United States Environmental
                     Protection Agency
United States Army
Corps of Engineers
July 1991
v>EPA     Engineering   Issue
                     DATA GAPS IN  REMEDIAL DESIGN
                     John E. Moylan
 The Regional Superfund Engineering Forum is a
 group of EPA professionals, representing EPA's
 Regional Superfund Offices, committed to the
 identification and resolution of engineering issues
 impacting the remediation of Superfund sites.
 The Forum is supported by and advises the
 Superfund Technical Support Project.

 This paper was prepared by the U.S. Army Corps
 of Engineers (US ACE), Kansas City District,
 Geotechnical Branch in cooperation with the EPA
 Engineering Forum. Engineering Forum review
 and comments were coordinated by Ken Erickson
 (Region IX), Forum Co-chair. For further infor-
 mation contact John Moylan, at FTS 867-3455 or
 Ken Erickson, FTS 484-2324.

 Introduction

 As the number of Superfund sites in the phases
 of Remedial Design (RD) and Remedial Action
 (RA) has grown, we have become increasingly
 aware of the adverse effects of inadequate or
 insufficient design  data. This paper is intended
 for use by RPMs as a checklist or reminder to
 consider certain aspects of RD/RA data needs as
 early as possible in the life of a Superfund site.
 Most of the items have been gleaned from the
 collective memory  of EPA and  USAGE personnel
 who have learned  some of the lessons the hard
 way.

 The items addressed include:

 • types of data where quality or quantity is most
   often found to be-inadequate or have the
   greatest effect on successful RD and RA;

 • examples of why these data are needed;
  • data needs for particular remediation features;

  Many design problems causing schedule slippage
  or cost overruns in either RD or RA can be
  attributed to site characterization data that are not
  sufficient for purposes of design. (No attempt is
  made in this paper to address potential
  inadequacies in interpretation of the site
  characterization data.) These data gaps affect
  not only highly technological treatment processes
  but also the more mundane aspects of
  remediation such as caps and liners.

  Technical work during the RI/FS phase has
  typically been considered the province of
  scientists whereas engineers are given the
  functional lead during RD/RA. The engineers
  have not always been requested to provide
  significant input during RI/FS. Similarly, the
  scientists have often not been made available
  during RD/RA. Consequently, some RODs or
  settlement agreements have dictated remedies
  that are only marginally appropriate or not
  effective at  all, much more costly than
  anticipated, or virtually impossible  to implement.

  The RPM should consider that a broad-spectrum
  technical assistance team is a key aspect to
  successful  management of the site from
  beginning to end.  Under ideal circumstances,
  this team should consist of members from
  organizations other than the contractors used by
  Superfund.  The starting point is the Regional
  Engineering Forum representative and other
  Regional technical specialists, such as the
  Technical Support Sections formed by some
  Regions or experienced RPMs,  who can
  participate  in the peer review process. Other
  resources include the Technical Support Centers
  (under the  auspices of the OSWER Technology
                     U.S. Army Corps of Engineers
                     Kansas City District
                     Geotechnical Branch
                  Technology Innovation Office: :  :
                  Ottfct of Solid Wait* and Emwgency-;
                  Rwpon»», USEPA, Wwrtilngton, D.C,

                  Walter W.Kovallck, Jr., Ph.D.
                  ::Director ;•:;: •'.'.•'•'-:.' '

-------
Innovation Office Technical Support Project), U.S.
Army Corps of Engineers,  U.S. Bureau of
Reclamation, U.S. Bureau  of Mines, state
agencies, and so forth.

It should be noted that typically design engineers
are accustomed to working from a clearly defined
problem. Unfortunately, most subsurface and
ground water contamination sites lack the degree
of site characterization needed to clearly define
the design at the beginning of the RD phase.
Therefore, it is important that the scientific
disciplines be available throughout the entire
RD/RA period to further define site conditions and
provide  interpretations for the engineer.  This will
help assure an expeditious and satisfactory
implementation  of the design.

A few examples of problems associated with
incomplete site characterization data follow:

1. Soils properties and handling characteristics
are often poorly evaluated or even ignored when
considering various remediation technologies
which require materials excavation and
processing.

2. Volatile emissions during  excavation and
handling of contaminated soils are often  not
anticipated.

3. Lack of information on temporal and spatial
variations in contaminant loading in ground water
remediation decisions can lead to inefficient
designs.

4. No pre-ROD consideration of availability  of
utilities resulting in underestimation of costs  or
schedule.

5. Poor understanding of the permeability of
slurry wall key layer leading to unacceptable
leakage.

6. Ground water treatment processes which
focus on the contaminants of interest but ignore
total ground water chemistry, especially the
anions and cations present, will impact the
effectiveness of the treatment process.

