Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 93
Groundwater Contamination
and Aquifer Reclamation at the
Rocky Mountain Arsenal,
Colorado
6
INTRODUCTION
LEONARD F. KONIKOW
U.S. Geological Survey
DOUGLAS W. THOMPSON
U.S. Army Corps of Engineers
AB STRACT
Groundwater contamination at the Rocky Mountain Arsenal, Colorado, is related to the disposal of liquid
industrial wastes and to industrial leaks and spills that have occurred during the 40-yr history of operation
of the Arsenal. From 1943 to 1956 the liquid wastes were discharged into unlined ponds, which resulted in
contamination of part of the underlying alluvial aquifer. Since 1956, disposal has been accomplished by
discharge into an asphalt-lined reservoir, which significantly reduced the volume of contaminants entering
the aquifer. In the mid-1970s toxic organic chemicals were detected outside of the Arsenal in the alluvial
aquifer. The Colorado Department of Health issued three orders, which called for (1) a halt to unauthorized
discharges, (2) cleanup, and (3) groundwater monitoring. Subsequently, a management commitment was
made to mitigate the problem. A pilot groundwater containment and treatment system was constructed in
1978; it consists of (1) a bentonite barrier and several withdrawal wells to intercept contaminated groundwater
along a 1500-ft length of the northern Arsenal boundary, (2) treating the water with an activated carbon
process, and (3) injecting the treated water on the downgradient side of the barrier through several recharge
wells. Because of the success of the pilot operation, it is being expanded at present to intercept most of the
contaminated underflow crossing the entire north boundary. However, boundary interception alone cannot
achieve aquifer restoration at the Arsenal. It is anticipated that the overall final program will also have to
include elements of source containment and isolation, source elimination, process modification to reduce the
volume of wastes generated, and development of alternative waste-disposal procedures that are nonpolluting.
A variety of alternatives have been proposed and are currently being evaluated to determine the most feasible
for implementation. The research, planning, and design studies that are necessary to achieve the reclamation
goal at the Arsenal illustrate that an effective aquifer restoration program is difficult to design and expensive
to implement.
The contamination of a groundwater resource is a serious prob-
lem that can have long-term economic and physical conse-
quences because in most cases the problem is neither easily
nor quickly remedied. Wood (1972) concluded, "The most sat-
isfactory cure for groundwater pollution is prevention." In many
93
cases where a serious groundwater contamination problem ex-
ists, the single most important remedial action that can be taken
is to eliminate the source of contamination. But even then,
contaminants already in the aquifer will continue to migrate
and spread unless some action is taken to immobilize, neu-
tralize, or remove them. Hence, there is often a need to clean
up or restore contaminated aquifers.
OCR for page 94
94
The "restorability" of a contaminated aquifer is dependent
on the hydrogeologic and geochemical properties of the affected
aquifer and on the chemical and physical properties of the
contaminant. Restoration of a contaminated aquifer is neither
technically nor economically feasible in many cases. Factors
frequently hindering restoration include (1) the slow diffusive
nature of groundwater flow, (2) the difficulty of defining sec-
ondary permeability effects, (3) the generally low oxygen con-
tent and lack of biologic reactivity in groundwater, (4) the re-
tention of some chemicals in the aquifer because they tend to
be sorbed by minerals in the rocks making up the aquifer, (5)
the lack of transferability of some restoration techniques from
one site to another, and (6) the lack of knowledge about the
source of the contamination.
Effective aquifer restoration programs, if technically feasible,
are both difficult to design and expensive to implement. Never-
theless, in response to public or governmental demands for
positive action in clearly documented cases where groundwater
contamination threatens public health, aquifer cleanup pro-
grams are being required and instituted more frequently. Some
programs are being financed and operated by the federal gov-
ernment. Examples include the Rocky Mountain Arsenal, Col-
orado, where irrigation and domestic water-supply wells in
adjacent areas have been contaminated from industrial wastes
stored at the Arsenal, and also Wurtsmith Air Force Base,
Michigan, where toxic organic solvents used in aircraft main-
tenance have entered and spread through the underlying aqui-
fer. Other programs may be implemented because of violations
of federal regulations. For example, a recent justice Depart-
ment suit was filed in North Carolina under the imminent
hazard provision of the Resource Conservation and Recovery
Art the Alit acts that the rl~f~ncl~nt~ "
~ an- 7A%~7 FUJI AM . . . permanently restore
the aquifer to a condition commensurate with safe human use"
(Hazardous Waste News 2~2), [an. 21, 1980, p. 12~. As an
example of an aquifer restoration program being initiated be-
cause of state regulations, a chemical company in northern
Michigan has come to an agreement with the state of Michigan
to remove the contaminants from the soil and groundwater at
their former dump site; the projected cost is $12 million to $15
million (The Wall Street fournal, Sept. 25, 1981, p. 48~.
General management options for restoring water quality in
aquifers currently available include the following: (1) eliminate
the source of contamination but allow restoration to proceed
only through natural flushing, dilution, and geochemical or
biological reactions; (2) accelerate removal of contaminants
through withdrawal wells, drains, or trenches; (3) accelerate
flushing with artificial recharge; (4) install "impermeable" bar-
riers to contain a contaminated area; (5) induce in situ chemical
or biologic reactions that would neutralize or immobilize the
contaminant; and (6) excavate and remove the contaminated
part of the aquifer. The selection of the best approach for a
particular situation requires the ability to predict changes in
flow and chemical concentration in the aquifer for each possible
management alternative. This in turn requires both adequate
field data to describe the aquifer systems and the development
of accurate simulation models to define the groundwater flow
system, pollutant-transport mechanisms, and nature and rate
of chemical or biological reactions.
LEONARD F. KONIKO\V and DOUGLAS W. THOMPSON
N
1
_q
o
o1 /
ol JO
u,1
Cal
~1
~1
r---
q
L ~
,/ ~ Brighton
DENVE;
COLORADO
A R A P A H O E C O
.
1!
