EARTHFILL DAM CONSTRUCTION AND FOUNDATION TREATMENT 341

ao

o B 7l

z
I

tl cqi
O*
-?.60 Soecrtic orotrtt
0 2.70 Soacifrc arorrtt
: <; ro
tl
t: Co3!\V
>a O
I
l-m IO
-lo
o
Jlo m
clr
zl)
-lr .t
.l;
\Sioaillc6l o6ounl ot I I
/
at< I wlumr chsga o tslu.ottoo. €s€
6lI Jrouxoettor ttt^rrtxr f !roul0 llxlt- ls
Els
1 rEourr Eo.\'-fJ-fT-ft

EIE ."
,,
o
Cosc II
vor05
I
----1
s0L r0 5
I
Dr
.E -O -7 o!.o1,
I
* IET FOPTIXUT t ORY
L oos! Voids rquol to the Dense
Oo- @- OPtl IUU IAIEF @ilE|{ll!a at oeY reEtsll-HIIURAL wAteR COtlEXTti 6t Oel wEElr I
wot?r volum! ot
ihe Iiquid limit
o Consiocring 'o-rotio: the rolio of noturol (inplocel dr, dcnsit, to
P.oclor morimum dry densitr. ord ul-(d, oplimum ,oter cont€nt minus
noturol wotrr conltnt. b. Considerrng nolurol dry den3ity ond liqurd limit

Figure l0-9. Criterion for treatment of relatively dry fine-grained foundations. Courtesy of Bureau of Reclamation.

Morimum Lood that of the foundation of Choke Canyon Dam. At this site
lnatiol ond Solurotrd
there are dry low-density alluvial soils up to 33 ft ( l0 me-
DrI dcnsitr(pcf)
ters) in depth, overlain by 3 to 6 ft (l to 2 meters) of lean
88.0 105.4
l{oisturr conlcnt(1} t3.9 zo.!
ocgree of soturofronGl {0.5 89.9 and fat clay soils. The very impervious clay precluded
prewetting the underlying silty and sandy soils by sprinkler
.! unless the clays were removed first. A scheme of using
.9 injection wells was field-tested and shown to be a viable
P solution (Fig. l0-l l). Construction started with excavation
of the cutoff trench, followed by backfilling to the original
ground surface. Holes 12 in. (30 cm) in diameter were
then drilled at 50-ft (I5-meter) centers, a 4-in. (10 cm)
Lod -r3i
slotted-plastic pipe was inserted, and gravel was packed to
form the injection well. Field tests showed that more effi-
cient wetting was attained by not having full well penetra-
tion. The best results were achieved by augering through
the layers of silty material, to determine their full depth,
G then backfilling with impervious clays to appmximately
one-half the thickness of the layer. Water was applied only
at gravity head. Some wells were spaced closer than 50 ft
( l5 meten) when moisture samples from surrounding au-
ger holes indicated a need for additional wells. The original
splcim.n,irorzt,.lIi-'' ohornifii.T' soil moisture was as much as 8% dry of Proctor optimum.
Closliftsrioo lrmM - rL
Moisture was increased only to approximately optimum.
Figurc l0-I0. One-dimensional consolidation tests at natural moisture Consolidation of just over 3% was measured during con-
(dashed) and at optimum moisture (solid). struction, with about 0.5% during reservoir filling.
342 ADVANCED DAM ENGINEERING FOR DESIGN. CONSTRUCTION. AND REHAEILITATTON

xol ?cxctrr?c0
6touilo Surr/lct txrrl?t^?tor rtLL

/r,'rorlnalf ArE/r or r€xgTt^Tto{

lrrEt 6 oirYs or wfTTtx6
CHOK g CAN YON OAM
PiEWETTIX6 FOUXOATIOX

Figure l0-ll. Foundation prewetting by infiltration wells.

Guidelines 6. The amount of moisture to be added can be esti-

mated from standard soil tests.
l. Flooding may seal the surface and limit penetration. 7. Field tests have been very useful for evaluating the
2. It is advisable to sprinkler-irrigate unstripped sur- efficiency of the final wetting method to be used.
faces. 8. A curing time is necessary to distribute moisture
3. Runoff is increased where the surface has been dis- within the deposits.
turbed by construction or farm equipment. 9. Instrumentation should be provided to verify the
4. Water should be applied continuously at the maxi- consolidation.
mum rate without runoff. 10. Flooding has been very successful in areas ofsevere
5. Saturation is not required to achieve consolidation collapse potential.
during embankment construction.

