No
EVALUATION OF
IMPLEMENTATION OF
RISK-BASED DECISIONMAKING IN RCRA
Annotated Briefing
Office of Management Systems and Evaluation
Office of Policy, Planning and Evaluation
U.S. Environmental Protection Agency
January 1987
Protection
-------
k>
-------
INTRODUCTION
The Program Evaluation Division (FED), Office of Management Systems and
Evaluation (OPPE), conducted a study of implementation of risk-based decision-
making in RCRA from May 1986 to December 1986 at the request of Marcia E.
Williams, Director of the Office of Solid Waste (OSW).
This annotated briefing presents the results of PED's evaluation,
including recommendations for facilitating the implementation of risk
assessment and risk management in the RCRA program in a way that supports
the Regional Offices and States.
The PED Project Team consisted of Marilyn Stone, who served as Project
Manager, as well as 'Judith Johnson, Eileen Sheehan and Walter Simon. Katherine
McMillan, PED Branch Chief, provided guidance and oversight of the project.
-------
-------
TABLE OF CONTENTS
PURPOSE OF PROJECT •
MAIN MESSAGES "
STUDY APPROACH
GENERAL FINDINGS
Broad Support Exists For Consideration Of Risk
But Support is Cautious and Qualified
Serious Concerns About Risk-Based Aporoach
Concern And Confusion About HO Intention?
CURRENT APPLICATION OF RISK
Risk Flays Only Minimal RDle In Current Decisions
FUTURE OPPORTUNITIES FOR USE OF RISK
Field Views About 'JSP. Of Risk In Future ?roaran Areas
Strong Support where flexibility Fror. Standards Needed
CAPABILITY OF RSGIO'J.-' AND STA.TE?
Field is Not Prepare;: For Making Risk-Based Decision?
5 u MMAR Y 0 F F I ND I N'G a '
RECOMMEHDATIQty
Set The Record Straight .
Apply In Limited Program Areas Vow
Build Performance And Credibility In ACL"
Build Capability ~,~ T';e Field
Develop Tools T.-; Sir.rlify Si te-Pnec if ic Decisions
Take I-. one stf-p At A ri-e
AFTERV.'ORD .
3
4
5
6
7
R
10
11
14
1 S
17
1 Q
22
27
-------
-------
-3-
PURPOSE OF PROJECT
STUDY TO PROVIDE AN EARLY LOOK AT IMPLEMENTATION
The Office of Solid waste (OSW) had expressed its intention to move
the technology-based RCRA program toward a more risk-based approach
to decisionmaking. Accordingly, OSW asked the Program Evaluation
Division (PED) to examine the implications for Regional Offices and
States of implementing risk-based decisionmaking in RCRA.
FINDINGS TO FEED INTO OSVv STRATFGY DEVELOPMENT -
- The findings from'this study should prove useful in guiding OSWs
development of a detailed strategy for integrating risk assessment
and risk management in the RCRA program.
STUDY ADDRESSED THREE AREA":
Openness and resistance to risk approach
0 vie explored the field's overall reactions and receptivity to the
use of a risk-based approach in RCRA. Region and State nersonnel's
perceptions of risk have to be understood if risk is to be factored
into decisions in the RCRA program.
Current application and future opportunities
0 PE;i sourjht to document where Regional Offices and States currentlv
consider risk in the RCP.A program and to identify other areas
where risk could be applied in the future.
Current capability and future needs
c The ability of Regions ann States to use risk assessnent tool? ar. •
make decisions based on risk greatly affects the prognosis for'
using risk in RCRA. The study explored the field's current ability
to co;:sj..:er risk an~ identified steps to build capability for
future.