~ - _ _. _ _. _ __
PLY O ~ 10 MILES
O 5 10 KILOMETERS
____—— .
/ 1
red 1
FIGURE 6.1 Location of study area (boundaries are approximate).
l
1
This chapter focuses on the groundwater contamination
problem at the Rocky Mountain Arsenal, which is located near
Denver, Colorado (see Figure 6.1~. This area is well suited for
serving as a case study to illustrate data requirements, inves-
tigative approaches, and management options related to the
reclamation of contaminated aquifers because (1) the 40-yr his-
tory of groundwater contamination is relatively well docu-
mented in the scientific and engineering literature; (2) the
geology and hydrology of the area are fairly well known; (3)
adequate, though limited, water-quality data are available to
calibrate numerical simulation models; (4) the locations and
strengths of contaminant sources can be approximately recon-
structed; (5) a management commitment has been made to
aquifer reclamation; and (6) construction, operation, and eval-
uation of a pilot reclamation system at the Arsenal have been
completed.
DESCRIPTION OF STUDY AREA
History of Contamination
The Rocky Mountain Arsenal has been operating since 1942,
primarily manufacturing and processing chemical warfare prod-
OCR for page 95
Contamination and Aquifer Reclamation
ucts and pesticides. These operations have produced liquid
wastes that contain complex organic and inorganic chemicals,
including a characteristically high chloride concentration that
apparently ranged up to about 5000 mg/L.
The liquid wastes were discharged to several unlined ponds
(Figure 6.2), resulting in the contamination of the underlying
alluvial aquifer. On the basis of available records, it is assumed
that contamination first occurred at the beginning of 1943.
From 1943 to 1956 the primary disposal was into pond A.
Alternate and overflow discharges were collected in ponds B.
C, D,andE.
Much of the area north of the Arsenal is irrigated, both with
surface water diverted from one of the irrigation canals, which
are also unlined, and with groundwater pumped from irrigation
wells. Some damage to crops irrigated with shallow ground-
water was observed in 1951, 1952, and 1953 (Walton, 19617.
Severe crop damage was reported during 1954, a year when
the annual precipitation was about one half the normal amount
and groundwater use was heavier than normal (Petri, 1961~.
Several investigations have been conducted since 1954 to
determine both the cause of the problem and how to prevent
further damage. Petri and Smith (1956) showed that an area of
contaminated groundwater of several square miles existed north
and northwest of the disposal ponds. These data clearly indicate
39°50t _
1
,/ D ~B~ ~ I
I r 1
i l i
L. .. |
J ROCKY MOUNTAIN ARSENAL BOUNDARY ~
~ _ . _, _ . _ . . _ _ . _ . _ .
O 1 2 MILES
1 I ~
1 1 1
O 1 2 KILOMETERS
EXPLANATI ON
............. 1
::::::::::::::::::1 Irrigated area do Unlined reservoir
-
· Irrigation well ~ Lined reservoir
FIGURE 6.2 Major hydrologic features; letters indicate disposal-pond
designations assigned by the U.S. Army (Konikow, 1977).
95
that the liquid wastes seeped out of the unlined disposal ponds,
infiltrated the underlying alluvial aquifer, and migrated down-
gradient toward the South Platte River. To prevent additional
contaminants from entering the aquifer, a 100-acre (0.045-km2)
evaporation pond (reservoir F) was constructed with an asphalt
lining in 1956 to hold all subsequent liquid wastes. Although
the liner eventually failed, even if the lining were to have
remained totally impervious, this new disposal pond in itself
would not eliminate the contamination problem because large
amounts of contaminants were already present in and slowly
migrating through the aquifer.
From about 1968 or 1969 through about 1974, pond C was
maintained full most of the time by diverting water from the
freshwater reservoirs to the south. This resulted in the inf~l-
tration of about 1 ft3/sec (0.03 m3/sec) of freshwater into the
alluvial aquifer. This artificial recharge had the effect of diluting
and flushing the contaminated groundwater away from pond
C faster than would have occurred otherwise. By 1972 the areal
extent and magnitude of contamination, as indicated by chlo-
ride concentration, had significantly diminished. Chloride con-
centrations were then above 1000 mg/L in only two relatively
small parts of the contaminated area and were almost at normal
background levels in the middle of the affected area (imme-
diately downgradient from pond C).
In 1973 and 1974 there were new claims of crop and livestock
damages allegedly caused by groundwater that was contami-
nated at the Arsenal. Data collected by the Colorado Depart-
ment of Health (Shukle, 1975) show that diisopropylmethyl-
phosphonate (DIMP), a nerve-gas by-product, has been detected
at a concentration of 0.57 part per billion (ppb) in a well located
approximately 8 mi (12.9 km) downgradient from the disposal
ponds and 1 mi (1.6 km) upgradient from two municipal water-
supply wells of the city of Brighton. A DIMP concentration of
48 parts per million (ppm), which is nearly 100,000 times higher,
was measured in a groundwater sample collected near the dis-
posal ponds. Other contaminants detected in wells or springs
~ in the area include dicyclopentadiene (DCPD), endrin, aldrin,
i dieldrin, and several organo-sulfur compounds.
The detection of these chemicals, which were manufactured
or used at the Arsenal, in areas off the Arsenal property led
the Colorado Department of Health to issue cease and desist,
cleanup, and monitoring orders in April 1975 to the Rocky
Mountain Arsenal and Shell Chemical Company, which was
leasing industrial facilities on the site. The Cease and Desist
i Order called for a halt to unauthorized discharges of contam-
i inants into surface water and groundwater just north of the
Arsenal. The cleanup order applied to all sources of DIMP and
DCPD located at the facilities. The third order called for a
groundwater monitoring program, the results of which would
be reported to the State Health Department on a regular basis.
Consequently, a program that included groundwater monitor-
ing and studies to determine a means to intercept contaminants
flowing across the north boundary of the Arsenal was estab-
lished by the U.S. Army.