EARTHFILL CONSTRUCTION
CLRuoe A. Frrzen

Control of earthfill construction is fully described in pub-
lications of the U.S. Bureau of Reclamationz and the U.S.
Corps of Engineers.3 This section will be limited to cov-
erage of critical items.
Operation of Borrow Sources
Borrow sources are considered to be all sources of mate-
rial, including required excavations and borrow areas.
Surface drainage should be diverted away from the bor-
row sources, and the excavation should be controlled to
assure positive drainage at all times. Where a high water
table is present, subdrainage should be provided to pull the
water table below the level to be excavated except where
free-draining sand and gravel can be excavated below
water. This subdrainage usually can be accomplished by
deep ditches, which, depending on the topography, may
have to be pumped. Spacing of ditches will be dependent
upon the permeability of the aquifer.
In arid areas where the borrow materials are more than
I or 2% below the minimum moisture content required for
the fill, the normal practice is to prewet the borrow sources
with sprinkler systems. Prewetting to significant depths re-
quires time, which must be taken into consideration by the
contractor in establishing the schedule ofoperations. In hu-
mid areas where the borrow materials arc 3Vo above the
maximum moisture content permitted. disking or deep
plowing should be carried out in the borrow areas; the sur-
face of the borrow area should be sealed when possible
before rains.
Inspection forces should make continuous observations
to assure that materials being excavated in the borrow
sources meet specified requirements for the zone receiving
the fill. Averages do not count-it only requires one lift of
 EARTHFILL OAM CONSTRUCTION AND FOUNOATION TBEATMENT 343
weak material to negate the validity of the stability anal-
yses, or one impenrious or semi-impervious lift to negate
the design function of a drainage layer or of an assumed
drained zone. Hence, timely action is needed to avoid
placement of detrimental materials in the embankment. All
filter, drainage, bedding, and stone-protection materials
should be sampled and tested prior to placement in the em-
bankment, and no deviation should be permitted from the
specified gradation Bnges; these materials should also be
tested after placement, as placement and compaction often
can cause degradation.
Control of Surface Drainage
As stated earlier, in Chapter 7, surface drainage from side
draws and from slopes above the dam area must be diverted
outside the area protected by the cofferdams. Provisions
must also be made for collecting and pumping out rain-
water that falls within the cofferdam area; otherwise, rain-
fall will cause undue delays in cleaning the foundation and
firr.
Removal of Dewatering SYstem
The foundation dewatering system should not be shut off
and removed until the embankment has more than adequate
weight to resist hydrostatic uplift, and until there is no dan-
ger of hauling and compaction equipment causing pumping
of groundwater into the embankment. Collector pipes and
surrounding drainage layers left in the foundation should
be backfilled with grout; and the system must be well
vented for the grout to flow throughout the pipes.
lnitial Lifts on Rock Foundations
Initial lifts for the impervious core on rock foundations must
be placed and compacted by methods that will assure ad-
equate compaction without damaging the rock surface. One
method to accomplish these two goals is (a) to use a small
dozer to push an extra-thick lift (approximately 150% of
the normal lift thickness) of fill material over the founda-
tion rock (numerous measurements of the loose-lift thick-
ness should be made to assure that it has a uniform thick-
ness), and (b) to compact the initial lift with heavily loaded
trucks or other suitable pneumatic equipment' Rolling is
continued until control tests indicate that the required den-
sity has been obtained at top of rock, as the top-of-rock
density is critical from both shear strength and seepage
control standpoints. Prior to placement of the next lift, the
top of the preceding lift must be scarified to assure bonding
with the next lift, but the scarification must be accom-
plished with a small farm+ype disk or other equipment that
does not break through to the foundation rock. Where the
foundation rock has an undulating surface, several lifts
compacted with pneumatic equipment may be requircd be-
fore the switch to the approved tamping (sheepsfoot) roller.
Compaction against Abutments and Structures
Special compactive effort is required for the earthfill against
abutments and structures, as uncompacted material at these
locations has been the source of numerous seepage and pip-
ing problems. These problems are usually caused by the
inability of self-propelled and towed tamping (sheepsfoot)
rollers to compact the materials adequately within I to 2 ft
(0.3 to 0.6 meter) of steep abutment and structure faces.
Experience on several projects has indicated the difficulty
of adequately compacting this gap with hand-held mechan-
ical tampers; these tampers can only compact very thin lifu'
and their production rate is so slow that the remaining fill
rises much more rapidly than they can compact the ga5
so the gap is often filled in with an extra lift of uncom-
pacted material without the inspector's knowledge. An open
pipe along the abutment or structure is created when, upon
saturation, the uncompacted layer settles. and the overly-
ing compacted layer arches over the void.
Good compaction against abutments and structures has
been obtained by rolling parallel to steep abutmens and
structures with the front tires of loaded, front-end loaders;
the tires can rub against abutment and structure faces with-
out damaging the rrrck or concrete. The production rate with
this equipment is sufficient to keep up with the remainder
ofthe embankment. In confined areas such as narrow back-
fill areas next to an outlet structure, hand-held mechanical
tampers or other special compaction equipment may be re-
quired. Regardless of the compaction equipment used, sev-
eral densities should be taken in each lift against the faces
of abutments and structures to ensure that the required den-
sity is being achieved.
Where embankment or backfill is being placed against
both sides of a structure, the fill of both sides should be
kept at the same elevation to equalize the thrust of the fill.
As yielding of the structure during compaction is less than
that required to develop active pressure, the forces exerted
by the fill on the structure may be considerably greater than
at-rest pressures.
Compaction around lnstrumentation
The compacted density around instrumentation riser pipes
should be equal to that of the contiguous fill' The riser pipes
must be protected from damage by the construction equip-
ment. On many projects, barriers are placed around the riser
pipe while the surrounding fill is raiscd I to 2 ft (0'3 to 0.6
meter) abovc the compacted zone around the pipe. The
barriers are then removed, the loose fill above the com-
pacted zone is removed, and the excavated area is back-
filled with compacted material using mechanical tampers.
 344 ADVANCED DAM ENGINEERING FOR DESIGN, CONSTRUCTION, AND REHAEITITATION
To prevent damage to the riser pipe from rammer-type
tampers, heavy vibratory plates have been successfully used
even for compacting impervious materials at West Point23
and Warm Springs?3 dams.
At Warm Springs Dam,3l a compacted mound 3 to 5 ft
(0.9 to 1.5 meters) high above the general fill level was
maintained around the riser pipes. A red flag was attached
to the riser pipes, but no protection other than the mound
was provided. The top of the mound was approximately 4
ft ( I .2 meters ) in diameter, and the conical side slope was
approximately I on l. As the general fill was raised adja-
cent to the mound, the loose material along the side slope
of the mound was bladed off and mixed with the general-
fill material; and the fill compactor was used to obtain good
compaction along the plane between the mound and the
general fill. Initially a rammer-type compactor was used
for compacting the mound, but the equipment was changed
to a heavy plate vibrator when it was found that the re-
quired density could be more easily obtained with the plate
vibrator. According to the record-test data,2a the imper-
vious and random fill materials in which the instruments
were located contained considerable granular material. The
impervious material had an average of 37 % sand size and
30% gravel size. The random material had an average of
36Vo sand size and 48Vo gravel size. Percentages are based
on weight. The fines for both materials were of low plas-
ticity and in general ploned just above the A-line on rhe
Atterberg-limits chart. Outer fringes of the mound were
compacted with the rubber tires of a backhoe. Lift thick-
ness forthe plate vibratorwas approximately 4 in. ( l0 cm).
To prevent twisting of the piezometer riser pipes and in-
clinometer casings, the forward direction of the plate vi-
brator was reversed after each pass, that is, from clockwise
to counterclockwise, and so on.
Placement of Lifts
Layers compacted or sealed with rubber-tired rollers must
be scarified prior to placement of succeeding lifts to assure
a good bond between lifts. Tamping rollers may leave a
well-indented liftsurface that does not need scarification;
however, even these lift surfaces will contain smooth areas
requiring scarification because the areas have been crossed,
sometimes repeatedly, by pneumatic-tired hauling equip-
ment. The thickness of each lift placed should be measured
and recorded; otherwise, the lifts will be placed exces-
sively thick. Disking and blading of each loose lift is usu-
ally required to break up large lumps and to provide a more
uniform surface for compaction.
The surface of the fill should be sloped to drain roward
the shoulders, and the slope should always be away from
inclined drains and filters to prevent fines from being
washed into the drainage system.
Moisture Control
The moisture content of fill material should be as close as
possible to the specified limits when the material is placed
on the embankment. On a mass production basis, it is dif-
ficult to appreciably raise the moisture conrent uniformly
through the lift, as considerable disking is required, which
interferes with the production schedule. Likewise, it is dif-
ficult to appreciably reduce the moisture content on the fill
in humid climates, and especially on the night shift. Ac-
cording to Hammer,2s "When disking for drying purposes,
it is especially important that the disk be pulled at a speed
sufficient to throw the soil off the individual discs, thus
enhancing the drying process by aerating the material."
The moisture content must be within the specified range
prior to compaction; otherwise, constant reworking of
compacted layers will be required. To provide moisture
data between the time when a lift is placed and it is com-
pacted, the tests must be made very quickly; but the stan-
dard laboratory method (ASTM D 22rc)26 reguires several
hours to dry the materials, making a more rapid method
necessary. Considerable variation in moisture contents has
been obtained by the hot-plate and pan method when com-
pared to the standard laboratory method. However, the mi-
crowave oven has been successfully used for drying sam-
ples quickly. The required cooking time for each type of
material is obtained by plotting the microwave time versus
moisture content to determine the time when the reduction
in moisture content levels off. The moisture content at that
time is plotted on a chart versus the moisture content ob-
tained by ASTM D 2216, and a 45" line is drawn through
the origin. The deviation from the 45o line determines the
correction factor that must be used for the microwave re-
sults for each type of material. All gravel particles larger
than the No. 4 sieve must be removed from the materials
prior to microwaving, as the stones may explode because
of an intemal water pressure buildup. Materials should not
be left in the microwave longer than the calibrated time, as
the microwave will start removing bound water.
Nuclear moisture-density equipment also is used to de-
termine the moisture content (ASTM D 3017) very rapidly.
A minimum of one check test (ASTM D 2216) per ten nu-
clear tests should be performed for assurance that the nu-
clear test values are correct.
Compaction Control
Specifications. Compaction is specified by either the
end-result or the procedural method. [n the end-result spec-
ification, the minimum percent of maximum density as de-
termined by a specific laboratory compaction test method
is specified; this type of specification would appear to be
the simplest way of controlling the fill compaction, but it
 EARTHFTLL
is not considered to be the best. The procedural specifica-
tion establishes the number of passes of specific equipment
for each tift; these requirements are based on past experi-
ence with similar materials or on analyses of test-fill re-
sults. In the procedural specification the required minimum
percent density is given to the resident engineer, but it is
not included in the specifications as a contract requirement.
However, provisions are usually included in the specifica-
tions for payment for extra roller passes as needed to obtain
the desired density. In both methods, the lift thickness and
the moisture-content range in relation to the laboratory op-
timum are usually included in the specifications. A slight
change in gradation, even from a borrow source with fairly
homogeneous materials, can result in a significant change
in the maximum density, and earth-rock mixtures cannot
be readily controlled by densities alone; so the procedural
specification is considered the best method for positively
controlling the compaction of the fill to uniformly achieve
the desired density, with density tests providing spot con-
firmation and as-built data. Also, the procedural specifi-
cation is fairer to the bidders, as the specifications provide
a definitive work effort with extra pay for any extra work.
The procedural specification requires continuous inspec-
tion for checking that materials placed in each zone are
within specification limits, for measuring the lift thickness,
and for counting roller passes; however, continuous in-
spection also is required under the end-result specification
for the same reasons except for counting passes. A labo-
ratory technician is required under both specifications to
perform moisture and density tests, but appreciably more
density tests should be performed with the end-result spec-
ification on the open fill than would be required for the
procedural specification. The same number of moisture and
density tests should be performed at the abutments and
against structures under either specification, and this num-
ber should be greater than that on the open fill.
Test Methods. Field density test methods and labora-
tory moisture-density and relative density test methods are
fully described in volume 04.08 of the 1987 Annual Book
of ASTM Standards.zo The sand displacement method
(ASTM D 1556) is commonly used to take densities in earth
fills. Nuclear equipment (ASTM D 3017) is often used to
measure density in place for shallow depths in commercial
work including dams. Prior to using nuclear equipment,
the technician should have training at the equipment man-
ufacturer's school; also, one check test by ASTM D 1556
for each ten nuclear tests should be performed to assure
that the results are consistent.
Field density tests of uniformly graded filter and drain-
age materials having less than 5% by weight passing the
No. 100 sieve are extremely difficult to obtain, as the upper
unconfined surface to depths of 6 to l2 in. (15 to 30 cm)
DAM CONSTRUCTION AND FOUNDATION TREATMENT 345
of these materials will not hold a density: hence, any re-
sults obtained will indicate a low relative density. One way
to check the relative density ol these materials, which
should be compacted in a saturated state with a vibratory
roller, is to perform standard penetntion tests (SPT) at
depth. The difficulty of obtaining accurate densities in
granular materials is another reason for using the proce-
dural specification for compaction.
Large-scale water ring displacement tests using a plastic
liner are employed to obtain densities for record purposes
in earth-rock fill mixtures and in rockfill zones. In earth-
rock mixtures the placement and compaction procedures
are normally established from analyses of test-fill densities
and/or level data from test fills, and the densities obtained
in the prototype are not often used for control purposes.
Selection of Correct Moisture-Density Compaction
Curve. The critical factor in the control of earth fill work
is to determine the appropriate maximum density and op-
timum moisture content in order to compare the field den-
sity and moisture content with these values. This compar-
ison is needed quickly to determine whether a completed
lift is satisfactory or needs to be reworked and recompacted
prior to placement of the next lift. On most projects, a fam-
ily of five-point moisture-density compaction (ASTM D
698 or D 1557) curves is developed in the early stages of
construction for each embankment zone. As construction
proceeds, additional moisture-density compaction cun'es
are developed when new-type materials are encountered.
As it normally takes at least 48 hours to prepare the spec-
imens and perform five-point moisrure-density compaction
tests, it is not possible to perform a five-point compaction
test for every density test taken on the fill. Hence, the crit-
ical step is to quickly relate the dry density and moisture
content obtained on the fill to the correct five-point com-
paction curve and to its accompanying maximum density
and optimum moisture content. In the visual-observation
procedure, the five-point compaction curve is selected by
comparing the soil obtained from the field density with jar
samples saved from the laboratory compaction tests. Ham-
mer,25 who considers this method to be unreliable, devel-
oped a quick method (2 houn) for Warm Springs Dam
whereby one-point compaction tests could be performed for
selection of the appropriate five-point compaction curve
where the curves are parallel, or two-point compactior: tests
could be made where the five-point compaction curves are
not parallel. In the one-point method the material from the
field density is pulverized and ntixed and allowed to dry to
a water content of approximately 2% dry of optimum. lt is
then compacted in a mold and weighed, and the dry density
is calculated from the wet weight and moisture content.
The compacted dry weight and moisture content are plotted
on the family of curves, as shown in Fig. 10-12. If the
 346 ADVANCED DAM ENGINEERING FOR DESIGN, CONSTRUCTION, AND REHABILTTATION

",oz aro
e uarw
efr.u
N o,att,f7
..taa cufa.7
hN Cad o, eiM)

