required, it is assumed that RO treatment will beprovided following conventional surface water treatment. Therequired RO treatment capacity will vary with each candidatewithdrawal site, depending on water quality characteristics (includingvariability) of the raw water. In each case, the RO treatment capacity isexpressed as a percentage of the conventional treatment capacity (from0 to 100 percent). Where desalting is required, a split-stream treatmentsystem will be provided because the degree of desalting requiredcould vary significantly with river flow or other factors. When surfacewater supply sources have variable raw water quality, this type ofoperational flexibility is essential to successful development.


44


Treated Water ASR System

The ASR system consists of a wellfield into which excess treated wateris injected for ultimate withdrawal and distribution as needed.Volumetrically, the ASR storage capacity is considered unlimited.However, the hydraulic capacity of the ASR wellfield determines therate at which water can be injected and recovered. In general, therequired recovery rate will control the size of the ASR facilities.

For the purpose of the facility requirements simulation, ASR injectionor recovery capacity should not limit the reliability of the water supplysystem. That is, sufficient ASR facilities will be constructed tomaximize the reliability of the water treatment plant, which is thefacilities component with the greatest cost. Therefore, ASR injectioncapacity was set equal to treatment plant capacity, and ASR recoverycapacity was set equal to 1.5 times the potential yield for thecorresponding diversion capacity. By doing so, ASR capacity will notlimit the reliability of the simulated water supply systems.

Once treatment plant capacity and a corresponding net reliable yieldare determined after the simulation and treatment requirementsanalysis, the required ASR capacity can be computed directly, basedon these values. The injection capacity will be equal to the maximumamount available for injection, which is equal to the treatment plantcapacity minus the minimum monthly demand. On the other hand,the maximum recovery rate is based on delivering the maximumdemand with the treatment plant shut down. Therefore, based on amaximum day demand equal to 1.5 times the ADD, the ASRwithdrawal rate will also be equal to 1.5 times the ADD. In this way,the ability to meet peak demands is fully satisfied by the ASR system.The ability to meet long-term demands and fill the ASR storage areamust be met by upstream diversion and treatment components.

FACILITY REQUIREMENTS SIMULATION RESULTSThe simulation results indicate that reliable gross water supply yieldsapproaching the potential yield can be developed for most candidatewithdrawal sites. In general, the gross reliable yield is equal to about95 percent of the potential yield. This is true for all sites except theSt. Johns River near Jacksonville, where adverse hydrologiccharacteristics result in a gross reliable yield of only 72 to 78 percent ofthe potential yield.


45


If there were not appreciable wastewater generated by the watertreatment process, then the gross reliable yield would equal the netreliable yield. Net reliable yield, as used in this application, is thewater actually delivered to the distribution system, while gross reliableyield is the water diverted from the stream and processed by the watersupply facilities. Gross reliable yield is identified by application of thewater supply systems simulation.

For conventional surface water treatment systems, volumetric lossesare generally in the range of 2 to 8 percent of the total raw waterinflow. The volume of wastewater generated will vary with thecharacteristics of the sludge produced and the dewatering systemprovided to treat the sludge. For the purpose of this preliminaryfeasibility analysis, a loss of 5 percent is assumed. This means that thevolume of treated water produced by the conventional treatmentcomponents is assumed to be equal to 95 percent of the total raw watervolume diverted for treatment.

For that portion of the diverted flow requiring desalting using ROtechnology, volumetric losses can be more significant. The wasteresidual, which is the concentrate produced in the desalting process,ranges from about 15 percent of the flow treated by the RO componentto more than 50 percent, depending on the quality of the raw water.

Therefore, the gross reliable yield identified by application of the long-term water supply systems simulation must be adjusted, based ontreatment requirements and raw water quality variabilityconsiderations to establish net reliable yield. Thus, net reliable yield,water quality characteristics, and treatability are interrelated.

WATER QUALITY AND TREATMENT REQUIREMENTSTable 2 presents a summary of the water quality characteristics forpreviously selected parameters for five USGS gauging stations ofinterest to this investigation. The summary provided includes theaverage, standard deviation, maximum value, minimum value, andnumber of observations for each parameter for each gauge.

Each candidate withdrawal site corresponds directly to a gaugingstation, with the exception of the St. Johns River at Sanford. TheSt. Johns River near Titusville is the nearest upstream station toSanford and is considered to be most representative. Therefore, waterquality data for Titusville is also used to characterize expected waterquality at Sanford.