As a result of continued monitoring, additional contaminants
have been identified in the groundwater at the Arsenal. The
most widespread of those found are Nemagon (dibromochlo-
ropropane) and various industrial solvents. Nemagon contam-
OCR for page 96
96
ination has been identified as probably resulting from Arsenal-
related activities, whereas the industrial solvents identified are
not unique to Arsenal activities. Extremely low concentrations
of Nemagon (< 2 ppb) have been found in wells located im-
mediately west of the Arsenal boundary. Other organic con-
taminants associated with pesticide manufacturing have been
found in wells located in a centrally located manufacturing plant
area known as the South Plants area. These contaminants prob-
ably entered the aquifer from accidental spills and leaks and
appear to be migrating from this area very slowly.
Hydrogeology
The records of several hundred observation wells, test holes,
irrigation wells, and domestic wells were compiled and anal-
yzed to describe the hydrogeologic characteristics of the alluvial
aquifer in and adjacent to the Rocky Mountain Arsenal. Kon-
ikow (1975) presented four maps that show the configuration
of bedrock surface, generalized water-table configuration, sat-
urated thickness of alluvium, and transmissivity of the aquifer.
These maps show that the alluvium forms a complex, sloping,
discontinuous, and heterogeneous aquifer system.
A map showing the general water-table configuration for
1955-1971 is presented in Figure 6.3. The assumptions and
limitations of Figure 6.3 are discussed in more detail by Kon-
ikow (1975~. The areas in which the alluvium either is absent
or is unsaturated most of the time form internal barriers that
significantly affect groundwater flow patterns within the aquifer
and, hence, significantly influence solute transport.
The general direction of groundwater movement is from
regions of higher water-table altitudes to those of lower water-
table altitudes and is approximately perpendicular to the
water-table contours. Deviations from the general flow pattern
inferred from water-table contours may occur in some areas
because of local variations in aquifer properties, recharge, or
discharge. The nonorthogonality at places between water-table
contours and aquifer boundaries indicates that the approximate
limit of the saturated alluvium does not consistently represent
a no-flow boundary but that, at some places, there may be
significant flow across this line. Such a condition can readily
occur in areas where the bedrock possesses significant porosity
and hydraulic conductivity or where recharge from irrigation,
unlined canals, or other sources is concentrated. Because the
hydraulic conductivity of the bedrock underlying the alluvium
is generally much lower than that of the alluvium, groundwater
flow and contaminant transport through the bedrock are as-
sumed to be secondary considerations compared with flow and
transport in the alluvial aquifer. Groundwater withdrawals in
the area are predominantly from wells tapping the alluvial
r
aquifer.
Contamination Pattern
Since 1955 several hundred observation wells and test holes
have been constructed to monitor changes in water quality and
water levels in the alluvial aquifer. The areal extent of contam-
ination has been mapped on the basis of concentration of chlo-
ride, DIMP, and other inorganic and organic compounds in
LEONARD F. KONIKOW and DOUGLAS W. THOMPSON
.~" ~~
tROCKY MOUNTAIN ARSENAL BOUNDARY of
O 1 2 MILES
1 ~ ~
O 1 2 KILOMETERS
EXPLANATION
5/00 - WATER-TABLE CONTOUR—Shows approximate altitude of
water table, 1955 - 71. Contour interval 10 feet (3 meters).
Datum is mean sea level
Area in which alluvium is absent or unsaturated
FIGURE 6.3 General water-table configuration in the alluvial aquifer
in and adjacent to the Rocky Mountain Arsenal, 1955-1971 (Konikow,
1977).
wells. Chloride concentrations ranged from normal background
concentrations of about 40 to 150 mg/L to about 5000 mg/L in
contaminated groundwater near pond A. Chloride data col-
lected during 1955-1956 indicate that one main plume of con-
taminated water extended beyond the northwestern boundary
of the Arsenal-and that a small secondary plume extended
beyond the northern boundary (see Figure 6.4~. The contam-
ination pattern shown in Figure 6.4 clearly indicates that the
migration of contaminants in this aquifer is also significantly
constrained by the aquifer boundaries.
Because chloride generally behaves as a conservative (that
is, nonreactive) solute in groundwater, it is often assumed that
chlorides can be used to indicate the maximum extent of con-
tamination from a source that contains chloride. But this as-
sumption is not always reasonable because chloride is also a
common natural constituent in groundwater. At the Rocky
Mountain Arsenal the extent of contamination as indicated by
chloride concentration reflects a dilution ratio of about 33:1
from the contaminant source to the definable downgradient
limit of contamination. However, the extent of contamination
as indicated by some of the organic compounds, such as DIMP
(see Robson, 1981), is much greater because they have a zero
background concentration and can be detected to trace con-
OCR for page 97
Contamination and Aquifer Reclamation
39°50'
1
r- --
'L
O 1 2 MILES
1 1 ,
1 1 1
O 1 2 KILOMETERS
EXPLANATION
Data point (Sept. 1955-March 1956)
500 Line of equal chloride concentration (in milligrams per liter).
Interval variable
4~ Area in which alluvium is absent or unsaturated
FIGURE 6.4 Observed chloride concentration in 1956 (Konikow,
1977).
centrations that reflect a dilution ratio of about 100,000:1. Other
organic contaminants exhibit a much smaller plume, or migra-
tion distance, than does the chloride because of reactions that
cause them to decay or to be adsorbed. Other differences among
shapes and locations of plumes of different contaminants arise
because they entered the aquifer at significantly different times
and (or) locations within the Arsenal. For example, the Nem-
agon plume occurs west of the chloride plume because the
source of the Nemagon was not from the disposal ponds but
apparently from a spill that occurred west of the ponds.
Contaminants have also been detected in several shallow
bedrock wells in or near the Arsenal. However, at present there
are inadequate data to define the areal extent, depth of pen-
etration, or rate of spreading of contaminants in the bedrock.
APPLICATION OF SIMULATION MODELS
The reliable assessment of hazards or risks arising from ground-
water contamination problems and the design of efficient and
effective techniques to mitigate them require the capability to
predict the behavior of chemical contaminants in flowing
groundwater. Reliable and quantitative predictions of contam-
inant movement can only be made if the processes controlling
97
convective transport; hydrodynamic dispersion; and chemical,
physical, and biological reactions that affect solute concentra-
tions in the ground are understood. These processes, in turn,
must be expressed in precise mathematical equations having
defined parameters. The theory and development of the equa-
tions describing groundwater flow and solute transport have
been well documented in the literature. Perhaps the most
important technical advancement in the analysis of ground-
water contamination problems during the past 10 yr has been
the development of deterministic numerical simulation models
that efficiently solve the governing flow and transport equations
for the properties and boundaries of a specific field situation.
