Embed Size (px)
Citation preview
CECW-EG
Engineer Manual
1110-2-1914
Department of the ArmyU.S. Army Corps of Engineers
Washington, DC 20314-1000
EM 1110-2-1914
29 May 1992
Engineering and Design
DESIGN, CONSTRUCTION, ANDMAINTENANCE OF RELIEF WELLS
Distribution Restriction StatementApproved for public release; distribution is
unlimited.
EM 1110-2-191429 MaY 1992
US Army Corpsof Engineers
ENGINEERING AND DESIGN
Design, Construction, andMaintenance of Relief Wells
ENGINEER MANUAL
DEPARTMENT OF THE ARMY EM 1110-2-1914U.S. Army Corps of Engineers
CECW-EG Washington, D.C. 20314-1000
Engineer ManualNo. 1110-2-1914 29 May 1992
Engineering and DesignDESIGN, CONSTRUCTION, AND MAINTENANCE OF RELIEF WELLS
1. Purpose. This manual provides guidance and information on the design, construction, and mainte-nance of pressure relief wells.
2. Applicability. The provisions of this manual are applicable to all HQUSACE/OCE elements, majorsubordinate commands, districts, laboratories, and field operating activities (FOA) having responsibilityfor seepage analysis and control for dams, levees, and hydraulic structures.
FOR THE COMMANDER:
Colonel, Corps of EngineersChief of Staff
Department of The Army EM 1110-2-1914US Army Corps of Engineers
CECW-EG Washington, DC 20314-1000
Engineer Manual 29 May 1992No. 1110-2-1914
Engineering and DesignDESIGN, CONSTRUCTION, AND MAINTENANCE OF RELIEF WELLS
Table of Contents
Subject Paragraph Page
Chapter 1IntroductionPurpose . . . . . . . . . . . . . . . . . . . . . . . 1-1 1-1Objective and Scope. . . . . . . . . . . . . . . 1-2 1-1Applicability . . . . . . . . . . . . . . . . . . . . 1-3 1-1References . . . . . . . . . . . . . . . . . . . . . 1-4 1-1General Consideration. . . . . . . . . . . . . 1-5 1-1
Chapter 2Relief Well ApplicationsDescription . . . . . . . . . . . . . . . . . . . . . 2-1 2-1Use of Wells . . . . . . . . . . . . . . . . . . . . 2-2 2-1History of Use. . . . . . . . . . . . . . . . . . . 2-3 2-1Other Applications. . . . . . . . . . . . . . . . 2-4 2-3
Chapter 3Basic ConsiderationsFoundation Investigations. . . . . . . . . . . 3-1 3-1Foundation Permeability. . . . . . . . . . . . 3-2 3-1Anisotropic Conditions. . . . . . . . . . . . . 3-3 3-1Chemical Composition of
Ground Waters . . . . . . . . . . . . . . . . 3-4 3-1Seepage Analysis. . . . . . . . . . . . . . . . . 3-5 3-1Allowable Heads. . . . . . . . . . . . . . . . . 3-6 3-6
Chapter 4Analysis of Single WellsAssumptions . . . . . . . . . . . . . . . . . . . . 4-1 4-1Circular Source. . . . . . . . . . . . . . . . . . 4-2 4-1Noncircular Source . . . . . . . . . . . . . . . 4-3 4-1Infinite Line Source. . . . . . . . . . . . . . . 4-4 4-1Finite Line Source. . . . . . . . . . . . . . . . 4-5 4-3Infinite Line Source and Infinite
Line Sink . . . . . . . . . . . . . . . . . . . . . 4-6 4-3
Subject Paragraph Page
Infinite Line Sink and Infinite Barrier . . 4-7 4-3Complex Boundary Conditions. . . . . . . 4-8 4-3Partially Penetrating Wells. . . . . . . . . . 4-9 4-3Effective Well Penetration . . . . . . . . . . 4-10 4-6
Chapter 5Analysis of Multiple Well SystemsGeneral Equations. . . . . . . . . . . . . . . . 5-1 5-1Empirical Method . . . . . . . . . . . . . . . . 5-2 5-1Circular Source. . . . . . . . . . . . . . . . . . 5-3 5-2Wells Adjacent to Infinite Line
Source with Impervious TopStratum . . . . . . . . . . . . . . . . . . . . . . 5-4 5-2
Infinite Line of Wells . . . . . . . . . . . . . . 5-5 5-2Top Stratum Conditions. . . . . . . . . . . . 5-6 5-2Infinite Line of Wells, Impervious
Top Stratum. . . . . . . . . . . . . . . . . . . 5-7 5-2Well Factors . . . . . . . . . . . . . . . . . . . 5-8 5-8Infinite Line of Wells, Impervious
Top Stratum of Finite Length. . . . . . . 5-9 5-8Infinite Line of Wells, Impervious
Top Stratum Extending toBlocked Exit . . . . . . . . . . . . . . . . . . 5-10 5-8
Infinite Line of Wells, DischargeBelow Ground Surface. . . . . . . . . . . 5-11 5-8
Infinite Line of Wells, No Top . . . . . . . 5-12 5-12Finite Well Lines, Infinite Line
Source. . . . . . . . . . . . . . . . . . . . . . .5-13 5-9
Chapter 6Well DesignDescription of Well . . . . . . . . . . . . . . . 6-1 6-1
i
EM 1110-2-191429 May 92
Subject Paragraph Page
Materials for Wells . . . . . . . . . . . . . . . 6-2 6-1Selection of Materials. . . . . . . . . . . . . . 6-3 6-1Well Screen . . . . . . . . . . . . . . . . . . . . 6-4 6-2Filter . . . . . . . . . . . . . . . . . . . . . . . . . 6-5 6-2Selection of Screen Opening Size. . . . . 6-6 6-3Well Losses . . . . . . . . . . . . . . . . . . . . 6-7 6-4Effective Well Radius . . . . . . . . . . . . . 6-8 6-5
Chapter 7Design of Well SystemsGeneral Approach. . . . . . . . . . . . . . . . 7-1 7-1Design Heads. . . . . . . . . . . . . . . . . . . 7-2 7-1Boundary Conditions. . . . . . . . . . . . . . 7-3 7-1Design Procedures. . . . . . . . . . . . . . . . 7-4 7-1Infinite Line of Wells, Impervious
Top Stratum . . . . . . . . . . . . . . . . . . 7-5 7-1Infinite Line of Wells, Wells in Ditch . . 7-6 7-2Infinite Line of Wells with Impervious
Top Stratum of Finite Length . . . . . . 7-7 7-2Computer Programs . . . . . . . . . . . . . . 7-8 7-4Head Distribution for Finite Line of
Relief Wells . . . . . . . . . . . . . . . . . . 7-9 7-4Well Systems at Outlet Works and
Spillways . . . . . . . . . . . . . . . . . . . 7-10 7-5Well Costs . . . . . . . . . . . . . . . . . . . . . 7-11 7-5Seepage Calculations. . . . . . . . . . . . . . 7-12 7-5
Chapter 8Relief Well InstallationGeneral Requirements. . . . . . . . . . . . . 8-1 8-1Standard Rotary Method . . . . . . . . . . . 8-2 8-1Reverse-Rotary Method. . . . . . . . . . . . 8-3 8-1Bailing and Casing. . . . . . . . . . . . . . . . 8-4 8-3Bucket Augers . . . . . . . . . . . . . . . . . . 8-5 8-3Disinfection . . . . . . . . . . . . . . . . . . . . 8-6 8-3Installation of well Screen and Riser
Pipes . . . . . . . . . . . . . . . . . . . . . . . . 8-7 8-5Filter Placement . . . . . . . . . . . . . . . . . 8-8 8-5Development. . . . . . . . . . . . . . . . . . . . 8-9 8-6Chemical Development . . . . . . . . . . . . 8-10 8-6Mechanical Development. . . . . . . . . . . 8-11 8-6Sand Infiltration . . . . . . . . . . . . . . . . . 8-12 8-8Testing of Relief Wells . . . . . . . . . . . . 8-13 8-8Backfilling of Well . . . . . . . . . . . . . . . 8-14 8-9Sterilization. . . . . . . . . . . . . . . . . . . . . 8-15 8-9Records . . . . . . . . . . . . . . . . . . . . . . .8-16 8-9Abandoned Wells. . . . . . . . . . . . . . . . . 8-17 8-9
Subject Paragraph Page
Chapter 9Relief Well OutletsGeneral Requirements. . . . . . . . . . . . . 9-1 9-1Check Valves . . . . . . . . . . . . . . . . . . . 9-2 9-1Outlet Protection. . . . . . . . . . . . . . . . . 9-3 9-1Plastic Sleeves. . . . . . . . . . . . . . . . . . . 9-4 9-1
Chapter 10Inspection Maintenance and EvaluationGeneral Maintenance. . . . . . . . . . . . . . 10-1 10-1Periodic Inspections . . . . . . . . . . . . . . 10-2 10-1Pumping Tests. . . . . . . . . . . . . . . . . . . 10-3 10-1Records . . . . . . . . . . . . . . . . . . . . . . .10-4 10-1Evaluation . . . . . . . . . . . . . . . . . . . . . 10-5 10-2
Chapter 11Malfunctioning of Wells andReduction in EfficiencyGeneral . . . . . . . . . . . . . . . . . . . . . . .11-1 11-1Mechanical . . . . . . . . . . . . . . . . . . . . . 11-2 11-1Chemical . . . . . . . . . . . . . . . . . . . . . . 11-3 11-1Biological Incrustation . . . . . . . . . . . . . 11-4 11-2
Chapter 12Well RehabilitationGeneral . . . . . . . . . . . . . . . . . . . . . . .12-1 12-1Mechanical Contamination. . . . . . . . . . 12-2 12-1Chemical Treatment with
Polyphosphates. . . . . . . . . . . . . . . . 12-3 12-1Chemical Incrustations . . . . . . . . . . . . 12-4 12-1Bacterial Incrustation . . . . . . . . . . . . . 12-5 12-1Recommended Treatment. . . . . . . . . . . 12-6 12-2Specialized Treatment. . . . . . . . . . . . . 12-6 12-3
Appendices
Appendix AReferences
Appendix BMathematical Analysis of Underseepageand Substratum Pressures
Appendix CList of Symbols
ii
EM 1110-2-191429 May 92
Chapter 1Introduction
1-1. Purpose
This manual provides guidance and information on thedesign, construction, and maintenance of pressure reliefwells.
1-2. Objective and Scope
The objective of this manual is to provide guidance andinformation on the design, construction, and mainte-nance of pressure relief wells installed for the purposeof relieving subsurface hydrostatic pressures which maydevelop within the pervious foundations of dams, levees,and hydraulic structures.
1-3. Applicability
The provisions of this manual are applicable to allHQUSACE/OCE elements, major subordinate com-mands, districts, laboratories, and field operatingactivities (FOA) having responsibility for seepageanalysis and control for dams, levees, and hydraulicstructures.
1-4. References
Appendix A contains a list of required and related publi-cations pertaining to this manual. Unless otherwisenoted, all references are available on interlibrary loanfrom the Research Library, ATTN: CEWES-IM-MI-R,US Army Engineer Waterways Experiment Station,3909 Halls Ferry Road, Vicksburg, MS 39180-6199.
1-5. General Considerations
All water retention structures are subject to seepagethrough their foundations and abutments. In many casesthe seepage may result in excess hydrostatic pressures oruplift pressures beneath elements of the structure orlandward strata. Relief wells are often installed torelieve these pressures which might otherwise endangerthe safety of the structure. Relief wells, in essence, arenothing more than controlled artificial springs thatreduce pressures to safe values and prevent the removalof soil via piping or internal erosion. The properdesign, installation, and maintenance of relief wells areessential elements in assuring their effectiveness and theintegrity of the protected structure.
1-1
EM 1110-2-191429 May 92
Chapter 2Relief Well Applications
2-1. Description
Pressure relief wells as used in this manual refer tovertically installed wells consisting of a well screensurrounded by a filter material designed to preventinwash of foundation materials into the well. A typicalrelief well is shown in Figure 2-1. The wells, includingscreen and riser pipe, have inside diameters generallybetween 6 and 18 inches (in.), sized to accommodate themaximum design flow without excessive head loss.Well screens generally consist of wire-wrapped steel orplastic pipe, slotted or perforated steel or plastic pipe.Slotted wood stave well screens, which are no longermanufactured, are found in many existing installations.Details of various well screens are given in Chapter 6.
2-2. Use of Wells
a. Relief wells are used extensively to relieveexcess hydrostatic pressures in pervious foundationstrata overlain by more impervious top strata, conditionswhich often exist landward of levees and downstream ofdams and various hydraulic structures. Placing the welloutlets in below-surface trenches or collector pipesserves to dry up seepage areas downstream of leveesand dams. Relief wells are often used in combinationwith other underseepage control measures, such asupstream blankets, downstream seepage berms, andgrouting. Horizontal stratification of pervious founda-tion deposits is not a major deterrent to the use of reliefwells, as each of the more pervious foundation stratacan be penetrated. The use of relief wells for leveesystems is discussed in EM 1110-2-1913; their use forearth and rock-fill dams is discussed inEM 1110-2-2300.
b. Relief wells provide a flexible control measure asthe systems can be easily expanded if the initial systemis not adequate. Also, the discharge of existing wellscan be increased by pumping if the need arises. Arelief well system requires a minimum of additional realestate as compared with other seepage control measuressuch as berms. However, wells require periodic mainte-nance and frequently suffer loss in efficiency with timefor a variety of reasons such as clogging of well screensby intrusions of muddy surface waters, bacterial growth,or carbonate incrustation. Relief wells may increase theamount of underseepage which must be handled at theground surface, and means for collecting and disposing
of their discharge must be provided (Turnbull andMansur 1954). Adequate systems of piezometers andflow measuring devices must be installed in accordancewith ER-1110-2-110 and EM 1110-2-1908 to providecontinuing information on the performance of relief wellsystems.
2-3. History of Use
a. The first use of relief wells to prevent excessiveuplift pressures at a dam was by the US Army EngineerDistrict, Omaha, when 21 wells were installed from July1942 to September 1943 as remedial seepage control atFort Peck Dam, Montana (Middlebrooks 1948). Thefoundation consisted of an impervious stratum of clayoverlaying pervious sand and gravel. Although steelsheet piles were driven to provide a complete cutoff,leakage occurred and high hydrostatic pressure devel-oped at the downstream toe with an excess head of45 feet (ft) above ground surface. The high pressurewas first observed in piezometers installed in thepervious foundation. The first surface evidence of thehigh hydrostatic pressure came in the form of dischargefrom an old well casing that had been left in place.Since it was important that the installation be made asquickly as possible, 4- and 6-in. well casings, availableat the site, were slotted with a cutting torch and installedon 250-ft centers in the pervious stratum with solid(riser) pipes extending to the surface. The excess headat the downstream toe was reduced from 45 to 5 ft, andthe total flow from all wells averaged about 4500 gal-lons per minute (gpm). However, the steel screenscorroded severely and in 1946 were replaced by17 permanent wells consisting of 8-in.-ID slotted red-wood pipe at a spacing of 125 ft.
b. The first use of relief wells in the original designof a dam was by the US Army Engineer District,Vicksburg, when wells were installed during construc-tion of Arkabutla Dam, Mississippi, completed inJune 1943. The foundation consisted of approximately30 ft of relatively impervious loess underlain by apervious stratum of sand and gravel. The relief wellswere installed to provide an added measure of safetywith respect to uplift and piping along the downstreamtoe of the embankment. The relief wells consisted of2-in. brass wellpoint screens 15 ft long attached to 2-in.galvanized wrought iron riser pipes spaced at 25-ft inter-vals located along a line 100 ft upstream of the down-stream toe of the dam. The tops of the well screenswere installed about 10 ft below the bottom of the im-pervious top stratum. The well efficiency decreasedover a 12-year period by about 25 percent primarily
2-1
EM 1110-2-191429 May 92
Figure 2-1. Typical relief well (after EM 1110-2-1913)
2-2
METAL WELL GUARD
PLASTIC STAND PIPE
--.---' '
z z I I
0 ..;N
VARIABLE-TO TOP OF----\ MINIMUM WATER TABLE __J _
RISER PIPE
2-FT.
"9'o. 0 0
'0 0 '3 ~ ,.-9
1-IN. WIRE MESH
RUBBER GASKET
CAST IRON TENON
BACKFILL
_,- SAND BACKFILL
OF GRAVEL FILTER
TOP OF WELL SCREEN
-BLANK PIPE THROUGH VERY FINE SAND OR SILT STRATA
o--~ •--- GRAVEL FILTER 4-IN. MIN. 0 ao
•• 0
.. c:c 0 .o
¢'., (]: ,<;1 ·0'
o a •.J<·v,
• • . • a 0 PERFORATED OR
SLOTTED SCREEN
6-lN. MIN.
EM 1110-2-191429 May 92
as a result of clogging of the screens. However, thepiezometric head along the downstream toe of the dam,including observations made at a time when the spillwaywas in operation, has not been more than 1 ft above theexcess head of 9 ft was observed (US Army EngineerWaterways Experiment Station 1958). Since these earlyinstallations, relief wells have been used at many leveelocations to control excessive uplift pressures and pipingthrough the foundation.
2-4. Other Applications
Pressure relief wells have also been used extensivelybeneath the stilling basins of spillways, outlet structures,
and other hydraulic structures. In addition, wells havebeen employed to control excess hydrostatic pressures inoutlet channels including areas immediately downstreamof navigation locks. Often wells incorporated in struc-tures have been located so that they discharge throughcollector pipes and manholes which are not readilyaccessible to cleaning and maintenance unless the struc-tures are dewatered. An example of a relief well systemincorporated into a toe drainage system for a dam isshown in Figure 2-2.
2-3
EM
1110-2-191429
May
92
2-4
@ORIGINAL GROUND
. ~EXCAVATION, PERVIOUS COMMON
t CHA~NEL
I
6'-0'' PROCESSED DRAIN
Figure 2-2. The drainaga systQfl'l wJih relief wei& at Cochiti Dam
5' -0" SELECT PERVIOUS FILL
3'-0"
~ TOE TRENCH ~ £ ~EUEF WEllS 5'-1)'' SELECT I PERVIOUS rill ' L
1 ~- 4'-0"
/
::::JI 1.5
/
TOE TRENCH INV.
2'-D"
15'-0"
EM 1110-2-191429 May 92
Chapter 3Basic Considerations
3-1. Foundation Investigations
The design of a relief well system should be precededby thorough field and geologic studies conducted inaccordance with EM 1110-1-1804. Sufficient boringsshould be made to define seepage entrance and exitconditions, the depth, thickness, and physicalcharacteristics of the pervious strata, as well as thethickness and physical characteristics of the top stratumin upstream or riverside areas and downstream orlandside areas. See Appendix B for further details.Particular attention should be given to the presence ofburied channels and pervious abutments which couldimpact on underseepage estimates. An example of ageneralized soil profile for relief well design along alevee reach is shown in Figure 3-1. The influence ofsurficial deposits on levee underseepage and on reliefwell design may be noted in Figure 3-2. High exitgradients and concentrations of seepage which mayoccur adjacent to clay-filled swales or channels willoften govern the locations of individual relief wells.Where soil conditions vary along the proposed line ofwells, the profile can be divided into a series of designreaches as shown in Figure 3-3. Additional borings, assubsequently described, should be made after completionof final design to ensure that a boring is located within5 ft of each final well location. In general, samplesshould be taken at intervals not greater than 3 ft or atchanges of soil strata, whichever occur first.
3-2. Foundation Permeability
Preliminary estimates of foundation permeability can bemade from laboratory tests or correlations with grainsize as described in EM 1110-2-1901. Because sam-pling operations do not necessarily indicate the relativeperviousness of foundations containing large amounts ofgravelly materials, field pumping tests are recommendedto verify the foundation permeability on all projectswhere the use of pressure relief wells is being consid-ered. The test well should fully penetrate the perviousaquifer, and a well flow meter should be used to deter-mine the variations in horizontal permeability withdepth. An example of data derived from a field pump-ing test conducted in this manner is shown in Fig-ure 3-4. Field pumping test procedures for steady stateand transient flow conditions are given in Appendix IIIto TM 5-818-5. Additional information, including pro-cedures for field permeability tests in fractured rock, is
given in EM 1110-2-1901. The vertical permeability ofindividual strata can be estimated from laboratory testson undisturbed samples or determined from field pump-ing tests (Mansur and Dietrich 1965).
3-3. Anisotropic Conditions
Analytical methods for computing seepage through apermeable deposit are based on the assumption that thepermeability of the deposit is isotropic. However,natural soil deposits are stratified to some degree, andthe average permeability parallel to the planes of stratifi-cation is greater than the permeability perpendicular tothese planes. Thus, the soil deposit actually possessesanisotropic permeability. To make a mathematicalanalysis of the seepage through an anisotropic deposit,the dimensions of the deposit must be transformed sothat the permeability is isotropic. Each permeablestratum of the deposit must be separately transformedinto isotropic conditions. In general, the simplest pro-cedure is to transform the vertical dimensions with thehorizontal dimensions unchanged.
3-4. Chemical Composition of Ground Waters
Some ground waters are highly corrosive with respect toelements of a pressure relief well or may containdissolved minerals or carbonates which could in timecause clogging and reduced efficiency of the well. Thechemical composition of the ground water, includingriver or reservoir supply waters, should be determinedas part of the design investigation. Sampling, samplepreservation, and chemical analyses of ground water iscovered in handbooks (Moser and Huibregtse 1976,Environmental Protection Agency 1976). Indications ofcorrosive and incrusting waters are given in Table 3-1.The chemical composition of ground water is a majorfactor in the chemical and biological contamination ofwell screens and filter packs as described in Chapter 11.
