g
(s
Cost-Effective Design of
Pump and Treat Systems
DISCHARGE TO
ATMOSPHERE
TREATMENT BUILDING
One of a Series on Optimization
-------
This page is intentionally left blank.
-------
Office of Solid Waste OSWER 9283.1-20FS
and Emergency Response EPA 542-R-05-008
(5102G) " April 2005
www.cluin.org
www.epa.gov/superfund
Cost-Effective Design of
Pump and Treat Systems
-------
This page is intentionally left blank.
-------
DISCLAIMER
Tliis document provides references to models and processes in use by outside parties and other Federal Agencies.
Mention of these models and processes does not imply endorsement for specific purposes.
Tliis fact sheet is not intended to be a detailed instruction manual. In addition, this fact sheet is not a regulation;
therefore, it does not impose legally binding requirements on EPA, States, or the regulated community, and may
not apply to a particular situation based upon the circumstances. The document offers technical recommendations
to EPA, States and others who manage or regulate ground water pump and treat systems as part of the Superfund
program or other cleanup programs. EPA and State personnel may use other approaches, activities and
considerations, either on their own or at the suggestion of interested parties. Interested parties are free to raise
questions and objections regarding this document and the appropriateness of using these recommendations in a
particular situation, and EPA will consider whether or not the recommendations are appropriate in that situation.
Tliis fact sheet may be revised periodically without public notice. EPA welcomes public comments on this
document at any time and will consider those comments in any future revision of this document.
-------
This page is intentionally left blank.
-------
PREFACE
Tliis fact sheet summarizes key aspects to consider for designing cost-effective pump and treat (P&T) systems. It
is part of a series of fact sheets that the EPA Office of Superfund Remediation and Technology Innovation
(OSRTI) is preparing as guidance for the ground water remediation community on effectively and efficiently
designing and operating long-term ground water remedies. This series is available at www.cluiii.org/optimization
and consists of the following fact sheets, plus others that will be available in the future.
• Elements for Effective Management of Operating Pump and Treat Systems
OSWER 9355.4-27FS-A, EPA 542-R-02-009, December 2002
• Cost-Effective Design of Pump and Treat Systems
OSWER 9283.1-20FS, EPA 542-R-05-008,'April 2005
• Effective Contracting Approaches for Operating Pump and Treat Systems
OSWER 9283.1-21FS, EPA 542-R-05-009, April 2005
• O&MReport Template for Ground Water Remedies (with Emphasis on Pump and Treat Systems)
OSWER 9283.1-22FS, EPA 542-R-05-010, April 2005
Access to a wider range of EPA documents is available at www.cliiin.org.
The recommendations contained in this series of fact sheets are based on professional experience in designing
and operating long-term ground water remedies and on lessons learned from conducting Remediation System
Evaluations (RSEs) at Superfund-financed P&T systems. The results of the first 20 RSEs conducted at
Superfund-financed P&T systems are summarized in Pilot Project to Optimize Superfund-Financed Pump and
Treat Systems: Summary Report and Lessons Learned (EPA 542-R-02-008a), and the site-specific
recommendations from the evaluations are available in the individual RSE reports (EPA 542-R-02-008b through
542-R-02-008u). The content of these fact sheets is relevant to almost any P&T system. Therefore, these
documents may serve as resources for managers, contractors, or regulators of any P&T system, regardless of the
regulatory program. Examples provided in this document have costs that are reasonable based on 2003 dollars
but do not reflect rigorous pricing through vendors.
-------
This page is intentionally left blank.
-------
A. INTRODUCTION
Remediation System Evaluations (RSEs) conducted at
20 Superfund-financed pump and treat (P&T) systems
have identified potential opportunities for reducing
annual operating costs without compromising
protectiveness at 17 of the 20 systems. On average, the
opportunities for each system translate to a potential
reduction of more than 30% in the annual operating
costs. The results of these RSEs are summarized in
Pilot Project to Optimize Superfund-Financed Pump
and Treat Systems: Summary Report and Lessons
Learned (EPA 542-R-02-008a). Cost reduction
recommendations included reducing labor costs,
simplifying the treatment plant, replacing treatment
components with more efficient units, and reducing
process monitoring. Many of these cost savings
opportunities arose because the parameters used to
design the P&T systems differed from the actual
parameters during system operation. This finding
suggests mat EPA project managers and other
environmental professionals would benefit from
guidance on designing cost-effective P&T systems.
An appropriately designed P&T system should achieve
the ground water remedy goals in a cost-effective
manner for the operating life of the system. Therefore,
the design of the P&T system should account for the
capital costs associated with system installation as well
as the annual costs for operation and maintenance
(O&M). In this instance and the remainder of the
document, the term "O&M" refers to activities
associated with operating and maintaining a P&T
system, and does not refer to any specific period of
time or regulatory status associated with the remedy.
For example, the Superfund program generally refers
to the first 10 years of a Fund-lead P&T system as
Long-term Response Action (LTRA), and the
subsequent period as "O&M". However, in this
document both of those time periods are considered to
be types of O&M.
System design generally occurs after site
characterization has been completed and usually
consists of the following steps:
considering remedy goals and associated
performance monitoring requirements
establishing design parameters (e.g., system flow
rate and influent concentrations)
selecting appropriate ground water
collection/extraction methods
selecting appropriate technologies for treatment of
each class of constituents
• determining an appropriate option for discharge of
treated water
incorporating appropriate system controls and
automation
Each of the above steps is discussed in this document
and design scenarios are provided in appendices for
two hypothetical sites as illustrative examples of cost-
effective P&T system design.
Because capital costs for installation and annual costs
for O&M are significantly higher than the costs of
designing a system, it is often appropriate to request a
design review from a third party. Once a system is
installed and operating reliably, the system
performance should be routinely evaluated to
determine if the performance and site conditions are as
expected. Changes to the system over time are
generally expected, due to evolving site conditions
and emergence of innovative technologies. The
system should be evaluated/optimized on a regular
basis by the site team each year and evaluated by an
independent party perhaps every five years.
TABLE OF CONTENTS
A. INTRODUCTION
B. REMEDY GOALS AND PERFORMANCE
MONITORING
C, SYSTEM DESIGN PARAMETERS 4
D. EXTRACTION SYSTEM 8
E. SELECTING THE APPROPRIATE TREATMENT
TECHNOLOGY 10
F. DISCHARGE OPTIONS 19
G. CONTROLS/REDUNDANCY/FAILSAFES .... 21
H. OTHER DESIGN CONSIDERATIONS 22
I. POST-CONSTRUCTION 22
J. CITED REFERENCES 23
APPENDICES A & B: ADDITIONAL ILLUSTRATIVE
EXAMPLES
-------
B. REMEDY GOALS AND PERFORMANCE
MONITORING
P&T systems are generally constructed and operated to
accomplish one or both of the following:
• Containment - prevent migration of a constituent
above a selected concentration to a receptor or
potential receptor
• Aquifer restoration - remove contaminant mass,
including non-aqueous phase liquid (NAPL) if
present, from an aquifer to achieve selected
cleanup criteria
In addition, a P&T system can be designed to meet
requirements for the discharged water, and possibly
discharged air, depending on the system.
The design of a cost-effective P&T system considers
these goals and requirements for the following reasons:
• The performance of each system generally
includes monitoring and evaluation with respect to
its goals, and a system can be designed to make
this performance monitoring easier and less costly.
As a remedy progresses and site conditions
improve, intermediate goals and milestones may
be achieved. A cost-effective P&T system is one
where operation of unnecessary system
components is discontinued.
Consideration of these two points during design is
facilitated by the concurrent development of an exit (or
closure) strategy and/or a performance-based
monitoring plan. An exit strategy generally is a
compilation of measurable milestones that indicates
progress toward remediation goals, specific conditions
that clearly indicate achievement of these milestones,
and a set of actions to occur (e.g., discontinuing a
component of the remedy) when these milestones are
achieved. If milestones are not achieved as expected,
this might be an indication that the remedial approach
and/or the goals need to be revisited. A performance-
based monitoring plan typically focuses on collecting
information that is necessary to document achievement
or progress toward goals.
The following three subsections describe how a system
design and elements of an exit strategy and/or a
monitoring plan can work together to result in a cost-
effective P&T system that addresses the remedy goals.
Containment, aquifer restoration, and meeting
discharge requirements are discussed separately.
Containment
Containment generally refers to hydraulic capture of
contaminants in a three-dimensional zone of the
subsurface and may pertain to a dissolved contaminant
plume and/or a NAPL plume.
Some containment remedies may be designed to
operate indefinitely, and may not be conducive to an
exit strategy. However, some of the contaminants at a
site may degrade over time or otherwise fall below
standards and not need further treatment. In such
cases, monitoring can be reduced, and components of a
treatment system may be discontinued if discontinuing
them does not compromise other treatment
components that are still in use. Therefore, although
an exit strategy may not be applicable for a
containment remedy as a whole, exit strategies for
some individual components (particularly those with
significant annual costs) are generally applicable.
Consideration of a performance-based monitoring plan
during design could lead to reduced annual costs and a
more cost-effective remedy. A performance-based
monitoring program for a containment remedy will
likely need to demonstrate plume capture through
interpretation of water level measurements and water
quality samples. In some cases, extracting more water
at a site may make evaluating capture easier and less
costly without substantially adding to treatment costs.
At those sites, it may be more cost-effective to
increase the extraction rate. Therefore, the scope of
monitoring needed to evaluate capture should be
considered for a variety of pumping scenarios. For
each scenario, the associated costs of this monitoring
and evaluation should be compared to the costs of
pumping and treating water.
"... it may be more cost-effective to
increase the extraction rate."
Aquifer Restoration
For aquifer restoration, an effective P&T system
design might have two components to the extraction
system: extraction wells dedicated to source control
(e.g., hydraulic capture of a NAPL source zone) and
extraction wells dedicated to restoration throughout
the rest of the plume. If feasible, this is a particularly
effective strategy because controlling the source area
allows the remainder of the P&T system to remediate
the downgradient portion of the plume. An exit
strategy associated with such a P&T system may
-------
include discontinuing the downgradient portion of the
extraction system when certain conditions are met but
continuing the operation of the source area extraction
system to maintain source control. With the majority
of the plume restored, source control and associated
monitoring can continue at a presumably lower cost.
This contrasts with an approach that aims to restore an
entire plume with little consideration to controlling the
source area. In such a case, the source area, which
may consist of elevated concentrations and/or NAPL,
can increase in size, and the entire system, rather than
just the source area component, will continue to
operate indefinitely.
"Control the source area to allow
the remainder of the P&T system to
remediate the rest of the plume in a
timely manner."
The specific conditions for shutting down any part of
the system or the whole system should be considered
during system design and clearly defined before
operation begins. If other remedial technologies that
can help meet these conditions are available at the time
of design or become available during O&M, a cost-
benefit analysis could be conducted to evaluate
continued P&T versus implementation of these other
technologies.
Given the above scenario, an appropriate performance-
based monitoring program would likely include limited
ground water quality sampling in the source area.
Because the source area would be controlled and not
necessarily restored, substantial water quality sampling
in the source area may not provide valuable
information regarding the performance of the remedy.