Although many of the processes that affect waste movement
are individually well understood, their complex interactions in
a heterogeneous environment may not be understood well
enough for the net outcome to be reliably predicted. Thus, the
analysis of groundwater contamination problems can be greatly
aided by the application of deterministic numerical simulation
models that solve the equations describing groundwater flow
and solute transport.
The solute-transport model described by Konikow and Bre-
dehoeft (1978) was used to simulate the movement of chloride
through the alluvial aquifer at the Arsenal in an effort to re-
produce the 30-yr (1943-1972) history of contamination, to help
test hypotheses concerning governing processes and parame-
ters to develop an improved conceptual model of the problem,
to aid in setting priorities for the collection of additional data,
and to evaluate possible management alternatives (Konikow,
1977~. The model included an area of approximately 34 mi2 (88
Imp. The stringent data requirements for applying the solute-
transport model pointed out deficiencies in the data base avail-
able at the start of the study. Specifically, it was found that the
velocity distribution determined from the water-table config-
uration mapped in 1956 (see Petri and Smith, 1956) was in part
inconsistent with the observed pattern of contaminant spread-
ing. The subsequent quantitative analysis and reinterpretation
of available hydrogeologic data, based partly on feedback from
the numerical simulation model, led to a revised conceptual
model of the aquifer properties and boundaries that incorpo-
rated the strong influence of the internal barriers within the
alluvial aquifer.
The solute-transport model of Konikow (1977) was calibrated
mainly on the basis of the chloride concentration pattern that
was observed in 1956 (Figure 6.4~. Computed chloride patterns
agreed closely with observed patterns, which during the 30-yr
history were available only for 1956, 1961, 1969, and 1972.
The calibrated model was then used to analyze the effects of
future and past changes in stresses and boundary conditions.
For example, comparative analyses illustrated that it would
probably take at least many decades for this contaminated aqui-
fer to recover naturally its original water-quality characteristics.
But it was also inferred that appropriate water-management
policies for aquifer reclamation can help to reduce this resto-
ration time to the order of years, rather than decades, for the
relatively mobile contaminants. Konikow (1974;) also noted that
the simulation results showed that a reclamation scheme using
a network of interceptor wells would aid in containing and
removing the contaminated groundwater.
OCR for page 98
98
Robson (1981) developed and calibrated a solute-transport
model for DIMP to help evaluate (1) the mechanisms and pa-
rameters controlling DIMP migration, (2) future DIMP con-
centrations in nearby municipal water supply wells, and (3) the
effectiveness of various groundwater barrier configurations de-
signed to halt off-Arsenal movement of contaminated ground-
water. The model included an area of about 90 mi- (230 km-)
and assumed that DIMP is conservative. Using the calibrated
model, Robson was able to reconstruct the historical movement
of DIMP in the aquifer between 1952 and 1975, to estimate
DIMP concentrations in the South Platte River resulting from
discharge of contaminated groundwater, and to predict future
DIMP concentrations under a variety of assumed management
alternatives.
To evaluate more fully the range of engineering approaches
or alternatives that would be feasible for construction along the
north boundary of the Arsenal, Warner (1979) modeled a smaller
part of the aquifer (2.5 mi2 or 6.4 ems) in that area in much
finer detail. He predicted the impact on DIMP concentration
of implementing a variety of interception schemes that incor-
porated variants of a basic plan that included elements of
groundwater withdrawal, a barrier, and reinfection of treated
water. Among other findings, Warner (1979) showed that a
properly operated hydraulic barrier, consisting of a line of
pumping wells, would be just as effective as a bentonite barrier
in stopping the movement of DIMP-contaminated groundwater
across the northern boundary of the Arsenal.
It is recognized that other organic contaminants of concern
may be sorbed or altered by chemical and biological reactions
as they move through the aquifer. The movement of a solute
that is sorbed will be retarded relative to the movement of a
conservative solute. This is beneficial in the sense that in a
given time a contaminant that is sorbed will not migrate as far
as a conservative contaminant. However, the sorption process
could pose a significant obstacle to aquifer reclamation because
even after the contaminant source has been eliminated, the
sorbed organics could later desorb and continue to migrate
through the aquifer, perhaps still posing a hazard after all con-
servative contaminants have been flushed out of the aquifer.
Sorption processes can and have been incorporated into solute-
transport models (see Grove, 1976), and this allows a more
realistic evaluation to be made of their behavior and response
to imposed aquifer reclamation stresses. Although this then
presents no great conceptual difficulty, in practice it is quite
difficult to determine the coefficients that describe the rates of
reactions and exchange capacity of the aquifer material for each
individual contaminant.
An overall systems-management model is currently in final
development under the sponsorship of the U.S. Army. This
computer model is expected to provide a valuable management
and decision-making tool to aid in evaluating aquifer recla-
mation alternatives at the Rocky Mountain Arsenal. The model
will be composed of numerous modules, including (1) ground-
water flow, (2) solute transport, (3) groundwater interception
and control, (4) surface-water control, (5) groundwater and sur-
face-water treatment, (6) cost estimation, and (7) report and
graphics output. The model will be evaluated and verified using
the Rocky Mountain Arsenal as a test case because of the abun-
LEONARD F. KONIKOW and DOUGLAS W. THOMPSON
dance of historical data there. After verification, selected al-
ternatives for contamination control and elimination at the Ar-
senal will be modeled with a goal of predicting long-term system
responses and costs. If successful, this model will be applied
to Installation Restoration programs under way at other loca-
tions.