-------
-------
-4-
MAIN MESSAGES
REGIONS AND STATES GENERALLY SUPPORT CONSIDERATION OF RISK IN
RCRA DECISIONS
BUT SUPPORT IS MIXED WITH REAL CONCERNS ABOUT WHERE IN THE PROGRAM IT
WILL BE APPLIED AND HOW RISK WILL BE FACTORED INTO DECISIONS, AS WELL
AS CURRENT ABILITY TO DO IT CONSISTENTLY AND WELL.
o RISK PLAYS i.ITTLF
IS LIKSLY TO I!,V'?.'
"LF IK C'lRREr.'T REGIONAL DECISIONS, BUT ITS
.^? A~ CO?.»FCTIV%: ACTION NOR" PROGRESSES.
O P.SGIO'!= f\\:'j STATES CURRENTLY LA"f AH EQUATE SKILLS AN~> KNOWLED~F T'
ASSESS: RISK AN"- SUPPORT ^ISK-EASF.D DECISIONS
O STRATEGY SHOULD BE "WALK BEFORE WE Rn»:," FOCf.'SINI OH A FF'v KEY
DECI3IO'::-" f,"i;-.' -•-.';:: D"VF.:^":T. (:•'-.?-.=••:• rrv, TOOLS, AND CREHI-TMTY FO-
XH*** r* f""'"^^"
-------
-------
1
-5-
STUDY APPROACH
REVIEWED AGENCY AND' OSW DOCUMENT?
- PED began by reviewing Agency and OSW policy statements and speeches
on risk assessment and risk management to gain a general understanding
of these terms within the Agency. The team also examined a number of
RCRA regulations to identify the current decisionmaking framework and
the role risk plays.
DEVELOPED BROAD DEFINITIONS
Risk assessment traditionally includes four steps: hazard identification,
dose/response, exposure assessment, and risk characterization. For
the purposes of this study, PED considered any one or combination of
these steps as risk assessment activity.
PED also defined risk assessment to encompass either quantitative or
quantitative consideration of health and environmental risks.
Similarly, a decision was considered risk-based if it was based
either entirely or partially on risk.
BEGAN KITH HO IN QSW, OPPF, OE?-
- PED interview^;? approxinately 25 managers and staff in Headriuarters
to discern specifically what they mean by risk-based decisionnakinc
anri where they see the opoortunities to incorporate risk into
INTERVIEWED __ft_LL 1 0 REGIONS
PED interviewed over SO Regional personnel, bv nhone or in
The tean visited "egicns IV, V and VIII and interviewe-3 the othe-
Regions by phone. In all Regional Offices, we spoke with t'-ie =C=-'.
Branch Chief and conducted separate interviews with one or more
staff members.
TALKED KITH 9 STVrE?
P'.:;j int^rviewc-c'; nine states, some of which were seen as leaner- in
risk-relatec a-:tivitie? ann others chosen at random, we visiter-
three of t :>* nir:*- States - ^ecrgia, Colorado an,i "ichican - an.-5
interview-?.': t'.-r others? v.' pho-:'-. T1*. total, we conducted over 2S
ir.tervie.;b with State oersonnel.
-------
-------
GENERAL FINDINGS
BROAD SUPPORT EXISTS FOR CONSIDERATION OF RISK IN RCRA
EPA Regions and States express support in general for a greater emphasis
on risk in RCRA. This support stems from a recognition of the advantaoes of
a risk-based approach and from practical needs in decisionmaking and in
responding to public demands.
A RISK-BASED APPROACH COULD PROVIDE APPROPRIATE REGULATION AND BETTER^
USE OF RESOURCES
Support is based in part on a recognition that risk-based decisionmaking
can provide appropriate regulation by tailoring the protection and
controls to the risks posed at a RCRA facility. '
A risk-based approach car. also result in a better use of resources,
both the regulator's resources in setting priorities and society's
resources, bv not requiring unnecessarv, and potentiallv costly
controls.
THE FIELD NEEDS FLEXIBILITY FROM PTANDAP.DR IN SOME ARFA =
A primary basis for support, however, is the practical need to varv
fron standards in so:r>» casos, oarticularly from the background stan^at
in clean-ups. Evaluation of the risks can provide an alternative
basis to determine what action? are needed.
