United States        Office of Research and      EPA/540/R-92/005
             Environmental Protection    Development         January 1992
             Agency          Washington, DC 20460
&EPA      Presentations EPA-State
            Soil Standards Conference

-------
                                  EPA/540/R-92/005
                                  January 1992
               PRESENTATIONS
EPA-STATE SOIL STANDARDS CONFERENCE
            Held at: Hyatt Regency Hotel
             Washington Rooms A and B
                Crystal City  Virginia
                on January 29  1991
                          U.S. Environrnei.lal ;
                          Region 5, Library  '""'
                          77 West Jacks-; r:  :
                          Chicago, IL  6C6L-.
This document was prepared from a transcript taken at
the conference by a court recorder.  It has been edited for
technical consistency only.
                                     vyo Printed on Recycled Paper

-------
                                     NOTICE

Development of this document was funded, wholly or in part, by the United States
Environmental Protection Agency.  This document has not undergone a formal USEPA
peer review.  Since this document is essentially a collection of papers presenting ideas of
individual authors, it has not been reviewed subject to USEPA technical and policy
review, and does not meet USEPA standards for USEPA document publication. The
views expressed by individual authors are their own and do not necessarily reflect the
views, policies, or ideas of USEPA.  Any mention of trade names, products, or services
does not convey, and should not be interpreted as conveying, official USEPA approval,
endorsement  or recommendation.

This document is not intended to and does not constitute any rulemaking, policy or
guidance by the Agency.  It is not intended to and cannot be relied upon to create a
substantive or procedural right enforceable by any party.  Neither the United States
Government nor any of its employees, contractors, subcontractors or their employees
makes any warranty, expressed or implied, or assumes any legal liability or responsibility
for any third party's use of or the results of such use of any information or procedure
disclosed in this report, or represents that its  use by such third party would not infringe
on privately owned rights.
                             ACKNOWLEDGEMENT

The U.S. Environmental Protection Agency (EPA) wishes to thank all of those who
participated in the development of the Proceedings Document, and attended the first
"EPA/State Conference on Determining Soil  Cleanup Goals at Superfund Sites", held on
January 29,  1991 at the Hyatt Regency Hotel in Crystal City, Virginia.  The information
received during and after the conference was extremely positive, and it is EPA's plan to
have another conference with an intent to incorporate international participation.

Several individuals played an important role in making the conference the success it was.
In particular, Alison Barry for organizing the conference and Jennifer Sutler and Randall
Breeden of USEPA, should be recognized. We thank these individuals, the authors,
speakers, and conference participants for their participation.
                                        11

-------
                                    PREFACE

                  EPA/STATE SOIL STANDARD CONFERENCE
                  JANUARY 29  1991 HYATT REGENCY HOTEL
                            CRYSTAL CITY VIRGINIA
The first U. S. Environmental Protection Agency (EPA)-sponsored national conference
on establishing soil cleanup criteria that would be protective of ground-water quality was
held on January 29, 1991.  Fourteen presentations were made from three panel sessions
which addressed the use of site-specific fate and transport modeling, the use of numerical
quality criteria, and the design and implementation of a regulatory program involving
both numerical criteria and site-specific assessment. The conference was attended by
over 120 representatives of 31 state regulatory agencies, EPA program offices, and all
EPA regional offices, as well as  the Department of Energy and the Nuclear Regulatory
Commission.

Included in this publication are questions and answers from the panel discussions, as well
as text from the presentations. The text was prepared from a transcript of the entire
conference proceedings by a court recorder.

This national conference was warranted and timely due to the complexity of issues
related to  this subject area and the growing number of hazardous waste sites  where
subsurface soil contamination may act  as a source for ground-water contamination.  On
June 19, 1991, EPA Administrator William K.  Reilly charged the Office of Solid Waste
and Emergency Response (OSWER) with the task of evaluating options for accelerating
the  rate of cleanups at Superfund sites, and the assumptions used when evaluating and
managing  the risks at Superfund sites,  in thirty days. The results of the 30-day study
included a recommendation to standardize the remedial planning and remedy selection
process, including developing standards or guidelines for contaminated soils.  In light of
this, establishing cleanup criteria for soils will move to the forefront of Superfund
attention.

The one-day conference was followed by a workgroup  meeting on  January 30, 1991
which was intended to initiate the  development of guidance for determining soil cleanup
levels at Superfund sites.  Workgroup members included the previous day's speakers, as
well as representatives from various EPA offices and Office of Research and
Development laboratories.  The workgroup produced a decision tree for the
development of soil cleanup goals  which protect ground-water quality, and also identified
a number  of issues associated with points on the decision tree that require further
research and/or decisions by the Agency.
                                       111

-------
As part of the overall Soils Project, the Hazardous Site Control Division has arranged
for the Office of Research and Development, through its Ada and Athens laboratories,
to conduct an evaluation of a variety of available soil/ground-water models in order to
identify their appropriate selection and use in determining remediation goals.

Future inquiries regarding this conference  or progress on this issue can be made in
writing to the attention of Mr. Loren Henning, U.S. Environmental Protection Agency,
401 M Street, SW, Mailcode OS-220W, Washington, DC, 20460, or by contacting EPA's
Hazardous Site Control Division at (703) 308-8392.
                                        IV

-------
CONTENTS


                           MORNING SPEAKERS

Speakers                                                                Page

Henry L. Longest, II      Opening Remarks	  1

Alison Barry             Introduction	  4

Dr. Ronald Sims         Technical Aspects of Establishing Soil Remediation
                        Goals	  8

Panel l--Site-Specific Applications of Fate and Transport Modeling and Soils
Regulations

Joe Williams             Introduction	  23

Jeff Rosenbloom         Phoenix-Goodyear Airport Site, Arizona
                        VLEACH Model and Site Characterization	  25

David Kargbo            Superfund Site, Region III
                        Site Characterization and Development of Soil
                        Cleanup Levels	  31

Richard Willey           Beacon Heights Site, Connecticut
                        Use of Summers Model  	  40

Christos Tsiamis         American Thermostat Site, New York
                        Use of Multimedia Model	  43

David Crownover         Pennsylvania DER--Legal and Technical Approach to
                        Setting Soil Cleanup Levels	  48

Questions for Panel 1	  53

-------
CONTENTS


                           AFTERNOON SESSION

Speakers                                                               Page

Panel 2--Non-Site-Specific Development of Numerical Standards for Various
Chemical Constituents

Randall Breeden   Introduction 	  57

Peter Kmet        Washington	  59

Randall May       Connecticut	  66

Steve Cochran     RCRA Program  	  72

Questions for Panel 2	  76

Panel 3--Combination Approaches-Standards With Site-Specific Options

Allen Wolfenden   Introduction 	  80

Lynelle Marolf     Michigan	  81

Jim Pennine       Minnesota	  89

Kate Joyce         New Jersey	  93

Dave Pagan       RCRA Corrective Action  	  94

Allen Wolfenden   California	  98

Questions for Panel 3	100
Appendix A: Proceedings Abstracts and Other Handouts

Appendix B: Decision Tree Generated by Work Group
                                    VI

-------
                           OPENING REMARKS
                            Henry L. Longest, II
MS. BARRY:  Good morning. Thank you all for coming.  We are certainly gratified
to see the number of people that have turned out to talk about this issue with us. My
name is Alison Barry. I'm representing the Hazardous Site Control Division of the
Office of Emergency and Remedial Response.

I'd like to introduce Henry Longest who is the Director of the Office of Emergency
and Remedial Response. He's going to offer some opening remarks before  we  begin
the program.

MR. LONGEST:  There was a meeting in San Antonio in October 1989 with EPA
and ASTSWMO, and one of the key issues that came out of that was the need for
soil cleanup levels.  We  had a lot of discussion about it, and we all agreed we needed
to do something in terms of coming up with  soil cleanup levels.

And today is a kick off to begin this process. It is an important process in Superfund.
We recognize that the risk issues  in Superfund are mainly based on site-by-site risk
assessments where we determine, looking  at  direct contact with individuals, what the
risk is and what the cleanup level should be.  It's now clearly time to focus on the
impacts to ground water of contaminated soil, look at fate-and-transport modeling,
and work  from there to establish  the level for cleanup in soils.

This is not an easy process when you consider the various hazardous waste
constituents we have to deal with. It's going to take some  time. It also is frustrating
at times to realize we have a program that is ten years old, and we're still talking
simultaneously about technologies, and cleanup levels to protect public health and
the environment beyond the direct contact, in areas like ground water and fate-and-
transport of contaminants from soils to ground water.

The Congress directs us, because  its what  their constituents want, to instantly
implement a program, with little or no time  to strategize how it should be
implemented or what needs to be done. For example, we just had to move forward
with the program while we concurrently looked at technologies and trained staff in
new disciplines and  protocols. And that brings me to another dimension of
Superfund—the fact  that  until recently there was no university program  for education
in hazardous substances  response.  A combination of many disciplines have had to
work together to come up with hazardous  waste expertise to support the Superfund
program.  I understand, however, that there  are now about three universities that do
have a program that provides an actual hazardous response curriculum.

-------
Just a couple of comments I'd like to make in general about Superfund.  In addition
to the technical exchange you're here for today, I think the program has made some
major steps forward particularly in the  last two years.  Most of you are aware of the
Administrator's Superfund Management Review, sometimes referred to as the 90-Day
Study.  We completed the study about  18 months ago, and now we are at a point
where the study recommendation have  been almost completely implemented.

There were a lot of charges in the 90-Day Study to come out with EPA guidance,
policies, and directives. Almost all of that has been accomplished. Now I think we're
back into more of an operational mode—we're moving beyond the individual guidance
and policy documents on  the program,  and dealing with scientific issues such as
cleanup levels and  soils.  I hope we can continue to concentrate our time on this type
of issue rather than spending our time  debating with Congress on various reports
criticizing the shortcomings of Superfund.  I  hope we've turned that around. Not that
we won't continue to see  some criticisms.  I'm sure we will.  That's the nature of the
Washington area and the groups that are in  Washington.  But I feel very comfortable
that we've turned the program around, and we're accomplishing a great deal,.

The strategy that came out of the Superfund Management Review specifically related
to the concept of "Enforcement  First".  And  that is taking place.  Sixty to 70 percent
of our new sites where we initiate RI/FSs are being taken over by responsible
parties.  This is as it should be.  Our long-term projection is that only 30 to 35
percent of the projects will be federally funded.  So I think we have a real success  in
terms of implementing the concept of Enforcement First.

Another key outcome of the 90-Day Study was the fact that we're moving out to
clean up acute threats first, and making sites safe as we use the removal program
more aggressively.  As many of you are aware, we put a lot of emphasis on this in  the
last year.  We identified 50 National Priority List sites where we needed fast track
removal action to make the site safe. To us that means remove the acute threats as a
first priority.  Lower level threats are addressed  as subsequent priorities.

A third major thrust of the 90-Day Study was to make sites clean.  Make sites clean
refers to the long-term remedial action, and I see this conference as one of several
that are directly related to the objective of making sites clean because we're talking
about cleaning up soil and debris which relates to long-term, not short-term, cleanup
action.

The last point I would like to make relates to our work with states and other federal
agencies.  We appreciate their participation  because we recognize in some cases
states and federal agencies may be out in front of us.  This  also applies to the
academic community.  But I'm pleased that we do  get together, and pool our
knowledge so we can share information with everybody who is working in the
program.

-------
So in closing I'd just like to say: welcome to this conference.




Thank you.

-------
                              INTRODUCTION
                                 Alison Barry

MS. BARRY: I'd like to present some of the background information for this
conference in order to frame the subsequent discussions.

Our goals for today are to facilitate the exchange of information about the range of
approaches to protecting ground-water quality through determination of soil quality
criteria at hazardous waste sites.

We'd also like to use this opportunity to express our interest in the issues of soil
remediation goals and to give you an idea of our ongoing and planned projects for
developing policy and guidance.

We have a very large group here today which includes representatives of 29 states,
8 EPA regional offices, several Office of Research and Development laboratories and
both the  CERCLA and RCRA headquarters program offices.  Given the variety of
technical and regulatory experience represented, I am confident that today's meeting
will stimulate policy development within the agency by offering a range of
perspectives on a common problem. I hope you also find this meeting useful to your
programs, and that you take subsequent opportunities to communicate with us about
this issue.

Let me begin by offering a brief summary about the context.  Within the Superfund
program, cleanup levels for contaminated sites are established on a site-specific basis
using a risk assessment approach.   The risk assessment usually considers  direct
contact exposure pathways. These  might include ingestion of contaminated ground -
water, inhalation  and ingestion of contaminated soils, and contact with surface water.
Usually we incorporate standardized exposure assumptions when we can  apply them.
However, risk assessments are tailored to reflect site- and chemical-specific
characteristics which may vary considerably from site to site and prevent  the usage of
our cleanup levels at this point. Consequently, soil remediation goals vary from site
to site as well as for a particular contaminant.

The interaction of environmental media such as soil and ground water should  also be
evaluated in the determination of soil cleanup goals.  In many cases contaminant
migration from soils to ground water will be a primary although indirect  pathway of
exposure. Contaminants leaching from soil particles in the vadose zone steadily
degrade ground-water quality causing contaminant levels in  drinking water supplies to
exceed health-based criteria such as MCLs or MCLGs or state equivalents.  This
pathway can  be technically difficult to evaluate since there are many subsurface
physical and  chemical processes which control leaching at a site.

-------
Currently there is no consistent approach to evaluating this pathway within the EPA
or among the state  regulatory agencies. Today you'll hear about a number of
methods for determining cleanup goals which account for this pathway in some way.

However, migration to ground water and determination of cleanup goals which
address this pathway reflect complex policy and technical issues. I'd like to outline
some of the larger issues before we begin our presentations.  The two main
categories of approach so far to developing soil cleanup goals have been non-
site-specific numerical standards, or the use of  a site-specific  evaluation.

Both approaches have their strengths and weaknesses.  Generic standards are easy to
use because they apply to all sites uniformly,  require little data collection or site
characterization,  and are consequently less expensive at this stage for both regulators
and  facilities.

However, because site conditions vary widely, standard criteria  may be inappropriate
for a specific site-either too conservative or not conservative enough as the case may
be.  Costs associated with reaching very conservative numbers might outweigh the
actual risks, or these numbers may be technically impractical to reach,  given available
technologies.  Standards also have to be developed that are technically defensible for
a great  range of sites.

Generic standards cannot reflect the issue of the  total mass of contaminants which
might potentially be released to ground water.  A site, therefore, with a relatively
small mass of contaminants and soils would be  subject to the same  standards as a site
with a large total mass of contaminants although  the former site represents a far
smaller threat to ground-water quality.

Site-specific approaches can have the advantage of greater accuracy because they may
account for at least some of the various physical and chemical processes  which
govern the behavior of the contaminants in the subsurface. These can include things
such as organic carbon content, partition coefficients, volatilization, biodegradation, et
cetera.  They may also be able to account for social and economic issues such as land
or aquifer use.

However, if fate-and-transport modeling is used to evaluate the potential for leaching
at the site, the utility of this approach depends  on the adequacy of the data available
for input as well as the appropriateness of the model selected, the input requirements
of a  model, the necessity of extensive site characterization, or collection of data that
is difficult to obtain or for which there are no standard operating procedures.
Meeting these requirements may be more costly than warranted by the degree of
greater  accuracy than can be achieved or in the cost associated with more
conservative cleanup criteria.

-------
In addition, not all models are suitable for all site types and hydrogeologic conditions.
The use of modeling entails decisions about which models to use and about which
conditions they should be used to evaluate.

Lastly, if site-specific modeling is employed, regulators need to decide how to use the
results in the  development of cleanup goals either as clean-up criteria or as risk-
management tools.

The third possibility which a number of states have recently implemented or are
considering is basically a hybrid of the previous two categories.  Standards may be
used in various ways: as triggers  for more site-specific analysis, as default cleanup
standards for certain site types, as screening tools for certain types of response action
or levels defining future site management areas.

Modeling or other site-specific risk analysis may be used where greater accuracy is
desired or for certain site types.  The advantage of this hybrid approach is the
flexibility afforded to the decision-making process.

So in designing an approach to soil remediation goals the following issues need to be
considered by regulators:

Do you want the  approach to reflect site-specific conditions that might be technical or
non technical  such as the land use issues I mentioned?

How closely do you want your goal to reflect actual risk?

If you're going to use some sort  of fate-and-transport modeling, how will you
interpret the results?

How accurate do you want to be?

What sort of guidance are you prepared to provide to select and use specific models?

If you're going to use numerical  standards, on what basis will you develop these
numbers?

How will you  use them?

How much data and what kind of data is necessary to support your approach?

What sort of resources do you want to commit to determining remediation goals or
site remediation?  Are some sites worth spending more on?

How much regulatory oversight are you prepared to provide?

-------
In the following presentations you'll hear a variety of responses to these questions
from the perspective of state regulators and EPA and regional staff. The first group
of speakers has been drawn primarily from EPA regional offices and represents the
most site-specific examples of developing soil cleanup numbers based on a leaching
pathway.
Our second group will consist of several non-site-specific approaches in which soil
standards are  directly tied to ground water standards or have been developed to
account for leaching to ground water.

Our third group will discuss the hybrid regulatory approaches in which a standard
process is designed to accommodate a variety of site types and exposure pathways.

Before we begin the presentations, I'd like to introduce Dr. Ron Sims, professor of
Civil and Environmental Engineering at Utah State University.  Dr. Sims has been
working very closely with the EPA through the Robert S. Kerr Laboratory in Ada,
Oklahoma  on issues such as vadose-zone processes, site characterization, and soil
remediation.  I asked  Dr. Sims to begin our meeting with a brief discussion of some
of the technical aspects of establishing remediation goals so as to form a kind of
context for our regulatory discussions subsequently.  Dr. Sims.

-------
              TECHNICAL ASPECTS OF ESTABLISHING
                      SOIL REMEDIATION GOALS
                Ronald Sims, Ph.D, Utah State University,
                   Civil and Environmental  Engineering
DR. SIMS: We are at the EPA-state conference on soil cleanup levels. As Alison
Barry just pointed out, the idea is to focus on the ground-water migration pathway.
So I'd like to welcome you to the first technical presentation and state that we're
going to explore for the next maybe half hour some issues involved in the migration
pathway to ground water dealing with  soil and soil cleanup levels.

The bottom line that we're talking about today is ground water protection.  If you
take contaminated soil and use some technology to reduce levels of constituents in
the soil, once the remedial technology has been completed, what's the potential for
those residual chemicals to move into  the water?  How fast, as Alison has mentioned,
and what types of concentrations will result in the ground water?  One term that I
use to talk about this is the level of contamination that will be "sponsored"  from the
soil into the ground water. The soil may act as a mechanism to sponsor a given
chemical concentration in the ground water so we want to look at  that connection
between the soil and the ground water for the next half hour.

One example of a Record of Decision in December 1988 concerns the Libby,
Montana, Superfund ground water site. I wanted to pull some quotations from that
document to illustrate what we're talking about today.  In the document it's
mentioned that contaminated soils present a public health threat via direct contact
and ingestion, and at many Superfund sites and in many Records of Decision the soil
contamination problem is looked at in terms of direct contact: workers at the site,
ingestion of the soil, and then  the risk assessment based upon that contact.

The Record of Decision also goes on to state this: "Contaminated soils also pose a
direct environmental threat and public health threat because they  act as source
materials by releasing contaminants to the ground water." And that's what this
discussion today is all about:  this pathway where the soil material acts as a source by
releasing contaminants to the ground water. What we're going to  discuss today is
how the soil actually goes about doing this, sponsors contaminants into the soil water
which can then enter the ground water.

If we expand our scale and look at the components of the soil-ground water system,
at the top of the  slide, there is a belt of soil water. This is where  water can actually
move upward as well as downward: upward by evaporation, and those plants that are
sitting at the top of the screen are also able to move water upward. Leaching can
move water downward. So when we have contaminants in the soil, they move into
the water, the water can move up as well as down and sideways.  It's a three-
dimensional problem.

-------
Then there is an intermediate belt below the soil root zone.  Then there is a capillary
fringe where the soil becomes wetter and water can actually be drawn up from the
saturated zone into the unsaturated zone.  So on the right hand side of the slide we
have the zone of saturation, and then the top of the zone of aeration. The zone of
aeration then is where our contaminated soil sits. If we have an underground storage
tank or leaky tank or some release to the soil, this can be either at the top of the soil
or farther down towards the capillary fringe.

In terms of the point of compliance, an interesting discussion now is that the point of
compliance for soil cleanup might be taken somewhere around the capillary fringe.  It
could be looked at anywhere within the zone of aeration. Note that the  zone of
aeration means the presence of air.  So in the system we  not only have soil and the
water which becomes contaminated, but we have air; and if we have oily materials,
then we have an oil phase.  So, in the zone of aeration we have several  phases, and
they would include the  soil phase and the water phase and then the air phase and
perhaps an oil phase. And so we have to look at the sponsoring of chemicals into the
water from air, from soil, and sometimes from oil.

If we go to the top of that previous slide and identify sources this could be a generic
site with a lot of problems at the site.  This site's been closed down, of course, and
it's a Superfund site; and we're now looking at source control.  So we remove the
septic tank on the left hand side of the slide. We removed hazardous waste piles and
the rock salt and the pesticide  contamination, and then we traditionally look at the
ground water. And if the ground water is contaminated, we go back and look for
sources on the surface,  and we move some more waste piles. So  this has become a
process where we oftentimes go to the ground water, go back to the surface looking
for more sources of contamination,  and after we remove all the sources  on the
surface, and we still have ground water  contamination, we're not quite sure what the
problem is.

This is what I call a black box. In this demonstration it's a blue box. On the output
side we see that the ground water is contaminated.  On the input side we go back
and look for sources on the surface.  Those are the rock piles and septic tanks,  and
we remove those and go back to the ground water.  What we want to do today is to
get inside  the box and look at the issues in the soil that are the bridge between the
input and  the output, meaning  the bridge between the material on the surface, the
hazardous waste, and the ground water that we see contaminated. And  oftentimes
we don't connect these.  We don't identify the issues in the soil that connect the input
to the output, and so we're continuously frustrated as the ground  water remains
contaminated, meaning above the maximum contaminant levels, but we have
removed sources on the surface; and we're still trying to figure  out what the problem
is.

This leads to a lot of confusion. It also leads to a lot of head spinning.  Sometimes
we go back out and sample more.  We take more samples, more samples.  Some

-------
people have called it "carpet sampling," where we go out and just sample the whole
site to try to figure out what the problem is and our belief and our faith is if we get
enough samples, if we take enough chemistry, if we do enough concentration
analyses, we'll see the problem.  And one of the lessons I think we've learned from
Superfund is that just collecting more data  does not ensure you will see the problem
more clearly.

Oftentimes collecting more data,  and in this case chemical data, leads to more
frustration, more confusion, and essentially becoming overwhelmed. Where I've got a
lot more data, I have a lot more sampling,  but that hasn't ensured that I now see
what the problem is at the site in terms of  soil contamination.  So what we want to
do is back up. And let's get inside the box, the blue box, and let's look at  site
characterization in terms of the processes that are  occurring  in the  soil so that we can
relate more effectively the soil concentration to the ground water concentration.

How do we get a handle on what's happening in the soil and the result in the ground-
water migration pathway?  So I'm going to  back us all up now and  start over again.
We're going to get back in that blue box and characterize a site in  order to assess soil
cleanup levels.  What are some of the issues involved?

Let's start with something called a mass balance approach.  This is something all of
us need in the engineering and science disciplines, and as part of our formal
education we all run into this concept of a  mass balance. And what I'd like to do is
use a mass balance in the soil and discuss what the mass balance means.

On the right hand  side of the slide there is a solid phase, and on the left hand side of
the slide there's a fluid phase; and I have taken the subsurface and in conceptual
form made it into a pie chart. On the right hand  side the solid phase we say is made
up of an inorganic component and an organic component.  The inorganic component
would be the sand, the silts, and the clays that make up soil. And the organic
component would be humic material. The material that you use in your garden to
make a good structure for  your soil to hold water.  It is comprised of organic carbon.
Alison Barry mentioned the importance of organic carbon, and we're going to talk
about why that's important.

We have two phases on the right hand side where  a chemical may be held up, may
hide, may reside, and on the left  hand side we have a fluid phase that's made up of
three compartments. We generally have water in the unsaturated zone.  But because
it's the zone of aeration, we also  have gas.  I use the term gas  instead of air because
in the soil microbial  activity can generate carbon dioxide levels that are 20 percent,
and in this room carbon dioxide in the beginning of this talk is much less than  half a
percent. By the time I'm done we'll have it up to 20 percent. Carbon dioxide is very
low in air but it's not so low in soil. At the bottom of the slide there is the term
N-A-P-L which is NAPL or Non-Aqueous Phase Liquid.
                                       10

-------
I mentioned sometimes in the zone of aeration we can have oil.  These can be
LNAPLs, the light non-aqueous phase liquids, or DNAPLs which dense non-aqueous
phase liquids, but they represent another fluid phase; they flow - under stress they
move.

Now, the idea is that if we have contamination in the migration pathway to ground
water, then those other compartments-the inorganic, the organic side, the NAPL
side, the gas side-all provide reservoirs for dumping contaminants into the water.  So
the contamination in the ground-water migration pathway may come from not only
the soil solid phase, but may come from the gas phase.  We may have TCE in the soil
volatilizing into the  air, and when it rains  it dissolves in  water. We know it dissolves
to a level of a thousand milligrams  per liter.  So it may be in  the air phase. The air
phase then receives  water and the TCE is then dissolved in the water and goes down
to the ground water.

In the non-aqueous  phase we may have TCE in some type of  oil and the partitioning
or the bleeding out  of the chemical is from the NAPL phase to the water phase.  So
when we say soil or the soil environment,  we're saying that  the chemical, the
chemical contaminant, the constituent may be in the inorganic side, the organic side,
the NAPL side, the  air side;  and all of the phases may be dumping chemicals into the
water phase which then travels down  to the ground water.

So let's look at each one of these phases to see how they work.  This is the inorganic
side of the subsurface compartments.  The inorganic side: according to the soil
texture tri-linear diagram the soil is made up of the inorganic components  of sand,
silt, and clay.  So you can have clay which is in the top part of the diagram, meaning
it's a larger percent  clay than silt or sand.  If we go to the silt end we have a lot of
silt, and in between we have sandy loams  and clay loams and  these-this inorganic
side of the soil, the  different particles, then have different implications in terms of
sponsoring chemicals to the ground water, to the soil water.

How do we measure the relationship  between texture and what occurs in the ground
water?  The texture of the soil whether you have a silt loam or sandy loam or clay
loam may indicate a different tendency to sponsor chemicals from the soil phase into
the ground-water phase.

This is the picture, a diagram or a schematic of the organic compartment in the
subsurface.  This doesn't move.  Generally the humic material sits on the soil. And
this is part of the soil matrix. When you look at it initially you might say it looks like
part of a hazardous  waste mixture.  There are benzene rings in there.  There are
carboxyl groups. It  looks structurally  similar sometimes  to the hazardous constituents
that end up in contaminated sites.

This represents the  organic carbon that  Alison Barry was talking about. There may
be a tendency  for chemicals in the soil to  reside in that humic material or organic
                                      11

-------
carbon. Conversely that may be where the contamination sits, and then the chemical
moves from there into the water phase.  So there's the organic side or organic carbon
phase.

Now here's the non-aqueous phase.  Notice the blue-green material that's trapped in
those oil particles. That's an oil. It has become physically trapped.  We refer to this
as the residual saturation. That non-aqueous liquid that's trapped in the soil
represents a source of contamination in the subsurface environment.

So now we have removed all the hazardous waste piles and the rock salt and septic
tanks. The ground water is still contaminated.  This is another source in the zone of
aeration that may be the  reason the ground water is contaminated.  What's going to
happen with the chemicals in that blue-green residual material?  Some of the
chemicals  may volatilize into the air phase.  So it  may have an upward migration,
and when  it rains the volatile chemicals dissolve into the water and go down to the
ground water. As it's moving down  and volatilizing it's also smearing around the soil
particles and, by adsorption, it's coating the soil particles. So when it rains again or
there is surface runon on to the site, the chemicals move from the soil into the
ground water. And then, of course,  we have direct leaching to the ground wa.ter.

So this non-aqueous phase liquid represents multiple pathways of sources to contami-
nate the ground water. By the volatilization pathway you contaminate ground water,
by desorption you can contaminate ground water,  and by direct leaching.  So what
we're talking about here are pathways by which the soil can sponsor chemicals into
the ground-water pathway.

Volatilization.  This slide shows a way to measure volatiles coming up through the
soil surface.  It's a hood we put across the surface and measure the flux of chemicals
coming out the top. That's the volatilization pathway on our mass balance wheel or
diagram.  So far we've looked at the mineral side, the inorganic side, the organic
compartment, the humic material. Then we went to the left side  of the pie  chart, and
we looked at the non-aqueous phase, the blue-green saturation, and the volatile
phase here; and we're saying all of those compartments in the subsurface can sponsor
or dump chemicals into the soil water.

So I presented the diagram this way: We have a hazardous constituent that enters
the soil. We've seen that there are solid, liquid, and gas phases in a soil.  The fate of
a constituent can move upwards into the volatilization phase, then go directly to the
ground water through dissolving in rain water.  We can have direct leaching (at the
bottom of the slide) of the hazardous constituent, and we can also have
transformation  or degradation of the chemical as well as biodegradation, that results
in intermediate products-let's say vinyl chloride from trichloroethylene.  Those
intermediate products then have their own leaching potential, their own volatilization
and degradation potentials.
                                       12

-------
So the idea when we talk about soils and the relationship between soil and ground-
water quality or water quality that enters the ground water is that, the soil can be
divided into compartments. Each compartment can be looked at as a reservoir of
chemicals that sponsors or dumps or bleeds or moves chemicals into the ground
water pathway.

Drawn another way on this slide: on the top there's the atmosphere going into the
top of the system.  On the bottom is saturated zone. We're interested today in
looking at the middle zone which is called  the vadose zone, and I mentioned four
elements we're looking at in this zone.  We can have oil in the soil. We have the soil
phase. We have air phase and water phase. As it rains and as the sun comes out the
water can move up and down, and with it the contaminants can move up and down.
So as we've just discussed chemicals in the oil phase, in the soil phase, and in the air
phase all have arrows going to the water. Therefore they can all dump chemicals
into the water.

Conversely the arrows go back from water  to air and soil and oil.  So in the
subsurface there are some type of processes, then, whereby the chemical goes back
and forth or among the phases from water to  soil to oil to air, and our question is
how do we get a handle on this.  Is this overwhelming?  Should we give up and just
excavate, forget the whole thing?  Conversely can we take a look  at tools or methods
to evaluate the quantitative relationship  and then be able to assess how much
chemical will be in the water  phase.

What we want to do is talk about ways to explain the distribution of a chemical
between water and soil, water and oil, water and air, and this is by the use of
partition coefficients.  There are several methodologies available to calculate or
perform or get a partition coefficient.  The first one says "Kd"—that's the
concentration in soil over the concentration of water. So the higher the Kd the
higher the concentration of the chemical is in soil compared to water.  The second
one "Ko" is the partition coefficient for oil  and that represents the concentration  in
oil over the concentration of water.  So the higher the Ko the more of the tendency
of the chemical to be in the oil phase, and on the bottom Kh is Henry's  law constant,
which is the relative ratio between the chemical in the air and the chemical in  the
water.

Notice that water is in the denominator in  all three equations, and water's what we're
interested in today. What we are interested in, then, is what are the values of Kd
and to some extent then I can get an idea how much would be in the water if I know
the Kd, and I know the concentration in soil. Right?

This represents an approach, a quantitative approach to getting a handle on how
much chemical's in the soil versus water, oil versus water, and air versus water in a
quantitative way.
                                      13

-------
Then we can use this information to look at the concept of retardation of the
contaminant in soil, "R".  Retardation is defined as the velocity of the water over the
velocity of the constituent.  So we put the constituent moving at one velocity and the
water at another velocity.  Why? The water moves in the water phase, but the
constituent moves between the water phase, the oil phase, the soil phase, and the air
phase. So it spends time in other compartments and  to that extent it may not travel
at the same rate as the water.  If R is greater than 1,  then the velocity of the water is
greater than the velocity of the constituent, so the constituent is lagging behind.
There may be  treatment techniques that encourage that, to increase the retardation,
to hold the constituent up in order to prevent the  constituent from getting to ground
water.

Notice the retardation, then, how fast the chemical will move compared to  how fast
the water will  move, on the bottom equation is related to  soil physical chemical
properties. The first one "P" (rho) is the soil bulk density.  So, if we want to get
some idea of the relative speed with which water versus the chemical enters ground
water, we need to know the bulk density of the soil.