/ \.t 'tf,''i,Ezi,*#i;il' ato

t 3tl srto l$ \"1 (Lnc a erwo

I
t
a
6
a
/ I
N E
2
I

I
a
o

Y(O \
<ux 'ltot cu*t ,
t',,\
vur trru& effffi
ft57 0 t&w ru rtPtaL-

20 25 to
utet cocrTti?, t srtt coitf xT, !a

'lfi*:me*,ff ;!$:H,:ffi,?,:iir:r E:tr Is,."? L*

' ffi *i :f i'ri.Titi.tflr:iE$::' #r,ffl i'i;':'
2. tsYrutEo co{Bcrloi culvc aasto ox lo.ils a aNo a
:Sraalr$cs
'H€ o olr txa l-xut, wxlcx l3 rHa f ttluaiEo
d a 2rud,q
P^Crrox cuiE3. - '..,:"*IffJ:f.::,::'*T::H:;- "orN7 aio errw qtlr coNrcflY of t*
Lilfli.",ff""rtr
Figurc l0-12. Illustration of one-point compaction correlation method. Figure lGl3. Illustration of two-point compaction correlation method.
After Hammer.r After Hammer.r

point falls between curves, a curve is projected to obtain
the maximum dry density and optimum moisture content.
A similar procedure is followed for the two-point method.
Before compaction, one specimen is dried to slightly below
optimum moisrure, and the second specimen is dried to a
lower moisture content. The results are plotted, as shown
in Fig. l0-13, to determine an estimated compaction curve
for comparison with the field density and field moisture
content. Hammel5 recommends: "Full five-point curves
should be continually run during construction (preferably
I forevery l0 or l2 one-point tests) to check the one-point
data and to supplement the family of curves. The family of
curves must be plotted to a large scale; otherwise, the plot
will become cluttered and errors will result." It is consid-
ered that five-point laboratory compaction tests should be
run at the same ratio used for the two-point method.
The U.S. Bureau of Reclamation has developed the
Rapid Compaction Control method, which is fully de-
scribed under Designation E-25 inits Eanh Manual.2
Earth-Gravel Mixtures. Where density control is re-
quired in random fill or earth-gravel mixtures, construction
control becomes even more difficult as larger field densities
and larger laboratory compaction specimens are required
than for fine-grained soils. It is commonly accepted that
the maximum particle size in the compaction mold should
not exceed one-sixth the diameter of the mold. Hence, full-
scale testing on material having a maximum diameter of 3
in. (7.5 cm) requires an l8-in. (46-cm)-diameter mold,
which has a volume of 2.65 ft3 (0.075 m3;. To reduce the
sizeof specimens needed in the field laboratory, consid-
erable research has been accomplished by Donaghe and
Townsend2T and others on scalping and replacing the plus
l-in. 1Z-cm; gravel for compaction in a 6-in. (15-cm)
mold, in compacting scalped minus ol-in. (2-cm) material
in a 6-in. ( l5-cm) mold, and in correlating the results with
full-scale testing of minus 3-in. (7.5-cm) gravelly material
in an l8-in. (46-cm){iameter mold. The Donaghe-
Townsend2T report is considered essential reading for em-
bankment engineers involved in construction control of
earth-gravel mixtures. Although their report's title is
"Compaction Characteristics of Earth-Rock Mixtures."
the rock sizes were actually gravel and not broken rock per
se.
At Warm Springs Dam2a the random material was re-
quired to have 100% passing the 4-in. ( lO*m) sieve and
not more than 50% passing the No. 200 sieve, and the ma-
terial was required to be compacted to an average percent
compaction of 95% of maximum dry density as determined
by the modified compaction test, ASTM D 1557. The ma-
terial was spread in lifts to produce an 8-in. (20-cm) lift
thickness after compaction with eight passes of double-
drum, self-propelled, vibrator mllers having a total static
weight of 45,500 lb (20,638 kg) and delivering an im-
parted dynamic force of 57,800 tb (26,218 kg) per drum
at 1700 vibrations per minute. As the material contained a
maximum of l0% by weight in gravel sizes in excess of 2
in. (5 cm), the particles larger than 2 in. (5 cm) recovered
 EARTHFILL OAM CONSTRUCTION AND FOUNDATION TREATMENT 347

in the density test holes werc placed back in the hole prior 2. Compaction tests on duplicate specimens by ASTM
to pouring in the density sand; hence, control was based D 1557 Method D in a 6-in. ( l5-cm)-diameter mold:
on the density of the minus 2-in. (5-cm) material, which that is, the experimenters scalped and replaced the
would require a l2-in. (30-cm)-diameter laboratory mold. plus l-in. (2-cm) material.
To reduce the amount of effort required for the field labo- 3. Compaction tests in a 6-in. ( l5-cm)-diameter mold
ratory, a testing Prcgram was performed to determine if on duplicate specimens by ASTM D 1557 lttethod C
results equivalent to the l2-in. (30-cm) compaction tests on scalped minus l-in. (2-cm) material, with results
could be obtained in a 6-in. ( l5-cm) mold (a) by scalping corrected upward for the effect of the oversize.
and replacing the plus f-in. (Z-cm) material in accordance 4. Use of the rtsults obtained from the l2-in. (30-cm)
with ASTM D 1557, Method D, or (b) by compacting the mold to evaluate each Procedure.
scalped f-in. (Z-cm) material in a 6-in. ( 15-cm) mold and
theoretically correcting the density upward and the mois- Some of the pertinent results are shown in Figs. l0-14
ture content downward by these commonly used equations: through l0-17. The gradation curyes for the 30% plus l-
in. (2-cm) material are shown in Fig. l0-14. The mois-
Densiry correction: ture-density compaction curves for the 30% plus ]-in. (Z-
cm) material are shown in Fig. l0-15. The optimum water
"Yf G^^Yn content versus the percent plus l-in. (2-cm) material is
,YA x 100 shown in Fig. 10-16. The maximum dry density versus the
11P, + G^"Y*PI percent plus l-in. (2-cm) material is shown in Fig. l0-17.
These data indicate that the compaction tests on the scalped
where: minus J-in. (2-cm) material with results theoretically cor-
rected provided better correlation up to 30% plus ]-in. 1Z-
?a = calculated dry unit weight of total sample cm) material than did the scalped and replaced tests. As
'yf : dry unit weight of finer fraction (from compaction most random material had 30% or less of plus l-in. 12-cm 1
test on scalped material) material, the procedure used at Warm Springs Dam to se-
G- : bulk specific gravity of coarser fraction (oversize lect the maximum density and optimum moisture content
material) was a one-point compaction test for each field density in a
7- : unit weight of water 6-in. (15-Cm)-diameter mold on scalped minus l-in. 12-
P" = percent of coarse material by weight of the sample
cm) material with the results theoretically corrected for
P, : percent of fine material by weight of the sample moisture and density for the oversize. A five-point, 6-in.
( l5-cm)-diametercompaction curve was run on each shift,
M oi s tu r e - c ont e nt co rr e cti on : and the bulk specific gravity ( G-) of the plus ]-in. 12-cm;
material was determined. Occasional compaction tests in
(wyPs * w,P,) the l2-in. (30-cm) mold were also run as a further check
WT:
r00 on the system.
The experimental data obtained at Warm Springs Dam
where: were very similar to those obtained by Donaghe and Town-
send.27 As shown in Fig. 16 of their report, the scalped
wr = water content of total sample in percent and theoretically corrected test results from the 6-in. ( l5-
l{,/ = p€rccnt water content of fine material cm)-diameter compaction tests up to 40% plus l-in. 12-
P, : percent of fine material by weight to total sample cm) content were more representative of the test results
r/c : percent water Content Of coarse material obtained from the full-scale tests in an l8-in. (46-cm)-di-
p" : percent of coarse material by weight to total sam- ameter mold than were the test results from the scalp-and-
ple replacement method. On every project where density con-
trol is required for earth-gravel or earth-rock mixtures. it
Note: The dividing line for coarse and fine fractions at is considcred advisable to conduct a laboratory experinten-
Warm Springs Dam was l-in. (2 cm), but it could also be tal program similar to the one at Warm Springs Dam to
No. 4 sieve size as used by Donaghe-Townsend.?7 correlate the results from different-size compaction molds.
The experimental work at warm Springs Dam23 in-
cluded the following: Compaction EquiPment

l. Compaction tests in l2-in. (30-cm)-diameter mold On large mass-production projects, self-propellcd, double-
on sfecimens with varying Percentages of plus ]-in. drum, sheepsfoot rollen are becoming the standard, rather
(2-cm) material. than towed rollers, for compacting the impervious fill. A
 348 ADVANCED DAM ENGlNEERING FOR DESIGN. CONSTRUCTION. AND REHAEILITATION

u. s slrr{)mo gtYt (Illffi n railfs U. i St^,.()^tO gw€ iUrBtEt rrqxltt

| ! .r,tt I tto t.l6 20 r lo !(, m tolao

I I
e r:
! E
6 6
G
z
6
z
I z
c
E

Gn^I'I EZE UT TILIITTTRS

Gn^rlt SArao
coSSLts 9[r m cuY
w fx IM

Figure l0-14. Warm Springs Dam. gradation curve for 30% plus ]-in. (2-cm) random material used in laboratory compaction studies. After Hammer.ls

recent development is Caron Compaction Wheels, which

IFCM can be used with many compacting machines. Caron Com-
o< 66Mdd.Mirulrakd
e--9 CGdGir-Mhs,46.CuN paction Wheels have interchangeable tamping caps of var-
\-{ 6hMd-Hs}4rrtUd
uFddRqtd(ASTM)
ious sizes. According to the manufacturer's literature.lE a
H l2-i&Md.Mhu!-irH Caterpiller 825 B Compactor equipped with Caron Wheels

r
I
l.trr
za {2
la

e FUU. sc^Il fl2.tN. Dtr MOt!)