Meeting Discharge Requirements
Monitoring of the process water, and in some cases,
treatment system off-gas, is often conducted to
evaluate performance and document mat discharge
standards are met. Because process monitoring can be
a significant portion of annual O&M costs, system
design should consider how system performance can
be cost-effectively demonstrated by process
monitoring.
Monthly or quarterly sampling is typically specified by-
discharge permits to demonstrate that effluent is
meeting standards, but additional process monitoring is
generally determined by what is needed to operate the
plant. Because frequent sampling and laboratory-
analysis can become expensive, treatment plants, when
possible, should be designed to operate based on
readings from sensors and less frequent sampling with
laboratory analysis. Turbidity, oxidation-reduction
potential, and pH (in conjunction with influent and
effluent samples) often provide sufficient information
to operate a metals precipitation system. The pressure
differential across filters is often used to indicate
fouling. Air flow rates and pressure are generally-
sufficient to indicate an air stripper's performance.
The use of sensors is generally more cost-effective
than frequent sampling with laboratory analysis, and
the sensor data are provided in real time.
System designers should incorporate various sampling
ports throughout the treatment plant. Ports should be
located for samples to be collected from the influent
and effluent of each major treatment component so
that the efficiency of each unit can be determined.
This is especially helpful during system startup, when
more frequent sampling is appropriate, and is also
helpful for less frequent sampling throughout the
duration of the remedy.
Because site conditions change, the influent
concentrations of some constituents may fall below
discharge standards over time, and/or some extraction
wells may be taken offline. System designers should
anticipate the conditions that would allow
discontinuing the use of treatment components (or a
treatment system) so that these components are not
operated unnecessarily. The conditions that would
merit restarting or reincorporating those components
could also be determined, if changing site conditions
lead to an unexpected increase in influent
concentrations.
An onsite laboratory for frequent chemical analysis is
costly to install and operate and is rarely appropriate
once system operation is stable. Designers should
thoroughly consider the use of sensors before
including an onsite laboratory in the design. During
system startup, more frequent sampling may be
appropriate, but a short-term solution, such as sending
samples offsite or using a temporary mobile
laboratory, is generally more appropriate and cost-
effective in the long term than installing and operating
a permanent laboratory.
"An onsite laboratory is costly and
is rarely appropriate..."
3
-------
C. SYSTEM DESIGN PARAMETERS
Proper evaluation and selection of treatment equipment
during design requires reliable estimates of the
extraction system flow rate and influent characteristics.
System designers, however, should note that changes in
the values of these parameters will likely occur during
system operation, particularly in the first five years.
These changes in parameter values might result from a
change in site conditions related to the operating
remedy, such as modifications to the extraction system
or source control/removal. Ideally, a system is
designed to minimize equipment that becomes
unnecessary soon after operation begins. Regardless,
system components that are initially needed may later
become unnecessary. In such circumstances, the
system should be sufficiently flexible to allow these
components to be bypassed so that unnecessary
materials and labor costs do not continue. In some
cases, it may be appropriate to lease a particular
treatment component if the designer believes that the
component will only be needed for a short time. In
other cases, it may be appropriate to install equipment
that is appropriate for the reduced mass loading that is
expected over the long term. In such cases, the system
would operate with increased cost and/or modified
extraction in the short term to accommodate the
initially elevated mass loading and meet discharge
standards.
Selecting appropriate design parameters reduces the
likelihood that installed treatment equipment will be
unnecessary. Common design parameters are
discussed below. Terms in bold are summarized in
Exhibit 1.
System Flow Rate
Design Extraction Rate
Based on the remedy goals for plume capture and
source control discussed above, a system design flow
rate and extraction well locations can be developed.
The design flow rate should be the actual flow
expected based on well yield data, pumping tests, and,
if appropriate, modeling using site-specific conditions.
Modeling can also help determine the optimal
locations for extraction wells. The design flow rate is
used for estimating initial contaminant mass loading,
and, depending on the system, items such as granular
Exhibit 1
Summary of Design Terms Used for Purpose of this Guidance
Design flow rate: expected flow rate of P&T system calculated from estimated extraction rates necessary to achieve
remedy goals (e.g., plume capture).
This value should be used to select treatment components and to calculate the design mass removal rate.
Hydraulic capacity: maximum expected flow rate of P&T system, generally calculated by multiplying the design flow
rate by a factor of safety greater than 1.0.
777/5 value should be used to size pumps, piping, and tanks but should NOT be used to calculate the design mass
removal rate.
Design influent concentration (for each constituent or class of constituents in system influent): expected blended
influent concentration from all extraction wells based on concentrations obtained from sustained pumping conditions (e.g.,
after more than 24 hours of pumping) and not from routine monitoring data.
This value should be used to calculate the design mass removal rate.
Maximum influent concentration (for each constituent or class of constituents in system influent): maximum
expected blended influent concentration from combined extraction, typically calculated by multiplying the design influent
concentration by a factor of safety between 1.0 and 2.0.
The treatment system should be able to handle this concentration. Therefore, this value should be used to help
select a treatment process but should NOT be used to calculate the design mass removal rate.
Design mass loading rate (for each constituent or class of constituents in system influent): estimated mass loading
rate (pounds per day) to the treatment plant of contaminants in extracted ground water, calculated by multiplying the
design flow rate by the design influent concentration (Exhibit 2).
This value should be used for estimating materials/utilities usage when analyzing costs of various treatment
options.
-------
activated carbon (GAC) usage, sludge production, and
chemical usage. It can also be used to screen potential
discharge options. For example, as the flow rate
increases, discharging to a publicly owned treatment
works (POTW) becomes more expensive, whereas the
costs for other discharge options, such as surface water
or reinjection, often remain the same.
Hydraulic Capacity
A maximum system flow rate (hydraulic capacity)
should be calculated using the design flow rate
multiplied by a factor of safety. The factor of safety is
site-specific and is determined by professional
engineering judgment, based on the importance of
hydraulic capacity for system cost and effectiveness
and on the reliability of the design flow rate calculation
(e.g., the calculation will be more reliable if the yields
of the extraction wells have been estimated with pump
tests on either extraction wells or nearby monitoring
wells). For example, if the design flow rate for a
system is 2 gpm, a factor of safety of 5 (for a hydraulic
capacity of 10 gpm) is usually reasonable, primarily
because the same size treatment components would
likely be used for any flow rate at or below 10 gpm.
However, for systems with multiple wells and flow
rates over 50 gpm, the factor of safety should generally
be 2 or less, according to the design engineer's
judgment.
The hydraulic capacity is used to size treatment
equipment, pumps, and piping to allow appropriate
throughput. Pump and piping design should consider
flow velocities and head loss. If any system
components are fed by pumps that cycle on and off, the
flow rate while the pump is operating should be
considered, not the average flow rate over a cycle. For
example, the average flow through an air stripper may
be 50 gpm, but if the treatment system is designed to
operate in batch mode (i.e., intermittently) at a flow-
rate of 100 gpm, the air stripper should be designed
with a flow capacity of 100 gpm. To avoid over-
design, the hydraulic capacity should not be used to
determine GAC usage, sludge production, disposables,
or the selection of a treatment process. These items are
best determined with the design flow rate, which is the
best estimate of the expected flow rate.
Design Concentration
In addition to the system flow rate, concentrations of
constituents expected in the system influent should be
determined. Site contaminants of concern, as well as
naturally occurring inorganics, such as iron, manganese
and hardness, should be considered.
Extracting ground water changes ground water flow
patterns, blends contaminated and relatively clean
water, and often changes the oxidation-reduction
potential of the subsurface. Concentrations of both
organics and inorganics under sustained pumping
conditions are sometimes orders of magnitude lower
than under non-pumping conditions. As a result,
concentration data used for determining the design
concentration of a P&T system should be collected for
each potential constituent during sustained pumping
conditions from operating extraction wells or from
pump tests at monitoring wells in the vicinity of the
proposed extraction wells. For inorganics, both
unfiltered and filtered samples should be analyzed to
determine the potential need for treatment and the
potential for physical filtration to be an effective
primary treatment. The longer pumping can be
sustained and the closer that rate is to the expected
extraction rate, the more accurate the estimate of the
influent concentrations will be. In general, allowing
24 to 48 hours of pumping prior to sampling is a
reasonable compromise between the cost for the test
and the value of obtaining data under pumping
conditions. However, test pumping for longer periods
of time may be beneficial for wells with a lower yield
(e.g., less than 5 gpm).
Typically mobile large-volume tanks are used to store
water from pumping tests. Treatment of the generated
water depends on the volume, influent characteristics,
and discharge standards. If discharge to a POTW is
feasible, treatment may not be needed and the
relatively low volume of water from pumping tests
make discharge to the POTW cost-effective. In some
cases, approval can be obtained for short-term
discharge to a nearby surface water body or to a storm
sewer. If a discharge point is unavailable or
testing/treatment standards are extremely stringent,
off-site treatment at an appropriately permitted facility
may be warranted, with costs at these facilities
typically ranging from $0.25/gallon to $0.50/gallon.
Small GAC units and mobile air strippers are available
for very reasonable costs. On-site treatment of the
pumping test water may be cost-effective and can be
useful for testing the effectiveness of preferred
treatment technologies.
Data from representative wells should be averaged
(weighted based on the flow per well) to give the
design influent concentrations for each constituent
(Figure 1). The use of concentration data obtained
during sustained pumping conditions, rather man
during non-pumping or low-flow conditions, reduces
the chance of over-designing the system to handle an
erroneously high mass removal rate.
-------
Figure 1
Estimating the Design Influent with Data Collected During Sustained Pumping Conditions
60,000
50,000
3* 40,000
c
o
c
0)
u
c
o
o
LLI
O
30,000
20,000
10,000
57,000
D Sampling Without Sustained Pumping
(e.g., Ground Water Monitoring During
Remedial Investigation)
• Sustained Pumping Conditions During
Design
10,230
16,300
4,200
4,200
1,932
MW-1
(20 gpm)
MW-2
(20 gpm)
MW-3
(30 gpm)
MW-4
(30 gpm)
Design
Influent
Expected Extraction Well Location Indicated by Nearby Monitoring Well
(Expected Sustained Pumping Rate from each Extraction Well)
Using analytical data collected without sustained pumping (e.g., from ground water monitoring during a Remedial
Investigation) leads to a design influent concentration of 18,230 ug/L and a design mass removal rate of 22 Ibs per day.
Using data collected during sustained pumping conditions leads to a more representative design influent concentration of
1,932 ug/L and a design mass removal rate of 2.3 Ibs per day. The use of concentration data obtained during sustained
pumping conditions, rather than non-sustained pumping conditions, reduces the chance of over-designing the system to
handle an overestimated mass removal rate.
"The use of concentration data
obtained during sustained pumping
conditions, rather than during non-
pumping or low-flow conditions,
reduces the chance of over-
designing the system,.."
The design influent concentrations should be used with
the design flow rate to estimate the system design mass
loading for each constituent (Exhibit 2), which should
be used to estimate GAC usage, sludge production,
chemical usage, etc. A factor of safety based on the
reliability of the design concentration estimate
(generally not exceeding 2) should be applied to the
design influent concentration to provide the maximum
influent concentration. These concentrations should
be compared with actual or estimated discharge
standards to determine what constituents may need
treatment. The treatment system should be capable of
treating the maximum influent concentrations.
