Presentations From The .1992 Coal Mining Impoundment.Informational Meeting .

••••U.S. Department of Labor.Mine Safety and Health Administration

Infonnatlonal ReportIR 12061993

PRBSBNTATIONS FROM THB 1992 COAL'MINING IMPOUNDMENTXQORMATIONAL HBE~ING

'I

"

MINE SAFETY AND HEALTH ADMINISTRATIONUNITED STATES DEPARTMENT OF LABOR

OFFICE OF TECHNICAL SUPPORT .

TABLE OF CONTENTSINTRODUCTIONRESERVOIR EVACUATION BY PUMPINGBill CannonEROSION PROTECTION FOR SPILLWAYSstephen W. DmytriwMSBA'S CONCERNS REGARDING SEISMIC ANALYSIS OPCOAL MINE REFUSE EHBANlQIENTSWade E. CooperPHREATIC SURFACEWade E. Cooper. .EFFECTS OFHINING ON DAKS AND IMPOUNDMENTSJohn W.FredlandUSE OF GEOTEXTILES AS A FILTERGeoz:geGardnerCONTROLLING SEEPAGE ALONG THE CONDUITSAbdul HamidDESIGN.OF PIPES FOR EXTERNAL LOADINGDonald KirkwoodPROBABLE MAXIMUM FLOODDaniel S. MazzeiCOMPACTION SPECIFICATIONSstanley Michalek

i. . ... /NEW ADMINISTRATIVE PROCEDURES/ALTERNA~EPLAN REVIEW PROCESSJohn J. MulhernIKPOUNDINGSTRUCTURES SAFETY D~SIGN PROCEDURES;PRESSURE TESTING SPILLWAY CONDUITSHarold OwensFREQUENCY OF MOISTURE DENSITY TESTING TOVERIFY COMPLIANCE WITHCOKPACTION SPECIFICATIONSTerence M. TaylorSHORT.TERM CRITERIALarry WilsonGRADED FILTERSLarry Wilson

l

PAGE NUMBER1

2

5

14

16

22

38

40

47

49

52

56

58

61

62

Appendix A:Appendix B:Appendix C:Appendix D:

Alternatives to the Plan Review ProcessProcedure Instruction LettersPart 77MSHA Contact Lists

ii

1I1'1'RODUCTIONOn May 20 and 21, 1992, the MSRA Coal Mining ImpoundmentInformational Meeting was held at the National Mine Health andsafety Academy in Beckley, West Virginia. Fifteen presentationswere given on key issues involved in the design and constructionof dams associated with coal mining. The sessions were attendedby approximately 180 people representing mining companies,conSUltants, and government. The format allowed seven groups of26 to 30 attendees each to rotate through seven classrooms todiscuss the issues.The attendees were told that to improve the consistency among theplan reviewers, engineers from the Denver and pittsburghTechnical support Centers meet twice annually to discuss specifictechnical issues. It was soon discovered that the topics beingdiscussed needed to be shared with anyone involved with coalwaste dam design, construction, or inspection. The only way toaccomplish that goal was through the issuance of ProcedureInstruction Letters. The Letters present a consensus of.engineering philosophy that could change over time. They do notpresent policy or carry the force of law.The first meeting was held in Denver, Colorado, in September1986. The first PIL, on compaction Specif~cations, was issuedwith an effective date of July 1, 1990. CUrrently, thirteenposition papers have been disseminated and more will follow asthe need arises.

1

RBSBRVOIR BVACUATION BY PUMPINGBill Cannon

aDc!

symmary of PresentationsMr.' cmytriw made a few brief remarks about Technical Support'sapproach toward evaluating open-channel spillways. He addressedthe appropriateness of location, alignment, material excavated orthe lining used to secure permanence during the life of thehydraulic conveyance. It was emphasized tnat overall spillwayquality is a function of facility hazard classification. Forexample, a large, high hazard structure where loss of thespillway could result in dam breach and possible loss of lifedemands there is no substitute for quality. The dam structurea~d hydraulic conveyance must function as designed under the,mostrigorous conditions without assistance from maintenancepersonnel. This position is firm, since it is unlikely thatworkers and equipment would be available when needed and mostlikely could not get to the site due to a myriad of storm relatedconditions and consequences. In this light, most MSHA reviewengineers insist that spillways' for large, hiqh hazard damsshould be excavated though competent rock or the channel linedwith. reinforced concrete to gain an approval recommendation.This example represents one end of the classification qualityresponse spectrum.On the opposite end of the spectrum are spillways designed inconcert with small, low hazard facilities where there is noexpectation for loss of life. Under this condition and inconjunction with discharge velocity, the channel could beadj.acent to or located over the d~m crest and formed fromembankment material. A lining, if required, could consist ofgrasses, riprap, or synthetic materials such as ArmorForm,FabriForm, Enkamat, articulated mats, gabions, pre-formed blocks,and the like. The Agency's reviewers consider these erosionprotection linings in the experimental stage similar togeotextiles for drainage systems more than a decade ago. As newproducts prove their worth and gain acceptance they can beapplied in more hazardous situations.Riprap of'course is not a new prOduct, but new calculativemethods have Deen developed which result in larger rock sizes andblanket thicknesses for equivalent velocities compared to olderdesign techniques. Early in the oversight responsibility, MSHAengineers relied on methods advocated by the Federal HighwayAdministration (FHWA). Through utilization, implementation, andfield performance, the Technical' Support Centers learned thatmany riprap linings failed at less than the design discharge. As

2

a result of these findings MaRA undertook a study to compare sixseparate design methods. The methods included FHWA',u. S. Bureauof Reclamation (USBR), Corps of Engineers (CE), CaliforniaDepartment of Transportation (CALTRAN), Columbia 'River Model(CRM), and Simons and Senturk. Using the USBR method as,thestandard, the various techniques yield,maximum stone sizevariations as follows: FHWA -0~76; USBR -1.0; CE -2.04;CALTRAN -2.67; CRM ~1.45; and Simons and Senturk-3.16.This normally concluded Mr. Dmytriw's considered remarks andthe floor was thrown open for questions. :Questions received,generally fall into six broad categories that are, addressed laterin this report. 'Mr. Bill Cannon addressed Reservoir Evacuation by Pumping bysimply reading the PIL inclUding several observational anecdotes.This SUbject is the least technical of the 13 Letters issued,contains no bibliography, and few questions were asked.During each session, either Bill or Stephen asked the 'attendeesif the PIL's have generally been helpfUl or simply an unnecessaryexercise. The general cons~nsus leaned toward helpful in thatthe Letters gave guidance and direction to the design engineer.Several attendees suggested additional PIL topics during theseven periods of lecture and discussion. They include (1) aguide to a more complete open-channel spillway alignmentexploratory investigation to define subsurface conditions androck quality, (2) the 'advantages and pitfalls of placingemergency spillways over the dam crest, and (3) reinforcedconcrete testing and inspection features. A non-Federalregulator suggested that MaRA prohibit installation of open-channel spillways over or adjacent to embankment dams whileanother stated that MSRA should insist upon greater concretetesting variety and frequency.

Questions Regarding presentation1. What is MSBA's policy toward multiple stage, open-channelspillways; changing, say, every ,four months?Spillway quality is a functio~ of facility hazard classification;that is, the consequences of failure dictate how substantial ordurable the spillway and/or lining must be constructed towithstand the erosional forces expected. If failure of aspillway can lead to failure' of the embankment, the channeland/or lining shall be desi~ned and constructed to preclUdefailure. Under the conditions described, if a company 'or an agentof the company chooses to design and construct numerous short-term spillways for a large, high hazard dam, each channel shallbe SUbstantial. In most instances, the channel will be eithercarved into competent rock or lined with reinforced concrete withsuitable seepage control measures. As the Agency develops more

3

Sadly, the answer is yes. It.is unfortuna~e that designers whopay att~ntion to details and companies that follow the approvedplan are often penalized by the delays in the, review processcaused by incompetent or shoddy professional work and operatorswho knowingly deviate from the drawings and specifications.4. Why is HSSA reluctant to adopt or permit utilization of newsynthetic spillway liners such as PabriPorm, ArmorForm, andarticulated mats?MSHA is utilizing essentially the approach the Mine WasteDivisions used when geotextiles were developed for drainagesystems. Our reviewers; for the most part, do not want toexperiment with unproven products where failure of the productcould lead to dam failure and probable loss of life. Most of theproducts being introduced have not been subj.ected to flood flowswith the attendant debris that is expected from the storms weanticipate.

4

MSKA'S CONCERNS REGARDING SEISMIC ANALYSIS OF COAL MINE ~FUSEBMBANlQIBNTS

Wade E. Cooper

I've been asked to talk about MSHA's concerns regarding seismicstability analysis of coal mine refuse embankments.· Morespecifically, MSHA is concerned with the procedures used toevaluate the seismic stability of moderate and high hazardimpoundments constructed using either centerline or upstreamconstruction methods. ~his talk will be concerned with onlythese types of embankments constructed· on nonliquefiablefoundations. For the limited time provided, this talk simplycannot cover all aspects of seismic s~ability analysis. The talkprimarily covers the most commonly used methods concerning onlyliquefaction and deformation, and·MSHA's concerns in this area.For example, this talk does not address pore p~essure dissipationafter the earthquake, reservoir seiches, and active faults withinthe embankment foundation. .Anyone who has designed a coal refuse. embankment using eithercenterline or upstream construction methods should, by now, knowthat the task of evaluating seismic stability is a complex,tedious, and potentially costly process with many uncertainties.The complexities and uncertainties make it extremely difficultfor both MSHA and designers to make a determination· as to whatconstitutes current, prudent engineering practice. It must berecognized that seismic stability analysis is of concern oniywhen there is the potential for a catastrophic failure of theembankment due to seismicity. This usually equates to asignificant loss or settlement of the embankment crest whichcould result in loss of life or significant damage to majorstructures.In performing a seismic stability analysis of an embankment thereare primarily three main components.:

1. Select appropriate seismic ground motion parameters;2. Liquefaction analysis; and3. Deformation analysis.

An accurate and ·reli~ble assessment in each of these componentsis vital for properly evaluating the seismic stability of theembankment. Each one will be discussed separately.

5

GROUND MOTION PARAMETERSThe objective of the design earthquake evaluation is to obtainthe ground motion parameters at the site to be used in thedynamic response analysis for assessing liquefaction anddeformation. Current Federal guidelines (~)' indicate thatextensive investigations and design analyses may be required forhigh hazard dams or where seismic conditions are ,severe. Theguidelines serve as an excellent tool' for the factors which needto'be considered in performing a local or si~e specificseismotectonic study. For moderate and high hazard dams, wherefailure of the embankment could cause loss of life or significantdamage to major structures, the controlling maximum credibleearthquake (CMCE) appears to be appropriate as the designearthquake. Procedures for determining the CMCE are presented inthe guidelines and will not be reiterated here. In regions ofthe united States where active faults are n9t well defined, suchas in many eastern 'states, the'Bureau of Reclamation (~)indicates that they determine earthquake loading in aprobabilistic manner. They also indicate that they exclude usageof earthquakes with a probability of occurrence of less than

.2E-S. MSRA is in the process of seeking current seismic criteria

.used by other Federal agencies involved' in earthen damconstruction.MSRA has some concern that the selection of the CMCE is a highlycomplex procedure and is usually out of the realm of expertisefor most registered engineers certify-ing design plans as well asMSHA reviewers. For this reason, it may be prUdent to leavedetermination of th~ CMCE up to recognized seismologists. In.addition, MSRA recognizes that it needs to perform more researchto determine acceptable probabilistic seismic design criteria.One of the more important aspects in any liquefaction anddeformation assessment is determining the dynamic response of theembankment. The dynamic response analysis determines the groundmotion parameters (such as acceleration, velocity, shear stressand shear strain) within the embankment during the earthquakemotion. These motions are needed for deformation analyses andmay be needed for application in laboratory tests used todetermine pore pressure development, strains, and whether or notliquefaction can be expected. .Many computerized models are currently available for performingthe dynamic response analysis. Selecting which model to use canbe a complicated task. Some of the computerized models proc~edeven further and contain models for determining liquefaction anddeformation based on soil input data derived from laboratorytests. Of the available computerized models for determiningdynamic response al,one, SHAKE is .probably the most common. It is

'Underlined numbers. in parentheses refer to the list ofreferences at the end of this report.

6

a one-dimensional linear elastic analysis method which assumeslevel ground conditions utilizing total stress concepts. Somereferences (~,~) indicate that it can significantly underestimateboth accelerations and cyclic shear stresses near the crest ofthe embankment because it does not take into account thegeometrical shape of the embankment. This appears 'to indicatethat two-dimensional dynamic response analyses may be moreappropriate for embankments. SHAKE may also result in a greaterresponse than any of the nonlinear analysis methods because ofresonance caused by the site period matching the period of theinput motion. Many of the more recent methods incorporateeffective stress concepts as well as nonlinear stress-strainproperties and may result in more reliable results. At thistime, there appears to be no consensus about which method is mostappropriate for coal refuse embankments.Dynamic soil parameters are needed'for input into all of thecomputerized dynamic response analysis programs. Theseparameters include the shear modulus, damping coefficient, bUlk,modUlUS, and how they change with shear strain. Many of theavailable programs contain typical soil input parameter value~.However, most of the values were obtained on natural' soils whichhave a higher specific gravity and different granular shape thanfine coal refuse. Their validity for coal 'refuse appearsquestionable. Rather thanusinq the software provided values, itappears necessary to perform adequate laboratory and field teststo determine these dynamic parameters~ Instead of relying onlaboratory tests to determine the in situ shear modulus, mostexperts recommend measuring in situ shear wave velocity viaeither cross-hole or down-hole techniques. Dynamic propertiesare usually determined using laboratory tests, such as eitherstrain or stress-controlled cyclic triaxial tests, on preferablyundisturbed samples. 'MSHA has some concern about (1),whether thelaboratory samples are truly undisturbed, (2) whether the effectsdue to disturbance are significant and how significant, (3)whether it would be better to reconstitute the samples, and (4)what are the uncertainties in response due to using reconstitutedsamples. ' ,The Committee on Earthquake Engineering (7) provides some'information indicating that pore pressure development isfundamentally more related to cyclic strain than to cyclicstress. This casts some concern as to Whether stress-controlledor strain-controlled laboratory tests should be performed forassessing dynamic soil properties.