4
ar I
d
a REPTACED{CC{, MOTJ))

d
F
, I

5
o.2.?0 sc^lrEDrrHE(nEnc,t .,)q, -.-
s.tffi cdlEffiilT'iliifiilor/
01020!0s0s0
2q68l012trl PERCEIfT PLUS 3/{.N' MATERIAL
WATER CONTENT (I DRY WT.)
Figure l0-16. Warm Springs Dam. random material compaction studies,
Figure l0-15. Warm Springs Dam, random materiirl compaction srudies. optimum water contcnt vs. pereent plus ]-in. (2-cm) matcrial. Aftcr
dry density vs. water content. After Hammer.:l Hammer.:t
 EARTHFITL DAM CONSTRUCTION AND FOUNDATION TREATMENT 349

Warm Springs for compacting thc impe rvious eanh fill. as

shown in Fig. l0-18.
142 **=,,, For random fill, earth-gravel and eanh-rock mixtures. a
g
variety of high-production types of compaction equipment
I has been used. As previously described, double-drum, self-
). l!0
propelled, vibratory rollers were used for the random zones
A at Warm Springs Dam,23 where the combined production
Ius
o rates for impervious and random zones ranged from 60.000
2
f to 120,000 ydr (45,873 to 91,141 mr; for two l0-hour
shifts per day. Rollen were used in series as shown in Fig.
IE t,t
A\A 10-19. As these rollers had smooth wheels, each lift had to
be scarified after compaction to assure good bond with the
ll4 A nexr lift.
SC LM & REPT CED
(6N. Dh MOLD) At East Fork Dam,2e Little Miami River, Ohio, produc-
B2 tion rates of 40.000 ya3 llO,saZ mrl were achieved in
ol020l0q0t0
PERCENT PLUS 3/4.IN. MATERIAL
compacting random fill consisting of soft shale and thinly
bedded limestone in S-in. (20-cm) loose lifts with four
Figure 10-17. Warm Springs Dam, random material compaction studies, passes of a self-propelled tamping roller followed by nvo
maximum dry density vs. percent plus ]-in. (2-cm) material. After
passes of a 50-ton (45,a00-kg) rubber-tired roller: the
Hammer.2l
tamping roller had tapered feet projecting 9.5 in. (24.1 cm)
from the drum, a tip contact area of 12 in.z (77.4 cm:),
with w/S-10 ( l0 in.2, 64 cm2; caps would have a weight and a contact pressure of 661 psi (46.5 kg/cmr1.
of 4747 lb/lin ft (706a kg/lin meter) of drum and a foot
contact pressure of 880 \bfin.z (61.9 kg/cm2) based on Utilization of lnstrumentation
two tamping feet per wheel only in full contact with the
ground surface. Equipped with w/S-14 ( 14 in.2, 90 cmz) Installation of the different types of instrumentation is de-
caps, the weight would be a788 lb/lin ft (7125 kg/lin me- scribed in Chapter 25. The responsibility for reading the
ter) of drum, and the foot contact pressure would be 634 instruments and interpreting the data should be given to a
lbfin.z (44.6 kg/cm2;. Self-propelled rollers in series specific individual. On large or complex projects, staffing
equipped with Caron Compactor Wheels were used at is usually provided for an instrumentation engineer. On