The second element is Kd. Kd is the partition coefficient between water and soil.
We have Kd divided by theta.  The symbol theta represents the amount of moisture
in the soil. So if I want to predict retardation or the  speed with which the  chemical's
moving  down to the ground water from the zone of aeration, I  need to know the soil
bulk density, partition coefficient, and  soil water content. Not a tremendous amount
of information, but it tells me what I need to use to evaluate the tendency of that
chemical to head towards  the ground water after cleanup; a quantitative relationship
to evaluate what's going to happen after treatment technology's been used.

Metabolic pathways.  Then we add on degradation. We have these critters, micro-
organisms (depending upon who you are you call them different things) that
biodegrade the chemical in the subsurface of the soil, and there's a lot of controversy
right now about the efficacy of bioremediation after the site is cleaned up.
Sometimes, in fact, we meet snake oil salesmen and people who sell magic bugs, foo
foo dust, elixirs, all kinds of things.  Some people  actually have kind of a belief in
bioremediation which is beyond the rational.  People  believe there's a bug  out there
that will clean up everything at any time and so it goes beyond controls, goes beyond
science; and I  think that they sometimes acquire the name "spray and pray."

I'd like to redefine what that whole idea of spray and pray is as opposed to
bioremediation, I call that "bioredemption".  Bioredemption of soils.  We put a
chemical in the soil.  We watch it disappear and we say it's biodegraded. In terms of
looking at the future fate and behavior after treatment, if the chemical is actually
sponsored into the air phase, we haven't really biodegraded:  it's gone into the air
phase. If it's leached  into the  ground water, we haven't biodegraded:  it's leached
out. So it's important to differentiate between biotransformation and  bioredemption
                                       14

-------
when we're looking at trying to assess the future fate and behavior of the site. So I'd
like to make that distinction for our discussion today.

In looking at degradation many people have seen this formula DC divided by Dd
equals negative Kc, then half-life (ta/2) equals negative 0.693 divided by k.  That's a
first-order equation that's commonly used to express biodegradation.  We need to
discuss today and tomorrow is what does the ta/2 include? Is it bioredemption? ta/2
meaning it disappears; it's gone so it  must have been biodegradation, or did that ta/2
account for what was volatilized?  What actually occurred due to microorganism
degradation versus what actually volatilized or what actually leached or what actually
degraded under  abiotic conditions or chemical hydrolysis. We need to know this in
order to accurately determine fate and behavior, and rate and extent of
bioremediation.  Because if it's really  due to volatilization, and we haven't  measured
that, when we try to  take into account the degradation and the volatilization and the
leaching we're going to get completely mixed up in terms of predicting subsurface
processes.

So ta//2-when you see biodegradation we need to focus in on:  Is it biodegradation or
is it simply disappearance of the parent compound? As I mentioned, it's important to
know how much is in the air phase, the soil phase,  the oil phase to account for
chemical fate so we don't attribute that to biodegradation.

The misunderstood world of saturated flow.  I  talked about chemical and physical
processes. Now let's talk about movement of water up and down. Let's zero in now
on the water phase.  We have some idea of the oil phase, the air phase, the soil
phase. Now let's talk about the water phase and the unsaturated zone. Jay Lehr is
the editor of Ground Water Monitoring Review, and Jay Lehr has written a very nice
discussion within the last couple of years called the "Misunderstood World  of
Unsaturated Flow," and he  defines the vadose zone there as "vadosis" meaning
shallow and "vadera" meaning to walk or wade. So it implies slow movement at a
shallow depth, meaning above the ground water.  But when we apply saturated flow
concepts to unsaturated flow, we often get into trouble.

We assume that water in the vadose zone will behave the same as water in the
saturated zone.  I'd like to point out that sometimes it doesn't do that.  In terms of
permeability-this is what we constantly think about, that permeability increases from
left to right on the slide. The numbers are not important but the relative ranking
there shows that the permeability  of the clay is much less than the permeability of
gravel.  So you would say gravel has the highest permeability, sand is the next highest,
silt the next highest, and clay is the lowest.  This is in a saturated environment.

We then sometimes apply this to the  unsaturated zone.  On the X axis is the
volumetric water content; how much moisture is in the soil. On the Y axis is the
hydraulic conductivity. Notice in the unsaturated zone the hydraulic conductivity is a
function  of the amount of water in the soil. The dryer the soil the lower the
                                       15

-------
hydraulic conductivity, meaning the slower the rate at which water will go through the
soil, and, therefore, the slower the movement of the chemical to the ground water.

This is different than in the saturated zone where we have a constant amount of
water which is equal to the porosity of the subsurface of the  environment.  So we're
saying here the dryer the soil the slower the water moves.

Now,  this slide illustrates a soil characteristic curve and again conceptually don't
worry about numbers. On the top, water content goes from left to right, increasing
from left to right.  Zero to 0.30, which means just 30 percent water in the soil. And
down the X axis as we go from top to bottom the soil becomes dryer.  I put a greater
suction on the soil to force water out.

And as we notice here we have different textures of soil we talked about before:  clay
on the right hand side,  sand on the left hand side.  What this means is: as I apply
more  and more force, as  I dry out the soil, as I push more and more water out, the
clay will hold much more water than the sand will.  There will be more water in the
clay than in the sand, and we said that hydraulic conductivity is a function of the
water content.  So we need to know the texture of the soil in order to have  some idea
of how fast water's moving through the soil, if the movement is a function of the
water content.

This is different than in the saturated zone. So if we get information on texture, we
can relate this back to the soil water characteristic curve in order to predict how fast
water will move in the water phase to the soil. Then we look at the sponsoring of
chemicals from soil, air,  and oil into the water. Thus we have some idea of how fast
we  are contaminating the ground water through the ground-water migration p>athway.

I'm going to show you something in the  unsaturated zone that is very hard to believe.
If we  put all this together, we have on the bottom on the X axis how dry the soil
might be.  That H in centimeters of water represents an energy and the soil gets
dryer as we go from the right-hand side  to the left-hand side. So zero, 100, 200, 300,
400 indicates a dryer soil and that Kh over K is the hydraulic conductivity under
unsaturated conditions  over hydraulic conductivity in the saturated conditions..

Notice the hydraulic conductivity of sand in the right-hand curve, goes to zero much
more  rapidly than the hydraulic conductivity in the  loam or the clay.  So we can
actually say, that, at a certain amount of dryness in the soil, the conductivity of water
through clay is faster than through sand.  So it's backwards, isn't it?  It turns the
whole world upside down and Jay Lehr points out what he calls the  "topsy-turvy"
world of unsaturated flow where it seems that day is night.

You say that doesn't make any sense. Water goes slower through clay than through
sand.  That's true in the saturated zone. In the unsaturated zone we need to have
some idea of the  texture  in order to relate this to the soil characteristic curve, and in
                                       16

-------
order to predict how fast water will move. Sometimes we put clay caps on top of
things to keep water out. We keep the surface water out, but we suck up the
contaminants from underneath the clay liner.  They get into the clay liner because
there's clay in there and clay can suck up water.  So what I'm pointing out here in the
unsaturated zone is that things don't move necessarily like they do in the saturated
zone. We can get  a handle on that through texture and through the soil
characteristic curve that we saw on the last page.  So this all relates now to ground
water.

Comparing porosities of the sand, silts, and clays indicates that clay has a higher
porosity than sands and silts.  Clay has more pore space; that means clay can hold
more water.  If a chemical constituent is in the water, clay can hold more
contamination than sand can hold. That's another reason for knowing soil texture.  If
we know the  texture, and we have some idea of the porosity, we have some idea  of
the amount of materials stored in the vadose zone. So  the more water we store
(meaning the more clay), the more pollutant we store in the clay.  We're now talking
about the water phase, and how we describe what's in the water phase, knowing
something about soil texture and the soil characteristic curve.

To summarize: in the subsurface environment if we take water, oil, soil, and air and
we put these  together, we have processes that spread the contamination around,
spread it out  and into the ground water.  And those  processes are called partitioning,
advection, and facilitated transport. Concerning facilitated transport, let's take some
of that NAPL and  dissolve it in the water so that the water's now not really water
anymore but  it contains things.  It contains particles.  It contains solvents that will
facilitate the  transport of an organic chemical.

Then we have dispersion. These are quantitative terms that can be used to evaluate
the spread of contamination.  Conversely, holding chemicals back from getting into
the ground-water migration pathway are things like ion  exchange for metals. These
are processes in the soil mineral and organic matter compartments that retain metals
and organic chemicals.  Biotransformation will retard the movement of the chemical
by destroying it, by removing it, by biological destruction. Precipitation of metals
generally in the soil will keep them from getting into the water pathway because
they'll partition into the solid compartment in the subsurface. And sorption moves
the chemical  from  the water phase to the soil phase, from the water phase to either
the mineral or the  organic side. Therefore in the subsurface we have things that
retard the contamination. What we are trying to do is put together in  an organized
manner, the mechanisms that spread contaminants down to  the ground water, and
mechanisms that hold the contamination up in the vadose zone or destroy it in the
vadose zone and therefore prevent it from getting into the ground water.

We can use this information for exposure assessment in terms of migration to the
ground-water pathway.  We're talking about routes of exposure.  We're zeroing  in on
the ground-water pathway, but I want to show you that the ground-water pathway is
                                       17

-------
influenced by the routes of exposure in the environment. What's the exposure in the
air pathway? That can dissolve into ground water.  What's the exposure through the
soil pathway?  That can dissolve into the water. What's the exposure from a rion-
aqueous phase?  That can dissolve into the water. We're looking at routes of
exposure in  the subsurface and we're going to zero in on the ground-water pathway
and this involves tracing environmental emissions from the source to the population.

After we remove all the sources on top of the vadose zone, all of the contaminated
materials, the rock piles, the pesticides, the underground storage tanks, et cetera,
we're looking at subsurface sources; and that is the pie chart we just looked at in the
subsurface.  If we integrate this information using the partition coefficients  Kd, Kh,
Ko, tj/2, texture, bulk density, how do we come up with a definition of the problem?
One way to  do that is through modeling these environmental processes.  It's difficult
for us to sit  there and line up all the Kd and t1//2 values and texture. We got this
information, but how do we now put it together to try to figure out what is going to
be the  result of the concentration in the water phase that moves, and how fast it
moves  to the ground water? We're going to talk about using modeling as a tool to
help us get there.

Let's talk about mathematical modeling and  again we can ask many questions of our
mathematical model.  We can say to the model:  Tell me how far constituent moves
in two  weeks.  We can say: I want to  know the exact concentration at two feet.  We
can also ask another question, and that is: Can the model rank for me, put in order
of priority, which  chemicals are going to volatilize first, and which chemicals are
going to leach first, and which chemicals are going to stay stuck in the soil by
adsorption?  How about doing that type of ranking so that I know which chemicals
are going to enter the water phase most rapidly.  Which chemicals are going 1o get
into  a ground-water migration pathway?

Another question we can ask the model is:  If you take the partition coefficient and
tj/2 and soil  structure and texture can the model tell  me:  Of the 60 chemicals at the
site,  which 10 are likely to go up, which 10 are likely to go down, and which 10 are
likely to be stuck? And prior to that,  rank them for me so I can tell which chemical I
should look  for first, I should monitor for first.  If I have ten dollars to trace what
happens at my site after clean up, I then know which chemicals to go after first in  the
air phase and the water phase and soil phase so I don't spread my money around
and get very poor quality data by getting pieces of information scattered across all the
chemicals of the site.

Mathematical  modeling can help me to analyze which pathways I should go after,
which chemicals I should monitor for first in a prioritization approach.   However, at
most Superfund sites Kd, Kh, Ko, t1/2 based on biodegradation, the soil  water
characteristic curve or relating the amount of water in the soil to the flow in the soil
is not measured.  Although there are methods to develop this information,  it's
                                       18

-------
generally not measured. So we have that first challenge: to get that information in
order to use the mathematical models.

The use of fate-and-transport modeling, then, can be based on a chemical mass
balance approach and, in fact, mathematical models do that.  The chemical mass
balance approach would say: Where is the chemical in those four phases: the air
phase, the water phase, the soil phase, and the oil phase. We need to know that in
order to know how much chemical is being dumped from the oil, air, and soil into the
water phase.  We are looking at a mass balance approach that models can help us to
achieve.

Unsaturated zone site-specific parameters would be:  the site-specific texture because
that's going to relate to how  much water's in there; the site-specific partition
coefficient because that's going to relate to how much organic matter's there; the
Henry's Law constant is not specific; we can look that up in the book. We can look
up Ko of the oil based on oil-water partition coefficients that we can look up in a
handbook.

We can get some information independent of the site, but we need some site
information, and Kd (the partition between soil and water)  is going to be a function
of the amount of organic carbon in the soil. And the texture's going to tell us the
hydraulic aspect of the soil.  We need  some site-specific data, and models and input
parameters, then, can be used to  take a mass balance approach to help us assess what
would be the sponsoring of contaminants to the water pathway down to the ground
water.

Let's zero in on data needs now,  on unsaturated zone model requirements.  There
are many unsaturated zone models.  I am not a mathematical modeler, but I have
used mathematical models. I need to know what goes into  the models in order to use
them and lean on the expertise of others who are experts on mathematical models.
There are a number of them.  Look at what most unsaturated zone models  require in
terms of information, that's important to generate for assessing the ground-water
migration pathway.  Physical properties we've already talked about.  Bulk densities
relate right back to retardation; we need the bulk density to get the retardation.
Porosity we talked about.  We need to know the porosity to know how much water's
going to be stored.  If it's clay then more water will be stored, more contaminants will
be stored.  If it's sand less water's stored, fewer contaminants.  We need to know the
soil characteristic curve we've talked about to know how much water is there
depending on how dry the soil is because we know that relates to hydraulic
conductivity.  Soil organic carbon content: we need to measure Kd partitioning and,
as you saw, retardation depends upon Kd and Kd depends upon the organic matter
content. And Kd was used to calculate the retardation factor.

Soil structure also gives us an idea of fracture flow in the zone of aeration.
Especially in the southeast and, in fact, around this area we have clays that crack.
                                      19

-------
We have red clays in the southeast that crack and open up when they get very dry
that affects the soil structure. If you have a large crack, you're going to have a large
input of contaminants into the crack.  So soil structure's important to let us know the
aeration of the soil and fracture flow possibilities.  Then unsaturated hydraulic
conductivity:  that's important because that tells us how fast water gets to the ground
water.

Here is one model, the RITZE, Regulatory and Investigative Treatment Zone
Enhanced, model pulled out because I had the information on this model. There are
soil properties, waste properties, and environmental properties and this is typical of
other unsaturated zone models.  Soil properties:  we need the saturated water content
of the soil.  We know that's important because water content determines transport or
influences transport.  Soil bulk density we've talked about.  Soil porosity we just
talked about.  Soil moisture coefficient:  We know why we need that, because that
influences how fast the water moves.  Those are the soil properties.

Waste properties.  Constituent concentration. How much is there?  Alison
mentioned that it is important to get an idea of the mass of material.  The
constituent concentration across the site can give us some indication of the mass.
The mass fraction of the oil applied because we need to know the oil compartment in
the subsurface zone.  The density  of the oil is important to know. Is that oil going to
move through the subsurface? The density of the waste and the detection limit in
terms of what can we measure in a soil.  Your  chemist will tell you  he can measure
much more accurately in the water phase than  the oil or soil phase. So there's a
tendency to measure things in the water phase, but if the soil and oil are sponsoring
chemicals into the water phase it's important to know your detection limit, because
99 percent of the contaminant may be in the soil phase, and that's kind of a reservoir
or pool where it bleeds into the water phase.  So you need to know the detection
limit in the soil, too.

Then the environmental properties, including the Van't Hoff-Arrhenius coefficient.
That is a temperature correction and if you're in the north or south and you have
hydrolysis and biological degradation, you can correct the rates of degradation for
temperature.  This as we saw can  go back to calculating the half life which you can
then put  into this model. Then the recharge rate and the temperature.  Temperature,
of course, relates to the Van't Hoff-Arrhenius correction factor.  Recharge rate.
What's the runon, runoff? How much rain are we getting?  We may have to control
that. In fact, part of the treatment up front may have been to control the runon and
runoff. What we're doing, then, is to integrate this information with a model so that
we can do a mass balance across the zone of aeration to tell how much chemical's in
the water phase as a result of all  these processes.

Superfund site soils data.  Generally defining the extent of contamination only.
Generally the information collected is chemistry data.  How much is in the soil?  As I
mentioned, carpet sampling across the whole site. How much chemistry in every foot
                                       20

-------
of the soil?  And as we saw, if you get that information you have one input
parameter-trial's concentration-but you don't know anything about the potential
movement either upward into the air, downward into the soil, or retardation in the
soil if you only measure concentration. While you're taking the soil core, you can
measure other things.

So while we are out there taking surface soil core, we may not want to spend all the
money measuring the concentration of contaminants; but we may want to get the soil
texture while we're there, the soil structure, the bulk density, the organic carbon
content, so that we can use these to put into mathematical models to get an
assessment of the mass balance in the subsurface so we can zero in on the ground-
water migration pathway.  So we're suggesting here, while we're talking soil cores,
let's do a little  more and perhaps let's take fewer soil cores and get more data, more
data now not meaning more  concentration data, but soil bulk  density or soil porosity,
soil structure, organic matter content, those parameters we just talked about.

Therefore the information requirements, then, in terms of the waste and the site and
looking at the interaction, is what we're trying to get.  The site waste interaction
characterization.  I started off saying: Let's back up and look at soil characterization
to predict or to assess the ground-water migration pathway.

We're at the end of our discussion now. What did we mean by that slide that said
soil characterization? We're talking  about transformation, both biotic and abiotic.
The rates of transformation and the extent. We need to know the rate  and extent in
order to see if the chemical's going to make it to the ground water or be destroyed
before it gets to the ground water.

The contamination profile: what is the profile concentration with depth in the soil?
So we know where we begin. Where does the chemical begin to migrate from? Two
feet, five feet, ten feet in the soil?  Then the distribution we talked about, the
distribution between soil  and air, soil and  water, soil and NAPL, because we need
that information to focus in on the water migration pathway. So this is  what we
mean by site characterization after treatment  in order to evaluate the migration in
the ground-water pathway.

What's it all about?  After all this discussion it's been about those processes-taking
those processes and trying to do a mass balance in the subsurface. And why are we
doing that in the subsurface? Taking an approach based on a chemical mass
balance?  That's what it's all about.  And using  a methodology that's wrapped around
subsurface processes.  That is distribution,  degradation, and soil properties.  Using
those simple words.  So when we finish the discussion just think mass balance
approach, subsurface processes.

At the top of the slide it  says we're doing  this all for protection of human health and
the environment. And the treatment of waste constituents to an acceptable level.
                                       21

-------
What we're trying to figure out is:  What is that acceptable level?  That's the goal on
top of the slide.  Protection of the public health and the environment. Treatment of
waste to an acceptable level.  We need to look at the ground-water migration
pathway, the surface water, and the atmosphere.  We're zeroing in on the ground-
water migration pathway today, but we said that you need to watch your surface
water because that will affect the ground-water migration.  Runon and runoff will
affect the downward transport into the ground water and atmosphere also.  Now let's
go down into the soil.

Atmosphere in the soil can be a  source of contamination to water going through the
soil. We may have trichloroethylene volatilizing into the soil atmosphere.  A rain
event dissolves the TCE into the water and back out into the ground water.  So we
can say, then, that  the surface water and soil atmosphere can be related right back to
the ground-water pathway.  Going down to the bottom of the slide; we're talking
about the soil system today. So it all goes back to the soil system, and within the soil
system all the subsurface processes we have talked about can be summarized as
degradation, transformation, and immobilization or retardation.

So we're talking subsurface processes, degradation, transportation, and mobilization
in the soil system, zeroing in on the ground-water pathway in order to protect human
health and the environment, and in order to specifically, for this conference, come up
with specific acceptable soil cleanup levels.

That's it.  The last line. That's a lot of stuff, I know, and I threw a lot of stuff at you;
and I appreciate your patience and hope you can hang in there for the rest of the
day.

Alison, I'm going to give up and turn it over to you again.
                                       22

-------
                    PANEL 1
SITE-SPECIFIC APPLICATIONS OF FATE AND TRANSPORT
       MODELING AND SOILS REGULATIONS

-------
                              INTRODUCTION
                                 Joe Williams
MS. BARRY:  This group is going to consist of a number of site-specific applications
of fate-and-transport modeling or regulations.  The group will be introduced by Joe
Williams who is a soil scientist with the Kerr lab in Ada, Oklahoma.  He's got
extensive experience in site remediation and fate-and-transport modeling.

MR. WILLIAMS: Thank you.  What I would like to bring up first is the fact that we
need to consider the things that Dr. Sims brought up in his presentation as we're
looking at these presentations and listening to what they have been doing at their
particular sites  or their regions or their states.

One of the questions I'd like for you to consider is "Has enough data collection and
site characterization been performed to support the use of simplified methods?"  This
is an issue of looking at this water movement and soil, the partition between the
various phases that need to be considered, not only in just getting that particular  data
but in  understanding the uncertainty and the variability of that data.  One thing in
dealing with soils that you'll come across is there's a lot of variability involved both in
the areal wise and depth wise.

The second issue is:  "Have the processes present at the site been defined well
enough to allow their exclusion from consideration when selecting a method or model
for determination of cleanup levels?" In other words, when you are  designing your
conceptual model of your site and  understanding what processes are there,
recognizing them, can the data that  you have already taken substantiate leaving out
some of processes as you try to develop or use your mathematical model.  So you
have to have some justification for that.  Our experience with the looking at RIs and
FSs as they come through our office is that many times the regions decide to use a
particular method, but they don't justify why they're going to use that method.  It's
just one that they pulled off the shelf.

The third question will be:  "Are processes, such as fractured flow, being considered
which might facilitate greater releases to the ground water than would be predicted
by simple approaches?"  This also has to do with the  NAPLs Dr. Sims talked about
and the structure of the soil. Are we opening up pores when the soil dries out
enough that when it does rain it flushes right on through the system? Is the model
capable of handling that?

"Are remediation technologies available and feasible for achieving cleanup levels
determined by the simple approaches?"

And the fifth one: "Does the cost of cleaning up to levels determined by simple or
non site-specific methods exceed the cost of obtaining more definitive
                                       23

-------
characterization data?"  There is a question that Alison raised a little earlier.  When
do we need to say we've spent enough on data or on site characterization?  .\nd
coming from a research and scientific background I have a hard time with that one.
In this session the five individuals who have been working in this area in the regions
and in the state will tell us what methods they have been using. The first speaker
this morning is Jeff Rosenbloom who is with the  Hazardous Waste Enforcement
Division with Region 9 in San Francisco.
                                       24

-------
          PHOENIX--GOODYEAR AIRPORT SITE, ARIZONA
          VLEACH MODEL AND SITE CHARACTERIZATION
                              Jeff Rosenbloom
MR. ROSENBLOOM:  I had planned to give a talk with a more technical bent, but
after listening to some of the presentations and discussing some issues during the
break, I wanted to share some other thoughts with you.  I've been an RPM, Remedial
Project Manager, for six years and recently became a manager.  So although I'm still
tied to the technical discussions I now have a management perspective.  The thought
kept going through my mind of how interesting all this was, but the management
perspective is like a wet blanket  in trying to set cleanup levels.  In talking with my
staff on what technical issues they are working on such as, what data are being
collected, and what models are being considered, there are a couple of other
constraints that need to be discussed.

One constraint is resources and another is enforcement since, as much as we all want
to do excellent science, we have  to sell it in a real world of shrinking resources  and
enforcement first that means having the PRPs sign up to implement the remedies.

From a resource perspective you're looking at your staff that are working on multiple
sites.  In addition, the amount of time they can work on each of these sites is limited.
In addition, EPA headquarters is always presenting the situation of trying to do more
with less. They want to move more sites through the pipeline, move them through
much faster which means less time collecting data.  But from the RPM perspective,
they want to collect as much  data as possible, to the extent of removal by sampling.
If you sample enough the problem's gone. Besides a concern with resources, we must
include the program initiative of an enforcement perspective.

PRPs these days are becoming increasingly sophisticated. They pay for extremely
expensive consultants and they have a phenomenally strong network both technically
and legally.

In addition, a lot of companies are extremely  recession conscious.  They are
concerned with  the amount of money to clean up these sites. So although we may
spend all this time on models to  get exact cleanup levels set, if we can't sell it to
them on a cost-effective basis, it  means nothing since we'll have no settlements.

So we need to emphasize straightforward data collection requirements for PRPs so
it's easy for them to implement remedies.  It's easy for them to sell it to their
management. It's also easy for EPA to oversee. We may have an extremely
sophisticated model but if the data requirements are too expansive or too difficult to
replicate, it's going to be difficult to implement in an enforcement world.  When you
look to implement this remedy, the companies are going to be concerned with how
much it costs and how long it will take, and EPA will also be concerned with seeing
                                     25

-------
how effective the remedy is which means how easy is it to go out back in the field
and find out if this remedy is working.

You look at all these issues from a management perspective and try to determine
what this clean up is actually going to entail and what are the cleanup levels?  Once
we get the levels, we need to determine how do you go out and verify that the site is
actually clean, which is the conversation I had with an RPM recently.  Is it when one
sample is left dirty? Is it one area? No one's gotten to that point.  That's one you
have to consider also.

One other thing that PRPs look at, which is one reason we're here, is consistency.  I
can't tell you how many times I've been in a situation where a PRP like General
Electric, General Motors, or Motorola says "Well, I understand what you're doing
here, but they don't do that in Region 5,"  or "In Region 6 we got a great deal." They
are very sophisticated  knowing what's going on across the country,  and as much as
within the regions we  try to network, we're so overloaded we have  a hard time doing
that.

Last thing we need to look at is from a communications perspective: "Is the public
going to accept this?"  Once again you're looking at technical  details. We're all
extremely advanced and we've been doing this for a number of years:  if you stand up
in front of a public meeting and say "We've used  this technical model and we figured
out the  residual contaminants of this." You start  talking  about Koc values you're
going to get blank stares. Someone's going to stand in the back room and scream at
you "I want that out of my backyard now.   I want it down to background."  'Well, we
can't get to background because of the following  reasons" and this person is seeing
red and they're not going to have anything to do with it.  So it is important to  realize
when you're presenting this information to keep it extremely simple.

And speaking of keeping it extremely simple, after all is said and done you have to
be able to sell this to your managers. I remember about three years ago I briefed my
branch chief about the subject of soil cleanup levels. I said "Listen, it would be fairly
detailed. I'm going to do a diagram of the soil and show different transport
mechanisms."  So we went through this whole long discussion of  "Okay, there's
evaporation, there's transformation, there's advection." We got to the end he said
"Well, that's very interesting, but now what's the soil concentration going to be?  Can
I get a number out of this?" That's what we have to deal with. Oftentimes if  we
don't come up with a  number it's difficult to sell.

Given that background I want to spend a  couple  of minutes with the site we've
worked on in Arizona that I had some excellent success with setting cleanup
standards.

The site is in Arizona. Disposal of solvents was through sewers, dry wells, and
evaporation ponds.  One method of getting rid of their solvents was taking waste
                                       26

-------
solvents and sending them down unlined canals and igniting them at night to burn
them off.  I suppose it was quite dramatic. They now have ground water
contamination stretching two and a half, three miles from the facility.  Soil levels
around 1,000 to 2,000 micrograms per kilogram of soil.  They have the usual soup of
VOCs, TCE,  and TCA.

Since this is an enforcement lead site the company is doing a lot of the work.  They
had to do a lot of soil borings and started to complain of the cost and  the need for
that kind of data  collection.  We started relying more on soil gas which is a technique
used at a lot  of the sites which involve organic compounds.

Soil gas is extremely effective for areas where you're unsure of the exact location of
the disposal over a large area. It involves collection of gas in the shallow soils at
about five feet. Basically, you drive a probe in the ground from the  back of a jeep.
So we're able to get a lot into usually inaccessible areas.  It's extremely fast.  We can
collect up to  20 samples a day at about $200 to $300 a sample.  And there's an
on-site mobile GC which we kept in a trailer nearby so we just cranked for two or
three weeks and covered the entire site, went back after we located sources, and did
some more detailed surface soil investigation.

What we found with this was that it did a phenomenal job at locating subsurface
residuals of the VOC.  It was tremendous.  So we got an aerial picture of where
sources were. We thought we needed to get a perspective on concentrations at
depth.

Once again we're faced with the problem of soil borings being really expensive.  We
went ahead with soil borings, but someone recommended that we don't back fill the
soil borings.  Instead, put in some PVC casing and  put in soil vapor probes at depth,
which turned out to be a wonderful recommendation.  It is about 80 feet to ground
water in alluvium. Basically, we found that by running a pilot study using these soil
vapor wells as extraction wells, they could draw a tremendous radius of influence in
the area.  So what we had was the aerial extent VOC contamination and then VOC
soil contamination at depth.
One of the slides I wish I had was a comparison of the volume of VOC's that are on
the soil matrix versus soil gas.  We had two orders of magnitude greater
concentrations and mass in the gas rather than on the soils itself. So we quickly
surmised that we had a  far greater mass in gas than on the soil matrix itself, which
lends itself very easily to collecting data quickly because you can do it in the field and
avoid some uncertainty because when you collect soils there's the handling problem.

After we got the soil gas information, we started dealing with cleanup levels, and
what we finally determined was that we had to look at a shallow cleanup level and a
deep cleanup level.  With the shallow level we were concerned with direct contact.
                                       27

-------
Were any workers exposed?  Was there any of the public that could be exposed? We
looked at those concentrations and were able to compare them to state standards, a
fairly straightforward process.  Where we got hung up was on deep soils, and what we
determined was that it was necessary to look at what was the affect of the
contaminants on ground water. And as was mentioned before we had to look at how
much contaminants were in the soil and what the transport was like.

We worked with a technical committee to help determine the soil cleanup levels.
The responsible parties company sat on that committee and they are part of the
process, and we found that was crucial  because we're now dealing with sites where
we've taken this process of doing soil gas at depth and looking at massive
contaminants and soil gas, and companies that  aren't up to speed on that are totally
blown away by what you're doing.

We looked at soil gas concentrations and other information and decided on a
transport model known as VLEACH, a model that was originally developed by
CH2M HILL.  Once again the committee decided that that was one of the better
transport models to use. It looked at VOC's in the soil and looked at transport to
ground water.  It was very interesting to use that model because we found it was
useful when we tried to justify this to the companies.  We said "Okay, you have a
choice.  You can pump and treat the ground water until you're blue in the face, or
you can remove the  source."  Their perspective was "What's the cost benefit here?
What am I getting for my buck?"  We did a quick run with the model and looked at
two scenarios: no source control "Let it all go, you'll eventually pump and treat it
and get it that way," and we looked at the cost  per pound of TCE removed and the
value was about $3,000 to $6,000 for pumping about 2,000 GPM  for 80, 100 years.

We then looked at source control: If you remove the source what was the benefit?
The cost difference was astounding.  Soil extraction cost about $30 to $80 per pound
of TCE removed.  Extremely cost efficient; small, skid mounted pads of carbon and
vapor extraction systems. It also cut the cleanup time on the order of a hundred
years.  I realize these numbers don't mean a whole lot because they're model results.
We don't have any site implementation factors  yet, but they're dramatic enough to
convince the PRPs that it's worthwhile  doing source control.

What we also found was that once we had the model that everyone agreed on, "Okay,
yes, we'll use this model, we'll look at the mass of contaminants and soils and soil gas
to see if there's any  effect on ground water," we needed a  decision tree. What
everyone needs to remember is you have these cleanup standards that have to be put
into a consent decree.  PRPs are going to sign  on the bottom line,  "Okay, we agree
we're going to clean this site up.  We're going to use this method." They're also
going to be looking very closely at a clause for  termination. The  CEO of the
company's going to say "When do I walk  away from this site?"
                                      28

-------
So once again we use the committee process to work out a decision tree, which is
very straightforward way of looking at using this model and the depth of the specific
soil gas data and say, "here's the residual mass of contamination.  Will it have an
adverse effect on ground-water quality?"  If so, we continue doing soil vapor
extraction.  It sounds very straightforward, but it needs to be put down on paper.

So by using this method, everyone was pleased because it's extremely inexpensive to
collect soil  gas data either from the surface or from depth. You have existing ports.
It's very easy for someone to go back and collect another sample, or to collect splits.
It's not like you're constantly going out collecting soil data.  This is especially
important because a lot of the sources were under buildings, and  they were not
thrilled with going and jack hammering through basements. The companies finally
realized that it was cost effective, which is something else you need to point out.

