United States
Environmental Protection
Agency
Solid Waste and Emergency
Response
(OS-110)
November 1990
EPA/540/2-90/018
v>EPA A Guide to Pump and Treat Ground-
Water Remediation Technology
to detemiBewhen, where, and
aM*tr^t&chftote^<^^s^
^ aMItevdojwent'stoto^^
OVERVIEW
While there are several ground-water containment and cleanup
options available to choose from, this fact sheet focuses on pump-
and-treat technology. The pump-and-treat process is the most
commonly used ground-water remediation technology at hazard-
ous waste sites. The objectives of pump-and-treat are to reduce
the concentration of contaminants to an acceptable level during
cleanup or to contain contaminants in order to protect the subsur-
face from further contamination. Pump-and-treat systems capture
contaminated ground water for surface treatment. This fact sheet
outlines the basic requirements for an effective pump-and-treat
system, which include identifying the contaminant, characteriz-
ing the subsurface, designing a capture system, installing extrac-
tion wells, and monitoring the remediation progress. Here the
"pump" portion of the pump-and-treat process is emphasized.
Recent research has identified complex chemical and physical
interactions between contaminants and the subsurface media that
may limit the effectiveness of the extraction phase. These
important limitations of pump-and-treat technology are also de-
scribed in this fact sheet.
CHOOSING PUMP-AND-TREAT
REMEDIATION
The first step in determining whether
pump-and-treat is an appropriate remedi-
al technology is to conduct a site charac-
terization investigation. If the risk assess-
ment shows the need for remedial action,
then site characteristics, such as hydraulic
conductivity, will determine the range of
remedial options possible. Sources of
ground-water contamination can include
leaky tanks, leachate from landfills, spills,
chemicals dissolving from nonaqueous
phase liquids (NAPLs), and chemicals
desorbing from the soil matrix.
Sites with ground-water contamination will
almost always include some form of pump-
and-treat remediation. Chemical properties
of the site and plume need to be determined
to characterize transport of the contaminant
and evaluate the feasibility of a pump-and-
treat system. To determine if pump-and-
treat is appropriate at a given site, one needs
to know the history of the contamination
event, properties of the subsurface, ground-
water flow characteristics, and biological
and chemical contaminant characteristics.
Identifying the chemical and physical site
characteristics, locating the ground-water
contaminant plume or NAPL in three di-
mensions, and determining aquifer and
soil properties are necessary in designing
an effective pump-and-treat strategy.
Several remedial methods may be com-
bined into a "treatment train" to attain
cleanup goals. The criteria listed below
outline the information necessary to de-
termine if pump-and-treat systems are
applicable to a site.
Criteria to Determine if Pump-and-Treat will be Effective
I) History of the contamination: II) Characteristics of the subsurface flow system:
A history of the contamination
event should be prepared to de-
fine the types of wastes present at
the site and quantify their loading
to the system.
Ground-water flow systems vary with time, season,
and pumping strategy. Understanding where ground
water recharges and discharges (mass balance), the
laws governing flow (Darcy's law), and geologic
framework through which the flow passes makes it
possible to determine ground-water flow character-
istics. Other subsurface flow system characteristics
include hydraulic conductivity, storage coefficient,
mineralogy, organic content, and aquifer thickness.
HI) Chemical and biological characteris-
tics of the contaminant:
Chemical characteristics of contaminants
include solubility, density, reactivity, ion
exchange capacity, and mobility in aque-
ous solution. Biological characteristics of
contaminants include the potential for
naturally occurring transformation and
biodegradation.
Printed on Recycled Paper.
-------
PUMP-AND-TREAT
REQUIREMENTS
Four basic components need to be devel-
opedforasuccessfulpump-and-treatsys-
tem.