7

LIQUEFACTION ANALYSISIn evaluating the stability of the embankment againstliquefaction, current practice allows the designer ess·entially.two choices: either assume liquefaction and assign a shear.strength to the liquefied soil or assess liquefaction bydetermining if the earthquak~ is of sufficient magnitude andduration to cause liquefaction. In this report, liquefaction ofa soil is considered to occur when the undrained steady-stateshear strength, as a result of strain sof~eninq, is the availableshear strength in the soil. Care must be taken to not confuse itwith cyclic mobility or limited liquefaction as defined by vaidand Chern (11). Soils that are dilative in situ need not beevaluated for liquefaction, but should be evaluated for strainand deformation. Dilative soils do not liquefy because theirundrained strength is greater than their drained strength.(A) Assume liguefaction:Assume that the fine refuse liquefiesas a result of the design earthquake and perform a post-earthquake static stability analysis to demonstrate that theembankment is statically stable after the earthquake. If theembankment is not statically stable (i.e., fails) in the.liquefied mode, then additional analyses are necessary todetermine if the design earthquake is of SUfficient magnitude andduration to cause liquefaction.This type of analysis appears rather straight forward, however,the difficulties or pitfalls of the method are not so apparent.One of the major difficulties of the method is in choosing theappropriate shear strengths for the soils assumed to beliquefied. Basically three shear strengths are commonly used forthe liquefied soils:

1) Assume the worst case; i.e., zero shear strength.Obviously, this case would not provide adequate factors ofsafety for embankments constructed using upstreamconstruction methods. However, it may provide adequatefactors of safety for centerline constructed embankmentswith wide crests.2) Another procedure, advocated by many experts, is to useresidual shear strengths derived empirically from standardpenetration test results (SPT)'or cone penetration testresults correlated with SPT. However, in using these, onemust recognize that they were determined from site specificdata on natural soils which. have a higher specific gravitythan fine refuse. Therefore, their applicability ·forcoalrefuse facilities is certainly questionable. In addition, aspecific blow count shows a wide range of residual shearstrengths. Choosing which residual strength ·over the widerange could pose a problem. .

8

3) Without the availability of appropria~e empiricalcorrelations of residual strength, many designers turn tothe field and laboratory test methods presented by Pouloset. ale (~,Z,d,12)to determine the undrained steady-stateshear strength of the liquefiable soil. This method requiresrelatively undisturbed samples from the field forconsolidated-undrained triaxial tests in the laboratory.The method is highly dependent on accurate measurement ofthe in situ void ratio. Although the method appears to berelatively straight forward for determining the undrainedsteady-state shear strength, there can be significant shearstrength uncertainties due to the relatively flat slope ofthe steady-state line (void ratio vs log of the steady stateshear strength) and the sUbjective amount of void ratioadjustment. The method is highly dependent on obtainingrelatively "undisturbed" samples. Some consultants contendthat "undisturbed" samples of fine refuse may be obtainedbecause of the slight cohesiveness and plastioity of mostfine refuse. MSHA still has some concerns whether or notrelatively "undisturbed" samples can be obtained forlaboratory analysis. In addition, some recent researchdevelopments suggest factors other than in situ density,such as stress path, dynamic pore pressure fluctuation, soilfabric, and stratification, may also play a role in theactual steady-state strength available in the field (~,~).Another method of estimating the undrained steady-stateshear strength that has been proposed is to use in situ vaneshear tests. MSHA has some concerns that this method mayoverestimate undrained shear strengths due to the effects ofplasticity and the potential for the presence of granularparticles (.a).

other concerns that arise include whether or not adjustments inavailable shear strength should be made for the nonliquefiablesoils and deciding on what is an acceptable factor of safety forthe stability analysis. The Bureau of Reclamation (~), as wellas some other, use a shear strength reduction of up to 20 percentdepending on the type of soil, its standard Proctor density; andconsolidation. Acceptable factors of safety for the iiquefiedmode appear to range from about 1.0 to 1~5 in the literature,depending greatly on the reliability of the shear strengths usedin the analysis.Although this method has some advantages over performing alaboratory based "triggering" analysis, it still must becarefUlly used before it can be ·considered a reliable analysis.(B) ASSESS LIQUEFACTION POTENTIALThere are basically two approaches to evaluating liquefactionpotential: either use empirical methods based on field tests oruse laboratory tests in conjunction with results of dynamicresponse analyses. .

9

1) Empirical Methods,: The majority of empirical methods,such as SPT, CPT, or liquefaction potential graphs, forassessing liquefaction potential are based on natural soilswith much higher specific gravities than fine refuse.Again, for this reason, the use of these empirical methodsfor assessing liquefaction potential of coal refuse .facilities is questionable.However, it may be possible to obtain a database which may .at some latter date show a correlation between the empiricalmethods and the laboratory methods used on fine refuse.MSHA believes that the empirical methods are not bythemselves adequate at this time for verifyingnonliquefaction of fine refuse. The empirical methodresults can provide additional information for making aninformed decision as to Whether or not liquefaction can beexpected. Some of the more recent updates in empiricalmethods for evaluating liquefaction potential are presentedby the Bureau of Reclamation (~) and Seed and Harder (6).2) Laboratory "Triggering" Method: Once the dynamicresponse analysis is completed and undisturbed samples ofthe fine refuse are obtained, laboratory testing to asse~spore pressure increase, strain, and liquefaction potentialcan be performed. The most common laboratory testing methodis the stress-controlled cyclic triaxial test under .undrained conditions on samples consolidated to their insitu effective stress. Other methods such as·torsionalshear and strain-control-led tests are reported in theliterature. In order to obtain the in situ effective stress(both horizontal and vertical), it may be necessary toperform a finite element analysis Which contains a model forthe stress-strain behavior of the soils. The shear stressesobtained from the dynamic response analysis are normallyconverted into an equivalent number of cyclic shea~ stressesfor use in the stress-controlled cyclic triaxial tests.Measurements in the test include stress, strain, and porepressure, which are used to determine pore pressuredevelopment and whether liquefaction, based on the'amount ofstrain,can be anticipated within th~ soil.

MSHA has many' of the same concerns with this laboratory test asit does with all·laboratory tests on "undisturbed" samples.These concerns are whether or not the sample is trulyundisturbed, whether the in situ effective consolidation stressesare properly modeled, and whether the initial states, such asvoid ratio, are properly modeled. It is well documented ·thatsample dis~urbance and confining effective stresses can havemajor effects on pore pressure and strain development.Anisotropic consolidation conditions can have opposite effectsdepending on the initial states of the soil (111; therefore,proper modeling of the initial states of the soil in the.laboratory is vitally important. Considerable uncertainties in

10

strain development from cyclic triaxial tests are appa~ent forsoils which have high permeability and no cohesion (11).The Bureau of Reclamation (~) presents the following concerningthe use of laboratory tests for assessing liquefaction potential:

"Finally, the triggering analysis using laboratory tests isthe last bastion of hope in tQe effort to disprove or proveliquefaction. It should only be considered for applicationto materials which have some cohesion and then should beused with considerable jUdgment and a comfortable margin· of

.safety with regard to strain (e.g., a factor of safety> 2)because if the phased approach is employed as describedpreviously, use of the triggering analysis as a basis forruling out liquefaction implies post liquefactioninstability."

In addition, the Committee on Earthquake Engineering (7) presentthe following concerning laboratory tests for predicting porepressure buildup:

"The major difficulty lies in the parameters relating to therate of pore pressure generation. ~he earliest efforts topredict pore pressures used results from laboratory tests onspecimens reconstituted to the in situ void ratio orsometimes "undisturbed" samples. The'practiceof predictingpore pressures solely on the .basis of laboratory tests,without the benefit of in situ measurements such as SPT orshear wave velocity, is now recommended only if great .precaution is taken to obtain samples with least·disturbance. such practice is particularly important .forsome types of soil for which there is as yet little or noexperience."

DEFORMATIONA deformation analysis must be performed for ~mbankments that donot liquefy, as well as those that do, to ensure the safety ofthe embankment. There are currently two main approaches used toevaluate deformation as a result of the design earthquake. TheNewmark approach (10) is commonly used and has been modeled inseveral computerized software programs. Meanwhile, thecomputerized finite element or finite difference methods arebecoming much more common. There are so many finite element andfinite difference models available that it is extremely difficultto access their validity. Some experts rely heavily on the Useof.a specific model for assessing deformation. Practically all ofthe computerized methods rely heavily on modeling the dynamicproperties of soils from laboratory test results such as thosefrom cyclic triaxial tests or consolidated-undrained triaxialtests. MSHA's main concern with all of these computerized

11

methods is that it is extremely difficult to determine if theyproperly or conservatively model the shear strengths and strains·of the soils during the earthquake motion. This causes problemsfor MSHA as.well as others in evaluating the appropriateness ofthe methods for evaluating deformation of fine refuseembankments.

CONCLUSIONSMSHA has some ~oncern that the empirical methods of liquefactionassessment may not be appropriate due to the difference inspecific gravity of the soils used in the empirical methods andthe soils used in coal refuse embankments. However, it appearsthey can provide additional insight into assessing the potentialfor liquefaction and possibly supply data which may show .correlation with laboratory methods. MS~ also has many 90ncernsabout the validity of laboratory methods for assessingliquefaction and deformation of coal refuse embankments becauseof the problems of obtaining "undisturbed" samples for testingand the apparently drastic effects of disturbance on testresults. This is particularly important since practically all ofthe computerized methods for assessing liquefaction andestimating deformation are based on models using parametersobtained from laboratory tests. There can be considerableuncertaint~es in liquefaction and deformation assessments basedon laboratory and field tests for coal refuse embankments.. ~t isevident that considerable judgement must be exercised inevaluating the seismic stability of coal refuse embankmentsconstructed using either c~nterline or upstream constructionmethods.

REFERENCES1.' Bureau of Reclamation, 1989, "Chapter 13, Seismic Design andAnalysis," Embankment Dams Design standards, U.S. Dept. ofInterior, 77p.2. Poulos, S.J., G. Castro, and J.W. France, 1985,"Liquefaction Evaluation Procedure," Journal of GeotechnicalEngineering, ASCE 111(b), pp. 772-791.3. Poulos, S.J., 1981,' tiTheSteady State of Deformation,Journal of the Geotechnical Engineering Division, ASCE, Vol. 107,No. GT5, May, pp. 513-562.4. Lo R.C., E. Klohn, and W.O. Liam Finn, 1991, "Shear strengthof Cohesionless Materials Under Seismic Loadings," IX PanamericanConference Proceedings,pp. 1047-1052.5. Holtz, R.D., and W.O. Kovacs, 1981, "An Introduction toGeotechnical Engineering," Prentice-Hall civil Engineering andEngineering. Mechanics Series, pp. 572-577. I

12

6. Seed, R.B., and L.F. Harder, Jr., 1990, "SPT-Based Analysisof cyclic Pore Pressure Generation and Undrained Residualstrength," Proceedings, H. Bolton Seed Memorial symposium, J.Michael Duncan (ed.), Vol. 2, May, pp.351-376.7. Committee on Earthquake Engineering, 1985, "Liquefaction ofSoils During Earthquakes," National Academy Press, Washington,D.C., pp • 89":'192• .8. Design Earthquake Task Group Subcommittee 1 Interagency Committeeon Dam Safety, 1985, "Federal Guidelines for Earthquake Analyses andDesign of Dams," Federal Emergency Management Agency, March, 45 p.9. Finn, W.D. Liam, 1990, "Analysis of Post-Liquefaction Deformationsin soil Structures," H. Bolton Seed Memorial proceedings, Vol. 2, J.Michael Duncan (ed.), May, 21p.10·, Newmark, N .M., 1965, "Effects of Earthquakes on Dams andEmbankments," Geotechnique, vol. 15, No.2, pp. 139-i60.11. Vaid, Y.P., and J.C. Chern, 1985, "Cyclic and Monotonic UndrainedResponse of Sa1;:uratedSands," Proceedings of Geotechnical EngineeringDivision, ASCE Convention, October 24, 1985, Advances in the Art ofTesting Soils Under Cyclic Conditions. pp. 121-147.12. Poulos, steve J., 1986, "Liquefaction and Related Phenomena,."Advanced Dam Engineering for Design. Construction. and Rehabilitation.

13

PHREATIC SURFACEWade E. Cooper

summary of PresentationThe phreatic surface within embankments must be conservativelydepicted or evaluated to calculate the minimum slope stabilityfactors of safety for both static and earthquake conditions.The phreatic surface must be cons~rvatively depicted or eva~uatedfor the long-term steady state seepage condition assuming thewater surface is maintained at the elevation of the lowe$tungated water outlet. If a rapid reservoir drawdown stabilityanalysis is provided, a phreatic surface evaluation for thiscondition may be required.Many different methods are currently available for determiningphreatic surface and seepage quantities. The computerizedfinite-element methods are becoming increasingly popular •.However, no matter which method is used, extreme care must beexercised to ensure that the assumptions inherent in the methodand procedures are fUlly satisfied or do not significantly af~ectthe results.For design purposes, ~ minimum horizontal to vertical permeabil-ity ratio of at least 9 appears appropriate. Lower ratios may beallowed provided they are adequately substantiated anddocumented.Drains which have been used to lower the phreatiq surface used inthe stability analysis must be properly designed. They must bedesigned for material compatibility, relative permeability,minimum thicknesses of at least 3 feet, and with seepage capacityfactors of safety of at least 10.