So in closing we'll  be getting some information out on more detail on the  VLEACH
model, but  the message here with  VOC's is to look at soil gas data extremely closely.
Lots of people dismiss it as not being a tremendous reservoir of VOC contamination.
And the other message is to get your PRPs on board, and if  you're working with
companies  try to convince them of the effectiveness.  Thanks a lot.

MR. WILLIAMS:  All right. Thank you, Jeff. Our next speaker is Dave Kargbo.
He's from the Hazardous Waste Management Division, Region 3 in Philadelphia.

A SPECTATOR:  While they're setting up could we ask you a couple quick
questions?  I never did hear exactly what overall remedial action was proposed  at the
site. Is it simply soil vapor extraction?

MR. ROSENBLOOM: We found that about 85 to 90 percent of the VOC's left in
the vadose  zone were in the soil gas and  found  that through a pilot study we moved a
tremendous volume.  So it was straightforward soil vapor extraction.  It was 80 feet to
ground water. So we had three or four extraction points at depth, and we monitored
the perimeter with a pneumatic piezometer to see how far the reach was of the vapor
extraction.

A SPECTATOR:  So in the abstract it says remedial action is halted to see if soil
concentrations rebound. Should that say soil gas concentrations rebound?

MR. ROSENBLOOM:  Yes. Thank you.

THE SPECTATOR: So there was no  actual contaminated soil cleanup goal ever set.

MR. ROSENBLOOM:  No.  There were contaminants detected actually on soil
matrk, but  it was such a small percentage we figured it wasn't worth being concerned
with.
                                      29

-------
A SPECTATOR: What was the time frame for VLEACH? How long do you have
to extract it?

MR. ROSENBLOOM:  The pilot study was in effect for about a couple of weeks.
We extracted upwards of 2,000 pounds of TCE.  We haven't started the final remedy
yet. We think it will take about three or four years to do the complete removal.
                                    30

-------
                     SUPERFUND  SITE, REGION III
        SITE CHARACTERIZATION AND DEVELOPMENT OF
                         SOIL CLEANUP LEVELS
                          David M.  Kargbo, Ph.D.
MR. KARGBO: My topic of discussion today is Soil Cleanup Goals Determination
at a Superfund Site in Region III.  This determination is part of an on-going RI/FS
study.

Before I begin, I want to first of all acknowledge Kathy Davies of Region III for her
significant input into the preparation of this presentation.

OBJECTIVE

The objective of the exercise was to calculate cleanup goals (maximum chemical
concentrations)  in soils that will prevent unacceptable risk to  human health via
pertinent current and/or future use pathways.

THE PROCESS

I shall now talk about the procedures used in calculating soil  cleanup levels at this
site. The first process involved the selection of the chemicals of concern.  Secondly,
exposure pathways were identified.  Finally, ARAR-based and health-based soil
cleanup levels were developed.

In this exercise, it was first of all necessary to identify the characteristics of the site.

The soils developed from colluvial parent material, 0-15 feet thick.  They grade from
silt loam (surficial soils) to clay (subsurface).  Below the subsurface  is thick saprolite
that was formed in place. The upper layer is mottled silty clay with the clay content
decreasing with depth.  Under the  saprolite is weathered rock and competent
bedrock.

To better illustrate characteristics of Site X, I developed a black box of the site
(Figure  1).  Ground water at the site occurs in the saprolite and bedrock.  Flow in
the saprolite is controlled by primary porosity while in the weathered rock  and
bedrock, ground water is limited in the interstitial spaces associated with well
developed fractures. It was determined that attempting to use the equivalent porous
medium approach to estimate flow (and hence contaminant dilution) in the fractured
bedrock would be inappropriate. Consequently, the saprolite  aquifer was modeled as
the aquifer of concern although drinking water was obtained from the bedrock
aquifer.
                                      31

-------
  u

*
o

                                  \
                  -  - -1
                        o
                        09
                                \
                                   \
                                       \-
                                        \   .
                                       •  •  \ •'
                                               \
£
                                                       \
                                                                           E
                                                                           O
 o


-------
CHEMICAL SELECTION

The chemicals of concern were subdivided into carcinogens and non-carcinogens.
The carcinogens were benzene, trichloroethylene (TCE), tetrachloroethylene (PCE),
methylene chloride, and arsenic. The non-carcinogens include cyanide,
chlorobenzene, PAH (naphthalene) and naphthalene acetic acid.

EXPOSURE PATHWAYS

In identifying exposure pathways, a couple of criteria had to be met by each pathway.
A pathway has to be a contaminant source, chemical release mechanism; medium of
environmental transport; point of potential human contact; and route of exposure.
Identified exposure pathways based on the above criteria were:

      •      Ingestion of contaminated ground water
      •      Direct contact of contaminated soils

My focus today will be the ingestion of contaminated ground water.

DEVELOPING SOIL CLEANUP LEVELS

For the development of ARAR-based and health-based soil cleanup levels, based on
the ingestion  of contaminated ground water, the mass-balance  approach which most
of us know as the Summers method was utilized. This approach estimates the
chemical concentration in the infiltrate  (Cw) at the saturated-unsaturated  zone
interface, which on mixing with ground  water beneath would result in the  ARAR-
based or health-based concentration of the chemical in ground water (Cgw). It is
expressed as:


                                =  Cgw (Qw + Qgw)                        (1)
                                        Qw
where

      Qw         volumetric flow rate (ft3/day) of recharge from infiltration into
                  ground water

      Qgw        volumetric flow rate (ft3/day) of ground water

Qw is the product of the area of the site (Ansite) and infiltration (I), expressed as:


I is calculated by the water balance equation.
                                     32

-------
                                                                            (2)
                                  Qw = Asite*!
      Qgw is obtained from Darcy's Law, expressed as:


                                                                            (3)
                               Qgw = A*K*(dh/dt)



where

      A           cross sectional area (ft2) of the saprolite aquifer

      A           W*b

             With W being the width of contaminated soils perpendicular to flow
             and b the thickness of the saprolite aquifer

      K           aquifer hydraulic conductivity (ft/day)

      dh/dl       hydraulic gradient (ft/ft)

The derived Cw would then be related to the soil cleanup goal (Cs) by the soil-water
partition coefficient, Kd,  expressed as:


                                                                            (4)
                                 CS  = Cw  * Kd
The derivation of Cw however requires values for Qw Qgw and Cgw (Equation 1).
Directions of water flow for calculations of Qw and Qgw are indicated in Figure 1.
Equations 2 and 3 were used to calculate Qw and Qgw, respectively.  ARARs (Tables
1 and 2) and health-based concentrations (for each carcinogen and non-carcinogen)
were used to represent concentrations in ground water that should not be exceeded
(Cgw).

For the organics, Kd was derived from literature values of Koc  (organic carbon
partition coefficient) and fraction of organic carbon (foe), expressed as:
                                       33

-------
CO
DC
<
DC
<
•D
03

•D
k *
^L
•D
0)
r+4
ro
£
— ^^
co
LLJ

•

0
.a
03


i
CO
" ^
c
0
0)
0
c
'o
*w
(0
o

1





^x
a
D
DC
<
^^^
DC
<
_J
o
•^•B
2
IO


3
CO
CO







_j
s
^
.<£
LJJ
I
O

0
C
s
c
a>
CO
o
^^
2
10


CO
cvi




0
C
0
•*T
0
O
jO
O
H
CO
0
1
o
•
o

_!.
^^
CO
CO



0
c
0
JIZ
0
o
_o

05
V-
•*-»
|2

<
z


*


0
;u
_o
f-
O

0
c
_0

0)
2
o
LO
o
CVJ
O)
 g
'c
 0
 CO
                    03

                   3

                   GC
                   <
                   DC
                   <
                    O
                   Z
                    • •
                   <
                     0)
                     •+•-
                     03
                     03
                     >
                     *D
                     cr
                     LU

                     CO

                     •a

-------
and ARARs

2
__
0
03
*~
CO
LU

CM
0)
.Q
•2

C
0)
O)
o
vx
c
'o
o
c
o
z
1




^
a
3
a:
QC
o
o
1
00
00

5
CO
CO





CO
— ••
o
^
LU
I
O
0
C
N
C
0
-Q
0
O
z

O)


+
0
C
^•^
0
CO
f—
Q.
CO

|
Z

o>





0 ^H
c 9
_0 <
co o
+•* /•%
r- 0
QO
CO <
Z
z

o>


a>
•^
±i
C
O
0
o
_CO
•p
a.
co
Z
Z

o




+
CO
D
*^T
•a
r~
CO
1
OOMV 003

o
o
M

c
0)
03
+•*
03
O
1

0)
E
CO
CO
•o
X.
1
o

-------
                                                                            (5)
                                 Kd = Koc * foe
foe was estimated at 10 percent from soils analysis.

For the inorganics, a shake test was carried out using a sample of onsite soil and
clean ground water at different soil-solution ratios.  After a period of time, the
distribution of contaminants in the soil and in the water was determined and Kd
expressed as:


                           v.     Concentration in soil
                           Ka
                                 Concentration in water
I consider the procedure for the determination of kd for inorganics as one that
measures desorption coefficients rather than adsorption coefficients given the fact
that the contaminants are desorbing from the soil into the water.  The assumption
here is that the desorption rate is the same as the adsorption rate. This assumption
will be discussed later.

As indicated  before, the product of Kd and Cw for each chemical gives Cs, the soil
cleanup goal.  Calculated Cs for carcinogens and non-carcinogens are as listed in
Tables 3 and 4, respectively, for both  ARAR-based and health-based ground-water
concentrations. For health-based, Cs was calculated at different risk levels  ranging
from l.OE-3 to l.OE-6. I have only  indicated the l.OE-6 Cs values.

DISCUSSIONS

I would like to spend some time now discussing the results, especially the method
used to generate such results.

Advantages of the Method

The advantages of the method include:  easy to use; simple data requirements;
applicable as a screening tool in comparing relative severity of sites; and may be
modified to suit site conditions.

Modification of the use of the method at this  site was related to the estimation of
aquifer properties.  The saprolite aquifer was used in the model even though the
bedrock aquifer was the aquifer of concern, supplying water for consumption.  This
modification was necessary because contaminants will  have to traverse the entire
                                       34

-------
E
a
Q.
c
^
O
"D
0
CO
05
.0
1
"05
I
*




CM
0
o"




CO
o
o"




10
o
o"




^
< °
Z 0
CO
Sg
o S
O)
Q. 0
D .£
C 0
$ «
^0
o ,

— •
'o
CO

CO
^)
03
<"~
T3
0
CO
05
JD
I
cc
CE
«<

CO
0
o"






i
O
LU
I
O

0
1
0
CD

CO
o"




0
c
0
™3
0
O
,0
H

h-
O
d


0
c
0
0
o

o
05
| —

<
z

CD
(•J
V
_O
6

0
c
_0
"0
2

q
cvi









'c
0
CO
^
 0)
+*
 CO
 E
'•^
 V)
LU
 co
ir
CO
LU
*J
 CO
  •
 o
 c
 o
o
 E
 o
LL
O

-------
E
Q.
Q.
•£
^^
03
O
1
3§
O CD
OO>
0
f—
§i
C CO
cd O
fl)
— C
73
0
CO
03
^ w
_Q
I
"03
0
I


O
O
to q

0
C
"co
^T
f-
CL
CO
          LU


          O   O
                            CO Q
          ^   p    e    3=
                            CO
                                     O
                                     O    O
                                                 o
                                                 o
                                                 CO
                                                 o
                                                 CO
                                                 CD
+
J0
±E
c
o
0
O
5
£
Q.
CO
Z

+
J—
2
D
2
|
CO
^-.
.2





0
'c
co
>
0
0)



_0


O
4->
o
Z
 • •
o
                                                           o
                                                           a
                                                           E
                                                           o
                                                           O

                                                           •o
                                                           0)
                                                           c
                                                           a>
                                                           •a

-------
thickness of the saprolite aquifer (which are assumed to be completely mixed) prior
to entry into the bedrock aquifer (Figure 1).

While the use of the entire thickness of the aquifer (required by the Summers model)
may not be  appropriate for some sites (e.g., sites with low density contaminants and
partially penetrating wells), the entire thickness of the saprolite aquifer was used
appropriately at this site.
Disadvantages

I would like to  talk now about the disadvantages of the method used at this site to
calculate soil cleanup goals.

1     Assumptions of the Method

(a)   Contaminants  are completely mixed throughout the entire thickness of the
      aquifer

As indicated before, this assumption was considered to be met given the site
conditions of Site X.  It was necessary to use the saprolite aquifer for modeling rather
than the bedrock aquifer since use of the equivalent porous medium approach was
inappropriate for the  bedrock fractures. Contaminants had to pass through the
saprolite aquifer to get to the bedrock aquifer and hence the assumption of complete
mixing in the saprolite aquifer.  This is however not the case in most of the sites that
we deal with in Region III.  Consequently,  the inherent assumption is that ARAR-
based and health-based risk levels in ground water must be met in the saprolite
aquifer in order for these levels to be met in the bedrock  aquifer. This may be  a
conservative, but acceptable and inevitable approach.

(b)   Flow of water is controlled by the Darcy velocity

Darcy velocity,  as we  all know, is volumetric flow rate per unit area. For Darcy
velocity,  the full cross-sectional area of flow considers both the solids and voids rather
than the true velocity (also called the average linear velocity) where the flow is
considered to pass only through voids.  Consequently, porosity (more appropriately
effective porosity) should be used in calculating velocity.

The  true volumetric ground-water flow rate (Qgwt) is related to  Qgw as:

                                  Qgwt   Qgw/n

where

      n           (effective) aquifer porosity (a fraction)
                                       35

-------
Consequently, Qgwt  > Qgw and the true Cw (Cwt) is greater than Cw calculated
from Equation 1. This therefore underestimates Cs from Equation 4. This again
provides conservative estimates of soil cleanup goal.
(c)     Equilibrium is assumed attained between the soil contaminants and
       contaminants in soil solution

Again this is a common assumption for a lot of soil cleanup goals calculation
methods, including this method. For the majority of contaminants, however,
attainment of equilibrium requires several days for the infiltrating water to be in
contact with the soil contaminants which does not usually occur in soils. Hence when
breakthrough studies are conducted, we find a tailing towards the right (desorption
portion of curve) indicating that adsorption rates (Kd)  and desorption rates (Kdes)
are not always the same even for the same chemical.  Hence, Ed >  Kdes. Literature
Kd values frequently provide no information on Kdes.  Given that the potential for
contaminants to desorb from solid mass into soil solution is directly related to their
potential to contaminate ground water, values of Kdes, rather than Kd should be
used in assessing soil cleanup goals.  Kd (rather than Kdes) for organics was used at
Site X and was estimated using literature values of Koc.

(d)    Amount  of water infiltrating into ground water is precipitation (P) less
       evapotranspiration (E) and runoff (R)  i e  I = P - [E + R]

I alluded to this assumption earlier on.  The objective is to obtain estimates  on how
much water percolates through soil layer(s) prior to entry into ground water.
Percolation through  soil layers differ according  to the physical, chemical, and
morphological properties of the layers. The soils at Site X differ with depth and can
be considered stratified.  This would therefore affect that rate at which water and
contaminants move down the profile into ground water. Effects of different  soil
layers were ignored in our modeling efforts.

The calculations for Site X assume that I is the "free water" (Ifw) that moves to
ground water following rain storms. It ignores the fact that following any rain storm,
water moves down the profile as free water only after the matrix forces (hygroscopic
water) and capillary forces (capillary water) have been satisfied. Consequently, in
most instances,  I > Ifw.  Consequently, infiltration into ground water was
overestimated (and hence Qe overestimated) by our method.

In the absence of field data, calculation of cumulative infiltration as a function of
measured soil moisture release characteristics, wetting front movement with time, and
saturated hydraulic conductivity (estimating gravitational component of flow) using
Phillip's equation is suggested.  This would allow separation of the vertical from the
horizontal component of infiltration.
                                       36

-------
2     Data Collection/Estimation Problems

The parameters used in the soil cleanup goals calculations are expected to be used in
most methods of calculating cleanup goals for soils.  The problems that will  be
discussed are therefore not unique to the method under discussion.
2 1   Input data

The parameters that are either measured or calculated include:

      •      ARAR-based or health-based chemical concentration is ground water
             (Cgw)

      •      Volumetric flow rate of recharge from infiltration into ground water
             (Qw)

      •      Volumetric flow rate of ground water (Qgw)

      •      Chemical concentration in the infiltrate (Cw) that corresponds to Cgw,
             given a dilution represented by (Qw + Qgw)

      •      Aquifer hydraulic conductivity (K)

      •      Chemical adsorption coefficient (Kd)

      •      Concentration of chemical in soil (Cs) at equilibrium with Cw

(a)   Chemical concentration in ground water

The MCLs used for ground water are "appropriate and relevant" but not "applicable."
Some argue that MCLs are not necessarily health-based and therefore may differ
from strictly health-based calculations.

As regards Cgw and  the use of MCLs, it is my hope that before this conference is
over, we would have discussed the appropriateness of using MCLs for calculating soil
cleanup levels.  Also the use of health-based goals and their method of calculations
should also be addressed.

(b)   Distribution coefficient (Kd)

While Kd was derived on a site-specific basis for inorganics, Kd (as alluded to earlier
on) for the organics was from already established equations of Koc which may have
been developed for a different environmental setting than that represented by Site X.
                                       37

-------
Generally for organic adsorption, we would want to collect data on the fraction of
organic carbon (foe) and use it to calculate Kd.  However, if foe is below a critical
level (foe*), and the mineral surface is high, adsorption of organics on inorganic
mineral surfaces should be considered.  I would  therefore suggest that for each
organic chemical for which a cleanup goal is to be determined, given the soil surface
area (or soil texture), foe* should be calculated.  Given measured foe, this would
determine the appropriateness of using the Kd equation that does  not consider soil
surface area.

I would also like to emphasize the significance of foe and suggest depth discrete
evaluation of this parameter since it varies with  depth.

Organic carbon was determined at Site X.  However, only surface soil values were
determined  and the need to incorporate surface  soil area given the contaminants  of
concern and measured foe was not evaluated.

For inorganics, Kd  determination seemed to be  appropriate because a shake test  was
carried out at different soihsolution ratios. However,  by  removing the sample from
the soil and disturbing it by shaking with water, you change the structure of the soil.
This causes  masking or destroying preferred flow paths; and increasing available
surface area of the  soils. Consequently, Kd derived may  be overestimated.  Also, the
shake test data cannot be used to explain long-term field behavior  of contaminants.
It is suggested that  other methods (such as column studies, field studies, etc.) that do
not destroy soil structure be considered as alternatives.

As regards the organics, given the contaminants  at this site, it may have been more
appropriate to  generate site-specific Kd (or Kdes) data for organics.  However, there
was not enough time to collect site-specific Kd data for organics.

2 2     Facilitated Transport

Facilitated transport is a relatively new concept and I  would like to talk a little bit
about it. It  has to do with  the fact that movement of  contaminants is sometimes
enhanced as a result of soil characteristics and the nature of contamination.  For
example, the presence of several organic solvents, acting  as primary solvents, could
increase flow of contaminants such as PCBs.  Also colloidal particles (organic and
inorganic) that are part of the transporting stream in the  geologic medium could
carry with them contaminants that are normally  considered to be immobile based on
their Kd values. At Site X, bulk hydrocarbons were not known to  be present.

The effects of colloidal transport compared to the effects of improper well
construction were difficult to differentiate.  This was not  considered to be a serious
problem, especially for metals since the concentrations of filtered samples were quite
similar to those of unfiltered samples.
                                       38

-------
2 3    Flow in stratified soils

As I mentioned earlier on, flow in soils is a function of the type of soils and whether
the soils are stratified or not.  The shake test used for the inorganics and the use of
Koc equations to estimate Kd do not account for the effects  of the different soil
horizons in  Site X.

3      Other Problems

Other significant  reactions that could have affected transport and fate of
contaminants  but not accounted for by the method used at Site X include:
biodegradation; dispersion; precipitation; redox reactions; volatilization; etc. The
method also ignores the  effects of:  mixed wastes; residuals from wastes; additive and
synergistic cancer risks from chemicals.

Conclusions

In conclusion, I wish to indicate that the method used to calculate soil cleanup goals
at Site X is  a  simple model that may be useful as  a screening tool in comparing the
relative severity of sites with similar site characteristics.  Data requirements are not
extensive and therefore cost-effective.

The model was appropriately modified and judiciously applied at Site X to reflect site
conditions.  This  is especially so given the fact that there is currently no adequate
model to deal with flow  and transport in fractured media especially when such a
media cannot be  approximated by the representative pore method.
There are however potential drawbacks with the use of the model both at Site X and
other sites especially given the complexities of Superfund sites in Region III.

In general, the method, when used to calculate Cw into the aquifer, does not
adequately consider soil  characteristics that affect Cw. These include:  presence of
diagnostic horizons and changes in these horizons with depth; porosity and pore  size
distribution; bulk density; effects of diffusion and hydrodynamic dispersion  (mixing)
of contaminants;  reaction (pH) and potential changes of Cw  with pH changes
volatilization; and effects of soil microbial  activities.

The method neglects the potential presence of mixed wastes  (which we frequently
encounter in many of our sites), residuals from wastes, colloidal transport of
contaminants, and effects of soil reactions (hydrolysis, redox  reactions, co-
precipitation,  and biodegradation, etc.)  on  Kd. It does not adequately account for the
possibility of non-attainment of equilibrium.
                                        39

-------
                BEACON HEIGHTS  SITE, CONNECTICUT
                       USE OF SUMMERS MODEL
                                Richard Willey
MR. WILLIAMS:  Our next speaker is Dick Willey.  Dick Willey is with Region I in
Boston, in the Hazardous Waste Management Division. Dick has also been working
with the Summers model, so he has other comments related to that.

MR. WILLEY: I would like to provide you with an  overview of some of the
important issues relating to selecting soils remediation levels for protection of
ground-water quality in New England. But first, I would like to digress for a
moment.

My professional responsibilities include providing technical assistance on
hydrogeology for 18 Superfund Sites.  Five of these are scheduled for Records of
Decision (RODs) this year.  I bring this up only to indicate that Regional technical
support staff and site project managers (RPMs) have work loads that allow little time
to research various models used to establish soils clean-up goals.  Even with lighter
work loads individuals within the Regional offices may lack the appropriate
professional background to  do  this research.  This leads me to conclude that the
Regions need additional guidance and assistance in the proper selection and
application regarding the most appropriate methods  to determine soils clean-up goals.
Dr. Kargbo (the previous speaker) has provided a detailed description of the
Summers Model and its application to a weathered bedrock setting in Region III.
Region I also uses the Summers Model to estimate soils remediation levels, and while
the model description will not be reviewed here, aspects of its application to a site
will be discussed.

The site is the Beacon Heights Landfill located in Connecticut.  It is on a hillside
underlain by glacial till.  Bedrock is relatively shallow and it crops out at  a number of
nearby locations.  The depth to the water table is less than 10 feet. The entire
landfill is scheduled to receive  an "impermeable" cap. However, several locations
beyond the perimeter of the proposed cap contain contaminated soils.
Contamination was transported to these areas (shown as hachured areas in Figure 1)
by erosion, slope failure and mass movement of materials off the sides  of the landfill.
The contaminants of concern are mainly organic compounds.

Northeast and downslope of the landfill are lightly populated residential areas.
Public health concerns for residents of nearby areas include incidental and ingestion
of contaminated soils and the leaching of soil contaminants to ground water, hence to
domestic water wells. Risk assessments at this and other hazardous waste sites
indicated that the potential leaching of soil contaminants to ground water usually
                                      40

-------
dictates the level of soil clean up.  Therefore, maximum drinking water limits
(MCLs) or risk based limits drive the soil clean up levels in the unsaturated zone.

The Regions point of departure is that contaminated areas lying beyond the
perimeter of a landfill fall within the zone of compliance.  And that these areas must
be brought into compliance with State and Federal laws so as to protect public health
and the environment.

The Summers Model is used to back calculate from MCLs or risk-based levels  in
ground water to obtain appropriate soils clean up levels.  It is a conservative model
that is relatively simple to use, and it does provide target clean up values that are
considered to be  protective.  Major site data requirements include:
             estimate of the quantity of water moving vertically through the
             unsaturated zone and laterally through the saturated zone,

             thickness of water saturated aquifer,

             partitioning coefficients for contaminants of concern,

             estimates of volume and concentration of contaminants.

Estimates of the quantity of water (ground-water flux)  moving laterally through the
system is obtained from monitoring well data. Hydraulic conductivity is estimated
from slug test data and hydraulic gradients are obtained from water levels in wells
using the 3 point method.  Water flux to the saturated zone (recharge) may be
obtained using the HELP model or U. S. Geological Survey reports on local ground-
water hydrology.  Recharge estimates should represent average annual conditions.

The saturated aquifer thickness is also derived from subsurface exploration data.
Saturated thicknesses in New England are seldom in excess of 100 feet. The need to
subdivide the saturated thickness based on the presence of low permeability units
requires site specific data.  However, our experience is that low permeability units are
usually not laterally extensive nor do they effectively prohibit downward migration of
contaminants (this also applies to the unsaturated zone).

The choice of partitioning coefficients for contaminants of concern are based on site
specific data.  Analysis of organic carbon content from representative materials
samples are used most often.  Frequently the resulting Foe values  are  a tenth of one
percent (0.001) and where  no data exist this is used as a default value. The use of
batch tests to determine  partitioning coefficients directly is also encouraged.

The bottom line of these data gathering and modeling activities  is to obtain a
defensible soils clean-up level. The selected soils clean-up level largely influences the
magnitude of the unsaturated zone clean up. For Region I this means the extent
(lateral and vertical) of soils  excavation, of the dimensions of a vapor extraction
                                       41

-------
system array. The consequences of selecting an inappropriate soils clean-up level is
either extraction of contaminants to a residual concentration that is not necessary
(level set too low) or lengthening the time required to pump and treat contamination
in the saturated zone (level set too high).  Unfortunately, there are no data in our
Region or any other Region that can be used to critically evaluate the "rightness" of a
particular soil clean-up level. All that can be said about the Summers Model or any
other method is that they are conceptually correct and they appear to account for
varying ranges of physical, chemical and/or biological processes.

From the foregoing discussion and other experiences I suggest that the following
issues be addressed:

1)     Evaluate current modeling methods used in determining soils clean-up levels
       to select a limited number of preferred methods. This list of preferred
       methods would then be periodically updated.

2)     Evaluate minimum data requirements needed to  establish soils clean-up levels.

3)     Track soils clean-up and subsequent ground-water remediation at selected sites
       to better understand the practical effects of unsaturated zone remediation on
       that for the saturated zone.
                                       42

-------

-------
       REGION II AMERICAN THERMOSTAT SITE, NEW YORK
                     USE OF MULTIMEDIA MODEL
                              Christos Tsiamis

INTRODUCTION

This presentation discusses the use of EPA's Multimedia Model in developing soil
clean-up levels for volatile organics at the American Thermostat Superfund site in
Greene County, New York.

In the American Thermostat plant's operations, organic solvents, such as
tetrachloroethylene (PCE) and trichloroethylene (TCE), were used within the
manufacturing process to clean the plant machinery. During the 1960's and 1970's,
waste PCE and TCE were poured down drains inside the building and were dumped
outside on the plant grounds for dust control.  Volatile organic contamination in the
soil is concentrated at only one location in the southwestern corner of the site (Figure
2).  It is estimated that the area  of contamination is approximately 30,000 square feet.
In the case of the shallow ground water aquifer, the contamination plume extends
over, approximately 26 acres in a northwesterly direction from the site at an average
depth of 50 feet (Figure  3).  In the bedrock,  the contaminated plume extends over
53 acres.  The maximum bedrock PCE concentration was detected at the residential
well adjacent to the  site  (Figure  1) and, during the  1989 remedial investigation
sampling,  was found to be 31,000 ug/1.

A risk  assessment conducted during the remedial investigation concluded that the
carcinogenic risks associated with the site, for both  the ground water and the soil,
exceeded the EPA acceptable levels. Remediation  for the ground water and the
contaminated soil at the  site was deemed necessary. In the case of the  soil, it was
calculated that in order to achieve an excess  carcinogenic risk  of 10 the soil would
have to be remediated to the following levels for the contaminant volatile organics
(VOCs):

             86.5 mg/kg for TCE
             18.6 mg/kg for PCE

If we cleaned up the  soil down to these levels, we would be protective of human
health.  At the same  time, we would be treating the ground water so as to  achieve
federal and state standards.  However, although protective, these levels could
represent a significant source of  recharge for the aquifer underlying the site
contributing to the increase  of contaminant concentrations above the ground-water
standards  and, in effect, compromising the ground-water remediation.

We, therefore, had to resort to some means of estimating the maximum concentration
of contaminants that would be allowed to remain in the soil  after clean-up so that
any leaching would not cause the ground-water concentrations  to exceed the ground-
                                      43

-------
water standards. We eventually selected the MULTIMED model developed by
EPA's Athens, Georgia, Laboratory to calculate the soil clean-up levels as it is
described below.

The MULTIMED model uses analytical and semi-analytical methods to solve the
mathematical equations that describe flow and transport through different media.  Of
particular interest to us in utilizing the model was flow through the unsaturated zone
and through the saturated zone underlying the American Thermostat site.

When the contamination in the soil is located above the water table, leachale can
migrate through the unsaturated zone and into the saturated aquifer. This process is
depicted schematically  in Figure 5.

The equations presented in Figure 6 are the governing equations for transport
through the unsaturated and the saturated zones.

Notice that the transport of the contaminants in the unsaturated zone is treated as a
one-dimensional problem with transport occurring only in the vertical direction. Fate
and transport mechanisms considered include dispersion, linear adsorption,  and first
order decay of the contaminant.

The equation for the saturated zone includes three-dimensional  dispersion, advection
in the transverse direction as well as first order decay.  In terms of boundary
conditions  at the source, the model allows you to specify the contaminant
concentration as a gaussian distribution in the  lateral direction (Figure 7) arid
uniform over the vertical mixing domain. A second alternative allowed by the model
is to assume a rectangular patch source of certain thickness and width.

Other boundary conditions specify that the background concentrations of the
contaminants in the aquifer are zero and that the concentration of  the contaminant
within the mixing zone is uniform (zero flux at z 0, b 0).

In addition, it  is important that the principle of conservation of mass is satisfied in
the transport calculations. The model can be used in the transient or steady-state
mode.

COMPUTATIONS

In using the model for the American Thermostat site,  we specified  the gaussian
distribution boundary condition.  In addition, we operated the model in the transient
mode.

The first 3  feet of the unsaturated zone, where most of the contamination was found,
is considered the source.  Contaminant transport was simulated  in the bottom 4 feet
of the unsaturated zone and in the saturated zone.
                                       44

-------
We assigned a leachate concentration of 1 mg/1 at the source and the model
computed a concentration at the specified receptor well distance.

A time series for concentrations at the well is calculated by the model.  For our
purposes, the maximum well concentration in the time series was used to back-
calculate the soil clean-up level.

As shown in this page (Figure 8), a linear factor called a dilution-attenuation factor
(DAF) is calculated next.

From the DAF we can then calculate the maximum allowable source leachate
concentration by assigning the ground-water standard as the acceptable concentration
at the receptor well.

To calculate the maximum allowable soil concentration (the clean-up level) we then
apply the partitioning equation.

RESULTS

This next table (Figure 9) presents the DAFs for PCE derived by the model. Notice
that  these values have been derived for transport along the center line of the plume
in the direction of the receptor  well.

Also, DAFs were calculated for average concentrations at the well over  a depth of 15
feet  and  for concentrations at the top of the aquifer.  In calculating the soil clean-up
levels we used the more conservative values associated with the top of the aquifer
concentrations. Similar results were obtained for TCE.

The  clean-up levels calculated from these dilution-attenuation factors have been
plotted as a function of the distance of the receptor well from the contamination
source and are shown in the following two graphs (Figures 10 and 11).

From these results, we adapted  for the American Thermostat site the  following clean-
up levels:

             for PCE 1 mg/kg
             for TCE 0.4 mg/kg

What were the model assumptions and how valid they prove to be in our case? The
major model assumptions are summarized in this page (Figure 12).

First, the saturated zone is assumed to be isotropic and homogeneous and the
ground-water flow is  assumed constant and uniform.  We felt that, given the short
distance of the receptor well from the source, perhaps there could be some validity to
these assumptions. Also, there is an assumption that adsorption follows  a linear
                                      45

-------
isotherm and there is local equilibrium at all times. Finally, mixing underneath the
unsaturated zone is assumed to be instantaneous and complete.  I would say that is a
very bold assumption.

To what extent did we use site-specific data?

The required data input to the model is listed in the following table (Figure 13). I
will go down the list and I will indicate the parameters for which we used site-specific
data and those for which values were assumed.