Q Goals and objectives
Q System design
Q Operational rules and
monitoring
Q Termination criteria
The first component consists of defining
the remediation goals and objectives (re-
medial action objectives) to be accom-
plished at a given site. This involves
gathering enough background site infor-
mation and field data to make assess-
ments of remedial requirements and pos-
sible cleanup levels. The first determina:
lion is whether cleanup or containment
will be the most appropriate pump-and-
trcat remedial action. If cleanup is cho-
sen, the level of cleanup must be deter-
mined according to maximum contami-
nant levels (MCLs) and alternate con-
taminant levels (ACLs), state laws, or
other criteria selected for the site. If
containment is chosen, pump-and-treat
technology is used as a hydraulic barrier to
prevent off-site migration of contaminant
plumes. The goals and objectives chosen at
this stage determine the course of the reme-
diation plan.
The next component consists of the design
and implementation of the pump-and-treat
system based on data evaluated in setting the
goals and objectives. The system must be
chosen and designed based on field data.
Selection of a system is also dependent on
whetherpump-and-treatis sufficientor more
than one remedial action will be used. The
criteria for well design, pumping system,
and treatment are dependent on the physical
site characteristics and contaminant type.
The system may then be installed, including
extraction wells, injection wells, drain inter-
cepts, and barrier walls, if necessary.
The third and most significant component
for ensuring the long-term effectiveness of
pump-and-treat is frequent monitoring of
progress to verify if the remedial strategy is
meeting remedial action objectives. Moni-
toring the remedial process with wells and
piezometers allows the operator to make
iterative adjustments to the system in re-
sponse to changes in subsurface conditions
caused by the remediation.
The final component in the pump-and-
treat process is determining the termina-
tion requirements. Termination require-
ments are based on the cleanup objectives
defined in the initial stage of die remedial
process. The termination criteria are also
dependent on the specific site aspects re-
vealed during remedial operations.
DATA COLLECTION
Collecting as much background site data
as possible initially may reduce theamount
of time spent gathering data in the field.
Accurate information on the type of con-
taminants present and their loading capac-
ity will promotea well-designed remedia-
tion plan. Contaminant information
needed consists of: 1) source character-
ization, including the volume released,
the area infiltrated, and duration of re-
lease; 2) concentration distribution of
contaminants and naturally occurring
chemicals in ground water and soil; and 3)
processes affecting plume development,
such as chemical and biological reactions
influencing contaminant mobility. Each
step of the pump-and-treat strategy is de-
pendent on the decisions reached in the
previous step. Therefore, it is vital that
each step is carefully planned and moni-
tored to allow for modifications.
1
Example of a Pump-and-Treat System
ff"
[TREATMENT FACILITY]
OVEItOUnOUl 3AMO
SLT
BEDROCK FLOW LINE
WATER TABLE
UNDER
CONDITIONS
-------
Understanding the hydrogeology and ex-
tent of contamination at a site are impor-
tant in planning successful field studies.
The hydrogeologic aspects listed in Table
1 (below) are vital in determining if a
pump-and-treat system would be an ap-
propriate remedial technology for a par-
ticular site. These aspects include deter-
mining the size of the contaminated aqui-
fer, depth to water table, hydraulic con-
ductivity of surrounding aquifer material,
and local ground-water use. Methods for
determining aquifer properties include a
slug test, pump test, and a borehole
flowmeter test. The pump test consists of
pumping one well and measuring the wa-
ter level response of surrounding wells.
Pump tests sample large aquifer sections.
A slug test measures the rate at which the
water level in one well returns to its initial
state after inducing a rapid water level
change by introducing or withdrawing a
volume of water. Theboreholeflowmeter
test measures flow direction and rate in a
borehole. These tests can indicate the spa-
tial variability of hydraulic conductivity.
Once data have been collected, the infor-
mation must be accurately interpreted.
There are numerous tools that can be used
to interpret data, including geochemical
analysis, geostatistical analysis, and
mathematical models. Geochemical
analysis uses ion speciation models to
interpret chemical changes in the aquifer.
Geostatistical methods may be used to
determine the relationship among various
parameters and define the statistical prob-
ability ofaparticular condition. Mathemati-
cal models may be used to simulate ground-
water flow patterns, contaminant transport,
and the changes resulting from a pump-and-
treat system.
Water hi a well bore seldom represents that
of the adjacent aquifer. Therefore, when
sampling ground water, pH, temperature,
and conductivity should be measured and
allowed to stabilize before a sample is taken
to more accurately reflect the ground-water
quality.