7.  Solvent extraction of explosives from soil is
feasible, however, the unrecognized instability of
the residue  can be disastrous.

8.  Cap designs which utilize the cost-
effectiveness of geosynthetics but require slopes
on which geosynthetics are not. stable or caps
which require the use of low permeability clays
but don't evaluate the availability of suitable clay
borrow material can be impractical to construct or
very costly.

9.  Excavation limits may be defined based on
chemical analysis of bore hole samples and not
account for contaminant migration along
secondary pathways, e.g:, root holes.  Trenching
and careful inspection and sampling may alleviate
this problem.

10. The presence of debris or boulders may not
be recognized or fully appreciated during the
analysis of remedial alternatives if bore holes
alone are used to investigate the limits of waste.

The term geochemical is used rather than the
more narrow chemical term in order to emphasize
the importance of understanding the chemical
processes operating in the geological
environment in order to implement effective
remediations.  The importance of quality
analytical chemistry is already well understood
and appreciated, however, our understanding of
ongoing chemical processes needs improvement.
The following paragraphs identify some
commonly overlooked data requirements and
include examples of problems resulting from the
data gaps.

Site Data needs are often overlooked  in the
pre-ROD/consent decree phase and even well
into design.  Unforeseen cost increases, time
delays, and  contract modifications can and do
result. Some common data needs include:

1.  Topographic Surveys - The need should be
readily apparent, however, this aspect is often
overlooked.  In some instances, available general
topographic mapping is used without verification.
Consequently during RA, excavation or fill
overruns or  underruns or impossible site drainage
are discovered which require contract
modifications.  Property boundary surveys and
adequate horizontal and vertical controls are also
included in this category.

2.  Utility Availability - Water, gas, power, and
sewer services required for remedy
implementation must be identified. In addition,
leaking industrial sewer lines might be
contamination sources and previously unidentified
utility lines crossing a remediation site can cause

-------
contract shutdown pending their relocation or
protection.

3. Borrow Availability - In most cases, site
remediation which includes earthwork, e.g., liners,
caps, slurry walls, etc., require the utilization of
earth materials to effect the remediation.  When
those materials are not available from required
on-site excavation, they must be  obtained
elsewhere and are referred to as borrow. In
some areas suitable borrow is scarce. The costs
of trucking suitable material from a distant borrow
pit will add  significant cost and transportation
problems if not recognized. As an example, a
5-acre cap with an average thickness of  3 feet of
soil,  requires  almost 25,000 cubic yards  or
approximately 1,400 truckloads of suitable earth
borrow material.  The availability  of admixtures
such as fly ash, cement, and lime for stabilization
should also be considered in this category.

4. Transportation Network - The proximity of
suitable roadways and/or rail lines is important to
remedies requiring the transportation of heavy
equipment and earth materials into the site or
contaminated or treated wastes from the site.
Local opposition to frequent heavy truck  traffic
and  damage to streets and roads, especially
through residential areas, must be anticipated.

Geochemlcal Data collection can often be
improved to more confidently  select effective
remedies and better effect quality RD and RA.
Some examples include:

1. Multiple Sampling Rounds - In too many
cases, remediation decisions are made  which
are based on single or poorly timed, multiple
ground water sampling rounds.  Time allowed for
RD often doesn't  provide for seasonal sampling.
As a result, chemical loading  may exceed
treatment plant capability, the plant may  be over-
designed, or the operating plan is not optimized
to accommodate variations in loading.

2. Anion/Cation Analysis - These analyses are
inexpensive, yet if they are overlooked in ROD
preparation, the designed treatment train may be
either more expensive than anticipated or
ineffective if not detected during  RD. Eh, pH, and
TOG are other chemical parameters that must be
considered for effective RD.

In addition, caution should be exercised  when
unusually high metals analyses are obtained from
turbid ground water samples. If these analyses
are used for treatment plant design, the design
and cost estimates can be unrealistic.  Total
chemistry also impacts well design.  Anion and
cation precipitates will prematurely clog injection
wells if not recognized and treated.

Geotechnlcal Data must be gathered  for many
types of remedies, both for purposes directly
related to the remedial process and for design
auxiliary to the actual remedial process, such as
building foundation design or excavations.

1.  Soil Moisture Content - The natural moisture
content of site soils, especially fine-grained soils,
is valuable information both in the pre-ROD and
RD phases. As  examples, the moisture content
of contaminated  soil that  will receive thermal
treatment affects fuel consumption  and the
moisture content of a fine-grained foundation soil
can be an indicator of the soils strength and
consolidation characteristics.