We had site-specific data as far as all the geometric parameters were involved, such
as the area of the source,  the thickness of the aquifer, the thickness of the
unsaturated soils.  We utilized precipitation data to estimate the infiltration rate at
the site and we assumed that the  recharge rate was equal to the infiltration rate. We
had  collected, during the remedial investigation, porosity data and carbon content
data at various depths at the site  and we also had hydraulic conductivity and
hydraulic gradient data from which we calculated the seepage velocity.  For
dispersivities, there were no data  available.  The only way, I am  aware of, to get
these data is by injecting dyes in the subsurface.  We did not want to do that at a
Superfund site. Instead, we calculated the dispersivities from a relationship that
existed in the literature, as a function of the distance.

Also, we would have liked to have had site-specific distribution coefficients.
However, we used the model after we had completed the RI and it would have been
quite expensive for us to go back to the site and collect additional data.  We used
Koc values from the literature and the carbon content values from site-specific data
to calculate the distribution coefficient.

REMEDIATION

How were the results obtained by the model used?

As you will recall, the clean-up levels for PCE and TCE were 1 and 0.4 mg/kg,
respectively.  You will also recall that the health-based levels were 18.6 and 86.5
mg/kg, respectively.

Upon comparing  these sets of numbers it was decided that we would be conservative
enough if we utilized the model derived levels to define the  areal extent of soils that
would require remediation.  Any  areas exhibiting soil contamination below the model
clean-up levels would be considered clean.

Secondly, these clean-up levels were compared to the levels  that could be achieved by
two  treatment technologies (incineration and low temperature  thermal extraction)
that were being considered for soil remediation at  the site.  It was concluded that,
based on these levels and the maximum contaminant concentrations found at the site,
                                       46

-------
the LTTE technology would be successful in treating the site soils at a much lower
cost (about $6 million lower) and without compromising the ground-water treatment.

It should be noted that the clean-up levels for the American Thermostat site derived
by use of the MULTIMED model were accepted by the New York State  and a
record of decision was signed in June 1990.

CONCLUSION

In conclusion, given the time constraints, the financial constraints, and the specifics of
this site, the MULTIMED model proved to be a useful  screening tool for arriving at
certain engineering judgments.  Our application  of the model for this site had less to
do with seeking an absolute answer to the question of what soil contaminant levels
would be protective of ground-water standards and more to do with gaining some
insight into an optimized system of combined ground-water and soil remediation.

In the near future we will attempt to verify these results in the field.

Finally, I would like to thank Terry Allison of the Athens Laboratory for  his valuable
help throughout this project.
                                      47

-------
                             FIGURES
1.  AT PLANT AND RESIDENTIAL WELL LOCATIONS




2.  SOIL CONTAMINATION




3.  CONTAMINATION IN SHALLOW AQUIFER




4.  AQUIFER CROSS SECTION




5.  SCHEMATIC OF LEACHATE MIGRATION




6.  GOVERNING EQUATIONS




7.  GAUSSIAN DISTRIBUTION DIAGRAM




8.  CALCULATION OF SOIL CLEANUP LEVEL




9.  DAF VALUES




10. PCE CLEANUP LEVELS GRAPH




11. TCE CLEANUP LEVELS GRAPH




12. MODEL ASSUMPTIONS




13. PRIMARY PARAMETERS USED

-------

-------

-------

-------
                                PLAN VIEW
                                     Contaminant RiumeB
                                SECTION VIEW
,57.
           Unsaturated Zone
                                                       Monitoring
                                                      -.Well           Ground Sur
                                                                      Water Tab
                                       Aquifer
                                                          B
(WW^W^^^
    Figure 3.1  A schematic of the waste facility and leachate migration
                through the unsaturated and saturated zones.

-------
              •KNUtl'1 MOI1VA111

                ?       I
                       '5
                      J±l
Is
!>
•|

f
5
r
cum 4»-j 1
:iKnssiciioN * *
• c
|



* t'Bl
            N *J
                  1!
                  I!
                           t§
                                     t    :i
I ji
III
I  £ -
I  . -
i  k c
\
z
i
I
•( n—i
] |
I 5 «
K 2 •
HI


-------
                GOVERNING EQUATIONS
UNSATURATED ZONE TRANSPORT MODULE
 R     _ D  aC  „  5C  .
            a?        "  v v
THE  SATURATED ZONE TRANSPORT MODULE
<52c      c2c  .    a2c   v  5c .    ac
   <         c    . n        v     . R       R A c   R

-------

-------
        CALCULATION OF SOIL CLEANUP LEVEL  ( Cs )
DAF = 1 / Concentration at the receptor  (mg/1)    ( 1 )


DAF = dilution - attenuation factor
Cw = DAF * groundwater standard              ( 2 )


Cw = leachate concentration at the source
Cs = Cw * Kd            ( 3 )
Cs =  cleanup soil concentration
Kd =  partition coefficient

-------
Table 3.
Dilution-Attenuation (DAFs) predicted by
Multimed for PCE at specified distances
downgradient from the contaminated soil zone.
Vertical location of receptor is at top of plume,
transverse location is on plume centerline.
Distance
         DAF
 (top, center of plume
in transverse direction)
        DAF
(average over 15 feet
depth, center of plume
in transverse direction]
100 ft
120 ft
140 ft
160 ft
180 ft
7
8
11
13
16
.2
.9
.0
.7
.8
7.2
9.4
11.8
14.6
18.0

-------
   CO

   v.
   0)

   "S
   C

   3

   O
2 .2

c °

g k

rco
O *- Q.
O -D ^
— 0)

      <*
 >  CD «
 CO — T3
CM
 G)

il
                        o
                        CO
                           o
                           o

                           T3  O
                           O  o
                           +* «^
                           o  o

                           1  s
                           O) =

                           18

                           ll
                       . O  SI
                                                                  o
                                                                  o  c
                                                                  c  o
                                                                  (0  O
                                                                  «••
                                                                  (A

                                                                  Q
                                                               O

                                                               CM
                       . O
                       - o
                  o
                  o
                  o
o
o
in
o
o
                            (qdd) |9Ae| dn -

-------
   03
  •o
   c
   3
   O
(0 Q


"52
O
O
asi
c?^ c

I Is
C3 — ""
CU

S


CO

0



O)
     Q.
     CO
                                                o
                                                00
                                                 o
                                                 o
                                                o O) £•»
                                                    -
                                                 o o

                                                 1 s
                                                 O) :=

                                                 I 8
                                                 E c


                                                §2 1
                                                T- _ <0
                                                 O -j-
                                                 o c
                                                 c o
                                                 (0 O
                                                 4-*
                                                 (A

                                                 Q
                                                o
                                                CM
                                                O
                                                o
                              §
                              to
                     (qdd) |9A9| dn -

-------



rH *
O =
TJ in
0 Q)
£ -P
in
in -P fd
fd C 12
3d) TJ
g in c
n in 3  in
(d in o H -H
in o
c H a Q) ^ in
O Q) X M 4-> -H
•H TJ W 3 M
-p o in fd M
fd g fd o a) a)
4J -H & C  C C
TJ Q) H fd fd 0)
£ S 3 i-3 n'
rd a; s «d o
4J = 0) M N
W id 43 -P i
C «P -P O TJ
Q O W g 43 
jg 4J tji M C fd fd
H a c w -H c/i M

>J 3 15 TJ fd -P
p in o o 3 •* fd
s in H N H H in
 QJ 0)
O H 


0)
43 (d
-P
in
^i »2
43 0
H
v i rpT ^^
^1 \^ i ^
0 J3 0

in
•H
\
O
in
(d
J3
04

TJ
(4-1 (d  fd fd
J3 4J
3 in in
0 C 3
0 0 O
0 -O O
fd c
in id
4J -H 4J
0 C
c >i fd
4J 4J
Q) -H in
M 0 C
fd O -H
H
in a) in
0) > -H
M
3 > C
4J 0 0
O rH -H
id  W

5 4J
in TJ c
3 C fd
0 3 • C
C 0 g -H
Q) • MM g
tTlH U> O fd
Oi Q) .
s.
M
a):
43 ',
4J
0 •
in g
•Hj 3
'••H
j3> M
0:43
•Hj-H
-P iH
Qj-H
M; 3
0, O1
in o)
TJ
(d rH
i fd
Mi 0
fd' o
0)irH
c
•H; c
H-H



TJ
0) 0)
4J C
fd o
C N
•H
g TJ
fd o
4J 4J
(3 fd
0 (3
U-H
C g
3 fd
4-)'
43 C
4J 0
•H O
jj i
0)
0*43
C4J
-H
Xi43i
•H 4->j •
g (di 0)
1 0) 4J
O C Q)
4-» M rH
o: a
Q) TJI g
3 CO
TJ 31 O

£3 ^i 'O
O^rH C
•H'4-) (d
4J'0
, 3 a) in
rH.M 3
-H'-H O
TJ TJ 0)
C
4J M fd
.C 0) 4J
fd. 4-^ C
Cj (d fd
•H1 J5 4J
STJ in
fd c c
4J 3-H
C. 0
a M in
Cjl O>«H

1

TJ
i TJ Q)
0 0 rH H 4->
•H H 3 fd
0) 4J id O C
CT» 3 U 43 -H
TJ JQ -H in g
0) -H 4J fd
MM H 4J
4J 4J 0) -H C
J3; W! > 00
o-Hi in u
•H 'oj e
TJ M 0) Q)
(d Ci O • 43 43
M fd  in 3 N -HO
0 3
TJ  in
' tJ»«H C C C
0) rH fd O
43 fd TJ ( X C 'H
4-» Ci-H -H W
w id g g c
4J > fd Q)
(0 O C (D 4J g
rH O 43 C -H
C rH-H 4J O TJi
O O 4J ' Oi
•H 
M C -H 43
4J O TJ,4J C TJ
c N a O' c
Q) (D Q) -H tf
Oi'TJ W TJ 4^1
C. Q) M 3, -
O'4->  43 . -H M
C IA 4-> M; O
1 4J -H C 4->' in,-H
fd  4J O , TJ 3
•H C 43
g O , 0) >n
fd U 43 3 43 H
4J 4-> O/' 4J M
Cj Q) Mi -H(
U 4J .H 4-> ' 0 «W=
in
1 H Q)

-------
TABLE  5-1.   PRIMARY PARAMETERS  USED  IN THE  SATURATED ZONE TRANSPORT MODI
Parameters
Units
Source-Specific  Parameters

      Area  of  the  land disposal  facility

      Leachate concentration at  the  waste  facility

      Either:
            Standard deviation of the source  (Gaussian)
      or
            Width of source  (Patch)

      Infiltration rate

      Recharge rate into  the plume

      Duration of  the  pulse

      Source decay constant

Aquifer-Specific Parameters

      Porosity

      Thickness  of the aquifer

      Thickness  of source

      Seepage  velocity

      Dispersivities (longitudinal,  transverse,
      v«r~ical)

      Retardation  coefficient

      Radial distance  from the site  to  the receptor

      Angle between the plume  center and the receptor

      Well vertical distance

      Time value at which concentration  is required

Chemical-Specific  Parameters
[m2]

[rng/1, g/nr


[m]

[m]

[m/yr]

[m/yr]

[yr]

[i/yr]
[cc/cc]

[m]

[m]

[m/yr]


[m]

[diT.er.cion'-

[m]

[degrees]

[fraction]

[yr]
      Effective first-order decay coefficient
[1/yr]

-------
    PENNSYLVANIA DER--LEGAL AND TECHNICAL APPROACH
                 TO SETTING SOIL CLEAN-UP LEVELS
                              David Crownover
MR. WILLIAMS: Our final presenter this morning will be David Crownover who is
with the Pennsylvania Department of Environmental Resources.

MR. CROWNOVER:  I just have some general comments. Some of them will tie
into some of the things that have already been mentioned this morning.  I want to
look at a couple of general issues, one of which  is the whole issue of the advantages
of a site-specific approach versus a standardized approach and an issue that Alison
raised this morning, which is:  How should the projections which are generated by the
models be used--as absolute clean up numbers or as a risk management tool that
feeds into the general process? I'll try to add some perspective  from the
Pennsylvania DER framework on this issue.

Dick Willey mentioned in his presentation  the importance of defensible clean-up
levels, and ultimately defensible clean-up levels are based on the underlying legal
authorities and essentially that means what can you enforce in court.  So I'd like to
talk some about the legal framework in Pennsylvania and the general technical
approach and how they tie in together.

Under Pennsylvania state law there's prohibition against disposal of waste without
DER permission. This is based on one of our general environmental statutes.  The
definition of waste disposal includes the release  of any hazardous substances.  This
general statutory provision, in  conjunction with other enforcement authorities,
provides DER with the authority to require responsible parties to clean up any
release to background  levels.

Pennsylvania state law also prohibits any release of contamination to the waters of
the Commonwealth,  including  the ground water.  This means that completed
responses must not allow a continuing release to the ground water. Therefore, in
Pennsylvania the soil-to-ground-water pathway is legally important, and it needs to be
factored into the Superfund remedies selection process.  These statutory requirements
relating to soil clean-up to background and ground-water protection essentially set up
a two-tier legal  framework.

First of all, contamination must be cleaned up to background concentrations.  If it is
not feasible to meet this first-tier requirement, then the selected remedy must assure
that there is not a continued release to ground water. If it is not feasible to meet this
second-tier requirement, then  the ground water  must be protected to the extent
feasible.  So that's our general legal statutory basis.
                                      48

-------
The feasibility determinations must be made at two tiers again, first regarding the
requirement to clean up to background and second regarding the requirement to
protect ground water.

For these purposes we define feasibility in a general way to include technical
feasibility, implementability, and economic feasibility, which includes both cost
effectiveness and absolute cost, and as mentioned earlier, costs are real in the real
world, and they can't be ignored.  I don't want to focus on this first-tier requirement
to clean up contamination to background concentrations.  Suffice to say that in the
case of many smaller-scale  clean-up projects it is feasible to remove all contamination
and thereby protect the ground water. But in the case of many larger scale Superfund
sites,  it is clearly not feasible to remove all contamination to background levels.  So it
really isn't an issue in many NPL-type sites.

In cases where it is not feasible to meet either the tier one or the tier two
requirement, then less stringent risk-based responses will be approved by DER, but
the responsible party retains all liability.  DER reserves the right to require
additional responses in the future, and DER will not  provide a release from liability
nor a covenant not to sue.  Thus our legal authority to require clean-up to
background legally places the burden on the responsible party, and thereby  provides
DER with broad  discretion to make technical decisions regarding clean-up issues. So
in some regard, our perspective in Pennsylvania is based on our statutory authority,
since  our legal framework gives us a strong support to move  off of background
without having to have a standard model  that we are going to need to try to enforce
in court or prove in court.

One note on this is that, while our technical approach fits very nicely with CERCLA
Section 121, it's very similar in concept to the National Contingency Plan approach.
Our statutory framework does not match  up  very well with 121 because our statutes
are the basis for our ARARs, so we are oftentimes issuing ARARs that are
essentially background, and they don't match very well; and we have problems with
the EPA sometimes on that.

At  this point I'd just like to make some general comments on this whole issue to try
to lend some perspective, just to define what we're looking at here today from this
perspective I just outlined.  In many cases where it is clearly  not feasible to meet the
first-tier requirement to remove all contamination to background concentrations,  it
may be feasible to meet the second-tier requirement to protect the ground water by
partial removal, containment, or control responses. These are the types of cases
where the soil-to-ground-water fate-and-transport projections do factor into  the
remedy selection process.

Another point I'd like to make is  from a technical approach: the ground-water
protection requirement can be accomplished by either removing the contamination,
containing the contamination, controlling  the contamination,  or some combination of
                                       49

-------
these actions.  The term "soil clean-up goal" implies clean-up by removal of
contamination (essentially excavation responses) but we need to remember that
ground water can also be protected by contaminant containment and control
remedies as alternatives to clean-up remedies.

Another factor is that when we look at the whole broad spectrum  of NPL sites,
oftentimes the fate-and-transport model is important to the decision-making process,
but there are a lot of cases where it is not important.  I just want to make a couple of
comments on that.

In many cases the remedy selection process is driven by feasibility factors.  And
consequently the soil-to-ground-water fate-and-transport projections are not crucial to
the decision. For example, in Pennsylvania about one-third of our NPL sites are
large-sized closed municipal  and industrial waste landfills.  For many of these large
landfills, the only feasible  option is to cap the landfill. This remedy often needs to be
selected for  feasibility reasons independent of the  impact on the ground water. This
remedy is often selected even though the cap may permit a continuing, though
generally reduced, release to  the ground water.  The ground water operable unit is
then addressed after the fact  to the extent necessary and feasible.

It is possible to have examples of both background clean-up feasibility and
infeasibility  driving the decision-making process at  a site independent of the fate-and-
transport projections. In a case at a state Superfund site, pits which were about ten
feet deep were filled with concentrated organic solvent waste. The bedrock is about
15 feet below the surface, i.e., 5 feet below the bottom of the waste. The waste is
being excavated.  The aerial extent of contamination has a clear boundary, and it is
feasible to excavate the waste and contaminated soil to a clearly defined boundary
independent of the soil-to-ground-water fate-and-transport  projections.

On the other hand, the vertical extent of contamination does have a clear boundary
and some of the contamination is in the bedrock.  In the vertical perspective it is not
feasible to excavate into the bedrock, therefore, the feasibility factors again drive the
decision independent of the soil-to-ground-water fate-and-transport models. We
generated fate-and-transport  numbers, but they really aren't going to influence our
final remedy selection because we can meet our standard of background laterally,  and
vertically we can't meet it and all that we can excavate to is to the bedrock.  So again
the decision will be made in both directions, and the fate-and-transport model really
will not define the excavation.  So now I've provided some examples of where the
models don't apply.  Now I'll try to focus on some  where they do.

Pennsylvania DER has not adopted a specific soil to ground-water fate-and-transport
model as a standard model.   Nor  have we adopted default soil clean-up numbers.  It
is my personal position that we should not adopt the standard official model in the
future. There is some debate within the  agency about this but at this point we do not
have any official model.
                                       50

-------
There is a large degree of site-specific variability among Superfund sites.  A specific
fate-and-transport model and specific assumptions which may be appropriate for
some sites may not be appropriate for other sites.  It is, therefore, preferable to
account for site-specific information and allow for professional discretion in selection
of the specific model and specific assumptions.

There is not a strong empirical basis to the fate-and-transport models. This was
touched on a little bit earlier. There may be some  debate about this, but I feel that
there is not a high degree of scientific certainty associated with the projections
generated by the fate-and-transport models.

A given model can be run using a range of reasonable parameters and the model will
generate a broad range of projections.  Expand the  exercise to include a range of
different models (or as mentioned earlier, sort of "model shop") and the resulting
projections will cover an even broader spectrum.

Generally there is no scientific supportable basis to prove that any specific model or
any particular set of parameters generates the correct result for any given site.  Ron
Sims  mentioned earlier in his discussion about "bioredemption"; I think that a similar
type of thing applies here where some people sort of adopt a religious faith in the
model, in the numbers that they generate and then  those become the absolute
numbers. I think we as scientists have to guard against that.

And as scientists and as regulatory decisionmakers it is important not to overstate the
degree of certainty associated with these projections.  Selecting a specific fate-and-
transport model as the official model creates a false sense of scientific certainty in the
mind of citizens, legislatures and senior officials and a reduced sense of credibility in
the mind of the regulatory community.

In order to provide a more accurate sense of the decree of scientific certainty or
uncertainty associated with the fate-and-transport projections,  we should use a range
of reasonable models and parameters to generate a range of reasonable projections
which can then be factored into the site-specific remedy selection process.  So the
ultimate results  that we report from these models I  suggest should be reported as a
range of reasonable numbers based on a range of reasonable parameters.

We need to look at the feasibility of achieving several  different clean-up goals across
the whole range of reasonably projected clean-up goals from the most conservative
and most costly  goals to less conservative and less costly goals. In some cases it may
not be feasible to clean up to the most conservative projected goals, but it may be
feasible to achieve less conservative projected goals which are still well within the
overall range of reasonable projections.  So in summing up the issue that was raised
this morning as to whether the numbers should be used as an  absolute clean-up
number or as a risk management tool, obviously I'm advocating that these should be
used as a risk management tool that is then fed into a  very complicated remedy
                                       51

-------
selection process which includes a whole lot of complicating issues such as some of
the enforcement issues that Jeff Rosenbloom mentioned and a variety of other issues.
                                      52

-------
QUESTIONS FOR PANEL 1

-------
                        QUESTIONS FOR PANEL 1
Question: Jim Harrington from New York. I have to thank Mr. Tsiamis for
comments on the negotiation process to find an acceptable solution, but one thing I'd
like to point out or ask him is about the utility of the model and can everybody use
the model.

MR. TSIAMIS:  No I don't think so.  I don't swear by the model. It was useful to us.
MR. TSIAMIS:  The person using the model has to be trained to use it, but no, it
cannot be used for any one site, I don't think. In our case the model was useful
because relatively speaking, we had a simple site.  We only had two contaminants:
trichloroethylene and tetrachloroethylene; and we had a very shallow and
consolidated zone, and as I mentioned during my  presentation the receptor well was
in a relatively close distance to the site. So we thought it suited our needs. However
in a case where you have a site with multiple contaminants, I  do not think that this
model could be used effectively since in my opinion it does not take into account one
important mechanism there, which is to facilitate transport.

Question: Follow-up question is: How do you justify not using the site boundaries as
point of compliance; you, in fact, are using the first public water supply well.

MR. TSIAMIS:  You have to  remember that our remediation consisted of soil
remediation and ground-water remediation. So the question there was not a  question
of protecting human health, which is the reason for defining a point of compliance.
We would have a ground-water remediation system there until the ground water was
cleaned up.  So we felt that in that case that first public well would be a reasonable
point of compliance.

MR. WILLIAMS:  Anybody else like to respond to that?

MR. WILLEY: In our region we usually have a compliance boundary that is the
point of departure that is:  The site will be cleaned up everywhere period, including
beneath the site.  The next point of departure is the perimeter of the site.  That
usually takes in landfills that appear to be ridiculous  to clean  up beneath the  site or
achieve ground water MCLs, and another  exception would be some of the PCB sites
that existed in the region. Beyond that we endeavor to choose the clean-up beneath
the site.  It's very conservative.  We have not drawn a compliance boundary beyond
the property of an individual source.

MR. SIEGRIST:  Bob Siegrist, Oak Ridge National laboratory. You have been
talking so far about cleaning up a site which is assumed to be dirty.  How do  you use
a site-specific approach to assess whether a site is dirty to begin with, and before you
                                      53

-------
go further to characterize it sufficiently to execute your site-specific approach? There
are many sites, but the first question is:  How bad is it? Do we need to do anything
further? How far do we go?

MR. ROSENBLOOM:  The question was:  How do you know if you need to actually
clean up a problem at a site?  I guess the approach that we took, at least at the sites
in Arizona, is that we had to get a handle on the mass of contaminants in the vadose
zone because we knew we had a ground water problem. We wanted to ascertain
whether that mass posed a threat to the ground water. I know it's a chicken/egg
problem, but we knew we had a ground water problem need.  We knew we had some
past source we had to determine if the vadose zone currently presented a problem.
So it's a matter of data  collection using the VLEACH model to ascertain whether
that mass would have adverse impact on ground-water quality.

MR. WILLEY:   In New England all of us-I can't think of an exception-all the sites
either have ground-water contamination or  a manifest environmental insult.  That
does not mean that the  hazardous ranking score is directly related to the priority of
the site or how bad the  site is.  It's simply to rank the site.

Now to a certain extent that is already done before we get to it.  We do have sites
where we are looking closely to de-list them.  We have not as of yet in our region
de-listed any sites, and we have not found any sites as of yet to be strictly speaking a
no-action site. That issue will be coming up within the next year or so also.

MR. LUCKEY:  Fred Luckey with the EPA Region 2. Is there a standard or
accepted method for total organic carbon analysis for soils? I've heard there are a
few different ways, depending on how you use it you get different values. I was
wondering if there was an opinion out there on which method is best.

MR. WILLIAMS:  We have been working on several methods there.  I'm not
personally involved with it, but we have several people at the lab that are working on
some methods and some of them are available.  In fact, contact me  later—we'll try
put you in touch with the right people to really answer your questions, but there's no
one standard as such.

MR. MCNEVIN: Tom McNevin, New Jersey DEP. Dave,  did you use ten percent
organic carbon for your particular site model?

MR. KARGBO: Yes.

MR. MCNEVIN: That was based on analysis?

MR. KARGBO: Yes.

MR. MCNEVIN: Would that be indicative simply of the surface layer?
                                      54

-------
MR. KARGBO:  That's right. As you probably remember I did allude to the fact
that that could be misleading because there is organic carbon content as you go down
the profile.

MR. MCNEVIN: Which would make your results then not conservative in the sense
that the ~ if you had a lower amount of organic carbon or a number that more
accurately reflected what was actually in the profile you would end up with lower
residual soil numbers.

MR. KARGBO:  Exactly. I indicated there's a possibility of distinct stratification
because of differences in soil texture, and if you have layers which are very fine-
textured, even though they may not have very high organic content, may also retard
contaminants. So I can't say absolutely that it would be very conservative.  That
brings us to the question of getting to know exactly what kind of layers you have.
Where the properties of those layers are and try to understand how flow and
contaminant retardation is going to be as contaminants move down the profile.

MR. MCNEVIN: I bring this up because one of the problems we've run across in
our modeling attempts--if you start to use more realistic organic carbon numbers or
try to average it across the profile, and you use point one or point two or point three
or something like that, you tend to get extremely tiny allowable numbers for residual
contaminant in the soil which tends to run contrary to our actual field experience in
terms of the prediction that you would then make from that, and that's so in our
mind  is a problem.

MR. KARGBO:  My suggestion to that is to use a step-wise approach wherein you go
from layer  to layer.  If you know the  existence of these layers, just like the HELP
model that calculates percolation from each layer to another, you can also calculate
as a result how much water's going to move down to get with a contaminant as a
result of your Kd you may calculate in these different layers.  I think that would
probably be a much better approach  than just averaging.

MR. WILLIAMS: One more question.

MR HAYES:  Bernie Hayes, Region IV. I had a question for Dave as well.  You
mentioned  something about a critical fraction of organic  carbon beyond which—

MR. KARGBO:  Well, the critical fraction is for almost all the contaminants you can
measure, or you can calculate w,hat the critical fraction is  based on surface area of
the soils that you are looking at.  In other words, if say you're looking at TCE and
your calculated critical fraction is probably eight percent, and if you have a soil or
layer horizon that's probably reading five percent as organic, fraction of organic
carbon, then that means that the relationship Kd equals to Koc times Foe may be not
be adequate. You may want to look at another relationship that includes not just
Koc and fraction of organic matter content but also the surface areas of the soils.
                                      55

-------
MR. HAYES: So you're saying that perhaps straight adsorption onto the surface
area of the soils rather than strictly under the organic fraction may offer-

MR. KARGBO:  Exactly. That depends also on the type of contaminants that you
have because obviously the structure of the contaminants is going to affect the type  of
adsorption that you're going to get and the magnitude of the adsorption and also
reactions of the contaminants.  You may have contaminants which, as a result of
reaction, probably end up having a positive charge which could adsorb on the clay
matrix.  So you have to also consider the type of contaminant as well as the surface
area of the soils that you're looking at, and, hence, the time in your critical fraction
organic carbon if  you don't have that based on your soils analysis, then you know the
simple relationship of Kd equals Koc times Foe is inadequate.

MS. BARRY: I'd like to wrap this up really briefly.  We've listened to a number of
site-specific approaches.  This afternoon we're going to take a different tactic and
explore a number of regulatory approaches as exemplified by several state programs
and two EPA programs outside of CERCLA.
                                       56

-------
            PANEL 2
NON-SITE-SPECIFIC DEVELOPMENT OF
NUMERICAL STANDARDS FOR VARIOUS
     CHEMICAL CONSTITUENTS

-------
                              INTRODUCTION
                               Randall Breeden
MR. BREEDEN:  As Alison mentioned before, our focus here today is to look at the
different approaches to establishing various soil cleanup standards, and the last panel
addressed some of the issues and problems associated with setting site-specific
cleanup levels using very sophisticated computer modeling techniques for fate-and-
transport analysis.

Right now we're going to take  a look at the flip side to establishing cleanup goals by
taking a look at non-site-specific development of numerical standards.

And by numerical standards, what is meant is that approach would establish cleanup
goals by setting specific concentration cleanup standards for various chemical
constituents.

As you can glean from that statement, there are obviously some major questions,
problems, and issues that must be addressed when the numerical standards approach
is used. Some of those I would like to present to you for thinking about while the
panel is presenting their material.

One is: What tools and methods do we have available at this point in time for
establishing numerical standards?  Two is: What is the basis for setting those levels?
And by that I mean, do we base them on risk management and risk assessment
procedures? And if so, what exposure pathway will be chosen in establishing the
standards?

As you can see, if you were to  choose direct ingestion of soil as your exposure
pathway, then that value could be  significantly different than if the exposure pathway
is via ground-water.

Therefore, does one establish numerous cleanup levels for each exposure pathway
and then end up with several different cleanup values for each constituent? And if
you do, how do you pick and choose within that range?

Perhaps of major importance is a more policy and procedurally oriented approach,
for example:  If one has chosen to use a numerical standards approach, does your
program and your statutes or your regulations allow you to have an out so that when
you come across the situation where the numerical standards do not really apply to
your site, do you have another  viable mechanism by which to come up with an
alternate value?

Well, Joe mentioned several points which must be considered when establishing site-
specific cleanup goals. Two of them are equally applicable to establishing non-site-
                                      57

-------
specific numerical goals.  The first is:  Are remediation technologies even available
and are they feasible for attaining your cleanup levels?

And, of course, does the cost of cleaning up to those levels as determined by non-
site-specific methods or numerical standards exceed the cost of obtaining site-specific
cleanup levels?

Well, those are just some of the factors to keep in mind.  I'm sure that the  panel
members who will be giving presentations will have some interesting stories that they
can tell us  from an historical perspective, and I'll  turn that over now to Peter Kmet
from the Washington Department of Environmental Conservation.
                                       58

-------
                               WASHINGTON
                                  Peter Kmet
MR. KMET: Thank you. Good afternoon.  I actually work for the Washington State
Department of Ecology, a futuristic sounding name. I am in the state's Superfund
program -  we call ourselves the Toxics Cleanup Program.

I am the section head for the policy section, and our section has just completed a
two-year process for adopting cleanup regulations  for our state to cover both
Superfund  sites as well as other state cleanup sites.

We have a new rule for just the cleanup standards portion,  adopted last Friday,
which I know many of you have been following. We plan to get copies of it out to
you as soon as we get some printed.

The regulation was  completed in response to a citizen's initiative called the Model
Toxics Control Act, Initiative 97.  In Washington state, citizens have citizen initiative
capability, where they can essentially petition state government to create a statute
through a citizen petition process  and  actually voting on  it.

The initiative was put on the ballot in November of 1988, and I think it is important
to talk about this because it gives you an idea of how people in general feel about
this issue.

The legislature had been debating this issue for years.  They got to the point where
they could  not reach agreement. The  environmental groups within the state got
frustrated and proposed the  citizen's initiative that set up essentially a state
Superfund  program.

Business and local government, immediately got nervous and went ahead and worked
legislation through the legislature,  and the governor called a special session and
within one day we had a Superfund law on the books essentially in response to that.
But the citizen's initiative went forward anyway.  They got enough signatures, got on
the ballot, and in the fall of 1988 the citizens of the state of Washington were faced
with a choice.

They had the citizen's initiative that they could vote on and they had the "business"
version of the cleanup law that they could vote on.

There were two  questions on the ballot hanging in the balance here. The first one
was: Do you want a state cleanup program?
                                       59

-------
Overwhelmingly the citizens of the state of Washington voted, "yes, we want a
cleanup program." Over 80 percent said "we want a program," and we're talking
about a program that was going to tax chemical feed stocks to the tune of an
estimated 50 to perhaps much more millions of dollars to create the program.

The second question was:  "Which one do you want." And the answer was fairly
convincingly the citizen's initiative. I think it passed 55 to 45 percent, which in terms
of elections is a fairly overwhelming citizen endorsement.

We found through our  discussions on this rule, that there was a great deal of concern
in the general public about these sites and the need to clean  them up, and I think
that's borne out on a national level and you're  finding it out in your individual states.
So that's the foundation that we were drawing upon to write this rule, a very strong
mandate from the citizens.

Before I go too much further, I'd like to acknowledge Dave Bradley of my staff who
is the principal  author of the  cleanup standards portion of this. He is not here today,
unfortunately, but has done a lot of the background work in this rule making, and I
know as some of you want to contact us, you'd probably be better off calling Dave
and getting your questions answered more directly from him.