SYSTEM DESIGN
Aneffectivepump-and-treatsystem depends
on careful design of the pumping and treat-
ment components based on the hydrogeologic
information gathered at the site. Design
considerations include type and location of
wells, pumps, and piping; drilling methods;
and well design and construction. Extrac-
tion wells may be used with injection wells
if the hydraulic conductivity of the site
material is high. Drains may be used if the
contaminated aquifer is close to the surface.
Intercept drains may be appropriate when a
shallow aquifer is surrounded by material
with low hydraulic conductivity. A long-
term aquifer test (longer than a few days)
can provide useful information and serve as
a prototype for the pump-and-treat system
design.
Table 1, Aspects of Sit© Hydrogeology
'I > Type, thickness, did sisal extartf of aquifer.
, , 2. Type of porpsily ^prinwrjs fat&gfam&ai jjpre spsc^ or seeojitfai
& Presence or absence of jmperawsafcte wats ami confining layers,
4> Depth to water table, iMcfcaess of vsdose zone.
Itydraalic
i , Pressure sortditforis? confined, qngoaSned, or leaky Existing or near-site use of ground water, -
Special care is required to avoid potential
problems with well-construction materi-
als, especially when dealing with NAPLs.
Wells should be designed so that screens
may be easily flushed and clogging prob-
lems commonly caused by oxidation of
manganese and iron can be treated. As-
pects to consider when selecting pumps
are failure rates, reactivity with contami-
nants, and ease of maintenance. Backup
equipment shouldbeavailableintheevent
of failure.
The types of pumping used at a site in-
clude continuous and pulsed pumping.
Continuous pumping maintains an inward
gradient, constantly drawing ground wa-
ter towards it. Pulsed pumping consists of
alternating periods of time when thepumps
are on and when they are off.
Depending on site characteristics and con-
taminant properties, injection wells may
be installed along with extraction wells to
reduce cleanup time. Injection wells in-
crease the hydraulic gradient by flushing
contaminants towards the extraction well.
The pump-and-treat system should be
evaluated periodically to determine if the
goals and standards of the design criteria
are being met. Monitoring the remedial
process allows for operational modifica-
tions to be made. One modification that
may improve the efficiency of contami-
nant recovery is to switch from continu-
ous to pulsedpumping. The non-pumping
period during pulsed pumping allows the
contaminants to diffuse and desorb from
less permeable zones into adjacent zones
of higher hydraulic conductivity, permit-
ting more efficient contaminant extrac-
tion when pumping resumes.
Another design modification is to cycle
pumping at selected wells in order to bring
stagnant zones into active flow paths for
remediation. When less soluble contami-
nants (NAPLs) are trapped in soil pores by
interfacial tension, the flow rates during
remediation may be too rapid for the con-
taminant to reach chemical equilibrium.
The non-pumping stage at selected wells
provides time for sorbed and residual
contaminants in the stagnant zone to reach
equilibrium with the ground water. Dura-
tionof pumping andnon-pumpingperiods
-------
are site-specific and can only be opti-
mized through continuous monitoring.
Ideally, the contaminant source would
be completely removed for proper aqui-
fer remediation using pump-and-treat
technology. Unremoved contaminants
\villcontinuetodissolveintotheground
water and prolong cleanup. It may be
advantageous to have multiple extrac-
tion wells pumping at a low rate rather
than one at a high rate. Analytical and
numerical modeling techniques can be
used to evaluate alternative designs,
optimal well spacings, pumping rates,
and cleanup times. Models can calcu-
late ground-water flow paths, locate
contaminant plume fronts, and attempt
to simulate contaminant transport.
Proper design will ensure that wells are
placed in the desired stratigraphic layer
so that the correct area will be
OPERATION AND
MONITORING
Once remedial action objectives are es-
tablished and a system is built to meet
these objectives, then a monitoring pro-
gram should be designed to evaluate the
success of the remedial system. Uncer-
tainties in subsurface characterization
make monitoring a necessary step in
pursuing a remedial strategy. Continual
monitoring of the pump-and-treat sys-
tem allows timely modifications to be
made when it is clear that the system is
not achieving prescribed goals.