46


Table 2. Water Quality Characteristics at Candidate Withdrawal Points

WaterTemperature

(deg C)

Haines Creek at Lisbon FL (2238000)Average 23.7Standard Deviation 5.3MaximumMinimumNumber of Observations

32.311.086

St. Johns River near Cocoa (2232400)Average 23.4Standard Deviation 5.9MaximumMinimumNumber of Observations

808.5293

St. Johns River near Christmas* (2232500)Average 23.4Standard Deviation 7.7MaximumMinimumNumber of Observations

796.5109

St. Johns River near De Land (2236000)Average 24.1Standard Deviation 6.7MaximumMinimumNumber of Observations

8110.5159

St. Johns River at Jacksonville (2246500)Average 22.17787Standard Deviation 5.244494MaximumMinimumNumber of Observations

3111.5122

Turbidity(JTU)

18154557

53

201

39

7412111

63101

11

94201

45

Color(PT-CO) p

36 740 0150 85 6

Alkalinity(mg/L

H CaC03)

3 101.55 10.77 1347 77

26 25 25

107 746 0420 93 5

1 36.74 17.02 1151 13

341 379 294

119 7.0 45.280 0.4 22.6

700 835 5

2 1208 17

96 125 95

98 767 0500 812 6

3 61.85 22.07 1332 2

134 188 159

76 7.3 68.655 0.4 17.0

400 8.2 9812 6 6 3360 73 47

Ammonia(mg/LNH3)

0.00700.00820.016

05

0.00300.0063

0.03037

0.00520.01740.09

026

0.00280.00460.024

051

0.00190.00210.008

013

Nitrate(mg/L N03-N)

0.00750.01500.03

04

0.06420.2110

1.40

45

0.10980.2867

1.80

45

0.18900.2604

1.10

39

0.22080.0968

0.40.0139

Total OrganicCarbon(mg/L)

18.58.134012

22.28.3490.139

20.25.026126

20.79.4466.523

12.26.421011

Hardness(mg/L

CaCO3)

16.16.625221

76.981.15708

325

154.2134.86722190

124.915448.13225

31246130

1316.1731088.345

33005252

Chloride(mg/L)

245

401725

12713596018

334

28024511503796

241865700

173

3551316411000

71104

XI

JTU = Jackson Turbidity Units.PT-CO = Platinum Cobalt Units."Assumed to be applicable at Titusville and Sanford withdrawal sites.



For this preliminary feasibility investigation, the available water qualityparameter of most concern is the chloride concentration reported in thelast column of Table 2. The magnitude and variability of the chlorideconcentration will define in large part the degree of RO treatmentrequired. This, in turn, will impact the net reliable yield and,eventually, the cost of developing the candidate water supply source.

The only truly fresh water source is Haines Creek. The maximumobserved chloride concentration is 40 milligrams per liter (mg/L), wellwithin the drinking water standard (DWS) of 250 mg/L. Developmentof this site would require only conventional surface water treatmentconsisting primarily of coagulation/sedimentation (clarification),filtration, and disinfection using ozone and chloramines.

Periodically, all the St. Johns River candidate withdrawal sites exceedthe DWS for chlorides. Thus, each site would require desalting of atleast part of the flow for part of the time. In each case, a split-streamtreatment system is envisioned. In a split-stream treatment system,only a portion of the total flow would receive RO treatment on an as-needed basis. All flow would receive conventional treatment, similarto that required for the Haines Creek source. When the inflow exceeds90 percent of the DWS for chlorides (0.9*250=225), then a portion of theinflow would also receive RO treatment so that the product water doesnot exceed 225 mg/L. Product water is the treated water that actuallyenters the distribution system and is delivered to the consumer.

As can be seen in Table 2, chloride concentration generally increases ina downstream direction. Mean chloride concentration ranges from127 mg/L near Cocoa to more than 3,500 mg/L above Jacksonville.However, the water quality at De Land is slightly better and much lessvariable than the water quality near Titusville, probably because of theinfluence of the Wekiva River, which contributes significant highquality flow between theses two stations. Because both Titusville andSanford are upstream of the Wekiva River, the Titusville data wereselected as most representative of conditions at Sanford.