Questions Re9ardinq Presentation1. In recent years, more emphasis has been placed on 95 percentstandard Proctor maximum dry density instead of 90 percent. Isthere a difference in the horizontal to vertical permeabilityratio due to the hiqher compaction requirement?There may be some change due to the higher compaction, however,we do not know for sure whether there is and how much it is,although we estimate it to be inconsequential.2. Is it aqooa idea to install piezometers near internal drains·to assess drain operation? .Generally "no" for several reasons. Proper operation of drainsis normally evaluated through monitoring effluent characteristicsof the drain such as quantity of flow and amount of fines.Piezometers are normally installed to monitor the phreatic

14

surface in the embankment. The readings'are used to evaluate thestability of the embankment. They may be used for assessing.drain operation, however, it is not considered comm~n practice.3. ODe person oommented that he had some problems with using aminimum horizontal to vertioal permeabil.ity ratio of 9.

with regards to this comment we could only reiterate what isstated in the Procedure Instruction Letter. The minimum ratio of9 is based on a thorough searoh of the available literature whichindicates the ratio can vary from 1 to over 100 •. It is based onwhat appears to be a reasonable minimum for design purposes. Alower ratio may be utilized provided it is fUlly substantiatedand depending on how'crucial the ratio is to the safety of theembankment. MSHA reviewers would be more inclined to accept alower ratio provided its use does'not significantly affect thestability or safety of the embankment.4. Why do we need to determine ooefficients of permeability forhomogeneous embankments without drains?There is really no need to determine coefficients of permeabilityfor this case, except when someone is interested in obtaining anestimate of the total anticipated seepage quantity.

Additional DisoussionIt was commented that one operator kept dozer operators and otherequipment operators from breaking piezometers by installing au.s. flag on each piezometer. This apparently kept the operatorsa significant distance from the piezometer. Bicycle pennants orreflective driveway markers on flexible poles serve equally wellas markers and raise no questions regarding flag protocol.

15

EFFECTS OF MINING ON DAKS AND IMPOUNDMENTS

John W. Fredland

Summary of presentationIf underground mining occurs near a dam, the ground disturbancescaused by the mining can damage the dam. In the design' andconstruction of a dam, a great deal of effort and expense istaken to ensure that seepage will be controlled, and piping, orinternal erosion, will not occur. In this regard, measures aretaken such as toundation exploration and testing, foundationpreparation, placement of drains, and controlled compaction ofmaterials. Furthermore, careful analyses are performed todetermine the required minimum freeboard, and to design·structurally safe decant pipes. The effects of mining can undoany or all of these measures and jeopardize the safety of a dam.The 'ground movements induced by mining (see Figure 1) can causethe .opening of joints in the foundation, cracking of embankmentmaterials, damage to decant pipes, loss of freeboard, and otheradverse effects. It is for these reasons that a "safety zone" isrecommended under and around dams. An example of a safety zoneis shown in Figure 2. This is a zone where no mining takesplace. ~f mining is proposed within the "safety zone," thendesigners 'should realize that 1.) the design and justificationwill be more involved than normal, and 2.) the dam design willneed to include design measures to compensate for potentialmining effects. The Procedure Instruction Letter includes alisting of some of these design measures.Because of the uncertainties involved in predicting both theeffects.of mining (pillar strengths, sinkhole development, mininginduced surface strains, etc.) and the response of the foundationand the dam itself to differential movements, designers must takea conservative approach to this issue.

Questions Regarding Presentation1. 'Does the Procedure Instruction Letter mean that nomining is.permitted under a dam?No. It is not the intent of the PIL to totally prohibit miningunder a dam. However, as indicated in the PIL, plans whichpropose mining under a dam will not be approved unless thepotential effects of the mining, and the associated uncertainty,are fully taken into account, and a complete, well-documentedanalysis is provided.Because of the nature of the ground disturbances created by fullextraction mining, such mining is norma11y.not permitted under adam and must be kept a safe distance away from a dam.

16

2. What about building a dam over an old room-and-pillararea?With room-and-pillar mining, MSHA would be concerned that thestrength of the pillars would be adequate to provide the requiredlong-term support. The possibilitY,of the pillars punching intoa moisture-softened fireclay layer beneath the coal seam would bea concern. Sinkholes can develop in areas of shallow rock cover.All of these potential effects would need to be specificallyinvestigated and analyzed. Testing should be performed as neededto Characterize the properties of the materials involved.In the case of old abandoned workings, an additional concernwould be for the accuracy of the older mine maps, especially withrespect to the robbing of pillars, Which may not be reflected onthe map. The site would,need to be explored toa sufficientextent to allow the accuracy of the mine map to be verified.3. What about buildinq a dam over lonqwalled areas?The mining of longwall panels will affect the surface atvirtually any depth of mining. Since total extraction miningcreates zones of tension on the surface, it can be particularlydang~rous in the vicinity of dams. Much uncertainty is ,associated with determining how the strains will be distributedon the surface, estimating how much strain will occur at aparticular location, and predicting the impact of the strains onthe foundation. The loading of the dam itself may causeadditional movements.If dam construction is proposed over an area which has alreadybeen longwalled, the designer would need to address issues suchas the following: are the mining induced movements completed andhas the area stabilized; how has the mining affected thefoundation with respect to its permeability: have cracks occurredor joints opened up which could provide a path for excessiveseepage, or for piping: how will the foundation exploration,program identify the effects of the mining; how will thefoundation preparation compensate for the mining effects; whatcompensating features should be included in the dam design; whattype of monitoring should be done to ensure that the design isworking as anticipated.4. our company is considering lonqwall mining under a highhazard dam. What's your opinion?As previously stated, longwall mining will affect the surface,and the impact cannot be predicted with a high degree ofconfidence. For these reasons, longwall mining under a dam isnot normally permitted. Mine planning, so that a slurry damwould be undermined after it is filled up and capped ,off, forexample, is recommended over mining near the impoundment When itis active.

17

5. Can sUbsidence be conside~ed to be over after a certainperiod of time, say 20 years?With·.room-and-pillar mining, .subsidence incidents have beendocumented to have occurred 50 years or more after the mining.In many older mines, long, narrow pillars were left in place.These pillars are particula~lY'susceptible to deterioration overtime because of the lack of confinement of the pillar core area,and because locally overstressed pillars can cause higher loadsto be transferred to neighboring pillars. Deterioration of roofsupports, especially timbers, is also a contributing factor.The majority of the subsidence which is related to longwallmining normally occurs within a couple of months of the mining.Further subsidence would depend on the na1:;.ureof the.overburden.strata, wat~r conditions, and, probably to a small extent in mostcases, the impact of surface loadings.In the case of total extraction, actual records of subsidence inthe area, under similar conditions, are probably the bestindicator of whether movement may have stabilized. Where thesafety of a high hazard dam is at issue, conservative assumptionsof the potential for movement must be made.6. In defining a safety zone around a dam, how is the drawa~qle determined?The angle of draw dellneates the surface area influenced byunderground mining. It is the vertical angle between a ·linedrawn vertically at the edge of the mining, and a line drawn fromthe edge of the mining to the nearest point on the surface whereno movement occurred. Draw angles from 5 to 45 degrees are foundin the literature, with the more common valu~s being from 15 to25 degrees. The draw angle will depend on the nature of theoverburden.preliminary estimates of draw angles can be obtained from thesubsidence literature. Draw angles should normally be based onmeasurements taken in the area of concern, under similarconditions of overburden and depth. Draw angles ~hould beverified by surveys at the dam site.7. How do you determine the rock properties needed in aSUbsidence analysis? .In dealing with a high hazard dam, site specific properties areneeded. This means that, in most cases, samples of the rock andcoal from the dam site wi.llneed to be obtained for testing.Coal isa difficult material to sample and test. Test resultsmay need to be adjusted to take into account the fact that onlystronger samples may survive the sampling and preparationprocess.

18

8. In determining the safety of pillars, what methods arerecommended by MSRA?A variety of pillar design methods have been developed.. For agiven set of conditions, these methods will predict a range ofsafety factors. In dealing with the safety o~ dams, aconservative approach needs to be taken. Therefore, MSHArecommends that one of the more conservative methods, such asHolland's method, should be used, with a conservative factor·ofsafety.9. Which subsidence prediction model does HSRA recommend?MSHA doesn't endorse a particular subsidence prediction model.The designer must investigate the available methods of analysesand determine which is most applicable to the particular site.Designers should keep in mind that the models are based onempirical data, and so they are generally considered to beapplicable only for the area.where the empirical data came from.10. What happens when one company has constructed animpoundment with an approved plan, and another companyhas the mineral rights for the coal under theimpoundment, and wants to mine i~?If the impoundment plan was approved without. the m1n1ng havingbeen taken into account, then mining near the impoundment wquldnot be consistent with the approved plan.· MSHA would require ~hecompany with the impoundment to show that the impoundment wouldbe safe with the mining, or to.indicate what measures would betaken to ensure the dam's safety. In other words, the companywith the impoundment would have to have their approved planmodified to account for the mining.MSHA would encourage the two companies to attempt to come to anacceptable arrangement between themselves, so·that the safety ofthe dam, and the mine, would both be safeguarded.

19

SURFA

CE

CRACKS

DUE

TOSU

BSID

ENCE

.ONEOFINCREASED

PERMEADIUTV

-OR

IGIN

AL·SUR

FACE

SUBSIDENCE

TROUGff'

-T

-"~

.•.••••

,•••

#•••

,•••

'#f·t~

•...

'..,

"f.'·-

.."..

..,...,..,"

".

.•,..'

,..,

,',

............!.~

•.••.Ill.•

•••••••

4"•••

"•

••,

'"."

•-

•••••••

~•.•.

•t"

•.•••••.•

,,:.,

.••.'

••'"

50FT

•

~BED

SEPA

RATI

ONS

AQUI

CLUD

EZON

E(0

-50

FT.

TO30

t)

~.~'l;~~~(,"l'.,.,,"~~

••..''~4~~~'8::\;~~,}-1

~:.",.,;,};;.>.:.1.,•'",.•,~.

...,..

•'...,...ro,."...,'."'.~.'.".,.....•r

~':<=';t\?f;~;W

~1;;!:

~:'::...~:';"""1>1';"':~~'::o?~_~fr;;!t~+S~~'~;·>

:;,.~:'.:

30t

TO58t

·'.'.--.

.•••

''''.._

ll~~~..

.''..

,...-.,~

•.-

.-.

I.'.•••

.'.··C···"·_'.'

•.••

.••..

.._

.".'....."....",e

:'......

..:'.

'.,.

.'.1'1

CAVE

DZO

NE0L'................'.#.

•·-

".'.~-"....•.'"..,,'..;.;.."

.•>.'

'.,~.

..";.'

'".1.'.

•.;:

".-.:

.;.~.'::'~.".:.:.~'.:.,""'."

.".

).°8...

3tTO

6tl't·()n

••.·..

..";,

....•;C-•.••'r.

""';'

'.

·•'...

.•.-,

...•

'.or,t'

I~..

....".

'.'..!..'

...:.','

".~.'

'i''-t'

\.,;:

••~

"~\.'

.,.,.-'.()

••00'0.-

..'.

'..."

""'''''''~~

_,,!!,~

~.

.•-'.

..-•..

.~=<1u.~~

.'COAL

SEAII

tI

GE.NE.8All~E.D_

SKEIc.I:fLIOJ"AL-.EXIBAcTION

FIGU

RENO

.1

:>jll18h-Water. Hark-1

%mpouudedSIS~faceWate~ ~r-_l_~_O_:--ir--.~----

HomalExt~actiol1 5t ••Pemitted ,., •••

~

Zoueof;foc !xt~actionlI'\ ••"" ...

Zone of NoExtractionTo P~oc.ct

))am

. Zone of !."(cractionUsing Guidelines

D is Oecer.ained From the iollo~~8table:

IJi'loom & PUla~ 58 or lOtPanel & 'tilar 30 or 270 fc.'roca! - 60c* Whichevell' Value 1sLa~8ei:

Safe~y Zone Benea~h SurfaceSt::uc:tureInrpoundinq LuqeBodY'of Surface Wa~er(af~er Babcock and Hooker, 1977.)

(Ie 8741) .