AQUIFER RESTORATION PROGRAM
Response to Cease and Desist Orders
As a result of the Cease and Desist Orders, an Installation
Restoration program was established at the Rocky Mountain
Arsenal under the direction of the Program Manager for Chem-
ical Demilitarization and Installation Restoration, Aberdeen
Proving Ground, Maryland. This office was later reorganized
into the U.S. Army Toxic and Hazardous Materials Agency
(USATHAMA), which currently directs the Installation Res-
"oration program at the Arsenal. The main objective of this
program is to limit the migration of contaminants from the
Arsenal to the degree required by applicable federal and state
regulations. The program is primarily concerned with contam-
ination problems resulting from historical activities on the Ar-
senal as opposed to ongoing operations.
The Installation Restoration program consists of three major
parts or subprograms that include regional groundwater mon-
itoring, contaminant migration control, and elimination of con-
taminant sources. This program had been organized to allow a
phased approach in developing and implementing contaminant
control systems, thereby accelerating the reduction of potential
environmental hazards. More than $25 million has been ex-
pended to date in the Installation Restoration program, ex-
cluding the costs associated with construction of the control
systems.
A comprehensive groundwater monitoring program was de-
veloped based on historical contaminant distribution infor-
mation and initiated late in 1975. It included sample collection
from both on-site and adjacent off-site wells. This monitoring
program has been continually updated since that time to in-
clude additional wells and analytical parameters as required.
Currently, it involves the collection and analysis of samples
from 90 to 100 wells on a quarterly basis. The information
generated from the monitoring program is used to define the
distribution and track the migration of known contaminants,
identify new contaminants, develop design criteria for contam-
ination control and treatment systems, and evaluate the op-
eration of existing systems.
The subprogram concerning contaminant migration control
at the Arsenal boundaries was initiated in late 1975 with the
goal of rapidly eliminating the migration of contaminants off
the Arsenal's grounds. Boundary control was the only viable
option because of the already wide distribution of contami-
nants, the long travel times associated with contaminant mi-
gration from the sources to the boundaries, and the lack of
precise definition of all source areas. Pilot and full-scale bound-
ary control systems have been implemented at the northern
Arsenal boundary, and plans have been developed to expand
OCR for page 99
Contamination and Aquifer Reclamation
the treatment system along the northwestern Arsenal bound-
ary. These systems will be discussed in more detail later in
this chapter.
Planning for the control and elimination of contaminant sources
evolved several years later as additional data became available
on specific source areas. The goal is to control or eliminate the
contaminant sources on the Arsenal grounds and thereby elim-
inate the need for boundary control in the future. Studies have
been undertaken to aid further identification and definition of
contaminant sources, to develop feasible source control and
elimination alternatives, and to develop control and treatment
systems. A summary of the strategy and progress of this sub-
program is given at the end of this chapter.
Contaminant Migration Control at Arsenal Boundaries
Because the contamination that resulted in the issuance of the
Cease and Desist Order was detected in surface water and
groundwater immediately north of the Arsenal, the primary
focus of the Installation Restoration program during 1976 and
1977 was the northern Arsenal boundary. A dike was con-
structed to stop the migration off the Arsenal of contaminated
surface water. Studies were initiated to determine a feasible
alternative for stopping the flow of contaminated groundwater
offthe Arsenal without significantly altering the normal ground-
water flow pattern in the area. The concept selected involved
interception of the groundwater a short distance south of the
northern Arsenal boundary, treatment of the water to remove
the contaminants, and reinfection of the treated water at the
boundary.
Two methods were proposed for intercepting the flow of
groundwater. The first method involved the use of a hydraulic
barrier, one or two lines of closely spaced pumping wells that
would provide for dewatering of the aquifer along or between
the lines. The permeability in the area is sufficiently high for
this concept to have worked, but the gradient is shallow and
concern was expressed over the potential for excessive recy-
cling of water from the reinfection wells back to the withdrawal
wells. As a result of this concern and to provide an additional
safety factor, a second method was selected that involved the
use of a slurry cutoff wall to form an impermeable barrier
between the withdrawal and reinfection wells.
Treatment Process
Late in 1975 a laboratory study was initiated to evaluate various
methods for removing organic compounds from representative
groundwater samples from the area. Treatment processes in-
vestigated include granular activated-carbon adsorption, pow-
dered activated-carbon adsorption, chemical oxidation using
ultraviolet (UV) light and ozone, and anionic exchange resins.
Key chemical parameters for analysis included DIMP and DCPD.
Extensive laboratory studies were conducted using standard
isotherm tests for evaluating the carbons and resins and using
batch reactor tests for evaluating the UV/ozone process. The
anionic exchange resins were dropped from further consider-
99
ation because of low efficiency and high cost. A series of field
studies was initiated on the carbon adsorption and UV/ozone
oxidation processes to permit further evaluation.
Powdered activated-carbon adsorption tests incorporating a
polymeric coagulant were conducted using a standard Army
Erdlator water-treatment unit (chemical addition, mixing, up-
flow clarification)(Sweder, 1977~. Granular activated-carbon ad-
sorption tests were conducted using a dynamic-flow, multi-
column system (Sweder, 1977~. UV/ozone oxidation tests were
conducted using a continuous-flow, mechanically mixed reactor
(Buhts et al., 1978~. Granular activated-carbon was found to
be more efficient (110 mg of carbon/L of water) in removing
the contaminants than was the powdered activated-carbon (200
mg of carbon/L of water>. Cost estimates were developed for
the carbon adsorption and UV/ozone oxidation processes based
on treating 10,000 gallons of water per hour (37,850 L/h). The
estimated cost of granular activated-carbon treatment was ap-
proximately $2 per 1000 gallons, powdered activated-carbon
treatment was approximately $4 per 1000 gallons, and UV/
ozone oxidation treatment was approximately $3 per 1000 gal-
ions. As a result of these studies and the immediate availability
of proven process equipment, granular activated-carbon was
selected for use in the proposed treatment system.
Installation and Operation of Pilot Containment System
Hydrologic and chemical data indicated that the highest con-
centrations of contaminants were crossing the northern Arsenal
boundary in the alluvial aquifer in an area associated with a
buried channel in the relatively impermeable bedrock of the
Denver Formation. This area is located approximately 1 mile
east of the northwest boundary and has a width of approxi-
mately 1000 ft (305 m). Because little operational information
was available on groundwater contamination control systems
similar to the one proposed, the Army decided to install a
limited pilot containment system in the area of high-contam-
inant concentrations and evaluate the possibility of extending
the treatment system across the entire affected part of the
northern boundary.