Even if discharge standards for iron, manganese,
hardness, and perhaps other inorganic compounds are
not anticipated, treatment of these constituents may
help prevent fouling of equipment, such as GAC units.
The designer should try to eliminate the need for
operator intensive metals removal by considering
equipment not prone to fouling or the addition of a
sequestering agent to keep metals dissolved. The
designer should also consider a temporary metals
removal system if inorganics concentrations are
expected to decrease soon after continuous pumping
begins, as they often do.
-------
Exhibit 2
Sample Method for Calculating Contaminant Mass Loading
Contaminant mass loading for water and air can be calculated for each chemical constituent in the extracted
water with the same basic equation. However, the units and conversion factors are different for air than they are
for water. To find the total mass loading for a class of constituents (e.g.. VOCs), the mass loading for that class
of constituents can be calculated by summing the mass loading rates for the individual constituents.
MH_n = Our, * CH
For Water:
3.785 L 1440mm. 2.2 Ibs.
. X X ; X '
gallon
dav
109 ug
MHO = mass loading in water (Ibs / day)
QKO
rate i11 water (gpm)
CHO = contaminant concentration (ug/ L)
For Air:
0.0283 m3 1440 mm. 2.2 Ibs.
ft3 * day " 106 mg
Mair = mass loading in air (Ibs / day)
Qair = flow rate in air (cfm)
Cair = contaminant concentration (mg / m3)
For air calculations, Cair in mg/m3 (with molecular weight, MWX, in grams per mole) can be obtained at 70°F and
a pressure of 1 atmosphere from parts per million by volume (ppmv) by the following steps:
Cair(mg/m3} =
Conc(ppmv) 1 mole air 1000 L 1000 mg
106
24.1 L
x MWY
m
Note: The conversion factor (1 mole air)/(24.1 L) vanes with both temperature and pressure. At a pressure of
I atmosphere and a temperature of32°F (0°C), the conversion is (1 mole air)/(22.4 L).
Approximate Molecular Weights (MW) in grams/mole of Common Volatile Organic Compounds (VOCs)
Benzene: 78
Carbon tetrachloride: 154
Chlorobenzene: 113
DCA: 99
DCE: 97
Ethylbenzene: 106
PCE: 166
TCA: 133
TCE: 131
Toluene: 92
Vinyl chloride: 62.5
Xylene: 106
Basing the influent concentrations on samples obtained
during sustained pumping is particularly cost-effective
if traditional monitoring well sampling techniques yield
organic contaminants with concentrations >1% of their
solubility or naturally occurring inorganic compounds
are present at levels that may need treatment. In both
cases, concentrations under sustained pumping may be
sufficiently low compared to traditional monitoring
well sampling results to allow for more cost-effective
treatment options.
Non-Aqueous Phase Liquid (NAPL)
The presence of NAPL (or even evidence that suggests
the presence of NAPL) in the subsurface complicates
remedial system design. NAPL can be found in three
general types at impacted sites:
lighter than water NAPL (LNAPL), such as
gasoline
• denser than water NAPL (DNAPL), such as
chlorinated solvents
• NAPL that has a similar specific gravity to water,
such as coal tar, that may be difficult to separate
from extracted water by gravity
In addition, NAPL can be found as either free product
or as residual product. Free product moves in a
separate phase from water and is recoverable via
extraction whereas residual product is trapped in the
pore spaces. Both free and residual NAPL serve as
continuing sources of contamination to the dissolved
contaminant ground water plume.
The presence of free LNAPL may impact extraction
and treatment component selection. If free LNAPL is
present in monitoring wells, it is an indication that
LNAPL is also present in the formation around the
well, and recovery tests using pumps or bailers should
-------
be conducted to determine the potential recovery rate
of LNAPL over a period of time. If the LNAPL
recovery rate is sufficient to merit automatic extraction,
rather than manual collection or another remedial
approach, product-only extraction pumps should be
considered in conjunction with water table depression
pumps. Recovered LNAPL would be handled
separately from recovered ground water. For example,
if the amount of recovered product is expected to be
more than about 10 gallons per day per well, total
fluids pumps can be used and a passive phase separator
at the head of the treatment system should provide
sufficient LNAPL removal for effective operation of
downstream units. When recovery rates diminish,
other removal technologies, such as soil vapor
extraction, can be used to address residual LNAPL.
Denser than water NAPLs may be identified by an
interface probe when product accumulates in the
bottom of a well, by specialized down-hole indicators
(e.g., FLUTe system), by inference based on
concentrations over 1% of solubility (Newell and Ross,
1992), or by other traditional or innovative methods.
DNAPL recovery tests should be considered at
extraction wells in areas where free DNAPL is
suspected in order to determine if the treatment system
influent will have DNAPL that may need removal.
Similarly, long-term pumping tests should be
considered at extraction wells in areas where either free
or residual DNAPL is suspected in order to estimate
dissolved influent contaminant concentrations. If
DNAPL is expected to be present in treatment system
influent, a passive phase separator at the head of the
treatment train is usually sufficient.
Coal tar and other NAPLs with a specific gravity near
1.00 may not be effectively removed from the extracted
water by a passive phase separator. Recovery tests are
often merited to estimate the NAPL recovery rate. If
that rate is less than about 10 gallons per day, an
oleophilic filter such as organo-clay can be used to
remove product, thus protecting downstream treatment
units. If a greater volume of NAPL is anticipated, a
dissolved air flotation (DAF) unit may be appropriate
for product removal. Because DAF units are expensive
to install and operate, designers should carefully
consider whether one is appropriate for the long-term.
If one is appropriate in the short-term, leasing the
equipment or using temporary pretreatment storage
tanks may be warranted so that if a decrease in product
recover^' occurs, an over-designed system will not be in
place for the duration of the remedy.
"Because DAF units are expensive
to install and operate, designers
should carefully consider whether
one is appropriate..."
D. EXTRACTION SYSTEM
This section presents typical components of a ground
water extraction system. The discussion for each item
includes rules-of-thumb for selection and design.
Selection of the appropriate components of an
extraction system should reduce capital expenses in
drilling wells and purchasing pumps. It should also
reduce long-term maintenance costs. Proper design of
an extraction system may also, in some cases,
eliminate the need for additional remedy components
such as barrier walls, if containment can be provided
cost effectively. Prior to installing barrier walls or
other remedy components that are meant to augment
the P&T system, a remedial alternative analysis should
be conducted. It may be more cost-effective, and
equally protective, to rely on P&T alone.
Vertical Wells/Angled Wells/Drains
Vertical wells are the type of wells most commonly
used for extraction systems. Normally, the vertical
well is screened in an impacted zone and pumped so
mat the water level is drawn down, causing impacted
water to enter the well. Vertical wells are the
preferred collection method where the ground surface
above the plume is accessible, and the aquifer provides
relatively high yields (>1 gpm per well) and saturated
thicknesses (>10 ft). Vertical wells are the only
reasonable collection method for plumes >100 feet
below the surface. When contamination is present at
distinct intervals along a well, the well can be designed
to focus extraction on those intervals that are most
contaminated and to reduce the amount of
uncontaminated water from other intervals.
Angled wells may be used below buildings or other
structures that do not allow access for vertical wells in
an appropriate area. Angled wells, however, have
limitations. For example, electric submersible pumps
are typically not used in angled wells because the
pump cannot be properly centered in the well, which
may result in damage to the well screen or overheating
of the pump. Therefore, pneumatic submersible
pumps or pumps with above-ground motors are
generally more appropriate in angled wells.
8
-------
Lateral drains can be effective for shallow plumes
where surface features and underground utilities are not
obstacles. Drains may be the only reasonable
alternative when the saturated thickness is < 10 ft and/or
hydraulic conductivity is low. Sump(s) are placed in
the low points of drains so that water may be collected
and pumped.
Wellheads
A wellhead should provide security and protection
based on the anticipated traffic loading. If an
underground vault is used, it should have access doors
that are easy to open and have adequate frost protection
for piping, including a heater, if necessary. Access in
the well cap should allow for water level measurement
or permanent transducer installation. Float switches
should be present to prevent a vault from flooding due
to a leak. A separate vault or above-ground installation
for the electrical junctions and controls is preferable to
locating them in a well vault. A separate vault may be
needed, depending on the flammability of the
contaminants or electrical codes. If possible, a flow
measurement device with a totalizer (for systems with
electric submersibles), a sampling port, and a flow
shut-off valve should be installed in each well vault (or
treatment building if each well has dedicated lines).
Confined space safety issues should be considered and
provide another reason to keep the well in a vault
separate from mechanical operating equipment.
Investment in properly designed well vaults facilitates
maintenance and monitoring in the future. Properly
constructing and developing wells will also improve
the performance of wells and minimize the need to
install additional wells to meet desired yield.
The installation of piezometers adjacent to extraction
wells is recommended to provide useful data for
development of accurate potentiometric surface maps
for performance monitoring. This is because water
level measurements in an operating extraction well are
generally not representative of aquifer water levels due
to well losses. These piezometers may also provide an
effective delivery point for well-rehabilitation
chemicals.
Pumps
Electric submersible pumps are the most commonly
used pumps in P&T systems and are appropriate for
wells yielding about 1.5 gpm or more. They consist of
a coupled electric motor and pump. The motor spins
the pump impeller, which pushes water out of the well.
Electric submersibles are the preferred choice for flows
>12 gpm from a well with depths to water exceeding 30
feet. Certain above-ground electric pumps (e.g., jet
pumps) can be used for shallow wells where the depth
to water during pumping is not greater than about 25
feet. Electric pumps can be operated with level
controls, amperage controls, or throttled to run
continuously in wells of any depth. Both submersible
and jet pumps are inexpensive (small ones are
<$ 1,000) and reliable in normal conditions. However,
if the pump is operating and the water level is below
the top of its intake, damage to the pump can occur.
Controls should be installed to prevent the pump from
operating if the water level in a well falls to the pump
intake. These pumps should not pump NAPLs and are
not as resistant to fouling as some pneumatic
submersibles.
Electric submersible pumps should be sized
appropriately. An oversized pump would need to be
throttled back, which adversely affects the pump motor
and wastes electricity (Example 1). This is another
reason why pump tests at extraction wells should be
conducted during the design phase. In general,
assuming 75% motor efficiency and $0.10/kilowatt-
hour (kWh),
1 horsepower = $900 / year
Example 1
Costs Associated with Oversized Extraction
Pumps
If 1 lip is required in each of four wells, switching from 5
lip pumps to 1 lip pumps can save approximately
$14,000 per year.
4 wells * 5 hp/well
4 wells x 1 hp/well
$900/year/hp
$900/year/lip
$18.000/year
- $3,600/year
Excess Cost $14.400/year
For flows greater than 100 gpm, an electric lineshaft
turbine pump, such as those commonly used for water
supply, may be warranted. These pumps are similar to
electric submersible pumps, but the drive motor is
mounted above-ground at the well head, and the pump
is easier to service than the large submersible pump
that would be needed to provide the same flow rate.