Using the water quality summaries in Table 2, conceptual treatmentrequirements for each candidate withdrawal site were established andare reported in Table 3. All the sites will require conventional surfacewater treatment, including chemical-aided sedimentation, filtration,and disinfection. In addition, all of the St. Johns River candidatewithdrawal sites will require varying degrees of RO treatment. For thepurpose of this preliminary feasibility analysis, RO treatment is


48


Table 3. Major Water Treatment Processes Required for Each Candidate Surface WaterWithdrawal Site

CandidateSurfaceWater

WithdrawalSite

Lake Griffin(HainesCreek)

St. JohnsRiver nearCocoa

St. JohnsRiver nearTitusville

St. JohnsRiver atSanford(LakeMonroe)

St. JohnsRiver nearDe Land

St. JohnsRiver nearJacksonville

Treatment Process8

Coagulation/Flocculation-

Sedimentation

/

/

/

/

/

/

Filtration

/

/

/

/

/

/

Reverse OsmosisSlightlybrackish

Cl<500 mg/L

/(50%)b

ModeratelybrackishCISOOto

1,000 mg/L

/(63%)

/(63%)

/(67%)

Highlybrackish

CM ,000 to5,000 mg/L

Saline Cl> 5,000mgA.

/(100%)

Concentratedisposal.

/(7.5%)c

/(12.6%)

/(12.6%)

/(13.4%)

/(55%)

aAII water treatment systems will require sludge thickening and dewatering facilities, finish water transfer pumping,ground storage tankage, disinfection (ozone) and miscellaneous facilities including offices, and site work.bDenotes required RO treatment capacity as a percentage of total conventional water treatment capacitycDenotes required concentrate disposal capacity as a percentage of total conventional water treatment capacity


49


grouped into the following four major classifications, which are basedon maximum expected chloride concentration:

• Slightly brackish. This classification applies when the maximumchloride concentration is less than 500 mg/L. For this application,the RO removal efficiency (salt rejection) is anticipated to be90 percent and the concentrate volume is anticipated to be about15 percent of the treated inflow.

• Moderately brackish. This classification applies when themaximum chloride concentration is greater than 500 mg/L, but lessthan 1,000 mg/L. RO removal efficiency is expected to be about90 percent and the concentrate volume will be about 20 percent ofthe treated inflow.

• Highly brackish. This classification applies when the maximumchloride concentration is greater than 1,000 mg/L, but less than5,000 mg/L. Expected RO removal efficiency is about 90 percentand concentrate volume is about 30 percent of the treated inflow.

• Saline. This classification applies when the maximum chlorideconcentration is greater than 5,000 mg/L. The expected ROremoval efficiency is greater than 98.5 percent and the concentratevolume is about 55 percent of the treated inflow.

These classifications are conceptual only; actual RO removalefficiencies and concentrate volumes may vary considerably withadditional planning and preliminary design analysis. However, thesevalues are considered appropriate for the purpose of initial feasibilityevaluation and cost estimating.

As can be seen from Table 3, the St. Johns River near Cocoa is slightlybrackish. Most of the time, raw water will be well within the DWS forchloride. However, the maximum chloride concentration of thediverted flow is expected to be about 400 mg/L. Flow with the highestchloride concentration (up to 570 mg/L) is associated with low riverflow and will not be diverted, based on the previously establisheddiversion rule. Therefore, under worst-case operating conditions, onlyabout 50 percent of the diverted flow would require desalting.

Under average or normal operating conditions, most flow would notrequire desalting. Over the long-term, only about 2 to 3 percent of thediverted flow would become waste concentrate, and the maximumconcentrated flow rate would equal approximately 7.5 percent of the


50


conventional treatment capacity. The net reliable yield is estimated tobe about 98 percent of the gross reliable yield.

The next three sites (Titusville, Sanford, and De Land) are similar.Each is considered moderately brackish, with maximum raw waterchlorides greater than 500 mg/L, but less than 1,000 mg/L. The split-stream treatment described previously for the Cocoa site would alsoapply to these sites. The estimated required RO capacity ranges from63 to 67 percent of the conventional treatment capacity and themaximum concentrate flow would be on the order of 13 percent of thetotal conventional treatment capacity. Net reliable water supply yieldfor these three sites will be about 94 to 95 percent of the gross yield,with only 5 to 6 percent of the diverted flow becoming wasteconcentrate.