FIGURE ~O. 2

21

USB OF GBOTBXTILBS AS A ~ILTBRGeorge Gardner

summa£! ot PresentationThe four primary d~sign ~~iteria were emphasized: soil.Retention,Permeability, Clogging Resistance, and Ability to Surv1ve.Installation Stresses·(i.e. ultraviolet resistance, resistance totearing, puncturing, etc.) (Fig. 1).Characterization of a fabric by Apparent opening Size (AOS) wasdiscussed (Fig. 2). Mechanisms of adequate filtration andclogging were illustrated with drawings from the literature(Figs. 3 and 4). Various semi-empirical criteria for particleretention (Fig. 5) and fabric permeability (Fig. 6) werementioned.Empirical clogging criteria were shown (Fig. 7). In addition tomeeting such criteria, in critical applications where clogging ofthe fabric may lead.to failure of the dam, MSHA believes that asoil-fabric interaction test is warranted. The .advantages andlimitations of using the Gradient Ratio Test (Fig. 8) to measureclogging resistance were discussed. While the Gradient RatioTest has anASTM standard, and can be run in a shorter timeperiod than a long-term flow .test, it is generally considered tobe more of an "index test" than a "performance test." Test data'from the literature was used to illustrate tha~ with somesoil/fabric systems there may be a significant increase in thegradient ratio with increasing silt content (Fig. 9).The Long-Term FloW Test simply involves testing the soil/fabricsystem in a permeameter. This type of test seems to bepreferable, however, the test may .take a consi~erable period oftime to run (depending on the permeability of the soil/fabricsystem), and may be influenced by laboratory conditions which maynot exist in the field. It appears that in justifying criticalinstallations, the Long-Term Flow Test is more appropriate thanthe Gradient Ratio Test for verifying that the fabric will notclog. The results of several Lonq-Term Flow Tests from theliterature are illustrated in Fig. 10. The second segment of thebi-linear c~rve is generally used as an indication of whether thesystem can be expected to clog over a long period of time.Although ASTM has not yet developed a standard for a Long-TermFlow Test, various procedures are available such as that used bythe Geotextile Research Institute. Figure 11 includes acomparison of Gradient Ratio and Long-Term Flow Test results. Itcan be seen that, under some cond~tions, the Gradient Ratio Test,when performed for a 24-hour period, may'not yet be simulatingthe long-term behavior. A note from ASTM D 5101 - 90, shown atthe bottom of this figure, suggests that this test may, at times,

22

need to be run for longer times. other newer tests such as theFine Fraction Filtration (F3) Test may also be considered as theybegin ~o gain aooepeanoe by ehe engineering community.Often, when considering coarse refuse materials, the potentialfor clogging will be much more of a concern than the possibilityof soil passing through the fabric. Therefore, to improve thefactor of safety against clogging, it is recommended that thelargest practical apparent opening size (AOS), which is alsoconsistent with the particle retention criteria, should be used.It was emphasized that MSHA requires the use of piezometers tomonitor the phreatic surface in cases where a geotextile fabricis used to wrap the underdrain (and failure of the underdrainwould effect the stability of the dam) •. The purpose ofpiezometers is to verify that the phreatic surface is being drawndown as anticipated in the design.Finally, the fabric must be able to tolerate installationstresses. The installation should be able to·be completed withminimal tearing or puncturing of the fabric. Any damage whichmay occur should be repaired. The'''ReconunendedMinimumProperties for Geotextiles Used in Non-Critical/ Non-SevereDrainage, Filtration, and Erosion Control Applications," asrecommended by Task Force 25, were presented (Fig. 12). Inaddition, manufacturers' recommendations should not be exceeded.Since MSHA often sees ·fabrics proposed in'critical applic.ations,these criteria should also be supplemented by field trials toverify the construction procedure, and.construction monitoring bythe designer, or his representative who is knowledgeable in thearea of geosynthetic construction and filter' criteria. Inaddition, construction specifications should be SUfficientlyrigorous to ensure a good installation. Limiting drop-height forrockfi11 and providing protective sand or gravel layers should beconsidered.In summary, it was emphasized that it is important· to recognizethat geotextiles are engineering materials and need to be treatedas such. The geotextile is part of a soil-fabric system and thecomponents of this system must be compatible. The same fabricmay not be suitable in all situations. While this seems obvious,it seems that a few designers have indiscriminately specified thesame fabric regardless of the application, and with little or notesting or analysis. MSHA's move toward site-specific testingseems consistent with current prudent engineering, as such testsare often pe.rformed even for far less critical applica'tions~The figures attached were taken from tran~parencies used in thepresentation and are intended·to illustrate typical guidelinesfrom the literature. They do not necessarily represent the onlyacceptable criteria. Other information sources may differ •.Ultimately, it is the designer's responsibility to make aconservative fabric selection based on thorough consideration ofeach of the design requirements. .

23

Questions Regarding Presentation1. Often material is not available suffiCliently aheac:fof time toconduct lonq-term flow tests prior to the initial desiqn. Howshould cloqqinq potential be evaluated in such cases?We would recommend that based on Index tests like the gradientratio test,various empirical relationships, and experience withsimilarmaterials~ a fabric could be selected. When material isavailable, a long-term test could be initiated. If the testpasses, ~he original,design is verified. If the test fails,construction would likely not be too far along for compensatingdesign modifications to be made to the facility. The risk ofhaving to perform modifications should be weighed against theuncertainty in the design,with consideration of how soon thedrain needs to be constructed following the refuse materialbecoming available.2. If a qeotextile is used as a separator, rather than as afilter itself, 'is it still necessary to test it for clogqinqresistance?If a fabric is in a location where if it were to become clogged,it would not affect the stability of the embankment,we would notrequire it to be tested for clogging resistance. If, on theother hand, clogging of the fabric (regardless of its perceivedfunction) could reduce the factor of safety against slopeinstability below acceptable levels, it should be tested.

Additional DiscussionThe participants were. asked for input regarding their experi~nceswith the use of geotextile fabrics in filtration applications.The response was generally favorable. The only negativeexperie~ce, which was relayed, involved one case where the fabricwas· contaminated by ~unoff of fines during construction. Careshould be taken to avoid this situation.A comment was made that the gradient ratio test is not real~yrepresentative of~the field conditions. We agree and r~commendthe use of a Long-Term FloW Test. The participant expressed thatthis test, also, may not be representative of field co~ditions.We expressed that we are certainly aware of its limitations,however, similar limitations exist ~ith all of the small-scalelaboratory tests which are tradition~lly performed (i.e. triaxialcompression, permeability, consolidation, etc.) This shouldalways be' accounted for in'interpretation of the results and byincluding reasonable factors of safety, consistent with theuncertainty. In closing, we indicated that, in other areas ofthe civil Engineering community, it appears that such testing isperformed for far less critical installations.

24

Proc

edur

eIDs

truc

t~To

perf

orm a

ccep

tabl

yasafI

lter

111adr

aina

geap

plIc

atio

n,age

otex

tilemu

stfu

nctI

onas

foll

ows:

1.re

tain

thepr

otec

tedsol

Ito

prev

entpip

ing;

2.ha

vesu

ffic

ientpe

rmea

bili

tytopr

even

tthe

b~ll

d-upof

wate

rpr

essu

re;

J.no

tbe

come

clog

ged;an

dN Ul

•4.

have

suff

IcIe

ntst

reng

thto

surv

Lvet

heco

nstr

ucti

onpr

oced

ures

.

:riq

ua:e1

4~ Designation: D 4751- 87

Standard Test Method forDetermining Apparent Opening Size of a Geotextile 1

100 ---"'I"'!""'--"-'-~-,----r----,95

80

20 '-

" ,'0 L.- __ ,.l.'_"';';"..;,J.,1 ~_~~l ~_~2 0.6 0.2 0.06 0.02 0.006 mm

Coarse MId. Fine Coarse Med.Sand Silt

Ll9tndA.7.4oz./yd? (250 gm/mz) ruin'bondedS. 8.8 oZJYd~ (300 gm/mz)netdle,pu',chedC. 4.1 oz./yd~ (140 gmlm2) melt-bondedO. 11.2ozJyd? (380 gm/mz) hessian wovenE. '5.5 ozJyd: (185 gm/m2) woven terv1ene

26

:rigure 2

Fabric

AlI9A9lte

(It

(et

(bt

Particles clogging

Averlgefabricthickness

Id)

Fllur. 2.2.3 Various hypothetical mechanisms invulvcd in long-term soil-to-fabric Row compati-bility (DRllI'McOown 1431). (a. Fumlllticm or an upstream soil Rller. (b) Upstrellm particles blockinglentexlile openinp. (c) Upstream particles llIChlnl over pme:ttilc openinp. (d) Soil paniclesclo'led within peJlclttilc structure.

Jrigure 3

27

Blinding

Fabric:

Clogging by Particle Deposition

Figure 3-1. Methods ,of. C10gg·ing and Blinding(Bell and Hicks, 1980) .'

28

SUMMARYOFGEOTEXTILEDESIGNANDSELECTIONCRITERIA

FOR

DRAINAGE.FILTRATION.A

NDEROSIONCONTROLAPPLICATIONS

I.SO

ILRETENTION(PIPINGRESISTANCECRITERIA)I

Soils

'-50%

Possing2

U.S.N

o.200sieve

Steady

SioleFlow

.AOS--.° 9

5k

B0 85

C'2

or~8

lJ 2'Cu "

"

B=I

B=0.5Cu

~~SO~Passing

U.S.N

o.200sieve

4L.CLa

B=~

uu

Woven:0'5

fDOS

Nonwoven:O'SL.

1.0D

OS

AOSNo.(fobrlc)~No.SO

sieve

Dvnom

lc,Pulsoling,

andCyclic

Flow

095f0

IS(If

soilcanmavebeneathfabric)

OR

050

'0.

5DO

S

050

'-0.

50 05

..

I.When''''~protec'ed

soUcontains

particles

from

Iinch

size

tothosepassingtheU.S.No.200sieve.

useonly.the

gradationof

soilpassing'heU.S.No."sieveinselectingthef(Jbric.

2.Select

fobricon

thebasisof

lorgestopeningvaluereouired(smollest

AOS)

:riq

ure5

II.

PERMEABILITYCRITERIA

I

A.Critical/Severe

Applications

k(fabric)~10k(soil)

-B.less

Critical/less

Severe

and(wUhClean

Medium

10CoorseSandsand

Grovels)

k(fobrici~

k(soU

)

I.Permeability

shouldbe

basedon

theactual

fabric

open

oreo

ovailableforflow.F'orexam

ple,If509&or

fabrIc

areo

Is10be

coveredby

(Ja'concrete

blacks,theeffective

flowarea

isreducedby

SO~.

~

Piga

r.,

III.

CLO

GGINGCRITER

IA

A.Critical/Severe

Applications

I

Select

fabricsmeeting

I,II.IIlB.and

perform

soillfo~ric

filtration

testsbefore

specification,

prequalifying

the

fabric,or

after

selection

before

bid

closing.

Alternotiv.,~

use

approved

listspecification

forfiltration

cppllcctlens,

Suggesledperformance

test

method:

GradientRatio

L.3

B.LessCri1icaI/Non-Sev~reApplications

Wheneverpossible,fabric

with

maximum

openingsize

possible(lowestAC3No.)fromretentioncriteria

should

bespecified.

Effective

OpenAreaQualifiers2 :

Woven

fabrics:

PercentOpenArea:

~4%

.No

nwov

enfab

rics:

f'oros

it},3

~30

'

Additional

Qualifier(Optional>:O

'S~3DI5

Additional

Qualifier(Optional):

015~3D

I5

I.Filtration

testsereperformance

lestsandcannot

beperformed

bythemanufacturer

asthey

depend

onspecl.ficsonanddesign

conditions.

Teslsto

beperformed

byspecifyingagency

orhisrepresentative.

·Note:

e~periencerequiredtoobtain

reproducible

results

ingradient

rotlo

test.

I'iq

ure7

~~I~D•• ignatlon: D 5101- 90

Standard Test Method for .Measuring the Soil-GeotextJleSystem Clogging Potential bythe Gradient Ratio'

It D5101

• n·IWCI 'Law ItATI -IIADINOS 'Ra" "THIS GUT'La"pair'Law- - - --- - -- - - ""..- - - INna"- - - IWIClNTU OHD- - - - --- - - 'aRr ",- - - 1an.!T- - -- - - / • VATElt aUTFl.aV- - -- - - 'DllT- - -- - -- - - III- - -- - - ~- - - 3- - .- -lit- - - ~- - - SOil.- - - ~- - - nov- - - 5- - - / -- - -- - -- - - roaOTiXTtL! aUT'l.aV ~- - / CHO

=- - -- - 1- J I- - 1- " b- - . -6 I / :c- 1= 1- " - ,- 1_

=j- i-I 1=1 I- ,- I MUT•- .- 11~1 I

- \=,- I- _.- ~-'t 2 3 4 5 5 1i. .., :::J' 1

/'tANOI'1ETE1lS P5~"E.~MEiE~ CQNSTA~T HEAD C~vIC£SJIIQ.3 Qeot.xtlI. Penn ••meter •••• tUp" Dlq •••m

Oradient racio • Air, .•.IslS,.•...l:Jl.I1J,u/S1.u

4.1t. 16 • tho h=I ~anse in inches from tho bonom of tho fabric co 1.0 in. (25. mm) of soil abovo the fabric. .Sf. 1.0 • tho fabric thicJcDessplus 1.0 in. (~ nun) of soil.4.hv • the bead change in inches between 2 in•. (SO 111m)of soil above the

fabric:.Su • 2.0 in. (50 mm) of soil.

J'iqure 8

32

10,..-- Nonwoven, melt· bonded..-_- Woven•• lit film

8

7

,9 6-••..-,~'i..~

3

2

-0 5

F1aur-e 2.11Wood (27».

Ottawa sand andViclcsburg sih loIS.soil mix~ure

Woven. monofilament >

Nonwoven needl •• punched

--------/' Woven. monofilamen~. fabric "a"

"-Woven, monofilament, >fabric "b"I I

10 4015 20 2SSoil silt content (%J

30 3S

OnllJient ratio test data u$ed te· illulItr.lte rabric clu~ging PQlenti:ll (ufter Haliburton :md

33

J1:Lgure ,

Long-Term Flow (Clogging) Test

Mica Silt Soli andVarious Fabrics

o Woven monofilamenta Nonwoven melt·bondedA Knit monofilament• Nonwoven l\iedie-punched

-.:.= 75!!~••SO~.i:i:

o

a ~_~..J..'....l.' ..l,~.:..;,:.:,:.:,~I__ :-..:,....:,~, ":,'":,,,,:,,,:.,,~,_-.I•......:,'....:.'_,~,,:.'.l.l~'~_..:...-.:.'-.:.'"':'''''o.:'~'~'0.1 10 100 1000

Time (hr.)

(b)

Figure 2.16 Lonc·term ftow tests Oft soil·fabrie: systems and typical response: curves(after Koerner and Ko [25]).