3-5. Seepage Analysis
The determination of whether relief wells are needed isbased on a seepage analysis which also provides theconditions for design of the relief well system. Theseepage analysis defines the entrance and exit conditionsand provides an estimate of substratum pressures whichmay exist under project flood conditions. On completedstructures where piezometric data are available, seepageanalyses are required to permit extrapolation of the datato the project flood conditions. The mathematicalanalysis of underseepage and substratum pressures iscontained in Appendix B.
3-1
EM
1110-2-191429
May
92
3-2
400
380
..J 360 Ill ::;:
,.: w w "-z 340 z 0 i= <(
> w 320 -' w
300 -
280
800+00
LEVEE STATIONING
81 0+00
NET GRADE OF LEVEE :::: EL 410.8
76 77
.PERVIOUS
Figure 3-1. Soli profile, relief w<~ll, and piezometer i"stallation data
820+00
RELIEF WELLS
81
400
380
360
340
320
300
-' 111 :l!i
t-w w "-:z
% 0
';;: > UJ ..J UJ
EM
1110-2-191429
May
92
3-3
LEGEND
SYMBOLS fOR TOPSTRATUI.t THICKNESS SYMBOL THICKNESS
JZ] 12 fEET OR MORE
D 8 TO 12 FEET
I : .. />.1 5 TO 8 FEET
D 5 FEET OR LESS
I· · I SEEPAGE AREA ..... L.t.L..U
-
SAND BOILS
FI!JUI'G 3-2. Sellpage thrwgh flOin1 bar deposits
EM
1110-2-191429
May
92
3-4
-':11 r -:11~11
1- DESIGN REACH 1
DAM
Zb = THICKNESS OF TOP STRATUM
D = THICKNESS OF SUBSTRATUM
F;gure 3-3. Profile of typical design reaches lor relief well analysis
DESIGN REACH 2
"'
_, IW
)'41=/
;:.,TLET STRUCTURE
ASSUMED IMPERVIOUS VALLEY FLOOR MATERIAL
EM 1110-2-191429 May 92
Figure 3-4. Coefficient of permeability and effective gain size of individual sand strata - Well FC-105
3-5
z 0
~
-
WELL WfC-i05
60~1t1G WB-105
p: cs
.·. ";:::
W'~"' ~""~ V.f.· 1.&1
370~--~---+--~----+----~--~----+----+---~
360~--+---~--J---+--~--+-~--+---~
I ~~ h\o 32 1.1 ~I" 350 1-::=:j::::;-t-t---' '¢.! • ~ -~ f- AVERAGE PERM£ ABILITY · · ~ i I--- OF SAND STRATA
I
@o.tz r%. .PK-\2 340 ,fl_ D K,r3000 • 10 -4CM/SEC -1-----J
;;;1~:: ~ : f--o_T~~~:::;_-_,++-,--+--+--1-----l--1------l fJ\0.37 Me E<i PK-13 ;; 0 l \.!.) ~ c
~ ®o.F ~· @0.04 I.IC G
@0.30 lift. (fj)0.76 t0
<::>· qj)1.20 •·····
@0.43 ~ G
cs @)0.30 ~
r'% @) 0.75 id"• @o.so-{<i o PK-14
~ ~ ~
NOTE:
5 :s1o H--+--~'----+----4----+----+--~----+---~ ~3oo5p~1===~=i==~==~--r--1---t--~
290 1--lp;:;;---t--+---+===~>, ,--+---·- -----i---1
280 ~-p
270~--~---+--~----+-+-+---4----+----+---1
i zso ,.0,----+--' --+--+-1--+--+--+-1 i
z5o t=:::±=±=t=:±:::__l__L__l _ _l__j 0 zooo •ooo sooo aooo
COEFFICIENT OF PERMEABILITY < 10 -•eM PER SE:C
l. WEU. FC-105 LOCATEO AT LEVEE STA 301+2'>. FT. CHARTRES l.EVEE DISTRICT. APPROX. 45 MilES SOUTH OF ST. LOUIS.
2. PERMEABILITIES SHOWN BY BAR GRAPH WERE COOIPUTEp .FROM WEU. t.IHER AND PIEZOMETER READINGS •
.3~ PERMEABIUTIES SHOWN BY 0 AfiD D WERE OEiiAIN£0 FROM LABORATORY T£STS ON BORING AND WELL DRIU.ING SAMI!'t.ES, RESPECTIVELY, .O,HC WERE ADN~T<:O TO THE EST1MA1ED NATURAL VOID RATIO BY THE !'ORMULA Kn=KL < ( •n;et) 2 AND CORRECTED TO T=20°C.
4. FtGUR£5 TO THE LEn OF BORINGS ARE 0 l O IN U!.l, OBTAINED fROM BORING SAMPLES: THOSE TO RIGHT OF BORINC fROI' WELL ORILLING SAMPLI;S.
5. CIRClED NUMBERS REFER TO GRAll! SIZE CURVES.
EM 1110-2-191429 May 92
Table 3-1Indicators of Corrosive and Incrusting Waters a
Indicators of Corrosive Water Indicators of Incrusting Water
1. A pH less than 7 1. A pH greater than 7
2. Dissolved oxygen in excess of 2 ppmb 2. Total iron (Fe) in excess of 2 ppm
3. Hydrogen sulfide (H2S) in excess of 1 ppm 3. Total manganese (MN) in excess ofdetected by a rotten egg odor a 1 ppm in conjunction with a high
pH and the presence of oxygen
4. Total dissolved solids in excess of 4. Total carbonate hardness in excess1,000 ppm indicates an ability to of 300 ppmconduct electric current great enough tocause serious electrolytic corrosion
5. Carbon dioxide (CO2) in excess of 50 ppm
6. Chlorides (CL) in excess of 500 ppm
Notes:a. From TM 5-818-5.b. ppm = parts per million.
3-6. Allowable Heads
Whenever a structure underlain by pervious deposits issubjected to a differential hydrostatic head, seepageenters the pervious strata, creating an artesian pressurebeneath the structure and downstream areas which couldresult in piping or failure by heave of the downstreamtop stratum. Pressure relief wells are designed to pre-vent piping and provide an adequate factor of safety,FS, with respect to uplift or heave. For this purpose,reduce the net head beneath the top stratum indownstream areas to an allowable value,ha. The equa-tion for FS is
(3-1)FSio
ic
γ ′/γw
ha/Zt
γ ′Zt
γwha
where
ic = critical upward hydraulic gradient, the ratio ofthe submerged weight of soil,γ ′, to the unitweight of water,γw
Zt = transformed thickness of downstream topstratum (see Appendix B)
The factor of safety with respect to uplift or heavenormally should be at least 1.5. In addition to providinga minimum factor of safety with respect to uplift ofheave (Condition a), relief wells may also be designedto ensure that piezometric heads in downstream areasare below ground surface, thereby preventing upwardseepage from emerging beneath the downstream topstratum (Condition b). The latter condition usuallyapplies to dams where visible seepage in downstreamareas is undesirable and can be prevented by installingthe wells with outlets in ditches or collector pipes alongthe embankment toe. The two conditions are illustratedin Figure 3-5.
a. Condition a. The allowable net head (ha) underthe top stratum of the downstream toe for this conditionis given by
(3-2)ha
ic
FSZt
b. Condition b. The maximum downstreampiezometric surface is defined by∆ hd which is thedifference between this surface and the elevation of thewell outlets corrected for well losses as subsequentlydescribed. For wells discharging into a collector ditch,
3-6
EM 1110-2-191429 May 92
Figure 3-5. Determination of allowable heads in downstream toe area
the factor of safety with respect to uplift below thebottom of the collector ditch should be at lease 1.5.The allowable net head under the top stratum below thebottom of the collector ditch for this condition is givenby the equation
(3-3)ha
ic
FSZc
whereZc is the transformed thickness of the downstreamtop stratum below the bottom of the collector ditch.
3-7
EM 1110-2-191429 May 92
Chapter 4Analysis of Single Wells
4-1. Assumptions
Analytical procedures for determining well flows andhead distributions adjacent to single artesian relief wellsare presented below. By definition, relief wells signifyartesian conditions, and equations for artesian flow areapplicable. In cases where wells are pumped, and gravityflow conditions exist, procedures for well analysis can befound in TM 5-818-5. It is assumed in the followinganalyses that all seepage flow is laminar or viscous, i.e.,Darcy’s Law is applicable. It is also assumed that steadystate conditions prevail; the rate of seepage and rate ofhead reduction have reached equilibrium and are not timedependent. Unless otherwise indicated, the well isassumed to penetrate the full thickness of the aquifer.
4-2. Circular Source
Certain geologic or terrain conditions may require theassumption of a circular source of seepage. The formulasfor a fully penetrating well located at the center of acircular source (see Figure 4-1) are
(4-1)hp HQw
2πkDln R
r
(4-2)hw HQw
2πkDln R
rw
where
hp = head at point p between the well and thesource
H = head at the source
Qw = well discharge
k, (kf) = coefficient of permeability of pervioussubstratum
D = thickness of pervious foundation
R = radius of circular source (radius ofinfluence)
hw = head at well
rw = radius of well
4-3. Noncircular Source
If geologic or terrain conditions indicate a noncircularsource of seepage, the radius of influence,R, may bereplaced byAc, defined as an effective average of thedistance from the well center to the external boundary.For a rectangular boundary of sides 2a and 2b, the valueof Ac is
(4-3)Ac
4abπ
4-4. Infinite Line Source
Conditions may arise where the flow to the well origi-nates from the bank of a river or canal reservoir oranother body of water. In such cases, the bank orshoreline may act as an infinite line source of seepage. Ifleakage occurs through the top stratum, the effective dis-tance to the infinite line source of seepage should becomputed as discussed in Appendix B. The solutions fora single well adjacent to an infinite line source (seeFigure 4-1) is determined using the method of imagesdescribed by Muskat (1937), Todd (1980), and EM 1110-2-1901. The formulas are
(4-4)hp HQw
2πkDln r′
r
(4-5)hw HQw
2πkDln 2S
rw
where
r′ = distance from pointp to image well
r = distance from pointp to real well
S= distance from real well to line source
A solution for hp is also presented in terms ofx and ycoordinates in Figure 4-1 (Equation 4-6).
4-1
EM 1110-2-191429 May 92
Figure 4-1. Summary of equations for artesian flow to single well
4-2
h ~ H Ow .!l_
p - 211kD In r
h ~ H Ow R
w - 21fk0 In 'w
CIRCULAR SOURCE
= • Q r'
h ~H--w-ln-p 2wk0 r
( h ~H-~InZS
w 2nkD 'w y X WELL IN TERMS OF X AND Y COORD!NA TES
s s
IMAGE WELL REAL WELL
= INFINITE UNE SOURCE
h = H- ~In 4S w 21Tk0
(c2 -r;f + 4S2c2 l
c2 - r; +}(c 2
- r~J + 4ic2 J [=I ,~LL
:L~ FOR WELL ON PERPENDICULAR BISECTOR, r0
= S
h = H - ~ In 25 G + £) w 21Tk0 r:\ f' 2
FINITE LINE SOURCE w ~
(4-1)
(4-2)
(4-4)
(4-5)
(4-6)
(4-7)
(4-8)
EM 1110-2-191429 May 92
4-5. Finite Line Source
In cases where the length of the source of seepage isrelatively small compared to its distance from the well,the source may be considered as a finite line source. Thesolution for a single well adjacent to a finite line sourcewas developed by Muskat (1937). The formulas, whichare available only in terms of head at the well, are shownin Figure 4-1 (Equations 4-7 and 4-8).
4-6. Infinite Line Source and Infinite Line Sink
As discussed in Appendix B, a semipervious landsideblanket can be replaced by a totally impervious topstratum and a theoretical line sink at an appropriateequivalent distance from the well. The theoretical linesink, parallel to the infinite line source, is referred to asan infinite line sink. A solution, based also on themethod of images, considering one of the infinite linesources as a sink, was developed by Barron (1948) and isshown in Figure 4-2.
4-7. Infinite Line source and Infinite Barrier
The method of images is an extremely powerful tool fordeveloping solutions to wells for various boundaryconditions. Solutions for various boundary conditionsincluding barriers are presented by Ferris, Knowles,Brown, and Stellman (1962), Freeze and Cherry (1976),and Todd (1980). For example, a typical problem wouldbe to calculate the discharge or heads for a single artesianwell located between a river denoted by an infinite linesource and a barrier such as a buried channel or rockbluff. In this case, the image well for the river wouldhave a second image well with respect to the rock bluffwhich in turn would have an image with respect to theriver and so on. A similar progression of image wellswould be needed for the impermeable barrier (seeEM 1110-2-1901). The image wells extend to infinity;however in practice, it is only necessary to include pairsof image wells closest to the real well because othershave a negligible influence on the drawdown. A solutionfor this case was presented by Barron (1982) and isshown in Figure 4-3.
4-8. Complex Boundary Conditions
Oftentimes, geologic factors impose conditions which aredifficult to simulate using circular or line sources andbarriers. In such cases, flow net analyses or electricalanalogy tests may be used to advantage especially when
the aquifer thickness is irregular and three-dimensionalanalyses are required. The use of flow nets for the designof well systems is described by Mansur and Kaufman(1962). Methods for conducting three-dimensional electri-cal analogy tests are described by Duncan (1963), Banks(1965), and McAnear and Trahan (1972).
4-9. Partially Penetrating Wells
The previous equations are based on the assumption thatthe well fully penetrates the aquifer. For practical rea-sons, it is often necessary to use wells which only par-tially penetrate the aquifer. The ratio of flow from apartially penetrating artesian well to that for a fully pene-trating well at the same drawdown is
(4-13)Qwp
Qw
Gp
or
(4-14)Qwp GpQw
2πkD(H hw)Gp
ln Rrw
where
Qwp = flow from partially penetrating well
Gp = flow correction factor for partiallypenetrating well
An approximate value ofGp can be obtained from thefollowing equation developed by Kozeny (1933):
(4-15)Gp
WD
1 7rw
2wcos πw
2D
where W/D is well penetration expressed as a decimal.An alternate equation developed by Muskat (1937)assuming a constant flow per unit length of well screen is
4-3
EM
1110-2-191429
May
92
4-4
= = P(x, Y)
"' " •
I ---------~ y '--"<;;"-'\ _.aw '" X WELL v 0 0 "" "' :::> z 'V -0 s x3 i7> r/ ./ ./>-:// //////// V///././///:1
"' "" I I
"' z I s II x3 I "' z :) z -::;
II -iil I
D k 1.., I I ;<; I I ---'
: I
I / //
= = x., PLAN SECTION
[ '""' ' ny rr(x+S)
] S + x3 - x Qw - CQ$ s + + '3 '3
hp ~ H s + - 47rkD In (4-9) y3 rry rr(x - S)
cosh 5 + - cos s + '3 '3
4ITk0 (H - H s ) s + '3
Q ~
w [2(S + •,)
2 ( 1- cos 5
2ns j ~ + ~ (4-10) In . 2 2
n 'w AFTER BARRON (1948) .
Figure 4-2. Dl'awdow~ tor well betw""'n lnlinim liN> sourco and do-stroam sink
EM
1110-2-191429
May
92
4-5
= =
• p(x, y)
y
X WELL
0
w s Ls 0<:
(.) "' "' ii? ::J L 0<: 0 < (11 ID
w w z z :::; ::l
= = PLAN
s : D k - BARRIER I I
SECTION
2 2 Q (SIN !12'. COSH !!!.. + SIN rrS) + (COS ~ SINH.!!!)
hp;H ___ w_ ln--~2~l _____ 2~L~----~2L~----~2~L~,--~2~L--4rrkD 2
(SIN.!!.! COSH !!:L - SIN rrS) + (COS rrx SINH!!¥. )2 2L 2l . 2L 2L 2L
0 (SIN rrS COSH rrZLrw + SIN 1rS )2 + (COS rrS SINH rrrw)
2
hw= H- ...::JL In __ ...,.:2)>.L ___ -=.!~---2~L,__~.-----22:!::L'---~2"'-L--4rrkD rrs 1T S 2 S 2
(SIN -2L COSH _!:ff- SIN:!;.__) + (COS 1!._ SINH 1rrw)
2L 2L · 2L 2L
Figure 4-3. Drawdown f<>r well batween infinite line scurce and inf.,ite barrier
( 4-11)
( 4-12)
EM 1110-2-191429 May 92
(4-
16)Gp
ln Rrw
D2w
2 ln 4Drw
G(T) ln 4DR
whereG(T) is a function ofW/D and approximate valuesfrom Harr (1962) are given in Table 4-1.
Table 4-1Partially Penetrating Well Function, G(T)
W/D G(T)
0.1 6.40.2 5.00.3 4.30.4 3.50.5 2.90.6 2.40.7 1.90.8 1.30.9 0.71.0 0.0
Values ofGp based on the above values for a typical well(rw = 1.0 ft) with a radius of 1,000 ft are plotted inFigure 4-4. An empirical method for calculating the headat any point for partially penetrating wells is described byWarriner and Banks (1977). Limitations of empiricalformulas for determining flows from partially penetratingwells are discussed in TM 5-818-5.
4-10. Effective Well Penetration
In a stratified aquifer, the effective well penetrationusually differs from that computed from the ratio of the
length of well screen to total thickness of aquifer. Todetermine the required length of well screen W to achievean effective penetration W
__in a stratified aquifer, the
procedure shown in Figure 4-5 can be used. It isassumed that the individual strata are anisotropic and eachstratum is transformed into an isotropic stratum inaccordance with the following equation:
(4-17)d dkh
kv
where
d = transformed vertical dimension
d = actual vertical dimension
kh = permeability in the horizontal direction
kv = permeability in the vertical direction
The horizontal dimension of the problem would remainunchanged in this transformation. The permeability of thetransformed stratum to be used in all equations for flowor drawdown is as follows:
(4-18)k khkv
wherek is the transformed coefficient of permeability.
4-6
EM 1110-2-191429 May 92
Figure 4-4. Flow to partially penetrating well with circular source
4-7
a_ <:.::><:.::>
z t= ~ 0 B ...: -' . o::W t--3: w Z<!> Wz "- -1->-<...JO:: -' 1- 0.6 -...:w -z t-w ~a.. "->-
-' ::,;_, 0=> E..._ o.4-.._ 3:0 01-~~ .._t--
oo 0 1- 0.2 ;:::-' <C-' a::W
3:
NOTE: CURVES ARE VALID FOR R = 1000 FT AND AND r = 1.0 H
w
OL-------~---------L--------L-------~--------' 0 20 40 60 80 100
WELL PENETRATION, W/D IN PERCENT
AFTER HARR (1962)
EM 1110-2-191429 May 92
Figure 4-5. Determination of actual and effective well penetrations
4-8
(r -lF dj : kv1 t khl
d2 I w I d3 I 0 d4 i
dl1l_ //W
ACTUAl SECTION
Actual wen penetration ~ W
Effective well penetration = W Actual well penetration in percent = W [D x 100 Effective well penelralion in pen::ent = W/D x 100
TRANSFORMED SECTION
1 Transform each layer into an isotropic layer of thickness d and penneabilily k
(4-17)
2. Calculate thickness of the 8-qufvalent homogeneous, isotroplc acquifer, 6
m-n ""'" D : L <imi<Hm L dmlf<vm
m-1 m-1
n "' number of strata, numbered from top to bottom
3. Calculate the elfective p!lrmaabilily of the transformed aquifer, k,
m-n
E dni<Hm ke =
,., m•n
E dmlkvnr m•1
4. Celculata tile eflective weD screen penetration into the transformed aquifer, WID
w w w E dk E dk E dkH
w - c . 0 = 0
D m-n D ke
m-n E dm km E rikH
m•1 m-1
!l
5. Determirw actual well pBnetratlon required to achieve a given effective wen penetration by succasstu# trials.
(4-18)
!4-19)
(4-20)
(4-21)
EM 1110-2-191429 May 92
Chapter 5Analysis of Multiple Well Systems
5-1. General Equations
In most applications, a system of pressure relief wells invarious arrays is required for the relief of substratumpressures or reduction of ground-water levels. In suchcases, analyses must be made to determine the numberand spacing of wells to meet these requirements. Thehead at any pointp produced by a system of fullypenetrating artesian wells was first determined byForcheimer (1914). His general equation as latermodified by Dachler (1936) is
or
(5-1)hp H1
12πkD
Qw1 ln
R1
r1
Qw2 lnR2
r2
. . . . .
Qwn lnRn
rn
(5-2)hp H1
12πkD
i n
i 1
Qwi lnRi
ri
where
H = gross head on system
n = number of wells in group
Qwi = discharge from ith well
Ri = radius of influence of ith well
r i = distance from ith well to point at which head iscomputed
The head,hwj, at any well, e.g. wellj, in a system ofnwells is determined from the equation
(5-3)hwj H1
12πkD
Qwj ln
Rj
rwj
i n 1
i 1
Qwi lnRi
ri , j
where
Qwj = flow from well j
Rj = radius of influence of well j
rwj = effective well radius of well j
r i,j = distance from each well to well j
The other symbols are as defined previously.Equations 5-1 and 5-3 as well as subsequent equationsfor multiple well systems are based on the principle ofsuperposition. Thus, the head at a given well in asystem of wells is equal to that resulting from this wellflowing as if no other wells were present minus thehead reduction caused at the well due to flow from theremaining wells. In most applications, the radius ofinfluence is large compared to the distance betweenwells and can be considered as constant. When wellsare pumped as in a dewatering system, the values ofQwi are known (or assumed). However, whenn wellsare used for pressure relief where they flow underartesian head conditions, the flow from each well mustbe computed taking into account the discharge elevationof each well. The procedure requires the solution ofnsimultaneous equations to determine individual wellflows.