Figure l0-18. Warm Springs Dam, self-propelled sheepsfoot rollers used in series for conrpacting inrpervious zone. Photo courtesy, U.S. Army Engineer
District, Tulsa.
 350 ADVANCED OAM ENGINEERING FOR DESIGN, CONSTRUCTION. AND REHABILITATION
,...
Figure l0-19. Warm Springs Dam, self-propelled vibratory rollers used in series for compacting random zone. photo courtesy, U.S. Army Engineer
District. Tulsa.
small projects, the embankment engineer may serve also as
the instrumentation engineer. The use of computers for
providing updated printouts of the instrumentation data is
most helpful in the timely inrerpretarion of the data. With
modern telecommunications equipment, the instrumenta-
tion data and printouts can be quickly provided to the de-
sign staff in the home office.
The on-site instrumentation engineer should always be
on the alert for abnormal readings or abnormal trends that
are beyond the guidelines provided by the design staff.
Careful and timely investigations are needed to determine
if abnormal readings are due to errors in instrumentation
readings, malfunctioning instruments or developments
within the monitored areas. A full set of tested backup in-
strumentation readout equipment should be kept on the
project at all times because critical readings would be
missed in the time required to obtain replacements. Fur-
thermore, backup equipment is often needed to determine
if readings are valid. If
the in-place instrumentation is
faulty, it should be replaced, if possible, as soon as a new
instrument can be obtained.
Verification of abnormal readings can also be deter-
mined by other observations. High pore pressures, as in-
dicated by piezometer data in the fill, may also be indicated
by heaving or weaving of the fill under the action of the
hauling equipment. The high pore pressures in the fill
shown by the piezometers may also be indicated by high
water levels in the inclinometer tubes. Movement indicated
by data from the inclinometers or from measurement points
may also be indicated by surface cracking in the fill where
the surface has not been recently disturbed. Most embank-
ment movements involve high pore pressure in combina-
tion with other factors.
A cautious, pessimistic interpretation should be made of
abnormal readings or trends. The questions should always
be: "What can this lead to, and what is the worst that can
occur?" This line of questioning will often show the need
for prompt correction. Possible corrective actions include:
construction of stability berms; drainage by installation of
horizontal drains in the embankment or abutments or by
installation of vertical drains in the embankment or foun-
dation, aided in extreme cases by electro-osmosis;3o partial
removal of the fill to remove a wet layer in the embankment
or foundation; cessation of construction until excess pore
pressures dissipate; and so on. Prompt action to prevent
excessive movement is crucial, as the reworked material
along the slide plane has a much lower strcngth than it had
before sliding. Once significant movement starts, it may be
impossible to stop the movement, and failure may occur.
At Carsington Dam3r one day after a crack opened above
the clay core of the nearly completed I l5-ft (35-meter)-
high dam, the upstream side slid away over about one-third
of the dam's 3800-ft ( I158-meter) length. As a resulr, the
dam was to be demolished and rebuilt.32
 EARTHFILL DAM CONSTRUCTION AND FOUNDATION TREATMENT 351
INSPECTION AND CONSTRUCTION RECORDS
Cleuoe A. Ferzen
lntroduction
The importance of adequate inspection in earth dam con-
struction cannot be overstated. Not one wheel should turn,
nor should one spade of dirt be turned over, unless an in-
spector is present. This rule applies to all shifts and not just
the day shift, as the level and competency of inspection
should be the same at all times. Verification of the ade-
quacy of the foundation treatment, the suitability of mate-
rials, and the placement and compaction over the entire
dam is in the eyes of the inspectors, as test records verify
only very limited areas that represent a minute fraction of
the total dam area and dam materials. Tests take time, and
several lifts could be placed before the test results indicated
that corrective action was needed; whereas.an experienced
inspector can quickly tell from the action of the hauling
and compaction equipment if the moisture content is in the
appropriate range, and if the proper compaction is being
achieved. If not, an inspector can stop the work before ad-
ditional lifts are placed and have tests performed to verify
his judgment. Inspecton are especially needed to watch the
placement and compaction of each lift against abutments
and structures; and they are needed as well to verify that
appropriate materials are placed in each zone and to watch
for unusual developments, which the design engineers
should be made aware of, in the materials or foundation
conditions.
Ouality Control
The responsibility for inspection normally lies with the
owner's resident engineer and staff, or with an engineering
firm retained by the owner. In the late 1960s, a concept
from the U.S. govemment's military procurement pro-
gram, known as contractor quality control (CQC), was
adopted by the Corps of Engineers and the Bureau of
Reclamation. CQC was designed to place the burden of
proof on the contractor for compliance with the require-
ments of plans and specifications. The owner's (the U.S.
government's) program was called quality assurance (QA),
and was intended to check the effectiveness ofthe contrac-
tor's quality-control system. As originally initiated, the
CQC was intended to reduce the number and cost of gov-
ernment inspection personnel. However, experience with
the CQC program has indicated that adequate quality as-
surance requires essentially the same level of inspection
and tests as were accomplished prior to introduction of the
CQC program. The current responsibilities of the contruc-
tor and of the Corps of Engineers for quality control are
described on pages 2-2,24, and 2-5 of reference 3.
According to Corso,33 the Federal Energy Regulatory
Commission (formerly Federal Power Commission) rc-
quires the owner to submit a quality control (QC) inspcc-
tion program for approval by the regional engineer before
construction may start; however, the QC progrdm may not
be administered by the contractor, but must be the respon-
sibility of the owner or a firm retained by the owner. Con-
tractors are not precluded from having their own QC pro-
grams for their own purposes.
A quality control program should be developed for each
project, and it should be tailored for the specific project
with appropriate personnel selected for the specific needs.
Clear lines should be developed as to the contractor's re-
sponsibility and the engineer's responsibility. Earth dams
are one ofa kind, and for them to function as designed, the
foundation and abutments must be prepared correctly
throughout; appropriate materials must be placed in each
zone; and each lift must be moistened and compacted prop-
erly. Hence, the owner must provide adequhte inspection
to ensure that all work is accomplished properly.
Construction History
If a dam is worth building, it is worth the cost of a written
record of how it was built. The construction history is an
invaluable aid in the evaluation of the dam's performance.
in analyzing problems, and in developing remedial mea-
sures; it is also a source of valuable information in le-eal
issues. The resident engineer and key staff members should
be retained at the site until the construction history has been
completed. To ensure accuracy and completeness, all rec-
ords should be kept curent as the job progresses. Daily
records are essential to preparation of the construction his-
tory and in establishing factual information for claims and
disputes.
Photographs are used to record progress and significant
events. High-quality color photographs are made on most
projects, and color processing laboratories can make good-
quality black and white prints from color prints if multiple
printing in black and white is needed. Video tape recording
and time-lapse photography are now used for unique op-
erations such as installation of relief wells, cutoff walls.
and so on. The tapes can be reviewed during preparation
of reports and are very useful training aids for similar op-
erations on future projects.
Records should be kept of all the project equipment, in-
cluding the brand nanre and model. A record should be
made of the downtime for each piece of equipnrent. De-
watering methods and their effectiveness should be re-
corded. Fill operations should be described with special
emphasis on the methods used to compact fill adjacent to
352 ADVANCED DAM ENGINEERING FOB OESIGN. CONSTRUCTION, AND REHABILITATIOT{
abutments and structures. All specialty work, such as fil-
ters, drains, relief wells, cutoff walls, placement of riprap,
and so forth, should be recorded.
The foundation-treatment report, including geological
mapping of the foundation, should be prepared by the site
foundation engineer and by the site engineering geologist.
This report should be prepared as the work progresses while
all the details are known. Numerous photographs should
be included to illustrate fully the foundation conditions and
the methods used to prepare the final surface. Grouting rec-
ords are an essential part of the foundation report. A profile
of the grout lines should be included, showing the takes for
the different stages and the mixes and pressures used. Grout
breakouts and foundation lifting or movements during
grouting operations should be fully described.
Disputes and Modifications
All oral orders given to the contractor should be followed
up immediately with a letter. This is particularly necessary
in case of slow progress, unsatisfactory work or materials,
or other matters that could lead to disputes.
Modifications in contract requirements are almost inev-
itable in earth dam construction. The terms and cost of a
modification should be agreed upon in writing before the
modified work is started; otherwise, the ultimate cost could
be excessive. Its impacts, if any, on the unchanged work
should be included in the agreement. Timely investigation
of changed conditions and early development of the mod-
ified design greatly lessen these impacts on the contractor.
Contract modifications for weather-delay time extensions
should be very carefully examined before being granted be-
cause major increases in cost can accrue to the owner if
such time extensions allow the contractor an additional
construction season to complete the.work. Weather is usu-
ally classified as normal, other than normal, or unusually
severe. Many contracts now are being written to provide
time extensions only for the classification of "unusually
severe," which is defined as weather that mrely happens
and/or is of such severity that it could not be anticipated
by the contractor. To provide a basis for evaluating a con-
tractor's claim for time extensions, a detailed record of the
job site temperature, precipitation, humidity, and wind ve-
locities should be kept for comparison with deviations es-
tablished for the month in which the claim delay occurred.
Appropriate similar data are also needed for evaluating
claims for time delays due to unusual stream flows.
As-Built Drawings
A current set of marked-up as-built drawings shoulci be kept
in the resident office. The official set of as-built drawings
should be drafted from the resident engineer's set when the
work is completed, or when each stage of the work is com-
pleted. Reduced sizes of the as-built drawings should be
included in the construction history.
Distribution of the construction-history reports would
normally include copies to the engineer, owner, pmject
records center, and applicable state and federal regulatory
commissions.
REFERENCES
l. Terzaghi. Karl, "Does Foundation Technology Really Lag?" En-
gineering News-Record, Feb. 15. 1962, pp. 58-59.
2. Bureau of Reclamation, Earth lvlanual, 2nd ed.. U.S. Department of
the Interior. 1974.
3. Corps of Engineers, "Construction Control for Eanh and Rockfill
Dams," EM I I l0-l-19l I, U.S. Depanment of the Army, Jan. t7,
1977.
4. Acker, Richard C., and Jones, Jack C., "Foundation and Abutment
Treatment for Rockfill Dams." tournal of the GT Division, ASCE,
Vol.98. SM 10. Oct. 1972.
5. Barmn, Reginald A., "Aburmentand Foundation Treatment for High
Embankment Dams on Rock," Jourzal ol the GT Dir.ision, ASCE,
Vol.98, SM 10, Oct. 1972.
6. Burke, Harris H., Content. Charles S., and Kulesza. Richard L..
"Current Practice in Abutment and Foundation Treatment," Joumal
of the GT Division, ASCE. Vol. 98. SM 10, Oct. 1972.
7. Pratt, Harold K., McMordic, Roben C.. and Dumas, Roben M.,
"Foundation and Abutments-Bennett and Mica Dams." Joumal of
the GT Division, ASCE, Vol.98, SM 10, Oct. 1972.
8. Stroppini, Elmer W., Babbitt, Donald H., and Struckmeyer, Henry
E., "Foundation Treatment for Embankment Dams on Rock."./oar-
nal of the GT Diyision, /SCE, Vol. 98. SM 10. Oct. 1972.
9. Wallace. B. J., and Iltihon. J., "Foundation Pracrices for Talbingo
Dam. Australia," Joumal of the GT Division, ASCE. Vot. 98. SM
10. Oct. 1972.
10. Walker. Fred C., and Bock, R. W.. "Tqeatment of High Embank-
ment Dam Foundations," Jounnl of the GT Division. ,.|SCE, Vol.
98, SM 10, Oct. 1972.
I l.
Committee on Embankment Dams and Slopes, ASCE. "Foundation
and Abutment Trcatment for High Embankment Dams on Rock,"
Joumal of the GT Division, ASCE, Vol. 98, SM 10. Oct. 1972.
12. Skempton, A. W., "Long Term Stability of Clay Slopes," Giotech-
nique, Yol. 14, No. 2, pp. 77-102. 1964.
13. Bjemrm, Laurits, "Progrcssive Failure in Slopes of Overconsoli-
dated Plastic Clay and Clay Shales," ASCE, Soil lvlech. &. Foun-
dation Division, Vol. 93, No. SM-5. pp. 3-49, 1967.
14. Peck. R. W.. "Srabiliry of Natural Slopes." ASCE, Soil Mech. &
Foundation Division, Vol.93, No. SM 4, pp.403-.117. 1967.
15. Knight, D. K.. "Oahe Dam Geology, Embankment and Cut Slopes."
ASCE, Soil Mech. & Foundation Division, Vot. 89. No. SM 2. pp.
99-t25.1963.
16. Wright, S. G.. and Duncan, J. M.. "Analyses of Waco Dam Slide."
ASCE, Soil Mech. & Foundation Division. Vol.9E. No. SM 9, pp.
869-877. Sept. 1972.
17. Insley, A. 8.. Chauerji. P. K.. and Srnith. L. 8.. "Use of Residual
Strength for Stability Analyses of Embankrnent Foundations Con-
taining Pre-existing Failures." Canaelian Geotech. Journal, Vol, 14.
pp.405-428, 1976.
18. Casagrande, Arthur. "Contrsl of Seepage thmugh Foundations and
Abutments of Danrs." First Rankine Lecture. Gioteehniepe. Sept.
t96r.
19. Houlsby, A. Clive, "Ccment Grouting: Water Minimieing Prac-
tices." /s.raes in Dam Grouring, ASCE, pp, 34-75, lvtay 1985.
20. Lombardi, Giovanni, "The Role of Cohesion in Cement Gmuting of
 EARTHFILL DAM CONSTRUCTION AND FOUNDATION TREATMENT 353
Rock." ICOLD-Fiftr.enth Inrernational Congress on Large Dams,
Lausanne, Swiuerland. 1985.
21. Deere, Don U.. and Lombardi, Giovanni. "Gmut Slurries-Thick
or Thin,': Issues in Dam Grouting, z{5C8. May 1985.
22. Corps of Engineers. "Grouting Technology." EM l tl0-2-3506,
U.S. Department of the Army. Jan. 20. 198.1.
23. Hammer. David P.. pesonal communication to Claude A. Fetzer.
24. Corps of Engineers. "Embankment Criteria and Performance Re-
port, Warm Springs Dam-Lake Sonoma Russian River Basin. Dry
Creek, Califomia," U.S. Army Corps of Engineem, San Francisco
District, Oct. 1983.
25. Hammer. David P.. "Methods for Effective Placcmeni and Cont-
paction Control During Eanh Embankment Construction." unpub-
lished paper presented at 1978 ASCE Convention, Chicago, IL. Oct.
l6-20,1978.
26. American Society for Testing and Materials, Annual Book of ASTM
Standanls, Vol. O{.08, Soi[ and Rock; Building Stones, ASTM.
Philadelphia, PA. 1987.
27. Donaghe. R. T., and Townsend, F. C., "Compaction Chlracteris-
tics of Eanh-Rock Mixrurcs, Report 1: Blcndcd lvlatcnal... Mircl-
laneous Papcr 5-73-25. U.S. Army Engincer Warcrways Expcrimcnr
Station. Vicksburg. MS, 1975.
28. "Caron Compaction Whccls," Bullctin No. CW.6-BI. Carun Com-
pactor Co.. Modcsto. CA.
29. Fetzer. Claudc A.. "Use of Compacted Shale as Dam Embank-
ments." Proceedings Seventh Ohio River Valtcy Soils Seminar.
Lexington. KY, Ocr.8, 1976.
30. Fetzer, Claudc A.. "Electro-osmoric Srabilization of Wesr Bnnch
Dam." Proceedings Confcrence on Stability and fr:rformance of
Slopes and Embankmcnrs, ASCE. Soil l{ech. & Foundations Divi-
sion, Berkeley. CA.pp. 95-12.1. Aug. 22-26. 1966.
31. "British Dam Slip Foretold." Engineering Ncu,s-Recorc!. June 28.
1984. p.22.
32. "British Dam to Be Scrappcd." Engineering Nexs-Rccorr/, Oct. 25.
1984, p. 18.
33. Corso, Ronald A., "Dam Safety Program of rhe Fedcral Eneqlv
Regulatory Commission." Associati<tn of State Dan Safert.O!ficials
(ASDSO) Newsleuer, Vol. No.2, N{ay 1985, pp.7,8, and 20.
 LL
EARTHFILL DAM PERFORMANCE AND
REMEDIAL MEASURES
Cueuoe A. Ferzen
Consuhing Geotechnical Engineer
Cincinnati, Ohio
RrcHeRo W. Knaupn
Civil Engineer
U.S. Bureau of Reclamation
Denver, Colorado
ESTABLISHMENT OF ACCEPTABLE NORMS
CLeupeA. Ferzen
Earth embankments are massive structures that inherently
have movements and seepage. Consolidation of the em-
bankment and the foundation occurs most rapidly during
construction and at a lesser rate for an extended period of
.time
thereafter. The initial filling and its accompanying sat-
uration may temporarily accelerate the consolidation of the
upstream section of the embankment, and initial filling will
also cause downstream seepage to develop. Consolidation
of the embankment and the foundation is accompanied by
transverse and longitudinal movements that may result in
transverse and longitudinal cracks. The predicted amounts
of consolidation, movement, and seepage should be deter-
mined by analyses during the design stage. These analyses
should be reviewed at the end of construction, and modi-
fied if the as-constructed engineering characteristics are dif-
ferent from those assumed during design. For instance,
354
Wrr-lrnu F. Swrcen
Consulting Engineer
Buhl, tdaho
Vice President and Senior Consuking Engineer (Retired)
Stone & Webster Engineering Corporation
Jacr G. Wulpr
President
Wahler Associates
Palo Ako, California
changes in the foundation treatment or achieving a higher
percent of maximum density than was specified may result
in less consolidation, movement, and seepage than were
computed during the design stage. There are no standard
guidelines that can be applied to the tolerable amounts of
movements and seepage, as each embankment has its own
characteristics; however, it is often helpful to compare data
from similar structures with similar foundation and abut-
ment conditions. The analyses and comparative data should
be used to establish the benchmark norms; however, the
inspectors, instmmentation personnel, and evaluators
should always be on the alert, even before the benchmark
norrns are reached, for adverse trends in data plots, cracks,
seepage concentrations, boils, evidence of piping, sink-
holes, and other signs of distress.
 EARTHFILL OAM PERFORMANCE ANO BEMEOIAL MEASURES 355
INITIAL FILLING
Clauoe A. Ferzen
Preparation
The initial filling represents a critical testing of the em-