Monitoring should consist of analyzing
the water quality and contaminant
movement, and supervising the me-
chanicaloperationofthepump-and-treat
system. If monitoring shows that the
cleanup objectives are not being met,
then changes to the pump-and-treat sys-
tem must be specified and implemented
to meet specified goals. There are three
basic components involved in monitor-
ing.
1) Design an appropriate monitoring
program to suit the pump-and-treat
system.
2) Actively monitor the system to
verify that the remedial strategy is
meeting the objectives and that
equipment is functioning properly.
3) Modify the remedial strategy to ad-
just for unexpected contingencies.
Specify alternate acceptable goals
or change the remediation strategy
to meet the original goals.
Monitoring criteria important in establishing
a successful monitoring scheme for a site can
be divided into three categories, chemical,
hydrodynamic, and administrative.
Q Chemical: A risk-based criterion,
including maximum contaminant
levels (MCLs), alternate contami-
nantlevels (ACLs), detection limits,
and natural water quality.
Q Hydrodynamic: Includes prevent-
ing infiltration through the vadose
zone, maintaining an inward gradi-
ent at the boundary of the contami-
nantplume, and providing minimum
flow to surface water bodies.
Q Administrative Control: Includes
reporting requirements, frequency
andcharacterof operational andpost-
operationalmonitoring, andlanduse
restrictions, suchasdrillingandother
access-limiting restrictions.
The location of themonitoring wells is critical
to any successful monitoring program. Water
level fluctuations and water quality should be
measured. Injection and extraction wells
change the subsurface in complex ways, re-
quiring continuous monitoring. Determining
the flow pattern generated by a pump-and-
treat system requires field evaluations during
the operational phase.
Apump-and-treat system may require modifi-
cations during the operational phase due to the
uncertainties involved in subsurface charac-
terization. Reasons for possible modifica-
tions resulting from operational monitoring
are:
• Improved estimates of hydraulic
conductivity requiring a change in
pumping rate or well location.
• Information on chemistry and
loading to the, treatment facility
requiring changes in treatment.
• Mechanicalfailureofpumps, wells,
or subsurface plumbing.
• Adjusting pumping rate or well
location to remediate a stagnant
zone (a hydrodynamically isolated
zone) as the contaminant plume is
remediated, or to enhance extrac-
tion if anticipated progress is not
achieved.
SOLUTE TRANSPORT
ASPECTS
Assessing the chemical properties of
the contaminant; plume is necessary to
eftamctfcriz&the transport of tfteehemi*
cal and evaluate mef easiblfty of pump/-
an.d«tteat, Movement of «on-ieactJve
dissolved contaminants in saturated
tion and somewhat by dispersion, Ad-
vecfJoacaiases the plume tomove at the
rateanddirectionofground-waterflaw.
Properties mat control the transport of
chemicals in ground water should be
treat systems.
Dispersion is the combined effect of"
mechanical mixing and chemical dif*
fusion. Dispersion causes the plume
volume to increase and its maximwm
concentration; to decrease. Reac|ive
contaminant transport is also affected
by sorption, desorption, and chemical
and biochsrnjcal reactions, Studying
sorption-desorption and transformation
processes Is essential in understanding
migration rates and concentration dis-
tributions Of contaminants. The plume
of: a reactive contaminant tends to ex-
pand more slowly than that of a non*
reactive contaminant, increasing
cleanup tune.
-------
Necessary modifications resulting from
the monitoring results should be imple-
mented. The system design should be
flexible enough to allow for easy adjust-
ments to quicken cleanup. Keeping the
possibility of modifications in mind when
constructing the pump-and-treat system
will promote the speed and efficiency of
remediation.
The pump-and-treat system is terminated
when the cleanup objectives are met.
Monitoring is needed to ensure that des-
orption or dissolution of residuals does
not cause an increase in the level of con-
tamination after operation of the pump-
and-treat system has ceased. Post-opera-
tional monitoring may be required for two
to five years after termination depending
onsiteconditions. Calculatingthecleanup
period for a site is necessary to estimate
termination time and potential length of
post-operational monitoring (Seeexample
on this page).