Jliqure 10

34

ii' , i iiJ

o 1~..J-.l...L.I..U.lWIO.L-""'--'L..,L.L.U.lWIClO:'--'--,L"",L...u.L.&..LL..;~I....L"""""~IO.OOO

TIMe , •••••••)

"f'"I'l-

S' t~••I-Z~D•300'

0.7 .

, 'i1'''( ~." OI"i

,. ~ -• !/. t

r -1~;.•_____ --./-1

, i , Iii "iii, i"I

F1S. 6. - Lons-Term Flow Curve of SUtyClay Soil and Nonwoven NeedledFabric (Upper) and CorrespondingValues of Gradient Ratio (Lower).

NOTE 4- This test can be run at hydraulic gradients other than thosespecified in this procedure .. for example .. i = 3 for 24 h. In all cases, thesystem hydraulic gradient should be increased gradually and in .incre-ments no greater than i = 2.5 and maintain those incremented levels fora minimum of 30 min. The test mav alsQbe run at longer interval$_Ulan24 h. until some recognizable equilibrium or stabilization of the systemhas occurred.

-:riqure 11

35

Task Force No. 25COMBINED TABLES 3-4 AND 3-5

REVISED 5-30-85

RECOMMENDED MINIMUM PROPERTIES FOR GEOTEXTILESUSED IN NONCRITICAL!'!) NONSEVERE(2)DRAINAGE,FILTRATION, AND EROSION CONTROL APPLICATIONS

I. MINIMUM SURVIVABILITY PROPERTIESA. Fibers used in the manufacture ofgeotexti1es shall consist of longchain synthetic polymers, composed of at least 85% by weight ofpolyolaphins, polyesters, or po1yamides.B. Geotextiles with low resistance to ultraviolet degradation (more than

30% strength loss at 500 hours exposure ASTM 0-4355) should not be.exposed to sunlight for more than 7 days.GeotextHes with higher resistance to ultraviolet degradation should notbe exposed for more than 30.days. .NOTE: Geotext11es can be manufactured or finished to resist degradationdue to prolonged exposure to ultra-violet radiat10nt i.e., fabricsresistant to exposure for multi-year periods (from 5 to 25 years) arenot uncommon.

C. Physical Property Requirements:Drainage(3) Erosion Contro1(3)

Test Method * Class A(4) Class 8(5) Class A(B) Class 8(7)

Grab Strength (TF #25 method 1)(Min. in either principle·direction) 180 lbs. 80 lbs. 200 lbs. 90 lbs.Elongation (TF #25 method 1)(Min. in either principle Not Notdirection) Specified Specified 15% 15%

Puncture Strength (IF #25method 4) 80 lbs. 25 lbs. 80 lbs. 40 lbs.Bu~st Strength (TF #25method 3) 290 psi 130 psi 320 psi 145 psi -

Trapezoid Tear (TF #25 method 2)(Min. in either principle -direction) . SO lbs. 25 lbs. 50 1bs. 30 lbs.

*Test method is in accordance with procedures inAppendix B of FHWA Geotextile Engineering Manual.FIGURE 12 (1 of 2)

3636

lCritica1 applications involve the risk of loss of life, potential forsignificant structural damage, or where repair costs would greatly exceedinstallation costs •.

2Severe applications include draining gap graded or pipab1e soil, high.hydraulic gradients, or reversing or cyclic flow conditions~

3A11 numerical values represent minimum average roll values, i.e., valuesmeasured for a sample (average of all specimen results) should meet or exceed-'specified values within a 2 sigma confidence level. Th~se values areconsiderably lower than those commonly presented in manufacturer'sliterature.

4C1ass A Filtration and Drainage applications for fabrics are whereinstallation stresses are more severe than Class B applications, i.e., verysharp angular aggregate is used, a heavy degree of compaction is specified, ordepth of trench is greater than 10 feet.

5C1ass B Filtration and Drainage applications are those where fabric is used withsmooth graded surfaces having no sharp angular projections, rio sharp angular'aggregate is used; compaction requirements are light, and trenches are lessthan 10 feet in depth.

6Class A Erosion Control applications are those where fabrics are used underconditions where installation stresses are more severe than Class B; i.e.,stone placement height should be less than 3 feet and stone weights should notexceed 250 pounds. Field trials are required where stone placement heightexceeds 3 feet or where stone ·weight exceeds 250 pounds.

7Class B Erosion Control applications are those where fabric is used instructures or under conditions where the fabric ~s protected by a sandcushion or by "zero drop height" placement of stone.II. MINIMUM HYDRAULIC PROPERTIES

A. piping Resistancg (soil retention) (1)

1. Soil with 50% or less particles by weight passing. U.S. No~ 200Sieve(21:AOS(3) less than 0.6mm (greater than #30 U.S. Std. Sieve)Soil with more than 50% par~ic1es by weightpass~ng U.S. No. 200 Sieve :AOS less than 0.3mm (greater than #50 U.S. Std. Sieve)

2.

B. pgrmgabilitvk of fabric(4) greater than k of soil

(l)Design values as determined by an engineering analysis which assures compati-bility between soil hydraulic conditions and geotexti1e are recommended(especially for critical/severe applications). Problem soils where the aboveguidelines may not apply are silts and uniform sands with 85 percent passing the#100 5ieve.(2)When protected soil contains particle sizes greater than #4 U.S. Std~ Sieve sizeuse only the gradation of soil passing the #4 U.S. Std. Sieve in selecting thefabric. '.(3)AOSdetermined for geotexti1es according to TF #25 Method 6.(4)Permeability determined for geotextiles according to TF #25 Method 5.

FIGURE 12 (2 of 2)3637

CONTROLLING SEBPAGB ALONG THB CONDUITSAbdul Hamid

summary of PresentationConduits are routinely installed through embankments by miningcompanies to control the design storm. This·creates anopportunity for seepage along the conduit. Uncontrolled seepage'along the ~onduit could cause piping of backfill material andpossible subsequent failure of the embankment. Therefore, it isvery i~portant that seepage along the pipe be discharged in'acontrolled manner to preclude piping and hazard to theembankment.Historically, anti-seep collars have been used around theconduits to control seepage. However, anti-seep collars arelabor intensive and require skillful labor and hand compactionalong the collars.' In ,spite of best efforts and quality control,these seepage control devices have been known to functionimproperly. For the last 15 to 20 years" the use 'of anti-seepcollars has been abandoned by many professionals in favor'offilters and drains axound the conduits at the downstream portionof the embankment. Drains and filters are able to controlseepage along the conduits better than anti-seep ~ollars.It must be emphasized that drains and filters for 'conduit seepagecontrol should meet the same criteria as the other drains andfilters in the embankment.

Questions Regarding Presentation1. What is a filter-drainage diaphragm?Filters and drainage .diaphragms are used around the pipes whichextend through the da~s to control and safely discharge seepagewater 'around the pipes. They are used in lieu,of anti-seepagecollars. The drainage material is designed to filter criteria,so that piping, and internal erosion of backfill material doesnot take place. The drainage material must be SUfficientlypermeable to allow seepage along the pipe to collect anddischarge safely in a controlled mariner. The outlet drain mayfollow along the pipe, or it may be tied to the dam's underdrainsystem. For more information, consult the references cited inthe Procedure Instruction Letter. specific requirements for sizeand location can be found in the references numbered 10 and 13.2. In designing a filter and drain system to be used instead ofanti-seepaqe collars, What filter criteria is preferred?When filters are used instead of anti-seepage collars aroundpipes to collect and safely discharge the seepage water, pipingand internal erosion are of primary concern. As explained in theProcedure Instruction Letter on "Graded Filters," there are two

38

design methods commonly used to ensure that filter criteria ismet and piping and erosion of surrounding material will not takeplace. Ei~hQr ~hQ method developed by Terzaqhi, or the methoddeveloped by the Soil Conservation Service can be used.3. What suggestions do you have for handling' leakage at a pipejoint?The author no longer allows the use of mechanical joints, sinceleakage in the joints cannot be prevented. If there are existingpipes with leaking joints, they present serious problems,particularly if the pipe is used in flood routing.. ·If the pipe.flows under pressure, the joints could separate completely andjeopardize the safety of the structure. The solution would be togrout and abandon the "pipe and install anew one with weldedjoints.4. Does a filter-drainage diaphraqm need an outlet drain?Yes, since the purpose the filter-drainage diaphragm is tocollect seepage that may occur around a pipe, and discharge it ina controlled manner.5. Are filter-drainage diaphragms used in earthen dams, or justcoarse refuse dams?They can be used in both earth dams and coarse "refuse dams.6. What is a good reference .forfilter-drain~qe diaph~aqms?Several references are listed in the Procedure InstructionLetter.7. What material is most useful for seepage collars?It is preferable that seepage collars be of the same material asthe pipe. otherwise, differences in thermal coefficients couldhave adverse effects.8. Which method is cheaper?We do not have information on the relative costs. It wouldappear that filter-diaphragm would 'be less costly because anti-seepage collars are very labor intensive and may not functionproperly. They may require remedial action.9. Comment: "Just because the Government"says that gradedfilters are the way to go does not mean it is state-of-the-art."As indicated in the procedure instruction letter, when properlydesigned, MSHA will accept either the filter and drain,approach orthe anti-seepage collar approach. .

39

DESIGN 01'PIPES FOR EXTERNAL LOADINGDonald Kirkwood

Brie~ summary o~ PresentationThe discussion began with an overview of the recommendationscontained in the Procedure Instruction Letter. The various·potential structural failure modes were discussed, as well asunder what conditions each failure mode might be expected togovern. The manufacturer's recommendations for fill height werethen discussed, inclUding the advice that these maximum fillheight recommendations should be used as 'only a r.ough guide. Ifthe installation is critical, detailed analysis· and calculationsshould be done. The discussion then turned to the lack ofconsensus among the experts with respect to the best method forstructural· design of flexible pipes. It was recommended· thatseveral methods be applied and the results. compared. The morecritical the installation is, the more sophisticated the analysisshould be. The finite element models were then discussed, and inparticular, the CANDE-89 program. Finally, monitoring as a vitalcomponent of design, approval, and model verification wasdiscussed.

Questions Regarding presentation1. what has ~een MSHA's experience'with failures of thedifferent types of pipes?We have seen numerous failures in corrugated pipe installations.Most of these are believed to have been caused by either improperinstallation' or by inadequate pipe couplings. There have alsobeen cases of welded steel pipe failures, but these have beenrelated to improper welds. We have yet to see a fallure of aplastic pipe due to excessive loading.2. Why isn't hydrostatic pressure used to test plastic pipe fortheir design deflections?Pipe deflections are a function of the pipe properties, the soilbackfill properties, and the pipe-soil interaction. Actual testson pipes that have not included the backfill, and consequently.the pipe-soil interaction, have not been able to predict fielddeflections. .3. What is the maximum amount of deflection that MSRA willaccept?Normally the manufacturer's recommendation for maximum deflectionbased on pipe SDR are considered the maximum allowable .deflection. However, the actual factors 'of safety built into themanufacturer's allowable deflection limit is not certain. ThePIL specifies a minimum factor of safety of 2.0 for wall crushing

40

and buckling while none is specified for deflection. In somecases, a factor of safety exceeding 2.0 results in· deflectionsthat are in excess of the manufacturer's recommendations. Itshould be noted that the allowable deflection is a performancelimit based on the pipe properties only, therefore, there shouldbe considerably less doubt concerning the allowable deflectionperformance limit than there obviously is about the actualdeflection under the maximum fill height. .4. The restrictions on maximum fill heights overflexi~le pipeshave led to designs with the periodic a~aDdonmeD~ of ~he pipe andthe installation of a Dew pipe at a higher eleva~ioD. ISD'~ this~uilding additional avenues for failure?The risk for failure from seepage through or around an abandonedflexible pipe is very small compared to the. risk of going beyondavailable technology relative to the performance of these pipesunder high fills. The technology for sealing these pipes is wellestablished, and their abandonment can be done with minimumpotential for future deflection given adequate care in theirabandonment.5. Do we have data, on ~he effect of high tempera~ures ODthe high density, polyethylene pipes?Yes, high temperatures can accelerate the softening process ofthe hdpe pipe material. We believe that has not been a problemat our facilities for several reasons. Coal waste is highlycompacted, eliminating the availability of oxygen which can causespontaneous combustion. In addition, a considerable·portion ofthe pipe is below the phreatic surface. This coupled with airand water moving through the interior of the pipes creates atemperature regulating effect.Normally hdpe pipe experiences some softening even withoutelevated temperatures, due to the effects .of time. Therefore,it's the soil envelope which provides most of the strength to thepipe-soil system.'6. Is ~he trench width critical?Yes, it can be. Trenching can significantly impact the loadingdistribution around the flexible pipe, because the trench wallsact as rigid abutments. If the trench gets too wide, the load isnot distributed to the rigid abutments. The load is carried as an"embankment" load, or a load on a pipe not installed in a trench.7. The Modified Iowa formula uses a modulus of soil reaction.How large of a modulus of soil reac~ion will MSHA allow?The modulus of soil reaction is a function of the pipe-soilinteraction. It is not strictly a soil parameter and'cannot,therefore, be taken directly from a soils test. The only knownpublished values of this·modulus are from the work ,of A. K.