REGIONS AND STATES RECOGNIZE MORE CONSIDERATION OF RISK IS
Regions and States acknowledqe that risk will' plav an increasinalv
inpsrtar.t rol» ir. decisi'-r.p in ve-.rc to co^.^. This recr>onitior. is
basec: on the Agency's statements about the need to focus limited
resources on t'^e areatest environmental problens "and to ensure cur
action? reflect ris"-:. In addLrior;, the field is increasinclv faceci
wit'n de.Tian'.:= fro~ the public for infnrnation about the risks facilities
p"se an"i t1":^ ri.^k0 t5";.^". .•.-.:.:] r^^^it fro1*1. Acre".:y actions.'
-------
-------
-7-
GENERAL FINDINGS
BUT SUPPORT IS CAUTIOUS AND QUALIFIED
While support for increasing the role of risk is widespread, that suooort
is contingent on where, how, and when a risk-based approach is adopted in
RCRA. In general, the field believes the shift to a more risk-based approach
should be implemented gradually and carefully.
LIMIT USE TO A FEW KEY PROGRAM AREP5, ESPECIALLY IN THE SHORT TERM
Regions ana states support risk-based decisions in a few program
areas, not imposed across the entire proqran. They do not believe
it is appropriate or necessary for all parts of the program, and fear
that significant change;; to the basic structure of the current regulatorv
framework would create too much disruption.
USE AS A TOOL
COMBINATION KITH TECWOLOGY
The field se'ss risk assessment as one tool in their arsenal, not
necessarily as t'.ie sole basis for making decisions. Regions a1")-*
States repeatedly stressed that they want to use risk infornation
in combination with technology considerations in. making decisions.
USE SPARINGLY WHILE GUIDANCE AN'O SKILLS ARE
Regions and Stares cautioned that EPA not get ahead of their abilitv
to do risk assessment and make decisions based on risk. They ey.pr<=>s
support for' ora-'uallv increasing ".'""• rroaran1 s reliance or. risk a= *
decisionnaking tool as the capability to reliably assess risk grows.
-------
-------
-8-
GENERAL FINDINGS
SERIOUS CONCERNS ABOUT RISK-BASED APPROACH AFFECT SUPPORT
Despite overall support, Regions and States, and many in Headauarters as
well, raised serious concerns about implementation of a risk-based approach.
The concerns described below represent their perceptions of the problems of
associated with risk-based decisions, which may be similar to the problems
encountered in any new and complex endeavor and may to some extent reflect a
resistance to change. However, our interviews suggest these concerns are
based on real implementation problems and on deeply held values and principles.
Regardless of their source, these concerns must be recognized and addressed
if risk-based decisionmaking is to be successfully incorporated into the RCRA
program.
o FIELD LACKS SKILLS TO ASSESS RISK
Regions and States currently do not have adequate capability to
assess risk reliably and make decisions based on risk. RCRA personnel
lack a general understanding of risk assessment and how to factor
risk assessment results into specific RCRA decisions.
They also lack specific expertise and knowledge in key areas, such as
toxicology and modeling of chemical fate and transport, that are
needed to conduct risk assessments.
TIME-CONSUMING AND RESOURCE-IVTE\'£IVF
RCRA personnel in the Recions and States exnress concern that a
risk-base.i approach introduces "one more step," and a potentially
time- ani resource-intensive step, into an already heavy workload.
Assessing risk to support decisions takes tine, in cart because it is
data-intensive. Risk assessment requires the up-to-date toxicoloaical
data on chenicals as well as a significant anount of data to character!?:':
the specific site.