LIMITATIONS OF PUMP-AND-
TREAT TECHNOLOGY
Reducing ground-water concentrations to
standards required by the Safe Drinking
Water Act or Land Disposal Restriction is
difficult using available technology for
many contaminants. There are several
inherent limitations that hinder effective
pump-and-treat site remediation. These
include the potentially long time neces-
sary to achieve the remediation goal; sys-
tem designs failing to contain the con-
taminant as predicted, allowing the plume
to migrate; and failure of surface equip-
ment
Research has substantiated other limita-
tions with the use of pump-and-treat tech-
nology. These limitations include con-
taminant residual saturation, chemical
sorption of the contaminant, and low hy-
draulic conductivity causing tailing ef-
fects.
1. Residual saturation
The presence of nonaqueous phase liquids
greatly complicates contaminant behav-
ior. Movement of contaminants in a sepa-
rate, immiscible phase is not well under-
stood either in saturated or unsaturated
1.0
8
UJ
1
0.5
| REMOVAL WITH TAILINOj
| THEORETICAL REMOVAL]
1
TIME
Effects of tailing on pumping time
zones. A less soluble contaminant moves in
response to pressure gradients and gravity
and is influenced by interfacial tension,
volatilization, and dissolution.
Residual saturation or irreducible satura-
tion is the limit of drainage, where a certain
pore volume will always remain. Both the
type of immiscible fluid involved and the
pore size distribution of the material deter-
mine the extent of residual saturation. Re-
sidual saturation reduces the overall amount
of contaminant that enters and migrates
within the saturated zone and acts as a
source of long-term miscible contaminant.
Additional datarequired to determineproper
remediation strategies for NAPLs include
fluid specific gravity, viscosity, and con-
taminant thickness and distribution. Sub-
stances that are particularly difficult to
remediate are halogenated aliphatic hy-
drocarbons, halogenated benzenes,
phthalate ethers, and polychlorinated bi-
phenyls. Dataonrelativepermeabilityare
readily available for many petroleum ap-
plications.butnotforh'quidsusuallyfound
at hazardous waste sites.
2. Sorbed chemicals
Mobile, non-reactive compounds are most
effectively treated using pump-and-treat
technology. Contaminants easily sorbed
onto the soil matrix are more difficult to
remediate effectively. The volume of
AN EXAMPLE OF CALCULATING CLEANUP TIMES
Assume that the area of ground-water contamination is ten acres; the aquifer is
perraeableaMSS feet thick; water in storage is 30% of the aquifer's volatile; and
the water is contaminated with a nonreactive substance. Under these conditions
it would1 be possible to exchangeone pore volume of water in this te» acre plume
in about a year with % pumping rate of 100 gallons/mmote:
Volume of contaminant^ 10 acres x 43,560 ft* /acre x 55 ft. x 7.48 gal/ff x 0.3 =
remove rats wlumeia<3neyeat,«5.4 g i07gattonsr/365days/
^= 102 gallons per minute. However, it will take longer tilian one
year te> completely remediate the cootaminattt due to tlie tailing effect oftet*
observed when using pump-and-treat technology. Tailing is the asymptotic
4e^eas«ofcontanauaRtcon^^^
Severalphenomena may cause tailing, including thepreseneeof a highly soluble
aMmobilecontamRantthatmigrateslato less permeable sones of the geologic
material, a reactive, easily sorbed compound, and desorpdon. Sites with {ailing
effects require longer pamping times and greater pumping volumes to reach the
same level of remediation,' • •
-------
pumpedwater required toremovethecon-
taminant depends on the sorption capac-
ity, the geologic material through which it
flows, and the ground-water flow velocity
during remediation.
If the ground-water flow velocity induced
by pumping is too rapid, the contaminant
concentration levels will not reach equi-
librium. This results in decreased effi-
ciency of contaminant removal (See dia-
gram on the right). The retardation factor
of a chemical (contaminant velocity rela-
tive to the water velocity) can be deter-
mined to estimate potential sorption ca-
pacity and remediation time.