41

Howard in 1977, where he constructed an apparatus to.test thepipe-soil system together. This work was done only for fillheights up to about 50 feet. Other work since' 1977, particularlywork back-figuring soil reaction moduli from large diameterinstallations', under relatively high fillS, suggests that thismodulus should increase significantly with increasing fillheight. This work is, however, inconclusive. Therefore, thesoil reaction moduli, to be used in the Modified Iowa formUla,should be taken from the acknowledged, conservative Howardvalues.8. Why isn't plastic pipe less sensitive to installation errors?Plastic pipe carries load by deflecting and mobilizing the.supporting strength of the pipe backfill material. This ideallyhappens in a uniform, symmetric fashion. Actual installationslikely rarely deflect either uniformly or SYmmetrically.However, if the deflection becomes too asymmetric, the pipe maybe subject to failure from localized buckling. The ability ofthe pipe to shed a large portion of the load to the surroundingbackfill is paramount to the adequate performance of the flexiblepipe under high fills. Therefore', if excessive deflection isinduced in the pipe during installation, or if there are hard orsoft areas in the backfill around the pipe, failure may resultlong before the design, based on uniform symmetric deflection,would predict it. steel pipe, on the other hand, has'considerably larger modulus of elasticity than does plastic pipe.The steel pipe, therefore, relies considerably less on thebackfill for support, although the backfill support is stillsignificant. Therefore, the steel pipe installation can actuallybe somewhat more forgiving relative to installation errors.9. How much fill height has been approved over flexible pipes,and has there been a requirement for monitoring of theseinstallations?MSHA.has yet to catalog specifics such as fill height over decantpipes for all of the approved plans. However, the maximum fillheight approved over a flexible pipe is believed to be between150 and 200 feet. When the fill height reaches the stage wheresome of the analyses, such as the Modified Iowa formula, suggestthat the factor of safety is less than 2.0, yet other. analyses,such as a finite element analysis, suggest .that the factor ofsafety is greater than 2.0, the installation can be approved.MSHA usually requires monitoring to verify. its performance. Thishas been the case with all known approved fill heights of between150 and 200 feet.10. Does MSHA have a position' OD maximum acceptable fill height.No, other than the position that until performance data isestablished for high cover situations, conservative designmethods need to be used and factors of safety of at least 2.0should be maintained.,

42

11. Will pipe deflection monitorinq data become available toeveryone from the MSIA approved installations?Although.the deflection monitorinq data will be avaiiableto andused by MSHA for its reviews, the details of these installationswill only be made available to others if the company permits itsrelease or if the information can,be obtained through the"Freedom of Information" act. However, MSHA will likely be ableto share the data in a general sense, relative to performancetrends if the installations are not identified.12. Are only plastic pipes beinq monitored?To date, very little actual monitoring has taken place. The onlyproposed deflection monitoring in MSRA approved installations isfor plastic pipes. That is.not to say that monitoring might' notbe requested or proposed for a steel pipe installation. Likeplastic pipes, steel pipes usually rely on backfill support for alarge part of their strength. If these installations approachthe limit of known, conservative design deflection estimates,monitoring will likely be requested.13. Has MSHA approved plans wi~h lower factors of safe~y wi~hprovisions for monitoring of the pipe?There is not one method of analysis that is, at this time; usedat the exclusion of all others. Each of the various methods offlexible pipe deflection analysis gives unique results. Somemethods give drastically different results than others.Therefore, what the factor of safety is depends on which methodwas used for predicting that deflection. All factors of theinstallation must be'considered in the review, including: theresults of various deflection prediction methods, theconsequences of failure of the installation, whether the pipe isrelied on for drawing down the design flood, the length of timethe pipe will be relied on for flood routing, the degree ofconservativeness in the design, the care in installation, andwhether the installation will be supervised by a qualifiedengineer. Given all of these factors, there are i~stallationsfor which some of the deflection prediction methods might notresult in an adequate factor of safety, yet the installation isapproved. In order for these installations to be approved, anacceptable deflection prediction method will. have to show anadequate factor of safety, and 'monitoring will likely berequired.

43

14. Can a flexible pipe be installed, and continually monitoreduntil it exceeds the allowable deflection, and then abandoned?The monitoring data, along with any research which may be beingdone, will determine how much fi11 these pipes can be approved .under. For the time being, we can approach the design limits andmonitor the pipe deflection. If the monitoring program shows t~edeflection to be considerably less than the predicted deflection,then it is possible for the pipe to be approved beyond this fillheight based on the performance data collected from th~monitoring program. This is most likely in non-criticalinstallations and for deflections below the performance limits.It seems highlY. unlikely, at this time, that MSHA would approveany pipe installations in critical areas and allow the deflectionto reach failure. '15. Have there been any pipe installations, proposed to HSBA,where the pipe is to be monitored with strain qaqes?I haven1t heard of any cases where monitoring pipes with straingages has been proposed.16. What are you solvinq for in the CANDE-89 proqram?The CANDE-89 program solves for stresses in the pipe wall anddeflections of the pipe. The results are stated in the form offactors of safety for pipe displacement, pipe wall o~ter fiberstresses, and elastic buckling.

17. I.e the use of the CANDB-89 buried pipe, structural design,finite element program acceptable?Yes, there is no one deflection prediction method, at this time,accepted to the exclusion of all other methods. The CANDE-89program has strong points and weak points, as,do all finiteelement programs. We do believe that when care is taken in theinput parameters, CANDE-89 can 'give an additional, usefulestimate of the maximum pipe deflection and pipe wall stressesunder high fill conditions. MSHA uses the CANDE-89 program asanothe~ indicator of the pipe performance.18. Does the CANDE-89 program take arching in the soil above thepipe into account?Ye~, the CANDE-89 finite element program incorporates loadreduction due to pipe deflection which is soil arching. The loadonce carried by the pipe is being transferred to the surroundingsoil, i.e. arching.

44

19. How does one determine the elastic/plastic soil propertiesfor input into the CANDE-89 program?There are soil types built into the CANDE-89 program. If yoursoil behaves like any of these soil types, the built in .properties can be chosen. If however, your soil is significantlydifferent than the built-in soils, the elastic/plastic soilproperties must be put into the program. These properties arehyperbolic properties and may be obtained from a triaxial test ifthe test is modified to carefully measure nonstandard details.The supporting CANDE-89 documentation should be referred to foran understanding of what is required.. .20. Is the CANDE-89 program more or less conservative than theModified Iowa formula?The output of the CANDE-89 program is highly dependent on theinput. Therefore, the CANDE-89 output can either be more or lessconservative than the Modified Iowa formula depending on whatparameters are input. In general, however, our experience hasbeen that if great care is taken in modeling both analyses thesame, the CANDE-89 program usually results in somewhat lessconservative values.21. Do all pipe installations require sophisticated analy~es?No, when pipe designs are taken very close to their statedperformance limit, more verification of their potentialperformance is requested and consequently, more the sophisticatedanalyses are often necessary. Installations which are shown tohave conservative factors of safety and which have conserVativedesign parameters such as backfill type and installationprocedures, do not require sophisticated analyses. Many times,showing that the manufacturers fill height limits have not beenapproached and that the Modified Iowa formula gives a factor ofsafety greater than 2.0 is all that is required.22. Do all approvals require deflection monitoring to verifydeflection?No, as with the required level of sophistication of the analyses,the installations which are obviously conservative in design andinstallation normally will not be required to be monitored.

45

23. Wha~ research is ~eiD9 done ~o check ~he accuraoy of thevarious models?We know of very little research being done of the performance offlexible pipes under high fills. As far as we know, none of theplastic pipe manufacturers are undertaking their own research inthis area. This is understandable since such a small proportionof their overall sales are for high fill applicatiqns. Theresearch that is being done, that we know of, is being done atvarious universities. Most of this research is not for smalldiameter, lo~ SOR, plastic pipes under high fills. It appe~rsreasonable to assume, at this time, that what performance datawill be forthcoming, will come from coal wast~ relatedinstallations.24. By taking the conservative route ,in every phase o~ thedesign, aren't we creating ultra conservative designs?Perhaps, ,but we won't know how conservative the installations areuntil the performance data is available. For the time being, weare choosing what we believe to be conservative parameters andmethods. We don't have the data to prove that all of thesemethods and parameters are conservative, and under whatconditions they might be non-conservative. We hope to becomemore liberal as the performance data, or research data, shows ushow conservative we've been. Even now, we are being somewhatless conservative than in the past by considering multipleprediction models, and.relying more heavily than ever before onmonitoring. Hopefully, this trend will continue towards moreexact and less conservative designs. For now, we must not forgetthat we are leading the way in using plastic, pipes under highfills and,we are often doing it in high hazard dams.

46

PROBABLE MAXIMUM FLOODDaniel S. Mazzei

summary of PresentationThe presentation of the methodology used in developi~g the design;jnT'l)'W'I>~rametersfor the,.,~,det.erm1nationof the Probable Maximum";F~a6dFbegan with 'a discussdtdrl';ofthe criteria involved. Thefactors that must be considered include: the principal storm,the antecedent storm, the loss rates or initial moistureconditions, the rainfall dis'tribution, the initial reservoirconditions, and windwaves. It was shown that many of theAgencies involved in Dam design vary specific parameters .'according to Agency policy and that it is important to beconsistent.The procedures and references used in selecting each of thehydrologic design parameters was discussed in depth. Particularattention was devoted to a discussion of the impact of the use ofAMC III. The use of the computer program IHMR-52" in obtainingthe temporal rainfall distribution for the principal storm was'also reviewed. .

Questions Regarding Presentation1. At which pool level should the storm routing begin?When a facility has an open channel spillway and a pipe spillway,the routing should begin at the invert of the open channel,unless a substantial justification is developed. That justifica-tion must address in detail what measures are in place to assurethat the primary spillway will not malfunction.2. Why donlt we develop a "cookbook" ,type guide for thehydrology and hydraulic portion of the impoundmentplan?A guide of this type is not practical in that every site isunique and the response to an input can vary considerably as aresult of small differences in design parameters. Additionally,such a guide would infringe on the engineering that must beexercised by a company or their consultant due to designconstraints that MSHA cannot consider, such as costs~ MSHA canspecify a desired end but does not have the right to dictate themeans.

47

3. Would it be practical to use the default rainfalldistributions in the HEC-l computer program?All hydrologic developments that are submitted to MSHA foapproval will be tested against the methodology which isspecified in HMR-51, and HMR-52, for sites covered by thosepUblications. o~~rHMR pUblications have been developed fordifferent regions" of the country and those references should beused. The criteria for the test will 'be the "hydrologically mostcritical" distribution. The HEC distributions may or may notfulfill this requirement, based on the watershed/reservoir'response.4. Can one use the minimum soil infiltration rates in thedevelopment of the Probable Maximum Flood?The use of a linear relationship in an extreme event raisesconcern. It seems to be more consistent to use the curve numberapproach. Evaluation of the HEC output indicates that a morereasonable build-up in losses results. The hydrologically mostcritical test would again apply.5. The use of extreme event, the PMF, seems to be unreasonablein this age of hazard and risk analysis.The Buffalo Creek disaster focused us as an industry on howserious a failure can be. In that light we have theresponsibility to provide for the safety of those living in theshadows of our dams'. Hence, the PMF is our design storm. Yes,the use of incremental hazard and risk analysis might beconsidered for structures for which a high degree of confidenceexists. However, in dealing with mine waste impoundments thereare too many uncertainties in too many critical areas, andconservatism is needed. In addition, current Federal guidelinesrecommend the PMF as the inflow design flood for high hazardsites.

48

COMPACTION SPECIFICATIONSstanley Michalek

Brief Summary of PresentationProper compaction of embankment material is one of the mostimportant elements in the construction of a safe dam. Compactionis performed to increase the density "and shear strength and todecrease the compressibility and permeability of the constructionmaterial. Specifications for compaction place limits on theminimum dry density, the range of placement water content, andthe maximum lift thickness.This Procedure Instruction Letter recommends that the material becompacted to at least 95 percent of the maximum dry density asdefined by the standard Proctor test (ASTM 0698)~ The placementwater content should not exceed the range -2 to +3 percent ofoptimum water content. The loose lift thickness should notexceed 12 inches for coarse coal refuse used in structuralportions of the dam. Refuse used in non-structural portions ofthe dam can generally be compacted to lesser requirements. Wherefine-grained soils are used in the embankment, a loose liftthickness of approximately 8 inches should be specified.

ouestions Regarding Presentation1. Why does HSHA require compaction to 9S percent of maximum drydensity as found by the standard Proctor test?" Can thisrequirement be lowered?Several statements were made that adequate embankment stabilitycan be achieved at lower compaction standards. MSHA'srequirement is that all structural fill (coarse coal refuse) becompacted to at least 95 percent of the maximum dry density foundusing the standard Proctor test (ASTM 0698). MSHA normallypermits less strict requirements in non-structural portions of anembankment. Non-structural fill could include material placed onthe downstream toe of the embankment to act as a buttress.Compaction, or densification, of the fill material achievesseveral goals: reduces settling, increases material str~ngth, andreduces permeability. MSHA believes the compaction practicesused by experienced dam builders are proper and therefore havebeen adopted by MSHA. This is partially true due to the factthat MSHA does not have the capability to conduct laboratoryanalyses to determine acceptable compaction densities. Allreferences cited in the PIL state that compaction to 95 percentof maximum dry density is proper.Several specific statements were made during the discussions.

49

statement:

Response:

statement:

The references deal with strictly soil materials.Coarse coal refuse is not comparable to soil andless compaction will still achieve aqequatestrength.Coarse coal refuse is similar to coarse grainedsoils., Some properties may vary; however,' aftercompaction and weathering the material is similar.One reference cited in the PIL reports results ofcompaction tests specifically on coarse coalrefuse. The accurateness of this reference wasquestioned during the discussions. However, MSHAsees no reason to dispute the work done in thestudy. .

To achieve 95 percent; an operator needs toaverage 97 to 100 percent compaction. This causesdegradation of the particles near the surface of alift.Excessive degradation of the refuse is notdesirable from the standpoint of permeability.Horizontal stratifications can occur if a layer offinely broken refuse is present. Some degradationhas to be expected due to the composition of thecoarse coal refuse. Therefore, compaction to alower density would ~till cause breakdown. MSHAinspection personnel commonly report that fielddensities well over 95 percent are being achieved.More care should be taken by the operator not toover-compact the material and cause unnecessarybreakdown. If excessive rolling is required toachieve the specified density at the bottom of thelift, then thinner lifts should be used.