PUBLIC RESISTANCE ' •
-. RCRA staff and managers virtually all agreed that public perceptions
about risk-based actions presents a serious obstacle. The public
does not unriersta"..: risk assessment or the use of probability in
decisionmaking. While there is increasina demand for risk information,
the public tends to want, risk to be re^uce.^ to zero; a cancer risk of
o'\°. ir. a rill i.or. is 0:1 »s too nany to the citizen who is convinced he
may be the "one." The public wants the most stringent controls
possible and is often more accepting of technology solutions that
represent' the best that can he achieved.
Public re--.i:-.~-z<;~~ ? = r. -,~".~y n-:tio.".«, requires extensive justification.
of decisions, an.i -.ay ^a->- ra^naa'ers reluctant to make decisions that
COM!--* potentially r•»<=;.;it in significant public controversy.
-------
-------
—9-
GBNERAL FINDINGS
SERIOUS CONCERNSAFFECT SUPPORT (cont.)
ADEQUACY OF ENVIRONMENTAL PROTECTION
Many in the Regions and states Questioned whether risk-based decision-
making will ensure adequate protection of health and the environment.
Repeatedly, we encountered the perception that risk-based decisionmakino
equates, in effect, to a "license to pollute." While such decisions
can result in greater protection, many believe that a risk-based
approach will almost always result in less stringent controls than
current standards mandate, and will in practice — and perhaps as
intended by its proponents — be used to subvert technoloay standards.
In practice, risk assessment tends to emphasize human health and deaths,
especially cancer deaths. As a result, there is concern that decisions
based on risk assessments tend not to give adequate weight to environ-
mental impacts or to recognize the value people place on a clean
environment, Quite apart from the threat of lethal human effects.
The field also takes RCRA's prevention goal quite seriously. They
see risk-based decisionmaking potentially at odds with this goal
since it can lead to allowing some pollution rather than seekinq to ,
minimize it. Many express a philosophic belief that it is both
equitable and appropriate to require companies, as a cost of doing
business, to meet minimum standards that reduce the likelihood of
damage and deal with pollution that results from the wastes they
generate and manage,
UNCERTAINTY OF RISK ASSESSMENT SCIENC?
Many of thosfi interviewed express a basic distrust of the "science"
of risk assessment. This skepticism that risk assessment can accurately
measure risk increases concern about the adequacy of protection
afforded by risk-based decisions.
Some feel that risk assessments rely on too many assumptions, leading
to unreliable conclusions. In one State, a model, using simplistic
assumptions, predicted that a plume of contamination at a site would
travel less than 200 feet in 20 years. In fact, the plume traveled
that distance in a matter of months. Incident-; like this, while not-
true for ail modeling efforts, increase distrust of risk assessments.
Some believe it i<= too easy to manipulate risk the assessment
2y changing one assumption or omitting certain data, the outcome of a
risk assessment can be altered. They fear that th*» reoulated communitv
and the public will manipulate mo^el^ to their advantage.
Gaps in data on many chemicals further erode confidence. If the
toxirclogical data are not available or are inconclusive, the ri«k
assessments reflect these problems.
GiVi-ri the- uncertainty of the ri?k assessment science, acknowledci*>J
even by its proponents, there is concern about inconsistent decisions,
::'i, t1"! s i ~ : : .•-I" ci^c.c Krs i n T f -S>J>>-P-: o i *f o "p-itl v . Hncertaintv n,-. •_• a2c-
ri '•:•.- ; -i t -i r. -7 • : of fi«e .- _:-• i o--.-- an-: er: f orce;-,^.-1. ^ actions more difficult.
-------
-------
-10-
GFNFRAL FINDINGS
CONCERN AND CONFUSIONABOUT HQ INTENTIONS
IS INCREASING RESISTANCE
While the concerns about implementation are real, the extent of the
problems may be magnified by confusion in the field, and among Headauarters
staff as well, about what "a move to a risk-based approach" really means and
what strategy OSW management intends to pursue.