3. Hydraulic conductivity
Hydraulic conductivity is another factor
influencing the effectiveness of pump-
and-treat remediation. Favorable condi-
tions for pump-and-treat activities are high
hydraulic conductivity—greater than lO^
cm/sec—and homogeneity of the sur-
rounding aquifer material. Determining
pump-and-treat feasibility is specific to a
site. The same range of hydraulic conduc-
tivities may allow a pump-and-treat sys-
tem to be applied at one site but not at
another depending on physical site char-
acteristics and chemical properties of the
contaminant.
These limitations are not insurmountable
if accurate data collection and careful
planning are employed when designing
and operating a pump-and-treat system.
This Guide to Pump-and Treat
Ground-Water Remediation
Technology is a publication of the
Technology Innovation Office,
Office of Solid Waste and Emergency
Response, US EPA, Washington, D.C.
Walter W. Kovalick, Jr., PhD.
Director
AOVECTION
LIQUID: LIQUID
PARTITIONING
QROUNDWATER VELOCITY -»•
Liquid partitioning limitations of
pump-and-treat effectiveness
(Keely, 1989).
Keely, J.F., 1989. performance of
pump-and-treat remediations, EPA
Superfund Ground Water Issue,
EPA/540/4-89y005, Cincinnati,
Ohio.
U.S. Environmental Protection Agency,
1984a. Case studies 1-23: Reme- *
dial response at hazardous waste "
sites, EPA/540/2-84-002a* Cincin-
nati, Ohio.
TJ.S, Environmental Protection Agency, „
1984b. Summary report: Remedial
response at hazardous waste sites,
EPA/540/2-84-Q02b, Cincinnati,
Ohio. , ',
U.S. Environmental Protection. Agency*
1985. Modeling remedial actions at
uncontrolled hazardous waste sites,,
EPA/540/2-85/001, Cincinnati, '
Ohio.
Sv Environmental Protection Ageney,
!986a* EOtA^ounfrwatermonl*
toring technical enforcement guid-
ance document, OSWBR-99SQ.1,
Washington, D,C
, Ettv&onmentalProteetion. Agency,
, $986b. Superfund public health
evaluation manual,
060, Cincinnati, Ohio,
-U.S, Environmental Protection Agency,
x " l9$7a, A compendium of technolo-
gies ased to fite treatment of hazard*
Cincinnati, Ohio,
, Environmental Protection Agency,
I987b. Ua»dfeook: Ground Water,
Cincinnati, Ohio.
"XXS,. Environmental Protection Agency,
, I987b, MEflEQAl, an equilibrium
metal spectation model} osejfs
manual, EPA/606/3-&7/012, Athens,
/ Oeorgia.
U.S, Environmental Proteclion Agency,
1988b. OaJdance on remedial'
actions for contaminated ground
water at Superfaad sites* EPA/540/
G-88/003, Cincinnati, Ohio,
, Environmental Ptotectloa Agency,
1988b. Guidance for conducting
remedial investigations and
feasibility studies under CERCLA,
OSW8R- 9&5$3+Qlt Washington,
D.C.
XJ.S> Environmental Proteclion Agency,
l$88c> Gfround-water modeling; an
overview and status report, BPA/
$00/2-89/02,8, Cincinnati, Ohio,
XJ.S> Environmental Protection, Agency,,
1989. Evaluation, of ground-water
extraction remedies, Vols , 1 and 1
OSWER, Washington, D,C*
To obtain the Basics of Puyp* fa freafGmund'Water Remediation Twhnefogy call or write ths Centet1 fiw, ,
Environmental Research lajforwatiottj €i»daoatirOI«io 452*8.
•U.S. Govotnmon! Printing Olfico: 1991 — 548-187/20529
.
-------
-------
United States Center for Environmental Research BULK RATE
Environmantal Protection Information POSTAGE & FEES PAID
Aflancy Cincinnati OH 45268 EPA
PERMIT No. G-35
Official Business
Penalty for Private Use $300
EEft/540/2-90/018
------- |