Well, we have the citizen's  initiative and we're  faced with a dilemma.  The citizen's
initiative said, "you shall adopt cleanup standards."  We didn't have a  choice. But we
knew, as everyone here knows, science isn't there, that there's still a lot of questions
on how we proceed.

Well, in the rule-making process we called this the regulatory dilemma, and  I want to
just digress a little bit more here and read a passage that I think puts this whole thing
in perspective.  It's some testimony by Dr. Bates, former science director for the
Food and Drug Administration testifying before Congress on a cancer policy In 1978
and his story goes something  like this:

       "A classic episode in the history of disease prevention  took place in London in
       1854. An epidemic of  cholera occurred in the neighborhood around Broad
       Street. John Snow, the hero of the story, studied the habits of the victims and
       found that almost all obtained their water from the well on Broad Street.
       Swift action was taken; the pump was closed down and the epidemic rapidly
       subsided.  This disease was caused by the exposure to  the bacterium Vibrio
       cholerae.  One can imagine the reaction that might occur today if it were
       proposed to close down a pump on the basis of evidence of the kind  that was
       obtained by John Snow.
       Many scientists would  point out that it had not been conclusively demonstrated
       that the  water was the cause of the disease.  They would be troubled because
       of the lack of satisfactory theoretical knowledge  to explain how the water
       could have caused the disease.  Furthermore, other habits  of those who had
       become  ill had not been adequately investigated, so it would not be possible to
                                       60

-------
      rule out other causes of the disease. The scientists would have been correct.
      Others would have pointed out that some members of the community who
      drank the Broad Street well had not succumbed cholera. Thus, even though
      there was something wrong with the water, there must be other factors
      involved and could we control these we would not have  to be concerned about
      the water.  The conclusions  are also correct.  Some who consumed the water
      from the Broad Street well would have objected to closing it because the taste
      of water from other wells was not agreeable.  Finally, if the pump had been
      owned by an individual who sold the water, he would certainly have protested
      against closing  down his business on the basis of inconclusive evidence of
      hazard."

And I think that pretty much sums up the regulatory dilemma that we're in here.

The process that we went about  laying out in our  rules is very similar to the federal
Superfund process as far as the administrative portion of the rule goes.

We do depart significantly in the cleanup standards  portion of the rule in one big
way.  The cleanup standards are determined first,  more or less independent of cost,
and then the remedy is selected  considering cost.

Basically people felt, we had advisory groups, and again, citizens groups as well as
business and others involved here,  was that the decision  is either it's clean or it isn't.
Cost shouldn't come into that. It's either clean or it isn't.

The rules address how to develop these cleanup standards for a variety of media,
including soils. The term cleanup standards as we use it in the rule includes the
number or  the level, concentration, whatever you  want to call it, that number in the
soil, and also it includes the point of compliance considerations that have been
touched on here.

For each media, including soil, there are three basic methods for doing this.  There's
a method A, tables, that you can go to.
There's a method B which is the standard method, and generally provides for a one
in a million cancer risk based on a residential site use scenario.

And then there's a method C, a conditional method allowed under only certain types
of conditions, which includes industrial sites which are quite narrowly defined in the
rule, and generally provides for a one in a hundred thousand acceptable risk level
based on an industrial site use scenario.

Okay, method A, as I mentioned, consists of a table of numbers. The numbers have
been based primarily--or derived primarily based on consideration of the direct
contact soil ingestion and ground-water protection.
                                       61

-------
They're generally felt to be fairly conservative numbers so that they can be applied to
a wide variety of site settings.  They're intended to be used at relatively simple sites
with very few hazardous substances.  The tables of numbers that you'll see in our rule
are fairly limited, maybe a dozen or so substances.

If these numbers aren't conservative  enough,  however, we do have a caveat in the
rule that we could require even more conservative numbers if necessary based on a
site-specific analysis.

Maybe it would be instructive to just read some of these.  I'm sorry I don't have an
overhead of this. I didn't  anticipate a great deal of interest here, not this many
people.

I'll just read off a few of these.  Arsenic 20.  These are parts per million.  Benzene
25,  cadmium 2, chromium 100, ethylbenzene 20, lead 250, PAHs 1, PCBs 1, tetra-
chloroethylene .5, toluene 40, TPH gasoline 100, TPH diesel 200, 1,1,1-
trichloroethane 20, trichloroethylene  .5, xylenes 20, just to give you an idea of the
numbers for soil cleanup levels.

Methods B and C  have detailed  equations and assumptions for determining the
cleanup level for the direct contact pathway that we've been talking about.

They generally follow the  EPA risk assessment guidance, although we have
prescribed in the rule the  assumptions that you're to use in most cases.

For determining cleanup levels that are protective of ground water, which is what
we're talking about here under methods B and C, you first have to determine the
appropriate level of ground-water protection and then you have to determine the soil
level needed to achieve that.

The rule specifies pretty specific criteria for what is an aquifer that is  to be protected
to drinking-water standards. And we didn't distinguish between actual or potential
uses, feeling that we couldn't predict future land use or ground-water use that
accurately.

So we have put technical criteria in terms of,  either it produces so much water or it
doesn't. For example, incidentally, we chose  a sustainable yield of a half a gallon per
minute as an "aquifer," which means  most ground waters in the state of Washington
have to be  protected for drinking-water purposes.

Knowing that  then, you can go through these  types of processes,  method A or B, for
example, and derive a ground-water number.

In many cases that will line up with the drinking-water standard.  If it turns out that
the assumption and the criteria that we provided do not lead you in the path of it
being a drinking-water aquifer, then you have to look  at nearby surface waters and
                                       62

-------
try to do some projection on what concentrations you need to protect nearby surface
waters.

Or if there are other ground water aquifers that are in the area that the
contamination is going to migrate into, you have to do some modeling to project into
those deeper aquifers.

For ground-water protection, we start out with the assumption that the  soil cleanup
number applies throughout the site, but that you can adjust that if it's not
"practicable", which is the key phrase we used there.

You can adjust it as far out as the property boundary, but it's supposed to be as close
as practicable to the source of the contamination.

Once an appropriate level of protection for ground water has been determined, two
options are provided.  There's the magic. The soil cleanup level is set at a hundred
times the ground- water protection level, or you can use an alternate method of
demonstration, the details of which haven't been provided in the rule, which is one of
the reasons why I'm here.

Okay, what's the basis for that hundred times factor?  It's not extensive.  Basically we
had a number of advisory committees and technical people looking at this.  We had
some historic practice as to what was going on in the state.

We were aware of what other people were doing in the country, and after talking to
and discussing this amongst a bunch of people and looking at how the numbers came
out, we thought it was a reasonable point of departure upon which to base this
discussion.

There was some limited modeling that was done as well, although I'm not familiar
with that enough to tell you exactly what was done there.

We also looked at the option of using say ten times for highly mobile, hundred times
for moderately, and maybe a thousand times for less-the least mobile contaminants
as a way of trying to break it down.  That was a suggestion by some of the people
involved in this.

When we started looking at mobilities of compounds as reported in the literature, we
couldn't really feel we could make a good break.

For any given compound you sometimes find two,  three orders of magnitude as to
just how mobile it is.  And so we chose not to do that. We may go back and revisit
that.
                                      63

-------
We also looked at the of using a leaching test. You've heard some people refer to
that.  For example, using the TCLP to perhaps project or predict what might be the
ground-water impact.

At this point we just do not feel we have enough information to know whether that
would be a conservative enough approach as a departing point, so we chose not to do
it.

There are several concerns--! have been involved with leaching tests for many, many
years, some of you are aware, and trying to interpret results.

The results are highly variable.  You're using basically a short-term test to predict
long-term impacts in the field. You're introducing oxygen, usually, and stirring the
stuff up, which is not what's occurring in the field necessarily.

And probably most importantly, your liquid-to-solid ratio has to be pretty high in
order to just get enough sample to analyze it in the lab, whereas in the field a drop
of water is slowing perking its way down over the surfaces of a considerable amount
more soil.  So that the actual concentration at the bottom of that column might be
considerably different than what would come out in a leach test.

Tests like the TCLP try to make up for that by using a moderately aggressive acid
leaching to try to adjust for that.  For organics I think there still needs to be some
work in that area.

Another issue that we considered in establishing the soil cleanup levels was the point
of compliance for soil. For  direct contact we chose to make the point of compliance
everything within the upper  15 feet of the soil.

That number was based on what we thought was protective, that if the soil at depths
below that, it was unlikely that it  would  ever be brought to the surface due to site
development activities.

But 15 feet, given the depths of basement, utilities, regrading of sites, seemed to be a
reasonable number.

Interestingly enough, we sent out  over a thousand copies of this rule and  few
commented on that particular decision.

I'll just tell you a couple other points we've addressed.  We looked at soil size. We
say sand  size or smaller particles  are the soil size  that we're  concerned about
primarily for ingestion or leaching.

We have specified total analysis as the method of analysis, not a leaching method
right now.  We have included some statistics, how to test, although I think there's a
                                       64

-------
lot yet to be done in that area as to just how you determine compliance with a
particular number.

I think we, through this rule making, narrowed the playing field a little bit and I hope
we'll start moving towards doing cleanup instead of arguing about numbers, but there
still are, as you can see, a number of areas where  we've got to do some work here.

And this is one in particular, and I can't pretend to say that we have the answers to
that.  We have an approach, something that at least lays out some ground rules for
people to go from, and certainly would be willing to share with you  in more detail
how we've approached it and would be willing to answer questions at the end.

Thank you.

MR. BREEDEN: Thank you, Peter.
                                      65

-------
                              CONNECTICUT
                                Randall May


MR. BREEDEN:  Our next speaker is Randall May from the Connecticut DEP.

MR. MAY:  My expertise is in the area of soils engineering and small-scale waste-
water management, and for 15 years I headed Connecticut's program in that area.
Then, our commissioner began assigning me to various department-wide or inter-
burea . tasks in many different areas. Those of you in EPA Region I will understand
that Commissioner Carothers  does not shrink from assigning difficult tasks. In this
case she assigned me to pull together our two Bureaus with somewhat conflicting
viewpoints on clean standards, assign goals and priorities, look into how we should
manage cleanups in view of the fact that we have several different units working on
them, and also write regulations for, "how clean is clean!"

For the past few months I've read a great deal and had extensive  discussions with our
staff.  Based on that I believe I can do what I was asked; describe the way clean up is
undertaken in Connecticut at  this  time.

All of us in the state program define our needs, our priorities, our regulatory needs,
our clean standard and everything else, based on what we're faced with, our universe,
our geology and so forth.

In Connecticut we have relatively  few federal Superfund sites.  We have about a
thousand sites that I'm simply going to call problem sites. These form the inventory
of our State Superfund, and are exclusive of the large number of problems with gas
stations.

About 98 percent of this universe  of a thousand sites that we are  dealing with involve
both solvents and metals. Metal finishing is the primary industry  in Connecticut and
has been for a long time.

Our staff considers 50 to 100  of these as very major complex sites. In about a
hundred cases total our staff believes that remediation will be very difficult.  In some
of those hundred cases they believe that there is no practical means of remediation
that comes close to the standard that we will probably adopt. That  creates a
dichotomy which I wish to discuss.

The driving force behind remediation in Connecticut is our Clean Water Act of 1967
which provides  our basic structure for permitting, enforcement, standards of water
quality.

We are determined that resource  protection is going to drive our  program and our
regulations.
                                       66

-------
Cost of remediation, numerical risk assessment and fear of disputes with people will
not be the principle driving force.  Resource protection will be.

I thought the gentleman from Pennsylvania did an excellent job of talking about the
dichotomy of what is feasible and the mandates for resource protection which are the
philosophical basis for our statutes and regulations.

Our law in Connecticut is very similar to the one  in Pennsylvania.  It says that no
person shall initiate a discharge into the waters of the state without the blessing of
the commissioner. As one former director said, that outlaws all activities of mankind
very effectively.  Our water quality standards are adopted under this law.  They were
first adopted in 1967 and were extended to a comprehensive set of ground-water
quality standards and classifications in 1980. I want to talk a bit about those
standards because they are an important element  in the  non-site specific clean up
program that exists.
The majority of our state is mapped as either GA or GAA.  These are ground water
areas where the  ground water is presumed suitable for drinking without treatment. A
GAA designation indicates that this is an area which is tributary to a public water
supply, either surface or ground water, and that no discharges whatsoever are allowed
other than those that  are human or animal in origin and non-contact cooling water.

Very shortly there will be a sort of super GAA designation which will reflect the
areas of contribution in our wellhead protection program.  I'm also awaiting the land
use regulations which will govern  in those areas.   This program is termed, "aquifer
protection" but it is actually a well head protection program. Within those areas state
regulations will control land uses  and will create a special  enforcement program.

Areas mapped GA are suitable for drinking without treatment.  These areas are the
host for a myriad of drilled bedrock wells.  One-third of Connecticut's population is
dependent on ground water for its water supply.  Of that total one-half comes from
bedrock supply.

Ninety percent of Connecticut is mapped as GA or GAA.  In other words, in  over
ninety percent of our land area our goal and presumption  is that the  ground water is
suitable for drinking without treatment.

Roughly six percent of the state is classified as GB. These are areas that are  known
or presumed to be degraded, generally urbanized  areas which are our historic
industrial sites.  These GB areas are the areas where we have internal debate about
the standard for  remediation.  Some of that is  due to the subtlety of the language of
the standards for ground-water quality.

I drafted the original ground-water quality standards  back  in 1980 and have never
thought of myself as subtle.  Apparently I was, because there is considerable debate
about what the policies for GB areas mean.
                                       67

-------
An example of this is the policy that states that the Commissioner may not seek to
restore ground water to drinking water quality in these areas. However, even if that
decision is made, the degraded ground water cannot adversely affect adjacent surface
water quality.  And remember, in Connecticut you can stand just about anywhere,
heave a reasonably sized rock, and hit a stream.  We are, therefore, very concerned
with in-stream water quality and aquatic toxicity as it is affected by discharges of
degraded ground water.

We also have an anti-degradation policy for ground waters within the GB areas.  No
discharge  is allowed which would further degrade the ground water or create
irreversible contamination.  That is designed as a discharge which should preclude a
further restoration to a quality suitable for drinking without treatment.

I think you can see some of the difficulties that can arise when you consider  these
policies in the context of site remediation. They are especially critical since the
majority of our major and complex sites are in areas which are mapped as GB.  The
problem is integrating realistic goals with anti-degradation and in-stream water
quality while dealing with especially difficult sites.

Of course we do have many sites  in areas which are mapped as GA or GAA.
However,  in the main there sites are somewhat smaller,  less complex or  pose less
threat to the resource.

Roughly three percent of our land area is classified as GC, which are areas that are
hydrogeologically suitable for waste disposal and unsuitable for significant water
supply.  As with the GB areas used for waste  disposal cannot further  degrade ground
water or adversely affect in-stream water quality. Public water supply must be made
available in these areas. Fortunately some of these areas have historically been used
for waste disposal.

In a general sense site remediation activities undertaken by our Bureau of Waste
Management and our Bureau  of Water Management have been divided  along the
lines of the water quality classifications.

In the areas classified as GAA and GA, remediation has primarily taken place under
the direction of  the Water Bureau.  These remediations  have had the highest priority
since they have had major impacts on drinking water supply.

We are a  densely populated State. It  does not take much  of a problem to
contaminate a fair number of wells. In fact we have thousands of such wells creating
an overwhelming problem.

For many years  our Water  Bureau had all the  expertise  in hydrogeology, ground
water and remediation within  the agency. In addition this Bureau is charged with
running our program of delivering potable water when a well is contaminated. Our
                                       68

-------
Waste Management Bureau has been in a growth process and is now staffed and up
to speed with numerous enforcement actions underway.

The goals of the water program have been driven from the standpoint of maintaining
and improving the quality of the resource, generally water being used for drinking
without treatment.  The protection of ground and surface waters are held paramount
to cost, difficulty of remediation and so forth.

The most common remediation scenario in these cases has been pollutants that can
not be tolerated in the unsaturated zone in an area where hydraulic control is
uncertain.  In these cases the staff has required use of analytical techniques that will
yield a conservative estimate of the mass and concentration of pollutants present.

The assumption is made that there will be an instantaneous loss of the total amount
of pollutants, discounting dilution, absorption or other mitigating factors such as
volatilization, or biodegradation.

The responsible party has then been required to remove or treat soil until the
foregoing analysis  indicates that the instantaneous release will not violate drinking
water standards or standards established by the toxic hazards section of our
Department of Health Services.  This standard has been controversial but has been in
effect for several years.

Aquatic toxicity is  considered in any  case where it is relevant, utilizing conservative
in-stream dilution  factors. In most cases the point of compliance with these standards
is the property boundary.

The staff of the Water Bureau has made empirical observation that the extensive use
of this technique of either "dig and ship" or "dig and treat" results in the most
immediate and dramatic improvement in water quality that can  be  obtained. Where
hydraulic control can be assured,  this standard approach is sometimes relaxed with  a
lesser reliance on soil removal and treatment.

Staff in both Bureaus are dissatisfied, however, with the state of where we are in
ensuring long-term management of hydraulic controls and pump-and-treat systems.
Both believe that we must find ways  to ensure operation for periods of more than
one hundred years. I do not plan to  be around to check up on some of that.

Before discussing our remediation problems in the GB areas, I would like to describe
the position of the staff on one of the principle topics of this conference, the
modeling of migration of contaminants from the unsaturated zone.

Our staff is uniformly skeptical of the modeling that is available. We have studied
this issue in conjunction with the Connecticut Agricultural  Experiment Station and
concluded that the models that we have seen are far too simplistic to be accurately
applied to the complex environment of the soil system.  We believe that there is a
                                       69

-------
better basis for the models that describe metals transport than those for organic
chemicals, but that both are poorly understood. We believe that there is a real need
for better three-dimensional models.
The principle reason for our skepticism is that  we have had considerable direct
experience that contradicts what the models predict. Our widespread contamination
problem with the pesticide EDB is a good example of  this. All the evidence was that
this pesticide would be lost in the 20 years since it was last used.  Yet we found that
significant amounts remained held in the soil voids and slowly discharge into the
ground water. The same appears  to be true for the pesticide chlordane.  So we have
had some very bad experiences that contradict  what the models would tell us.

Our position is that if a compound is persistent in the  soil,  then you are going to have
to deal with it in ground water sooner or later.

Predictive models are used in Connecticut as a check on issues such as monitoring
well placement and fluid movement, basically used as part  of a judgement process but
not to arrive at a numerical clean standard.

The staff--and I stress this-is extremely concerned about the technology-transfer
aspects of this modeling issue.   We have excellent staff, all graduate geologists and so
forth, but they are frustrated because very little information on modeling cornes down
to them. They are concerned that there are worthwhile developments that they are
unaware of.  They feel that most state regulatory agencies have to rely on some
consultant walking in and  trying to bag them with a new model to find out it exists.

I alluded to  difficulties in  remediation in the areas classified as GB, due  to the
complexity of sites, the standards of water quality and  internal debate on what is
required.  The staff dealing with the standards  for the  drinking water supply areas
(and it is a very stringent standard) have primarily been dealing with smaller sites.
The larger, more complex sites, primarily in the areas  mapped as GB are being dealt
with jointly by the Waste Management Bureau and the Water Bureau and there is
some internal debate about the clean standard in those areas.

On the GB sites the same type of impact analysis is done and the  same conservative
assumptions are made, however removal or treatment  takes place  until impacts would
be ten times the drinking water standards. The validity of  this number is a matter of
some debate, particularly in light of the water quality goals. I do not, frankly,
understand the derivation of the ten times standard, but I understand that is was
worked  out with Region I of EPA.

There are several other issues  which are the subject of staff debate with regard to
cleanups.  Our hazardous  waste staff believes that the  days of "dig and ship" or
extensive soil removal are essentially over.

There is also a dispute over the value of soil removal when the majority  of volatiles
may already have migrated two to three hundred feet into  the bedrock.
                                       70

-------
I believe that the way we have to look at this is not to get so overwhelmed with the
problem of the ground-water contamination that we overlook the continuing source
of pollution in the unsaturated zone.  It is not an either/or situation.  We must not
abandon the principle of halting the illegal discharge and we must practice ground-
water remediation.

We have, just yesterday,  set some of the policies that we will follow in the drafting of
the regulations that I am working on.
We will continue to utilize the water quality standards as the basic tool for
determining the degree of remediation. To the maximum extent we will utilize the
relatively simple standard in the areas suitable for drinking water.

We will divide even further, or more precisely, the standards that will apply in the
different water quality  classifications.  We will provide a hierarchy of desired
solutions, with less desirable solutions available only i.fter stringent tests. We will use
modeling and risk assessment  as information tools, not as a gate to a lesser standard
of remediation.  Simple containment will be our last option.

And we will emphasize in these regulations a process  to  arrive at a clean standard,
rather than a number.
                                       71

-------
                              RCRA PROGRAM
                                Steve Cochran
MR. BREEDEN:  I'm going to turn it over to Steve Cochran of EPA.

MR. COCHRAN:  I want to discuss the implications of the RCRA hazardous waste
program on the cleanup of contaminated soil. As a general rule, the hazardous waste
program is not concerned with determination that a solid waste is also a regulated
hazardous waste.  Through the ARARs, the RCRA hazardous waste identification
process, and the management standards tied to it, has potentially significant
implications for cleaning up contaminated soils at Superfund sites.

I'm going to talk briefly about the two mechanisms by which a solid waste becomes a
regulated hazardous waste-the hazardous waste characteristics and listings.  Then I'm
going to discuss our efforts to develop de minimis exemption levels for listed
hazardous wastes.

The first mechanism-referred to as characteristic waste-involves identifying
properties or "characteristics" which, if exhibited by a waste, indicate a potential
hazard if the waste is  improperly managed.  The Agency has to date identified four
such characteristics: ignitability, corrosivity, reactivity, and toxicity.  Of prime interest
here is the toxicity characteristic.

The toxicity characteristic (TC), which was revised on March 29,  1990, identifies
wastes that contain one  or more of 39 hazardous constituents in sufficient
concentrations such that there is a potential for  hazardous  concentrations to leach
into the ground water. The TC combines fate and transport modeling, a leaching
protocol (the Toxicity Characteristic Leaching Procedure or TCLP) and health-based
values (i.e., MCLs  RfDs and RSDs) to  identify these wastes.  Thus the TC  is
designed to prevent ground-water contamination by requiring specific management of
those wastes  that clearly pose a threat to ground water.

The hazardous waste characteristics are based on  threshold values or properties.
Once a waste no longer exhibits a hazardous waste characteristic, it is no longer a
regulated hazardous waste. These thresholds are designed to identify wastes that are
clearly hazardous.  Wastes that are below these  thresholds, particularly for  toxicity,
may still pose a threat to human health and the environment but would not be
regulated as characteristic hazardous wastes. This is an important point.

The point to be made here is that the combination of the fate and transport model,  a
leach protocol, and the  health-based  numbers may be an appropriate approach for
determining cleanup levels for contaminated soils.  The TC levels, however, are not
appropriate cleanup levels since they are designed to reflect concentration  levels  in
waste that are clearly  hazardous (i.e., threshold levels), not the levels at which there
                                       72

-------
is no significant risk.  In general, the TCLP is a fairly good indicator of the potential
mobility of hazardous constituents in contaminated soil. It is important to note,
however, that the TCLP was designed to mimic what occurs (e.g., the mobility of
constituents) when wastes are disposed in a municipal landfill.  Because it assumes an
acidic environment due to decomposing organic matter, the TCLP would not be
appropriate for determining levels that are safe to leave in the ground without a
cover (e.g., "walk-away" levels).

The other mechanism for identifying wastes as hazardous is to specifically list them as
hazardous.  This process involves a determination by the Agency, after studying
specific industrial sectors, that specific waste streams are hazardous (i.e.,  listed)
because they typically and frequently contain hazardous constituents at levels that
pose a threat to human health and the environment.

Listed wastes are not threshold based hazardous wastes as  are the characteristic
wastes.  The only method for a listed waste to exhibit Subtitle C control is for it to be
specifically delisted through a formal rulemaking  process.

Further complicating the scope of listed hazardous wastes are the "mixture" and
"derived-from" rules, and the "contained-in" regulatory interpretation as they relate to
listed wastes. The "mixture rule" states that if you mix a listed waste with a solid
waste, the entire mixture becomes the listed waste;  The "derived-from rule" state that
any waste derived from the  treatment, storage or  disposal  of a listed hazardous waste
is also that listed hazardous waste.  The "contained-in" interpretation states that soil
(or other environmental media such as ground water or sediment) that contains a
listed hazardous waste must be managed as if it were the hazardous waste until it no
longer contains the listed waste. Thus once a waste is a listed waste, it remains a
listed unless delisted by rulemaking.

These "rules" can significantly impact Superfund cleanup activities involving soil
contaminated with listed wastes  because subsequent management (either on-site
treatment or off-site disposal) of those contaminated soils  trigger full Subtitle C
controls. The one exception is that Superfund can avoid the administrative
procedures (i.e., formal rulemaking process) and delist wastes managed on-site
through the ROD process.  However, the substantive requirements (e.g., levels that
provide appropriate protection of human health and the environment) must still be
met.

On top  of these requirements, Land Disposal Restrictions  (LDR) Program treatment
standards (termed Best Demonstrated Available Technology or BDAT) have been
established for  all hazardous wastes. These treatment standards must be met before
a listed hazardous waste (including mixture, derived-from,  and contained-in wastes)
can be disposed in a Subtitle C landfill (i.e., as a  hazardous waste). This has
potentially significant implications  for Superfund or RCRA corrective action cleanups
since the generation  of hazardous wastes (including environmental media
contaminated with listed waste)  as part of a remediation project could trigger LDR.
                                       73

-------
The BDAT treatment standards are strictly technology-based; that is, there is no
consideration of risk in setting the levels.  The LDR program was mandated by
Congress with a specific charge to reduce toxicity and mobility to the maximum
extent possible. Thus there is the potential for requiring treatment beyond the point
of significant risk.

The Office of Solid Waste (OSW) has been trying to solve some of the problems
outlined above with respect to the difficulties of hazardous wastes, contaminated
media, and treatment  residuals exiting Subtitle C control. Specifically, OSW is
developing a Notice of Proposed Rulemaking (NPRM) on de minimis exemption
levels.  As currently drafted, this proposal would set up a generic process by which a
generator of a hazardous waste could "demonstrate" that their listed waste (including
environmental media contaminated with listed waste  and treatment residuals)  is  no
longer hazardous.

The goal of the de minimis NPRM would be  to set up a self-implementing process by
which otherwise regulated hazardous wastes that clearly do not pose a significant risk
could exit Subtitle C without a formal delisting rulemaking.   At  the same time, the
NPRM would propose to cap  the BDAT treatment standards based on risk, thus
raising many of the BDAT levels and preventing unnecessary treatment.

The de minimis proposal would involve two steps; 1) sampling and analysis and 2)
the de minimis  demonstration. The first step would be the submittal of a sampling
and analysis  plan to the appropriate EPA  Regional Office.  The Region would have
60 days to review the plan, after which time the generator would proceed with
sampling and analysis.  The NPRM specifies roughly 200 constituents for which
testing would be required.

Once the sampling and analysis  is completed, the generator would submit the
demonstration package, which would  certify that all constituents of concern in their
waste were at or below de minimis levels. The generator would again submit  the
package to the appropriate EPA Regional Office for review. After 60 days, if no
comments are received from the Region, the  waste (contaminated media or BDAT
treatment residual) would no  longer have  to be managed as  a hazardous waste.

The de minimis proposal addresses hazardous waste and treatment residuals
differently from contaminated soil. For waste, the generator would evaluate the
mobility of constituents in the waste.  Thus a leach test  (the  TCLP) would be
required and the generator would have to demonstrate the waste or treatment
residue did not contain constituents of concern at or above de minimis levels.  (Note
that if the waste of concern was contaminated ground water, then the total levels for
detected constituent in the ground water would be compared to the TCLP levels). In
the case of waste or treatment residuals that  meet the de minimis levels, management
as a solid waste would still be required (i.e., the waste would be disposed in a
municipal landfill).
                                      74

-------
For soils contaminated with listed wastes, the generator would be required to
demonstrate that the hazardous constituent levels in the soil were below total levels
and leachate levels.  In this case, however, a different leach protocol is used, one
based on the acidity of rain water (the Synthetic Acid Precipitation Leach Test). In
the case of contaminated soils, if the demonstration can be made that the soil is
below both the total and leachate levels for detected constituents, then the soils
would no longer be subject to either Subtitle C (hazardous) or Subtitle D  (solid)
waste management requirements. In other words, the soil could be left in place. In
this case the exposure routes of concern are both consumption of contaminated
ground water as well as direct contact (ingestion or inhalation) of the soil.

Since the de minimis exemption levels for soil include both a total and leachable
level, and the leachable level is based on infiltration of rain water, this approach may
be appropriate to use in determining the adequacy of soil cleanups at Superfund and
RCRA corrective action sites. Clearly, the actual levels would vary depending on
site-specific considerations.  In instances were  detailed risk assessments aren't
appropriate, the generic de minimis levels may be an appropriate surrogate since they
are designed to be conservative. The de minimis program thus is not based  on risk
assessment but risk management where the goal is to avoid any significant exposure
resulting from wastes allowed to exit Subtitle C control.

On that note I think I'll stop and answer any questions you might have.  Thank you.

MR. BREEDEN:  Thank you, Steve.
                                       75

-------
QUESTIONS FOR PANEL 2

-------
                        QUESTIONS FOR PANEL 2
MR. PHILLIP: Stan Phillip, the state of California, a question for Steve and then
one for Connecticut Randy.

Steve, in all you said I didn't hear anything where one can make some decisions
about how to collect the samples to run TCLP at a Superfund site in a soil and debris
case.

Is it like  one worst case sample  of something at the site? How do you design a
sampling strategy to come up with what constitutes a representation  of whether there
is or isn't hazardous waste to manage?

MR. COCHRAN:  Okay. Well, we're taking as a given-let me answer the last
question first.  We take as a given that there is a hazardous waste, otherwise-in other
words, from the RCRA perspective the question we're answering is:  "When is it no
longer a  listed waste derived from a mixture or contained in waste?

So the presumption is that, in fact, there is a listed waste.  Okay. We will specify in
the rule for certain kinds of--like for surface impoundments and for landfills and for
just generally contaminated areas, we specify in there what we think are the absolute
minimum requirements for the number of samples.

And we also discuss a little bit what representative means, but generally we defer to
our SW-846, our sort of Bible of methods where we discuss what representative
sampling is supposed to mean in there.

But I'll be honest with you.  The whole business  of the word "representative" to me is
a very big problem, particularly  in the characteristics program where somebody goes
out and--with respect to the TC, whether or not you have a TC waste is very much
dependent on where you sample.

And the burden's on the agency to prove that, in fact, what you did was correct, and
my own view is I'd like burden and I'd like to move it.

But the answer to the question is that we require that they make sure that  they have
sampled  their waste in a representative way, and then we talk about different kinds of
units that would hold waste or contaminated soil and we talk about grid and we talk
about random sampling and that kind of thing.

But I think you know that we really don't have any answer to that one and  it's really,
when they submit the sampling and analysis plan to you, that is the time at which
                                      76

-------
somebody is going to have to take a look at it and try and make a judgment as to
whether or not that's reasonable.

And we're not going to try and say what that  is right now.  I mean, in the rule itself
we give some general information.

Question: Stan, I'd just like to add that in some cases, also, we're talking about
cleanup criteria after treatment.

So I don't know that I would necessarily say that would be used as a screen across
the site right off the bat.  We might be using  them after the wastes or contaminated
media had already been subjected to some form of treatment.

MR. PHILLIP:  What we're faced with is, we've  got the  site out there and there's dirt
and there's waste  and there's, you know, kind of a mixture.

And the question  is:  Are we-when we deal with this stuff, if we push it up in one
place and consolidate it with some other material, are we dealing with a  RCRA
placement?  Do we  have RCRA waste? And how much dirt you throw in with your
sampling program that  dilutes the results?  So I get the answer though.

MR. BREEDEN: You had one for Randy?

MR. PHILLIP:  Randy, maybe I missed it, but you take the volume of dirt  and you
ring the contaminant out of it and you throw  it into the ground water, right?  How
much ground water  are you throwing it into?

MR. MAY:  To be very honest with you, I'm  not too sure what dilution they're
allowing.  They use up-slope recharge and they use infiltration.

And my understanding is that it's over a very  short period and a very conservative
dilution model.  I am only familiar directly with a model that we used~or a standard
we used to evaluate leachate from incinerator ash and where  would tolerate that in
terms of aquatic toxicity, and it was incredibly conservative. It was like a very large
amount with one-base dilution.

MR. BREEDEN: Another question?