The North Boundary Pilot System (NBPS) was constructed
and placed in operation in July 1978. It included the following
five subsystems: a barrier, dewatering wells, reinfection wells,
treatment plant, and monitoring wells. A schematic diagram
of the system is provided in Figure 6.5.
The barrier was constructed by filling a 3-ft-wide, 1500-ft-
long trench, averaging 25 ft in depth, with a mixture of soil
and bentonite clay. The barrier was anchored approximately 2
ft into the bedrock all along the alignment.
The dewatering wells were installed south (upgradient) of
the barrier approximately 225 ft apart on a straight line parallel
to the barrier. There were six 8-inch-diameter wells placed
within 30-inch-diameter gravel-packed holes. Each well was
screened throughout the entire saturated portion of the alluvial
aquifer. A submersible pump and flow control system were
installed at each well site. Water from the wells was pumped
through an underground manifold to a single sump at the treat-
ment plant.
OCR for page 100
100
The injection wells were installed north (downgradient) of
the barrier, approximately 100 ft apart on a straight line be-
tween the barrier and the northern Arsenal boundary. There
were twelve 18-inch-diameter wells, which were installed in
36-inch-diameter gravel-packed holes. The recharge wells were
screened to a point above the water table. Treated water was
continuously injected into the recharge wells by gravity flow
through an underground manifold system. Sensors and flow
control valves were installed in the wells to prevent overflow
or surface discharge in the event that a well experienced an
excessively high buildup of hydraulic head because of clogging
of well screens or other factors.
The treatment plant subsystem was designed to treat 10,000
gallons of water per hour. It consisted of two mixed-media
pressure filters, each 4 ft in diameter, and two adsorber vessels
(or columns), each 10 ft in diameter and 11 ft high, designed
to contain about 20,000 lb (9100 kg) of granular activated car-
bon. Water from the collection sump was pumped through the
filters in parallel to remove suspended material, then through
the carbon absorbers, and finally to the injection wells. Only
one carbon adsorber was in operation at any one time. When
the DIMP concentration approached 50 ppb, the carbon was
replaced. During 1978-1981, replacement was required ap-
proximately once every 9 months. The exhausted carbon was
transported offsite for regeneration by a commercial vendor.
Carbon usage rates ranged from 100 to 150 mg of carbon/L of
water. The treatment system was designed to be largely au-
tomatic and simple to operate by incorporating automatic back-
washing of the filters and sensors for control of pumps and
valves.
Ten monitoring wells were installed both upgradient and
downgradient of the pilot containment system. They were cased
with small-diameter PVC pipe and screened in the alluvial
PILOT TREATMENT
PLANT
INFLUENT
FROM
DEWATERING
WELLS
.
... _ _ ..... _
REINJECTION
- DEWATERING WELL
WELL GROUND SURFACE
~ ,,,~,,~, t~ ~ ,,,~-,;--
_ /~
FIGURE 6.5 Schematic diagram of north boundary contamination
control pilot system.
LEONARD F. KONIKOW and DOUGLAS W. THOMPSON
~ ~ ~ ~ ~ ~ ~ ~ I ~ ~ ~ , An- T T
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
RETENTION TIME
FIGURE 6.6 Typical gas chromatography/mass spectrometry scan of
north boundary pilot treatment system influent.
aquifer. Water levels and chemical quality were monitored
periodically to provide information on the effectiveness of the
operation of the system.
The cost of the barrier and the wells as constructed in 1978
was $450,000. The facility for housing the treatment system
cost approximately $40,000. The treatment equipment was ob-
tained under a lease/service contract agreement with a com-
mercial vendor with an initial cost of approximately $100,000
and a yearly fee ranging from $135,000 to $150,000.
The NBPS operated successfully for a period of approxi-
mately 3 yr. For example, during fiscal year 1979, downtime
was less than 1 percent of operating time. The granular acti-
vated carbon effectively removed the organic contaminants from
the groundwater, generally to a level of less than 10 ppb, as
illustrated by a comparison of typical gas chromatography/mass
spectrometry analyses of the influent (Figure 6.6) and effluent
(Figure 6. 7) of the treatment system. The flow of groundwater
downgradient from the NBPS was essentially unchanged
(D'Appolonia Consulting Engineers, Inc., 1979~. Preliminary
data indicate that the concentration of organics in the ground-
water downgradient from the pilot system has diminished sig-
nificantly.
Expanded Containment System
As a result of the successful operation of the pilot containment
system, construction of the expanded containment system was
begun in early 1981. The expanded system consists of a 6800-
ft barrier ranging from 25 to 50 ft deep, 54 withdrawal wells,
and 38 reinjection (or recharge) wells. The expanded barrier
effectively intercepts all the contaminated groundwater flowing
across the northern Arsenal boundary in the alluvial aquifer.
The expanded treatment system is designed to treat 36,000
gallons (136,000 L) of water per hour. The adsorbers used in
the pilot operation have been replaced with three pulsed-bed
adsorbers designed to contain 30,000 lb (13,600 kg) of carbon
OCR for page 101
Contamination and Aquifer Reclamation
LU
In
g
cat
to
a
UJ
Tl
-
a) ~ ~a
_
~o A n Q ~
.. 1 ~ _ ma_
I r r ~T~TT~
1 2 3 4 s 6 7 8 9 10 11 12 13 14 15 16
RETENTION TIME
FIGURE 6.7 Typical gas chromatography/mass spectrometry scan of
north boundary pilot treatment system effluent.
each. The new adsorbers should be much more efficient than
the old ones because the anticipated carbon usage rate is only
25 to 30 mg of carbon/L of water. The mixed-media filters have
been replaced with cartridge filters, which are easier to main-
tain. The whole system is highly automated and will require
only intermittent monitoring by a single operator. The esti-
mated cost for the expanded system is approximately $6 million.
The expanded system became operational in 1983.