2. Why does'HSHA require moisture range specifications?

Response:

Several statements were made that specified densities wereachieved with moisture contents well outside the specification.MSHA'~ requirement is that all structural fill (coarse coal,refuse) will be compacted with a placement water content notexceeding the range of -2 to +3 percent of optimum.As in the selection of required minimum density, MSHA has adoptedthe practices and recommendations of experienced dam builders.Each of the references cited in the PIL stress the importance ofplacing material wit~in the proper moisture range. The Bureau ofReclamation's Earth Manual states that "securing the maximumbenefit from compaction requires that the moisture in the soil becontrolled. The specifications will require that the watercontent be uniform throughout the layer to be compacted and thatit be as close as practicable to that content which will resultin the maximum densification of the material." The fa'ctthat

50

coarse coal refuse exhibits a definite Proctor curve demonstratesthat moisture plays a part in the material's compaction.statement: Material does not pass moisture specification when

the weather is wet.Material should not be placed when moisturecontent exceeds the specified range. Material maybe placed and spread in loose lifts during wetweather. However, the material should be allowedto dry to the proper moisture content beforecompaction. Some operators commented thatconstruction of the embankment must continue inall weather or coal production will suffer. Thisis not considered adequate justification to cutcorners when constructing high hazardimpoundments.

3. Why does MSHA require placement of material in lifts notexceeding one foot?

Response:

Comments were made th~~ adequate densities were being achievedwhen lifts exceeding one foot were used. In addition, it waspointed out that MSHA did allow two-foot thick lifts in the past.This practice was abandoned when seepage problems at the sitessurfaced. MSHA's requirement is that all structural flll (coarsecoal refuse) will be compacted in lifts not exceeding 12 inches.When fine-grained soils are involved, lift thickness should notexceed 8 inches.The references state that adequate densities can not be achievedwhen lift thickness exceeds approximately 12 inches. Over-compaction of the surface materials usually occurs in order toobtain specified densities at the bottom of the lift. Over-compaction results in a severe breakdown of material therebyleading to horizontal stratification of the material.No sUbstantive proof has been submitted demonstrating that liftsin excess of 12 inches can be used. Any test would have to showthat proper densities are being achieved at depth without causingexcessive breakdown of surface material.

51

NEW ADMINISTRATIVE PROCEDURES/ALTERN~TE PLAN REVIEW PROCESSJohn J. Mulhern

summary of PresentationThis session, presented by John J. MUlhern, Assistant Directorfor Safety, Technical Support, discussed possible changes to theexisting plan approval process. Assisting at these sessions wereRoger Schmidt from Coal Mine Safety and Health, Kelvin NU, Chiefof the Mine Waste and Geotechnical Engineering Division,Pittsburgh Safety and H~alth Technology Center ~nd John Odell,Chief of the Mine Waste and Construction Divislon, Denver Safetyand Health Technology Center.The introduction to this discussion began with a synopsis of theprogram's background initiated by the Buffalo Creek disasterwhere 125 lives were lost. This disaster started a series ofevents that included review of the design and constructionprocedures used at that time by the mining industry., inspectionand evaluation of all coal mine waste disposal sites, anddeveloping and promulgating stringent regulations for theconstruction of water, sediment or slurry impoundments andimpounding structures. After the initial review for each siteminimal plan review by Technical Support was expected. A copy ofthe presentation is included in Appendix 1.In today's program, approximately 20 years later, plans are stillbeing reviewed. However, there are relatively few plans for newsites being submitted for review and approval. Most of thereviews are for modification to existing sites. This past yearMSHA reported to the Intergovernmental Committee on Dam Safety(ICODS) that four new sites and 66 existing sites requiring majormodifications were approved.MSHA's present program requires plans to be submitted to theDistrict Manager, Where they are administratively reviewed, thenforwarded to Technical Support for technical review. Thesetechnical reviews have requested additional or more completetechnical information on all dams classified as high hazard.Those classified as low hazards have a 9 percent rejection rate.The Federal Emergency Management Agency's (FEMA) role withrespect to dam safety was discussed. MSHA represents theDepartment of Labor as a member of the ICODS which is chaired byFEMA. As required, by Pre$idential Order, MSHA prepares abiennial report on the Agency's compliance with rrFederalGuidelines for Dam Safety". This report, and those of the otherFederal Agency members of ICODS, is forwarded to FEMA for·preparing the report on the Nation's Dam Safety Program. Thisreport is sent to the President and Congress.MSHA's plan backlog has been steadily increasing. Someconsiderations to possible changes in the plan review process

52

were discussed. One option presented a proposal to haveadministrative changes. This proposal suggested a limitedadministrative review by the District prior to transmittal toTechnical support. If Technical support did not recommendapproval, then the plan would be returned to the company. Theywould have 60 days to provide the requested information. If thatreply is also found unacceptable, the company would have 30 daysto reply. Should the third review be returned as "not approved"any resubmittal would be considered a newly submitted plan andreviewed in the order received for all plans. The second optionaddressed alternative procedures for the technical review. Thesewere:

1. If the company needs the plan sooner than ~SHA canconduct the review, an accredited independent reviewerwould be retained at the expense of the company.

2. The plan would be reviewed to conform to establishedcriteria set by MSHA.

3. Technical Support's role would be to resolvedifferences between the designer and reviewer.

4. The accredited independent reviewer would makerecommendations to Technical Support. TechnicalSupport would recommend approval to the District.Some limits on this option would include:1. New dams (sites) would not be included in this

program.2. A size limitation of the dam (i.e. not exceeding

twice the size of the existing site).3. The designer must inspect the dam after any major

rain event.4. Initially, limit the program to modifications of

low hazard sites.Included in the appendix are copies of the presentation thatdiscussed obvious concerns. Also included in the appendix areother technical review options that were part of thepresentation.

Questions Regarding Presentation .1. If industry will have a 60 day and 30 day limitation on

resubmittal of plans, will MSRA have a limitation on theturnaround time for the review of plans?

53

Procedures are being reviewed in an attempt to become moreresponsive. One of the goals of this conference is to seekways to improve our turnaround time. We are her~ solicitin9'ways tha~ y6u can suggest will improve our review time.

2. comment: HSHA needs to be more responsive to the industryand streamline the review process.~gain, we agree that current procedures need improving. Themajor objective of this ses~ion is to seek your ideas andassistance in improving the design review process.

\<

3. If the operator has a conceptual plan, can they come in fora discussion?Yes, they can. However, because of current backlogs wewould want to minimize "philosophical design discussions"and limit these discussions to specific site/designconsiderations.

4. Several questions were raised about tripartite contracts inwhich the mine operator would pay a consultant to review thedesign under HSRA supervision.Many variations of this idea were discussed. No firmprocedu~e was recommended by the group.

S.Why doesn't HSRA hire independent reviewers?A consultant was hired to provide independent. reviews ofplans in the late 1970s. These reviews were unsuccessful.and to the best of everyone's memory, were not completedbefore the funds were spent.

6. A comment repeated several times was "hire more reviewers".This would solve the review time problem, however, currentbUdget constraints preclude any major personnel increases.

7. Why doesn't MSRA allow the company to start constructionprior to approval?The regt;llationsrequire the District Manager to approve theplan "prior to the beginning of any work ••• "

8. Why not work closer with the states?.Most states wait until MSHA reviews and approves the plan.We welcome an opportunity to work with states.· To this end,we will .investigate possible ways ~o work together.

9. Can plans be approved by stages?Yes, we are currently doing this.

10. Why no~ have an in4epen4en~ reviewer have ~o~al approvalresponsiJ)ili~y?At least for the initial stages of any independent reviewprogram, Coal Mine Safety and Health prefers TechnicalSupport be re~ponsible for the plan approval process.

11. Why no~ use o~her governmen~ money, such as ~hat co1lec~edJ)y the Office of Surface Mining (OSM) and ~he EnvironmentalProtection Agency (EPA), and money collected from industryto hire more people?Government monies are obligated by each Agency's budget,which is a law, referred to as General ~ppropriations.Its language on how monies are spent is specific. Itrequires enabling legislation called "Authorizations".Any industry money collected without specific enablinglegislation (Authorization) would go into ~he general fundand could not be used by MSHA.

12. what are some of the major causes for the plan to berejected?Some causes of rejection are deficiencies in the seismic,pipe cover, phreatic line, spillway erosion protection, andcompaction analyses.

14. Why 4oesn'~ MSRA establish a priority list for plan reviewsthat would be based on a high fee, guaranteeing the planwould J)ereviewed within a month?Again, the fees could not be used by MSHA, and withoutadditional reviewers, it would not help shorten the reviewprocess.

15. Why doesn't MSHA charge a review fee?As stated above, this fee would go to the General Treasuryand could not be used to add personnel.

16. Comment: If plans are in the litobe reviewed" stack fO'r1 1/2 or more years, the guidelines change and the pla~s arerejectea. .Plans are seldom that long in the preliminary review cycle.We are attempting to develop solutions to the long reviewtimes that would resolve this concern.

55

IMPOUNDING STRUCTURES SAFBTY DBSIGN PROCBDURES:PRESSURE TBSTING SPILLWAY CONDUITS

Harold Owens

summary of Presen~a~ionIn general, a buried pipe must be capable of toleratinq both themaximum internal and,external pressures which it may experienceduring its service life in order to eliminate the potential forinfiltration and exfiltration. It was pointed out thatinfiltration may cause loss of pipe backfill material which maylead to structural collapse of the pipe or it may cause pipingwithin the embankment. Exfiltration may cause saturation of thesurroundinq embankment and associated slope stability problems.Hence, MSHA recommends pressure testing of conduit spillways inaccordance with the provisions of Procedure Instruction Letter190-11-4, pertinent manufacturers' recommendationsi and currentprudentenqineering practice. Testing should be .at, or somefactor of safety above, the maximum anticipated hydrostatic headunder desiqn storm conditions •. Any leaks need to be repaired.The recommended allowances for "apparent leakage" based oncriteria from other agencies were discussed. The limitations ofsuch criteria were pointed out. For instance, it was indicatedthat it would be improper to apply an allowable "apparentleakage", intended to account for surface absorption in concretepipe, to'a fused polyethylene, or welded steel pipe.The relationship between temperature and pressure, and itsimplications to pressure testing, was discussed. Changes intemperature during the test may result in dangerous or damaging.pressures, or may need to be accounted. for in interpreting thetest results. An experience in District 4 involving a very largepressure increase in a pipe during pressure testing was relayed.Pressure relief valves are useful in protecting the pump and pipefrom high pressures. Temperatures shoUld be monitored in casecalculation becomes necessary.Pressure-testing procedures, which have been used in the past,were discussed. Photographic slides were used to illustrate someof these typical procedures and possible problems.When pressure testing is necessitated by lengthening an existingpipe, there are possible complications arising from air which isentrapped in risers which have been abandoned at lowerelevations. Pressure test results become difficult to interpret,presumably due to compression and dissolution of the entrappedair. This should be considered when determining the cappingprocedure for inlets.

56

Questions Regarding Presentation1. Would MSRA require pressure testing of that portion of a pipewhich is within the impoundment area, or only that part which isunder the structural embankment?We consider pressure testing necessary if the pipe is to be in, .or in close proximity to, the foundation area of future upstreamconstruction. If it is in the impoundment, away from the .structural embankment (and no future upstream construction willoccur), in general, we do not consider pressure testing necessaryfrom a dam-safety standpoint. However, pressure testing wouldseem prudent, in order to minimize the potential for a blackwater discharge.

Additional DiscussionSeveral of the participants expressed similar experiences withpressure test results being difficult to interpret due totemperature and pressure effects.

57

FREQUENCY OF HQISTURE DENSITY TESTING TO VERIFY COMPLIANCEWITH COMPACTION SPECIFICATIONS

Terenoe M. Taylor

Summary of PresentationIn this presentation the following issues were' disoussed:the reasons for moisture density testing, reoommended minimumtesting frequenoies, exoeptions to the reoommended minimumfrequenoies, looations for testing, 'and the keeping of testreoords.Also disoussed as an additional topio was the use of rock-correotion formulas for oorreoting field readings for oversizedpartioles. The reasons for using correctiQn faotors along withthree empirical approaches to,correoting for oversize partioleswere presented. The three formulas were the ASTM equation basedon the research of Shockley, the FAA equation, and the WestVirginia Department of Transportation equation. Comparisons ofthe equations to the results of'actual experimental 18" fullscale specimens were made to show which equation is the bestpredictor based on the percentage of gravel content. Alsomentioned during the sessions, was the problem of companiesreporting measured densities that appear to be above the 95%maximum dry density limit, when in faot the results are reallybelow the suggested 95% minimum, onoe the proper rock-correctionfactor is applied.

ouestions Regarding presentation1. Why did HSHA select a testing frequency of one test every2000 CUbic yards?MSHA has adopted a frequenoy consistent with the minimum testingfrequenoies established by the Department of the Navy and theBureau of Reclamation for mass earthwork. A listing of thesereferences can be found on the final page of the PIL.2. Can the restriction of one test every 2000 cubic yards bealtered? How?As stated in the PIL, in cases where a record of consistent testresults is established, or in cases involving low-hazard dams,less frequent testing may be considered if justification isprovided. In the,case of a new impoundment, the operator wouldbe responsible for initially following the 1 test /,2000cydsrequirement. After a history of consistently meeting themoisture-density criteria is established the company could thenpropose a variance to the above interval. Proof of test resultconsistency should be established using statistical controls.In regard to the testing interval, at this time MSHA does not

58

have an exact interval number. Although the variation wil~ have.to be reviewed on a site by site basis, intervals of say 1 test J3000 cyds with at least 1 test per lif~ ~ be consideredacceptable by the reviewer.Achieving consistency in meeting the moisture-densityspecification will be related to·the relative homogeneity of thematerial. It can often be expected that consistency will be mucheasier to achieve with coarse refuse than it would be with .material taken from a borrow pit.3. Can the frequency of testing in the non-structural zone berelaxed?The frequency of testing in the non-structural zone can berelaxed, assuming the non-structural zone has been properlydelineated. If a non-structural zone is to eventually becomepart of a structural zone it would be prudene to test morefrequently, as this wouid simplify future modifications greatly.At this time, MSHA does not have an alternative testing intervalfor the non-structural zone. As mentioned above the variationwill have to be reviewed on a site by site basis.4. Where should tests ~e performed?The tests should be conducted at random locations throughout thelift. In addition, any area thought to have been ineffectivelycompacted should be tested. Areas SUbjected to the greatestvehicle compactive effort (primary haul roads) should be avoided.In cases where two-foot lifts have been specified, tests shouldbe conducted at mid-depth of the lift and at the top of the lift.The same is true for 18" thick lifts. The question was raisedduring one of the sessions by an operator who mentioned the top6" of each 12" lift were disturbed due to dozer'tracking. Inthis case, MSHA recommends testing 6" down in each lift whichwould essentially be the same as testing every 12" depth ofmaterial placed.5. Where shoUld new material for laboratory compaction tests beo~tained?Assuming this question pertains to the 1 Proctor test / 20 fielddensity tests, MSHA normally recommends that the material betaken directly from the plant.