Question: This is for Randy.  You mentioned that you supported the idea of a range
of cleanup goals as in Pennsylvania. How do you enforce a range of cleanup
standards?

MR. MAY:  Of course, I think again that's a  fundamental regulatory difficulty
question.  My personal philosophy is that good people are infinitely superior to good
regulations and that it's a reasonable thing  to use a variety of techniques and then to
entrust to quality  regulatory people to make that decision.
                                      77

-------
And we've designed programs like that I'm very pleased with.  Now, admittedly both
the industry and Environmental Defense Fund don't like that because it doesn't say
which end of the range you're normally going to get to.

We found though in a number of cases in this--I don't want to spend too much time
in this, but our staff has found in several different areas that very often when you get
the range from reasonable procedures, the range is not so great you can't go with the
conservative end and be happy with it.

MR. BREEDEN:  John?

Question: The fellow from Connecticut again. You had mentioned or alluded to
land-use controls in the context of ground-water-protection regulations.

Could you elaborate on that? Is that enabling legislation at the state level? That's
usually a function that's associated with municipalities or lower levels of government.

MR. MAY:  Yes.  We do have enabling legislation at the state level.  It's highly
controversial.  Needless to say, Connecticut is a very Yankee home-rule state.  I
really do like it. I was kidding about the climate, and it goes very much against the
grain of our state government and all of New England certainly, if not  all of the
United States, to allocate to the  state any kind of land-use  control.

We've chosen to do it in these major contributing areas to  our major public-water
supply, ground water areas, stratified drift aquifers.

As I have drafted the regulations, they are very restrictive in terms of what are
allowable new land uses and they place an enormous amount of regulatory control on
many existing land uses.

For example, an automobile-repair facility could never work on an automobile
outside.  I mean, down to that level of restriction.

Now, those have not yet gone to public hearing.  They're in a public process, though.
It's a very interesting process.

MR. BREEDEN:  A couple more questions before we call time? Yes, sir.

MR. SANTOS:  Chris Santos from EPA.  You mentioned the question with regard to
the health-based, the new proposal,  the health-based standards.

When the official rule was derived, EPA used plans of about 500  feet from the soil.

MR. COCHRAN:  That's correct.
                                      78

-------
MR. SANTOS: Do you intend to use the same plans or will it be in compliance at
500 feet from the site?

MR. COCHRAN:  Let me back into an answer.

The reason for having the 500 feet was to try and take into consideration some
dilution-attenuation, and we use the model to figure out what happens in that
500-foot span.

And by the way, that 500 feet was determined by the landfill data that we had and
where there might be sources of-where people were actually look at drinking-water
sources.

But the answer to your question is, basically it's at the point where the constituent
would enter the ground water because we make no dilution or attenuation
assumptions.

So it's straight at the health-based level.  So the effect of that is, at the point where
the constituent would enter the ground water, that's the point at which the regulatory
level comes into play.  So there's no dilution-attenuation.

Question: For Peter of Washington state.  What's the basis for the Table A values?

MR. KMET: There are footnotes on there explaining each number. Some of the
numbers are based on a direct contact as a control.

In many cases, the hundred times ground-water number actually controls, which has
been alluded to here by others.  So a number of those numbers.

You'll notice when you look at the rule there's a table for industrial and residential,
and those numbers are identical in many cases. It's because ground water's
controlling.

Some are based also on, I believe, plant toxicity.  Dust inhalation is another one.  So
it varies depending on what we thought.

They're also adjusted for the practical-what is it, PQL, practical quantitation limit.
So in the  case of PCBs, the actual 106 risk level would be lower than that.  But when
you adjust it to the PQL, that's where we went.
                                       79

-------
                PANEL 3
COMBINATION APPROACHES-STANDARDS WITH
          SITE-SPECIFIC OPTIONS

-------
                             INTRODUCTION
                              Allen Wolfenden
MS. BARRY:  Okay, I'd like to start the third panel.  Our moderator is going to be
Allen Wolfenden. He's the chief of the Technical Services Branch, Department of
Health Services in the California Toxic Substance Control Program.

MR. WOLFENDEN:  Thank you.  We have five speakers today to talk about some
approaches dealing with their state programs and case identification of some soil
cleanup levels.

And we'll have Mr. Pagan, Dave Pagan, from the RCRA program talking about the
RCRA program corrective action, which is a developing program and very analogous.

And you'll have half a talk from the state of California at the end about the standard-
setting program, and we'll identify some of the gaps that we have in our program and
the overall framework which we are putting together to develop some site-specific
cleanup numbers and approaches to assessing sites.

So without any further discussion of the state program, I'd like to introduce  Lynelle
Marolf from the state of Michigan.  She's the assistant to the  deputy director from
the Michigan Department of Natural Resources.
                                     80

-------
                                 MICHIGAN
                                Lynelle Marolf
MS. MAROLF:  By way of background, I'm working right now in the director's office
as assistant to the deputy for environmental protection.  But until this past July I
worked in our site cleanup program and was responsible for drafting our state
cleanup regulations and that is the reason that I was drafted to be here today.

The history behind the development of our regulations really began in November of
1988.  The voters passed a $425 million bond issue to provide for publicly funded
cleanup of sites in our state that wouldn't be taken care of by responsible parties.

We felt a really strong need to provide accountability in the expenditure of that
money by explaining  to people through the rule-making process how far we would
take the cleanups at the sites that we were going to be paying for with public money.

It's also true that program maturity both in the Superfund program and otherwise
really called for the setting of cleanup standards at that point in time. We had a lot
of sites that had gone through RIs and partially through the FS phase of the
Superfund program and we needed to be explicit with EPA about what our ARARs
would be at those sites.  And some of the state regulations that we were holding out
as ARARs were getting a lot of static from EPA about whether or not they really
constituted ARARs under the Superfund program.

So to  make clear exactly how we intended to approach cleanup at Superfund, state-
funded, and privately funded sites in the  state, we undertook the development  of
these regulations. And, really, we were faced with the same dilemma that almost
every  other speaker here today has talked about. There is, I think, a kind of
dichotomy between a group of people who, when faced with a cleanup situation say:
"Give me a number, just tell me how far to clean this up;" and other people who
demand flexibility and want to be able to take into account site-specific circumstances
in setting a cleanup standard.

For your information, there are excerpts  from our rules that include the cleanup
standards in the packet of materials that are in your folder.  I don't know whether
you're  really going to want to read along with me, but the material is there if you
want to refer to  it. If you're interested in getting the complete set of regulations
which has some  other administrative process steps in them, I'd be happy to mail  them
to anybody who  leaves me their card or calls me or what have you.

What I'd like to do this afternoon is describe to you how our rules work, what the
different approaches  are that we've laid out in our regulations, and then reflect a
little bit on how the issue that we're focusing on in this conference really ties into the
way we've approached our rules.  I'll do that primarily by focusing on one approach
                                       81

-------
to soil cleanup that is, I think, really the most important part of the regulations that
we developed.

In general terms the rules require remedial actions to be protective of the public
health, safety, welfare and the environment and natural resources. And that's an
outgrowth of all of our environmental statutes in the state. The way that we have
approached this flexibility concept and the dichotomy between giving people firm
numbers to work with and allowing flexibility is to provide for three different types of
cleanups, which is quite similar to, I think, the way that Washington is approaching
this.

And I should tell you too that when we were doing our rules, we were probably about
six months ahead of Washington and I was talking regularly with the fellow  there who
was their primary author, and at one point he called and he said:  "I'm just,  you
know, phoning to check in and see where things are."

And we said: "Oh,  well, we got a major redraft done and that will go into the mail
tomorrow."  "Well, can you mail mine tonight, because we  need to go public with our
next revision in three days and if you've changed anything  in a big way, I'd like to
know?"

So I think that there was a lot of information exchange between a number of states
that were all involved in this same process at the same time.

We did account for what we call three different types of remedial actions cleverly
labeled  A, B and C. The type of remedial action plan is proposed by the party
undertaking the cleanup, but is subject to approval by our  agency. A remedial action
plan can incorporate a combination of different types, which I think will become clear
as I go through the explanation of the different  types.

We also have some general provisions relating to the cleanup of ground water.  The
legal  framework in  our state is quite similar  to that in Pennsylvania where essentially
ground-water discharges which have the potential to become injurious  to the public
health, safety, and welfare are prohibited.  And so, given that framework, from our
water resources protection statute we added some general  provisions that apply
regardless of the type of cleanup that's being proposed.  And in a really quick sense,
those general ground-water-protection provisions require cleanup of any material
present  in an aquifer, either through active remediation  or through some sort of
biological or chemical process that can be  documented to be occurring at the site.

We also require that, at the time that ground-water remediation is undertaken, that
the vertical and horizontal extent of the  plume be stopped from migrating any
farther.   That plume extent is defined by our type B levels which are essentially
thresholds for contamination definition.
                                       82

-------
In developing the three types of cleanups, we really try to approach both type A and
type B as what I think of as a cookbook approach to cleanup.

We also  have a site-specific approach to cleanup, very much like the Superfund
process, which we call type C. We did that in anticipation of more complex sites and
the fact that a number of the PRPs that we're dealing with really want to be able to
undertake detailed site evaluations and site-specific solutions.

Our approach to cleanup prior to the time that these regulations were implemented
was to require all sites to be cleaned up to background or to uncontaminated levels,
which really isn't a very realistic policy in the long term.

Clearly there are some sites that can't be cleaned up to those levels: massive
landfills, lots of extensive ground-water contamination sites.  It's  really going to be
difficult to achieve the kind of standard which had  been our public posture up to that
time. So our approach under these regulations largely represents a shift from that
sort of single standard to a risk-based approach to cleanup.

Briefly describing the three types, our type A cleanup is to either naturally occurring
levels or method detection limits.

Type B is a risk-based approach that relies on some presumed exposure assumptions.
The concept behind a type B cleanup is, at the completion of the remedial action, we
would be able to walk away from a site and feel comfortable that its future use could
be for  any purpose without restriction. So that's basically taking into account a
residential scenario.

And then the type C approach is the most complex.  It allows for a site-specific
assessment of risk and allows for some kinds  of waste containment alternatives  that
really aren't possible under the other two scenarios.

I don't want to spend a whole lot more time on type A. It basically is just a
restoration of the environment to its natural conditions. I'd  like  to spend  the most
time on type B.

Our type B criteria for cleanup of ground water in aquifers,  is to a one in  a million
risk level for carcinogens,  to a health-based standard of human life cycle safe
concentrations for non-carcinogens.  That equates to the process used by EPA in
setting an MCLG under the drinking-water program.  And then we also have criteria
that take into account aesthetic impacts on ground water.  So if a health-based
standard would result in some sort of taste or odor  restriction on the use of that
ground water, we require an additional degree of cleanup to protect against that
impact on aesthetics.

Our approach to type B criteria for soils, I think, is really unique among the states as
far as I know. Because we don't regulate the soil directly in Michigan, and, again,  I
                                       83

-------
don't think that any other state does either, we took an approach for type B soils of
looking at the impacts of contaminants in soil on ground water, on surface water,
through direct contact, through air emissions, and then in a category that we call
"other" that can take into account things like  potential food chain impacts,
phytotoxicity, agricultural impacts, things like that.
The approach that we've taken to the issue of protecting ground water from
contaminants in soil is to look at essentially a leachate-generation process.  We
require that the contaminant levels left in soil have to be below a level that produces
leachate equaling our ground-water standard.  TCLP as specified in the rule is one
test that would always be acceptable to the department.  We accept others  on a
case-by-case basis as long as they are documented by the party proposing the remedy
to accurately represent in situ conditions.  And I think in certain circumstances TCLP
is more aggressive than a typical in situ test and other people proposing alternative
tests.

I'm looking forward to the 8020 acid test that I think the fellow from EPA  spoke
about just awhile ago.  That's one that was in developmental stages when we were
writing our regulations, and it really appeared to have a lot of promise for the
purpose that we are applying it.

We get some criticism for using TCLP in this  context because it's not intended as a
test for soil, but rather for waste material.  We recognize that. We know that TCLP
is a relatively aggressive test, but we feel that  that's appropriate for us as a posture
for regulators because it gives us conservative results.

Interestingly, we also chose sort of an arbitrary factor  to relate our ground-water
standard to soil concentrations that are measured in total. Our factor is 20 as
opposed to Washington's factor of a hundred. Ours is just about as scientifically
based as theirs, which is to say that we chose 20 basically because the water-soil
dilution ratio in most of the commonly used leachate tests is 20 to one.  And with
that water-to-soil dilution  ratio, you're not going to see measurable results in any
circumstance where you've got  a level of more than 20 times your ground-water
standard.

I probably should explain more clearly that the way we approach this is to look at the
five different potential pathways for impacts of contaminants in soil.  The one of
these that produces the most restrictive limit is the one that  controls.  In almost every
case, that is the ground-water protection number.  For  certain kinds of materials like
PCBs and some of the other metals that have serious concerns for direct contact and
which are not highly mobile, I think that the direct contact numbers are going to be
controlling.

We're  also doing some evaluations right now about air emissions of metals  in dust,
and it would appear that that's going to be a controlling factor in some of the
                                       84

-------
circumstances as well, which is something that none of our technical staff really
expected that's taken us somewhat by surprise.  We'll have to see how that develops
as we go on.

The second general criteria that we apply is that we don't allow the transport of
contaminated soil into surface waters that would result in a violation of our water
quality standards. We allow no mixing zone. Again, a conservative assumption, but
because of our need to make this a really kind  of cookbook approach to cleanup, we
chose not to use a mixing zone.

The next criteria for soils is that we allowed no air emission that results in greater
than a one in a million risk on-site.  And for non-carcinogens, a no-harmful-effect
level, which would be comparable to the no-adverse-effect level that we allow in
ground water.

The fourth criteria is no dermal contact or ingestion risk, again, greater than a one in
a million risk or human life cycle  safe concentration.  The algorithm that we  used to
calculate that direct contact hazard is actually provided in the rule.  It is a
combination of a lot of work that had been done previously by EPA in the Superfund
program, pretty much reduced and simplified into  a single equation.

That was one of the most controversial areas that we approached during our rule
making.  There was a lot of resistance on the part of the regulated community and
initially even from the environmental advocacy community and the advisory group
that we used. They didn't like the idea of committing to rule something that wasn't a
broader scientific consensus. But we were able to achieve agreement on the
algorithm in our rules now. We don't have a comparable algorithm for the air
emissions or  for the movement of contaminants into surface water.

I think our rules would be better if we had those kinds of algorithms,  but we were
really under some time pressures to develop our rules and didn't feel  like  we had
algorithms that would have the degree of confidence and applicability that we had for
the other media.

The last category that I mentioned earlier is what  we call the "other-injury category".
We can take into account here things  like  toxicity  and agricultural impacts.  There's
one example I can think of, actually two.  Sodium  and chloride are  common
contaminants in a number of contamination sites in Michigan.  We've got  about 3,400
sites.  Sodium and chloride are not common contaminants at any of our Superfund
sites,  but we  have some brine production wells in Michigan and pipeline problems
and production site problems at those kinds of facilities that have given us some
pretty serious sodium and chloride contamination problems.  The most significant
effect there is the effect of sodium and chloride on the soil structure relative to the
ability to grow crops.  It's something substantially less than a health-based level based
on either direct contact or leaching or any other kind of health-based impact. You
see a collapse of the soil structure in a way that prohibits the growing of beans, and
                                       85

-------
beans are the primary crop in the area where all this brine extraction is taking place.
So for those particular contaminants, this other category would be the controlling
factor in setting our soil cleanup levels.

For our type B criteria for surface water; now, this is apart from impacts that
contaminants in soil have.  Our type B criteria for surface water is no violation of our
water-quality standards  as a result of the natural migration of ground water into
surface water.  We don't have more specific criteria  for surface water sediment under
type B.  Any site that involves a significant amount of surface water sediment as part
of a cleanup plan will have to be handled as a type C.   We looked at trying to put a
more cookbook approach for surface-water sediment into these rules but really
weren't able to achieve  consensus on that matter either. We have 14 areas in the
Great Lakes and Michigan which will present really  substantial issues relative to
surface-water sediment  cleanup, the Areas of Concern that have been identified by
the International Joint Commission and EPA and  all of those sites will have to go
through our type C process. This is the most simplistic approach really to looking at
the impacts of ground water on surface water.  We don't allow a violation of water-
quality standards, and again, no mixing zone.

Our type B criteria for air quality are the same off site  as on site.  No violation of
our air pollution act.

And then finally our type C criterion which  apply to all media (they are not media
specific) is simply that the cleanup criteria are determined on the basis  of a
site-specific risk assessment.  Under type C we  can allow for containment or
exposure-control measures, which are not allowable under type A or type B.

Again, the idea behind the type B cleanup is that we can allow for any kind of use
without restriction. A type C cleanup could be something like containment of a
landfill or at an industrial site,  could involve something like an alternative exposure
scenario which I think we'll commonly see in industrial  sites where the residential
setting is really not an appropriate thing  to base the  calculations on.

There is not explicitly a range of risk specified for type  C cleanups. We would
entertain proposals that moved away from the basic  level of risk that's specified in
the type B approach, one in a million and human life cycle safe concentration. We
wanted to provide for that not so much to open up the  possibility of dramatically
different cleanups under type C, but to acknowledge that in certain circumstances the
technical limitations of a particular approach to cleanup may take you very close  to
our one in a million risk level, say 1.2 x 106. We wanted the flexibility to
acknowledge that as a complete cleanup  without having to otherwise tamper with our
standards under type B.

I think that we'll quite often see alternative  exposure scenarios presented by
responsible parties at  industrial sites and for sites where containment  really is the
only alternative as well.
                                       86

-------
Consideration of cost effectiveness is also part of our analysis for type C criteria.  It's
not something that will be taken into account for type A or type B.  Those require
cleanup to the fixed standards that are produced by the algorithms in the rule or by
background or method detection limits without consideration of cost.  So if cost is a
factor which makes it impracticable to achieve those levels, then it's the responsibility
of the party doing the cleanup to propose a type C approach.

We also have some additional requirements for type C actions that involve primarily
containment,  and that is that the party proposing the remedy has to  assure long-term
monitoring to demonstrate the effectiveness of the remedy. And in  most cases we
would require financial assurances, deed restrictions, or other institutional controls if
that's necessary to assure the integrity of the remedy.

We've done those things in part as a disincentive.   I think it may force people to look
at the economics of their cleanup in a somewhat different way.  If at the outset of a
cleanup they've got to put up enough  money to provide for operation and
maintenance or in other ways assure the integrity of their remedy, they may look
differently at  the cash flow associated with doing a cleanup.  We may get a type A or
a type B cleanup without having to sacrifice the restricted use of that land over the
longer term.

I had hoped to be able to give you a couple of case studies.  We talked about how
our rules are  really working in practice.

They've been effective since July this past year, and surprisingly, we've only gotten
about 20 plans from  the regulated community proposing to apply our rules at
cleanups.  I think that's in part because the consulting engineering industry is; a little
bit behind the curve  in terms of understanding what these  regulations  require and
allow for.  And consequently, I think what's really happening is that a lot of people
are doing type B cleanups but they're going to come to us  after the  fact and
document what they've done as a cleanup as opposed to coming to us for approval
before they proceed with their cleanup.

That's something that will work well under our current regulatory framework, but in
July of this coming year we'll have some major changes to our cleanup statute  take
effect which will allow us to require a plan to be submitted before the responsible
parties proceed with  the cleanup. So  that will give us a little bit more leverage to get
the horse  before the  cart, and right now we've got  a little bit of the  cart before the
horse.

With respect  to the issue of modeling, which has been obviously a subject of some
discussion today, our technical staff is really uncomfortable with the idea of modeling
in the vadose zone in particular.  They've had  some bad experiences similar to the
ones that  were described from Connecticut, I believe, where what we saw in reality
just didn't match up  to what had been predicted by the model.  Our rules allow for
models; actually it's not excluded under our rules, but the receptiveness of our
                                       87

-------
technical staff to making decisions on the basis of any kind of vadose-zone modeling
in particular is really very limited. I think that they would be open to those kinds of
things if they had some more confidence that-particularly that it wasn't going to cost
more to develop and assess what the model tells you than it  does to just do a more
aggressive cleanup to begin with.
That is a position among our field staff, that the cost of acquiring data for anything
less than the most complex cleanups, in order to run a model, is less than the cost of
just doing a more extensive soil cleanup.

Thanks very much.
                                       88

-------
                                MINNESOTA
                                Jim Pennine
MR. WOLFENDEN:  Thank you very much.  The next speaker will be Mr. Jim
Pennino from the state of Minnesota.  He's a hydrologist with the Site Response
Section, Minnesota Pollution Control Agency.

MR. PENNINO:  About 18 months ago I was  assigned to a committee working on
developing soil cleanup guidelines for the Superfund program for the state of
Minnesota. And for the benefit of two of my supervisors who are in the audience, I
know I promised that I would finalize this thing last spring, and I know I promised I
would finalize it this past December, but I really, really will finalize it by April.

In Minnesota we have our own Superfund regulations similar to CERCLA, and we
have in our approach, each hazardous waste site is assigned to a project team and the
project team is responsible for overview of the site investigation feasibility studies and
coming up with cleanup goals.  Unfortunately, there were no specific guidelines  on
what kind of cleanup goals should be made, so consequently each project team came
up with their own cleanup goal.  We haven't achieved cleanup  or we haven't arrived
at the point where we're developing cleanup goals for many of the sites. We have
about 160 sites on our state program.  A number of those are federally funded sites
or on the  federal NPL.

But a few of our sites have have had some cleanup goals developed, and in the first
few years  of the program, some of these  cleanup goals were, you go out and you dig
until you don't see any more visible product in the hand samples in the field, and
that's your cleanup goal.  And that sort of works when you have an oil or a gasoline
spill because some of the contamination  is kind of obvious. But it's uncertain
whether that's really protective of ground water.  In some cases it may not even be
protective of other routes of exposure.

More recently, the Minnesota Department of Health has published some draft
standards  for soil ingestion that have been used at those  sites.  It's based on if a child
were to be in a playground exposed to contaminated soil or ingest the soil, you have
a cleanup level that-or a concentration level that's allowable in the soil that would
not harm the child.  A few people have tried to use those standards.  Most of them
are for metals, but we have a couple for  organics.  And they've tried to adopt the
standards  as overall cleanup goals even though they weren't intended to be protective
of ground water.  They may not be protective in terms of inhalation of soil
particulates contaminated with the organics or metals.

So this draft document that we've been developing is designed to provide a little
more scientific basis for our cleanup goal decisions. Basically we're using a risk
assessment general approach and this follows some of the guidelines in U.S. EPA risk
                                      89

-------
assessment documents.  The project team is required to evaluate each of the possible
routes of exposure, including drinking water, soil ingestion, volatilization in the air
into the breathing zone, particulates, runoff into surface streams, and other ARARs.

After they examined all those routes of exposure, they picked the most restrictive one
and they used that as an overall cleanup goal for the site.

This draft document focuses primarily on cleanup in terms of being protective of
ground water. In the past, using the project team approach, a lot of our cleanup
goals have been somewhat inconsistent.  We have tried to evaluate some of those and
I've looked through some of the historical records on that, and it appears that a lot of
our cleanup goals are sometimes very restrictive. They seem to be driven more by
cost and expedience, and in  some cases the negotiating skills of the project team
when they face the responsible party for the cleanup.  And that ends  up  causing a lot
of inconsistency in cleanup goals.

What we're trying to do now is  provide a process by which we can derive cleanup
goals for each site, and it will be basically one process, hopefully flexible, that will be
used at  all sites. But the cleanup goals will vary for a given chemical from site to
site.

Basically the procedure involves a set of several equations, and what we're trying to
do is  account for some processes in the soil.  I'll try to draw the diagram here. If this
is the surface of the ground, many of our sites have had interim responses where
there may have been some contamination in the soil and  the surface contamination
has been removed and possibly incinerated or  shipped off to another site, but we
didn't know what to do with all the  contamination so there's still some contamination
left buried.  In some cases we still have contamination all the way to  the surface.

But the conceptual model we're using here basically assumes that there's a known
extent of soil contamination under the site and that there are certain processes going
on. There's a certain amount of adsorption which is occurring at the site that the
contaminants, particularly organic contaminants, are permanently adsorbed to the soil
particles. So some of those  are going to stay there. Another process is volatilization,
where the assumption is that some volatilization will occur for organic compounds so
that some of these will volatilize into the soil air space or possibly to the atmosphere.
There is also the assumption that there's probably some bacterial activity in there
that's going to degrade some of the  contaminants at the site.  After these processes
have taken their toll on the contaminants, what's left is available to form leachate
that will move down to the water  table.  This  is leachate formed by percolating rain
water that will move through the contamination zone, dissolve some of the
contaminant, and carry it down  to the water table.

The cleanup  goal is to be protective  of ground water, and we're assuming the point of
use is the point immediately underneath the contamination zone, and that's limited to
the extent of the contamination. Since most of the ground water is used for drinking
                                       90

-------
water purposes, or that's basically the assumption, we use the EPA maximum
contaminant levels or recommended allowable limits, which are tentative standards
developed by the Minnesota Department of Health.

They're a more comprehensive list of organic compounds and metals and other
substances that the Minnesota Department of Health has developed for us and
they're health-related values. We're assuming that since the point of use is
immediately beneath the site, that in order to be protective of this point, we
therefore calculate the amount of leachate that's going to be produced.

We plug the MCL, the maximum contaminant level or recommended allowable limit;
or in some cases if the ground-water resource is pristine and a very important aquifer,
we may adopt a non-degradation standard for that aquifer. In that case we would use
the quantitation limit or the detection level as the cleanup goal or the basis for the
cleanup goal.

What we then do is  take the cleanup goal and the ground water, MCL, plug it into
the leachate equation, and back-calculate through the contaminant zone to a
concentration of contaminant in the soil that, if left in place after other remedial
measures are taken, produce sufficient leachate to contaminate the ground water
above the maximum contaminant level or other ground-water criteria.

We've used these equations for a couple of sites in Minnesota to see how they work,
and I can give you a couple of typical values. We came up with a cleanup goal for a
sandy soil with a minimal amount of organic matter, contaminated with PCBs, in an
area where the water table is relatively shallow, only about five or six feet below the
surface, and we came up with a cleanup goal for PCB at the site of 2 to 6 milligrams
per kilogram (mg/kg).  And the range is based on the assumptions.  We didn't have
actual site data on the amount  of organic matter in the  soil, so we figured there was
a range of organic matter that may be there.  And based on that,  on the lower end
you'd have a two milligram cleanup goal for PCB, and at the higher end, a six
milligram.
In some cases  we're using default values when we don't have enough site-specific
information. And where you don't have enough site-specific information, our
guidance, I believe,  will be that you use the lower end of the range. And what I
think we intend to do, and it's going to be up to management to some extent, is that
when the responsible party wants to know what cleanup goal we want, depending on
how much money they were willing to spend on getting  site-specific data for our
calculations, we'll give them a number we feel comfortable with.  We won't give them
a range and let them pick something in there.  We'll work out a range  and then we'll
give them a number we feel comfortable  with in terms of how much confidence we
have in the data based on the complexity of the site, based on how much confidence
we may have that the responsible party will be able to monitor the site. So we
probably will be dealing with ranges, but  the public may not see those.   It's going to
depend on the type  of confidence we have in our data.
                                      91

-------
I don't know if I really explained it, but we're intending to use as much site-specific
data as possible.  Many of the parameters that were discussed earlier today, including
things like soil density, organic matter content; will be taken into consideration.

One area that we have the least amount of information on is biodegradation rate
constants, and that's one of the things that we're trying to use in our calculations.
There's very little published data.  There's even less field data on biodegradation
rates. And  one of the things we're thinking about doing is requiring test studies done
at all sites where there are biodegradable compounds in the soil and have the
responsible  party develop a biodegradation rate for the site and also some more
specific information on volatilization.

But some of these may be kind of a wish list.  I don't know how practical it's going to
be to require these things, but we're hoping that there will be more of this
information developed by Ada and other research facilities. Right now we're trying
to account for as much of the site-specific characteristics as possible in our cleanup
goals. And  that's all I had.

MR. WOLFENDEN: Thank you very much, Jim.
                                       92

-------
                               NEW JERSEY
                               Dr. Kate Joyce
MS. JOYCE: New Jersey's Department of Environmental Protection and Energy will
not be submitting a manuscript from the presentation at the Workshop in January.
In May 1991, New Jersey DEPE released draft preliminary cleanup standard
regulations to solicit comments prior to formal proposal.  The cleanup standards
contain numeric criteria for building interiors, surface and subsurface soil and ground
water.  Over 1,000 pages of comments were received from the regulated community
and environmental consultants. The draft standards were also peer reviewed by the
regulatory and academic community. Substantial changes have been incorporated
into the standards, rendering moot much of the approach presented at the January
workshop. Target date for proposal as regulations in the New Jersey Register is
December, 1992. Information concerning the cleanup standards can be obtained
from Dr. Kate Joyce, Division of Publicly Funded Site Remediation, (609) 633-1348.
                                      93

-------
                       RCRA CORRECTIVE ACTION
                                 Dave Fagan
MR. WOLFENDEN: Our next speaker is Dave Fagan.  He's the chief of the correc-
tive action policy section.  I guess Dave has had an opportunity to watch Superfund,
so he can pick and choose all the good stuff he wants to take out of it for the
corrective action program and hopefully leave behind the stuff that didn't work so
well.

MR. FAGAN:  What I want to talk about is corrective action.  Hopefully we have
learned a few things from Superfund.  The corrective action rule making, proposed
rule making was proposed last July. I have a certain sympathy with the people in the
states who are still struggling with getting things in their publications.  This rule took
us--well, it had a fairly long gestation period.  Not only did it take us a while to get it
through the  agency, it took us  20 months to get it through OMB.  And so as a result,
some of the main policy things in there are, say, three to four years old.

Subpart S deals with a number of issues having to do with soil; and in some
specificity, the soil-to-ground-water issue. And so I'll run through a few of these
things, and I think what I'll also be able to do is identify some of the big issues that
remain and  some of the things that hopefully this group  or this meeting will put into
motion and  give us  some answers for as we finalize this rule  over the next five  to ten
years.

Okay, a couple of main themes in the Subpart  S rule.  I  want to talk about action
levels and cleanup levels and exposure assumptions and  deep soils in no particular
order.

Action levels-this is a regulatory concept that I think is still unique to Subpart S, is
still unique to RCRA. This is a set of specific numbers that  are concentration  levels
for each medium based on lifetime exposure, and the idea here is that these are
constant, conservative, lifetime-based numbers  that constitute, in essence, a rebuttable
presumption. So when we find contaminants above these action levels at a RCRA
facility, we at least are able to stand up and say that there is presumption here that
there is a potential risk.  And the reason why we thought we needed these action
levels,  these trigger  levels, is so that we can say that there is at least a potential risk
here so we can require the next phase of the study, the  corrective measures study,
which is analogous to the Superfund feasibility study.

These numbers are, in essence, trigger levels.  They're not cleanup levels. They're
trigger levels for making an owner-operator at  a RCRA facility go the next nine
yards.
                                       94

-------
In terms of the numbers themselves, they're based on the agency peer-reviewed
numbers, the dose-response numbers for toxicants and carcinogens. The reference
doses we're using for the toxicants, and cancer slope factors for carcinogens.  For
class A and B carcinogens, we're using 106 as the action level, and 105 for class C
carcinogens.  We're getting a lot of comments on the 105 for class C's.  We'll
probably be looking  at that issue again.  But that's where that worked  out.

The exposure assumptions  that lie behind these numbers are really the key to them.
And I think the exposure assumptions are really the main issue behind these
numbers. What we have done in setting up these action levels is assume basically
residential  exposure  as our starting point. So we're using the 0.2 grams per clay for
toxicants, 0.1 grams per day for carcinogens.

I need to underscore the fact these are not cleanup levels. The  other point about
action levels  is the issue of location,  or "where".  In fact, the where has been a much
bigger issue in developing this set  of regulations than the issue of "how much". So for
the action levels we are saying that they  are measured in a surficial layer, and that
brings up this issue, how surficial is surficial, and I'll talk about that in a minute.

Not to be confused with action levels, the cleanup levels are specified in the rule.
Basically we're giving ourselves, the regulators, and the states and the regions, as well
as the regulated community, a little more flexibility. We're basically using the action
levels as the  point of departure. This is  Superfund terminology.

The mix range is  hinged on the presumption that no one level of risk is necessarily
appropriate for the incredibly wide variety of facilities that we have. We are finding
new ones all  the time and new facilities every time we check the data base,  but
currently we're dealing with somewhere between, I'd say,  4,000 to 6,000 facilities
when all is said and  done.  So we've got lots of little problems, lots of big problems,
and lots and  lots of in-between problems. Each one is different and I  think 1hat still
argues for going with a  system that allows a certain amount of site-specific flexibility
in setting these how-clean-is-clean numbers.