5-2. Empirical Method
An empirical method developed by Warriner and Banks(1977) using the results of electrical analogy studies byDuncan (1963) and Banks (1965) can be used to deter-mine the head at any point within a random array offully or partially penetrating wells. The method,described in EM 1110-2-1901, is also valid forarbitrarily shaped source boundaries. A FORTRANcomputer code is provided by Warriner and Banks(1977).
5-1
EM 1110-2-191429 May 92
5-3. Circular Source
a. General case.The general equations for a groupof fully penetrating wells subject to seepage from acircular source with radiusR are shown in Figure 5-1.It is assumed that the radiusR is large with respect tothe distances between wells and that the flows fromeach well are equal. As indicated previously in the caseof variable well discharges, the procedure requires thesolution of n simultaneous to solve for individual wellflows.
b. Circular array of wells. A special case consistsof a circular array ofn wells equally spaced along thecircumference of a circle of radiusrc, the center ofwhich is also the center of a circular source of seepageof radius R. The general equations are shown inFigure 5-2.
c. Other well arrays. For other multiple-well sys-tems within a circular source, see Muskat (1937), Banks(1963), and TM 5-818-5.
5-4. Wells Adjacent to Infinite Line Source withImpervious Top Stratum
Where wells are located adjacent to a source which canbe approximated as an infinite line source and thepervious stratum is overlain by an impervious topstratum extending landward to a great distance, a solu-tion for heads and well flows is obtained using themethod of images. The equations are shown in Fig-ure 5-3 for the case of (a) equal well discharges and(b) variable well discharges. As noted previously,case (b) requires the solution ofn simultaneous equa-tions to determine individual well flows.
5-5. Infinite Line of Wells
An infinite line of wells refers to a system of wells thatconforms approximately to the following idealizedconditions:
a. The wells are equally spaced and identical indimensions.
b. The pervious stratum is of uniform depth andpermeability along the entire length of the system.
c. The effective source of flow and the effectivelandside exit or block, if present, are parallel to the lineof the wells.
d. The boundaries at the ends of the system areimpervious, normal to the line of the wells, and at adistance equal to one-half the well spacing beyond theend of the well system. For the above conditions, theflow to each well and the pressure distribution aroundeach well are uniform for all wells along the line.Therefore, there is no flow across planes centeredbetween wells and normal to the line, hence no overalllongitudinal component of the flow exists anywhere inthe system. The term infinite is applied to such a sys-tem because it may be analyzed mathematically byconsidering an infinite number of wells; the actual num-ber of wells in the system may be from one to infinity.
5-6. Top Stratum Conditions
The permeability and lateral extent of the top stratumlandward of an infinite line of wells can have apronounced effect on the performance of the well sys-tem. The assumption of a completely impervious topstratum extending landward to infinity is a convenientassumption for which theoretical solutions are available.However, this condition is rarely realized in practice. Amore general condition occurs when the impervious topstratum extends landward a finite distance terminating ata line sink. This condition is also applicable withrespect to results at the well line to the case of asemipervious top stratum which can be converted to anequivalent length of impervious top stratum using appro-priate blanket formulas. The two conditions are illus-trated in Figure 5-4 together with assumed head distri-butions with and without relief wells including theeffects of well losses. Calculation of the corrected nethead on the well system,h, should also take into consid-eration any extension of the well riser above tailwaterelevation. A third condition occurs when the pervioussubstratum is blocked at some point landward of thewell line. Theoretical solutions for the three conditionsfollow.
5-7. Infinite Line of Wells, Impervious TopStratum
The head midway between wells and the well flows forthe case of an impervious top stratum extending land-ward a great distance (L3 = ∞) may be calculated usingthe method of multiple images (after Muskat 1937,Middlebrooks and Jervis 1947). Solutions are shown inFigure 5-5 for the case of no well losses. Equa-tions 5-14 through 5-17 are applicable to both fullypenetrating and partially penetrating wells. The lattermake use of the so called well factors,Θa andΘm.
5-2
EM 1110-2-191429 May 92
Figure 5-1. Random array of fully penetrating wells with a circular source
5-3
CIRCULAR SOURC[
p
'3 H1 " WEll. 1
'2 3
2
PLAN
Th& hea<l at Po!ot P io:
hp - H1 - ....,;,. (ow1 In R • Dw2 In !!. • .... 0.., In !!.) ~1uu.1 r1 r2 r n
or
i .. n - , E ~i-1
Owi In R) f;
II 11
II II II II
srCTION
R
h,
PERVIOUS I SUBSTRATUM
.I
If all wruls have tha same tadius, and discharglil at ths sams elevation, hw, then-ths WlilU discharye~ are equal and given by:
(5·4)
(5·5)
(5·6)
(5-7).
EM 1110-2-191429 May 92
Figure 5-2. Circular array of fully penetrating artesian wells with a circular source
5-4
EM 1110-2-191429 May 92
Figure 5-3. Multiple wells adjacent to infinite line source - general case
5-5
INfiNITE LINE SOURCE
2' I' a...___ 0-._
------- r' • r' ' --
' ( j )
2
r,
p r,
IMAGE WELLS REAL WELLS
A. For e'fJal weH discharges
B. For variable well discharges
whera
' f· Qwiln..!..
r;
ri ""' distance from Point P to roof W~i I rj = distance from Point P 10 image Well I
'i = distance from Well j to real WeH i r~ = distance from WeU j to image Well ~ J
JNF'INITE UNE SOURCE
0
(5-11)
(5-12)
(5-13)
EM 1110-2-191429 May 92
Figure 5-4. Infinite line of wells with infinite or finite impervious top stratum - general case
5-6
SUBSTRATUM
s D
HEAD WITHOIJT WELLS
a. IMPERVIOUS TOP STRATIJM EXTENDING TO INFINITY
h-, H
0 s
H ; NET HEAD ON WELL SYSTEMS INCL
h ; CQRR. N<:T HEAD ON WELL SYSTEM
Hm= NET HEAD MIDWAY BETWEEN WELLS hm= CORR. NET HEAD MIOWAY BETWEEN WELLS
H0,f AVERAGE N<:T HEAD IN PLANE OF WEllS
h = CORR. NET HEAD IN PLANE OF WELLS ov
H w = WELL LOSSES W = WELL PENETRATION
HEAD WITHOUT WELLS
TOP STRATUM
SUBSTRATUM
L 3 OR x3
b. IMPERVIOUS TOP STRATUM OF FlN!TE LENGTH
EM 1110-2-191429 May 92
Figure 5-5. Infinite line of wells parallel to infinite line source - impervious top stratum
5-7
EM 1110-2-191429 May 92
5-8. Well Factors
The well factor, Θa, is the "extra length" or averageuplift factor, andΘm is the midwell uplift factor. Forfully penetrating wells,
(5-18)
(5-19)
Approximate solutions for the well factors for variouswell penetrations were developed by Bennett and Barron(1957). More theoretically exact solutions weredeveloped by Barron (1982) and verified by electricalanalogy tests. The theoretical results are shown inTable 5-1 and plotted in Figures 5-6 and 5-7 togetherwith the data from the electrical analogy tests. As thereis a linear relation between the well factors and loga/rw
for values ofa/rw greater than about 20, the well factorsare shown in terms of values ata/rw = 100. The wellfactors at any other value ofa/rw are given by thefollowing equations:
(5-20)
(5-21)
where ∆Θ is obtained from Table 5-1. Values of thewell factors may also be obtained from the nomographfrom EM 1110-2-1901 shown in Figure 5-8 (after Ben-nett and Barron 1957). The nomograph though basedon approximate solutions, is reasonably accurate for wellpenetrations greater than 25 percent. A computerprogram for well design based on the Figure 5-8 wasdeveloped by Conroy (1984).
5-9. Infinite Line of Wells, Impervious Top Stra-tum of Finite Length
In many instances, the impervious top stratum landwardof a line of wells is of finite length, and the boundaryedge can be considered as a line sink. The presence ofexposed borrow pits or other seepage exits landward of
the well line can be simulated by a line sink. The headdistribution beneath the top stratum without wells varieslinearly from 100 percent of the net head at the effectivesource of seepage to 0 percent at the line sink. Theconditions are illustrated in Figure 5-4 (b). Theseconditions are also applicable to the case of asemipervious landside blanket after conversion to anequivalent length of impervious blanketx3 . Equationsfor the head midway between wells and well flows areshown in Figure 5-9. The equations are applicable toboth fully penetrating and partially penetrating wellsystems. The equations in Figure 5-9 apply to the caseof no well losses. If well losses are considered, sub-stituteh for H as shown in Figure 5-4 (b).
5-10. Infinite Line of Wells, Impervious TopStratum Extending to Blocked Exit
Pervious foundations seldom extend landward to a greatdistance. Blockades generally occur because of thepresence of old clay-filled channels or upland forma-tions. If the distance from the line of wells is large,then the approximation of an infinite landward extent isreasonable. If the distance from the line of wells is lessthan the well spacing, then the error due to theapproximation may be significant. The equations for thehead midway between wells and well flow are shown inFigure 5-10 with exact equations for the case of fullypenetrating wells and reasonably accurate equations forboth fully and partially penetrating wells where the dis-tance to the blocked exit is greater than one-half timesthe well spacing. The presence of a blocked exit can beignored if the equivalent length of landside impervioustop stratum is less thanLB .
5-11. Infinite Line of Wells, Discharge BelowGround Surface
In many well installations, the well outlets are locatedbelow the ground surface to prevent any seepageupward through the top stratum. Under this condition,the blanket formulas are inapplicable and the top stra-tum is assumed to be impervious. Solutions areobtained using equations in Figure 5-5, withhd at orbelow ground surface, assuming∆hd = Hav .
5-12. Infinite Line of Wells, No Top Stratum
A special case may exist in which there is no landsidetop stratum and wells are needed to lower the headsbelow the landward ground surface. The flow in thiscase is a combination of artesian and gravity flow, and
5-8
EM 1110-2-191429 May 92
Table 5-1Theoretical Values of Θa and Θm
W/D D/a a/rw Θa Θm ∆Θ
100% All values 100 0.440 0.550 1.00
75% 0.250.501.02.03.04.0
100 0.5230.5630.6060.6780.7480.818
0.6330.6670.6810.6820.6820.682
0.489
50% 0.250.401.02.03.04.0
100 0.7420.8570.9831.1751.3611.547
0.8510.9551.0121.0241.0241.024
0.733
25% 0.250.501.02.03.04.0
100 1.2251.5691.9262.3902.7983.199
1.3351.6221.9082.0242.0472.075
1.466
15% 0.250.501.02.04.0
100 1.6622.3102.9703.7474.941
1.7722.4012.9383.2933.432
2.077
10% 0.250.501.02.04.0
100 1.9082.9343.9775.1396.814
2.0183.0253.9414.6495.071
3.298
5% 0.250.501.02.04.0
100 1.7783.8796.0638.377
11.144
1.8873.9696.0217.8649.283
6.963
the equations shown in Figure 5-11 (Johnson 1947) maybe used to estimate heads midway between wells andwell flows for design.
5-13. Finite Well Lines, Infinite Line Source
The essential difference between finite and infinite welllines is the presence or absence of an appreciablecomponent of flow parallel to the line of wells, resultingin nonuniform distribution of heads midway betweenwells and well discharges.
a. Impervious top stratum.Where the landside topstratum is impervious and extends landward to infinity,
the solution for a linear array of equispaced wellsparallel to an infinite line source can be obtained usingthe equations shown in Figure 5-3.
b. Impervious top stratum of finite length.In thecase of an impervious top stratum extending to a finitedistance landward of the well line or in the case of asemipervious landside top stratum converted to anequivalent length of impervious top stratum, theoreticalsolutions for finite well lines are not available. Empiri-cal solutions based on electrical analogy tests arepresented in EM 1110-2-1905. The application of thesesolutions for design is discussed in Chapter 7.
5-9
EM 1110-2-191429 May 92
Figure 5-6. Theoretical values of average uplift factor (after Barron 1982)
5-10
EM 1110-2-191429 May 92
Figure 5-7. Theoretical values of midwell uplift factor (after Barron 1982)
5-11
0 0
!I lt ~ ...... 0
.... <(
E "'
10.0 9.0 f-
1 LEGEND '
8.0 1- THEORY 7.0 f---&- ELECTRICAL v--=-6.0 1--
5.0
4.0
3.0
2.0
1.0 0.9 0.8 0.7
0.6
0.5
0.4
0.3
0.2 0.2
r:r
:) v ~
G-c;r-
~
ANALOGY TESTS ~l
_... -- :: v /./'
v v /'
~ ::;:: f:.--" 1-" £
/v .,........ ~ ~ v . ./"'
~ ~-r:.---' ......
-
0.4 0.6 0.8 1.0
D/o
' I _... f W/0=57. ___..... '-- I
'
lit - __ ., :1----1
~ 1--
W/D 15%
)
"'"""" --; . ..f )W/0=257.
W(D=1so:;
-- --- I -L - --·--1 --! W/0=75%
~ :! I I I 1--- -W/D~1DO%
2.0 4.0 6.0 8.0 10.0
EM 1110-2-191429 May 92
5-12
e 0 ~ 0 > -e ~
w a
" "' % 0
" ., ~
~
~ 0
~ "' 0 w z ... ~ z " "' " "-... ?
~ " ~ 0
~ - z ij 0
,;; - ;.; .. ~ "' "' ~ 15 L L L L L L L
J !); ~ !!! ... 1:' 1:' "' ~ " ~ < X w
we 'i "' <l
C> ~ C> "' 0 - - N ...i ,.; 0 z
O> i\
0 s
00~ 0 (H ~ 1\ • L
' ....._ D ooz
\ 0 oz 0
00~ 0 O> ocv 0 Ot oo,; 0 OS
000 ~ ~~,~~\ ~\ ~ ~ .~,
0 001 ~ .. , -"c>'
% '(0/M) NOI1VHBN3d 113M lN30~3d
EM 1110-2-191429 May 92
Figure 5-9. Infinite line of wells parallel to infinite line source, impervious top stratum of finite length
5-13
EM 1110-2-191429 May 92
Figure 5-10. Infinite line of wells with blocked landside exit
5-14
EM 1110-2-191429 May 92
Figure 5-11. Infinite line of fully penetrating wells, combined gravity, and artesian flow
5-15
"" LINE
0 a
0
H1 I 0 I
I ..iL_ 0 a I
0 I
s I
"a 00 s
SECTION
Gravity llow, 0 11 • arta•ian f!ow Q•
0" • 1tk (h! - h~ : a • • k (H1 - hi/ aD !!!n;xg Xa
In ae <f
~
(5-31)
(5-32)
r '""") where ~ • !E. In ,.,-.-nx8 ~
(5-33)
hp = height of saturation al any point in gravity flow zone
' (h2 _ h2\ cos~ (x • x;, -cos 2ffy
h - g ~ lnl---~sc------~a~ g 2us cosh 2rr (x ~ x .. ) - cos 21iy
-.- a a In ae ~
(5-34}
llH m .:. excess head above the wetl outlet midway between w-ells
(5-35)
EM 1110-2-191429 May 92
Chapter 6Well Design
6-1. Description of Well
While the specific materials used in the constructionvary and the dimensions and methods of installationsdiffer, relief wells are basically very similar. Theyconsist of a drilled hole to facilitate the installation; ascreen or slotted pipe section to allow entrance ofground water; a bottom plate; a filter to prevent entranceand ultimate loss of foundation material; a riser to con-duct the water to the ground surface; a check valve toallow escape of water and prevent backflooding andentrance of foreign material; backfill to prevent rechargeof the formation by surface water; and a cover and sometype of barricade protection to prevent vandalism anddamage to the top of the well by maintenance crews,livestock, etc. Figure 2-1 shows a typical relief wellinstallation. The hole is drilled large enough to providea minimum thickness of 4 to 6 in. depending on thegradation of the filter material as subsequentlydescribed. The hole is also overdrilled in depth to pro-vide for the fact that initial placements of filter materialmay be segregated. The amount of overdrilling requiredis variable depending upon the size of tremie pipe usedfor filter placement, the total depth of the well, andmost importantly on the tendency of the selected filtermaterial to segregate. The backfill indicated as sand inFigure 2-1 normally consist of concrete sand or other-wise excess filter material. Its only function is to fillthe annular space around the riser pipe to prevent col-lapse of the boring; these granular materials are easilyplaced and require a minimum of compaction. Thebackfill indicated as concrete in Figure 2-1 forms a sealto prevent inflow of surface water from rains andflooding.
6-2. Materials for Wells
Commercially available well screens and riser pipes arefabricated from a variety of materials such as black iron,galvanized iron, stainless steel, brass, bronze, fiberglass,polyvinyl chloride (PVC), and other materials. Howwell a material performs with time depends upon itsstrength, resistance to damage by servicing operations,and resistance to attack by the chemical constituents ofthe ground water. Wood has proven to be very stable inmost environments in well installations, as long as it iscontinuously submerged in water; however wood wellscreens and risers are no longer commercially available.
Stainless steel is apparently a very stable material inmost environments; however it is relatively expensive.Type 304 stainless steel has excellent corrosion resis-tance; whereas Type 403 stainless steel has moderatecorrosion resistance. Low-carbon or other-type steelwire-wrapped screen may be more economical in manyinstances; however it has no corrosion resistance. Brassand bronze are extremely expensive and are not com-pletely stable in some acid environments. Fiberglass isa promising material; however its performance history isrelatively short. PVC appears to be completely stable,and it is easy to handle and install; however it is a rela-tively weak material and easily damaged. The life ofiron screens is extended by galvanizing, which may notprovide permanent protection. Ferrous and nonferrousmetals should never be placed in direct contact witheach other, such as the case of a brass screen and a steelriser; the direct contact of these dissimilar metals mayinduce electrolysis and a resultant deterioration of thematerial.
6-3. Selection of Materials
Since pressure relief wells are designed and installed toprotect the foundations of structures, selection of mate-rials for the well should be based on costs and perfor-mance over the life of the structure which it protects.Generally, the choice of well screen material willdepend on three factors: (a) water quality, (b) potentialpresence of iron bacteria, and (c) strength requirements.A water quality analysis will determine the chemicalnature of the ground water and indicate whether it iscorrosive and/or incrusting (see Table 3-1).Enlargement of screen openings due to corrosion cancause progressive movement of fines into the well,therefore it is essential that the well screen be fabricatedfrom corrosion-resistant material where corrosive watersare expected. Similarly, if incrusting ground water isexpected, future maintenance which may require acidtreatments as described in Chapter 12 necessitates theuse of material that can withstand the corrosive effect ofthe treatments. When the presence of iron bacteria isanticipated, the well screen should be selected whichcan withstand the damaging effects of the repeatedchemical treatments described in Chapter 12. Thestrength of the well screen is usually not a major factorwhen commercial well screens designed for deeper wellinstallations are employed. The screen sections shouldbe able to withstand maximum compression and tensileforces during installation operations as well as horizon-tal forces which may develop during installation andpossibly later because of lateral earth movements.
6-1
EM 1110-2-191429 May 92
6-4. Well Screen
a. Slot type. A variety of slot types are available inmost types of well screens. PVC screens with openslots of varying dimensions consisting of a series of sawcuts are typically available. Metal and fiberglassscreens are available with open slots, louvered or other-wise shielded slots, or "continuous slots." The"continuous slot" screens consist of a skeleton of verti-cal rods wrapped with a continuous spiral of wire. Thewire can be a variety of cross-sectional shapes. Thetrapezoidal-shape wire provides a slot that is progres-sively larger toward the inside of the screen. This shapeallows any filter gravel that enters the slot to fall intothe well rather than clog the screen. The open-typeslots are advantageous in developing the filter. Theyallow the successful use of water jets; whereas shieldedslots deflect the water jet and reduce or destroy itseffectiveness in the filter. Machine cut slots typicallyhave jagged edges which facilitate the attachment ofiron bacteria making screens difficult to treat later.Continuous slot screens are commercially fabricated ofType 304 and 316 stainless steel, monel, galvanized orungalvanized low-carbon steel, and thermoplasticmaterials, mainly PVC and ABS or alloys of thesematerials. Couplings and the bottom plate for the wellscreen may be either glued, threaded, or welded andshould be constructed of the same material as the wellscreen.
b. Dimensions.The size of the individual openingsin a well screen is dictated by the grain size of thefilter. The openings should be as wide as possible, yetsufficiently small to minimize entrance of filtermaterials. Criteria for selection of screen opening sizeare presented subsequently. The anticipated maximumflow of the well dictates both the minimum total open-slot area of the screen (the spacing and length of slots)and the minimum diameter of the well. The open areaof a well screen should be sufficiently large to maintaina low entrance velocity of less than 0.1 ft per second(fps) at the design flow. Representative areas and maxi-mum well capacities for various well diameters withdifferent continuous slot sizes are shown in Table 6-1.Well screen manufacturers should be consulted for morespecific information. The well diameter must be largeenough to conduct the maximum anticipated flow to theground surface and facilitate testing and servicing of thewell after installation. Head loss in the well should alsobe taken into consideration in selecting a well diameter.