bankment. The preparation for the initial filling should in-
clude a review of: the design reports, the construction rec-
ords, the operations manual, and the instrumentation data
obtained during construction and during the intermediate
time since the embankment was completed.
Particular emphasis should be given to data obtained
during temporary impoundments or longer-term lower im-
poundments. All springs and weeps that existed before the
dam was constructed and that developed during any im-
poundments should be located on the ground, numbered,
and plotted on large-scale maps; the discharges should be
measured and plotted against pool elevations and rainfall.
Adequate staffing should be arranged for inspection of
the dam, for reading the instruments, and for plotting the
data. To conserve manpower on large projects, the instru-
mentation data should be transmitted to computers with
plotting printout capability at the job site. Lighting should
be set up for that portion of the dam and downstream area
that is to be monitored on a 24-hour basis. A spare set of
instrumentation readout devices should be purchased and
site-tested, as critical data could be lost through failure of
a readout device during the monitoring program. A check
inclinometer should be installed in stable ground on the
abutment (preferably in bedrock), and the accuracy of the
readout device should be verified by taking readings in the
check inclinometer each time readings are scheduled for
the site inclinometers.
On-Site lnspection
The inspection force should include lead personnel who
have a thorough understanding of the embankment design
and construction. The lead personnel should hold training
sessions for the entire force to explain: the design of the
embankment, the operations manual, the locations and pur-
poses of the instruments, the monitoring Program, the per-
formance of the embankment to date, and the duties and
schedules of each inspector.
Personnel should be taken into the field and shown each
instrument and each monitoring point. The operation of all
measurement and recording devices should be explained,
and each instrumentation reader should make trial runs with
the measurement and recording devices. The embankment
and downstream area shoutd be inspected by the entire
party, and current defects such as cracks, depressions, and
so on, should be noted and photographed. Possible signs
of distress should be thoroughly explained, and the most
vulnerable locations for potential distress should be pointed
out. Individual members of the inspection team should be
thoroughly oriented on the importance of their observa-
tions, which may indicate distress not detected by the in-
strumentation.
lnstrumentation Data
The validity of the instrumentation data should be checked
constantly. Instrumentation readers should take a check
reading anytime there is significant divergence from the
previous reading. When significant movements are indi-
cated since the last inclinometer reading, the accuracy of
the readout device should be verified in the check inclino-
meter, and additional inclinometer readings should be ob-
tained. Alignment and settlement data from surveys are es-
pecially subject to elrors' particularly on long
embankments. Site observations are helpful in evaluating
data. For instance, if the alignment survey data indicate a
spread of 0.5 ft (15 cm) between the upstream and down-
stream crest monuments, there should be a wide crack in
the crest. Cross checks should also be made between in-
clinometer data and surveY data.
The schedule for obtaining instrumentation data and for
site observations should be based on site conditions and the
filling schedule. For most projects, a reasonable schedule
may initially require: daily readings on all piezometers, re-
lief-well discharges, and spring discharges; twice-daily
walk-over inspections of the embankment and downstream
area; weekly readings on inclinometers and extensometers;
and monthly surveys of alignment and settlement monu-
ments. As the filling continues' the monitoring schedule
should be revised to meet the needs of the evaluators. On
rapid-filling projects, such as pumped storage' the instru-
mentation readings must be obtained at a much morc fre-
quent rate.
Evaluation
Evaluation of the data should be made daily. Abnormal
data should not be discarded until they are proven to be
erroneous. The data evaluation must give due considera-
tion to the geological site condltions and the embankment
design; and, in this context, the data should be projected
to determine the ultimate condition or worst{ase condi-
tion. For instance, a rise in the downstrcam valley-bottom
piezometers at various intermediate pool levels should be
projected to the scheduled operation pool levels to deter-
mine whether excessive uplift would exist at those levels.
Further projections of the piezometric levels should be
 356 ADVANCED DAM ENGINEERING FOR DESIGN, CONSTRUCTION, AND REHABILITATION
made to the maximum flood-pool level for the same pur-
pose. Projections of the data in this manner will indicate
whether or not the filling should be stopped at some inter-
mediate level until remedial action, such as installation of
relief wells or drains, is accomplished. During the filling
operation, sudden increases may occur in the piezometric
levels, in the seepage flows or in the sediment content of
LONG.TERM MONITORTNG
Cleuoe A. Ferzen
Ifno incidents have occurred during the initial filling, there
is a tendency to downgrade or de-emphasize. long-term
monitoring. However, as long as water is impounded in the
reservoir, the potential exists for incidents to occur from
slope failures due to saturation and/or drawdown, under-
mining or piping failure from long-terrn seepage, and so
on. Hence, it is necessary to establish a long-term moni-
toring program to ensure the safety of the dam.
Value of Continued On-Site Monitoring
The powerplant, the outlet-works gates, and the spillway
gates at many dams are operated remotely from some other
central station; and it is possible to collect instrumentation
data electronically and transmit the information to a central
computer without having anyone at the site. However, therc
is no substitute for on-site monitoring to ensure the safety
of a dam. Even monitoring with a television camera is not
considered an adequate substitute for on-site inspections.
In practice, the dams at dry flood-control reservoirs only
require periodic walk-over inspections except during stor-
age periods; however, for storage reservoin, frequent in-
spections of the dam, abutments, and downstream area are
advisable. In many cases, on-site personnel have alerted
authorities to problems that could not have been detected
by instrumentation. For example, the detrimental cloudy
seepage at Wolf Creek Dam,r which retains the largest-
capacity reservoir east of the Mississippi River in the
United States, was first observed by a powerplant operator
on a Sunday moming; the flow through the turbines had
the seepage, and it may be advisable temporarily to lower
the pool while investigations are made to determine the
cause of the increases. The evaluators should have full au-
thority to intemrpt the filling schedule ro make investiga-
tions or to do remedial work, as the basic purpose of the
monitoring program is to ensure the safety of the dam, and
not to meet a predetermined filling schedule.
been shut down, and the cloudy seepage was observed dis-
charging into the shallow tailwater from joints in the lime-
stone bedrock.
Piezometers cannnot detect the grain-by-grain removal
that leads to underground piping. In many cases, the first
knowledge that a hole has eroded beneath or through an
earth dam occurs when a sediment-laden spring erupts
downstream. The sooner the spring is discovered, the
sooner it is that action can be taken to save the dam. It is
impossible to observe a dam for 24 hours each day on a
long-term basis, but the chances ofobserving distress signs
in time to take corrective action are greatly increased if
observations are made daily rather than weekly or monthly
when piezometers are read. All site personnel including
grounds keepers should be trained to look for distress signs
and to report such signs promptly to appropriate authori-
ties. The better trained observers are, the more likely it is
that distress signs will be observed.
Monitoring and Evaluation
The training of personnel, detection of criticat factors, va-
Iidity of instrumentation readings, and evaluation of data
were covered above under "Initial Filling." Continued
timely evaluation of the instrumentation data is mandatory;
otherwise, the manpower and cost of collecting the data
and maintaining the system are wasted. Collecting data and
filing them in a drawer may give a false sense of security
to the owner, and to the downstream public, and may be
indefensible in a court of law after a failure.
 EARTHFTLL OAM PERFORMANCE AND REMEDIAL
REMEDIAL MEASURES
CLeuoe A. Fe'rzrn
Necessity for Timely Action
The amount of movement and seepage that an embankment
will tolerate without failure depends on a complex inter-
action of the site geology and the characteristics of the as-
constructed embankment. Embankments constructed of
flexible materials and founded on soft sediments can sus-
tain appreciably greater movements without failure than
brittle embankments founded on rigid foundations.
The critical factor is to determine whether the move-
ments are normal movements or are on a continuing or ac-
celerating path that could lead to failure. Once the move-
ments pass a critical point, it may not be possible to stop
them because of the massiveness of the structure; hence, it
is imperative that timely efforts be made to reduce the driv-
ing force or provide additional support before the critical
point is reached. For example, if detrimental movement of
the upstream slope is being caused by drawdown of the
reservoir, the drawdown should be stopped immediately,
and an underwater free-draining berm should be placed at
the upstream toe to provide stability for the future draw-
down. If movement of the downstream slope is being
caused by a high seepage line, the reservoir should be
drawn down, if possible. and drains should be installed in
the fill to reduce uplift, and/or a free-draining stabilization
berm should be provided for the downstream slope. Other
options include removal of a few feet from the embank-
ment crest to reduce the driving force on the slide.
With regard to seepage, earth embankments founded on
well-graded pervious materials may experience considera-
ble underseepage without detrimental effects: whereas a
small amount of seepage carrying fines from a critical sec-
tion of the embankment or foundation can cause failure if
quick action is not taken. Wolf Creek Damr was saved by
an emergency foundation grouting program. Both Fonte-
nelle Dam2 and East Branch Dam3 were saved by emer-
gency drawdown of the reservoir. However, the erosion of
material out of Teton Dama was so rapid, and the river
inflow was so large, that the reservoir could not have been
drawn down to prevent failure. Although not all dams can
be saved after detrimental seepage is detected, many dams
can be saved by quick and appropriate action.
Seepage Problems
Detecting and locating the path of detrimental scepage are
very difficult problems. Seepage occurs at all earth dams,
whether visible or not. At most dam sites it is difficult to
collect and measure all the seepage, as much of it may be
occurring underground through cavernous formations or
MEASURES 357
through pervious'foundation materials with littlc or no sur-
face discharges. Seepage carrying fines from the embank-
ment or foundation may continue for an extended period of
time without any visible effects on the dam until sulficient
material is removed to permit formation of an upstream
sinkhole. This sinkhole permits the reservoir to have free-
flow access under full reservoir head into the eroded pipe,
thus causing a dramatic increase in flow through the pipe
probably accompanied by development of a large down-
stream spring. The increased flow accelerates the removal
o[ material from the dam and foundation, and can result in
a very rapid failure of the embankment, as happened at
Teton Dam.{
The difficulty facing the monitoring staff is to detect det-
rimental seepage before it develops into a major irreversi-
ble failure mode. Without any surface expression of a de-
veloping piping problem, the monitoring staff may assume
that no problem exists. However, once impounded in the
reservoir, water will forever seek out the easiest flow path
to the downstream area; and if there are any defects in the
embankment or foundation, the water will find those de-
fects that afford the easiest seepage path to the lower ele-
vation. Hence, the monitoring staff, in cooperation with
the designers, should review the geological site conditions,
the foundation treatment, the embankment design, and
construction records to determine whether conditions are
favorable for piping and detrimental seepage to occur. If
conditions do favor piping, it should be assumed that pip-
ing is occurring until proved otherwise. To prove that pip-
ing is not occurring, extensive investigations may be re-
quired, including but not limited to the follow'ing:
l. Sediment and quantity measurements in all down-
stream springs, seepage flows, and flows in under-
ground rock fissures.
2. Periodic walk-over and aerial surveys of the down-
stream area to determine if springs and wet areas are
increasing in size.
3. On-site inspections to determine if depressions or
sinkholes are developing.
4. Equipotential plots of the piezometric data to deter-
rnine if relatively high piezometric hcads extend
downstream in any area, as this condition may indi-
cate an opcn passageway either through or beneath
the core trcnch.
5. Devclopment of temperature-contour plots ftom
tempcrature measurements in the piezomcters to de-
temrine if the colder water frotn the deepest part of
the reservoir is feeding any piezometers.
6. Dye and saline tests to check the path and velocity of
 358 ADVANCED DAM ENGINEERTNG FOR DESIGN, CONSTRUCTTON, AND REHABILITATION
seepage flows from the reservoir to piezometers and
to downstream springs.
7. Drilling in suspect areas to determine the presence of
cavities or soft spots and to install additional piezo-
meters.
8. Partial removal of the embankment to locate a defect,
as was accomplished at one dam in the Conewango
Creek Watershed.5
If it is determined that piping is occurring, the foregoing
procedures will help establish the most likely seepage path,
which may deviously pass through various materials in the
embankment and foundation. The conditions will dictate
whether the reservoir needs to be drawn down to avoid an
emergency and/or to complete remedial action. If the
seepage path is through the foundation rock (valley bottom
or abutments), the following alternatives may be consid-
ered:
l. Grouting-cement, chemical or asphalt.
2. Cutoff walls-concrete-element wall as installed
at
Wolf Creek Damr or concrete-panel wall as installed
at Walter F. George Dam.6
3. Removal of embankment and treatment of rock to
permanently plug the voids in the rock, as was done
for one dam in the Conewango Creek Watershed.s
Grouting may be only a temporary solution. The grout
fills only the existing voids in the rock fissures, and the
grout in many cases may be surrounded by soil or highly
weathered rock, which could be susceptible to removal in
time by the seepage. Installation of a concrete cutoff wall
provides a nonerodible barrier through the weathered, cav-
emous, or open-jointed rock; this type of wall is consid-
ered permanent. AII work through the embankment must
be done with extreme care to prevent damage from exces-
sive hydrostatic pressure that could cause splitting or lifting
of the embankment. The watchword is: "First, do no
harm."
Removal of a section of the embankment and proper
treatment of the foundation have been used for seepage
problems at the aforementioned dam on Conewango Creek
Watershed and for the right abutment at Fontenelle Dam.?
Where the seepage path passes through overburden be-
neath the embankment and is not related to bedrock defi-
ciencies, the following alternatives may be considered:
l. Relief wells along the downstream toe to reduce the
pressure and to collect the seepage in a controlled
manner.
2. Reverse filters graded to prevenr piping and weighted
to prevent heaving ofthe foundation.
3. Cutoff wall through the embankment and the seepage
strata, tied into a thick impervious stratum or into
bedrock-
4. Cement or chemical grouting, depending on the size
of void and type ,rf foundation material.
In many cases, reverse filters are used to fill in low-lying
areas where seepage occurs: and relief wells are used to
relieve the pressures at depth. For earth embankments
founded on deep pervious alluvial deposits, impoundment
of the reservoir often raises the downstream groundwater
to the ground surface; this often causes concentrated seep-
age areas, particularly in the outlet-works discharge chan-
nel, which can be controlled by lining the bottom and chan-
nel side slopes with stone prc,tection underlain by properly
graded filters to prevent removal of fines from the foun-
dation. For bleeding abutments, horizontal drains may be
used to collect the seepage back of the face.
Cutoff walls arc very expensive and are used only as a
last resort. Steel sheetpile walls have been used as cutoff
in some dams; but if the piling is driven to a depth of more
than 20 ft (6.1 meters), or even less deep in highly com-
pacted materials, it is apt to be driven out of interlock and
thus would not be effective. Because of corrosion, sheet-
pile walls may not be permanent. A slurry-trench cutoff
consisting of graded soil placed into a bentonitic-slurry-
filled trench is a possible solution, but extreme care is
needed to prevent pockets of slurry from being trapped in
the backfill; furthermore, the portion of the slurry trench
that is located above the saturation line in the embankment
may dry out and crack. Another altemative is to use a ce-
ment-bentonite slurry to support the trench walls and as
the permanent seepage barrier; however, after initial hard-
ening the portion of the wall above the embankment satu-
ration line may dry out and crack, and would not provide
an impervious barrier for higher reservoir levels. A per-
manen[ type of wall would be a concrete wall as described
for bedrock foundation-seepage problems.
Where the seepage path passes through the embankment
and is not related to foundation conditions, the following
alternatives may be considered:
l. Reverse filters on the downstream seepage area to
control piping temporarily until permanent repairs can
be made.
2. Cutoff walls as described above for foundation-
erburden seepage.
3. Cement or chemical grouting, depending on the size
of the void and the type of embankment materials.
4. Removal and replacement of the faulty embankment
section after drawdown.
5. Reverse filter on downstream slope overlain by but-
tress fill.
 EARTHFILL OAM PERFORMANCE ANO REMEDIAL MEASURES 359