FIELD FEARS BROAD CHANGES IN CURRENT PROGRAM
Regional and State RCRA staff expressed a fundanental concern about
the use of risk in RCRA decisions. Speeches and strategy documents
from the Agency and the RCRA program that imply a wholesale shift to
a risk-based approach are fueling fears that management wants to
alter the basic structure of the RCRA program.
Few support broad alterations to the current regulatory structure of
the RCRA program. Major changes would seriously disrupt ongoing work
in RCRA and significantly threaten oroaress towards statutory deadlines
that Regions an;5 States are struggling to meet.
FIELD VIHV;S RISK ASSESSMENT AS ALWAYS QUANTITATIVE
We found a t-^rv.iency in the field and in Headauarters to define risk
assessn-i.it as a strictlv quantitative process, rather than recognizincr
that decision? could be base.", on qualitative assessment of risks in
sone case5?. This perception increases concerns about the tin*», data,
and capability required to implement a risk-based approach.
RECENT ME:r.-AGE£ SUPPORT THESE
Recent speeches, articles, and new risk initiatives seem to supoort
the concerns of RCRA staff that a wholesale and rapid shift to a
risk-based program is envisioned. The messages often contain quali*i-
cations that limit the scope and timeframe of the initiative to
incorporate risk into the program. However, oiven the dearee of
apprehension in the fielo about n-~t»*'.tial rhanoes, the caveats are
not getting through.
The field's relative lac'-: of experience in applyina risk and doino
risk assessments, combined with the broa."5. statenents about making
RCV- a risk-be:'-1^' pr-jcru', are in 7 rising resistance to the crograr'=
in
-------
-------
-11-
CURRENT APPLICATION
RISK PLAYS ONLY MINIMAL ROLE HJ CURRENT RCRA DECISION?
FEW REGIONS HAVE DONE QUANTITATIVE RISK ASSESSMENTS
Except for a few risk assessments done in response to public pressure,
often for incinerators, few Regions have conducted quantitative risk
assessments for the RCRA program. Some have undertaken steps in the
assessment process such as toxicological evaluations for alternate
concentration limit (ACL) petitions or examination of exposure pathways
for RCRA Facility Assessments. In general, however, experience with
quantitative risk assessment is very limited, particularly in perforninn
toxicology assessments and modeling.
- Several States interviewed frequently perform risk assessments to
guide decisions. In most cases, quantitative assessments of risk are
done ad hoc when questions arise. However, one State routinely uses
assessments of the risks from specific chemicals to set limits on
discharges to air and surface water; methods for using risk to issue
discharge permits for ground water are under development. However,
even in these States, risk assessments are rarely done for the RCRA
program.
USE _OF^ QUALITATIVE ASSESSMENTS MORE COMMON
The roost connon use of risk factors to guide action is in orioritv-
setting. For this purpose, risk is evaluated Qualitatively to determine
which RCRA facilities to address first, whether for permitting,
enforcement, or other activities. Virtually every Region has SOT?
type of system that qualitatively compares the risk presented bv
-different facilities to set priorities.
Oth«r factors, however, limit the extent to which priorities are
based oi. risk, most notably national priorities which emohasi^ft
statutory dea.iline= ani permitting land disposal facilities. Several
Regions and States felt that the priority placed on meetina on stfl^utTv
deadlines to some extent impairs their ability to target actions to
the highest risks.
For example, one Region, after evaluatinc its facilities, found that
one-third o* their highest risk f«~ilities were not operating lan^
disposal facilities, but, because of the 1988 deadline, the Region
could not make these these facilities a top priority. On" State
believes' that ha2'irdou
-------
-------
-1 2-
CURREHT APPLICATION
RISK PLAYS ONLY MINIMAL ROLF (cont)
The ability to determine the risks posed by facilities is also limited
by the amount of data the Region or State has on each facility. Most
Regions have not systematically identified or organized data available
from sources such as permit applications or inspection reports on all
facilities, while knowledgeable about the conditions at a subset of
facilities that have received attention in the past, such as commercial
land disposal facilities, they have only a limited understanding of
the universe of waste facilities. Schemes to rank facilities accordina
to risk are therefore biased towards those facilities where data is
most comprehensive and thus may overlook some facilities with significant
problems.