Pneumatic submersible pumps are typically $1,500 to
$3,000 and are powered by compressed air. When the
-------
pump intake is below the water level in the well, the
water enters the pump body. The compressed air then
evacuates the pump body and pushes the fluids to the
surface through tubing. If the water level drops below
the pump intake, the pump can remain online, and no
compressed air is used. Many pneumatic submersibles
can handle NAPLs and are very resistant to fouling and
chemical degradation. They are typically suitable for
wells with yields from 0 to 12 gpm where water is up to
300 feet deep. They are the preferred choice for
extracting water at flow rates below 1.5 gpm when the
water is more than 25 feet deep.
Vacuum pumps generally are suitable for extracting
water when more than five well points are necessary to
achieve extraction goals, the water is less than
approximately 25 feet deep during pumping conditions,
and ground water is in a relatively low permeability
formation. One pump draws a vacuum on a common
header system that has drop tubes in several wells.
Vacuum pumps are especially cost-effective when 15
or more well points are necessary. A vacuum pump
with an air/water separator, vapor controls, and sump
can cost $10,000 to $25,000. The vacuum pump
typically has greater maintenance needs than the
electric or pneumatic submersibles, but for cases where
it replaces many pumps, its life-cycle costs may be
lower.
Piston pumps have been used at deep, small-diameter,
low yielding wells, especially deep two-inch
monitoring wells that have been converted to extraction
wells. These pumps often need significant operator
attention. Often drilling a larger well and installing a
different pump results in lower life-cycle costs than
installing a piston pump.
Piping
High-density polyethylene (HOPE) is the pipe material
typically used for extraction systems, and it has
properties that often make it preferable to alternatives.
The pipe is flexible, non-brittle, and relatively chemical
resistant. Installation demands specialty equipment and
training, but HDPE typically provides for a reliable
pipeline. In very unusual cases, polypropylene, PVC,
steel, fiberglass, or other materials may be preferred.
Single-contained piping (i.e., common piping) is
typically employed for conveying impacted ground
water and should provide sufficient reliability.
Double-contained piping includes a second wall for
additional containment and typically leak detection
sumps and probes. It may be requested by regulatory
agencies if the lines cross property boundaries or
uncontaminated areas. It may also be requested by site
stakeholders if leak detection is preferred. Leak
detection systems, however, may have false positive
alarms due to precipitation, condensation, or other
malfunctions that can shut down the P&T system
unnecessarily. Double-contained piping can also cost
1.5 times as much as single-contained piping installed
below ground. Therefore, systems should typically be
designed to minimize the use of double-contained pipe
because effective installation and testing of single-
contained pipe generally provide a maintenance free
piping network.
E. SELECTING THE APPROPRIATE
TREATMENT TECHNOLOGY
Designing a cost-effective treatment system typically
involves selecting the most appropriate technologies to
treat contaminants. Generally, selection involves
using design parameters defined in Section C to
determine the life-cycle costs associated with each
treatment technology, comparing the costs and benefits
for each technology, and selecting the most
appropriate one, with the understanding that
concentrations will likely decrease over time and
potentially within a few months of operation. The cost
comparison should account for the capital expense for
installing the system, annual costs for system O&M,
and replacement or maintenance costs. For system
components with an expected life span that is shorter
man the expected remedy duration, include
replacement costs. For other items, an annual
maintenance allowance as some small percentage of
the installed capital cost should be included. A sample
calculation of the estimated cost for a single treatment
approach is shown in Example 2.
This section provides common treatment options for
various phases or classes of contaminants as well as
benefits, drawbacks, and general cost information for
each technology.
NAPL
NAPL is a potential form or phase of contamination.
It may be present in extracted ground water in an
emulsified form or in a free flowing form. In either
case, the product should be separated from water in
order to meet discharge standards and maintain the
effectiveness of other treatment processes.
Design Parameters to Consider for NAPL:
type of NAPL (light, dense, neutral)
• maximum influent flow rate to treatment svstem
10
-------
• estimated NAPL recovery rate
• form of NAPL (emulsified or free flowing)
Option 1: Phase Separator
• appropriate for free flowing LNAPL or DNAPL
• easily maintained with little labor
does not remove emulsified product
low capital ($ 15,000 for unit up to 50 gpm) and
operating costs
Option 2: Oleophilic Filter (e.g., organoclay)
removes both free flowing and emulsified product
removes LNAPL, DNAPL, and NAPL with a
specific gravity near 1.0
• easily maintained with little labor
• costly for large volumes of NAPL (e.g., more than
10 gallons of NAPL per day)
• removes approximately its weight in petroleum
product (~7 pounds per gallon) at approximately
$1 per pound
Option 3: Dissolved Air Flotation (DAF)
• appropriate for large volumes of NAPL with a
specific gravity near 1.0
operator intensive
produces emissions that may need control
technologies
• relatively high capital cost, only cost-effective for
large volumes of product
• consider leasing large fractionalization tanks
(phase separation) or DAF units on a short term
basis to avoid a permanent DAF installation
Organic Compounds
Organic compounds can be subdivided into a number
of categories, including volatile organic compounds
(VOCs), semi-volatile organic compounds (SVOCs),
and non-volatile organic compounds. Other
classifications also exist, including chlorinated vs. non-
chlorinated compounds. However, volatility is of
particular interest because those contaminants that
volatilize often can be removed from the extracted
ground water and transferred to the vapor phase where
they are addressed more economically.
Removal of organic compounds from ground water is
generally accomplished by partitioning them to air and
collecting the vapors, by partitioning them to solids, or
by destroying them. The most cost-effective and
reliable approach will depend on site-specific
conditions. The following sections outline the more
prevalent treatment approaches and the factors that are
used in determining the most appropriate choice. A
summary of general advantages and disadvantages for
various technologies is provided in Exhibit 3.
Example 2
Sample Cost Analysis of a Potential Treatment
Option
Hypothetical Site Conditions
contaminant: VOCs
treatment technologies:
air stripping
GAC for offgas treatment (off-site regeneration)
design mass removal rate: 2 Ibs/day
• system highly automated
Estimated Capital Costs for Air Stripping Option
Category
Air stripper (installed)
GAC units (installed)
Electrical, controls, building, etc.
Engineering, startup,
contingency
Total Capital Costs
Estimated
Cost
$60,000
$10,000
$50,000
$40,000
$160,000
Estimated Annual O&M Costs for Air Stripping Option
Category
Labor
O&M operator
Engineering
Electricity
Materials
• Vapor GAC
Chemical analysis
Maintenance
Estimated
Units
< 8 hours/wk
32 hours/yr
5HP
7,500 Ibs/yr
3 samples/mo.
varies
Total Annual Costs
Estimated
Cost
$25,000/yr
$5,000/vr
$20,000/yr
$5,000/yr
<$5,000/yr
$60,000/yr
Estimated Life-Cycle Costs for Air Stripping Option
Capital Costs
Annual Costs (30 years)
30-year life-cycle costs
30-year life-cycle costs (discounted
at 5%)
$160,000
$1,800,000
$1,960,000
$1,082,000
This analysis for air stripping would be compared with
an analysis for a competing treatment technology, such
as UV/Oxidation. The extraction system, reporting,
long-term monitoring, project management, oversight,
and contracting costs are not Included In the above
costs because they are assumed to be the same for each
competing treatment option.
11
-------
Exhibit 3
Summary of Key Aspects of Treatment Technologies
Technology
General Advantages
General Disadvantages
Organic Compounds
Air Stripping
GAC
Polymeric Resin
Biological Treatment
UV Oxidation
appropriate for most VOCs and some
SVOCs
low operator labor needs
relatively low O&M costs
relatively low capital costs
appropriate for many organic
compounds, including VOCs, SVOCs,
and other non-VOCs
remove some metals and other
inorganics (generally < 90% efficient)
low operator labor needs
relatively low O&M costs (when used
for appropriate situations)
relatively low capital costs
appropriate for some constituents that
are not effectively treated by GAC
low operator labor needs
• effective for high contaminant
concentrations, especially with on-site
regeneration
often effective for constituents that are
not easily removed by air stripping or
GAC (e.g., ketones, ammonia)
fixed-film units reduce nutrient usage
destroys all types of organic compounds
on-site and does not need off-gas
treatment
appropriate for some constituents that
are not easily removed by other methods
(e.g., 1,4-Dioxane)
• not appropriate for many SVOCs and
other non-VOCs
• off -gas often needs treatment
• not appropriate for some organic
compounds (usually low molecular
weight VOCs)
• more likely to need pretreatment for
solids removal than air stripping
contaminant specific and not
appropriate for process water with
many constituents
• more expensive than GAC for many
common contaminants at typical
concentrations found in ground water
• relatively operator intensive relative to
air stripping or GAC
• generates solids that may need disposal
high capital costs relative to air
stripping and/or GAC
often has higher O&M costs than air
stripping or GAC
more likely to need pre-treatment for
metals and/or solids (to prevent
fouling) than air stripping or GAC
Inorganic Compounds
Filtration
Settling and/or Metals
Precipitation
Solid Phase Partitioning
and Ion Exchange
low operator labor needs
often relatively low capital and O&M
costs
effective and reliable for metals removal
including chromium and arsenic (with
proper pH adjustments)
low operator labor needs
available for various metals
may not be sufficient to remove solids
and/or metals to discharge standards
bag or cartridge filters might need
frequent replacement and disposal
operator intensive
relatively high capital and O&M costs
relative to other treatment components
may generate a substantial amount of
solids that need disposal
not cost-effective for high
concentrations
not appropriate for multiple
constituents needing removal
12
-------
Air Stripping
Air strippers are widely used to remove VOCs, and. to
a limited extent, SVOCs. from ground water. Systems
have been in place and reliably treating ground water
for decades. The systems are generally reliable and can
operate unattended. O&M costs for air stripping are
generally low because associated labor and materials
are minimal. Power consumption is generally limited
to the operation of a blower or fan. Exhibit 4 presents
additional information on selecting appropriate air
strippers. Appendix A presents an example with a cost
analysis for selecting an air stripper as the treatment
technology for VOCs. Computer software programs to
select an appropriate air stripper based on flow rate,
influent concentration, and air and water temperature
are readily available from vendors. Selection of an air
stripper model can be somewhat conservative as
capital and O&M costs do not increase significantly
with more efficient models.
Because air stripping transfers contaminants from
water to air, the air stripper off-gas may need
treatment, which is discussed later in this section.
Exhibit 4
Considerations for Selecting an Appropriate Air Stripper
Design Parameters to Consider when Selecting an Air Stripper:
hydraulic capacity
maximum influent concentration
contaminants of concern
potential for fouling
design mass removal rate (to determine potential for off-gas treatment)
height or space limitations
visual impact or zoning restrictions
Diffuser: Air is injected into a tank through piping with small holes or slots allowing small bubbles to exit and contact the
water. This method is the least susceptible to fouling by iron precipitation, biological growth, or hardness. However,
stripping efficiency is relatively low; it is difficult to achieve >95% efficiency. Flow rates above 50 gpm likely need large
tanks and high horsepower blowers. This type of stripping should be chosen only for flow rates <50 gpm where lower
stripping efficiency is acceptable. It is especially suitable when flow rates are <10 gpm and significant fouling is expected.
It is relatively more difficult to capture emissions from this type of stripper if off-gas treatment is included. This aeration
method can also be useful as part of a metals precipitation process.
Tray Stripper: Water flows over a series of trays within a chamber where air is blown in and moves upward through small
holes in the trays. The water froths on the trays as it flows down to the stripper sump; water does not pour through the holes.