Other Contaminant Migration Control Systems
Concepts have been developed for two additional boundary
contaminant migration control systems located along the north-
western Arsenal boundary (Figure 6.8~. One system will be
located at the southern end of that boundary and the other
midway along that boundary. Both systems have been devel-
oped primarily to control the migration of low concentrations
of Nemagon across the boundary. Both systems will be similar
in size to the NBPS and will incorporate granular activated-
carbon treatment of the groundwater. The system to be located
on the southern end of the boundary (Irondale System) was
constructed under the direction of Shell Chemical Company
and incorporates a hydraulic barrier for interception of the
groundwater, along with the injection wells. It became oper-
ational in 1983. The other system, to be constructed by the
Army, will incorporate a slurry cutoff wall, withdrawal wells,
and reinfection wells, similar to the pilot system. It is scheduled
to be operational in 1985.
Planning for Control and Elimination of Contaminant
Sources
Contaminant migration control at the boundaries of the Rocky
Mountain Arsenal was initiated to stop or severely limit the
migration of contaminants off the Arsenal grounds as soon as
possible. Owing to the size of the Arsenal and the extent of
101
the source areas, the boundary control systems could be re-
quired to operate for an indefinite period of time. The only
way to limit this requirement and the associated cost is to
control or eliminate the contaminant sources. Therefore a study
was initiated in 1980 to identify and assess existing and inno-
vative control or elimination alternatives that are capable of
bringing the Arsenal into compliance with all applicable federal
and state environmental laws and regulations. Another study
objective was to develop preliminary cost data and technical
data for use in a subsequent detailed evaluation and comparison
of alternatives. A study team made up of 12 government and
independent scientists and engineers was established to con-
duct and manage the study. A review of historical operations,
past study reports, and data from ongoing studies was made to
identify, where possible, potential sources of contaminant mi-
gration problems.
The next phase of the study involved the development of
control strategies. Guidelines and criteria for development of
the strategies were required because of the complexity of and
relationships between the contaminant sources and migration
characteristics. In addition, some degree of commonality of
structure or organization among the strategies was needed to
enable a comparison and ranking of the alternatives to be de-
veloped. As a result, a hierarchical approach and structure for
generation and classification of control strategies were devel-
oped incorporating five levels of detail ranging from concept
to unit operation (Rocky Mountain Arsenal Contamination Con-
trol Study Team, unpublished report, August 1981~. Each team
member individually developed a number of strategies using
the hierarchical approach and determined the problem defi-
nition and technical data-base deficiencies associated with each
scheme. The schemes were then submitted to the group as a
whole for integration and evaluation.
Screening criteria were developed to aid in evaluating and
comparing the alternative schemes. The goal was to produce
EXPLANATION
· · ·Dewatering wells
/ ~ _ ~
E] Liquid Treatment / tt NORTH BOUNDARY ~
'\ A Recharge wells ~~ SYSTEM 7
Slurry trench NORTHWEST BOUNDARY '/
/9~,/ SYSTEM (Proposed) ,~1
/ ~ Basins
,,* ~ Basins
B ~*7
v//////////////////////////////////,
SOUTH PLANTS AREAWAY
Lakes
I ~iRONDALE SYSTEM
(Shell Chemical Co.)
l
l
~1\
COMPLEX ~
r- l
I 1 0 ~ 2 KILOMETERS
0 1 2 MILES
l
1
l
1
am,. 1
!
_ ~
FIGURE 6.8 Location of existing and proposed boundary control
systems.
OCR for page 102
102
a set of criteria that could be applied at the various hierarchical
levels, thereby enabling a general screening of the schemes
rather than a detailed evaluation of each one. The major criteria
selected for use are as follows:
1. Availability of technology,
2. Amount of additional data required,
3. Cost and time needed to fill data gaps,
4. Life cycle costs capital and O&M,
5. Compatibility between systems,
6. Degree of risk environmental and technological,
Compliance with regulatory requirements.
The individual schemes developed by the study team mem-
bers were integrated, evaluated, and screened by the study
group as a whole. This work resulted in the presentation of 14
alternative schemes that were recommended for detailed eval-
uation by the Contamination Control Study Team. The schemes
incorporate various aspects of the technologies listed in Table
6.1. The schemes address only the known contaminant sources
at the Arsenal and therefore may have to be expanded if ad-
ditional sources are identified in the future.
In addition to the development of the alternative schemes,
the study team identified a number of data gaps concerning
both problem definition and technology development that must
be filled before final selection of a control or elimination al-
ternative can be made. Studies have been included in the
overall Installation Restoration program to fill these data gaps.
They include additional hydrogeologic definition of certain areas
on the Arsenal, surface-water hydrology definition, technology
development for water treatment, and technology development
for disposal of contaminated soil and residue. As the data from
these additional studies become available, the study team will
further evaluate and revise the alternatives as required with
the goal of selecting one alternative for implementation.
The implementation of the selected alternative will be con-
ducted using a phased approach. As soon as a particular part
TABLE 6.1 Contaminant Source Control and Elimination
Technologies
Groundwater Interception
Hydraulic barrier
Slurry trench
Dewatering trench (French drain)
Water Treatment
Adsorption (carbon and resin)
Chemical addition/coagulation/precipitation
Filtration
Membrane separation
Chemical oxidation
Activated sludge
Volatile stripping
Ion exchange
Contaminated Soil and Residue Treatment
Incineration
Fixation/stabilization
In situ forced leaching
Excavation and disposal
LEONARD F. KONIKOW and DOUGLAS W. THOMPSON
of the alternative is defined and design criteria are developed,
construction will be initiated. For example, the elimination of
Basin F will probably be one of the first major actions initiated
because it is known to leak and because the extent and nature
of the contamination associated with this area of the Arsenal
have been better defined than elsewhere. The control and
elimination of known contaminant sources at the Rocky Moun-
tain Arsenal are currently expected to involve a 5-yr construc-
tion program that is scheduled to start in 1985. A final cost
estimate for the construction program has not been developed,
but preliminary estimates range from $50 million to $100 mil-
lion.