To some extent, a qualitative sense of the risks posed by a facilitv
influences the actions taken at that facility, although situations in
which risk information is used in this way are difficult to identify.
Several permit writers believe that their sense of the seriousness of
a facility's problems affects the extent of data they reauire and the
conditions they place on the facility. Several States and Regions
mentioned that risk plays a role in the actions taken at closures,
because the regulations and guidance on closures are relatively
general and thus allow some flexibility to consider risk.
ACLS ARE EXPOSING REGIONS AKD STATES TO RISK CONCEPTS
In discussions about current application, the field tended to focu=
on ACLs, because this is o.=u- area of the program that is risk-based
and with which the Regions and States have some experience.
Experience to date with ACL petitions is limited; the Regions
receiver, about 20 petitions, but the number of ACL petitions is
likely to grow dramatically as the number of facilities facincr corre.-tive
action increases, makino ACLs a more attractive alternative for the
company because of the high costs of cleanup.
Currently, most ACL petitions are in the early stages. State= an^
Regions are either conducting initial reviews or have reaueste.-l
additional information. Thus, many Regions and States have not
really begun to analyze risk information in depth or act on petitions.
At the time o£ our interviews, there had been no final decision? en
ACLs, although "at least one proposed decision on a relatively simple
ACL petition was nearinc publir review.
In a few Regions, however, experience with ACL petitions was confirnina
the field's fear? about implementing the ACL provision anri, for some,
about risk-based decisionns>ing in general. For exanole, thev had
found that reviewing oetitions t.V'.^s considerable time .anr resource?:
one petition n i.'it-- ; :>•* ~. :- 1 ••: vo ", i_:->'-
-------
-------
-1 3-
CURRENT APPLICATION
RISK PLAYS ONLY MINIMALROLE (cont)
In one Region, an ACL petition is sparking disagreements, raising
tension among the staff, management, and between Region and State.
Differences of opinion about exposure pathways and future uses of the
land around the site illustrate the difficulties in reaching consensus
about how to characterize risk and act on that analysis.
In another Region, an ACL petition is confirming fears that companies
will attempt to use the ACL provision as a means to subvert their
responsibility to clean up contamination their activities have created.
The company had purchased 1000 acres surrounding its 80-acre site,
and argued that, since it controlled access, health effects need not
be considered. They were asking that ACLs be set at or above current
contamination levels, in some cases three orders of magnitude higher
than current levels. The Region and State were very uncomfortable
with the argument, which would in effect write-off future use of the
land, but had difficulty rejecting the petition strictly on the basis
of risk.
Draft guidance in place at the time of the study allowed the use of
such "institutional controls" as a basis for granting an ACL; revised
guidance will invalidate the argument described above. However, such
a petition, whose premise was oosftd in terms of risk and appeared to he
in conformance with Agency policy and guidelines, has encouraged
concerns that the risk-based approach will be •apolie'l in a wav that
compromises environmental protection. While this specific issue has
b-.'en resolved, it illustrates the need to design risk-based decisions
carefully, in a way that will not create distrust in the field of t'->e
decisionnakinc approach.
FIELD HANTS TECHNOLOGY AND RISK TO GUIDE CLEAN-UPS
- While Regions expressed support for the concept of ACLs in general,
most prefer that cleanup decisions not be based solely on risk, but
consider available technology and what it can achieve as well. In
effect, 'they prefer that facilities undertake some corrective actior
to see how much contamination can be reduced, then consider risk
through the ACL process to see whether additional corrective action
is worthwhile. Several Regions are'following this aporoach with some
success, at least with operating facilities that want a RCR» oerTit
and thus have an incentive to comply with Reqional request1?.