The tray holes are susceptible to fouling but can be easily cleaned on a schedule (usually monthly to annually). Stripping
efficiencies >99% can easily be achieved and maintained with proper maintenance. The unit size and high horsepower
needs compared to packed towers (due to greater air pressure) usually limit the effective range to flow rates of 5 gpm to 500
gpm. Emissions can be easily captured from these units, and they can be easily placed within buildings minimizing any
weather issues. The tray stripper is usually the best choice for treating 10 gpm to 300 gpm, especially for sites in colder
climates or sites with potential fouling issues and/or needing emissions treatment.
Packed Tower: Water flows from the top of a 15 to 35 foot high tower downward though plastic packing material while air
is sent up from the bottom of the tower. The tower packing is susceptible to fouling, especially when influent iron
concentrations are above 5 mg/1, and may need regular acid washing, cleaning with peroxide or biocide, and periodic
packing changeouts. Stripping efficiencies >99% can easily be achieved and maintained with proper maintenance. The
packed tower can be effective for flows from 1 gpm to 5000+ gpm. Capturing emissions from a packed tower is relatively
expensive and can lessen stripping effectiveness. The packed tower is usually the best choice for treating >500 gpm and
may be chosen for lower flow rates if fouling is minimal.
Common VOCs with Low Air Stripping Efficiencies
Acetone t-butyl alcohol (TEA)
Naphthalene 1,2-Dichloroethane
2-Butanone (MEK) MIBK
MTBE
Tetrachloroethane (1,1,1,2 or 1,1,2,2)
1,1,2-Trichloroethane
13
-------
Solid Phase Partitioning
Option I: Granular Activated Carbon
In this approach, ground water contaminants are
removed by adsorption to GAC. Two units are
generally placed in series with regular sampling
(perhaps monthly or quarterly) after the first unit to
determine when that unit has reached its chemical
loading capacity (i.e., when chemical breakthrough
occurs). Any contamination that breaks through the
first unit in between sampling events is adsorbed in the
second unit. When breakthrough of the first vessel is
detected, the carbon in the first vessel is changed out
and the vessel configuration is changed so that the
previous "backup" unit is now used as the lead unit.
Many VOCs, SVOCs, and non-volatile organics are
readily adsorbed by GAC while others, such as MTBE,
TEA, and ketones, are not. The carbon usage depends
on the contaminant characteristics and concentrations,
(see Exhibit 5) as well as other constituents in the
water that may compete for active sorption sites.
Estimating usage and sizing vessels can be
accomplished with vendor software.
The cost for GAC typically ranges from $1 to $3 per
pound for a complete change out, though unit costs
may be higher for smaller change outs. Used GAC is
usually taken off-site for regeneration so that the GAC
can be used again. In some circumstances, used GAC
may be disposed of in a landfill. In either case, the $1
to $3 per pound estimate noted above includes the
disposal or regeneration costs.
GAC is prone to biological or mineral fouling if
process water is not properly filtered or pre-treated.
Regularly backwashing the GAC can help alleviate this
fouling. If GAC fouls it may need replacement before
it has reached its capacity for contaminant removal. In
many cases, early replacement is not cost-effective;
however, in some limited cases, it may be more cost-
effective to increase the replacement frequency than it
is to provide pretreatment.
Appendix A presents an illustrative example with a
cost analysis for selecting GAC as a treatment
technology for VOCs.
Option 2: Polymeric Resin
This technology works in a similar fashion to GAC and
can be adapted for on-site regeneration. At high
concentrations, the on-site regeneration may make this
technology more cost-effective than GAC. Some
polymers may also be more effective for some
contaminants that are not easily adsorbed by GAC,
such as 1,2-dichloroethane, methylene chloride, and
ketones. Polymeric resins are engineered for specific
chemical characteristics, and resins are available for a
variety of contaminants.
Contaminant Destruction
Option 1: UV/Oxidation and Ozonation
This technology destroys organic contaminants by
oxidizing them. Therefore, when comparing costs of
this technology to that of air stripping, the treatment of
air stripping off-gas treatment associated with air off-
gas associated with the air stripping should be
included. UV/Oxidation is more prevalent man
ozonation. Because other constituents can compete for
oxidants, obscure the UV light source, or otherwise
foul the reaction chamber, pre-treatment and filtering
is often needed, more so than for other technologies.
Frequent cleaning or replacement of a UV/oxidation
system or its components is generally not a potential
alternative to pre-treatment, as it is for air stripping or
GAC.
UV/Oxidation can be used for a variety of organic
compounds, including some that are not easily
removed by air strippers or GAC, such as 1,4-Dioxane.
Treatment efficiencies can be as high as 99%, but are
much lower for some constituents.
O&M costs can be relatively high due to replacement
parts (lamps, seals, and other parts), high concentration
hydrogen peroxide, and electricity needed to power the
lamps (often 30 kW). Utilities and maintenance items
alone may cost over $50,000 per year. Appendix A
presents an illustrative example with a cost analysis for
considering UV/oxidation as a treatment technology
for VOCs.
Option 2: Bioreactors
Bioreactors utilize microbes to degrade organic
contaminants and consist of a fixed-film media (such
as sand or activated carbon) within a tank or canister.
The organic contaminants serve as nutrients to the
microbes, and the fixed media provides a surface
where the contaminants accumulate and the microbes
can grow. By accumulating contaminants near the
microbes, the fixed media allows bioreactors to
function at lower contaminant concentrations than
suspended growth activated sludge systems that are
typically used for municipal wastewater treatment. In
general, activated sludge systems that have been
designed for P&T systems are often ineffective
because concentrations of organic constituents are too
low to maintain the microbial population.
14
-------
Exhibit 5
Sample Method for Calculating Preliminary GAC Design Estimates
There are two values that are typically used to determine the practicality and cost-effectiveness of GAC: the GAC usage
and the size of the GAC units. Both of these can best be estimated by a GAC vendor, but the following calculations allow
for a rough estimate of both values for a single chemical constituent in the process water. The estimated total GAC usage
and GAC unit size is the sum of the usage and unit size for each constituent.
GAC usage for organic compounds is both chemical and concentration dependent and can be estimated using a Freundlich
isotherm as follows:
1. Calculate the ratio of pounds of each contaminant to pounds of GAC needed, R, given the design influent concentration.
R =
1000
x C
UN
where
K = is a tabulated partitioning coefficient (see below) with units (mg/g)(L/mg)1/N
\IN= a tabulated dimensionless parameter (see below)
C = design influent concentration with units mg/L
2. Calculate the GAC usage rate (GUR) in pounds per day for each contaminant based on design mass removal rate.
GUR =
MRR
R
where
MRR = the mass removal rate (see Exhibit 2) with units pounds/day
R = ratio of pounds of each contaminant to pounds of GAC needed (no units)
Sizing a GAC unit should account for empty bed contact time (EBCT) between the process water and the GAC as well as a
convenient GAC changeout schedule. The EBCT should generally be between 15 and 30 minutes per vessel. The
approximate vessel size in "pounds of GAC" can then be calculated as follows:
1ft3 SOlbs of GAC
Vessel Size = EBCT x Hydraulic Capacity x x
7.48gal
ft3
where the hydraulic capacity is provided in gpm.
Ideally, the vessel should be sized to allow changeouts to occur on a quarterly or semi-annual basis when other site
activities are conducted.
The (K) and (1/N) parameters for some common compounds are provided below.
Contaminant
^(mg/g)(L/mg)1/N
1/N
Toluene
100
0.45
Chlorobenzene
100
0.35
Lindane
285
0.43
PCE
51
0.56
TCE
28
0.62
Methylene
Chloride
1.3
1.16
Parameters from Dobbs and Cohen, 1980 (EPA-600/8-80-023).
See Appendix A for an example calculation of GAC usage and vessel sizing.
15
-------
Bioreactors may be sensitive to oxygen concentrations
and may be difficult to start up. They may be most
appropriate for some organic contaminants that need
treatment but are not easily stripped or adsorbed to
GAC, such as ketones, and for some inorganic
contaminants, such as perchlorate, nitrate, or ammonia.
Systems capable of treating up to several hundred gpm
are readily available. Solids generated from microbial
biomass are generated during operation and need
proper disposal. As with UV/Oxidation and ozonation,
bioreactors typically become less cost-effective as
contaminant concentrations decrease overtime. The
cost-effectiveness should be periodically evaluated
against GAC and other alternatives.
Inorganics in Ground Water
Removal of inorganics in ground water, primarily
metals, is generally achieved by some combination of
metals precipitation, settling, and filtering or by ion
exchange. In some cases, filtering media may be
enhanced by activated alumina or carbon to adsorb
metals. In limited cases, reverse osmosis (a form of
filtering) can be used, but due to the high potential for
fouling, other technologies are generally more cost-
effective. Exhibit 3 provides general advantages and
disadvantages for common metals removal approaches.
Metals in ground water may be dissolved in the water
or sorbed to suspended particles, including colloids and
larger particles. The removal of dissolved metals
generally is performed with chemical precipitation,
whereas removal of metals that are sorbed to particles
might be achieved through filtration alone.
With the exception of filtration alone, metals removal
is typically expensive, both in terms of capital and
annual costs, relative to the technologies for other
contaminants of concern. Therefore, installation and
long-term operation of metals removal systems should
be avoided, when possible. If filtering alone does not
provide sufficient metals removal during treatability
tests, site teams should thoroughly consider alternate
discharge options, pumping locations, and even
alternative remedial strategies not using ground water
extraction prior to installing a permanent metals
removal system with a high expected mass removal rate
(> 10 pounds per day of dry solids).
Metals removal is often not needed for ground water
remediation but is needed to preserve other treatment
processes or to meet discharge standards. Iron, for
example, may be present in ground water at sufficient
concentrations to foul a UV/oxidation system, an air
stripper, or GAC, or may simply exceed the discharge
standards. In such instances, cost-effective alternatives
to metals removal may include adding sequestering
agents to keep the metals in solution, increasing
maintenance and cleaning of other treatment processes,
or utilizing other discharge options.
Metals removal, at least in the short term, is generally
unavoidable when the constituents of concern are
metals or when the design influent concentrations for
iron or other nuisance constituents exceed 15 mg/L or
mass loading exceeds 10 pounds per day of dry solids,
respectively.
Aquifer conditions change once pumping begins. As a
result, metals removal may appear necessary based on
initial data but may only be needed for the short-term
(i.e., less than a year) or not at all. For this reason, the
need for a permanent metals removal system should be
thoroughly evaluated prior to installation. Leasing of a
system over several months may be more cost-
effective than installing a system, operating it, and
heating the large space used to house it over the long-
term. In addition, leasing the system may allow
various approaches for metals removal to be pilot
tested under site-specific conditions.
Exhibit 6 presents a potential framework for evaluating
the need for metals removal system. In addition,
Appendix A presents an example of how high metals
concentrations at a hypothetical site may be addressed
during design.
Filtration/Reverse Osmosis
Filtration of process water can be used for pre-
treatment to remove suspended particles, after metals
precipitation/settling to remove precipitated particles
that have not settled, or near the end of the treatment
process to prevent injection wells from fouling due to
solids loading.