The final candidate withdrawal site, the St. Johns River nearJacksonville, has poor water quality characteristics. Net outflow at thispoint can be relatively good quality fresh water, or it can be sea wateror a mixture of the two. Chlorides have ranged from 71 mg/L to11,000 mg/L. Chloride concentrations can also be high at any flowrate. Sea water will flow upstream during incoming tides and be partof the outflow on the outgoing tide. The Jacksonville site is consideredsaline, because the diverted raw water will often exceed 5,000 mg/1chlorides, and would require construction of sea water RO facilities.Over the long term the reliable net yield would be about 65 percent ofthe reliable gross yield. That is, about 35 percent of the total divertedflow would become waste concentrate.

Given the very poor hydrologic and water quality characteristics ofthis site and the expensive and relatively under-used facilities requiredto develop the potential water supply, this site is not considered afeasible surface water supply alternative and should not be evaluatedin future phases of the SJRWMD alternative water supplyinvestigations.

These adverse hydraulic and water quality conditions apply only tothe downstream portion of the main stem of the St. Johns River. It ispossible that tributaries to the lower St. Johns River could bedeveloped as viable water supply sources if the withdrawal sites arelocated above tidal influence or behind constructed salinity barriers.Candidate watersheds may include Black Creek in Clay County orDeep Creek in St. Johns County. Evaluation of these hydrologicsystems would be hampered, however, by a lack of basic long-termstreamflow and water quality data.


51


RELIABLE NET YIELD AND FACILITY REQUIREMENTSThe results of the long-term systems simulation and the water qualityand treatment requirements analysis were used to establish estimatesof the reliable net yield for each of the trial water supply systems. Thereliable yield, along with the facility capacities required to develop thatyield, are reported in Table 4.

There are six reliable yield estimates and six water supply systemsdefined for each of the candidate water withdrawal sites. For example,at the Lake Griffin site, reliable yields of up to 28 mgd could bedeveloped. Facilities required to develop the maximum 28-mgd yieldinclude a 40-mgd water diversion structure, a 38-mgd conventionalwater treatment plant, and a 42-mgd ASR system. An off-line rawwater reservoir with a 200-million-gallon (MG) capacity would also beneeded. This water supply system will fully meet the desired monthlydemand at least 98.3 percent of the time.

As previously discussed, the planned minimum flow and levels analysisfor Lake Griffin may further restrict the reliable net yield available fromthis candidate withdrawal point. Therefore, to provide a relativelyconservative estimate of Lake Griffin water supply potential, themaximum net reliable yield to be evaluated in the University of Floridadecision model will be limited to 50 percent of the maximum net reliableyield estimated in this analysis, or 14 mgd (0.5 x 28).

If Lake Griffin appears to be a desirable water supply source, as determinedby application of the University of Florida decision model, then it may benecessary to re-evaluate Lake Griffin facility requirements and costs. Thisre-evaluation would be based on a new diversion rule established after theplanned minimum flows and levels analysis is complete.

The St. Johns River sites would require additional facilities. Themaximum reliable net yield for the St. Johns River at Sanford is about279 mgd. To fully develop this yield, substantial facilities would berequired including a 392-mgd river diversion structure, a 2-billion-gallon off-line storage reservoir, a 372-mgd conventional watertreatment plant, a 223-mgd RO treatment facility, a 419-mgd ASRsystem, and a 45-mgd concentrate disposal system.

The data reported in Table 4 can be used to estimate the approximatefacility requirements for any desired yield at any of the candidatewithdrawal sites by linear interpolation. Figures 22 through 27illustrate the relationship between net reliable yield and requiredfacilities for each of the six candidate withdrawal sites.


52


Table 4. Surface Water Facility Requirements Summary for Range of Total Reliable NetYield

WaterReliable Net River Diversion Treatment PlantYield (mgd) Capacity (mgd) Capacity (mgd)

RO TreatmentCapacity (mgd)

ASR RecoveryCapacity (mgd)

RO ConcentrateDisposal

Capacity (mgd)

a) Lake Griffin (Haines Creek)

b)St

c)St

d)St

e)St

f)St.