Cracking and Slides
Cracking sometimes occurs along the crest and along the
slopes during initial filling. These cracks are usually caused
by adjustment of the embankment to the water load and to
differential settlement of the embankment zones and/or
foundation as the embankment becomes saturated. An ex-
cellent description of the cracks and analyses of their causes
for a high earth and rockfill dam is presented in a paper by
Patrick on Round Butte Dam.8 Most cracks are longitudi-
nal, but transverse tension cracks often develop near the
abutments as the higher valley section of the impervious
core settles more than the abutment sections settle. Trans-
verse cracks often occur across the crest where there is an
abrupt change in the underlying abutment contact, that is'
where the descending slope abruptly changes from a gentle
to a steep slope. Cracks, slides, and crest slumping may
also be caused by earthquakes. As previously discussed,
cracks and slides may be caused in the upStream slope by
reservoir drawdown and in the downstream slope by a high
seepage line. Other possible causes of instability of the
downstream slope include excavations that remove toe sup-
port and excessive hydrostatic uplift in a critical foundation
layer. Cracks and slides in either slope can also be caused
by excessive stress on a weak layer in the embankment or
in the foundation.
The evaluator must determine the cause of any observed
cracks, or of movements detected by the instrumentation,
and determine whether these adverse factors could lead to
a failure of the embankment. As prcviously indicated,
timely action may be needed to Prevent failurc. Investiga-
tions that may be needed to evaluate these conditions in-
clude: daily measurements of the width and length of the
cracks; measurement of joint movements in conduits pass-
ing through or beneath the embankment; measurement of
movements of bridges or other structures located on the
embankment; surveys of stakes on the embankment, abut-
ments, and downstream area: frequent readings of the
available inclinometers; and, where inclinometers are not
available, quick installation of thin-walled rigid plastic pipe
in a drilled hole for probing with a straight rod to determine
the location of the slip plane. Plots of the measurement
data versus other factors such as reservoir rise or fall and
pertinent piezometric data are helpful in projecting the ul-
timate progression of the cracks. A graph of conduit move-
ments and height of fill versus time for West Branch Dame
is shown in Fig. ll-1.
If it is determined that.the cracks are not indicative of
potential slides, action should be taken to obliterate or fill
the cracks to prevent them from filling with water. For
cracks that are only l0 to 12 ft (3 to 3.7 meten) deep'
excavation can be made along the crack with a backhoe,
and the excavated material can be used to refill the trench
in a compacted state. Deep cracks can be filled with a high-
fluid nonshrinkable grout, preferably a chemical grout that
remains somewhat flexible.