The support for a "technology first, risk later" annroach in most
case<= is baseH, first, or. the belief that facilities shTj'M He hel T
responsible for environmental damage. Second, while it may take a
year or more tc assess risk- and make decisions on ACLs, some correctivo
. actior. can begin within a few months to reduce contamination an^
minimize further migration of the plu^e. Third, the public is more
l:"<.= ly to rio'e^i <=• go-;1, basel on risk if so^e corrective action, an^
thas so-.v r»;;u:-.L->n in dairuvje, hns taken plac^. In fact, several
States say they will not aonrove anv ^CI-p because the comrvinitv will
-------
-------
-14-
OPPORTTJMITIES
FIELD VIEWS ABOUT APPROPRIATE USE OF RISK
IN FUTURE PROGRAM AREAS
To identify opportunities where a risk-based approach could potentially
be applied, PED asked the Regions and States to react to the use of risk-based
decisionmaking in a few selected RCRA program areas.
SELECTED A SAMPLE OF FOUR POTENTIAL PROGRAM AREAS
Staff at Headquarters identified a number of possible areas where a
risk-based approach could be used. From this, PED selected the
following program areas to gauge field reaction:
0 Corrective Action for soli-i waste management units and
regulated units;
0 Closure and Post-Closur
-------
-------
-15-
OPPORTUNITIES
STRONG SUPPORT WHERE FLEXIBILITYFROM STANDARDS NEEDED
RISK SHOULDINFLUENCE"HOW CLEAN ISCLEAN" DECISIONS IN CORRECTIVE
ACTION AND CLOSURES
There is nearly universal support for using«risk as a factor in
determining how much cleanup is necessary, specifically in corrective
action and closures.
Support is based on a recognition that cleanup to background is not
technically or economically feasible at all sites. At those sites,
an alternative basis is needed to determine how much cleanup is
warrented.
Despite support for factoring risk into corrective action and closures
decisions, Regions and States feel that a risk-based approach is not
appropriate or needed at all sites. Moreover, it should be used in
combination with a technology-based approach, and applied conservativelv
to ensure adequate environmental protection.
FIELD SKEPTICAL ABOUT RISK-BASED WAIVERS
- While Regions and States generally support ACL waivers as a way to
factor risk into clean-up decisions, they do not support risk-based
waiver and exemption provision* in general, and particularly for the
examples provided of secondary containment for tanks and test burn*
for incinerators.
Most Regions and states believe that waivers from these technical
standards would not be worthwhile because the cost of complying with
these requirements is not prohibitive, but the implementation costs
of assessing risk for each waiver provision would he significant,•
taking valuable time and resources away from more important work.
One Regional manager commented that reviewing these oetitions would
be a very low priority in his Region.
Regions an3 States are also concerned about the'difficulties an--1
resources required to cope with public opposition to waivers fror-
regulatory standards. The public is opposed to waivers in general
because they, bv definition, exempt facilities from basic technical
requirements.
-------
-------
OPPORTUNITIES
RELISTING FOUND STRONG SUPPORT FROM SOME AMD 'SERIOUS OBJECTIONS
FROM OTHERS ~ , "
There is wide disagreement about Headquarters' effort to redefine
hazardous waste, with opinions varying according to how each respondent
weighed the potential disruption of the program against the need to
improve the definition of hazardous waste.
Those who oppose relisting are concerned that it will cause too much
disruption, changing the basic structure of the program bv, for
example, altering the universe of Subtitle C facilities subject to
RCRA. They also fear it would make an already complex program even
more complex.
Those who favor relisting believe a redefinition of hazardous waste
is worth the disruption because the present definition is too illogical
and problematic, regulating wastes that, in fact, present little
hazard. «
In general, the Regions were concerned primarily with the potential
disruption to the program, while the States placed more weight on the
to inorove the definition.