5.410.515.319.223.728.0

Johns19.939.558.075.091.5108.4

Johns27.050.974.497.4120.8142.5

Johns48.697.4149.9192.0239.2279.1

Johns62.6124.4189.3242.9304.7350.8

6.713.320.026.733.340.0

River near Cocoa25.751.377.0102.7128.3154.0

River near Titusville34.268.3102.5136.7170.8205.0

River at Sanford65.3130.7196.0261.3326.7392.0

River near De Land82.0164.0246.0328.0410.0492.0

6.412.619.025.431.638.0

24.448.773.297.6121.9146.3

32.564.997.4129.9162.3194.8

62.0124.2186.2248.2310.4372.4

77.9155.8233.7311.6389.5467.4

0.00.00.00.00.00.0

11.623.134.746.357.969.5

19.438.858.377.797.1116.6

37.174.3111.4148.6185.8222.9

49.699.2148.8198.3247.9297.5

8.115.822.928.835.542.0

29.959.286.9112.5137.3162.6

40.476.3111.5146.1181.1213.8

73.0146.1224.8288.0358.8418.6

94.0186.5283.9364.4457.0526.2

0.00.00.00.00.00.0

1.73.55.27.08.710.4

3.97.811.715.519.423.3

7.414.922.329.737.244.6

9.919.829.839.749.659.5

Johns River above Jacksonville75.1152.1229.1302.7381.0418.9

245.8491.7737.5983.3

1,229.21 ,475.0

233.5467.1700.6934.1

1,167.71,401.3

221.8443.8665.6887.4

1,109.41,331.2

112.6228.1343.8454.0571.5628.3

122.0244.1366.1488.1610.1732.2


53

I

45

40

35

30

Off-Line Reservoir Volume = 5 Days SupplyBased on River Diversion Capacity

No RO Treatment orConcentrate Disposal Required

I

O 25J£ 20uCO

15

10

10 15 20

Net Reliable Yield ADD (mgd)

25 30

River DiversionTreatment PlantASR System

Figure 22. Surface Water Facility Requirements for Lake Griffin (Haines Creek).

54

180

1601E 140

a 120

100

£ 80


20 40 60 80


100 120

River DiversionTreatment PlantASR SystemRO CapacityConcentrate

Figure 23. Surface Water Facility Requirements for St. Johns River near Cocoa.

55


250

" 200

8-150Uw9

0,00

60 80 100


120 140 160

River Diversion— Treatment Plant- - ASR System

RO Capacity— Concentrate

Figure 24. Surface Water Facility Requirements for St. Johns River near Titusville.

56

450

400

IE 350

300

O 250

S£ 200


21

150

100

100 150 200


250 300

River Diversion• Treatment Plant

ASR SystemRO CapacityConcentrate

Figure 25. Surface Water Facility Requirements for St. Johns River at Sanford.

57


600

•o 500

o 400

IO8 300

200

SFiooDC

50 100 150 200 250


300 350 400


Figure 26. Surface Water Facility Requirements for St. Johns River near De Land.

58

1,600

1,400

•5- 1,200

8.o

1,000

§ 800£

% 600

S.fc 400

200cc


50 100 150 200 250 300


350 400 450

NOTE: Conventional treatment plant capacityand RO treatment capacity are equal forthis candidate withdrawal site.


Figure 27. Surface Water Facility Requirements for St. Johns River above Jacksonville.

59

Proposed Cost Estimation Procedure

PROPOSED COST ESTIMATION PROCEDUREThe surface water supply cost estimates will include development of acomplete array of planning-level cost estimates for six different sizefacilities at five different candidate withdrawal sites. Cost functionsfor each withdrawal site (net reliable yield [ADD] vs. cost) will also bedeveloped and a TM will be produced to report the methods used andresults obtained.

SITES CONSIDEREDAll candidate surface water withdrawal sites, with the exception of theSt. Johns River above Jacksonville, will be included in the costestimating phase of the investigation. As discussed previously, theJacksonville site does not provide a practical alternative for municipalwater supply development.

COST PARAMETERS AND CRITERIACost parameters to be considered have been previously established bythe project team and include the following:

• Construction cost• Non-construction capital cost• Land cost• Land acquisition cost• Total capital cost• Operation and maintenance (O&M)• Equivalent annual cost• Annualized set-up cost• Annualized unit cost

Economic criteria, including cost basis, non-construction capital costfactor unit land costs, interest rate, and facilities life expectancies, havebeen previously established for all cost estimates developed as part ofthe SJRWMD alternative water supply strategies investigations. Thesepreviously established criteria will be used to develop the requiredcost estimates for the surface water facilities.