6-5. Filter
a. In order to prevent infiltration of foundationsands into the filter, the filter gradation must meet thestability requirement that the 15 percent size of the filtershould be not greater than five times the 85 percent sizeof the foundation materials. As shown in Figure 6-1,the design should be based on the finest gradation of thefoundation materials, excluding zones of unusually finematerials where blank screen sections should be pro-vided. If the foundation consists of strata with differentgrain size bands, different filter gradations should bedesigned for each band. Each filter gradation must alsomeet the permeability criterion that the 15 percent sizeof the filter should be more than three to five times the15 percent size of foundation sands. Either well gradedor uniform filter materials may be used. A uniformfilter material has a coefficient of uniformity,Cu, of lessthan 2.5 whereCu is defined as
(6-1)Cu
D60
D10
where
D60 = grain size at which 60 percent by weight isfiner
D10 = grain size at which 10 percent by weight isfiner
The Cu of well-graded filter materials should be greaterthan 2.5 and less than 6 to minimize segregation. Thegrain sizes should be reasonably well distributed overthe specified range with no sizes missing. Well-gradedfilter materials used with proper well development pro-cedures increase efficiency and permit the use of largescreen openings; however they are subject to segregationduring handling and placement. Well-graded filtersshould have an annular thickness of 6 to 8 in. Uni-formly graded filters permit a lesser annular thickness offilter (4 to 6 in.) and are not subject to segregation,thereby reducing the amount of overdrilling.
b. The filter should consist of natural material madeup of hard durable particles. It should contain nodetrimental quantities of organic matter or soft, friable,
6-2
EM 1110-2-191429 May 92
Table 6-1Properties of Wire-wrapped Continuous Slot Screens(Manufactured by Johnson Division, SES Inc.)
Shipping Weight Intake Areas (square inches per foot of screen)
Nom.Lb/Ft Diam.
Slot Opening Size
10-slot 20-slot 40-slot 60-slot 80-slot 100-slot 150-slot 250-slot
43 15 26 41 52 59 65 73 82
5 3 1/2 18 31 49 61 70 77 88 99
6 4 20 35 57 71 81 88 101 115
6 4 1/2 23 40 64 80 92 100 114 129
7 5 26 45 72 90 102 112 112 132
8 5 5/8 28 49 79 99 113 123 141 159
10 6 30 53 85 106 100 112 132 156
15 8 28 51 87 113 133 149 160 194
19 10 36 65 108 141 166 186 200 243
22 12 42 77 130 143 171 195 237 265
35 14 37 68 97 132 161 185 232 292
41 16 42 60 108 148 180 208 261 327
47 18 36 69 124 169 206 237 298 375
57 20 41 77 139 189 229 264 280 366
71 24 61 113 131 182 226 265 343 449
72 26 63 118 138 191 237 278 360 471
81 30 75 138 161 224 278 325 422 552
91 36 84 157 184 255 317 371 481 629
Notes:1. Open areas may differ somewhat from these figures. Extra-strong construction, for example, reduces open areas in some cases
because heavier material is used to increase screen strength.2. The maximum transmitting capacity of the screen can be derived from these figures. To determine gpm per ft of screen, multiply the
intake area in square inches by 0.31. It must be remembered that this is the maximum capacity of the screen under ideal conditionswith an entrance velocity of 0.1 fps.
thin, or elongated particles. Crushed carbonateaggregates should be avoided because they tend to breakdown with a loss in permeability. Furthermore, theywill tend to dissolve if the wells require future acidtreatment as part of future rehabilitation operations. Itis often difficult to purchase material that meets therequired gradation, and it may be necessary to have thematerial specially blended. The special blends areexpensive and sometimes difficult to acquire, but
essential to the installation of acceptable permanentrelief wells.
6-6. Selection of Screen Opening Size
In general, the slot width (or hole diameter) of thescreen should be equal to or less than the 50 percentsize of the finest gradation of filter. Application of thiscriterion is demonstrated in Figure 6-1. Use of the
6-3
EM 1110-2-191429 May 92
Figure 6-1. Typical design of filter for relief well
50 percent size criterion for the selection of screen slotsize appears to provide reasonable assurance against in-wash of filter materials during well development andsurging and furthermore results in suitably largeopenings to minimize the effects of incrustations andblockages which may develop during the life of the well(Hadj-Hamou, Tavassoli, and Sherman 1990).
6-7. Well Losses
a. Head losses within the system consist of entrancehead loss in the screen and filter (He) plus friction headlosses arising from flow in the screen, riser, and
connections (Hf) plus velocity head loss (Hv). The totalhydraulic head loss in a well (Hw) is given by
(6-2)Hw He Hf Hv
b. The entrance losses in the screen and filter for aproperly designed and developed screen and filter willgenerally be relatively small at the time of wellinstallation. Installation techniques resulting in smear orundue disturbance of the drill hole walls, however, canresult in relatively large initial entrance losses. Entrance
6-4
EM 1110-2-191429 May 92
losses can be expected to increase with time for avariety of reasons discussed in Chapter 11. For exam-ple, as shown in Figure 6-2, the entrance losses for 8-in.-ID slotted wood well screens, based on piezometerdata at the time of installation, amounted only to about0.10 to 0.25 ft for a flow through the screen of 10 gpmper foot of screen. However, as shown in Figure 6-2,entrance losses for the particular wells increased signifi-cantly with time. The initial entrance losses for wire-wrapped screens should be even less. Both field andlaboratory tests indicate that the average entrancevelocity of water moving into the screen should notexceed 0.1 fps. At this velocity, friction losses in thescreen openings will be negligible and the rates ofincrustation and corrosion will be minimal. The averageentrance velocity is calculated by dividing estimatedwell yield by the total area of the screen openings. Ifthe velocity is greater than 0.1 fps, the screen lengthand/or diameter should be increased accordingly. Thelong-term value of entrance loss is difficult to predict,and unless experience in a specific location is available,conservative values based on Figure 6-2 should beselected.
c. Friction losses in the screen and riser sectionsmay be estimated from Figure 6-3. The head loss in thescreen section should be computed for a distance ofone-half the screen length. More accurately, frictionlosses can be calculated according to the Darcy-Weisbach formula as described in EM 1110-2-1602.
The resistance coefficient in the formula is solved bythe Colebrook-White equation also given in EM 1110-2-1602. This equation requires the input of an effectiveroughness parameter for the material comprising thewell screen and riser pipe. A computer code for thesolution of the Colebrook-White equation is given inUSAEWES (1973).
d. Velocity head losses,Hv, should be computed bymeans of the equation
(6-3)Hv
υ2
2g
where
υ = the velocity of the water in the riser pipe
g = acceleration due to gravity = 32.2 ft/sec2
Losses due to elbow connections should be includedwhere applicable.
6-8. Effective Well Radius
The effective well radius to be used in design computa-tions is calculated as the outside radius of the wellscreen plus one-half the thickness of the filter.
6-5
EM 1110-2-191429 May 92
Figure 6-2. Entrance losses versus inflow for 8-in.-ID slotted wood well screens in St. Louis District (afterMontgomery 1972)
6-6
EM 1110-2-191429 May 92
Figure 6-3. Friction head losses in screen and riser sections
6-7
0.2 0.4 0.6 2 3 4 5 10
8
6
4
3
"' z 0 u 2 w
"' "'
/
,_s / , .... ;:::,. v v
o~" '/;.-/ ?&?' L
"-"'""'"" 0~ / v <;)~.,... v/ v v
, .... ¥-'-"- v v / ...- / v
w a. .... w w u.. u 0.8 a; ::::> '-' 0.6 w <:>
-v y,s""L v v v / /
~ / / ,o ,.,- / ./ / / I'" v ,. r:i·~ /
'0~ v -
"' """ I 0.4 u IJ)
Cl
...J
...J w
0.2 :;:: l'
/ / v
/
c."~ / .. ~
v /
0
0.1 Hf =
303 v 1.85 FT/100 FT OF PIPE c 1.85 d 1. 167
HAZEN - WILLIAMS. C = 1 00 0.08 FOR OTHER VALUES OF C MULTIPLY
FRICTION VALUES BY (1 00/C) 1.85 0.06 '
0.04 0.1 0.2 0.4 0.6 2 3 4 5
Hf FRICTIONAL HEAD LOSS (rT/100 FT OF PIPE)
EM 1110-2-191429 May 92
Chapter 7Design Of Well Systems
7-1. General Approach
The design of relief well systems consists essentially ofdetermining the location and penetration of wells thatwill reduce the piezometric surface of the substratumpressure,ho, in landside or downstream areas to anallowable head,ha. Analyses are made using formulaspresented in Chapter 4 and 5. Where wells are requiredalong the toe of a levee or dam, the wells will generallybe located along a line so that their locations are definedby a well spacing. The well spacing is first determinedassuming an infinitely long line of wells, and then thespacing is reduced where necessary to allow for thereduced efficiency of a finite number of wells comparedto the infinite number. For given boundary conditionsand the same allowable head, there are any number ofcombinations of well spacing and penetration that willsuffice. The final selected spacing and penetrationshould be based to a great extent on the most economi-cal design. The presence of natural topographic featuresmay require adjustment in the design well spacing toensure that well outlets are located at the lowest practi-cal elevation.
7-2. Design Heads
The design of relief well systems for dams are based onsteady state conditions which would prevail with thereservoir pool at the maximum design level. Thisreservoir pool normally is taken as the top of thesurcharge pool. The design net head is the differencebetween the latter elevation and downstream tailwaterelevation, usually taken as downstream ground surfaceor lower, if appropriate. In the case of relief welldesign for levees, the design net head is usually taken asthe difference in elevation between net grade of thelevee and tailwater.
7-3. Boundary Conditions
Boundary conditions which must be determined includethe distance to the effective source of seepage entry,S;the distance from the line of relief wells to the effectiveseepage exit,x3; and the distance to a blocked exit,LB,if such exists. Procedures for the determination of thesevalues are given in Appendix B.
7-4. Design Procedures
Direct application of the formulas in Chapter 4 and 5 isnot possible as they are based on the assumption thatthe hydraulic head losses in the well are zero. Asshown in Figure 5-4, the head losses must be deter-mined on the basis of the computed well flow andadded to the maximum landside head with wells, whichin turn would result in a lower factor of safety withrespect to uplift. If the tops of the well risers extendabove tailwater, the difference in elevation should beadded to the well losses in determining the maximumlandside head with wells. The maximum landside headwill always occur midway between wells for fully pene-trating wells. For partially penetrating wells, there maybe a difference as the average head may exceed thehead midway between wells. To maintain the requiredfactor of safety, a reduction in well spacing is requiredso as to lower the maximum landside head with wells.Thus, an iterative procedure must be utilized to find thewell spacing which satisfies the condition that whenwell losses are considered, the head midway betweenwells or the average head, whichever is greater, equalsthe design values. Procedures involving this concept arepresented below.
7-5. Infinite Line of Wells, Impervious TopStratum
The general procedure for designing a system of reliefwells along an infinite line with an impervious topstratum extending to a great distance landward follows.The procedure is valid for both fully and partially pene-trating well systems.
a. Compute the allowable head,ha, under the topstratum at the downstream toe of the dam or levee fromEquation 3-2. Assume tailwater elevation coincideswith ground surface (or as appropriate).
b. Assume thatHm, the net head midway betweenwells, is equal toha and that well losses,Hw, are equalto zero (see Figure 5-4a).
c. For a given well penetration,W, computeHm, forvarious trial values of well spacing,a, based on Equa-tion 5-14. Interpolate to determine the required wellspacing forHm = ha.
d. Calculate the well flow,Qw, for the above wellspacing and penetration using Equation 5-17.
7-1
EM 1110-2-191429 May 92
e. Assume the well dimensions, and calculate thewell losses,Hw, corresponding toQw.
f. Repeat step c usinghm = Hm - Hw in place ofHm in Equation 5-14, and determine a new value ofa.
g. Repeat steps (d) through (f) until relativelyconsistent values ofa are obtained on two successivetrials. The value of a derived in this manner satisfiesthe design requirement for fully penetrating wells.
h. If the wells are not fully penetrating, repeat step(b) assuming thatHav, the average net head, is equal toha.
i. For a given well penetration,W, computeHav forvarious trial values of well spacing,a, based on Equa-tion 5-15. Interpolate to determine required wellspacing forHav = ha.
j. Calculate the well flow,Qw, for the above wellspacing and penetration using Equation 5-17.
k. Assume the well dimensions and calculate thewell losses,Hw, corresponding toQw.
l. Repeat step (i) usinghav = Hav - Hw in place ofHav in Equation 5-15 and determine a new value ofa.
m. Repeat steps (j) through (l) until relatively con-sistent values ofa are obtained on two successive trials.For design, select the lesser value of well spacingdetermined from steps (g) and (l) for a given wellpenetration.
n. Repeat for various well penetrations to develop arelation between well penetration and spacing thatsatisfies design requirements.
7-6. Infinite Line of Wells, Wells in Ditch
The design of an infinite line of wells with the well out-lets located in a collector ditch to lower landsideground-water levels below ground surface should bebased on the assumption that the top stratum isimpervious regardless of its permeability. The designprocedure is essentially similar to that described in thepreceding paragraph with the following exceptions:
a. Assume an elevation for the well outlets,hw, andcompute the allowable head,ha, under the top stratumbeneath the collector ditch from Equation 3-3. Proceedwith steps (b) through (g) to obtain a design well
spacing that satisfies Equation 3-3.
b. Select a design ground-water level landward ofthe wells defined byhd = hw + ∆D. Assumehd = Hav.
c. Proceed with step (i) usinghd in place ofHav.
d. Continue steps (j) through (n) to obtain the designwell spacing.
7-7. Infinite Line of Wells with Impervious TopStratum of Finite Length
The procedure for design of an infinite line of wellswith a landside impervious top stratum of finite lengthis presented below. The procedure is also applicable tothe case of a semipervious landside top stratum afterconversion to an equivalent length of impervious topstratum as discussed in Appendix B. The spacing for aninfinite line of relief wells for a given penetration isdetermined using an iterative procedure. For small wellspacings, the average uplift factorΘa will be equal to orlarger than Θm and will control. For large wellspacings,Θm will be equal to or larger thanΘa and willtherefore control. A summary of the equations used isshown on Figure 7-1. The procedure for computing thewell spacing for both conditions is as follows:
a. Compute the allowable head beneath the top stra-tum at the downstream toe of the dam,ha, fromEquation 3-2.
b. Assume that the net head in the plane of thewells, Hav, is equal toha and calculate the net seepagegradient toward the well line,∆M, substituting in Equa-tion 7-6 as follows:
(7-11)∆Mh ha
S
ha
x3
where
S = distance from effective seepage entry to line ofwells
x3 = distance from line of wells at the landside toe toeffective seepage exit (length of landsideimpervious top stratum)
c. Assume a well spacing and compute the flowfrom a single well using Equation 7-7.
7-2
EM 1110-2-191429 May 92
Figure 7-1. Nomenclature and formulas for design of relief wells at toe of dam
7-3
h H
s
(s + x3 ) h=H-H w X
3
h = hfl,
av s s .. x, r _ .. a x3
Hav = Hw + hav
9, h9,. hm "'"'hav-a a s s + x3
- + a x3
Hm = Hw + 11,
w
(H)
(7·2)
(7-3)
" (7-4)
(7-5)
HEAD WITHOUT WELLS
//// /' TOP STRATUM
/
SUBSTRATUM
/;M:
X 3
H -Hav
s -
Qw = al!.M kt D
orQ• hkrD
Hav
XJ
s s + x3 • +
a x3
hm = (1/;Mfjm
hav = al!.MBa
D
(7-6)
(7-7)
(7-8)
(7-9)
(7-11})
EM 1110-2-191429 May 92
d. Assume the well dimensions and calculate thewell losses,Hw, corresponding toQw.
e. Compute the net average head in the plane ofwells, hav, using Equation 7-3.
f. Substitute values of∆M and hav in steps (b) and(e) and solve forΘa using Equation 7-10.
g. Find Θa from Table 5-1 and Figure 5-6 orFigure 5-8 for the given well penetration using thevalues ofa used in step (f) and the correspondinga/rw
andD/a values.
h. The first trial well spacing is that of valuea forwhich Θa from step (f) equalsΘa from step (g).
i. Find Θm from Table 5-1 and Figure 5-7 orFigure 5-8 for the given well penetration and first trialwell spacing and the corresponding values ofa/rw andD/a values.
j. If Θa > Θm repeat steps (c) to (i) using the firsttrial well spacing in lieu of the spacing originally usedin step (c), and determine the second trail well spacing.This procedure should be repeated until relativelyconsistent values ofa are obtained on two successivetrials. Usually the second trial spacing is sufficientlyaccurate. If in step (j), Θa < Θm, a modified procedureis used for a second trial using steps (k) through (t).
k. AssumeHm = ha and computeQw from Equa-tion 7-7 using the value of∆M obtained in step (b) andthe first trial well spacing from (h).
l. EstimateHw from Qw of step (k).
m. Compute the net head midway between the wellsashm = Hm - Hw.
n. Using Θa and Θm from steps (h) and (i),respectively, computehav from Equation 7-4.
o. Using Hw and hav from steps (l) and (n),respectively, computeHav from Equation 7-3.
p. Compute∆M from Equation 7-6 usingHav fromstep (o).
q. Using hm and ∆M from steps (m) and (p),respectively, computeΘm for various values ofa fromEquation 7-9.
r. Find Θm from Table 5-1 and Figure 5-7 or Fig-ure 5-8 for the values ofa used in step (q) and thecorrespondinga/rw andD/a values.
s. The second trial well spacing is that value of awhich Θm from step (q) equalsΘm from step (r).
t. Find Θa from Figure 5-6 for the second trial wellspacing and the corresponding values ofa/rw andD/a.
u. Determine the third trial well spacing by repeatingsteps (k) to (t) using the second trial well spacing in lieuof the spacing originally assumed in step (k), and instep (n) using the valuesΘm and Θa from steps (s) and(t), respectively, instead of those from steps (h) and (i).This procedure should be repeated until relatively con-sistent values ofa are obtained on two successive trials.Normally, the third trial is sufficiently accurate.
v. Repeat steps (g) through (u) for various wellpenetrations to develop a relation between wellpenetration and spacings that satisfies designrequirements.
7-8. Computer Programs
A computer program for design of relief wells systemsbased on the above procedures was developed byConroy (1984). Comparisons of the computer and handsolutions are presented by Cunny, Agostinelli, andTaylor (1989).
7-9. Head Distribution for Finite Line of ReliefWells
In a short, finite line of relief wells, the heads midwaybetween wells exceed those for an infinite line of wellsboth at the center and near the ends of the well systemas shown in Figure 7-2. With an infinite line of wells,the heads midway between wells are constant along theentire length of the well line. Many well systems maybe fairly short; thus, it will be necessary to reduce thewell spacing computed for an infinite line of wells sothat heads midway between wells will not be more thanthe allowable head under the top stratum. The ratio ofthe head midway between wells at the center of finitesystems to the head between wells in an infinite line ofwells, for various well spacings and exit lengths, isgiven in Figure 7-3. The spacing of relief wells in afinite line should be the same as that required in aninfinite line of wells to reduce the head midwaybetween wells toha divided by the ratio ofHm /Hmn ∞
7-4
EM 1110-2-191429 May 92
from Figure 7-3. In any finite line of wells of constant
Figure 7-2. Variation of pressure relief along a finite line of relief wells (after EM 1110-2-1905)
penetration and spacing, the head midway between wellsnear the ends of the system exceeds that at the center ofa system. Thus, at the end of both short and long wellsystems, the relief wells should generally be madedeeper to provide additional penetration of the pervioussubstratum in order to obtain the same head reduction asin the central part of the well line. In the case of fullypenetrating wells, the same head reduction can beobtained by additional wells using gradually decreasingwell spacings near the ends of the line. The above-mentioned procedures for designing finite relief wellsystems, although approximate, are usually sufficient.
7-10. Well Systems at Outlet Works andSpillways
When well systems for outlet structures and spillwaystructures are being designed, the problem is to design agroup with a finite number of wells with proper spacingand penetration which will reduce the head at the centerof the well group to the allowable design head. In thistype problem, usually the pressure at the well is knownbecause such wells normally will be discharging intotailwater elevation, possibly through a collector systemeither under the stilling basin or along the channel-sideslopes. Thus, the head at the well is equal to tailwaterplus the hydraulic head loss in the well and collectorsystem. This elevation when subtracted from the reser-voir pool represents the net head acting on the system.For such a system of fully penetrating wells, the
equations using the method of images for fullypenetrating wells with a line source and imperviousdownstream top stratum are utilized, and the head at thecenter of the well group is calculated by superposition.
7-11. Well Costs
The design of relief well systems will normally producevarious combinations of well spacing and penetrationwhich satisfy the design criteria. The optimum designshould be based on initial cost as well as overall costsincluding maintenance and possible replacement costsover the life of the structure. Costs should be calculatedper 100-ft stationing as shown in Figure 7-4. Elementsincluded in the estimate of initial costs are the cost ofdrilling or other installation technique, as well as thecost of well screen, riser pipe, and filter, all of whichare on a foot basis. Additional fixed costs include back-filling, well development and testing, plus the costs ofwell guards, check valves, and horizontal outlet pipes ifused. As shown in Figure 7-4, the well spacing andscreen penetration should be selected that will result inthe minimum well cost per station over the life of thestructure.