59

6. CanUBA approve liD "exceeclencecriteria" for moisture-clensity testing?

This question relates to an acceptable amount of failed tests per"X" passing tests. At this time MSHA does not have such acriteria. The decision on whether a particular amount of'failures is acceptable has been, left up to the MSHA fieldinspecting personal.

. :'':<':'~>-:,e..~..:,:,,' .:.':,,:::;_,:~~7. In relation to the lproctor"t.s':'-/20 moisture clensitytests, is it appropriate to us., the '''one-point Proctor" test?

MSHA has .not established a criteria concerning usage·of the "one-point Proctor". If results of your moisture-density tests showconsistency in meeting the established values, the company maypropose to periodically check a point on the Proctor curve. MSHArecommends that the "one-point Proctor" not be used exclusively.Perhaps the one-point Proctor test could be used every other orevery third time the full Proctor curve is developed. Again,this would have to be addressed on a site by site basis, andwould be at the discretion of the reviewer.Also, this would be related to the overall homogeneity of the~aterial. Coarse refuse materials may tend to be more uniformthan material taken from a borrow pit.

60

SHORT TERH CRITERIALarry Wilson

Summary of Presentationoccasionally, impounding structures cannot safely handle the fulldesign storm for short periods of time. 'The implementation ofShort Term Design criteria for a particular site may be necessarybecause of procedural changes at the facility or trans~tions fromone stage to the next. During these unavoidable periods, a lesssevere design storm may be considered. ~ ,The maximum length of time considered under Short Term criteriais two years. This does not mean that in all instances two yearsis appropriate. Short Term criteria is limited to the shortest'possible time required to complete the transition.The appropriate Short Term design storm is listed in MSHA'sDesign Guidelines and is based on size and hazard rating.

ouestions Regarding PresentationThere was only one question concerning Short 'Term Design criteriathat was not addressed in the PIL. .1. When does the short term clock start when a site is to beabandoned?When the impoundment is no longer able to handle the full designstorm.

61

GRADED P:ILTI;RSLarry Wilson

summary of PresentationThe main,purpose of graded filters is to prevent basematerials from flowing into internal drainage systems. Thisis accomplished by either of two widely accepted methods.The first method was developed by Mr. K. Terzagl1!'in'the early-1940s. The second method was developed by the SCS in the.mid-1980s. The second method is recommended by the Bureau of,Reclamation. Both are based on the r~lative grain sizes of thefilter and base material.'The primary difference between MSHA's requirements and thesemethods is that MSHA uses maximum/minimum gradation sizes to makethe comparisons in the formulas rather than the average values.The reason for this difference is that the mining industry hassuch a wide range of possible materials that will be allowed inan impounding facility. A very coa~se material may not becompatible with a very fine material.

oue&tions Regarding presentation1. :Isit acceptable to use coarse refuse as the filter medium ina graded filter?Coarse refuse will deteriorate with time and exposure to theelements. It is therefore not generally used as a granularfilter medium. If, however, the coarse waste material is veryclean, sound, and is buried immediately, coarse waste materialmay sometimes be used in a filter zone. This zone must beSUbstantially larger than a conventional filter zone tocompensate for degradation of the coarse waste material. Thecoarse waste material may also need to be enclosed in a filtermedia to help prevent fines migration.2. :Isthe use of limestone material in a drain ever permitted?There are many drains with limestone material in them. Most ofthese, however, are in underdrains below refuse piles. There aresome instances 'of limestone in drains in impounding structures,but these are rare.Drains and filters are critical design features of many dams;Drains are used to maintain internal seepage patterns atprescribed levels. If they were 'tomalfunction, the phreaticsurface could exceed design levels and result in a significantreduction in the slope stability' of an imp~unding facility.

62

Filters are designed to control seepage so that piping ofembankment material is prevented. The failure of a:filter couldresult in uncontrolled seepage and the development of piping. Itis therefore obvious that if a drain or filter is necessary inmaintaining internal seepage at a certain level, or controllingseepage exiting a facility, their performance over the long termmust be ensured.The acidic environment that drains and filters in mine "wastestructures are exposed to dictate special considerations.Dissolution of calcareous limestone and sandstone will occur ifexposed to mine acid. Seepage flowing through structuresconstructed of coarse mine waste will generally become acid as aresult of pyrite oxidation.Specification for anx aggregate to be used in drains and filtersmust detail soundness, durability and resistance to dissolution·resulting from exposure to acidic environments. Appropriate testand acceptable performance limits should also be specified. .3. How do you know when a filter is clogged?There are several warning signs that a drain may have becomeplugged or is not functioning properly.

A. A decrease (or increase) in flow.from the outlet of thedrain.

B. An increase in the piezometr~c level in the area of thedrains.c. Seepage flowing around the outlet to the drain.

D. Discoloration or fine particles being transported withthe water being discharged from the drain.

ALTERNAT\VESTO THE PLAN ..

REViEW······ .PROCESS

U.S. Department of Labor Mine Safety and Health Administration

4015 Wilson BoulevardArlington, Virginia 22203-1984

EFFECTIVE DATE: 07/01/90 EXPIRATION DATE: 03/31/92

PROCEDURE INSTRUCTION LETTER NO. 190-11-1

MADISON McCULLOCH /veI'~"',- - ,tt..cCU e~d-Director of Technical Support

FROM:

SUBJECT: Impounding Structures Safety Design Procedures:Compaction Specifications

ScopeThis procedure instruction letter has been prepared to guide MSHApersonnel who are responsible for reviewing impoundment plans.It is MSHA' s intent that this information also be shared with theoperators and designers of coal mine impounding structures.

PurposeDesign and construction plans for active and proposed impoundingstructures are submitted by coal mine operators to MSHA forreview and approval. It is MSHA' s intent to maintain its designand construction requirements to the latest proven pr inciples of"current, prudent engineer ing practices."

Mine waste dams and impoundments have some features which maydiffer from typical earth dams. When proposed design plansdeviate from standard design cr iter ia for dam safety, the plansand supporting documentation must clearly indicate the reason forthe deviation, and provide a technical basis for the proposeddesign. While this information is presented to guide thereviewer, it remains the responsibility of the designer to keepabreast of changes in technology for impoundment structures andto design accordingly.

This letter is the first in a series addressing important designissues. MSHA' s goal is that these issues be thoroughlyconsidered so that plans have a sound engineering basis, withproper documentation. The issues addressed have been selectedbased on MSHA' s experience with reviewing submitted plans. Theyrepresent items of particular concern or points of contention.They do not address all of the issues which must be considered indesigning an impounding facility. MSHA intends to update andexpand on the issues as needed.

The following information presents MSHA' s current consensus on animportant design issue.

2

Procedure InstructionsProper compaction of embankment material is one of the mostimportant elements in the construction of a safe dam. As statedin Enqineer inq and Desiqn Manual - Coal Refuse DisposalFacilities, E. D'Appolonia Consulting Engineers, Inc., 1975, "Anysoil placed as a constructed structural fill, including coalrefuse embankments, is normally compacted to increase density andshear strength and to decrease compressibility and permeability."Testing has shown that a small change in the density of coarsecoal refuse can have a significant impact on some of itsproperties.Compaction specifications need to place acceptable limits on thefollowing: 1) the minimum dry density, 2) the range of placementwater content and, 3) the maximum lift thickness. In arriving atthese specifications, it is prudent that the recommendations andpractices of authoritative, experienced dam builders be used forguidance.

Some pertinent references on compaction specifications are asfollows:

1. Naval FacUities Engineer ing Command, NAVFAC DM-7. 2,May 1982, Table 4, page 7.2-46, for earth dam greaterthan 50-feet high the required density is 95 percentof modified Proctor, moisture limits of -1 to +2percent of optimum, and 12(~)-inch compacted liftthickness.

2. Corps of Engineers, Earth and Rock Fill Dams, EM 1110-2-2300, March 1971, pages 5-13, "Selection of designdensities, while a matter of judgement, should be basedon the results of test fills or past experience withsimilar soils and field compaction equipment. The usualassumption is that field densities will not exceed themaximum densities obtained from the standard compactiontest nor be less than 95 percent of maximum densitiesderived from this test."

3. Bureau of Reclamation, Desiqn of Small Dams, ThirdEdition, 1987, Table £-1, page 657. Cohesive soilscontrolled by Proctor test having 0-25 percent plusNo. 4 fraction by weight should have a minimumacceptable density of 95 percent and a desirable averagedensi ty of 98 percent; and 26-50 percent plus No. 4fraction by weight should have a minimum acceptabledensity of 92.5 percent and a desirable average densityof 95 percent. More than 50 percent plus No. 4 fractionby weight should have a minimum acceptable density of 90percent and a desirable average density of 93 percent.These percentage densities are based on the minus No. 4fraction and limit moisture content to -2 to +2 percent

3

of optimum. Permeability testing should be per formedon cohesive soils which contain more than 50 percentgravel and are used as a water barrier.

4. S. K. Saxena, D. E. Lourie, and J. K. Ras, CompactionCr iter ia for Eastern Coal Waste Embankments, Journal ofGeotechnical Engineer ing, Volume 110, No.2, February1964, "Recommendation. - Based on the findings of thisstudy, it is recommended that coarse coal refuse,typical of eastern United States coal regions, becompacted near the optimum moisture content to a densitygreater than 95 percent of maximum dry densitydetermined in accordance with ASTM D-698. Compactedlifts should not be greater than 1 ft. (0.3m) inthickness. "

The following recommendations are made for the structural fillportions of impounding structures:

1. Mater ial should be compacted to at least 95 percentof the maximum dry density as defined by the standardProctor test, with the placement water content not exceedingthe range of -2 to +3 percent of optimum.

2. In compacting coarse coal refuse, the lift thicknessshould not exceed 12 inches. When fine-grained soils areused for embankment construction, lift thickness should notexceed 8 inches.

3. For materials where the Proctor moisture-densityrelationship does not apply, specifications should bebased on relative density test values.

Less stringent compaction specifications than those cited abovewould not generally be consistent with current, prudentengineer ing practices. Plans with such specif ications cannot berecommended for approval unless a detailed technicaljustification, which demonstrates that the proposed practicewould have no adverse affect on the safety of the dam, can beprovided by the designer. The designer would need to showthrough testing and analyses that all potential problems,including settlement, cracking, piping, instability,stratification, and seepage, have been taken into account in thedesign and that compensating design features have beenincorporated. It should be noted that less stringent compactionspecifications can generally be used in areas which can be shownto be "non-structural" portions of the dam.

4

References

1. Bureau of Reclamation, Desiqn of Small Dams, U.S.Department of the Interior, 1987.

2. Chen, C. Y ., Enqineer inq Properties of Compacted CoalMine Waste, American Society of Civil Engineers,New York, New York, May 1981.

3. Corps of Engineers, Earth and Rockf ill Dams, EM 1110-2-2300, U.S. Department of the Army, March 1971.

4. Corps of Engineers, Enqineerinq and Dam Desiqn -Stability of Earth and Rockfill Dams, EM 1110-2-1902,U.S. Department of the Army, April 1, 1970.

5. E. D' Appolonia Consulting Engineers, Inc., Enqineer inqand Desiqn Manual - Coal Refuse Disposal Facilities,U. S. Department of the Inter ior, Mining Enforcement andSafety Administration (MESA), 1975.

6. Naval Facilities Engineering Command, NAVFAC DM-7.2,U.S. Department of the Navy, May 1982.

7. Saxena, S. K., et. al., Compaction Cr iter ia for EasternCoal Waste Embankments, Journal of GeotechnicalEngineering, Volume 110, No.2, February 1984.

8. Shah, N. S., et. al., Compaction Cr iter ia for Coal Wast~Embankments, U.S. Department of the Interi.or, Bureau ofMines, Contract No. JOI00031, August 1981.

AuthorityThese procedures were developed as an aid to compliance with theprovisions of 30 CFR 77.216 through 77.217.

Filinq InstructionsMSHA personnel should file this letter behind the tab marked"Procedure Instruction Letters" in the binder labelled "ProgramPolicy Handbooks and Procedure Instruction Letters."

Issuinq Office and Contact PersonOffice of Technical SupportJohn J. Mulhern (703) 235-1590, FTS 235-1590

Distr ibutionCoal Special Interest GroupsCoal & All Volume Program Policy Manual HoldersAll Coal Mine Operators