60


CONSTRUCTION AND O&M COST COMPONENTSConstruction and O&M cost estimates will be developed at thepreliminary planning or cost curve level for the major componentsrequired. Major components required for each candidate withdrawalsite may include the following:

• River diversion structure- intake structure- bar screen- pumping station- raw water transmission to off-line reservoir

• Off-line raw water reservoir- perimeter levee- emergency spillway- site preparation

• Conventional surface water treatment plant- raw water pumping- coagulation / flocculation-sedimentation- filtration- sludge handling facilities- disinfection (primary ozone, residual chloramines)- instrumentation and controls (I&C)- transfer pumping- ground storage tankage- off ice/laboratory- general site work

• RO treatment (in addition to conventional facilities)- RO membranes- chemical addition- concentrate disposal system (deep wells)

• ASR system- wells- pumps- piping and I&C

Applicable cost curves will be identified in the literature. Wherenecessary, cost curves or unit costs for individual items (e.g., filtration)or major systems (e.g., conventional surface water treatment) will bedeveloped. Using the identified or developed construction and O&M


61


cost curves, a spreadsheet application will be developed. Thisspreadsheet will be used to develop the required cost estimates for sixdifferent water supply yields at the five candidate withdrawal sites.The facility capacities to be used in the cost estimating procedure arepresented in Table 4, parts a through e.

LAND COSTSLand requirements will be estimated for the off-line raw water storagereservoir, the water treatment plant, and ASR wellfield. Total landcosts, including the cost of acquisition, will then be estimated, based onthe estimated total land requirement for each of the 30 facilitiesconsidered.

OTHER COST ESTIMATESOther cost parameters, including total capital cost, equivalent annualcost, set-up cost, and unit cost, will be estimated, based on theconstruction, land, and O&M costs computed for each facility. Thesecost estimates will be included in the cost estimating spreadsheet andwill be developed in accordance with the cost estimating guidelinesand economic criteria established for the SJRWMD alternative watersupply investigations.

COST FUNCTIONSThe cost analysis will result in the development of estimated costs foreach of the eight cost parameters for six different water supply systemsat each of the five candidate water supply withdrawal sites. Theseindividual cost estimates will be used to develop cost equations foreach cost parameter applicable to the individual withdrawal sites. Theresulting cost equation will likely have the following general form:

COST(p,1) = a(p,1) + b(p/1)*RNYc(p,l)

Where:

COST(p i) = estimated cost of parameter p, at withdrawal site location1, in total dollars or dollars per year, depending on the units of the costparameter.

RNY = reliable net water supply yield in mgd. This is also the ADDdelivered to distribution.


62


a(p,l) b(p,l) and c(p,l) are best fit parameters for each equation asdetermined by regression analysis.

To provide added flexibility, the costs associated with the off-line rawwater reservoir will be considered separately from the other watersupply system components, including the river diversion structure, thewater treatment plant, and the ASR system. Surface water supplysystem costs with and without the reservoir component may then beevaluated if desired.

These equations may be used to estimate any cost parameter for anydesired water supply yield, within the range investigated, for eachcandidate withdrawal site. These equations will be used in theUniversity of Florida Decision Model to define the cost characteristicsof the surface water supply alternative.

TECHNICAL MEMORANDUMA TM reporting the methods and results of the surface water supplycost estimating and cost equation development will be prepared.


63

Summary and Recommendations

SUMMARY AND RECOMMENDATIONS

SUMMARYThis preliminary surface water availability and yield analysisaddresses the municipal water supply development potential of sixpreviously selected candidate surface water supply withdrawal sites(CH2M HILL, 1996b). The locations of the six sites are as follows:

• Lake Griffin (Haines Creek)• St. Johns River near Cocoa• St. Johns River near Titusville• St. Johns River at Sanford(Lake Monroe)• St. Johns River near De Land• St. Johns River above Jacksonville

In evaluating these sites, similar analyses were conducted thatincluded the following elements:

• Development of flow duration curves

• Determination of minimum streamflow requirements andmaximum water supply withdrawal rate

• Evaluation of the effect of water supply withdrawal on the flowduration relationship

• Evaluation of the potential water supply yield as a function ofmaximum installed withdrawal capacity

• Determination of streamflow water quality characteristics andwater treatment requirements

• Estimation of the type and size of water supply facilities requiredas a function of the total reliable water supply yield developed

Flow Duration Analysis

There were two primary objectives for the streamflow durationanalysis. First, the flow duration curves for each candidate withdrawalsite were established to determine the frequency of allowable watersupply diversion, the minimum streamflow requirements, and themaximum diversion rate to be investigated.


64


For this project, minimum streamflow requirements are defined as thepositive streamflow rate observed 95 percent of the time. Using thisrate, at least 5 percent of the time (during low-flow periods) watersupply withdrawal will not be allowed. In addition, the maximumwithdrawal rate considered is equal to 25 percent of the long- termaverage flow rate. Lesser withdrawal rates are also considered.