7-12. Seepage Calculations
As previously noted, the presence of relief wells willtend to increase the total quantity of seepage beneath alevee or dam,Qs. The seepage per foot of structurewith no wells is computed by the equation
7-5
EM 1110-2-191429 May 92
Figure 7-3. Ratio of head midway between relief wells at center of a finite well system to head midwaybetween wells in an infinite system (after EM 1110-2-1905)
7-6
EM 1110-2-191429 May 92
Figure 7-4. Determination of optimum well design
7-7
EM 1110-2-191429 May 92
(7-12)Qs
kDHS x3
The seepage with wells is equal to the flow from thewell system ΣQw, plus the seepage beyond the wellsystem,Qsw , which is computed by the equation
(7-13)Qsw
kDHav
x3
where Hav is computed from equations in Figure 7-1.The estimation of the total quantity of seepage passingbeneath a structure with or without wells can be ofimportance in selecting a well spacing which will inter-cept the desired amount of seepage.
7-8
EM 1110-2-191429 May 92
Chapter 8Relief Well Installation
8-1. General Requirements
Proper installation of relief wells is essential to thesuccessful protection of the structures for which they aredesigned to protect. Before installation is begun, allmaterials required for completion of the installationshould be on hand at the worksite. The well screen andriser should be checked for proper material, length,diameter, and slot openings. The filter material shouldbe inspected and checked against gradation specifica-tions. Successful completion of a well installation isoften dependent upon time, and many installations havebeen aborted because of delays. An open boring ofsufficient size and depth is necessary to facilitate theinstallation of a well. The hole should be vertical sothat the screen and riser may be installed straight andplumb. As previously discussed, the hole is drilled largeenough to provide a minimum thickness of 4 to 6 in.,depending on the gradation, of the filter material. Themethods of providing an open boring in the ground arenumerous; however not all are acceptable for the instal-lation of permanent relief wells, and those consideredacceptable are discussed in the following paragraph.
8-2. Standard Rotary Method
One method of drilling for well installation which hasgained popularity in the well drilling industry is stan-dard rotary drilling using a biodegradable, organic dril-ling fluid additive. No bentonitic clays are used in thedrilling fluid. Standard rotary drilling consists of rotat-ing a cutter bit against the bottom of a boring, while afluid is pumped down through the drill pipe to cool andlubricate the bit and return the cuttings up the open holeto the ground surface. The required size of bit isgoverned by the screen diameter and the thickness offilter. The ability of the fluid to carry the cuttings isdependent on its velocity and viscosity. The velocity ofthe returning fluid is reduced with increased boringdiameter, and the reduction is compensated by increasedviscosity of the drilling fluid. One such drilling fluidadditive is marketed under the trade name "Revert," so-called because the fluid reverts to the viscosity of water,normally in about three days. Chemicals can be addedto speed up or delay the reversion of such fluids as"Revert." Ground-water temperatures may effect rever-sion times.
a. Equipment. A rotary-type drill rig of sufficienthoisting and torque capacity is required. The cutter ordrill bit can be of either drag or roller design. The drillpipe should be as large as practicable to increase thevolume of fluid at the drill bit and, consequently, thevelocity of the fluid returning up the open hole.
b. Problems. The reverting process of the drillingfluid leaves a small amount of slimy ash which,unavoidably, is mixed into the filter material; however alarge percentage of this ash is removed during develop-ment of the well. Testing to determine the extent ofdetriment caused by this ash residue has not beensufficient to evaluate the effectiveness of this method;however it has been used successfully in installation ofpermanent relief wells. Chemical development of thewell is required as subsequently described.
8-3. Reverse-Rotary Method
This method is generally considered to provide the mostacceptable drill hole and should be used whenever pos-sible for the installation of permanent relief wells. Inthe reverse-rotary method, the hole for the well is madeby rotary drilling, using a similar cutting process asemployed in standard rotary drilling except the drillingfluid is pulled up through the drill pipe by vacuum andthe drilling fluid reenters the top of the open boring bygravity. Soil from the drilling is removed from the holeby the flow of drilling fluid circulating from the groundsurface down the hole and back up the hollow drill stemfrom the bit. Since the cross-sectional area of the bor-ing is many times larger than that of the drill pipe, theslow downward velocity of the fluid acting against theopen boring does not erode the walls. The drilling fluidconsists of water and, unavoidably, a small amount ofthe finer fraction of the natural material being drilled.A high velocity is attained with the fluid returning upthrough the drill pipe, thus eliminating the need for ahigh viscosity. The drill water is circulated by a centrif-ugal or jet-eductor pump that pumps the flow from thedrill stem into a sump pit. As the hole is advanced, thesoil particles settle out in the sump pit, and the muddywater flows back into the drill hole through a ditch cutfrom the sump to the hole. The sides of the drill holeare stabilized by seepage forces acting against a thinfilm of fine-grained soil that forms on the wall of thehole. A sufficient seepage force to stabilize the hole isproduced by maintaining the water level in the hole atleast 7 ft above the natural water table. Figure 8-1shows schematically the circulating system for
8-1
EM 1110-2-191429 May 92
Figure 8-1. Schematic diagram of circulatory system (after EM 1110-2-1913)
reverse-rotary drilling. No bentonitic drilling mudshould be used because of gelling in the filter andaquifer adjacent to the well. If the hole is drilled inclean sands, some silt soil may need to be added to thedrilling water to attain the desired degree of muddiness(approximately 3,000 ppm). A biodegradable organicdrilling fluid additive such as "Revert" or equivalentmay also be added to the drilling water to reduce waterloss.
a. Equipment. Reverse-circulation rotary drillingrequires somewhat specialized equipment, most of whichis commercially available or easily fabricated. Anyrotary-type drill rig large enough to handle the load andhaving sufficient torque capability can be adapted tocirculate water through an eductor to create a vacuumon the drill pipe. Drill pipe and hoses should be of aconstant inside diameter throughout the system to assure
that material entering the system can be circulated com-pletely through it. In alluvial deposits, a drag-type bitsimilar to the cutter head for a dredge is sufficient.Roller-type bits are commercially available for use inconsolidated deposits. The eductor consists of a pipe Ywith a nozzle fitted into one end of the Y.
b. Problems. It is necessary to maintain an excesshydrostatic pressure on the drill hole to stabilize thewalls. In most materials, a minimum excess head of7 ft is required and greater is desirable. When the staticwater level is very near the ground surface or artesianconditions prevail, it may be necessary to elevate thedrilling rig on temporary berms. Some success has beenexperienced by lowering the water level with wellpoints, but if the pressure is derived from a deeper,artesian source, it is necessary to lower the pressure inthe aquifer with deep wells. Since the formation in
8-2
EM 1110-2-191429 May 92
which a well is installed consists predominately of gran-ular material, the loss of water into the formation pre-sents a problem during drilling. An almost unlimitedsupply of water can be necessary to maintain acompletely filled, open boring. A large sump is re-quired to supply adequate water. During the drilling, allcuttings from the boring are deposited in the sump andmust be provided for. A sump three times the antic-ipated volume of the completed boring is adequate, if itcan be kept filled with water from another source.Consideration should be given to the required thicknessof the natural impervious clay blanket when constructinga sump. An instantaneous loss of water resulting in lossof excess head can cause failure of the boring walls.Often, if the rotation of the drill bit is stopped, the waterloss is greatly reduced. The boring must be kept full ofwater until the well screen, riser, and filter are installed.
8-4. Bailing and Casing
In cases where standard or reverse-rotary drilling is notsuccessful, an equally acceptable method of drilling con-sists of bailing while driving a steel casing into the holeto stabilize the boring walls. This method is economicalin some materials, and it does not inject deleteriousmaterials into the formation. Loose to medium dense,clean, granular materials can be bailed economically.Often the granular materials are overlain with a cohesiveoverburden which does not yield easily to bailing, and itis more economical to auger through this overburden.
a. Equipment.A drill rig with a wire line hoist anddriving capability is adaptable to this method of wellinstallation. It should be remembered that large casing,heavy enough to sustain driving, presents a sizable loadto be handled by the drill rig. The use of a vibratorypile driver can greatly facilitate the driving and subse-quent removal of the casing. The casing should beflush-joint, or welded-joint steel pipe. Two types ofbailers are commonly used for this purpose (Figure 8-2).Details are given in EM 1110-2-1907. The bailer isoperated on a wire line by lowering to the bottom of theboring and quickly pulling, or snatching, up a shortdistance a number of times to fill the bailer.
b. Problems. This method of drilling producesgood results but often presents problems in operations.Thin layers of cohesive materials, or cemented materialswithin the formation, can preclude the advance by
bailing and may also produce smear along the sides ofthe drill hole which could impair free flow into the well.Penetration of the casing can be retarded by friction ofthe granular formation against the outside of the casingunless vibratory hammers are used. After the casing isset, the boring completed, and the well installed, thecasing is removed. The casing should be pulled, as thefilter material is placed, to prevent disturbing the wellinstallation by the friction of the filter material insidethe casing. Using a vibratory pile hammer to drive andextract casing can densify loose foundation materialsand filter materials. Generally, when material isdensified, the hydraulic conductivity is reduced. Thevibratory hammer cannot be used in wells that havemore than one filter pack. As densification in the filterpack occurs, the material settles. This settlement, com-bined with settlement which occurs as the filter fills thevoid left by removal of the casing, results in uncertain-ties regarding the final position of the top of the filter.There are many uncertainties associated with thismethod of installation which makes it very difficult toestimate time and costs.
8-5. Bucket Augers
Under certain conditions drill holes for relief wells canbe made with a bucket auger. The method has beensuccessfully employed where cobbles up to 10 in. havebeen encountered. A bucket with side cutters isemployed, and only water is used as the drilling fluid.The rate at which the bucket is inserted or withdrawnmust be carefully controlled; thus close inspection isobligatory. A steel casing is installed through the topstratum to prevent smearing of fine-grained materials onthe walls of the drill hole.
8-6. Disinfection
Before drilling begins, all tools, rods, bits, and pumpsshould be thoroughly washed with a chlorine solution tokill any bacteria remaining from previous well installa-tions. Water used in the drilling process and filtermaterials should also be treated with a chlorine solution(Driscoll 1986). The strength of the chlorine solutionshould not be less than 100 ppm, which means a pro-portion of 100 lb of chlorine to 1 million lb of water.Calcium hypochlorite which contains 65 percent avail-able chlorine is commonly used for this purpose. Therequired weight (wt) of calcium hypochlorite to produce
8-3
EM 1110-2-191429 May 92
Figure 8-2. Bailer and sand pump assemblies (after EM 1110-2-1907)
8-4
5'
1/4"
BAlLER ASSEMBLY
WELD
SAND PUMP
EM 1110-2-191429 May 92
a given strength in N gallons (gal) of water is given bythe equation
(8-1)
Wt (lb) N (gal) × 8.33
× solution strengthavailable chlorine
where both solution strength and available chlorine areexpressed as a decimal. Thus, for a chlorine solution of100 ppm in 1,000 gal of water, using calcium hypochlo-rite with 65 percent chlorine, the required weight ofcalcium hypochlorite is
Similarly, for chlorine products such as sodiumhypochlorite which is available in gallons, the requiredvolume (V) to produce a given strength in N gal ofwater is given by the equation
(8-2)
Thus, for a chlorine solution of 100 ppm in 1,000 gal ofwater using sodium hypochlorite with 10 percent avail-able chlorine, the required volume of sodium hypochlo-rite is
V (gal) 1,000 (gal) × 0.00010.10
1.0 gal
8-7. Installation of Well Screen and RiserPipes
Once the boring is completed and the tools withdrawn,the boring should be sounded to assure an open hole tothe proper depth. The well screen and riser pipe can befabricated at the factory in varying lengths. Thecontractor will determine these lengths based on thecapacity of his equipment. The bottom joint of the wellscreen should be fitted with a cap or plug to seal thebottom of the screen. The lengths of screen areconnected together as they are lowered into the hole.Each length must be measured to determine its total
made-up length, and the bottom of the screen should beset at the designed depth, or as field conditions require.The method of connecting the lengths of screen andriser vary: metal screen and riser have threaded orwelded joints; plastic and fiberglass screens usually haveeither mechanical or glued joints. Each joint should bemade up securely to prevent separation of the wellduring installation and servicing activities. Each jointshould be kept as straight as possible to facilitate easeof servicing and testing. The riser and screen sectionsof the well should be centered in the drill hole by meansof appropriate centering devices to facilitate a continu-ous filter around the well screen. If materialsappreciably finer than anticipated in design are encoun-tered, design personnel should be notified. In suchcases, it may be necessary to replace the screen by asolid pipe or blank screen to prevent piping of founda-tion materials into the well. Immediately after instal-lation of the well screen and riser, the total inside depthshould be sounded. The exact inside depth of the wellmust be known to determine whether damage occursduring development and servicing of the well.
8-8. Filter Placement
Caution in proper design, control of manufacture, andhandling of filter materials to the jobsite can becompletely negated by improper placement in the well.Acceptable construction of permanent relief wellsdemands that the filter be placed without segregationbecause widely graded filters when placed in incrementstend to segrate as they pass through water, with coarseparticles falling faster than fine particles. A tremieshould be used to maintain a continuous flow ofmaterial and thus minimize segregation during place-ment. A properly designed, uniform (D90/D10 < 3 to 4)filter sand may be placed without tremieing if it ispoured in around the screen in a heavy continuousstream to minimize segregation. The tremie pipe shouldbe at least 2 in. in diameter, be perforated with slots1/16 to 3/32 in. wide and about 6 in. long, and haveflush screw joints. The slots allow the filter material tobecome saturated, thereby breaking the surface tensionand preventing "bulking" of the filter in the tremie. Oneor two slots per linear foot of tremie is generally suffi-cient. To avoid contamination by iron bacteria, thefilter should be washed through the tremie pipe using a100-ppm chlorine solution. The tremie pipe is loweredto the bottom of the open drill hole, outside the wellscreen and riser pipe. The presence of centering deviceswill interfere with the proper use of the tremie by pre-venting uniform filling to some extent. The use of dualdiametrically opposed tremie pipes will ensure more
8-5
EM 1110-2-191429 May 92
uniform placement. After the tremie pipe or pipes havebeen lowered to the bottom of the hole, they should befilled with filter material and then slowly raised to keepthem full of filter material at all times. Extending thefilter material at least 2 ft above the top of the screenwill depend on the depth of the well to compensate forsettlement during well development. The top of thefilter should also terminate below the bottom of theoverlying top stratum if present. The level of drillingfluid or water in a reverse-rotary drilled hole must bemaintained at least 7 ft above the natural ground-waterlevel until all the filter material is placed. If a casing isused, it should be pulled as the filter material is placed,and the bottom of the casing kept 2 to 10 ft below thetop of the filter material.
8-9. Development
A well is at best inefficient until properly developed.Development procedures include both chemical andmechanical processes. Development of a well should beaccomplished as soon after the hole has been drilled aspracticable. Delay in doing this procedure may preventa well being developed to the efficiency assumed indesign.
8-10. Chemical Development
Chemical development is applied usually in the casewhere special drilling fluids are utilized and chemicalsare injected into the well to aid in the dissolution of theresidual drilling fluid in the filter. The chemicals shouldbe of a type and concentration recommended by themanufacturer of the drilling fluid. They should beplaced starting at the bottom of the well and dispersedthroughout the entire screen length by slowly raisingand lowering the injection pipe. After the chemicalshave been dispersed, the well should be pumped and theeffluent checked to ensure that the drilling fluid hascompletely broken down.
8-11. Mechanical Development
The purpose of mechanical development is to removeany film of silt from the walls of the drilled hole and todevelop the filter immediately adjacent to the screen topermit an easy flow of water into the well. The resultof proper development is the grading of the filter fromcoarsest to finest extending from the well. The effect ofproper development is an increase in the effective sizeof the well, a reduction of entrance losses into the well,and an increase in the efficiency of the well. Manyfactors, including but not limited to development
methods, well design, and filter installation, affect thetime it takes to fully develop a well. Basically there arethree methods used in development as discussed below.
a. Water Jetting. A water jet, consisting of a seriesof small nozzles at the end of a pipe, lowered into thewell screen, is very effective in developing thecontinuous slot-type, wire-wrapped screens. A typicalwater jet is shown in Figure 8-3. Water is pumpeddown and out through the nozzles at a high velocity.Nozzles are directed toward the screen slots in smallconcentrated areas, as shown in Figure 8-4. The waterjet equipment can be fabricated in local welding shops.The size and number of nozzles must be consistent withthe size and length of the pipe through which the wateris pumped to ensure a high-pressure and high-velocityjetting action. This method requires a high-pressure,relatively high-volume water pump. The lowest effec-tive nozzle velocity for water jetting is about 100 fps.Better results are obtained with nozzle velocities be-tween 150 and 300 fps. Normally, development with awater jet is started at the bottom of the screen. Jettingis accomplished at one depth with the jet rotated for afixed period of time. The jet is raised approximately0.5 ft; rotation and jetting is continued for another fixedperiod of time. For the most effective jetting, the wellsshould be pumped or airlifted during jetting to removethe fines as they are dislodged by the jetting. Thisprocess is continued until the entire well screen hasbeen jetted. The jetting tool should be continuously inmotion since a small amount of sand is disturbed andmay cause localized erosion of the screen. Jetting mustbe repeated a number of times to ensure optimum devel-opment of the well.
b. Surging. A surging block is a plunger consistingof one or more stiff rubber or leather discs attached to aheavy shaft. These discs should be about 1 in. smallerin diameter than the screen ID. A typical surge block isshown in Figure 8-5. Surging consists of moving waterin and out of the screen using the up and down motionof the surge block through short sections of the wellscreen. The well should always be pumped or bailed toensure a relatively free inflow of water prior to surging.Surging should begin with a slow and gentle motionabove the well screen and continue with more vigorfrom the top of screen downward. This method is lesseffective than the water jet described above in contin-uous slot screens and more effective in screens withwidely separated slots and louvered or shielded slots.The surging block should be pulled at approximately2 fps for effective surging. For record keeping
8-6
EM 1110-2-191429 May 92
purposes, it is convenient to use 15 round trips as one
Figure 8-3. Schematic of four-nozzle jetting tooldesigned for use unside 8-in well screen for jet devel-opment
cycle. The amount of material deposited in the bottomof the well should be determined after each cycle (about15 trips per cycle). Surging should continue until theaccumulation of material pulled through the well screenin any one cycle becomes less than about 0.2 ft deep.The well screen should be bailed clean if the accumula-tion of material in the bottom of the screen becomesmore than 1 to 2 ft at any time during surging,thenrecleaned after surging is completed. Materialbailed from a well should be inspected to see if anyfoundation sand is being removed. If the well is over-surged, the filter maybe breached with resulting infil-tration of foundation sand when the well is pumped.
c. Pumping. One of the least effective and slowestmethods of developing a well is simply pumping fromthe well. Pumping should be accomplished at a suffi-cient rate to effect maximum drawdown in the well.
Figure 8-4. Well development by high-velocity jetting(after Driscoll 1986)
Figure 8-5. Development with a surge block (afterDriscoll 1986)
8-7
EM 1110-2-191429 May 92
The water passing from the formation through the filterinto the well removes part of the finer fraction of thefilter material. The pumping equipment required de-pends on the size, yield, and anticipated drawdown inthe well. Surging produced by repeatedly starting andstopping a pump is only effective where the static waterlevel is well below the ground surface. Pumping, con-tinued over a long period of time, is a reasonably effec-tive method of well development. Pumping of the wellis normally accomplished by inserting a pipe in the welland forcing compressed air to the bottom of the well. Ifthe depth of submergence of the pipe is at least50 percent of its length, air bubbles reduce the weight ofthe water column and will cause a flow to the groundsurface. If 50 percent submergence is not possible, thewater column which must be physically blown out ofthe well as it accumulates will require a large supply ofair. Pumping can be accomplished using a mechanicalpump, but granular material in the water can causedamage.
8-12. Sand Infiltration
During the development process, sand and silt will bebrought into the well. When the depth of sand collectedin the bottom of the screen reaches 1 ft, it should beremoved by bailing. The accumulation of sand in thescreen prevents development of that portion of thescreen. A properly developed well will not produce anappreciable amount of sand, and entrance losses throughthe filter will be reduced to a minimum. In each of themethods discussed above, the actual amount of develop-ment must be recorded: the length, diameter, speed, andnumber of cycles of a surging block; the volume, pres-sure, and diameter of water jets; and the rate andmethod of pumping and length of time pumped. Inaddition, the amount of filter and foundation materialsbrought into the well and bailed out should be recorded.Upon completion of the development of the well, allmaterial infiltrated into the well should be bailed out.The well should be pumped to achieve a drawdown inthe order of 5 ft in the well. If the well produces sandduring pumping in excess at approximately 2 pints perhour (as determined from sounding and from collectionof well flow in a 10-gal container) the well should beresurged or developed further and repumped. Wellscontinuing to produce excessive amounts of sand after 4to 8 hours or surging or pumping should be abandonedand properly plugged.