If a clarifier or settling tank is included in a treatment
plant, filters comprised of sand, anthracite, and some
other media can be used. These filters can be
backwashed rather than replaced. Backwashing can be
automated, based on the pressure differential across
the media, and some continuously backwashing units
are also available. Rebedding of the media is likely
needed periodically (on the order of years). The
backwash, with participates, is directed to the clarifier
for settling and eventual removal as sludge, which then
needs proper disposal.
Cartridge or bag filters with disposable media can be
used to remove participates of 100 microns down to <1
micron. During regular operation, bag filters become
increasingly clogged with particulates and need
16
-------
Exhibit 6
Potential Decision Tree for Addressing Metals in Ground Water
Discharge standards for all constituents vary with discharge location (see Section F). To cost-effectively
design a ground water remedy when metals are present, consider the following example decision tree for
EACH discharge option.
Estimate influent metals concentrations by sampling
during sustained (e.g., 24 hours) pumping conditions
i
r
Are metals above discharge standards'?
yes
Are concentrations high enough to
affect other treatment components?
Are mass loading and ground water
quality favorable for ion exchange?
yes
Is frequent cleaning, use of filters,
and/or use of sequestering agents
more cost-effective than pre-
treatment for metals?
yes
No specific action
required for metals
Metals
precipitation and/or
flocculation
Evaluate costs for all options for each discharge location, and proceed with pilot test of most favorable option(s) as
necessary. If metals precipitation or another large scale system is required, consider leasing equipment for several
months to determine if problem is short-lived, or if a permanent system is required.
regular replacement, with used bags taken off-site for
disposal. As a result, these filters are generally not
appropriate to remove particles directly downstream of
a metals precipitation unit because replacement is too
frequent to be practical.
Reverse osmosis and nanofiltration are typically
potable water treatment technologies. They remove
dissolved inorganics and small particles from process
water and may be appropriate as a polishing step or for
primary treatment of ground water with very low levels
of suspended solids. The systems usually have very
high associated maintenance and waste disposal costs if
solids levels are elevated.
Metals Precipitation/Settling
This process uses chemicals to turn dissolved metals
into solids (i.e., metals precipitation) that can then be
removed from the process water by settling or filtering.
This approach includes the addition of multiple
chemicals, including oxidants, acids and bases for pH
adjustment (optimal pH range varies depending on the
metal), and polymers. For some applications, the
addition of iron is also performed.
Filtration may be sufficient for removal of solids, but
settling or clarification is typically needed when
influent solids loading to a system exceeds 10 pounds
17
-------
per day. In most ground water applications, a lamella
(inclined-plate) clarifier is used to minimize space
needs, and filtration is generally performed after the
clarifier.
Precipitation/clarification and associated maintenance
need regular operator monitoring. Because metals
precipitation results in the generation of solids.
additional operator labor may be needed for sludge
handling. Therefore, operator labor is a large cost
driver for metals precipitation systems. The disposal of
solids, particularly if they are considered hazardous,
will also contribute to the cost of the remedy. Sludge
dewatering is often conducted with a filter press to
reduce the amount of sludge needing disposal. In many-
cases, the concentration of metals in the influent is
overestimated, and the filter presses are used
infrequently. In such cases, it may not be cost-effective
to purchase, operate, and maintain a filter press. If
sludge generation is likely, consideration should be
given to using engineered sludge drying beds or
potentially disposing of sludge in liquid form.
Solid Phase Partitioning/Ion Exchange
Activated alumina, activated carbon, and other media
may be used for solids filtering and will also provide
some adsorption of metals and other inorganics. Using
adsorption alone is potentially applicable for treating
water that does not need 90+% removal of a specific
compound.
With ion exchange, the ground water contaminant is
removed by passing ground water through a canister
containing a resin. This resin adsorbs a higher valent
inorganic by exchanging it for a lower valent one
(sodium is used in many home systems). The resin
needs periodic regeneration (generally done off-site)
based on the contaminant mass removed. Compared to
metals precipitation systems, these units can operate
unattended and need little maintenance. However,
periodic regeneration can be costly if the mass removal
rate is high. The resin can also foul if solids are
present and not sufficiently filtered. In general, these
units are preferable to metals precipitation for
hexavalent chromium or other metals
with relatively clean water (low suspended solids). Ion
exchange can also be effective for non-metal
inorganics, such as perchlorate.
Off-Gas Treatment
Treatment of off-gas from a P&T system may be
needed due to regulatory requirements or a preference
for not transferring contamination from ground water to
the atmosphere. The technologies in this category can
be used to effectively treat off-gas from air strippers or
other systems. Selection is dependent on the analysis
of life-cycle costs. When comparing the costs of air
stripping with other technologies, the cost of off-gas
treatment should be included.
Granular Activated Carbon
GAC also adsorbs the volatile organics present in off-
gas, though it should be noted that GAC units designed
for water treatment are not interchangeable with GAC
units designed for vapor or off-gas treatment. The
GAC usage rate can be estimated using software
programs available from vendors. GAC usage is
usually much more favorable for vapor phase
treatment with GAC than it is for liquid phase
treatment with GAC. In general, a given mass of GAC
adsorbs 4-10 times more contaminant mass in the
vapor phase than in the liquid phase. Depending on
the amount of GAC use projected, it may be
regenerated on-site or off-site. A reasonably accurate
estimate of contaminant mass loading (see Exhibit 2)
should be used for determining carbon usage and the
best option for regeneration.
Off-site Regeneration'Disposal: This is the most
common method of GAC regeneration. Capital costs
for installation are relatively low, involving only the
carbon container and duct work (on the order of
$10,000 for mid-size units). This technology is
generally preferred at long-term carbon usage rates up
to 200 pounds per day or more. GAC replacement
costs (including regeneration or disposal) may range
from $1 to $5 per pound.
On-Site Regeneration: This technology includes a
capital investment of at least $150,000, as well as
significant operator attention and energy costs when
regeneration is needed. It may be appropriate if
carbon usage is projected to be over 500 pounds per
day for long-term operation and the BTU content of
the off-gas is low. Given the difference in capital
costs between on-site and off-site regeneration, the
design mass removal rate should be carefully estimated
with concentration data that are representative of
pumping conditions. On-site regeneration is rarely
cost-effective, especially when designers consider
decreases in influent concentrations overtime.
Because the adsorbed contaminant is removed on-site,
equipment and procedures for properly storing and
disposing of the recovered product are necessary.
Polymeric Resins
This technology is similar in capital cost and operation
to activated carbon with on-site regeneration. It is
18
-------
appropriate for treating contaminants, such as 1,2-
dichloroethane, that are not readily adsorbed by carbon.
When fewer regenerations are needed, the operator
labor and energy costs are reduced. As with on-site
regeneration of GAC, equipment and procedures for
properly storing and disposing of the recovered product
are necessary.
Thermal/Catalytic Oxidation
VOCs are destroyed at a temperature of 800 (catalytic)
to 1500 (thermal) degrees Fahrenheit. Because
contaminants are destroyed, there are no treatment
residuals that need further handling or disposal.
However, if chlorinated hydrocarbons are present in the
treated vapors, an acid gas scrubber may be needed to
meet emission standards, and equipment replacement
schedules will likely be accelerated due to corrosion
issues.
A small unit (e.g.. for a gas station SVE system
operating at 100 cubic feet per minute) costs about
$50,000, but most off-gas oxidizers cost at least
$500,000. Natural gas is used to supplement the
contaminant vapor during operation. Few ground water
treatment systems have off-gas concentrations high
enough to warrant this technology, and natural gas
costs may be as high as $20,000 per month. When
combined with an SVE system, the technology may be
cost-effective for a short term; however, as mass
loading decreases, another form of off-gas treatment
generally becomes more cost-effective. This
technology is generally considered for short-term
leasing until contaminant concentrations decrease so
that a more cost-effective technology may be used.
F. DISCHARGE OPTIONS
The treatment system design should consider the
discharge standards, which will vary based on the
discharge method selected. Four common discharge
methods for treated water are discussed below. The
discussion includes how each option may impact the
design.
Surface Water/Storm Se\ver (NPDES)
Under the National Pollution Discharge Elimination
System (NPDES), treated water may potentially be
discharged directly to a nearby surface water body or
indirectly through a storm sewer.
Potential Advantages
Discharge typically is not subject to a flow-based
fee, but some storm sewer systems will.
Potential Disadvantages
Discharge standards are based on ambient water
quality and may be comparable or more stringent
than drinking water standards. Reporting may
also be more rigorous than other discharge
options.
• Natural constituents in ground water may need to
be removed before treated water is discharged.
• Environmental toxicity testing may be needed.
• Access to a nearby surface water body or storm
sewer is needed.
Cost Information
• Piping effluent from the treatment system to a
surface water body or storm sewer is needed.
This discharge option is preferable in many cases
and is commonly selected. It is usually
significantly more cost-effective than POTW
discharge for high flows.
• Monthly sampling of influent and effluent for
many parameters is typically needed.
Publicly Owned Treatment Works (POTW)
Extracted water is discharged to the POTW, where it is
treated prior to discharge to a surface water body. In
some cases, discharge can occur with no other
treatment needed. Appendix B presents an example
with considerations for selecting either surface water
or a POTW for discharge.
Potential Advantages
This option often has less stringent discharge
standards and monitoring requirements, especially
for organics. Generally, municipalities use a
discharge limit of 2.13 mg/L for total toxic
organics, which is a higher concentration than the
influent to many P&T systems.
• Ketones and ammonia, which are difficult to treat
with P&T systems, are easily treated by POTWs.
The POTW provides a secondary treatment for
some constituents to prevent (except in extreme
cases) damage to surface water receptors.
19
-------
Potential Disadvantages
Potential Disadvantages
If the POTW is near capacity, it may not accept
treated ground water discharge.
• In certain areas where ground water is the sole
source of drinking water, reinjection of treated
water may be necessary.
The POTW may be reluctant to accept water that
has certain constituents or is relatively "clean"
compared to typical sewer water.
Cost Information
Capital expenditures involve piping discharge to a
sanitary sewer connection point.
• The POTW typically charges $0.003 to $0.03 per
gallon (2003 dollars). Discharge to the POTW, if
available, may be a good option for low (<30 gpm)
flows and in situations where meeting discharge to
surface water criteria is difficult and/or expensive.
Regular sampling of constituents may be needed,
but the frequency may be reduced (to quarterly or
semi-annually) once the P&T system has been
proven reliable, and conditions are stable.
Reinjection
Treated water is reinjected to the subsurface through
wells, galleries, or basins.
Potential Advantages
• Discharge standards are typically similar to
drinking water standards, but for some
constituents, such as ammonia, standards may be
more relaxed than those for discharging to surface
water.
Unlike discharge to surface water, there generally
are not requirements to remove some natural
ground water constituents prior to reinjection.
• Reinjection can be used for assisting hydraulic
containment or flushing of a contaminant source.
In these situations, the effects of reinjection should
be considered in the extraction system design.
Reinjection of treated water helps conserve ground
water as a natural resource, which is particularly
beneficial in areas where ground water serves as a
sole source for drinking water or where dewatering
is a concern.
Reinjection into the heart of the plume may
compromise plume capture by spreading the
plume. Additional hydrogeological analysis
(perhaps including modeling) may be necessary in
designing the extraction/injection systems.
• Reinjection systems may need more maintenance
than oilier discharge options, especially due to
solids or biological fouling.