These minimum streamflow and maximum diversion rate criteria, aswell as the source evaluation methodologies in this TM, werepreviously established for the purpose of this preliminary feasibilityanalysis (CH2M HILL, 1996a). If these criteria change, then all facilityrequirements reported in this TM will also change. This is becausesurface water facility requirements are a direct function of the alloweddiversion criteria; thus, the results reported in this TM are valid onlyfor the previously selected diversion criteria.

Once the withdrawal parameters were established, the maximumeffect on the flow duration relationship for each candidate withdrawalsite was investigated. From this effort, it was determined that thereare little noticeable effects if the entire streamflow regime isconsidered. However, if only the normal and low-flow portions of thestreamflow regime are considered, the water supply withdrawaleffects are more apparent. The water supply withdrawal tends toflatten the flow duration curve at moderate flow rates. At flow ratesless than the minimum withdrawal rate, the flow duration relationshipwould be unchanged because water supply diversion would not beallowed.


The objective of the potential yield analysis is to establish arelationship between the installed river diversion capacity and themaximum long-term volume of water that may be withdrawn forwater supply. This parameter is termed potential yield because itrepresents the maximum possible water supply yield, if sufficientstorage and treatment facilities are constructed to fully use thediverted flow and assuming there is no waste in the treatment process.

The maximum potential yield for the Lake Griffin site is approximately75 percent of the maximum diversion capacity. For the four upstreamSt. Johns River sites (Cocoa to De Land), the maximum potential yieldranges from 81 to 84 percent of the installed maximum diversioncapacity. For the downstream St. Johns River site (above Jacksonville),


65


the maximum potential yield is only 64 percent of the maximuminstalled diversion capacity.


Facilities required to develop the surface water supply at eachcandidate withdrawal site were determined by a combination offacilities simulation and water quality and treatability analyses. Ineach case, required facilities included the following:

• River diversion structure• Off-line raw water storage reservoir• Water treatment plant• ASR system

The simulation analysis estimates the gross reliable yield, which couldbe developed for each site for a selected array of facility sizes. Thewater quality and treatability analysis established the treatmentprocesses needed in the water treatment plant component of the watersupply system. Also, the analysis provides a preliminary quantitativeestimate of the waste water volume produced. These volumetric lossesinclude losses associated with conventional surface water treatment, aswell as the waste concentrate produced by the required desalting. Thetotal volumetric loss must be subtracted from the system gross yield toobtain the net reliable yield, which is the final estimate required. Thenet reliable yield is defined as the average daily demand that can bereliably delivered to the distribution system by the surface watersupply facilities.

The results of the water availability and yield analysis are summarizedin Table 5. Lake Griffin is the only true freshwater site among thecandidate withdrawal sites considered. Desalting will not be requiredat this site and a waste concentrate stream will not be produced.Therefore, the estimated maximum reliable yield for the Lake Griffinwithdrawal site, 28 mgd, is nearly equal to the maximum potentialyield of 30 mgd.

The planned SJRWMD minimum flow and levels analysis for LakeGriffin may further restrict the reliable net yield. Therefore, to providea relatively conservative estimate of Lake Griffin water supplypotential, the maximum net reliable yield, to be evaluated by theUniversity of Florida decision model, will be limited to 50 percent ofthe maximum net reliable yield estimate in this analysis, or 14 mgd(0.5 x 28).


66


Table 5. Water Supply Availability and Yield Analysis Results Summary

CandidateWater SupplyWithdrawal

Site

Lake Griffin (HainesCreek)

St. Johns River nearCocoa

St. Johns River nearTitusville

St. Johns River atSanford

St. Johns River nearDe Land

St. Johns Riverabove Jacksonville

AverageAnnual

Streamflowcfs (mgd)

248(160)

958(619)

1,273(823)

2,425(1,568)

3,043(1,967)

9,129(5,901)

MaximumDiversion

Rate(mgd)

40

154

206

392

492

1,475

MaximumPotential

Yield(mgd)

30

124

166

331

415

953

MaximumReliable

Yield(mgd)

28a

108

143

279

351

419

Note: The maximum yield values for the St. Johns River sites are not independent. These values representcumulative amounts for the candidate withdrawal site and all upstream sites. For example if a 100-mgd reliablewater supply were developed near Titusville the maximum reliable yield at De Land would be reduced by 100 mgdfrom 351 mgd to 251 mgd.

aMaximum reliable yield for Lake Griffin considered in the University of Florida decision model will be limited to70 percent of this value or 19.6 mgd.