SUMMARY AND CONCLUSIONS
Removing pollutants from a contaminated aquifer may seem
to be an almost impossible task. While this may be true for
some contaminated aquifers, others may be amenable to one
or more plans for artificial reclamation that could significantly
accelerate the rate of water-quality improvement in the aquifer.
The feasibility of any such reclamation plan would be strongly
dependent on the hydraulic and chemical properties of the
aquifer, on the type and source of contamination, and on the
duration and areal extent of contamination. Because a variety
of reclamation plans can be proposed for any one problem, an
accurate model of flow and contaminant transport in the aquifer
is an invaluable tool for planning an efficient and effective
program.
The control and elimination of contaminant migration and
contaminant sources at the Rocky Mountain Arsenal represent
a large, complex, and costly undertaking (over $25 million has
been spent in the Installation Restoration program). An exten-
sive well-monitoring program has been required to define the
extent of the contamination and the relationships between the
sources and contaminant migration patterns. Control of con-
taminant migration at the Arsenal boundaries has proved fea-
sible using a system involving groundwater interception,
treatment, and reinfection. The system was operated success-
fully without adversely affecting the flow and distribution of
groundwater downgradient from the treatment system, and it
has resulted in a significant decrease in the concentration of
organic contaminants in groundwater downgradient from the
pilot system.
Although boundary-control systems can be used successfully
to stop or restrict the migration of contaminants off the Arsenal's
grounds, they cannot solve the problem of continued contam-
inant migration from the source areas to the environment. The
overall solution thus involves the control or elimination of the
contamination at the sources. A program has been successfully
initiated at the Rocky Mountain Arsenal to develop and assess
source control and elimination strategies. Through additional
data collection and feasibility studies, a single strategy will be
selected and implemented using a phased construction ap-
proach. The ultimate goal of these activities is to bring the
Arsenal into compliance with all applicable federal and state
environmental laws and regulations.
The great difficulty and expense involved in mitigating
OCR for page 103
Contamination and Aquifer Recla7nation
groundwater contamination problems do not lessen the need
to do so; they do illustrate the long-term benefits of planning
and designing waste-disposal activities to prevent or minimize
future contamination hazards.
AC KN OWLE D G M E NTS
The Installation Restoration program at the Rocky Mountain
Arsenal (RMA) is being funded and directed by the U. S. Army
Toxic and Hazardous Materials Agency, Aberdeen Proving
Ground, Maryland, in cooperation with the Rocky Mountain
Arsenal, Denver, Colorado. The authors wish to thank the
personnel from these organizations for their support. Special
thanks are extended to Carl Loven, Chief, Process Develop-
ment and Evaluation Division, RMA, and Donald Hager, Ru-
bel and Hager, Inc., Tucson, Arizona, for providing operational
and cost data on the RMA contaminant control systems.
REFERENCES
Buhts, R. E., P. G. Malone, and D. W. Thompson (1978). Evaluation
of ultraviolet/ozone treatment of Rocky Mountain Arsenal ground-
water, U.S. Army Engineer Waterways Experiment Station, Tech-
nical Report Y-78-1, 78 pp.
D'Appolonia Consulting Engineers, Inc. (1979). Evaluation of north
boundary pilot containment system, RMA, Denver, Colorado, Proj-
ect Number RM79-389, 60 pp.
Grove, D. B. (1976). Ion exchange reactions in groundwater quality
models, in Advances in Groundwater Hydrology, Am. Water Re-
sour. Assoc., pp. 144-152.
Konikow, L. F. (1974). Reclamation of a contaminated aquifer, Geol.
Soc. Am. Abstr. Programs 6, 830-831.
Konikow, L. F. (1975). Hydrogeologic maps of the alluvial aquifer in
103
and adjacent to the Rocky Mountain Arsenal, Colorado, U.S. Geol.
Surv. Open-File Rep. 74-342.
Konikow, L. F. (1977). Modeling chloride movement in the alluvial
aquifer at the Rocky Mountain Arsenal, Colorado, U.S. Geol. Surv.
Water-Supply Pap. 2044, 43 pp.
Konikow, L. F., and J. D. Bredehoeft (1978). Computer model of two-
dimensional solute transport and dispersion in ground water, U.S.
Geol. Surv. Techniques of Water-Resources Inv., Book 7, Chap.
C2, 90 pp.
Petri, L. R. (1961). The movement of saline ground water in the vicinity
of Derby, Colorado, in Ground Water Contamination Symposium,
Robert A. Taft Sanitary Eng. Center Tech. Rep. W61-5, pp. 119-
121.
Petri, L. R., and R. O. Smith (1956). Investigation of the quality of
ground water in the vicinity of Derby, Colorado, U.S. Geol. Surv.
O pen-File Rep ., 77 pp.
Robson, S. G. (1981). Computer simulation of movement of DIMP-
contaminated groundwater near the Rocky Mountain Arsenal, Col-
orado, T. F. Zimmie and C. O. Riggs, eds., in Permeability and
Groundwater Contaminant Transport, American Society for Testing
and Materials, Philadelphia, Pa., pp. 209-220.
Shukle, R. J. (1975). 1974-75 Groundwater Study of the Rocky Moun-
tain Arsenal and Some Surrounding Area, Colorado Department of
Health, Denver, Colo., 20 pp.
Sweder, R. G., Jr. (1977). Carbon adsorption treatment of contami-
nated groundwater at Rocky Mountain Arsenal, Rocky Mountain
Arsenal, Denver, Colo., 97 pp.
Walton, G. (1961). Public health aspects of the contamination of ground
water in the vicinity of Derby, Colorado, in Ground Water Con-
tamination Sy7nposium, Robert A. Taft Sanitary Eng. Center Tech.
Rep. W61-5, pp. 121-125.
Warner, J. W. (1979). Digital-transport model study of diisopropyl-
methylphosphonate (DIMP) ground-water contamination at the Rocky
Mountain Arsenal, Colorado, U.S. Geol. Surv. Open-File Rep. 79-
676, 39 pp.
Wood, L. A. (1972). Groundwater degradation causes and cures, in
Proceedings 14th Water Quality Conf., Urbana, Ill., pp. 19-25.
Representative terms from entire chapter:
mountain arsenal