So what we've got is a risk range for carcinogens, 104 to 106.  This is fairly  familiar
stuff.  I think the agency is pretty much wedded  to this idea of the risk range.  At
least I haven't heard any serious talk recently for trying to narrow down to 105 or
something like that.  I think we're going to maintain this kind of flexibility.

Again, the  point of compliance, and  this is the where in this program,  is the surficial
layer.  And in the proposal we stuck our neck out a little, and I  was heartened to
hear that New Jersey did the same thing and said that the top two feet is "surficial".

I guess  the timing too, and maybe this is outside the scope of this particular
gathering, but the timing of how soon is soon, if you will, when do these facilities
have to get cleaned up to these levels, is maybe  the biggest issue of all in the
corrective action program.  We have owner-operators that have  been operating, in
                                        95

-------
some cases, facilities for over 200 years.  We have pre-revolutionary solid waste
management units.  I think the timing is really where the RCRA program is
struggling with trying to accommodate some of these very real concerns about how
fast can we expect all this to happen, how fast can the agency get it to work, and how
soon is really soon enough.

So enough said on that. That is, as I see it, really the threshold issue as we work to
finalize the Subpart S rule.

A few other thoughts on exposure assumptions. Again, exposure is really tough to
nail down.  We started talking in putting this rule together about using
non-residential or other than residential exposure scenarios in setting these numbers
up for cleanup.  Again,  these are the health-based surficial cleanup levels.  These
RCRA facilities  primarily are industrial properties, very often located in large
industrial areas,  and so  it  doesn't make sense to always assume that this industrial
property in the middle of  this industrial area is going to go condo. That's another
thing that's kind of hard to sell. In some cases they do though, and that's what makes
this issue so difficult.

And we don't have any  magic wisdom for how to  make these decisions, but we think
that as we get a  little better handle on this, get some more experience with it, we
should be able to make some common-sense decisions about what kind of exposure
to assume when  we're dealing with facilities in certain locations.  It's not going to be
easy.  In fact, our regions  who are by and large implementing the program at this
point still are very uncomfortable about saying: "Here's a facility and we think that it
will always be industrial; we don't think this will ever be developed". I anticipate a
lot of difficulty in loosening this up, at least in all but a few  cases.  But we'll see.

Getting back to  the deep  soils,  how-deep-is-deep issue.  What  the rule says, what
Subpart S says, and this gets to the ground-water interface, is that we will deal with
non-surficial or deep soils essentially as a source.  We will deal with that not from  the
kids eating-dirt-exposure-scenario perspective, but we will deal with it as a transfer
medium.  So we will deal  with deep soils and require some cleanup or control of
them in such a way as to protect ground water or surface water from being
contaminated at above health-based levels.  Not background, but health-based levels.

How do we do that?  We  don't have any good clean answers.  Now, I'm heartened to
see states like New Jersey and others taking a whack at this, and I think hopefully by
the time  we finalize Subpart S, we will have enough experience under  our belt to get
more definitive and perhaps put something in regulations as  to what kind of models
to use, what kind of formula are appropriate.

That's it  for Subpart S.  One final thing.  Maybe it's of interest to this  kind  of crowd.
Some of you may be familiar; in fact, I think there's been some discussion about the
TC rule that got finalized  not too long ago.  There was a deferral for petroleum
releases from underground storage tanks while the agency tries to figure out how big
                                       96

-------
that problem is and what the implications are of regulating what could be a very
large quantity of contaminated soil.  Just the implications of having suddenly these
huge volumes of hazardous waste come into the RCRA system is something that-
well, it causes people to think twice.

We got a petition from the state of New York that pointed out that giving this
exemption to soil contaminated from petroleum releases from underground storage
tanks was only a small part of the problem, that there's soil all over the place
contaminated from all kinds of sources of petroleum, pipelines and transfer stations
and transportation spills and this and that and the other.  And so they put together
some fairly persuasive arguments as to why we ought to perhaps broaden this
exemption for petroleum releases from different sources.  We are in the process of
looking at that and trying to collect some data and trying to make some decisions
about how to go with this, but it's likely that you will see  a Federal Register
sometime soon that addresses this particular issue.

Allen, I think that's all I have to say.

MR. WOLFENDEN:  Thank you very much.
                                       97

-------
                                CALIFORNIA
                               Allen Wolfenden
MR WOLFENDEN:  At the January conference I provided a summary of two
California State Department of Toxic Substances Control (Department) initiatives
that address the State site remediation program and soil cleanup.  Since my
presentation at the conference, there have been significant changes in our approach,
thus making the presentation out of date.  These changes have been the result of
several public workshops.

The initiative has move forward in two parts.  The Integrated Site Mitigation Process
(ISMP) describes how the site  remediation process will be managed, commencing
with site discovery and concluding with site certification.  This process is described in
a document titled "Draft Integrated Site  Mitigation Process".

The second portion of the initiative was  the development of Scientific and Technical
Standards for site remediation. Initially, a comprehensive set of standards were
developed for site characterization, exposure assessment, toxicity assessment and
selection of soil remediation levels.  The intention was to promulgate significant
portions of these standards as  regulations.

Many concerns were raised by industry and environmental concerns about this
approach.  As a result of these concerns, it was decided to redraft most of the
standards as guidance documents and promulgate only a set of performance
requirements as regulations.

The Department is currently in the process of reviewing these Draft Standards to
select out the critical performance standards that will go into regulations.  The
proposed regulation package will focus on three critical areas:

       1.     Site Characterization
      2.     Toxicity Assessment
      3.     Risk Assessment

The proposed site characterization regulation, which is currently in early draft stages,
will set  forth the minimum requirements for characterizing the areal and vertical
extent of contamination.  It will establish criteria by which to determine which
potential routes of exposure must be evaluated and how much data must be collected
for each route of exposure. The Toxicity Assessment regulation will set forth
procedures for determining the relative toxicity of cancer  and noncancer producing
compounds.

The third component is the Soil Remediation Levels (SRL) development process.  It
is set forth as a computer program based on the exposure estimation algorithms in
                                       98

-------
the "Environmental Protection Agency Risk Assessment Guidance for Superf und
Volume I Human Health Evaluation Manual (Part A)", December 1989. The
program has been structured to evaluate ten exposure pathways.  Based on data
developed pursuant to the site characterization regulation, the SRL program will
calculate conservative soil cleanup levels.  Responsible parties can choose to
remediate their site to these standards or background in many cases without
completing a full health risk assessment.

A summary of the technical basis for each of the regulation packages is provided in a
document titled "Summary of Draft Scientific and Technical Standards".  The
guidance documents will,  where possible, reference existing EPA guidance, and
professional literature.  Guidance documents will be prepared for each regulation
package.

The Department anticipates having drafts of the proposed regulations and guidance
documents  available for review in early 1992.
                                      99

-------
QUESTIONS FOR PANEL 3

-------
                        QUESTIONS FOR PANEL 3
MR. WOLFENDEN:  Any questions for myself or the rest of the panel members
who I think are all still here?

Question: I have two questions.  In your table of metals or standards for metals, a
standard for the background was added to the list. My question is, has DEP
developed a statistical value  pertaining to background?

MS. JOYCE:  We haven't developed a statistical method,  although we do have a
large amount of data that tells us what the background is for the state for various
land uses. But we haven't established a method for determining what background is
at a site.

Question: So you'd use the available data?  You don't do it on a site-specific data
basis?

MS. JOYCE:  Right.  Eventually we will determine what that background level is and
put that in the chart.  It's one part per million for cadmium.

Question: My second question was for Lynelle. You mentioned in setting your
regulatory levels you assumed no mixing in the saturated zone, in the aquifer. Given
the fact that we set those to  maybe protect the ground water in the wells, and given
the fact that when the state agencies go out there to  sample and make a
recommendation, they draw a diluted sample of the well, don't you think that your
assumption is too conservative?

MS. MAROLF: It certainly  is a conservative approach. In part, the philosophical
basis for that is not so much a question of whether we want to take into account
dilution or not as it is the water resources law that underlies what we're doing with
our cleanup standards. We basically are trying to find a way to do a  cleanup program
within a context of ground-water protection law that  says non-degradation is our
ground-water standard.  So for us to allow some sort of mixing zone into ground -
water or a dilution to be considered is really inconsistent with the water resources
law that we have as a basis to work  from. So  it's  that as much as a technical
solution.

Question: Kate, your decision matrix lays out all the different pathways and you had
different sets of numbers that you had thrown up, like 17,000 for toluene and  100 in
another case. Do you go with  the more stringent standard, or for the first two feet
you get  17,000 and for the rest of it  that goes down to the  water you get a 100 parts
per million? How do you determine that?
                                      100

-------
MS. JOYCE:  Initially what we had in that comparison with field data box was
exactly that, that you would select the lowest pathway, the lowest criteria for the
contaminant. But it falls apart when you realize that the direct contact numbers are
only operative for the first two feet.  So if that number was lower than the air or the
ground-water number, you wouldn't use that number throughout the entire soil
profile.  You would only use that for the first two feet. But other than that, we
haven't made any blanket recommendation that the lowest number should be used.

Question: Will you make that determination later when you finalize the rule for that
lower number being used?

MS. JOYCE:  The question was, are we eventually going to make the determination
that the lowest number for any of the pathways that are operative will be used? I
would say, you know, leaving aside the direct contact pathway, that that's likely that
we would enforce the lowest number.  Because to get a ground-water number, you
have determined that ground water is a concern at the site, you know, previously.  So
I would say that's likely, yes.

Question: Carol Fox, Montana.  Lynelle, I have a question for you. You mentioned
that you had 20 companies that have come or proposed to come under your different
scenarios. Of that, how many proposed to do like  method B versus method C?  I
think a lot would just go with C.

MS. MAROLF:  Yes, I would have expected that too initially. In fact, when we came
back from the public hearing where our rules were approved, I fully expected that
there would be a pile of type C proposals on my desk. We've gotten two type C
proposals.  I believe that both of those were Superfund sites and the balance are
equally divided between type A and type B.

Question: Another question for Lynelle. This is Jennifer from Ohio EPA.

Regarding the A, B and C, are there any criteria laid out in the rules that dictate how
you are supposed to evaluate which one is a reasonable choice by the PRP?

MS. MAROLF:  There  are specific criteria that apply to  all different types that we
have to evaluate and then there are some additional criteria that apply to type C
proposals.  Our evaluation of a type C proposal is  really a balancing of a lot of
factors, including technical feasibility, cost, whether or not they've complied with the
overall standards that would-or the overall criteria that would apply for type A and
B, too.  We have criteria in our rules comparable to CERCLA in that off-site
disposal of untreated waste was the least preferred option, those kinds of things.  So,
yeah, there are criteria there to look at. They're in part six of the rules if you want
me to send you the whole package and let you look at those.  Because we did leave
the type C proposals so open-ended, it's the responsibility of the party proposing the
remedy to demonstrate that what they're doing is protecting the public health and
environment.  We did put criteria in here so it was kind of broad.
                                      101

-------
Question:  Bruce Nicholson, North Carolina Superfund. This question is for Lynelle
or any of the other states who want to respond to it.  But, Lynelle, you mentioned
that often air emissions have controlled risk, and from working with the EPA Office
of Air Quality from the emissions from lead smelters and lead piles and whatnot,
there's controlled risk in that situation also. I was just wondering, you mentioned
before there were no algorithms that you would use to measure those air emissions.
Is there anything that you're considering or some of the states are doing?  I know
that EPA has the AE42 document that does have future emissions in place.

MS. MAROLF:  The approach  that the rules require us to take is to look and see
whether or not we have a violation of the air quality criteria specified under our state
law. And just recently our staff did some calculations based on applying our
proposed air toxics rules to dust at the National Air Quality Standard, 650
micrograms per  cubic feet or something like that. And what those calculations show
is that using what we would  apply as a regulatory approach for a permitted source,
that even background concentrations of metals in soil would produce an unacceptable
level of risk if you assume those same exposures, 24 hours a day,  365 days a year and
whatnot. So we're in the process right now of examining those exposure scenarios to
try to establish whether or not that's really an appropriate thing to do when you're
judging a cleanup setting as  opposed to a regulatory approach.

It's really a hot topic for us right now for another reason. People are looking  to use
contaminated soils they recover at landfills. Before we approve that kind  of use, we
want to be sure  that the kind of fugitive dust that we get at landfills, some of which
are unfortunately located near residential areas, is not going to pose an unacceptable
risk.

So that's the context right now that we're looking at this issue, but it will have
implications for  the way we make our cleanup decisions. If you want to give me your
card or something, I'd be happy to try and keep you up to date on how we decide
about that.

Question:  I have another question. This one's for  Kate. My notes are kind of hard
to read now with my scribbling.   I'm not sure exactly if I wrote this down right, but
you had said that for the ground-water goals, the goals were 106.  Does that mean
you did not use MCLs  or you did use MCLs and 106 if there was no MCL?

MS. JOYCE:  Right.  The ground-water numbers are the MCLs or 10 6 if  there are
no MCLs.

Question:  Kate, I think it was you who had the overhead about sandy  soils, silt, and
silty loams. Have you done  any work with clay soils, standards for clay or—

MS. JOYCE:  No. The three types, sand, silty loam, and silty clay loam were the
three  categories  that were picked to put into the  model as far as porosity goes. But,
no, we haven't delineated any finer than those three categories.
                                      102

-------
MR. WOLFENDEN:  If there are no other questions, we can go back to Alison for
some closing remarks. Alison.
MS. BARRY: First of all, to follow up on some of the concerns that have been
raised today relative to the use of fate-and-transport modeling, EPA has just begun a
research project to evaluate a number of vadose and saturated zone models for use at
a site.  It's going to be  a comparative study which will tell us what kinds of
assumptions and what kind of data input are necessary to use these particular models.

And then secondly we're going to do some laboratory validation of certain vadose
portions of the models. So that we intend to produce some guidance on the use and
selection of models in the future that won't be coming out in the immediate  future,
but there will be intermediate guidance offered to you, I'm sure, on the part of the
labs as well as my office.

Second  thing is the workshop tomorrow is intended to focus on initiating a Superfund
policy for developing soil chemicals that protect ground-water quality.  I know that
the memorandum had specified that attendance would be limited to speakers, but we
realized since then that there's been some confusion about that. For the record, if
you presented material today or you've been specifically contacted about the work
group, we'd like you to be here tomorrow at 8:30 so that we can begin.  Most of you
already have an agenda. The few of you that do not,  I will provide one when we get
there in the morning.  The rest of you, if there's been any inconvenience associated
with the confusion, I apologize for myself and the agency.  Please see me if you need
anything.

Thirdly, we'll be issuing a transcript  of these proceedings and a memorandum
summarizing the work group conclusions.  That should be coming along  in the next
month or so and we will be issuing it and sending it to the addresses that you
provided to the contractor outside.  So if your address is incorrect,  please make sure
that you've given us the right one before you leave.

And lastly,  we'd like to thank everybody, the speakers and the attendees for  their
timely and  well chosen questions. We'd like to maintain our communication with
most of you on these issues,  so feel free to call us, my office, and we will certainly be
in contact with you.

For those of you who want to continue discussions, we can stay here for awhile but
I'd recommend we move up  to the lounge where it's a little bit more convivial and I
think some of us would prefer to be.
                                      103

-------
        Appendix A
PROCEEDINGS ABSTRACTS AND
     OTHER HANDOUTS

-------
                                   ABSTRACT

                                         by
                               Joe Williams, U.S. EPA
               Robert S. Kerr Environmental Research Laboratory
Presentations in this session will cover site-specific issues and how they are addressed
through a sampling of approaches. Due to regulatory, scientific, and economic issues involved
in the determination of soil cleanup criteria, or soil standards, it is extremely important to
cover as many of these standpoints as possible when selecting a specific approach. There are
several methods currently available which are highly conservative with regard to protection
of ground water. These methods require little to no site specific data, and do not account for
site variabilities, mixed waste processes, and chemical properties as effected by site specific
parameters.  However it becomes imperative when considering the economics of obtaining
these levels, that a more site specific approach should be used. A few basic questions should
be asked when weighing the  efficacy of utilizing simplified approaches as opposed to more
site-specific approaches. Some of these questions follow:

      1.    "Has enough data collection and site characterization been performed to support
           the use of simplified methods?"

      2.    "Have the process present at the site  been defined  well enough to allow their
           exclusion from consideration when selecting a method, or model for determination
           of cleanup levels?"

      3.    "Are processes, such as fractured flow, being considered which might facilitate
           greater releases to the ground water than would be predicted by the simple
           approaches?"

      4.    "Are remediation technologies available and feasible for achieving cleanup levels
           determined by the simple approaches?"

      5.    "Does the cost of cleaning  up to levels determined by simple, or non site-specific
           methods exceed the cost of obtaining more definitive characterization data?"

The speakers will be addressing these, as well as other questions in their presentations as
they present information related to obtaining cleanup criteria at  hazardous waste sites.
Speakers will be discussing their experiences with site-specific approaches, and the degree
of information required for utilization of these approaches. Also at issue will be consideration
of technical and policy perspectives.

-------
                                 ABSTRACT

         Setting Clean-up Levels for Volatile Organic Soil Contamination
                                       by
                                Jeff Rosenbloom
	EPA Region IX	

A vast majority of the Superfund sites throughout the nation are contaminated by volatile
organic compounds.  Although these compounds represent a significant threat to both public
health and the environment, experience at two Arizona Superfund sites can streamline the
investigation and development of soil clean-up levels.

Poor handling of solvents at both the Phoenix-Goodyear Airport (PGA) and Indian Bend Wash
(EBW)  Superfund sites has resulted in high levels of volatile organic compounds (VOCs)
contaminating soil and ground water.  Disposal methods include wastewater sent to dry
wells, surface impoundments, sewer outfalls feeding unlined ditches, and spills.

During the remedial investigations, it became clear that residual VOCs in soils were found
on both the  soil matrix and in the gas between soil particles.  In fact, there were large
courses of low carbon soil, sands and gravels, where the concentrations of VOCs in the soil
gas were orders of magnitude greater than on the soil. Therefore, soil gas sampling became
the predominate investigation tool  for both shallow and deep soils.

For VOCs, as well as other contaminants, establishing soil clean-up levels has become a two-
part process.   Shallow levels  tend  to  be evaluated using direct-contact  health  risk
calculations.  In contrast, since  deep soils do not contact surface flora or fauna, we have
begun to evaluate deep soil contaminant migration to the water table resulting in an adverse
impact on ground water quality.

A  subsurface transport model  known  as VLEACH has been used at PGA and IBW to
determine if a VOC source area presents a threat to ground water. Using both soil and soil
gas concentrations, the mass of VOCs in the vadose zone is calculated. The VLEACH model
is  then used to determine if concentrations in ground water will rise above applicable or
relevant and appropriate standards, e.g. MCLs or risk generated concentrations, given the
downward VOC migration.

As the soil clean-up levels are based on mass transport rather than a specific concentration,
a decision tree was created to evaluate the effectiveness of the remedial action. During the
remedial action, the mass of VOCs in the vadose zone  should decrease. As the  mass
decreases, the model will be run to determine if that current VOC mass in the vadose zone
continues to present a threat to ground water. Once the threshold of ground water protection
is  achieved,  the remedial action is halted to  see if soil concentrations rebound and mass
estimates once again pose a threat. If not, the remedial action is complete.

-------
                                  ABSTRACT

               Soil Cleanup Goals at a Superfund Site in Region III
                                        by
                              Dave M. Kargbo, Ph.D.
Soil cleanup goals for inorganics and organics were calculated for a superfund site (Site X)
in Region III as part of an ongoing RI/FS study.  The objective was to estimate soil cleanup
goals that would prevent unacceptable risk to human health resulting from the ingestion of
contaminated ground water and direct contact of soil by residents.   Chemicals of concern
identified include arsenic, cyanide, benzene, methylene chloride, tetrachloroethene (PCE),
trichloroethene   (TCE),   chlorobenzene,   naphthalene,  naphthalene   acetic  acid,
naphthylacetonitrile, and tetrahydofuran.

The ground water underneath Site  X occurs in the saprolite and bedrock.  The saprolite
varies from silty clay to coarser textured materials downwards and flow is controlled by
primary porosity. Flow in the bedrock is controlled by fractures.  Both saprolite and bedrock
aquifers are hydrologically connected. Depth to ground water ranges from < 5 to 35 ft. The
soils above the saprolite also grade from silt loam to clay downwards.

For the ground water ingestion pathway, a mass balance approach was used to estimate the
chemical concentration in the infiltrate (Cw) which on mixing with ground water would result
in ARAR-based (MCLs, & AWQS) and Health-based concentration of the chemical in ground
water.  Cw was then related to the soil cleanup goal (Cs) through the partition coefficient
(Kd).  For the carcinogens, estimated Cs ranged from .07 to 92 ppm (ARAR-based) and .02
to .08 ppm (Health-based).  The non-carcinogen Cs ranges are .07 to 680 pmm (ARAR-based)
and 1 to 5,000 ppm (Health-based).

Several problems were noted with the model. Influence of soil physical properties (texture,
structure, vertical variation of inorganic  as  well  as organic soil colloids, water and
contaminant flow regime, etc.), chemical properties (eg.  soil catalysis of chemical reactions),
and effect of soil microbes especially on organics was not accounted for by the model. It was
difficult to assess the potential influence of facilitated transport and additive and synergistic
cancer risk on estimated Cs at Site X.

The potential disadvantages of the model were recognized prior to its application.  However,
in the absence of an alternative model, the model was judiously utilized with modifications
to better simulate the transport processes.  For example, by assessing flow in the saprolite
aquifer rather than the fractured  bedrock aquifer where water  is  actually  used  for
consumption, the assumption of complete mixing in ground water was considered appropriate
for this site. Also while the original model requires background concentration (Cb) input, Cb
was not included in the  Cw estimation method because background concentrations for  the
chemicals of concern were zero.

-------
                                ABSTRACT

   Use of the Multimed Model to Determine Soil Cleanup Levels for VOCs at a
                                Superfund Site
                                      by
	Christos Tsiamis, USEPA, Region 2	

This presentation will discuss the use of EPA's Multimedia Model (MULTIMED) in the
development of soil cleanup  levels for Trichloroethylene  and Tetrachloethylene at the
American Thermostat Superfund site in Greene County, New York. Soil cleanup levels were
developed based on compliance with New York State drinking water standards at the first
residential well (the first receptor) downgradient from the site.  The model was used to
simulate contaminant transport through a portion  of the unsaturated  zone and in the
saturated zone.  Data variations as a function of the "first receptor" distance from the site
are presented.  The assumptions inherent in the model, the utilization of site specific data,
and the uses of the results as a screening tool in the selection of the remedy for the site are
discussed.

-------
                                  ABSTRACT

                                        by
                              Dave Crownover, Chief
             Hazardous Sites Cleanup Act (HSCA) Response Section
          Pennsylvania Department of Environmental Resources (DER)

Pennsylvania state law prohibits any release of contamination to the groundwater, therefore
soil to groundwater fate and transport  projections  must be factored into the superfund
remedy selection process.

The focus of this meeting is on the specific types of superfund cases where the soil to ground
water fate and transport projections are central to the remedy selection decision. However,
in order to place this issue into an overall perspective it is useful to note that, in many cases,
the remedy selection process is driven by feasibility factors, or other factors, and the soil to
groundwater fate and transport projections are not central to the remedy selection decision.

Pennsylvania DER has not  adopted  a specific fate and transport model as the standard
model, nor have we adopted default soil cleanup numbers. It is my personal position that we
should not adopt a standard "official" model in the future.

There is a large degree of site specific variability among Superfund sites. A specific fate and
transport model, and specific assumptions, which may be appropriate for some sites may not
be appropriate for other sites.  It  is therefore preferable to account for site specific
information, and allow for some professional discretion in selection of the specific model and
specific assumptions.

There is not a strong empirical basis to the fate and transport models. There is not a high
degree of scientific certainty associated with the projections generated by the fate  and
transport models. Selecting  a specific model as the "official" model creates a false sense of
scientific certainty.  As scientists, it is important not to overstate the degree of certainty
associated with these projections.

In order to provide a estimate of range of uncertainty, it is useful to run more than one model
using a range of reasonable assumptions and parameters. This provides the decision maker
with a range of projections  (from the most conservative projections to less conservative
projections) which can be factored into the site specific decision making process.

-------
                                  ABSTRACT

             Washington State's Approach to Soil Cleanup Standards
                                        by
                      Pete Kmet, Environmental Engineer
                                       and
                     Dave Bradley, Environmental Specialist
       Toxics Cleanup Program, Washington State Department of Ecology

The Washington State Department  of Ecology is in the process of promulgating cleanup
standards for contaminated sites. These standards address cleanup requirements for several
different media, including soil.

The soil cleanup levels are determined primarily by consideration of potential exposure via
direct contact and potential ground water impacts. For simple sites, a set of table values has
been provided that consider these two pathways (as well as others). For more complex sites,
detailed equations and requirements for the direct contact pathway have been provided in
the proposed rule.

For groundwater, the standard setting process for more complex sites consists of determining
the appropriate level of protection for the groundwater resource and then determining what
level of soil protection is  necessary to  achieve this. The rule provides specific:  criteria for
determining if a groundwater bearing  zone would be a potential source of drinking water.
Under these criteria, most groundwaters in the State of Washington would be protected for
drinking water use.  For groundwater not considered drinking water, the  cleanup levels
would be based on the potential for  contamination of surface waters, and/or groundwater,
near the site that would  require protection for drinking water use.  Once the appropriate
level of groundwater protection has been provided, two options are provided for determining
the appropriate soil  cleanup level.  In the first method, the soil cleanup level is set equal to
100 times the groundwater protection level. The rule also provides for an alternative method
of demonstration, the details of which have not been specified.

Another issue that  has been considered in the proposed rule, is the point at which the
cleanup level applies.  For direct contact, the proposed rule states that soil within  15 feet of
the ground surface must meet the  cleanup level. This is to address the potential for soil to
be brought to the surface by site development activities. For groundwater  protection, the
point of compliance is throughout the site, since precipitation would leach through the entire
soil profile.

The rule also prescribes  that soil cleanup levels will be based on  total analysis of that
fraction, 2 mm or smaller in size (sand sized). A total analysis has been specified to provide
for more consistent test  results (rather than a leach test).  The particle  size  has been
specified since this is the  range of particle sizes most likely to be ingested and contribute to
leaching. It also facilitates obtaining reproducible laboratory analyses.

Once the cleanup standards (cleanup level and point of compliance) have been determined,
the rule provides for a process for selecting an appropriate cleanup action that will meet the
standards.  The cleanup  action must meet a number of requirements, including utilizing
"treatment to the maximum extent practicable." Where it is not practicable to treat the site
(including soil) sufficiently to achieve the cleanup levels, the use of containment is allowed.

-------
                                  ABSTRACT

    Connecticut's Current Practice for Remediation of Hazardous Waste Sites
                                        by
                   Randy May, Supervising Sanitary Engineer
              Connecticut Department of Environmental Protection

Connecticut's standards for the remediation of hazardous waste sites are in a period of
transition from operation within broad Statutes  and general policy to the adoption of
comprehensive regulations. The process of drafting the regulations is underway and has
served  to  highlight  existing  program  elements and inconsistencies.    While these
inconsistencies exist, the current program has dealt successfully with the remediation of a
diverse group of hazardous waste sites.  These have ranged from sites where substantial
quantities of hazardous waste were generated and discharged, to the diverse distribution of
the pesticide EDP across the tobacco growing region of the State.

The fundamental driving force behind Connecticut's remediation program is the Clean Water
Act of 1967 which sets forth requirements for permits, enforcement and standards of water
quality. The Water Quality Standards, first adopted in  1973, and expanded  to include all
groundwater in 1980, are a principle element in remediation decisions. The Water Quality
Standards classify some 91% of Connecticut's  groundwater as GAA  or GA, suitable  for
drinking without treatment and impose the highest standard for remediation in those areas.
Roughly 6% of our land area is classified as GB which generally indicates urbanization and
proven  or presumed degradation  of the groundwater  resource.  Not  surprisingly, the
preponderance of our roughly 900 problem sites are located within these areas. The balance
of the area, 3%, is classified as GC, meaning that it appears to be geologically suitable for
proper waste disposal.

The Department has been organized to deal with remediation based on historic, legislative
and water quality standard based issues. The lead for remediation in areas classified as GA
or GAA has rested with the Bureau of Water Management, and been driven by a primacy of
absolute resource protection. This unit also administers the State's program for the delivery
of alternate drinking water supplies to affected parties.

The remediation standard in these areas has not utilized models of pollutant migration. The
staff position has been that existing models are  too simplistic for utilization in this critical
process.  The staff has  also had direct experience with models that completely failed to
predict the actual dispersion of pollutants from the unsaturated zone.

The standard of remediation in these groundwater areas has  been  to utilize analytical
techniques that will yield a highly conservative  estimate of the concentration of pollutants
in soil. The assumption is made that the entire volume present is discharged, at once, to the
groundwater system without removal mechanisms. Remediation, normally soil removal, has
taken place to the level that will result in attainment of the relevant drinking water standard
or guideline issued by our toxic hazards section of the Department of Health Services. Soil
removal has been  viewed as an integral part of the program, resulting in immediate
improvement in groundwater quality. Soil removal has been combined with hydraulic control
where such techniques are feasible.

-------
Remediation in areas classified as GB has generally had a lower priority and these programs
have been administered by the Bureau of Waste Management. The remediation of these sites
has caused issues of substantial difficulty.  The Water Quality Standards do not require that
these waters be "necessarily" returned to a quality of, "suitable for drinking without
treatment." As a result the Department determined that a lesser standard was permissible.
In these cases a methodology identical to remediation in a drinking water area is followed;
except that the allowable concentration is 10 times the drinking water standard.

There is some difficulty in fitting this standard in the basic anti-degradation concept inherent
in the water quality standards. This may be offset by the reality that the GB sites tend to
be the largest and most difficult to correct. The size of the areas of contamination are also
believed  to exceed any practical limit for soil  removal as  a principle element in  the
remediation process.  Hydraulic and containment control are believed to be the appropriate
requirements, but it must be stressed that the final rule has not been drafted.

As with the GA and GAA sites, modeling based on  potential  pollutant migration is not
directly utilized in determining clean up goals. The reasons for this are largely as previously
stated. The Department's staff does express concern that newer modeling techniques may
be developed but that limited technology transfer may impede their knowledge  of these
procedures. Existing migration models are considered as an element in determining if factors
such as well  position and measured concentration accurately reflect the progression of a
plume.

Connecticut's methodology  can be  considered non  site specific in the sense that  the
underlying clean up standard keys to mapped water quality standards.  The actual extent
of clean up and the methodology of remediation remain very specific to the individual site.

-------
                                 ABSTRACT

  Cleanup of Contaminated Soils Under Administrative Rules for the Michigan
                          Environmental Response Act
                                       by
                                 Lynelle Marolf
    Assistant to Deputy Director Michigan Department of Natural Resources

The Michigan Department of Natural Resources (MDNR) has promulgated administrative
rules (effective July 11, 1990) which  specify cleanup standards applicable to all sites of
environmental contamination in the state.  The rules provide, in general, for three types of
cleanup criteria, designated Type A, Type B,  and Type C.  Type A criteria are based on
reduction of hazardous substances to background or to analytical limits. Type B criteria are
based on reduction of hazardous substances to an acceptable risk level using standardized
exposure assumptions.  Type C criteria  are developed on the basis  of a site-specific
assessment of risk to the public health, safety, and welfare and to the  environment and
natural resources. The cleanup "type" is proposed, as part of a remedial action plan, by the
party responsible for site cleanup, subject to review and approval by MDNR.

Discussion in this presentation will focus on the Type B soil cleanup criteria. The unique
approach of the Type B soils cleanup criteria was developed by MDNR staff to account for the
threats which contaminants in soil may pose to groundwater, to surface water, through direct
contact, through inhalation, and through other pertinent pathways (e.g., phytotoxicity). The
rules specify the manner in which threats to groundwater and through direct contact must
be addressed. Consideration of other pathways is required, when information is available to
allow an assessment to be made, but specific algorithms are not provided in rule.

-------
                                 ABSTRACT

        Minnesota Draft Procedures for Establishing Soil Cleanup Goals
                                       by
                                  Jim Pennine
                       Minnesota Pollution Center Agency

Minnesota has developed a draft procedure for establishing soil cleanup goals. The approach
is to use site specific data acquired during the Remedial Investigation (RI) to derive cleanup
goals for a given hazardous waste/Superfund site.