8-13. Testing of Relief Walls
Performance of relief wells properly installed anddeveloped is determined by pumping tests. The pump-ing test is used primarily to determine the specificcapacity of the well and the amount of sand infiltrationexperienced during pumping. The information from thistest is required to determine the acceptability of the welland will be used to evaluate its performance and loss ofefficiency with time. The results of this pumping testmust be made a part of the permanent record concerningthe well.
a. Equipment. The equipment required for apumping test consists of a pump of adequate size toeffect a substantial drawdown. If the water level in thewell is near enough to the ground surface, and thespecific capacity of the well is high enough to produce asubstantial flow with a small drawdown, a centrifugalpump may be used for this purpose. If the water levelin the well is lower than about 18 to 20 ft, a deep-wellpump will be required to effect substantial drawdown.A flow meter is required to measure the flow rate. Aflat-bottom sounding device and a steel tape are requiredto determine the amount of sand infiltration deposited inthe bottom of the well. A suitable baffled stilling basinis used to determine the amount of sand in the effluent.A sounding device suitable for determining the depth tothe top of the water is needed to find the exact draw-down in the well. A well flow meter is desirable tomeasure the amount of flow at various depths within thewell to define flow from various zones.
b. Pumping. The well must be pumped to obtain aspecified drawdown or flow rate. Drawdown measure-ments in the well should be made to the nearest 0.01 ftand recorded with the flow rate at 15-minute (min)intervals throughout the duration of the tests. Sufficientsand infiltration determinations are necessary to esta-blish an infiltration rate for each hour of the pumpingtest. The rate of sand infiltration may be determinedfrom sounding and measurements of sand in the efflu-ent. For most properly developed wells, the amount ofsand deposited in the well will be negligible and sandinflitration in the effluent can be recorded in terms ofparts per million (Note: sand infiltration in parts permillion is approximately equal to pints per hour times3,000 divided by the pumping rate in gallons per min-ute) as measured with a centrifugal sand tester or other
8-8
EM 1110-2-191429 May 92
approved sediment concentration test (Driscoll 1986).The length of time that the pumping test must be contin-ued is normally specified for the particular project. Ifthe rate of sand infiltration during the last 15 min of thepumping test is more than 5 ppm, the well should beresurged by manipulation of the test pump for 15 min;then the test pumping should be resumed until the sandinfiltration rate is reduced to less than 5 ppm. If after6 hours (hr) of pumping the sand infiltration rate ismore than 5 ppm, the well should be abandoned.
8-14. Backfilling of Well
After completion of the well testing, the annular spaceabove the top of the filter gravel should be filled withfilter gravel if necessary to achieve design grade. Theremainder of the hole should be filled with either acement-bentonite mixture tremied into place or concretewhere the height of drop does not exceed 8 ft. In bothcases, a 12-in. layer of concrete sand or excess filtermaterial should be placed on top of the filter beforeplacement of grout or concrete. A tremie equipped witha side deflector will prevent jetting of a hole through thesand and into the filter.
8-15. Sterilization
Upon completion of the pumping tests and before instal-lation of the well cover, each well should be sterilizedby adding a chlorine solution with a minimum strengthof 500 ppm. Sufficient solution should be added to thebottom of the well to provide a volume equal to threetimes the volume of the well based on the outer diam-eter of the filter. Before the solution is introduced intothe well, all flow from the well should be stopped withinflatable packers or riser extensions. The solutionshould be injected into the well through a jetting tool byslowly raising and lowering the tool through thescreened portion of the well. The well should be gentlyagitated at 10-min intervals every 2 hr for the first 8 hrand then at 8-hr intervals for at least 24 hr. As thechlorine will dilute with time, the concentration shouldbe periodically checked; if it falls below 500 ppm, addi-tional chlorine compound should be added. It should benoted that calcium hypochlorite may combine withnaturally occurring calcium in the ground water to forma precipitate of calcium hydroxide which can plug the
pores of the foundation soils. Therefore, chlorine in theform of calcium hypochlorite should not be used inwaters containing high calcium content.
8-16. Records
Permanent records of the installation, development, test-ing, and sterilization of a permanent relief well must bekept for evaluation of future testing. To monitor theefficiency and performance of the installation, the recordmust include identification of the well, method of drill-ing, type, length and size of well screen, and slot size.The filter should be defined as to grain-size character-istics, depth, and thickness. Elevation of the top of thewell and the ground surface should be recorded. Anabbreviated log of the boring should be included todefine the depth to granular material, the thickness ofthat material, and the percent penetration of the well.Development data should include the method of devel-opment, the amount of effort expended in development,and the amount of materials pulled into the well duringdevelopment. The record should show the final soundeddepth of the well in case some fines remain at the bot-tom. The pumping test data should include the rate ofpumping, the amount of drawdown, the length of timethe pumping test was conducted, and the amount of sandinfiltration during pumping. Installation and pumpingtest data should be recorded on forms similar to thatshown in Figures 8-6 and 8-7. Forms should be filledin completely at the time each operation is completedand any additional observations should be recorded in a"remarks" section.
8-17. Abandoned Wells
Wells that produce excessive amounts of materials dur-ing pumping tests or that do not conform to specifica-tions and can not be rehabilitated should be abandoned.Abandoned wells should be sealed to eliminate physicalhazards, prevent contamination of ground water, con-serve hydrostatic heads in aquifers, and prevent inter-mingling of desirable and undesirable waters. Primarysealing materials consist of cement or cement-bentonitegrout placed from the bottom upward. In general,abandoned wells should be sealed following proceduresestablished by local, state, or Federal regulatoryagencies.
8-9
EM 1110-2-191429 May 92
Figure 8-6. Relief well installation report 28
8-10
PROJECT:
WES fQR~ 7'¥1 MAR ';3
REVl:iE:~ OCT. '~3
RELIE:F WELL INSTALLATION REPORT
ro:
PERFORAfiC:HS:
LOG OF
HOLE: P!P£
EM 1110-2-191429 May 92
Figure 8-7. Relief well pumping test report
8-11
RELIEF WELL PUMPING TEST REPORT
~I:OJECT: I LEVEE
DISTRICT: lOCAT!OH T i:LEV TOP (ST4)' or RISER,
Oi..T£: ,.nwE ttsT STdttD~
£U.PSE:O TIUE ""' DEPTH' TO DRAWDOWN n.ow llll<
MINUTES WAf~ .. I'm" tN GP'-i
' •
-~~L--SAND INnLTRA.T!OH TEST
DEPTH Of" WEU:: ~''.!!!. ,To.~~ ~. DEPTH SAND QJJ+I OR SAND
1£ST Till£ NO.
~J DiPTJ.t IO SAND IN WELL AFTrR TEST~
REMARKS:
WD fORM 798 l.l.f..R '!:3 R EVIStD OCT 'S3
TO
~~
~
IN LOSS or f'Ul.IP£0 WEU SAND,. IN<> our or (PTS) WEll.. PTS WELL. (PTS)
-
T I I I UltD JN Wil..L
AfTER TEST:
I WELL NQ;
lllLtE ltST CDMPUT"EO!
tU.FStD ll£1'Tll ~woow FI..OW TillE TG WATER 1H F"££T IN G-PW WINUTES
f ~~gRr T~~t;-~TAl FNF"L.OW L(NOTH RAT£ or SAND
:!AN~JO 01' TEST Ulnl.TRA.TION W£LL (WIN) (PlS. Hl!)
I
I
--
I SlHO IN Wi:LL AfTER CI...EAH!NG~
-
!HSPE:CTOR
EM 1110-2-191429 May 92
Chapter 9Relief Well Outlets
9-1. General Requirements
Relief wells should always be located where they areaccessible by a drill rig for pump testing and cleaningand provided with outlets for this purpose. The outletsshould be designed to minimize maintenance and toprovide protection against contamination from back-flooding, damage from floating debris, and vandalism.When wells are to discharge into a collector ditch orbackwater which may contain organic matter, debris,and fine-grained sediment in suspension, or where highvelocities may be expected while the wells are flowing,they should be installed off to the side and should dis-charge into the ditch or area through a tee connectionand horizontal outlet pipe protected against corrosion.A flat-type check valve should be installed on the wellriser with a flap gate on the end of the horizontal pipe.An example of this type of installation is shown inFigure 9-1.
9-2. Check Valves
Control of backflooding, which greatly impairs wellefficiency, is best implemented by flat-type check valvesconstructed of aluminum (see Figure 9-1). The checkvalve is supported by a soft rubber gasket which fitssnugly over the top of the riser or cast iron tenon set inthe concrete backfill. Other types of check valves maybe used but should be thoroughly tested under controlledconditions before application in the field.
9-3. Outlet Protection
For wells discharging at ground surface, the tops of thewells should be provided with a metal screen to safe-guard against vandalism, accidental damage, and theentrance of debris. Details of a conventional metal wellguard are shown in Figure 9-2. A suitable alternativeconsists of a section of stainless steel wire wound screenas shown in Figure 9-3. In the case of a T-type wellwhere the top of the riser pipe is more than 5 ft belowground, the well guard should be 42 in. in diameter topermit safe access by a ladder. A guard screen consist-ing of a wire mesh with 1 in.-square openings may beinstalled at the end of the outlet pipe to prevent animalsand debris from entering the outlet pipe in the event theflap gates do not close properly.
9-4. Plastic Sleeves
Where relief wells are provided for underseepage con-trol at levees, the well flows at relatively low riverstages will be somewhat in excess of natural seepage.In cases where the additional seepage is consideredobjectionable, each well can be provided with a plasticsleeve, 1.0 or 1.5 ft in length, which will raise the dis-charge elevation of the well accordingly. The sleevesprevent well flow at low river stages when no pressurerelief is necessary. At higher river stages or as soon assubstratum pressures develop to the extent that waterbegins to spill over the top of the sleeves, they shouldbe removed so that the well can function as intended.
9-1
EM 1110-2-191429 May 92
Figure 9-1. Typical detail of well top, check valve, and outlet
9-2
EM 1110-2-191429 May 92
Figure 9-2. Metal well guard details
9-3
5/8' BRASS-BOLT AND \J ASHER c 4 REQ' D>
2" X 2' X 3\16" ---~
. 0
L
I
-ffi-
1-------21' MIN.
31!6' STEEL ROD LIFTING RING ATTACHED BY wELDING
22' DIA lO GA. GALV. STEEL COVER PLATE
21' DIA. 10 GA. GAL V. CDRR. ME;T Al
PIPE. BITUMINOUS COATED
2' X 2' X 3/16"
BDL T OR RIVET
w!RE
NATURAL GROUND
~--------36~·-------~
EM 1110-2-191429 May 92
Figure 9-3. Alternative metal well guard
9-4
PLAN
rT 20" 16"
.if
""'. '· ,
l
' '
- Pl.ATE THREADED FOR FASTENING
SECTION A-A
' --8'' I.D.
.if -·"' ': .... ~ .. . : ...
~ 21" MSN ·l 36''
.3,B " SOCKET HEAD CAP SCREWS 304 SS ( 4 REQ'D}
2 !" QlA COVER PLATE 304 ss
20" DIA JOHrtSON SCREEN .250 SLOT - .304 SS
S<E OETAIL
~ ·. . :" 10"
___l
ELEVATION
.I METAL WELL GUARD DETAIL
(NTS)
(.235)
ROO
DETAIL
I'IOTE: 1 - FINISH TO BE FLAT
NOT BUFfED
2 - 51% OPEN AREA
EM 1110-2-191429 May 92
Chapter 10Inspection Maintenance and Evaluation
10-1. General Maintenance
Relief wells require a certain amount of nominal mainte-nance to ensure their continued and proper functioning.Any trash or obstruction in the well or well guardshould be removed immediately. Sand or other materialthat may have accumulated in and around flap gates toobstruct the flow or prevent functioning of the gatesshould be removed. Outfall ditches, bank slopes, orberms should be properly maintained in the vicinity ofhorizontal outlet pipes. The area in the immediatevicinity of the wells should be kept free from weeds,trash, and debris. Mowing and weed spraying should beextended at least 5 ft beyond the well, and the groundshaped and maintained for inspection and servicing ofthe wells.
10-2. Periodic Inspections
a. Periodic inspections of relief well systems shouldbe carried out as described in ER 1110-2-100 andER 110-2-1942. Relief well installations in readilyaccessible locations at dams and appurtenant structureswhere well flow is continuous should be visuallyinspected weekly. Observation should be made forevidence of wet spots on the dam or on the groundaround the wells and structures, for evidence of slough-ing or piping, for indications of discharge of sand orother materials from the wells, and for surficial signs ofdamage. The inspection should detect whether vandal-ism, theft, abuse by carelessness, unauthorized use ofthe wells or associated piezometers, or other irregulari-ties have occurred. The inspection should include anexamination of check valves, gaskets, well guards, coverplates, flap gates on tee outlets, and other appurtenances.Malfunctioning or damaged items should be repaired orreplaced. At yearly intervals, piezometric levels andflow quantities should be measured, and wells should besounded for evidence of deposition of sand or othermaterial in the wells. Where relief wells penetrate twoor more aquifers, the well flows at various depthsshould be checked at yearly intervals to determinewhether flows between aquifers are occurring.Piezometric levels and flow quantities should also bemeasured approximately one week after the attainmentof an unusually high reservoir level. Wells in relativelyinaccessible locations, as beneath stilling basins, shouldbe inspected whenever the structure is unwatered for ageneral maintenance inspection, or when there is
evidence of significantly decreased effectiveness, asshown by changes in flow quantities or piezometriclevels for a constant combination of reservoir level andtailwater level.
b. Flowing wells located in areas in which failurewould not constitute a hazard to life or property, as onexcavated slopes of canals, should be visually inspectedat monthly intervals. Measurements of piezometriclevels and flow quantities should be made annually.
c. Relief wells located along the toe of levees and atlocations where they flow infrequently should beinspected annually, preferably immediately prior tonormal high-water seasons and more often during majorhigh waters. Flow quantities and piezometric levelsshould be measured approximately a week after a peakin the reservoir level or in the river level at a levee.Pumping tests should be performed at five-year intervalson wells that flow infrequently. The tests should beperformed to determine the specific capacities and theefficiencies of the wells. The amount of sediment in thewells should be measured before and after performanceof the pumping tests.
10-3. Pumping Tests
All wells should be pump tested every five years usingprocedures described in Chapter 8. Wells in relativelyinaccessible locations should be pump tested wheneverthe structure is unwatered or when piezometric dataindicate that well efficiency has decreased significantly.Wells should be checked for sanding before and afterpumping. All wells requiring removal of sedimentshould be pump tested after cleaning to see if any ap-preciable loss of efficiency has resulted from foreignmaterial entering the well. In the case of continuouslyflowing wells, the discharge should be measured at highpool levels to determine specific capacity and indica-tions of sanding. If the pumping tests indicate that thespecific capacity is less than 80 percent of that deter-mined at the time of installation, then corrective mea-sures should be employed. Investigations as describedin Chapter 10 should be conducted prior to initiating therehabilitation methods described in Chapter 11. If therehabilitation are unsuccessful in restoring the wells toat least 80 percent of their original efficiency, consi-deration should be given to replacing these wells.
10-4. Records
A record should be kept of all inspections andmaintenance performed on each well. The record
10-1
EM 1110-2-191429 May 92
should include all pumping test data, descriptions ofrehabilitation efforts, and summaries of well flows andpiezometric data during periods of high river stages orpool levels.
10-5. Evaluation
a. It should be noted that a reduction in well dis-charge accompanied by a fall in piezometric levels indownstream areas probably indicates a decrease inseepage due to siltation in the reservoir, riverbed areas,or riverside borrow pits, which is a favorable condition.It is possible, however, that such a reduction was causedby erosion or excavation of an impervious top stratum ata point downstream of the line of wells, thus permittingexit of seepage to tailwater much closer to the wells.This condition would be unfavorable, because it wouldindicate a higher value of the seepage gradient and anincreased potential for piping immediately downstreamfrom the well line. A reduction in well dischargeaccompanied by an increase in piezometric levels indi-cates clogging or obstruction of the relief wells, andrequires immediate remedial action. Observation ofchanges in flow and piezometric levels must be relatedto changes or lack of changes in both reservoir level andtailwater level. Often, variation in tailwater level at
a dam has greater influence on well performance thanvariation in reservoir level, because the point at whichthe tailwater has access to the aquifer is considerablycloser to the well than the point at which the reservoirpressure can enter the aquifer.
b. The values obtained from measurement orpiezometric levels and flow quantities should beextrapolated to predict the values that would be pro-duced by a maximum design reservoir or river elevation.If these values are greater than those for which thestructure was designed, or if the specific capacities orthe efficiencies of the wells are less than 80 percent ofthe values that were obtained at the time of installationof the wells, additional investigations should be per-formed to determine the cause of the inadequacies.Investigations may include the examination of the wellscreen by means of a borehole camera, sounding thewell with a caliper, and the performance of chemicaltests on the water and on any deposits or incrustationsfound in the well. If there are any inclinometer tubesinstalled in the foundation in the vicinity of the wells,they should be read to determine if there has been anyhorizontal movement of the foundation that would causedisruption of well screens or risers.
10-2
EM 1110-2-191429 May 92
Chapter 11Malfunctioning of Wells andReduction in Efficiency
11-1. General
Relief wells may not function as intended and may alsobe subject to reduced efficiency with time. Failure ofrelief wells to function as intended can be attributed to anumber of causes. Deficiencies in design can usually beassessed during initial operation of the well system.Based on piezometric and well flow data, an assessmentof the effectiveness of the well system can be made andif considered inadequate, additional relief wells may beinstalled. Relief wells may malfunction for a variety ofreasons including vandalism, breakage, or excessivedeformation of the well screens due to ground move-ments, corrosion or erosion of the well screen, and agradual loss in efficiency with time. The reduced effi-ciency generally determined as a percentage loss inspecific capacity based on the specific capacity deter-mined from pumping tests at the time of installation is ameasure of increased well losses, which in turn result inhigher landside heads. Thus, reduced well efficiencywill result in hydrostatic heads larger than those antici-pated in the design. The major causes of reduced spe-cific capacity with time are (a) mechanical, (b) chemi-cal, and (c) biological.
11-2. Mechanical
Most relief wells undergo some loss in specific capacityprobably due to the slow movement of foundation finesinto the filter pack with a corresponding reduction inpermeability. The process occurs more commonly incases of poorly designed filter packs, improper screenand filter pack placement, or insufficient well develop-ment. Generally, the major cause of reduced efficiencyby mechanical processes is the introduction of fines intothe well by backflooding of muddy surface waters.Normally, backflooding can be prevented by the use ofcheck valves at the well outlet; however if not properlydesigned and maintained, the valves may not function asintended. The introduction of fines into the well andsurrounding filter pack under backflow conditions canresult in serious clogging which will result in reducedspecific capacities.
11-3. Chemical
Chemical incrustation of the well screen, filter pack, andsurrounding formation soils can be a major factor inspecific capacity reduction with time. Chemicaldeposits forming within the screen openings reduce theireffective open area and cause increased head losses.Deposits in the filter pack and surrounding soils reducetheir permeability and also increase head losses. Theoccurrence of chemical incrustation is determinedchiefly by water quality. The type and amount of dis-solved minerals and gases in the water entering the welldetermine the tendency to deposit mineral matter asincrustations. The major forms of chemical incrustationinclude: (a) incrustation from precipitation of calciumand magnesium carbonates or their sulfates, and(b) incrustation from precipitation of iron andmanganese compounds, primarily their hydroxides orhydrated oxides.
a. Causes of carbonate incrustations.Chemicalincrustation usually results from the precipitation ofcalcium carbonates from the ground water of the well.Calcium carbonate can be carried in solution in propor-tion to the amount of dissolved carbon dioxide in theground water. For a well discharging from a confinedaquifer, the hydrostatic pressure adjacent to the well isreduced to provide the gradient necessary for the well toflow. The reduction in pressure causes a release ofcarbon dioxide which in turn results in precipitation ofsome of the calcium carbonate. The precipitation tendsto be concentrated at the well screen and surroundingfilter pack where the maximum pressure reductionoccurs. Magnesium bicarbonate may change to magne-sium carbonate in the same manner; however incrusta-tion from this source is seldom a problem as precipita-tion occurs only at very high levels of carbonateconcentration.
b. Causes of iron and manganese incrustation.Many ground waters contain iron and manganese ions ifthe pH is about 5 or less. Reduction of pressure due towell flow can disturb the chemical equilibrium of theground water and result in the deposition of insolubleiron and manganese hydroxides. The hydroxides initial-ly have the consistency of a gel but eventually hardeninto scale deposits. Further oxidation of the hydroxidesresults in the formation of ferrous, ferric, or manganese
11-1
EM 1110-2-191429 May 92
oxides. Ferric oxide is a reddish brown deposit similarto rust, whereas the ferrous oxide has the consistency ofa black sludge. Manganese oxide is usually black ordark brown in color. The iron and manganese depositsare usually found with calcium carbonate and magne-sium carbonate scale.
11-4. Biological Incrustation
a. Iron bacteria are a major source of well screenand gravel pack contamination. They consist of organ-isms that have the ability to assimilate dissolved ironwhich they oxidize or reduce to ferrous or ferric ions forenergy. The ions are precipitated as hydrated ferrichydroxide on or in their mucilaginous sheaths. Theprecipitation of the iron and rapid growth of the bacteriacan quickly reduce well efficiency. Iron bacteriaproblems in ground water and wells are recognizedthroughout the world and are responsible for costly wellmaintenance and rehabilitation.
b. Despite the widespread familiarity with ironbacteria problems in wells, relatively little is knownabout their growth requirements. One reason for thelack of research on iron bacteria is that these organismsare difficult to culture for experimental study and thatpure cultures of many of these organisms have neverbeen obtained. Available information on the nature andoccurrence of iron-precipitating bacteria in ground wateris summarized by Hackett and Lehr in Leach and Taylor(1989).
c. In order to determine which genus of ironbacteria is contained in a particular water sample, asystem of classification based on the physical form ofthese organisms has been employed by the water wellindustry (Driscoll 1986). The three general forms re-cognized are:
(1) Siderocapsa. This organism consists of nume-rous short rods surrounded by a mucoid capsule. Thedeposit surrounding the capsule is hydrous ferric oxide,a rust-brown precipitate.