67


The four upstream St. Johns River candidate withdrawal sites (Cocoato De Land) are similar to each other. The worst-case water qualitydiverted for use at these sites is classified as either slightly ormoderately brackish. Under many flow conditions, desalting will notbe required. However, some desalting would be required for some ofthe diverted flow. Also, a waste concentrate would be produced,reducing the net reliable yield to about 108 mgd to 351 mgd, or about85 percent of the maximum potential yield. Concentrate volumewould be on the order of 5 percent of the total diverted flow.

The final candidate withdrawal site, the St. Johns River aboveJacksonville, has poor water quality characteristics because ofsignificant tidal mixing. The worst-case water quality at this site issaline or near seawater conditions, with observed chlorideconcentrations of more than 10,000 mg/L. This water would requireextensive desalting, generating a large amount of concentrate. Theestimated maximum reliable yield that could be developed at thislocation is only 44 percent of the maximum potential yield. About35 percent of the diverted flow would become waste concentrate.

These adverse hydraulic and water quality conditions apply only tothe downstream portion of the main stem of the St. Johns River. It ispossible that tributaries to the lower St. Johns River could bedeveloped as viable water supply sources if the withdrawal sites arelocated above tidal influence or behind constructed salinity barriers.Candidate watersheds may include Black Creek in Clay County orDeep Creek in St. Johns County. Evaluation of these hydrologicsystems would be hampered, however, by a lack of basic long-termstreamflow and water quality data.

RECOMMENDATIONSIt is recommended that five of the six candidate surface waterwithdrawal sites be considered further in the cost estimating phase ofthese investigations. These sites are as follows:

• Lake Griffin (Haines Creek)• St. Johns River near Cocoa• St. Johns River near Titusville• St. Johns River at Sanford(Lake Monroe)• St. Johns River near De Land


68


Because of its poor hydrologic and water quality characteristics, thesixth candidate withdrawal site, the St. Johns River above Jacksonville,should not be considered further. The results of this analysis indicatethat this site is not a viable municipal surface water supply source.Thus, it is unlikely that this site would prove to be an economicallyviable municipal water supply source under any foreseeable set ofwithdrawal criteria or other operating conditions.

The cost estimating procedure recommended in this TM should beapproved and applied to the five remaining candidate withdrawalsites.


69

References

REFERENCESCH2M HILL. April 1985. Program documentation Peace River water

supply simulation, prepared for General Development Utilities, Inc.,Miami, Florida. Gainesville, Florida: CH2M HILL.

CH2M HILL. October 1987. Needs and sources report for the PortCharlotte service area, prepared for General Development Utilities,Inc., Port Charlotte, Florida (Draft). Gainesville, Florida:CH2M HILL.

CH2M HILL. June 1993. Peace River regional water supply facilityexpansion plan, prepared for Peace River/Manasota Regional WaterSupply Authority. Gainesville, Florida: CH2M HILL.

CH2M HILL. September 1994. Analysis of water supply potential forarea B, the Everglades buffer strip, and the Hillsboro basin: Phase3b, East coast buffer implementation (C-4123-A3), prepared forSouth Florida Water Management District (Draft). Gainesville,Florida: CH2MHILL.

CH2M HILL. January 1996. Surface Water Data Acquisition andEvaluation Methodology, Technical Memorandum B.l.f AlternativeWater Supply Strategies in the St. Johns River Water ManagementDistrict. Gainesville, FL: CH2M HILL.

CH2M HILL. April 1996. Surface Water Withdrawal Sites, TechnicalMemorandum B.l.h Alternative Water Supply Strategies in theSt. Johns River Water Management District. Gainesville, FL:CH2M HILL.

McMahon, T. A. 1992. Chapter 27: Hydrologic design for water use inHandbook of Hydrology. D. R. Maidment, Editor-in-Chief.New York, NY: McGraw Hill, Inc.

Pyne, R. David G. 1995. Groundwater recharge and wells, a guide toaquifer storage recovery. Boca Raton, Florida: Lewis Publishers.

Wycoff, R. L. 1985. Simulation of a water supply system with aquiferstorage. Paper presented at a stormwater management modelusers' group meeting. Gainesville, Florida.


70

water supply needs and sources assessment...

Documents