2.   Atterbero Limits - These parameters define
the plasticity of fine-grained soils, give the
geotechnical designer an early indication  of the
strength of that soil, especially when evaluated
with moisture content, and can be an indicator  of
contaminated soil handling and processing
characteristics.  The test  is  relatively inexpensive
but the results can be very useful.

3.   Soil Strength Parameters - Generally not
needed prior to the RD phase. Some design
features requiring soil strength testing  include
structure or building foundations, significant
excavations, dredging, and  slurry wall  trenches.
Blow counts from Standard Penetration Tests can
be  used as an early  indication of soil strength.

4.   Gradations -  Some representative gradations
or particle size distribution analyses done in the
RI/FS phase can be very helpful in estimating
approximate permeability and for designing
efficient monitoring wells. Gradations  are
required for the  design of such things  as
collection drains and withdrawal wells and in
evaluating soils  handling and processing
characteristics.

5.  Excavatabilitv - While there is no one  test or
set of tests to define this design  parameter,
evaluations and judgments should  be  made in  the
pre-ROD phase concerning excavatability when
excavations ot any kind are required in the

-------
remedy.  Excavatability includes such factors as
whether the material can be machine excavated,
the necessity for blasting, the  existence of
boulders or cobbles, the need for dewatering, etc.
None of the excavatability factors should be
minimized since all of them can greatly impact
the final cost of remediation.

6. Landfill  Subsidence - Remediations often
include capping an existing landfill and perhaps
incorporating a gas  collection  and venting system.
Many such landfills  are still subsiding with
attendant surface disruption capable of adversely
impacting the effectiveness of the cap and vent
system.  Carefully surveyed settlement data
collected throughout the RI/FS phase are
invaluable for remedy selection and to support
design.  Settlement  data collection should
continue through  RD and RA  and into the
operations  and maintenance phase if
displacements are continuing  and significant.

Hydrogeologlcal Data are routinely collected
both during the RI/FS  and RD phases.  However,
several  aspects will be discussed which are
sometimes slighted  but can be very important to
selection of an effective remedy and to proper
design and implementation.

1;  Multiple Water Levels - In  order to understand
the hydrogeological character of the site in
sufficient detail to select an effective remedy, it is
important that enough water levels be obtained to
define both the vertical and horizontal flow
directions seasonally and as they respond to both
natural and manmade recharge and discharge.
For example at a ground water contamination site
in the Plains States, the regional flow is severely
distorted locally by irrigation pumping during
several months of the year.

2.  Detailed Stratigraphy - In too many cases,
stratigraphic detail has not been well developed
due to poor sample recovery  often coupled  with
too infrequent sampling intervals, lack of
geophysical logs, improper sampler selection,
field geologists poorly trained in logging methods,
or combinations of the above. Even relatively
minor variations in lithology have a strong
influence on contaminant migration and plume
development. This is an important factor during
pre-ROD,  RD, RA,  and even  into the operation
and maintenance phase of both ground water
and vadose zone remediation.
 3.  Secondary Porosity Features - Joints,
 defoliation planes, bedding planes, root holes,
 etc., often strongly influence the overall gross
 permeability of bedrock materials and
 fine-grained soils, especially clays.  In too many
 cases these features are not targeted during site
 exploration or if they are, the vertical features are
 difficult to intercept and analyze. Careful
 consideration of these features is warranted
 during the RI/FS phase and remedy selection for
 problems such as contaminated bedrock  aquifers,
 multiple stacked aquifers, and  slurry walls keyed
 into an impermeable layer.  For sites such as
 these, additional characterization will also be
 needed during RD.

 The various types of site characterization data
 discussed in this paper are not needed or at least
 not to the same degree for all features of site
 remediation. The following remediation features
 were considered:

 1.  Withdrawal & injection wells
 2.  Internal drains
 3.  Slurry walls
 4.  Slurry wall key layer
 5.  Caps
 6.  Chemical Stabilization
 7.  Ground water treatment
 8.  Landfills
. 9.  Thermal treatment
 10. Soil washing
 11. Excavations
 12. Dredging
 13. Vapor extraction

 Table 1 presents a summary of site
 characterization data determined to be useful or
 needed for remediation. The  table also suggest
 in  which phase or phases of the remediation
 process it is advantageous to acquire the data.

-------
                                 Table 1. Site Characterization Data needed for Remediation
                                                 REMEDIATION FEATURE
          DATA
       Topographic Surveys
                                   1,3
              1,3
             1,3
             1,3
                    1,3
                     1,3
      1,3,4
(/) Q
       Utility Availability
       Borrow Availability
       Transportation Network

-------