(2) Gallionella. This organism is composed oftwisted stalks or bands resembling a ribbon or chain. Abean-shaped bacterial cell, which is the only living partof the organism, is found at the end of the stalk.
(3) Filamentous Group. This filamentous groupconsists of four genera: Chrenothrix, Sphaerotilus,Clonothrix, and Leptothrix. The organisms are structur-ally characterized by filaments which are composed ofseries of cells enclosed in a sheath. The sheaths arecommonly covered with a slime layer. Both the sheathand slime layers or these organisms typically becomeencrusted with ferric hydrate resulting in large masses offilamentous growth and iron deposits.
d. Identification. The presence of iron bacteria isusually indicated by brownish red stains in well collec-tor pipes or ditches. Television and photographic sur-veys can pinpoint the locations of screen incrustation,and samples of the incrustations can be obtained by asmall bucket-shaped container. Samples can be sent tothe USAE Waterways Experiment Station, or a privatefirm familiar with iron bacteria for identification. Identi-fication is best accomplished by scanning electron ortransmission electron microscopy and phase contrasttechniques. Correct identification is necessary for selec-tion of an appropriate treatment method.
e. Prevention. It is not clear whether iron bacteriaexist in ground water before well construction takesplace, or whether they are introduced into the aquiferfrom the foundation soils or in mix water during wellconstruction. Evidence exists that iron bacteria may becarried from well to well on drill rods and other equip-ment and therefore every effort should be made to avoidintroducing iron bacteria into a well during installation,maintenance, or rehabilitation operations. After comple-tion of operations on a well, all drilling equipment,tools, bits and pumps, should be thoroughly disinfectedby washing with a chlorine solution (100 ppm) beforeinitiating work on another well.
11-2
EM 1110-2-191429 May 92
Chapter 12Well Rehabilitation
12-1. General
The analysis of well discharge records andaccompanying piezometric data will often indicatewhether the relief wells are functioning as intended. Adecrease in well discharges with time for similar pool orriver stages with rising piezometric levels between wellsis usually indicative of decreasing well efficiency. Aquantitative measure of the loss in efficiency is onlydetermined by carefully conducted pumping tests aspreviously described. Should the pumping tests indicatea reduction in specific capacity of more than 20 percentcompared to that measured at installation, a detailedstudy should be made of the consequences of the reduc-tion and what remedial measures should be employed.Generally, it may be possible to restore the wells toabout their original efficiency by means of rehabilitationtechniques.
Rapidly developing technology in the fields of chemistryand microbiology, as they are related to wells andaquifers, could negate portions of the following rehabili-tation techniques but the items covered are at leastbroadly covered and represent present practice. Envi-ronmental concerns (past and present chemical usage)also require that certain Federal, State, and local laws befollowed and rehabilitation techniques may have to bemodified to comply with these laws.
12-2. Mechanical Contamination
Plugging of relief wells by silts, clays, or other particu-late media entering the filter pack either from the forma-tion or through the top of the well is usually difficult todetermine except as indicated by periodic pumping tests.If significant reductions in specific yield are noted,rehabilitation of the well is in order. Mechanical rede-velopment of the well similar to that used to develop anew well should be the first step. Overpumping orpumping the well at the highest rate attainable is gener-ally advantageous. Surging and the use of horizontaljetting devices also may produce beneficial results.
12-3. Chemical Treatment with Polyphosphates
Mechanical plugging of relief wells is corrected mostoften by chemical treatment with polyphosphates. Thesechemicals act as dispersing agents which causes silt andclay particles to repel one another and calcium,
magnesium, and iron ions adhering to the particles toremain in a soluble state. The most widely used chemi-cals for this purpose are the glassy sodium phosphateswhich are inexpensive and readily available. The chem-icals are usually applied in concentrations of 15 to 25 lbper 100 gal of water in combination with at least50 ppm of chlorine (about one-half gal of 3 percenthousehold bleach or chlorox in 100 gal of water). Phos-phate solutions are mixed in a barrel or tank adjacent tothe well. The material is best dissolved in smallamounts in a wire basket or perforated container inagitated or swirling water. If the material is droppeddirectly into the tank or well, it will sink to the bottomand form a large gelatinous mass that could remainundissolved for some time. One of the most effectivemeans of introducing the phosphate and chlorine solu-tion into the well is by means of a horizontal jettingdevice. The well should then be surged vigorously priorto pumping. Three or more repetitions of injecting,surging, and pumping over a 2 to 4-hr cycle will bemuch more effective than a single treatment with a lon-ger detention time.
12-4. Chemical Incrustations
If the cause of reduced well efficiency is determined tobe chemical incrustation, more frequent cleaning andmaintenance should be initiated. If the efficiency re-mains low, consideration should be given to treating thewell with a strong acid solution which can chemicallydissolve the incrusting materials so that they can bepumped from the well. Acids most commonly used inwell rehabilitation are hydrochloric acid, sulfamic acid,and hydroxyacetic (glycolic) acid. Acid treatmentshould be used with caution on wooden screen wells asthe acid may tend to attack the lignin in the wood andcause severe damage. Methods for acid treatment ofwells are described in detail by Driscoll (1986). Themethods require great care and only experienced person-nel with specialized equipment should be employed.Specialized firms with experience in this field should beutilized for this purpose.
12-5. Bacterial Incrustation
Incrustation of wells by iron bacteria is best controlledby a combination of chemical and physical treatments.Many chemical treatments have been suggested andapplied in practice but their success has been variable asevidenced in many cases by recolonization or regrowthin the treated wells. A strong oxidizing agent such aschlorine is widely used to limit the growth of iron bac-teria. Chlorine, in the form of a gas, is used in the
12-1
EM 1110-2-191429 May 92
restoration of commercial wells; however safety andexperience requirements limit its general application. Amore convenient alternative is the use of hyperchloriteor other chlorine products (see Table 12-1). A discus-sion of procedures for the use of the various products isgiven by Driscoll (1986). Physical methods for controlof iron bacteria are available, however sufficientresearch has not been accomplished to justify their usein relief wells. A survey of new techniques is presentedby Hackett and Lehr in Leach and Taylor (1989).
12-6. Recommended Treatment
As clogging of well screens and filter materials iscaused not only by the organic material produced by thebacteria but also by oxides and hydroxides of iron andmanganese, better results are usually obtained by treat-ing the well alternately with a chlorine compound toattack the organic material and a strong acid to dissolvethe mineral deposits. Between each treatment the wellis pumped to waste to ensure that chlorine and acid arenot in the well at the same time. A recommended pro-cedure using the two procedures is:
a. Inject a mixture of acid, inhibitor, and wettingagent. The addition of a chelating agent such as
hydroxyactic acid may sometimes be beneficial. Aninhibitor is needed only if the well screen is metal. Theamount of acid should be typically one and a half totwo times the volume of the well screen. If a chelatingagent is not used, iron will precipitate out if the pH risesabove 3. The precipitate can result in clogging; there-fore the pH should be monitored throughout the acidtreatment and not be allowed to rise above 3 regardlessof whether a chelating agent is used.
b. Gently agitate the solution with a jetting tool at10-min intervals for a period of 1 to 2 hr.
c. Pump out a volume of solution equal to the vol-ume of the well.
d. Determine the pH of solution removed from thewell. If the pH is more than 3, repeat steps (a) to (c).
e. Allow the acid to remain in the well for a mini-mum of 12 hr and then pump to waste.
f. Inject a mixture of chlorine and one or morechloric-stable surfactants (detergents and wetting agents,for example). The concentration of the chlorine shouldexceed 1,000 ppm.
Table 12-1Quantities of Various Chlorine Compounds Required to Provide as Much Available Chlorine as 1 lb of Chlorine Gas 1
Chemical% AvailableChlorine
Number of lbEquivalent to1 lb Cl2
Chlorine Gas 100 1.0
Calcium Hypochlorite 65 1.54
Lithium Hypochlorite 36 2.78
Sodium Hypochlorite 12.5 8.0
Trichlorisocyanuric Acid2 90 1.11
Sodium dichloroisocyanurate2 63 1.59
Potassium dichloroisocyanurate2 60 1.67
Chlorine Dixoide 4 25.0
Chlorine Dioxide 2 50.0
Notes:1. From Driscoll (1986).2. Chlorine compounds that incorporate isocyanuric acid stabilize the chlorine against degradation from sunlight. Except for storage, the
advantage offered by the addition of isocyanuric acid is less valuable in water wells.
12-2
EM 1110-2-191429 May 92
g. Gentle agitate the solution with a jetting tool at10-min intervals every 2 hr for the first 8 hr and then at8 hr intervals for at least 24 hr.
h. Pump out a volume of solution equal to thevolume of the well.
i. Determine chlorine concentration. If the concen-tration is less than 10 percent of the original concentra-tion, repeat steps f to h.
j. Perform a pumping test on the well. If thespecific capacity has improved by more than 5 percent,repeat the entire procedure until the specific capacitydoes not improve by 5 percent.
12-7. Specialized Treatment
The USAE Waterways Experiment Station personnel,funded under a repair evaluation maintenance and resto-ration (REMR) work unit, developed a field procedure(Kissane and Leach 1991) for cleaning water wells thatprovides initial kill of the active bacteria in the well,dissolves the biomass in the screen, in the gravel pack,and some distance into the aquifer, and provides someinhibition of future growth. The procedure was devel-oped using a patented process known as the AlfordRodgers Cullimore Concept (ARCC). The proceduresin general include an initial well diagnosis performedwith a prepackaged field microbiological test kit whichis designed to give a qualitative indication of the typesof bacterial and chemical agents at work in the wells,and a very general indication of the bacterial concentra-tions. The initial water chemistry is also measured priorto treatment. A treatment is then designed with the
information from the tests, targeting the problematicagents with an appropriate set of chemicals. Redevelop-ment of the wells using the ARCC method is based onthe use of blended chemicals and high temperature(BCHT) and is divided into three principle elements oftreatment:
a. Shock. This phase is achieved by adding hightemperature chlorinated water to the well and surround-ing aquifer to "shock" kill or reduce the impact of dele-terious algae and bacteria. The water is chlorinated to>700 ppm with gaseous chlorine to avoid binders foundin powdered chlorine and is applied to the well as steamuntil the well temperature is brought above 120 deg Ffor massive bacteria kill. The chlorine treatmentremains in the well for a specified period of time;mechanical surging is used; and pumping follows forremoval of the initial loosened biomass.
b. Disrupt. This phase is achieved by the additionof chemical agents, acids and surfactants, and steam tothe well and surrounding aquifer while the well is pres-surized. Mechanical surging to break up organic andmineral clogging in the system is also used. Themechanical surging and chemical set time are importantduring this phase to achieve dissolution of the remainingbiomass.
c. Disperse. This phase of treatment consists ofremoval of the material that has been clogging the welland aquifer. Acceptance criteria for the well arechecked and further cycles are considered or a final coldchlorination treatment is applied for inhibition of anyremaining bacterial colonies.
12-3
EM 1110-2-191429 May 92
APPENDIX AReferences
A-1. Required Publications
TM 5-818-5Dewatering and Ground Water Control (Nov 83)
ER 1110-2-100Periodic Inspection and Continuing Evaluation ofCompleted Civil Works Structures (Apr 88)
ER 1110-2-110Instrumentation for Safety-Evaluations of Civil WorksProjects (Jul 85)
ER 1110-2-1942Inspection, Monitoring, and Maintenance of ReliefWells (Feb 88)
EM 1110-1-1804Geotechnical Investigations (Feb 84)
EM 1110-2-1602Hydraulic Design of Reservoir Outlet Works (Oct 80)
EM 1110-2-1901Seepage Analysis and Control for Dams (Sep 86)
EM 1110-2-1905Design of Finite Relief Well Systems (Mar 63)
EM 1110-2-1906Laboratory Soils Testing Ch 1-2 (Nov 70)
EM 1110-2-1907Soil Sampling (Mar 72)
EM 1110-2-1908Instrumentation of Earth and Rock-Fill Dams - Part 1(Aug 71), Part 2 (Nov 76)
EM 1110-2-1913Design and Construction of Levees (Mar 78)
EM 1110-2-2300Earth and Rock-Fill Dams General Design and Con-struction Considerations (May 82)
A-2 Related Publications
Banks 1963Banks, D. C. 1963 (Mar). "Three-Dimensional Electri-cal Analogy Seepage Model Studies; Appendix A:Flow to Circular Well Arrays Centered Inside a CircularSource, Series G," Technical Report No. 3-619, USArmy Engineer Waterways Experiment Station,Vicksburg MS.
Banks 1965Banks, D. C. 1965 (Mar). "Three-Dimensional Electri-cal Analogy Seepage Model Studies; Appendix B: Flowto a Single Well Centered Inside a Circular Source,Series H," Technical Report No. 3-619, US Army Engi-neer Waterways Experiment Station, Vicksburg, MS.
Barron 1948Barron, R. A. 1948. "The Effect of a Slightly PerviousTop Blanket on the Performance of Relief Wells,"Pro-ceedings of the Second International Conference on SoilMechanics and Foundation Engineering,Rotterdam,Netherlands, Vol 4, p 342.
Barron 1982Barron, R. A. 1982 (Sept). "Mathematical Theory ofPartially Penetrating Relief Wells," Unpublished reportprepared for the US Army Engineer Waterways Experi-ment Station, Vicksburg, MS.
Bennett and Barron 1957Bennett, P. T., and Barron, R. A. 1957. "Design Datafor Partially Penetrating Relief Wells,"Proceedings ofthe Fourth International Conference on Soil Mechanicsand Foundation Engineering,London, United Kingdom,Vol II.
Conroy 1984Conroy, P. 1984. "Computer Program for Relief WellDesign According to TM 3-424," US Army EngineerDistrict, St. Louis, St. Louis, MO.
Cunny, Agostinelli, and Taylor 1989Cunny, R. W., Agostinelli, V. M., Jr., and Taylor,H. M., Jr. 1989 (Sept). "Levee Underseepage SoftwareUser Manual and Validation," Technical Report REMR-GT-13, US Army Engineer Waterways ExperimentStation, Vicksburg, MS.
A-1
EM 1110-2-191429 May 92
Dachler 1936Dachler, A. E. 1936. "Grundwasserstromung,"Julius Springer, Vienna.
Driscoll 1986Driscoll, F. G. 1986. "Groundwater and Wells" 2d.,Johnson Division, SES, Inc., St. Paul, MN.
Duncan 1963Duncan, J. M. 1963 (Mar). "Three-Dimensional Elec-trical Analogy Seepage Model Studies," TechnicalReport No. 3-619, US Army Engineer WaterwaysExperiment Station, Vicksburg, MS.
Environmental Protection Agency 1976Environmental Protection Agency. 1976 (Jul). "Manualof Methods for Chemical Analysis of Water andWastes," EPA-625-6-76-003a. Available from Environ-mental Protection Agency, 26 W. St. Clair Street,Cincinnati, OH 45268.
Ferris, Knowles, Brown, and Stellman 1962Ferris, J. G., Knowles, D. B., Brown, R. H., and Stell-man, R. W. 1962. "Theory of Aquifer Tests,"Geological Survey Water-Supply Paper 1536-E,US Government Printing Office, Washington, DC.
Forcheimer 1914Forcheimer, P. H. 1914.Hydraulik, Teubner, Berlin.
Freeze and Cherry 1976Freeze, R. A. and Cherry, J. A. 1976.Groundwater,Prentice-Hall, Englewood Cliffs, N. J.
Hadj-Hamou, Tavassoli, and Sherman 1990Hadj-Hamou, T., Tavassoli, M., and Sherman, W. C.1990 (Sept). "Laboratory Testing of Filters and SlotSizes for Relief Wells,"Journal of Geotechnical Engi-neering, American Society of Civil Engineers, Vol 116,No. 9.
Harr 1962Harr, M. E. 1962. Groundwater and Seepage,McGraw-Hill, New York.
Johnson 1947Johnson, S. J. 1947. Discussion of "Relief Wells forDams and Levees," by Middlebrooks, T. A., and Jervis,W. H., Transactions of the American Society of CivilEngineers, Vol 112.
Kissane and Leach 1991Kissane, J., Leach, R. 1991 (Mar). "Redevelopment ofRelief Wells, Upper Wood River Drainage and LeveeDistrict, Madison County, Illinois," Technical ReportREMR GT-16, US Army Engineer Waterways Experi-ment Station, Vicksburg, MS.
Kozeny 1933Kozeny, J. 1933. "Theorie and Berechnvng derBrunnen," Wasserkroft W. Wasser Wirtschoft, Vol 29.
Leach and Taylor 1989Leach, R.E. and Taylor H.M. Jr. 1989 (Jul). "Proceed-ings of REMR Workshop on Research Priorities forDrainage System and Relief Well Problems," FinalReport, US Army Engineer Waterways ExperimentStation, Vicksburg MS.
Mansur and Dietrich 1965Mansur, C. I., and Dietrich, R. J. 1965 (Jul). "PumpingTest to Determine Permeability Ratio,"Journal of theSoil Mechanics and Foundation Division, AmericanSociety of Civil Engineers, Vol 91, No. SM4.
Mansur and Kaufman 1955Mansur, C. I., and Kaufman, R. I. 1955 (May)."Control of Underseepage, Mississippi River Levees, St.Louis District, Corps of Engineers," US Army EngineerWaterways Experiment Station, Vicksburg, MS.
Mansur and Kaufman 1962Mansur, C. I., and Kaufman, R. I. 1962. "Dewatering,"Foundation Engineering, edited by G. A. Leonards,McGraw-Hill, New York.
McAnear and Trahan 1972McAnear, C. L., and Trahan, C. C. 1972 (Jan)."Three-Dimensional Seepage Model Study, OakleyDam, Sangamon River, Illinois," Miscellaneous PaperS-72-3, US Army Engineer Waterways ExperimentStation, Vicksburg, MS.
Middlebrooks 1948Middlebrooks, T. A. 1948. "Seepage Control for LargeEarth Dams,"Transactions of the Third InternationalCongress on Large Dams, Stockholm, Sweden.
A-2
EM 1110-2-191429 May 92
Middlebrooks and Jervis 1947Middlebrooks, T. A. and Jervis, W. H. 1947. "ReliefWells for Dams and Levees," Transactions of theAmerican Society of Civil Engineers, Vol 112.
Montgomery 1972Montgomery, R. L. 1972 (Jun). "Investigation of Re-lief Wells, Mississippi River Levees, Alton to Gale,Illinois," Miscellaneous Paper S-72-21, US Army Engi-neer Waterways Experiment Station, Vicksburg, MS.
Moser and Huibregtse 1976Moser, J. H. and Huibregtse, K. R. 1976 (Sept)."Handbook for Sampling and Sample Preservation ofWater And Wastewater," EPA-600-4-76-049; Availablefrom Environmental Protection Agency, 26 W. St. ClairStreet, Cincinnati, OH 45268.
Muskat 1937Muskat, M. 1937. The Flow of Homogeneous FluidsThrough Porous Media, J. W. Edwards, Ann Arbor, MI.
Todd 1980Todd, D. K. 1980. Groundwater Hydrology, 2d ed.,John Wiley & Sons, New York.
Turnbull and Mansur 1954Turnbull, W. J. and Mansur, C. I. 1954. "Relief WellSystems for Dams and Levees,"Transactions of theAmerican Society of Civil Engineers, Vol 119,Paper No. 2701.
US Army Engineer District, Albuquerque 1967US Army Engineer District, Albuquerque. 1967 (Jul)."Cochiti Lake, Rio Grande, New Mexico, Embankmentand Conveyance Channel," Design Memorandum No. 9,Albuquerque, NM.
US Army Engineer Waterways Experiment Station1956US Army Engineer Waterways Experiment Station.1956 (Oct). "Investigation of Underseepage and ItsControl, Lower Mississippi River Levees," 2 Vols,Technical Memorandum No. 3-424, Vicksburg, MS.
US Army Engineer Waterways Experiment Station1958US Army Engineer Waterways Experiment Station.1958 (Jun). "Analysis of Piezometer and Relief WellData, Arkabutla Dam," Technical Report No. 3-479,Vicksburg, MS.
US Army Engineer Waterways Experiment Station1973US Army Engineer Waterways Experiment Station.1973. "Computer Program H2030- Discharge in Pres-sure Conduits Using the Darcy-Weisbach Formula,"Vicksburg, MS.
Warriner and Banks 1977Warriner, J. B., and Banks, D. C. 1977. "NumericalAnalysis of Partially Penetrating Well Arrays," Techni-cal Report S-77-5, US Army Engineer WaterwaysExperiment Station, Vicksburg, MS.
A-3
EM 1110-2-191429 May 92
APPENDIX BMathematical Analysis ofUnderseepage and SubstratumPressure
B-1. General
The design of seepage control measures for levees anddams often requires an underseepage analysis withoutthe use of piezometric data and seepage measurements.Contained within this appendix are equations by whichan estimate of seepage flow and substratum pressurescan be made, provided soil conditions at the site arereasonably well defined. The equations contained hereinwere developed during a study by the US Army Engi-neer Waterways Experiment Station (1956)1 of piezome-tric data and seepage measurements along the LowerMississippi River and confirmed by model studies. Thefollowing discussion is presented in terms of leveeunderseepage; however the analyses and equations areconsidered equally applicable to dam foundations. Itshould be emphasized that the accuracy obtained fromthe use of equations is dependent upon the applicabilityof the equation to the condition being analyzed, theuniformity of soil conditions, and the evaluation of thevarious factors involved. As is normally the case, soundengineering judgment must be exercised in determiningsoil profiles and soil input parameters for these analyses.