Cost Information
Reinjection systems typically involve greater capital
expenses than other discharge methods, unless other
discharge locations are unusually distant. Therefore,
reinjection may be selected for discharge if the
distance to other discharge points is excessive or if
discharge back to the aquifer is necessary to meet
system goals.
Reinjection systems may need frequent cleaning and
maintenance because they are prone to fouling by
solids, iron, biological growth, and calcium carbonate.
Reuse
Treated water is reused at an active industrial facility
or is used for irrigation or potable water supply.
Potential Advantages
• Reuse of treated water reduces or eliminates the
need for a facility or organization to use water
from other sources, thereby conserving water as a
natural resource.
Reuse of treated water may also eliminate costs
associated with discharging the water and the
costs of using water from other sources.
Potential Disadvantages
For use in irrigation or drinking water supply,
specific Federal and State regulations may be
applicable, or relevant and appropriate.
Additional testing relative to other discharge
options may be needed.
• Additional system conservatism (e.g., an
additional GAC vessel) may be desired compared
to treatment systems without direct discharge to
human receptors.
Reusing water in industrial processes may involve
additional treatment relative to discharging the
water elsewhere. Reused water should be treated
20
-------
to meet the facility standards and any downstream
discharge standards. Treating to facility standards may
be more costly than discharging to another location and
using public water at the facility.
Facilities may only operate on a part-time basis,
and the P&T system may need to operate
continuously. If continuous extraction and batch
treatment during facility hours is not available,
reuse may not be feasible.
• A backup discharge point should be available in
case the needs of the facility change.
Contaminants that are undetected using current
analytical techniques or contaminants that are
present as tentatively identified compounds (TICs)
may not be removed by treatment, causing a
potential risk to end users of the water.
Cost Information
Cost information is facility and site dependent.
G. CONTROLS/REDUNDANCY/FAILSAFES
Exhibit 7
An effective P&T system should have the following
characteristics:
• down time of <5%
no releases of untreated water to the environment
• appropriate automation of system controls so that
minimum operator time is needed without
excessive capital expense
However, unlike a municipal waste water treatment
system, ground water moves relatively slowly, and a
P&T system typically can be stopped temporarily
without adversely affecting human health or the
environment. The amount of time that a P&T system
can be non-operational due to maintenance or other
issues depends on the natural ground water flow and
other site-specific factors that can be addressed through
hydrogeological analysis that potentially includes
ground water flow modeling.
Exhibit 7 provides typical labor needs for different
types of treatment systems. The amount of labor is
indicative of the degree of failsafes, alarms, and
automation involved. For some systems, sufficient
maintenance and supervision is needed full-time.
However, 24-hour attention is rarely, if ever, necessary.
i
General Guidelines for Labor Typically Needed
for Various Types of Treatment Plants
Treatment Plant
air stripping
vapor phase GAC
for offgas
treatment
• GAC
filtration
• UV/Oxidation
• GAC
metals removal
filtration
• metals removal
filtration
and one or more of
the following
air stripping
• GAC
biotreatment
UV/oxidation
Estimated Labor
weekly checks by local
operator (8-12
hours/week)
quarterly checks by
engineer
weekly checks by local
operator (8-12
hours/week)
quarterly checks by
engineer
weekly or semi-weekly
checks by local operator
(8-16 hours/week)
quarterly checks by
engineer
one full-time operator
(1* 40 hours/week)
one full-time operator
with potential for part
time assistance
(40 - 60 hours/week)
Redundant and Spare Equipment
Process pumps, blowers, and filters that are
continually in use and key to system operation should
have redundant units piped in parallel to allow
servicing without downtime. These devices are
typically a small percentage of the treatment system's
cost, and a redundant unit is worthwhile to avoid
system downtime. It is not typically worthwhile to
have redundant or spare treatment units that do not
have active mechanical/electrical parts. For example,
redundant units are not needed for tray air strippers
(excluding the blower), equalization or reaction tanks,
and clarifiers. Also, redundancy is not appropriate for
units that are used infrequently (e.g., most sludge
handling systems) and units that are a high percentage
21
-------
of a system's cost (e.g., UV/Oxidation system). It is
also not typically worthwhile to have a redundant
power supply for a P&T system, given that power
outages are typically limited to less than 24 hours. It is
worthwhile to install appropriate surge protection and,
depending on the system location, a lightning arrester
to limit damage from line surges and lightning strikes.
Spares for key mechanical and electrical equipment
that have a specific service life or operate in harsh
environments (e.g., UV/Oxidation system lamps,
transducers, double diaphragm pump diaphragms,
submersible pumps, small GAC units) should be kept at
the treatment system and replaced by the system
operator as necessary.
Failsafes/Alarms
P&T systems should have alarms for any parameter
value that is outside of the typical operating range.
These alarms should interface with the system controls
to prevent a release of untreated water or any other
problem from occurring by shutting down system
components or the entire system as appropriate. The
following are typical parameters for which failsafe
alarms are installed:
• high tank levels
• high differential pressure across a filter
• high or low air stripper blower air pressure
well vault or building sump water accumulation
• low water flow in the treatment system
• other system specific items
Providing these failsafes and alanns should allow
operator attention to be reduced and protectiveness to
be maintained cost-effectively.
Automation/Remote Monitoring
Except for the simplest systems (e.g., liquid GAC,
oil/water separator, filtration), an unmanned P&T
system should have an autodialer to inform the system
operator of an alann condition. In many cases, remote
monitoring of the system is worthwhile to reduce
operator labor. There is a wide array of remote
capabilities available depending on the price, such as
the ability to view the water levels in wells or tanks,
check on alarm status, start and stop motors, and open
and close valves remotely. The degree of automation
appropriate for a specific site should be based on the
amount of operator labor time and expense (including
travel) that is saved by the automation capabilities
(with equivalent protectiveness) for the expected life of
the system.
H. OTHER DESIGN CONSIDERATIONS
Other factors that may affect the construction and
operation of a P&T system are described below.
However, these factors generally do not affect the
selection or sizing of treatment components.
Climate
Sites in colder climates may be exposed to freezing
temperatures. Most treatment components are housed
in a climate-controlled building, but some components,
such as tanks, well heads, and some piping may be left
exposed. Proper insulation and heating should be
considered to reduce the potential for upsets and
additional maintenance in the future.
Community Issues
The surrounding community may also affect long-term
operation of the system. Appropriate consideration of
noise and potential emissions or odors will reduce the
impact of the system on the surrounding community
and facilitate community relations. Appropriate
security measures will reduce the potential for
vandalism, which can lead to system upsets and
additional costs during O&M. However, excessive
security can add substantially to annual O&M costs.
Remoteness
For sites that are fairly remote, designers might
consider additional automation or remote telemetry
capabilities to minimize the need for operator visits
and the associated expense. Designers might also
include additional system redundancies that allow the
system to keep running until the operator can visit the
site.
I. POST-CONSTRUCTION
Although design and construction of a P&T system
marks a significant milestone in remedy
implementation, it typically does not represent the end
of the cleanup process. The post-construction period
generally involves operation, maintenance, and
monitoring that may last decades or indefinitely.
Therefore, the success of a P&T remedy is not based
only on the design and construction of the system.
Because operating a remedy inherently changes the
site conditions (i.e., changes ground water flow
patterns and removes mass), O&M should be a
dynamic process and should include periodic
optimization evaluations.
-------
A discussion of O&M and optimization of the O&M is
beyond the scope of this document, but the reader is
referred to other fact sheets in this series, which are
available at the www.cluin.org/optimization website.
Elements for Effective Management of Operating Pump
and Treat Svstems, December 2002
(EPA 542-R-02-009, OSWER9355.4-27FS-A)
Effective Contracting Approaches for Operating Pump
and Treat Systems, April 2005 (EPA 542-R-05-009,
OSWER9283.1-21FS)
O&M Report Template for Ground Water Remedies
(with Emphasis on Pump and Treat Systems), April
2005 (EPA 542-R-05-010, OSWER 9283.1-22FS)
For further information on topics covered in this
document and for access to a broader spectrum of EPA
documents, the reader is directed to www.cluin.org.
J. CITED REFERENCES
Dobbs, R.A. and J.M. Cohen, Carbon Adsorption
Isotherms for Toxic Organics, EPA 600/8-80-023,
April 1980.
Newell, C.J. and R.R. Ross, Estimating Potential for
Occurrence ofDNAPL at Siiperfund Sites. EPA 9355.4-
07FS, January 1992.
APPENDICES
ADDITIONAL ILLUSTRATIVE EXAMPLES
Two examples are provided to illustrate design
considerations for a cost-effective P&T system.
Hypothetical sites and conditions are used.
Approximate costs are provided for the purposes of the
example. Although the costs are reasonable based on
2003 dollars, they are estimates for illustrative
purposes only, and do not reflect rigorous pricing
through vendors. The site conditions are typical of
what may be found at a particular site, but the
description is simplified so that the design aspects
discussed in this document can be effectively and
clearly illustrated.
23
-------
Appendix A (Page 1 of 3)
Sample: Design Considerations for TCE Treatment at a Hypothetical Site
with High Iron Concentrations
Remedy Goal and Strategy:
hydraulic containment at site property boundary and eventual cleanup of site ground water to MCLs
control source area with pumping to allow other wells to restore remainder of the property to MCLs
Remedial Investigation Results: The TCE source area, a former lagoon, is 200 feet upgradient from the
property line. The maximum TCE concentration in a source area well is 50,000 ug/L. Monitoring well data
from locations outside of the source area have TCE concentrations of 5,000 ug/L. TCE breakdown products are
present at levels <500 ug/L and iron concentrations average 15 mg/L.
Extraction System Parameters: Ground water modeling indicates that one well pumping with an extraction
rate of 10 gpm is sufficient to control the source area (i.e., monitoring data shows TCE concentrations of 50,000
ug/L). Two downgradient wells, each pumping 25 gpm are likely to be needed to provide containment at the
property hue and mass removal (i.e., monitoring data shows TCE concentrations of 5,000 ug/L).
Design Considerations:
The TCE and iron concentrations are extremely high for a relatively transmissive aquifer. Extraction wells
should be installed and pumping tests conducted to better estimate influent concentrations prior to selection
of the treatment technology. It is very likely that influent iron and TCE concentrations will decrease under
sustained pumping.
Air stripping with off-gas treatment GAC, and UV/oxidation are the competing technologies for TCE. At
this flow rate a tray aerator air stripper could be used, which is easier to clean than a packed tower.
UV/oxidation will very likely need pretreatment if the influent iron concentration is above 1 mg/L. GAC
will also have a tendency to foul. The tray aerator may perform well with a frequent, but reasonable
cleaning schedule.
Both reinjection and surface water are potential discharge options. Discharge to surface water may be
contingent on an iron concentration below a specified standard (e.g., 600 ug/L). Fouling would be a
concern for reinjection through reinjection wells, but can be easily addressed with an infiltration gallery.
During the pmnping tests half of the produced water should be treated with a leased tray aerator and the
other half should be treated with GAC to determine how much impact iron fouling will have on each
technology if a metals precipitation system is not installed.
Pumping Test Results:
Iron concentrations during 24-hour pump tests drop to below 5 mg/L. Further decreases are expected with
long-term sustained pumping, but iron concentrations at 5 mg/L may need pre-treatment for any of the
technologies considered for VOC treatment.