s
c '5
9
t
U
o o
c co
.9
o
,u, .\ | Stsolir
15-16
o
o
-+-Johl
E
F

U' /-lora s.paoriLn

Figure ll-1. West Bmnch Dam-conduit movements, Oct.22 to Nov.25' 196'{.

 360 ADVANCED DAM ENGINEERING FOR OESIGN, CONSTRUCTION, AND REHABILITATION
If it is determined that the cracks do indicate potential
slides, timely actions as previously described should be ini-
tiated immediately to reduce the driving force and/or to
provide additional support. [f th'e movement plane is in the
bedrock, in either the foundation or the abutments the fol-
lowing remedial actions should be considered:
l. Vertical or angled relief drains into the foundation.
2. Horizontal or slightly inclined drains into the dis-
tressed abutments.
3. Berms to buttress the embankment or the abutment.
4. Temporary removal of the embankment crest to a
depth where the movement stops.
5. Removal of the embankment and the weak rock strata
and reconstruction of the embankment.
If the movement plane is in an earth foundation or an
earth abutment, the following remedial acrions should be
considered:
l. Installation of retief wells in the foundation, with
pumping used on a temporary basis to increase their
effectiveness.
2. Installation of horizontal drains in the abutment.
3. Installation of stone columns in the foundation.
4. Electro-osmosis to reduce the pore pressures and in-
crease strength.
5. Placement of berms to buttress the embankment or
the abutment.
6. Temporary removal of the embankment crest to
depth where the movement stops.
7. Removal of the embankment and removal of the weak
shallow strata, as planned for Carsington Dam.l0
Where the weak strata are too thick or too deep to be
removed, other alternatives may be used to strengthen
the base of the embankment, as described in the case
history for Mohicanville Dike (below).
If the movement plane is in the earth embankment, the
following remedial actions should be considered:
1. For a submerged upstream slope, a granular upstream
toe berm can be placed through the water. In an
emergency, a portion of the crest could be removed
temporarily.
2. For movement within the downstream section:
a. For a high seepage line, install horizontal drains;
or for surface sloughing, install drainage trenches
backfi lled with free-draining material.
b. For a weak layer in the embankment, install sand
drains or stone columns through the layer to pro-
vide drainage and to accelerate consolidation.
c. Install a berm, preferably of free-draining mate-
rial.
d. Temporarily remove the embankmcnt crcst to a
point where the movement stops.
Settlement
The evaluator must determine whether large areas of set-
tlement are due to normal consolidation of the embankment
and foundation, or the settlement is a forerunner of a deep-
seated slide. Essentially the same investigations or checks
should be made as previously discussed for cracks and
slides. The rate of normal consolidation should decrease
with time, whereas settlement due to shear displacement in
the embankment or foundation may increase with time,
particularly if the causative loading is holding steady or is
increasing. Corrective measures may not be needed for set-
tlement due to normal consolidation unless the crest has to
be raised to maintain the required freeboard. However,
corrective measures for shear displacement would be in ac-
cordance with those previously described for slides.
Small areas of settlement may result from many factors,
including but not limited to: piping into cavernous rock;
piping into improperly graded drainage layers; piping along
animal burrows; piping along tree roots or along rotted-out
tree roots; piping of dispersive clays into drainage layers;
piping of embankment materials into conduits; settlement
of poorly compacted backfill along structures or abut-
ments, possibly accompanied by piping into the voids cre-
ated by the settlement; and localized zones of poor com-
paction. The investigations needed to determine the cause
a of the local settlement would be based on an assessmenr of
the site conditions; but they could include measurement of
the rate of settlement, drilling of test holes to determine the
relative density of the materials or to locate voids, drilling
of large-diameter holes for visual inspection of the em-
bankment and foundation materials, dye tests in drill holes,
and so forth.
The appropriate remedial measures would depend on the
causative conditions, and could range from excavation and
recompaction of a local area to removal of a substantial
part of the embankment for proper foundation treatment or
replacement of faulty filters.
CASE HISTORIES
Mohicanville Dike No. 2
Cleuoe A. Frrzen
The Huntington District of the U.S. Army Corps of Engi-
neers used a layer of steel mesh at the base of N{ohicanville
Dike No. 2 to provide the srrengrh needed to complete rhe
24-ft (7.3-meter)-high dike. Construcrion of the original
dike was halted in 1937 at a height of 8 ft (2.4 meters)
after repeated failures occurred before the design height
 EARTHFILL DAM PERFORMANCE AND REMEDIAL MEASURES 361
Impervloua FIII
zold Dtke rllt F
Membrane - -
F-Cutoff--__
\
)sott
/
Figure I l-2. Ivlohicanville Dike No. 2-reconstructed section. After Frrnks.rl
could be achieved. Recent investigations and analyses in-
dicated that the failures were primarily due to a weak foun-
dation consisting of peat and soft clay.rr lSee the typical
section in Fig. ll-2.)
Because of real estate limitations at the site, sufficient
width for adequate-stability berms could not be provided
to develop an end-of-construction factor of safety of 1.3.
Limit equilibrium stability analyses performed by Law En-
gineering and Testing Company indicated rhat an addi-
tional maximum force at the levee base of 30,000 lb
( 13,608 kg) per lineal foot (0.3 meter) of levee was re-
quired to provide a factor of safety of 1.3. A finite-element
analysis was performed by Waterways Experiment Station
to determine the most appropriate reinforcing material and
to develop instrumentation predictions for monitoring con-
struction. Steel wire mesh was chosen as the reinforcing
material in preference to polyester filter cloths, fiberglass,
and Kevlar Geo-fabrics. Steel mesh was the only material
that would provide the needed strength in one layer. Ad-
ditionally, much smaller movements would occur within
the embankment and foundation with steel mesh, compared
with using the other materials.
An upstream membrane-lined slurry trench was provided
through the peat and into the soft clay to cut off foundation
seepage. A downstream horizontal drainage blanket was
provided with a collector pipe at the downstream toe to
intercept any seepage that bypassed the cutoff. The cutoff
trench was considered necessary because of observed
seepage during the flood of 1969, when the reservoir was
impounded against the dike to a depth of approximately 3
ft (0.9 meter). Based on permeability data obtained in the
peat at the centerline, the seepage should have been I
gal./min. (0.06liters/sec); whereas rhe observed secpage
was approximately 500 gal./min. (31.5 liters/sec). Most
of the seepage was believed to have occurrcd through
cracks, fissures, or pervious zones in the peat and in the
embankment. These seepage paths may have been partly
caused by large displacements during the original dike con-
struction. A high-density polyethylene, 100-mil (0.25-cm)-
Er. 984
Steel Mesh
El. 960
ctay < 60,
thick membrane, was provided on the upstream side of the
slurry trench to maintain the integrity of the cutoff for the
settlements and displacements expected during reconstruc-
tion of the dike. The 2.5-ft (0.76-meter)-wide by 30-ft
(9.1-meter)-deep slurry trench was backfilled with soil ob-
tained from the existing dike that had been mixed with ben-
tonite.
The old existing dike was degraded to a point approxi-
mately 2 ft (0.6 meter) above the surrounding ground, and
the steel mesh was laid on the degraded surface. The steel
mesh was 2 x 6-in. (5 x l5-cm) Dl2/D4.5, fabricated
in S-ft (2.4-meter)-wide rolls. The rolls were strai_ehtened
at the job site and cut into 160-ft (48.8-meter) lengths, the
length needed to extend from the upstream toe to the down-
stream toe. The 2-in. (5-cm) spacing was laid perpendic-
ular to the dike axis along the direction of maximum stress,
with the wire having an area of 0. l2 in.z (0.77 cmr;. The
6-in. ( l5-cm) spacing was laid parallel to the axis and had
an area of 0.045 in.: 10.29 cm21. The foundation and the
first overlying lift were treated with lime to reduce the risk
of the wire's corroding. Long-term stability analyses in-
dicated that the wire mesh would not be required for sta-
bility after ten years, when consolidation of the foundation
soils would have increased their shear strength sufficiently
for stability.
The cutoff trench was constructed in 1983. In 1984 the
dike was degraded to El. 960 ft (292.6 meters), the wire
mesh was installed, and the dike was raised to a height of
I I ft (3.35 meters). The dike was cornpleted between June
and September 1985. The embankment, foundation, and
wire mesh were extensively instrumented to measure hor-
izontal and vertical movemcnts, pore pressures. and stresses
within the steel wire mesh. The movemcnts were in good
agreerncnt with thosc predicted from finite-element anal-
yses. The maxirnum stress recorded for the wire mesh was
34,200Ib ( 15,513 kg) per lineal floot (0.3 meter). By Feb-
ruary 1986, all instrumentation d.lta had peaked, and their
values were decreasing. A construction and pedormance
report will be issued by the Huntington District.
362 ADVANCEO DAM ENGINEERING FOR DESIGN, CONSTRUCTION, AND REHABILITATION
Logan Martin Dam
WrLLrnu F. SwrcER
Logan Martin Dam is located on the Coosa River southeast
of Birmingham, Alabama. Before construction of the dam
the river was flowing across exposed bedrock at about El.
395 ft (120.4 meters). It was bordered on the right by a
narrow floodplain that overlapped moderately rising cherty
residuum developed by deep weathering of the underlying
dolomite. On the left of the river there was a broad flood-
plain about 900 ft (275 meters) wide, and then again cherty
residuum. These floodplain deposits are about 20 to 30 ft
(6.1 to 9.1 meters) thick and grade downward from silty
clays at the surface to sand and gravel above the top of
rock. A number of sinkholes were present in the flood-
plains upstream of the dam site. The bedrock is nearly flat-
lying dolomite of the Lower Knox Formation. It contains
some limestone, less than l0%, and minor impurities in-
cluding chert, silt, and sand. There has been extensive so-
lution activity, which has developed along several joint sets
and along steeply dipping minor faults. The upper portion
of the rock is pinnacly, with many isolated blocks sur-
rounded by residuum and alluvial deposits. Solution ap-
\-a,
EF
'FF
Be o Rocx
-ri-;-
---_/
Figure I l-3. Logan Martin Dam-plan and sections.
pears to have been dominantly vertical with local widening
at more readily soluble beds. Solution effects extend to
depths of more than 400 ft (122 meters).
The dam was completed in 1964. lts plan is shown in
Fig. I l-3. It includes a powerhouse on the right bank with
adjoining spillway. The right floodplain was removed to
construct these structures. An embankment section founded
on residuum adjoins the powerhouse on the west (right)
side. Extending east (left) from the spillway is an embank-
ment section constructed across the rest of the river chan-
nel. This is approximately 600 ft ( 183 meters) long. This
embankment continues across the floodplain and then on
cherty residuum. The total length of the east embankment
is about 4650 ft ( 1417 meten). Typical cross sections of
the embankments as built are shown in Fig. I l-3.
Shortly after the reservoir was filled, a number of springs
and discharges in the river were noted downstream of the
dam. In 1968 a sinkhole developed in the dam. It was lo-
cated just downstream of the crest and about 400 ft ( 122
meters) east of the end of the spillway. This sinkhole was
treated by filling it with quarry-run stone. Through borings
around the chimney sink, sand was sluiced into voids in
the embankment. Discharge from a spring at the toe of the
dam and opposite the sinkhole, over a period of time, had
Oct c r ual CoxsrRucrroto
- --Mo o r Fl cA Tro N 117 8'71
A Comcecreo Lneert CuaY
BFr-ooo Puarn DePosrrs
C QuaqRv Rurt Rocrc
D Peocr,ssep l?ocr<
-Srcrrox A-A
@
Sca ue Fec r
VrB,r c s g6'
-'1
Aecrrox B-R