B-2. Assumptions
It is necessary to make certain simplifying assumptionsbefore making any theoretical seepage analysis. Thefollowing is a list of such assumptions and criteria nec-essary to the analysis set forth in this appendix:
a. Seepage may enter the pervious substratum atany point in the foreshore (usually at riverside borrowpits) and/or through the riverside top stratum.
b. Flow through the top stratum is vertical.
c. Flow through the pervious substratum is hori-zontal. Flow in the vertical direction is entirelydisregarded.
d. The levee (including impervious or thick berm)and the portion of the top stratum beneath it areimpervious._______________________________1 References cited in this appendix are listed in
Appendix A.
e. All seepage is laminar.
In addition to the above, it is also required that thefoundation be generalized into a pervious sand or gravelstratum with a uniform thickness and permeability and asemipervious or impervious top stratum with a uniformthickness and permeability (although the thickness andpermeability of the riverside and landside top stratummay be different).
B-3. Factors Involved in Seepage Analyses
The volume of seepage (Qs) that will pass beneath alevee and the artesian pressure that can develop underand landward of a levee during a sustained high waterare related to the basic factors given and defined inTable B-1 and shown graphically in Figure B-1. Otherterms used in the analyses are defined as they are dis-cussed in subsequent paragraphs.
B-4. Determination of Factors Involved inSeepage Analyses
Many of the factors necessary to perform a seepageanalysis, such as exploration and testing, havepreviously been mentioned in the text; however they arediscussed in more detail as they apply to each specificfactor. The use of piezometric data, although rarelyavailable on new projects, is mentioned primarily be-cause it is not infrequent for seepage analyses to be per-formed as a part of remedial measures for existinglevees in which case piezometric data often areavailable.
a. Net head, H. The net head on a levee is theheight of water on the riverside above the tailwater ornatural ground surface on the landside of the levee.His usually based on the net levee grade but is sometimesbased on the design or project floodstage.
b. Thickness, Z, and vertical permeability, kb, of topstratum. Where the thickness of the riverside blanketdiffers from that of the landside blanket, the designa-tions, ZbR and ZbL are used. Similarly the permeabilityof the riverside and landside blankets are designatedkbR
andkbL , respectively.
(1) Exploration. The thickness of the top stratum,both riverward and landward of the levee, is extremelyimportant in a seepage analysis. Exploration to deter-mine this thickness usually consists of auger boringswith samples taken at 3- to 5-ft intervals and at every
B-1
EM 1110-2-191429 May 92
Table B-1Examples of Transformation Procedure_________________________________________________________________________________________________________________
Strata
ActualThicknessft
ActualPermeabilitycm/sec
Transformed Thickness,ft
Ft = kb for kb = 1 × 10-4 cm/seckn
Clay 5 1 × 10-4 1 5.0
Sandy Silt 8 2 × 10-4 1/2 4.0
Silty Sand 5Z=18
10 × 10-4 1/10 0.5Zb = 9.5
Figure B-1. Illustration of symbols used in Appendix B
B-2
EM 1110-2-191429 May 92
change of material. Boring spacing will depend on thepotential severity of the underseepage problem butshould be laid out for sampling the basic geologic fea-tures with intermediate borings for check purposes.Landside borings should be sufficient to delineate anysignificant geological features as far as 500 ft awayfrom the levee toe. The effect of ditches and borrowareas must be considered.
(2) Transformation. The top stratum in most areasis seldom composed of one uniform material but ratherusually consists of several layers of different soils. Ifthe in situ vertical permeability of each soil layer (kn) isknown, it is possible to transform the top stratum to anequivalent stratum of effective thickness and verticalpermeability. However, a reasonably accurate seepageanalysis can also be made by assuming a uniform verti-cal permeability for the top stratum equal to the perme-ability of the most impervious strata and then using thetransformation factor given in Equation B-1 to deter-mine a transformed thickness for the entire top stratum.
(B-1)Ft
kb
kn
where Ft is the transformation factor. Some examplesusing this procedure are given in Table B-1 and in Fig-ure B-1. A generalized top stratum having a uniformvertical permeability of 1 x 10-4 cm/sec and thickness of9.5 ft would then be used in the seepage analysis forcomputation of effective blanket lengths. However, thethicknessZbL may or may not be the effective thicknessof the landside top stratumZt that should be used indetermining the hydraulic gradient through the top stra-tum and the allowable pressure beneath the top stratum.The transformed thickness or the top stratum equals thein situ thicknesses of all strata above the base of theleast pervious stratum plus the transformed thicknessesof the underlying more pervious top strata. Thus,ZbL
will equal Zt only when the least pervious stratum is atthe ground surface. Several examples of this transfor-mation are given in Figure B-2. To make the finaldetermination of the effective thicknesses and perme-abilities of the top stratum, conditions at least 200 to300 ft landward of the levee must be considered. Inaddition, certain averaging assumptions are almostalways required where soil conditions are reasonablysimilar. Existing landward conditions or critical areasshould be given considerable weight in arriving at suchaverages.
c. Thickness D and permeability kf of pervious sub-stratum. The thickness of the pervious substratum isdefined as the thickness of the principal seepage-carry-ing stratum below the top stratum and above rock orother impervious base stratum. It is usually determinedby means of deep borings although a combination ofshallow borings and seismic or electrical resistivitysurveys may also be employed. The thickness of anyindividual pervious strata within the principal seepage-carrying stratum must be obtained by deep borings. Theaverage horizontal permeabilitykf of the pervious sub-stratum can be determined by means of a field pumptest on a fully penetrating well as described in the maintext. For areas where such correlations exist, their usewill usually result in a more accurate permeability deter-mination than that from laboratory permeability tests.In addition to the methods above, if the total amount ofseepage passing beneath the levee (Qs) and the hydraulicgrade line beneath the levee (M) are known,kf can beestimated from the equation
(B-2)kf
Qs
M
d. Distance from riverside levee toe to river, L1 .This distance can usually be estimated from topographicmaps.
e. Base width of levee and berm, L2. The distance,L2, can be determined from anticipated dimensions ofnew levees or by measurement in the case of existinglevees.
f. Length of top stratum landward of levee toe, L3.This distance can usually be determined from borings,topographic maps, and/or field reconnaissance. Todetermine this distance, careful consideration must begiven to any geological feature that may affect the seep-age analysis. Of special importance are deposits ofimpervious materials, such as clay plugs which canserve as seepage barriers. If the barrier is located nearthe landside toe, it could force the emergence of seep-age at the near edge and have a pronounced effect onthe seepage analysis.
g. Distance from landside levee toe to effective seep-age exit, x3. The effective seepage exit (Point B,Figure B-1) is defined as that point where a hypotheticalopen drainage face would result in the same hydrostaticpressure at the landside levee toe and would cause the
B-3
EM
1110-2-191429
May
92
B-4
f "
II
"
i Ill
ACTUAL STRATI\.
GROUND SURFACE <C.S)
SANDY $JL T K1 "" C X 10-4
K CLAY 2 = .tX.lD-4
SIL 1Y SANO _4 K...., = 10 X 10
" SAND·
Figure 8-2. Transformation of tops strata
112
l/10
l/10
1/2
112
1/1[)
TRANSFORMED STRATA FOR COHPUT.r!ON OF SEEPAGE
AND SUBSTRATUM PRESSURE
' " " " " N
*
t z b3., 0.5'
z ,_ o.s·
"'!
z4
:F b2-=5" .L
f z b3 "'0.5'
CLAY
SANDY Sll T
SA.ND
SANDY StLT
CLAY
SANDY SILT
CLAV
'= S lL TY S:AND
TRANSFORMED STRAT~ FOR CD~PUTATION OF" MAXIMUM
ALLOIJABLE UPLirT
N
*
i Sll H SAND
SANDY SILT
N
1 cuw
t SANDY SILT
N CLAY
* f z 3 ;;;;; 0.5' SILTY SAND
EM 1110-2-191429 May 92
same amount of seepage to pass beneath the levee aswould occur for actual conditions. Point B is locatedwhere the hydraulic grade line beneath the levee pro-jected landward with a slopeM intersects the groundwater or tailwater. If the length of foundation and topstratum beyond the landside levee toeL3 is known, x3
can be estimated from the following equations:
(1) For L3 = ∞
(B-3)xc
1C
kf ZbL D
kbL
where
(B-4)CkbL
kf ZbL D
(2) For LB = finite distance to a seepage block
(B-5)x3
1c tanh cLb
(3) For L3 = finite distance to an open seepage exit
(B-6)x3
tanh cL3
c
h. Distance from effective source seepage entry toriverside levee toe, x1. The effective source of seepageentry into the pervious substratum (Point A, Figure B-1)is defined as that line riverward of the levee where ahypothetical open seepage entry face fully penetrates thepervious substratum. An impervious top stratum be-tween the seepage entry and the levee would producethe same flow and hydrostatic pressure beneath andlandward of the levee as would occur for the actual con-ditions riverward of the levee. Effective seepage entryis also defined as that line or point where the hydraulicgrade line beneath the levee projected riverward with aslope,M, intersects the river stage.
(1) If the distance to the river from the riversidelevee toe,L1, is known, and no riverside borrow pits orseepage blocks exist,x1 can be estimated from the fol-lowing equation:
(B-7)x1
tanh cL1
c
where C is calculated from Equation B-4 using appro-priate properties of the riverside top stratum.
(2) If a seepage block (usually a wide, thick depositof clay) exists between the riverside levee toe and theriver in order to prevent any seepage entrance into thepervious foundation beyond that point,x1 can be esti-mated from the following equation:
(B-8)x1
1c tanh cL1
where L1 equals distance from riverside levee toe toseepage block andc is calculated from Equation B-4.
(3) The entrance conditions often are such that anassumption of a vertical entrance face is not reasonable.Two limiting cases are shown in Figure B-3. The addi-tional effective length,∆ L1, may be obtained for eitherCase A which assumes a uniformly sloping entranceface or Case B which assumes a combined infinitehorizontal entrance face with a vertical entrance face,D′, varying from0 to D (see Figure B-3).
i. Critical gradient for landside top stratum,ic. Thecritical gradient is defined as the gradient required tocause boils or heaving (flotation) of the landside topstratum and is taken as the ratio of the submerged orbuoyant unit weight of soil,γ ′, comprising the top stra-tum and the unit weight of water,γw, or
(B-9)ic
γ ′γw
Gs 1
1 e
B-5
EM 1110-2-191429 May 92
Figure B-3. Corrections for nonvertical entrance face (after Barron 1982)
B-6
\ \ \
/ 0 0.0
I J
/ ~V
1/ 1\. cY
""' /
/ ~" / ~
~ (), 1 0.2 0.3 0.4
l>L 1/Il
EXTRA LENGTH RATIO DUE TO SEEPAGE ENTRANCE CONDITIONS
WATER SURFACE
PERVIOU~.
/ / /
1.0
0.8
~ 0.6 ::::>
u ~
.e_ " 0.4 "-0 0 ;;;:
0.2 "'
0.0 0.5
FOUNDATION
L2.
//
AFTER BARRON ( 1982)
EM 1110-2-191429 May 92
where
Gs = specific gravity of soil solids
e = void ratio
j. Slope of hydraulic grade line beneath levee, M.The slope of the hydraulic grade line in the pervioussubstratum beneath a levee can best be determined fromreadings of piezometers located beneath the levee wherethe seepage flow lines are essentially horizontal and theequipotential lines vertical. The slope of the hydraulicgrade line cannot be reliably determined, however, untilthe flow conditions have developed beneath the levee.If no piezometer readings are available, as in the casefor new levee design,M must be determined by firstestablishing the effective seepage entrance and exitpoints and then connecting these points with a straightline, the slope of which isM.
B-5. Computation of Seepage Flow andSubstratum Hydrostatic Pressures
a. General
(1) Seepage. For a levee underlain by a perviousfoundation, the natural seepage per unit length of levee,Qs, can be expressed by.
(B-10)Qs s∫kf D
wheres∫ is the shape factor. This equation is valid pro-vided the assumptions upon which Darcy’s law is basedare met. The mathematical expressions for the shapefactor s∫ (subsequently given in this appendix) dependupon the dimensions of the generalized cross section ofthe levee and foundation, the characteristics of the topstratum both riverward and landward of the levee, andthe pervious substratum. Where the hydraulic grade lineM is known from piezometer readings, the quantity ofunderseepage can be determined from
(B-11)Qs Mkf D
(2) Excess hydrostatic head beneath the landsidetop stratum. The excess hydrostatic headho beneath thetop stratum at the landside levee toe is related to the nethead on the levee, the dimensions of the levee and
foundation, permeability of the foundation, and thecharacter of the top stratum both riverward and land-ward of the levee. The headhx beneath the top stratumat a distancex landward from the landside levee toe canbe expressed as a function of the net headH and thedistancex, although it is more conveniently related tothe headho at the levee toe. Whenhx is expressed interms ofho it depends only upon the type and thicknessof the top stratum and pervious foundation landward ofthe levee; the ratiohx/ho is thus independent of river-ward conditions. Expressions fors∫, ho and hx forvarious boundary conditions are presented below.
b. Case 1 - no top stratum. Where a levee isfounded directly on pervious materials and no top stra-tum exists either riverward or landward of the levee(Figure B-4a), the seepageQs can be obtained fromEquation B-12. The excess hydrostatic head landwardof the levee is zero andho = hx = 0. The severity ofsuch a condition in nature is governed by the exit gradi-ent and seepage velocity that develop at the landsidelevee toe which can be estimated from a flow net com-patible with the value ofS computed from Equa-tion B-12.
c. Case 2 - impervious top stratum both riversideand landside. This case is found in nature where thelevee is founded on thick (15-ft) deposits of clay or siltswith clay strata. For such a condition, little or no seep-age can occur through the landside top stratum.
(1) If L3 is infinite in landward extent or the pervioussubstratum is blocked landward of the levee, no seepageoccurs beneath the levee andQs = 0. The head beneaththe levee and the landside top stratum is equal to the nethead on the levee at all points so thatH = ho = hx.
(2) If an open seepage exit exists in the impervioustop stratum at some distanceL3 from the landside toe (i.e.,L3 is not infinite) as shown in Figure B-4b, the distancefrom the feet toe of the levee to the effective seepageentry (river, borrow pit, etc.) isL1 = L2. The equation forthe shape factor is given by Equation B-13, and the headsho andhx can be computed from Equations B-14 and B-15,respectively.
d. Case 3 - impervious riverside top stratum and nolandside top stratum. This case is shown in Figure B-4c.The condition may occur naturally or where extensivelandside borrowing has taken place resulting in removal ofall impervious material landward of the levee for aconsiderable distance. The shape factor is computed fromEquation B-16. The excess head at the top of the sand
B-7
EM 1110-2-191429 May 92
landward of the levee is zero, and the danger from pipingmust be evaluated from the upward gradient obtained froma flow net.
e. Case 4 - impervious landside top stratum and noriverside top stratum. This case is more common thanCase 3 and occurs when extensive riverside borrowing hasresulted in removal of the riverside impervious top stratum(Figure B-4d). For this condition, the shape factor iscomputed from Equation B-17; the headsho and hx arecomputed from Equations B-18 and B-19, respectively.
f. Case 5 - semipervious riverside top stratum and nolandside top stratum. This case is illustrated inFigure B-5a. The same equation for the shape factor aswas used in Case 3 can be applied to this conditionprovided x1 is substituted for L1 in Equation B-16
resulting in Equation B-20. Since no landside top stratumexists,ho = hx = 0.
g. Case 6 - semipervious landside top stratum and noriverside top stratum. This case is illustrated inFigure B-5b. The shape factor is given by Equation B-21and the headsho and hx are computed from Equa-tions B-22 and B-23, respectively.
h. Case 7 - semipervious top strata both riverside andlandside. Where both the riverside and landside top strataexist and are semipervious (Figure B-6), the shape factorcan be computed from Equation B-24. The headho isgiven by Equation B-25. The headhx beneath the semi-pervious top stratum depends not only on the headho butalso on conditions landward of the levee and can becomputed from Equations B-26 through B-30.
Figure B-4. Equations for computation of underseepage flow and substratum pressures for Cases 1 through 4
B-8
EM 1110-2-191429 May 92
Figure B-5. Equations for computation of underseepage flow and substratum pressures for Cases 5 and 6
Figure B-6. Equations for computation of underseepage and substratum pressures for Case 7
B-9
EM 1110-2-191429 May 92
APPENDIX CList of Symbols
a Well spacing
a,b Dimensions defining boundary of non-circularsource
Ac Effective average distance from well center toexternal boundary of source
Ae Equivalent radius of a group of wells
c (1) Conversion factor used in determining the effec-tive length of a pervious foundation covered bya semipervious blanket
C (2) Hazen-Williams coefficient
C (3) One-half the length of a finite line source
Cu Coefficient of uniformity
d Thickness of a pervious stratum
d_
Transformed thickness of pervious stratum layerwith thickness = d
dm Thickness of pervious foundation layers (summa-tion of m = 1 to m = n)
D Thickness of pervious foundation
D_
Transformed thickness of pervious foundation
D’ Thickness of sloping entrance face of perviousfoundation
Dn Grain size for which n percent of the sample issmaller
e Void ratio
Ft Permeability transformation factor
FS Factor of safety
g Acceleration due to gravity
Gp Flow correction factor for partially penetratingwell
G(T) A function used in analysis of partially penetrat-ing well
h Net head on well system corrected for well losses
ha Allowable net head beneath top stratum at landsidetoe of levee on dam
hc Piezometric head midway between wells in circulararray
hd Maximum head landward from a slot or line ofwells
hg Head at boundary between artesian and gravityflow
hm Net head midway between well corrected for welllosses
ho Excess hydrostatic head beneath the top stratum atlandside levee toe
hp Net head at any point p
hw Head at well
hx Head beneath top stratum at distance x from land-slide toe of levee
hav Average net head in plane of wells corrected forwell losses
hwj Head at well j in a system of n wells
∆hd Difference in elevation between landslidepiezometric surface and well outlets
H Net head on well system. Difference betweenriverside pool and landslide tailwater
H1 Total head measured from bottom of perviousfoundation
He Entrance loss in screen and filter
Hf Frictional head loss
Hm Net head midway between wells
Hv Velocity head loss
C-1
EM 1110-2-191429 May 92
Hw Well losses
Hav Average net head in plane of wells
HmnNet head midway between n number of wells
Hmoo Net head midway between wells in infinite line
∆Hm Excess head above the well outlet midwaybetween wells
ic Critical hydraulic gradient
jo Downward force acting against uplift pressure
k Coefficient of permeability
k-
Coefficient of permeability of transformedstratum
kb Vertical permeability of top stratum
k-
e Effective permeability of multistratum trans-formed aquifer
kf Horizontal permeability of pervious substratum
kh Coefficient of permeability in horizontaldirection
kL Coefficient of permeability from laboratory tests
kv Coefficient of permeability in vertical direction
km Permeability of pervious foundation layers(summation of m = 1 to m = n)
kn Vertical permeability of individual layers com-prising top stratum (n = layer number)
kbL Vertical permeability of landside top stratum
kbR Vertical permeability of riverside top stratum
L Distance from source to seepage exit
L1 Distance from source to landside toe of levee ordam
L2 Base width of impervious levee and berm
L3 Length of pervious foundation and top stratumbeyond landside toe of levee
LB Distance from line of wells to blocked exit
Le Distance from line of wells to seepage exit
∆L1 Extra length of pervious foundation due to slopingentrance face
n Number of wells in group
M Slope of hydraulic grade line (at middepth ofpervious foundation)
∆M Difference in slopes of hydraulic grade line river-side and landside of toe of levee
p A point
Qa Artesian component of seepage flow
Qg Gravity component of seepage flow
Qs Total amount of seepage beneath levee
Qw Discharge from a single well
Qwj Flow from well j
Qwp Flow from partially penetrating well
Qsw Seepage flow beyond well system
r Radial distance from well (distance from point pto real well)
rc Rows of circular array of well
ri Distance from ith well to point p
rij Distance from well i to well j
r’ Distance from point p to image well
ro Distance from well to center of finite line source
rw Radius of well
rwj Effective well radius of well j
R Radius of influence
Ri Radius of influence of ith well
Rj Radius of influence of well j
C-2
EM 1110-2-191429 May 92
S Distance from effective seepage entry to line ofwells
Sj Distance from infinite line seepage to multiplewells
v Flow velocity in well
W Actual well penetration
__W Effective well penetration
x,y,z Cartesian coordinates
xa Distance from effective seepage entry to pointwhere gravity flow occurs
xg Length over which gravity flow occurs to wellline
x1 Distance from effective seepage entry to river-side toe of levee
x3 Distance from landside toe of levee to effectiveseepage exit
Z Thickness of top stratum
Zb Transformed thickness of top stratum
Zc Thickness of top stratum below collector ditch
Zn Thickness of individual layers comprising topstratum (n = layer number)
Zt Transformed thickness of landside top stratum foruplift computations
ZbL Transformed thickness of landside top stratum
ZbR Transformed thickness of riverside top stratum
α Angle of entrance face
γ Unit weight of soil
γw Unit weight of water
γ’ Submerged unit weight of soil
Θa Average uplift factor
Θm Midwell uplift factor
∆Θ Change inΘa andΘm per 1 log cycle of a/rw
δ Offset distance between well and center of circularsource
s∫ Shape factor
C-3