The complex nature of soils makes the assignment of chemical specific numerical soil cleanup
standards that are applicable to all sites either too conservative or too liberal for most sites.
Minnesota uses  a project team approach in its supervision of the cleanup of Superfund sites.
This approach, combined with the existing regulatory framework, facilitates the development
of cleanup goals on a site by site basis. This is consistent with current practice in a number
of other states. Historically, site by site cleanup goals have been largely driven by costs, time
constraints, the negotiating skills of the project team, and expedience rather than sound
scientific understanding of the site conditions and exposure mechanisms.  Cleanup goals in
Minnesota and  elsewhere have been very inconsistent, covering a wide range of criteria
including visual examinations for visible products,  animal testing, leachate  tests and
arbitrary numerical standards chosen because they were used at another site where the same
contaminant was found.

This procedure is designed to provide a consistent approach to setting cleanup goals on a site
by site basis.  The approach will be the same but the cleanup numbers will vary from site to
site.

The procedure involves performing a risk assessment for the various  routes of exposure to
the contaminants that exist at a given site and then setting cleanup goals that will protect
the public health. The focus of the procedures in this draft document are to protect ground
water used for  drinking purposes.  Other routes of exposure are briefly discussed in the
document.  Minnesota plans to develop a  similar procedure for calculating cleanup goals
designed to protect threatened ecosystems.

All ground water beneath sites is assumed to be capable of providing drinking water or is
assumed to be  hydraulically connected to a nearby aquifer that is  capable of providing
drinking water.  Therefore, National Interim Primary Drinking Water Standards (NIPDWS)
Maximum Contaminant Levels (MCLs), Minnesota Department of Health (MDH) Recom-
mended Allowable Limits (RALS) for  private drinking  water wells, or non-degradation
standards are assumed to be the exposure criteria for the drinking water route of exposure.
The soil cleanup goals are back calculated from the MCL, RAL or non-degradation standard.
Where soil contamination results in a degraded ground water discharge to surface water,
aquatic criteria can also be used to calculate the soil cleanup goal.  The cleanup goal is a
threshold concentration in the soil that would not leach sufficient amounts of contaminants
to contaminate  the ground water above the MCL, RAL or other criteria.

-------
The procedure involves several calculations which are based on a conceptual model.  The
conceptual model  consists of  a  set  of conditions that  are typical of  hazardous
waster/Superfund sites with contaminated soil.  The conditions  include:   an area of soil
contamination has  been defined in the subsurface;  a  portion  of the contamination  is
permanently adsorbed to the soils; in the case of organic contaminants,  a portion of the
contaminants are being degraded by organisms and another portion is lost to the soil air
space or to the atmosphere via volatilization; finally, the remaining contaminants are subject
to solution into percolating rainfall which carries the contaminants to the ground water.

Calculations  are  performed to  determine the  amounts of  adsorption,  biodegradation,
volatilization and leaching that will occur for a given site. Using a drinking water MCL or
other criteria, the equations are rearranged to back calculate a soil concentration which is
the soil cleanup goal. Using this procedure a polychloinated biphenyl (PCB) cleanup goal of
six milligrams/kilogram (mg/kg) was calculated for  a site with a  shallow  water table,
moderately permeable soils  and predominantly sandy subsurface soils. A trichloroethylene
cleanup level of 0.044 mg/kg was calculated for another site, assuming vinyl chloride  as a
breakdown product and using drinking water criteria for vinyl chloride.  This latter site has
a shallow water table, moderately permeable soils with low organic matter content and clayey
to sandy, subsurface soils. These cleanup goals are generally comparable to values found in
the U.S.  EPA publication "Determining Soil  Response Action Levels Based  on Potential
Contaminant Migration to Ground Water:  A Compendium of Examples".

-------
                                 ABSTRACT

           Summary of TSCP Draft Scientific and Technical Standards
                                       by
                State of California Department of Health Services
                       Toxic Substances Control Program

The Toxic Substances Control Program (TSCP) of the Department of Health Services has the
responsibility of managing California's hazardous waste program.  A primary goal of TSCP
is to ensure the protection of public health and the environment  from adverse impacts of
hazardous waste sites.  The Technical Services Branch, in addition to providing technical
training  and  guidance for  the TSCP, is developing  scientific and  technical  standards
(standards) to set forth the requirements for site characterization, exposure assessment,
toxicity assessment  and risk characterization, and soil remediation levels for hazardous
substance release sites.

This document contains a summary of the draft standards. These standards will be applied
to sites being remediated under the direction of TSCP. Standards will also be applied, where
and  when  applicable,  to facility  closures and  corrective actions administered by  the
Permitting Program and to abatements or small site cleanups ordered by the Surveillance
and Enforcement Program.

                                BACKGROUND

The  Site  Mitigation Program is proposing a  revised process that will be more efficient,
thereby expediting hazardous waste site remediations.  As a part  of this process, TSCP is
implementing standards to clarify and prescribe  the scientific and technical requirements
for conducting a site investigation, and to establish site specific remediation goals. Because
TSCP oversight of the site remediation process may be limited to selected sites, performance
standards are needed to set the minimum requirements  that must be met by responsible
parties to properly conduct  a remedial investigation, and to adequately remediate a site.
Standards are being  developed to address the scientific and technical aspects of the remedial
investigation (RI) and portions of the feasibility and remediation phases of the site mitigation
process.  If the standards are followed and implemented correctly,  the RI and the selection
of a cleanup level should meet the regulatory objectives.

Standards Development and Regulation Promulgation

It is the  intention of the Department  to promulgate as regulations  the draft standards
summarized in this paper and other standards currently  under development.  The process
will involve a series of public workshops to solicit comments and  suggestions. Due to the
volume and complexity of these standards, a series of workshops will be scheduled to focus
on specific sections of the standards.  The purpose of the workshops is to enable the public
to-have significant input to the standards. Based on the input from the workshops, these
draftvstandards will be submitted as draft regulations. It is anticipated that portions of the
draft standards may not be amenable to promulgation as  regulations.  We will solicit input
on this subject during the public workshops. Those standards not promulgated as regulations
will be developed as  guidance documents to guide TSCP staff and  responsible psirties.

-------
                                    OVERVIEW

Specific technical  standards were developed based  on the fundamental components of
implementing a site investigation and remediation.  This includes considering all aspects of
project scoping, site characterization, and risk assessment in relationship to the contaminants
present in the various media (i.e., air, soil ground water, and surface  water), as well as
specifying a method for determining soil cleanup values for contaminants. The standards
have been organized into the following volumes:

       1)   Site Characterization
       2)   Exposure Assessment
       3)   Toxicity Assessment and Risk Characterization
       4)   Soil Remediation Levels

This organization is consistent with the basic components of the RI process and portions of
the feasibility study process as developed by the U.S. Environmental  Protection Agency
(EPA).

Site Characterization

During site characterization, field data is collected  to define the nature and extent of
contamination present at, or migrating from a waste site, and identify  the processes that
serve to transport contaminants.

Standards have been drafted that will provide direction in:

       •   developing an overall site characterization strategy,  as well as the individual
           techniques and procedures for characterizing contaminants in the various media
           (i.e., air, soils, ground water, and surface water),

       •   selection and use of techniques to define the physical characteristics of a  site (i.e.,
           geology, hydrogeology, soils, meteorology, surface features, etc.), and

       •   selecting appropriate chemical analysis methods and evaluating the accuracy of
           laboratory analytical results.

Performing an adequate site characterization is essential since the information gathered will
form the basis of exposure and risk assessment activities, and will serve to determine the
feasibility of various remediation measures.

Exposure Assessment

Exposure assessment is described by the EPA Human Health Evaluation  Manual (1989) as,
"the determination or estimation (qualitative or quantitative) of the magnitude, frequency,
duration and route of exposure." Further, "the objective of the exposure assessment is to
estimate the type and magnitude of exposures to chemicals of potential concern  that are
present at or migrating  from  a site."  The exposure  assessment draws  on  the site
characterization data to identify the occurrence of contaminants in environmental media, and
to identify the significant pathways for transport  of contaminants  to  the population of
concern. This information is integrated with environmental fate and transport predictions
to yield estimates of the amount of contaminant contacted by the population of concern over
a period of time.

-------
The scientific and technical  standards for exposure assessment are  intended  to  be
complementary to EPA's Superfund Exposure Assessment Manual (1988) and Human Health
Evaluation Manual (1989). Draft standards developed to date provide direction in:

       •  handling of analytical chemistry results that serve as input data for exposure
          assessment measurements and/or modeling,

       •  developing and documenting relevant exposure scenarios, and

       •  developing and documenting estimates of exposure.

Toxicity Assessment and Risk Characterization

The purpose  of a toxicity assessment is to  identify the health hazards  associated with
exposure to hazardous wastes. The scientific literature is reviewed to determine the potential
for each chemical contaminant to cause adverse health effects in exposed individuals, and to
evaluate the relationship between the degree and duration of exposure to each chemical and
the likelihood of occurrence of adverse effects in the exposed population.

Risk characterization is the final  step  of the risk assessment  process.   The toxicity
assessment and exposure  assessment are integrated into quantitative expressions of the
health risks posed by exposure of the populations of concern to each and every chemical at
or emanating from the hazardous waste site.  A quantitative health risk characterization of
a  facility  or  site  is  an integral part of the regulatory  decision-making process.  The
characterization of the health risks posed by hazardous waste sites will eventually form a
part of the public record, and contributes to the justification of regulatory decisions taken to
ensure that public health and the environment are protected against significant risk. The
risk characterization provides information about the nature and magnitude of the potential
health risks  associated with the project, provides a  basis for judging  the  necessity of
instituting remedial measures, and can be used to compare the reduction of risks afforded
by different remedial or control strategies.

The draft standards developed to date provide direction in:

       •  the preferred format, documentation, and analysis of site specific data in the
          determination of associated adverse health impacts,

       •  characterization of potential adverse health impacts from chemical mixtures, and

       •  identification of reproductive and/or developmental toxicants.

Further scientific  standards will be  developed  to  address 1)  low  dose  cancer risk
extrapolation, 2) chemical specific carcinogen identification,  3) dioxin/furan mixture risk
estimation, 4) identification of genotoxicants, 5) toxicokinetics, and 6) the  role of other
regulatory exposure criteria.
     ^ -
Soil Remediation Levels

The standard for soil  remediation levels will provide EPA recommended risk assessment
formulas to determine conservative remedial goals that are protective of human health. They
provide uniform methods and acceptable EPA recommended parameters for assessment and
characterization of risk.

-------
 Risk management decision making at waste sites or facilities must not simply be concerned
 with the concentration of some toxicant, but must consider the total mass of toxicant present
 at a site, and the extent of the interface between a contaminated soil reservoir and other
 media of exposure.  For a given contaminant, the greater the potential extent of exposure,
 the greater the need to reduce the reservoir of contaminant.

 Risk management through soil remediation may be described as the process by which the
 health effects resulting from exposure to soil contaminated with hazardous waste are reduced
 through reduction of the mass and the concentration of contamination to levels that protect
 human health and  the environment.  Various regulatory agencies have proposed several
 markedly different criteria for the determination of an "acceptable" soil remediation level,
 including reducing  concentrations  to background levels, establishing fixed levels  of
 contamination for different chemicals, and using a risk assessment approach to determine
 a site and chemical-specific remediation level. Depending on the nature of the site land use,
 and the type  of chemical(s) present, the first two criteria may result in either over  or
 underprotection of human health, and limit risk management flexibility. Consequently, the
 TSCP is proposing standards for soil remediation which use a risk assessment methodology
 that focuses on the protection of human health.

 In this standard, conservative risk management decisions regarding soil remediation will be
 made by the calculation of a soil concentration which,  if humans were to be  chronically
 exposed over long periods of time, would not be expected to produce adverse health effects.
 However, this calculation can be done only when several conditions are met. First, the dose
 received from each exposure pathway must be estimated in a scientifically defensible manner.
 This implies that a site-specific exposure scenario has been defined that includes all potential
 exposure pathways.  The pathways  are dependent on the  types of human receptors that
 might be present at a site. The types of receptors are determined primarily by the  site land
 use, both before and after remediation (e.g.,  workers at an industrial facility vs.  children
 playing in  a residential backyard or playground).   Second, mathematical relationships
 between soil concentrations and concentrations in other media of exposure must be available,
 or conservative assumptions must be made to relate soil concentration to other media, since
 soil serves as a reservoir of contamination for other media to which humans may be  exposed.
 Third, scientifically demonstrated acceptable levels of chronic daily intake of individual toxic
 chemicals must be known.

 The following exposure parameters are needed to provide an adequate assessment of dose
 from a reservoir of contaminated soil:

          nature and extent of soil contamination,
          current and/or proposed land use,
          current and/or proposed receptor population,
          potential exposure pathways, and
          soil, meteorological, geological and ground water characteristics.

When the  general characteristics of these parameters are known, EPA recommended
parameters, as supplied in  the standards, can be used as conservative  estimates in the
exposure calculations.  The equations do require  certain  site-specific values in  order to
calculate a health-protective soil remediation  level.

-------
            SUMMARY OF SITE CHARACTERIZATION STANDARDS

The following are descriptions of the individual scientific and technical standards documents
developed to date for site characterization.

Project Scoping and Data Quality

The following standards address the planning of a site  characterization, and the quality
assurance of data collected during a field investigation.

      Remedial Investigation Project Scoping. These standards represent a revision
of the Scoping Section of the EPA RI/FS Guidance Manual. The EPA guidance was selected
as a starting point for the development of this standard because it contains the essential
elements of the scoping process.

Scoping is an organizational process that is the first step  of site characterization.  It guides
all subsequent RI/FS activities and is essential to the timely and economical performance of
any hazardous waste site remedial investigation.  Scoping provides background data and
strategies that clearly define what has been done at the site, what needs to be done, and how
it will be accomplished.

The standards provide the requirements for planning and organizing a RI/FS, including 1)
establishing a site management strategy, 2) identifying possible remedial strategies and risk
standards, 3) developing a sampling and analysis plan, 4)  preparing a site health and safety
plan, and 5) preparing a community relations plan. The standard stresses the need to
analyze existing data and identify data  gaps, the development of a conceptual model of site
conditions and potential exposure pathways, and establishing data quality objectives.

      Data Validation. It is essential to verify the quality of analytical data that is  to be
used for  site characterizations,  enforcement  actions, hazardous waste determinations,
exposure assessments, and regulatory compliance. These standards establish procedures that
ensure that the analytical data generated during site characterization will be of known
quality.  The  first portion  of the standards, presented  as a checklist,  identifies the
information that must be collected and presents the process for defining Data Quality
Objectives (DQOs) for this information.  The second portion of the standard, presented as a
flow chart, presents the requirements for comparing analytical quality control data to the
DQOs to assess the overall quality of the analytical data.

      Surveying and Mapping Hazardous Waste Site Sampling Locations.  These
standards present the  surveying and mapping techniques most appropriate for locating
sampling stations (for soils, ground water, and air) during site characterization and specify
the accuracy required when using each of the techniques. The accurate location of sample
points horizontally and vertically is important for defining the extent of contamination,
depicting the subsurface geology,  and performing contaminant fate and transport analysis.
The standards also provide criteria for determining when a licensed surveyor should be used.
Part of the standard is modified from EPA guidance documents addressing surveying and
mapping  procedures.  Proper implementation of the standard will assure that  sampling
stations are accurately surveyed and mapped.

Geology and Ground Water

The following standards address the techniques used to characterize the geology and ground
water conditions at a hazardous waste site.  ,.

-------
       Hydrogeologic  Site Characterization at Hazardous  Waste Sites.   These
 standards direct the investigator in determining the scope and detail necessary in conducting
 a hydrogeologic site characterization. The standards cover the general areas necessary to
 perform a comprehensive hydrogeologic investigation of a hazardous waste site, and reference
 the detailed standards to be followed in performing field work. The standards stress the fact
 that hydrogeologic investigations of hazardous waste sites must be conducted by or under the
 direction of qualified professionals.  The standards are based on several EPA guidance
 documents that have been developed for conducting a site characterization.

       Surface Geophysical Techniques at Hazardous Waste Sites. Surface geophysical
 techniques are  generally safe to  use  at hazardous waste site because they do  not spread
 contamination or create pathways for  contamination.  In addition, surface geophysical
 techniques may help reduce the cost  and duration of hazardous  waste site investigations.
 These standards establish criteria for  the use of surface geophysical techniques that may be
 useful in acquiring information about subsurface geologic features, contaminant plume and
 container locations,  and ground  water occurrence.   Methods discussed include  seismic
 refraction and reflection, resistivity, frequency domain electromagnetic techniques, ground
 penetrating radar, gravity, and magnetometric techniques.  The standards provide criteria
 to ensure the proper collection  and reporting of geophysical data, and to assure that
 interpretations  are reasonable and independently verifiable.

       Drilling. Coring. Sampling, and Logging at Hazardous Waste Sites.  These
 standards specify criteria for drilling and the collection of samples of geologic materials. The
 standards specify criteria for logging, i.e. the recording of field observations during drilling,
 resulting in the collection of data that is systematic and consistent. These standards are
 based on methods common to the environmental drilling industry. In addition to assuring
 accurate data collection, implementation of these standards at hazardous waste sites will
 minimize the potential for spreading contamination.

       Borehole Geophysical Techniques at Hazardous Waste Sites. Standards are
 presented for geophysical techniques that are used to  acquire  lithologic and  hydrologic
 information from wells and borings.  Methods include electrical logs, nuclear logs, caliper
 logs, and acoustic televiewer logs. Borehole geophysical techniques are a means to obtain
 more information from a boring than can be obtained by coring and/or sampling alone. The
 standards provide criteria to ensure the proper collection and reporting of geophysical data,
 and to assure that interpretations are reasonable and independently verifiable.

      Design and Construction of Monitoring Wells and Piezometers at Hazardous
Waste Sites.   Standards  are provided  for the construction  of monitoring  wells and
 piezometers to  ensure that the  data  collected from these  monitoring installations  is
representative, accurate and consistent.  The standards specify criteria for the borehole in
which a well is  to be constructed, for well materials, and for well construction techniques.
Standards for well decommissioning are  also provided to ensure protection of ground water
resources. Proper installation of monitoring wells and piezometers is particularly important
because many decisions regarding the risks posed by a site are based on descriptions of
ground water flow, and the type, extent,  and concentrations of contaminants present. Each
of these elements is dependent on the design and construction of the wells and piezometers
that provide access to the subsurface for sampling of ground water and measurement of
hydraulic head.

-------
      Design and Construction of Extraction Wells at Hazardous Waste Sites. These
standards provide criteria for the construction of extraction wells at hazardous waste sites,
installed for the purpose of removing contaminated ground water.  Proper design  and
construction of extraction wells is essential to ensure that the wells serve their intended
purpose and do not exacerbate ground water contamination. Topics covered are similar those
for monitoring wells.

      Ground Water Sampling at Hazardous Waste Sites. These standards establish
procedures to properly collect, transport and preserve ground  water  samples.  Topics
addressed include decontamination of equipment, well purging, sample withdrawal, filtration,
record keeping, containers, and chain of custody. Implementing these standards will ensure
that representative and consistent samples are collected. Collection of this data is important
because of the direct effect of contaminant concentrations measured in ground water on the
exposure and risk assessment processes.

      Conducting  and Interpreting  Multiple and Single-well Aquifer Tests at
Hazardous Waste  Sites.   These standards establish the criteria for  desiring  and
conducting aquifer tests (pumping and slug tests), and for interpreting and reporting the test
results.  The standards address pumping well design and pumping rate, observation well
design and placement, water level measurement and recording, and data presentation and
analysis. Proper implementation of these standards will help define the hydraulic properties
of geologic materials underlying a hazardous waste site. Accurate values of these piarameters
are necessary to estimate the rate of contaminant migration in ground water, to identify
connections between aquifers, and to design extraction systems for remediation.

      Graphical Representation of Ground Water Conditions at Hazardous Waste
Sites. These standards establish requirements to be followed in the presentation of ground
water hydraulic head data from hazardous waste sites. The standards will ensure that
ground water contour maps and flow nets  are properly constructed, and provide an accurate
representation of the  ground water regime at a site.  When constructed according to the
standards, the graphical presentations will not only depict the ground water regime, but will
help ensure  that all relevant hydrogeologic factors have been considered during the  site
characterization.

Soils

The following standards apply to the characterization of soil properties and the concentration
and extent of contamination in soils.

      Determination of Physical and Chemical Properties  of Soils at Hazardous
Waste Sites for Environmental Assessments. Environmental and risk assessments often
require the determination of various properties of soils to predict the fate and transport of
contaminants. In this standard, methodologies and procedures for determining some of the
major physical and chemical properties of soils as part of the site characterization process are
presented. These properties include soil moisture, particle density, organic carbon content,
specific surface, soil mineralogy, and soil pH.  The standards specify criteria for defining the
data quality objectives,  assuring soil sample representativeness, and documenting the
analytical method in the sampling and analysis plan.

-------
       Estimation of Sorption Properties of Soils at Hazardous Waste Sties for
Environmental Assessments.  Sorption estimates are used in evaluating the  fate and
transport of contaminants from soil to ground water for risk assessments and remedial design
evaluations. These standards provide for obtaining quality data which consider soil sorption
mechanisms and factors affecting sorption.  The standards include defining the data quality
objectives,  documenting the best  scientifically valid approach,  assuring soil sample
representativeness, determining soil pH effect, and determining  the  dominant sorbent
fraction in the  soil.   Methods and procedures for estimating sorption coefficients are
presented and specific references cited. Standards for conducting batch and column studies
are also presented.

       Selection of a Soil Sampling Design for Hazardous  Waste Sites. These
standards present requirements for designing a soil sampling program that will assure that
the source and extent  of contamination will be adequately characterized.  Standards will
address three types of sampling strategies, 1) search sampling, used for detecting  evidence
of contamination,  2) probability sampling, which is necessary  to  test hypothesis about
contaminant levels present at a waste site, and 3) sampling for spatial  analysis,  which is
necessary to determine the area  and vertical extent of contamination.   Selecting  an
appropriate  sampling  strategy is important since the information obtained in  the soils
investigation will serve as a basis in the selection of a remedial alternative or guide further
site investigation activities.

       Estimating Parameters of Log-Normally Distributed  Populations. These
standards provide criteria for estimating the true population mean, variance, and coefficient
of variation of data collected in the investigation of a contaminated media. These estimates
can be used to calculate the quantity of a contaminant in a specified media or the average
contaminant media. These estimates can be used to calculate the quantity of a contaminant
in a specified media or the average contaminant concentration that may be used as  an input
parameter to an exposure assessment model.

       Sampling of Soils and Solid Media for Volatile Organic Compounds.  The
sampling of soils or solid media for volatile organic compounds (VOC's) must be conducted
in a manner designed  to minimize the loss of volatiles from the sample.  This standard
prescribes the sampling procedures necessary to reduce the loss of contaminants and improve
the quality of data generated from sampling activities.

Air Monitoring

The following standards apply to the characterization of meteorological conditions on and
surrounding waste sites, and to the measurement of contaminant concentrations in air.

      Methods for Collection of Meteorological Information  at Hazardous Sites.
These standards specify procedures for conducting a meteorological survey to determine the
local wind flow patterns, specify criteria for a screening or refined air monitoring plan, and
set criteria for  the sitting of meteorological stations and the monitoring duration.  The
standards are  designed to be consistent  with EPA's Air / Super fund National Technical
Guidance Study Series, (1989).

-------
       Operating Procedures for Air Sampling and Analysis. These standards provide
methods for the monitoring of ambient air at hazardous waste sites. Specifically, standards
specify criteria for 1) the number and location of monitoring stations, 2) the proper placement
of the sampling inlet probe, 3) determination of monitoring duration and sampling frequency,
4) identification of sampling techniques and analysis methods, and 5) qualifications and
training of sampling personnel. The standards identify refined air sampling techniques.

             SUMMARY OF EXPOSURE ASSESSMENT STANDARDS

The following are descriptions of the individual scientific and technical standards documents
developed to date for exposure assessment.

       Concentration Data in Exposure Assessment for Hazardous Waste Sites. The
standards delineate a uniform  approach to the treatment of concentration values in the
exposure assessment portion of a risk assessment.   The approach ensures that exposure
assessments will incorporate concentration estimates that adequately protect public health
and the environment. The standards do not describe in detail specific statistical methods for
evaluating sample  data  nor methods for modeling  the  rate  of loss or  transport of
contaminants.

       Documentation  and Assumptions Used  in the Decision to Include and
Exclude Exposure Pathways. These standards describe assumptions and documentation
required to make the choice of pathways to include in a site-specific exposure assessment.
Pathways that should be considered for inclusion in an exposure assessment are listed.  The
standards describe procedures for selecting, at a specific site and for a specific chemical,
which of these pathways are relevant for an exposure assessment.  The standards encourage
the  use  of  simple multiplicative  and  additive  models  for  making the  pathway
inclusion/exclusion decision.  These types of models expedite the process of ass essing both
relative magnitude and uncertainty of many potential exposure  routes associated with a
single environmental medium.

       Documentation   of Methodologies.  Justification,   Input.   Assumptions,
Limitations, and Output for Exposure Models.  These standards describe models,
assumptions and documentation required to assess  exposure.  The exposure assessment
process consists of two steps 1) the inclusion/exclusion analysis, by which the significant
exposure pathways at a site  are selected and 2) the exposure models that are applied to the
selected pathways. These standards specify the subset of exposure pathways that will be
employed for a specific site and chemical, and the methods, assumptions, and inputs that are
acceptable for making realistic exposure assessments.

       Mathematical Modeling of Ground Water Flow and Contaminant Transport
at Hazardous Waste Sites. These standards establish procedures to be followed in the
application of ground water modeling to hazardous waste sites.  The standards  apply
primarily to models used to predict the future location and concentrations of contaminants
in ground water,  and to  design  extraction  systems  for ground  water remediation.
Implementation of the standards will ensure that data requirements are met, and that model
selection, model implementation, and model application are appropriate to the site and the
problem to be solved.  The standards require that  uncertainties in model results to be
assessed, and that the uncertainties be properly communicated so that decision makers can
determine the usefulness of the model  results.

-------
       Methods for Estimating Baseline Air Emissions at Hazardous Waste Sites.
 These standards provide methods for estimating baseline air emissions to identify the impact
 of air emissions from a hazardous waste site. The standards 1) provide criteria to assess the
 potential for air emissions, 2) identify the important site conditions and parameters  that
 affect air emissions, and 3) identify predictive  emission models  and field measurement
 techniques.  The emission estimate is used to determine whether further studies, including
 air monitoring or dispersion modeling, are required.

       Methods for Dispersion Modeling at Hazardous Waste Sites.  The standards
 provide requirements for dispersion modeling  to estimate downwind air concentrations of
 contaminants, and to determine the locations of maximum impact. The standards specify
 screening and refined models to be used in dispersion modeling as well as the meteorological
 data required as input to the recommended dispersion models. The results of modeling may
 be used in exposure assessment or in  designing an ambient air monitoring program.

    SUMMARY OF TOXICITY ASSESSMENT AND RISK CHARACTERIZATION
                                   STANDARDS

 The following are descriptions of the individual scientific and technical standards documents
 developed to date for toxicity assessment and risk characterization.

       Documentation and  Format of a Multimedia Baseline Risk Assessment for
 Hazardous Waste Sites and Permitted Facilities.   This standard  provides  a brief
 overview of the contents of a risk assessment  prepared  for a hazardous waste  site or
 permitted facility. The recommendations in the standard are intended to aid RPs and their
 contractors in  preparing risk assessments and TSCP Project Managers  in planning  and
 evaluating risk assessments.  A suggested outline for a baseline risk assessment report is
 presented.  This outline is based upon the risk assessment process  described  in the EPA
 Human Health Evaluation Manual. Specific written guidance for preparing and documenting
 a risk assessment is presented. This guidance briefly highlights some  of the more important
 aspects of this effort. It is not intended to be comprehensive, since comprehensive guidance
 is available in the EPA Human Health Evaluation Manual.

      Risk-Appraisal;  A Procedure for Determining the Contribution of Individual
 Chemicals to Total Adverse Health Effects. These standards provide the risk manager
 with a set of procedures  to determine  whether  a site poses an unacceptable level of risk to
 human health from the presence of chemicals that are either noncarcinogenic, carcinogenic,
 or both.  On the basis of  the determination the risk manager can then decide if the site  is a
 candidate for remediation.  Moreover, the methods presented here will permit the risk
 manager to  assess the contribution of each contaminant  to either  the site-specific total
 carcinogenic risk or the total noncarcinogenic hazard associated with the site.

 The approach  described here differs  from previous regulatory guidance  because  it was
 developed to provide a more accurate characterization of site-specific carcinogenic risks and
"noncarcinogenic hazards by accounting for multimedia, multiple-pathway estimates of
 potential human exposure. The basis of our procedure is a relatively straightforward process
 that is applicable to carcinogenic and noncarcinogenic chemicals.

-------
      Assessment of Developmental and Reproductive Risks. This standard provides
a framework for the evaluation and interpretation of potential adverse effects to male or
female reproduction and to the developing organism for the purposes of conducting site
specific human health risk assessments. Evaluation of reproductive or developmental toxicity
of an agent entails review of pertinent data on exposure, distribution,  metabolism,
pharmacokinetics, and toxicity from human and animal studies. Decisions on the potential
for agents of concern to cause reproductive or developmental harm require an understanding
of the underlying principles of toxicology and developmental and reproductive biology, as well
as determination of the significance of animal studies in the assessment of risks to humans.
While agents may be identified as  male or female reproductive toxicants or developmental
toxicants, the actual risk to people must be characterized based on the type of exposure,
including route, duration, and levels.

      Selection, Use  and Limitation of Surrogate  Chemicals  for Evaluation of
Exposures to Complex Mixtures.  While  quantitative evaluation of all chemicals of
potential concern is the most thorough approach in a risk assessment, this standard provides
procedures for conducting a risk assessment when consideration of all chemicals of concern
is infeasible. For some sites or facilities there may be a large number of individual chemicals
of a particular chemical class for which toxicity  data and/or health-based exposure criteria
are not available, thus necessitating the use or a surrogate chemical to provide an estimate
of potential health risks associated with exposure to these substances. At certain sites or
facilities, the list of identified chemicals remaining after quantitation limits, qualifiers, blank
contamination, and background have been evaluated may be very lengthy.  Since carrying
a large number or chemicals through a quantitative risk assessment may be complex, and
may consume significant amounts of time and resources, a surrogate chemical procedure may
be considered for these rare instances.

It is important to  recognize that the time required to implement and  defend the selection
procedures discussed in this section may exceed the time needed to simply csirry all the
chemicals of concern through the  risk assessment. Therefore, it is anticipated that the
procedures described in this standard may be needed only in rare instances.

           SUMMARY OF SOIL REMEDIATION LEVEL STANDARDS

The following are descriptions of the scientific and technical standards for the determination
of soil remediation levels.

      Determination  of Soil Remediation Levels (SRL). The SRL standard utilizes
uniform procedures that permit the user to: 1) estimate the chronic  exposure  to  toxic
materials in contaminated soil using a multipathway risk assessment methodology, and 2)
calculate a soil concentration for a  single chemical which will be health-protective for sites
with localized, clearly defined soil contamination, provided that no significant contamination
of ground water or food has occurred.

These procedures  present  the  TSCP interpretation  and application of those equations
endorsed by EPA which estimate the chronic daily intake for a single chemical via multiple
exposure pathways. In the absence of comprehensive sampling information the Standard
defines those conservative assumptions about environmental transfer of soil contaminants
into other media of exposure which will be required in  order to utilize the SRL process.
Similar conservative assumptions  about human intake rates will also be required in the
absence of site-specific information. The equations in the SRL may then be rearranged to
calculate a final soil concentration which would be protective of human health,  A sample
spreadsheet illustrating the process is supplied.

-------
               Appendix B
DECISION TREE GENERATED BY WORK GROUP

-------
                        DECISION TREE FOR DEVELOPING SOIL CLEANUP GOALS BASED ON
                                               GROUNDWATER MIGRATION PATHWAY
   DECISION FACTORS
1 Aden L»v»li include but
  are not tinted to
  •  Background
  •  MCU<100
  •  L«acNngte«t
  •  Based on pjrftortng
    coefficient (Ke of
 models.

 Exarnpleiof rnnrrun
 additional data requirement*.
 ndude but are not limited to:
 •  Bodegradaion
 •  Volatlizaton
 
-------
m
TJ

S
*.
o
53
CO
"  c
   
~o ^
 >  m c
CD  3  3
 H>  <  «
 3  -i  (B
 n  O  Q.


"355
    (I)  Q)

    2  *
    01  <"
o
Ul
      OJ
      O
      o
                              o
                              3
                          O 5" O

                          S°I
                          5  3  
                          en
                          o

                          ro
                           3
                           (D
                                  
                     o

-------