TCE concentrations in source area well decreases to 20,000 ug/L during pumping. TCE concentrations in
the downgradient locations drop to 2,500 ug/L. Further decreases will likely occur during long-term
sustained pumping. 1,2-DCE (TCE breakdown product) concentrations decreased during pumping to less
than 100 ug/L.
24
-------
Appendix A (Page 2 of 3)
Sample: Design Considerations for TCE Treatment at a Hypothetical Site
with High Iron Concentrations
(continued)
Design Parameters:
Design flow rate: 60 gpm
Hydraulic capacity: 120 gpm (assuming a factor of safety of 2.0)
Design TCE influent concentration (weighted average from pump test data): -5,500 ug/L
Maximum TCE influent concentration: 11.000 ug/L (assuming a factor of safety of 2.0)
Design mass removal rate: 3.96 pounds per day (see Exhibit 2)
TCE discharge criteria: 5 ug/L (for discharge to surface water or reinjection)
Iron discharge criteria: 600 ug/L for discharge to surface water
Iron discharge critera: "none" for reinjection
Technology Evaluation:
Capital costs for the potential VOC treatment systems (air stripping with off-gas treatment, and UV/oxidation)
are all comparable and are low relative to O&M over 10 or more years. The costs of each remedy are similar
with the exception of utilities and consumables.
Air Stripping Costs: A tray aerator with a 10 HP blower will remove the TCE to MCLs. The blower has a
motor efficiency of-75%, and electricity cost per kilowatt-hour (KWh) is $0.10. Carbon vendor for off-gas
treatment reports carbon adsorbs 8.9% TCE by weight at a cost of $2.50 per pound.
0.75KW 1 24h 365 days SO. 10 $8,760
Power Cost = 10HP x ——— x . x -— x x =
HP 7s< % efficiency day year KWh year
3.96 Ibs TCE 100 Ibs carbon 365daVs $2.50 $40,600
x x x =
day 8.9 Ibs TCE year Ib carbon year
$8,760 $40,600 $49,360
Total Comparison Cost = + =
year year year
GAC Costs: The parameters for calculating GAC usage based on chemical loading are provided in Exhibit 5.
Iron fouling, however, may result in more frequent replacement of GAC and therefore a higher usage rate. Costs
for GAC to treat dissolved contamination are assumed at $2 per pound. Ratio of pounds of TCE to pounds of
GAC needed (R)
_ 1kg 28gTCE( L \°62 (5.5mg\°62 _ 8.06kg TCE _ 8.06lbs TCE
R ~ lOOOg X kg GAC ( mgl * ( L ) ~ 100kg GAC ~ 100Ibs GAC
GAC Usage Rate (GUR)
3.96 Ibs TCE 365days 100 Ibs GAC 17,933 Ibs GAC
GUR = x — x =
day year 8.06 Ibs TCE year
Off - gas Cost =
25
-------
Appendix A (Page 3 of 3)
Sample: Design Considerations for TCE Treatment at a Hypothetical Site
with High Iron Concentrations
(continued)
UV/Oxidation Costs: A system with (2) 30 KW lamps could likely treat the influent to MCLs. In addition to
the power, lamp replacements and seals will cost about $15,000/yr and hydrogen peroxide about $25,000/yr.
24h 365days $0.10 $52,560
Power Cost = 60 KW
Total Comparison Cost =
day year KWli year
$52,560 $15,000 $25,000 $92,560
+ : + =
year year year year
Polymer resin adsorption systems generally would become competitive if the flow rate and/or VOC
concentrations were higher so that consumable costs exceed $200,000/yr, but it would be worth verifying this
status with vendors.
Recommend ation s:
GAC offers the cheapest annual cost for TCE treatment at this site. Air stripping is a 37% increase in cost
($14,000/yr more), and UV/oxidation is a 158% increase in cost ($57,000/yr more).
Iron removal will likely be needed for any of these three remedies with an estimated iron influent
concentration of 5 mg/L, for an additional cost of at least $ 150,000/yr in labor and materials. However, if
the iron influent decreases to 3 mg/L over time, air stripping may become effective with frequent cleaning
but without iron removal, whereas GAC and UV/oxidation would likely still need iron removal.
Over time, the VOC influent concentrations will decrease. This decrease will lower consumable and utility
costs for the air stripping and GAC systems by a much greater amount than it will lower UV/Oxidation
system consumable and utility costs.
Given a relatively strong possibility that iron concentrations will drop below 3 mg/L within the first year of
operation, design should include air stripping with off-gas treatment and leasing of an iron removal system
(metals precipitation with clarification and filtering) for several months. Iron pretreatment will occur over
this time period allowing the air stripper to function. If iron levels decrease sufficiently, iron removal can
be discontinued.
If iron levels do not decrease sufficiently, the iron removal system can be purchased. Pilot studies can be
conducted to determine if metals precipitation is needed or filtering is sufficient.
Wliile all three of the alternatives can be operated with periodic site visits, the UV/Oxidation system is
more difficult to operate than the alternatives, will perpetuate metals treatment indefinitely, and is more
prone to downtime (especially if there are upsets with the iron removal system).
If additional extraction wells will improve progress to reaching site goals, they can be added as die
concentrations decrease at the initial wells so that consumable costs do not increase substantially. The
stripper can easily be designed to accommodate additional flow.
Discharge should be to the reinfiltration gallery or basin so that iron removal can be discontinued even if
influent iron concentrations are above 600 ug/L.
Consider excavation or in-situ treatment of the source area.
26
-------
Appendix B (Page 1 of 2)
Sample: Design Considerations for Treatment of Ketones and Constituents at a Hypothetical
Site with High Iron Concentrations and the Potential for Discharge to the POTW
Remedy Goal: The goal at this landfill site is to provide hydraulic containment of the contaminated ground
water at the property line.
Remedial Investigation Results:
Monitoring wells in likely extraction areas have the following contaminants and average concentrations.
Contaminant/Constituent
Ammonia
Acetone
MEK
MffiK
Benzene
Toluene
1,2-DCA
Iron
Concentration
20mg/L
5,000 ug/L
1,000 ug/L
700 ug/L
200 ug/L
500 ug/L
300 ug/L
40 mg/L
Extraction System Parameters: Simplistic analytical modeling based on historical potentiometric surface
maps and a hydraulic conductivity estimate from one pumping test suggest a total pumping rate of 45 gpm from
three wells (15 gpm each) is sufficient for capture. A factor of safety of between 2.0 and 3.0 would be
recommended to determine the hydraulic capacity.
Design Considerations:
Discharge to surface water would likely need treatment of VOCs to Federal MCLs and treatment of
ammonia to 1 mg/L; there is no iron concentration discharge limit. A treatment system consisting of metals
precipitation, a fixed film biological treatment unit, filtration and liquid GAC at a minimum would be
appropriate to treat this water to meet MCLs. Capital costs of this system including the building would
exceed $1,000.000.
Alternatively the POTW would agree to a long-term contract to take the untreated water for $0.02/gallon.
The cost of this approach is heavily dependent on the design flow rate. More accurate estimates of the flow
rate should be obtained. Pump tests should be conducted near each well. Investment in a numerical ground
water model is worthwhile. Capital costs would be on the order of $250,000.
Pump Test Results: Interpretation of the pump test results and the development and use of a numerical model
suggest that an extraction rate of 30 gpm from three wells (10 gpm each) is sufficient for capture. A factor of
safety of approximately 1.5 can be used to determine the hydraulic capacity. Concentrations decreased during
pumping, but not significantly.
Technology Evaluation:
POTW Discharge: Water would be pumped and discharged directly to the POTW. At 30 gpm and a discharge
cost of $0.02 per gallon, the cost for discharge to the POTW is approximately $315,000/yr. Additional costs,
on the order of $50,000 per year would likely be needed for maintenance, discharge monitoring, and reduced
project management relative to the treatment system.
27
-------
Appendix B (Page 2 of 2)
Sample: Design Considerations for Treatment of Ketones and Constituents at a Hypothetical
Site with High Iron Concentrations and the Potential for Discharge to the POTW
(continued)
Pump and Treat System: Water would be pumped to a treatment system and then discharged to surface water.
Such a system would have the following estimated costs for O&M.
Cost Category
Operator labor (two full-time employees)
Process monitoring and analytical
Chemicals and materials (including GAC)
Non-hazardous sludge disposal
Power for system and building heating
Project management and reporting
Total Estimated Annual Cost
Estimated Annual Cost
$200,000/yr
$30.000/yr
$30,000/yr
$20.000/yr
$60,000/yr
$50.000/yr
$390,000/yr
Recommendations:
• The O&M for the P&T and POTW options are within 10% (the POTW option costing less). The capital
cost for the POTW option is $750,000 less than the P&T option.
The POTW cost estimate is flow dependent. If 45 gpm (rather than 30 gpm) is needed, an additional
$160,000 per year would be incurred for discharging. In this "worst case scenario", the POTW option is
still more cost-effective for the first five years due to the lower capital costs.
The POTW discharge provides much greater security against problems. Contracts to discharge at the same
rate for 10 or more years, without a minimum payment, are common. This would protect against inflating
operating costs. No startup costs, discharge limit excursions, system downtime and other problems inherent
with operating a complex treatment system would occur with POTW discharge.
Because the landfill is a continuing source of the contaminants present, it is unlikely that the P&T system
operation costs would decrease due to discontinued treatment system components. If conditions unproved
so that a simple treatment system (e.g., GAC only) could be installed, the POTW discharge could be
discontinued.
If, over time, conditions allow select extraction wells to discontinue pumping, the overall flow rate may
decrease and the POTW option would provide greater savings. The conditions that would allow
discontinuing pumping from a well (at least for a trial period) should be clearly stated ahead of time.
Lesson Learned: Had the pump tests and modeling not been conducted, the estimated design flow rate would
have made the P&T option more attractive. It still may not have clearly been the best option, but many parties
would have chosen it. The investment in the model and pump tests helped avoid excess capital expenditures of
$750,000 and a myriad of problems that may have been encountered while running a complex water treatment
system.
28
-------
NOTICE:
The U.S. Environmental Protection Agency funded the preparation of this document by GeoTrans, Inc. under General Service
Administration Contract GS06T02BND0723 to S&K Technologies, Inc., Bremerton, Washington; EPA Contract No. 68-C-
02-092 to Dynamac Corporation, Ada, Oklahoma; and EPA Contract No. 68-W90-0065 to ICF Consulting, Inc., Fairfax,
Virginia. Mention of trade names or commercial products does not constitute endorsement or recommendation for use.
This document may be downloaded from EPA's Clean Up Information (CLUIN) System at http://www.clum.org. Hard copy
versions are available free of charge from the National Service Center for Environmental Publications (NSCEP) at the
folio wing address:
U.S. EPA NSCEP
P.O. Box 42419
Cincinnati, OH 45242-2419
Phone: (800) 490-9198 or (513) 489-8190
Fax: (513) 489-8695
29
-------
&EPA
United States
Environmental Protection
Agency
Office of Solid Waste and
Emergency Response
OSWER9283.1-20FS
EPA 542-R-05-008
April 2005
www.cluin.org
www.epa.gov/superfund
------- |