-------
-------
-1 7-
CAPABILITY
FIELD IS NOT PREPARED FOR MAKING RISK-BASED DECISION?
RCRA STAFF AND MANAGERS LACK GENERAL UNDERSTANDING AND SPECIFIC SKILLS
- Risk-based decisions represent a significant change from technology-
based decisions in the kind of analysis and weighing of factors it
requires. Given the predominence of engineering and chemistry hack-
grounds among field personnel, RCRA staff are not equipped by education
or experience to readily understand and use risk assessment tools.
They do not really have a sense of what risk numbers mean. As one
permit writer exclaimed, " 'ten-tc—the-minus-six* is voodoo to me."
Managers as well are unfamiliar with risk concepts and can be reluctant
to defend risk-based actions to a skeptical public, especially given -
the wide bands of uncertainty associated with many risk estimates,
when they themselves do not fully understand the basis or implication?;
of the estimates.
Specifically, Regions and States personnel mentioned lack of capability
and experience with interpretation of toxicity data and evaluation of
fate an:3, transport models.
Given the lack of chemical standards that establish a uniform risk
level for a chemical, the field needs qualified toxicologists to
interpret toxicity data for risk assessments. Currently, toxicoloqv
expertise in the Regions is very limited. Two Regions have no toxico-
iogists, and only one Region has as many as three. Expertise in RC?.&.
is even mor» limited. Only one Region has a toxicoloaist in its RCRi
-ranch. When guestions arise about the toxicity of a chemical, RCRA
staff currently turns to a toxicologist in another part of the Reaion
or to Headquarters, but this expertise will not he sufficient to meet
P.CR^'s increasing denand for toxicity assessment.
t-'hile most Regional R.?RA Branches have hydroaeologists to characterize
sites, they have little experience with fate and transport models.
They lac*; confidence in their ability to accurately assess a proposed
mod^l ann also question whether models can in fact predict movene.it.
at all types of sit»<;. The Regions need specific technical guidance
to help the^i evaluate which models are aopropriate for different
sites.
State capability ranges from good risk assessment sunport to virtually
none. A few States have adequate support for their risk assessment
needs, either centralized in one orouo that supports all environmental
programs or located within earh me-lia progran. Most States, however,
have very limited capability. They obtain some suooort, particularly
i>: toxicology, fron u'iversitie? or other state agencies. Because of
their ow:, li-nit.-,tions, the Regions cannot provide much toxicology
to th^ir States.
-------
-------
-1 8-
CAPARILITY
FIELD IS NOT PREPARED (cont)
LACK OF GOOD DATA AND STANDARDS IS A MAJOR PROBLEM
Both Regions and states expressed repeated concern about the adeauacy
of data for doing risk assessments. Basic data on the health and
environmental effects of constitutents are not always available, which
decreases the reliability of risk estimates and erodes confidence in
decisions.
- Some, particularly in the States, believe that the development of
national standards for commonly found constituents that establish
"acceptable concentrations" is an important, even a critical step in
enabling risk-based decisions to be made consistently and with relative
ease, without such standards, each Region or State must conduct its
own assessment of toxicity studies, with potentially different results,
and must defend its conclusions against legal challenge.
The field is also concerned about the adequacy of scientific knowledge,
for example, about the effects of mixtures or the impact of an anaerobic
environment on chemical fate. There are also concerns about the
•
scientific basis to accurately assess potential exposure and predict
movement of constituents, especially in soils and in ground water at
certain sites.
Equally important, Regional staff has limited access to existing
chemical data and research results in the field; data and scientific
information are not well organized or readilv accessible. The access
problem in some states is m^re acute. One State relies only on a few
handbooks for toxicity information, and has no budget to exnand its
library, or to gain access to computerized data bases that car. provide
more up-to-date information.
Finally, RC?.^ staff has difficulty obtaining the site-specific data
they need to fully characterize sites for risk assessments. In
contrast to the CERCLA procran, in which in-depth site assessments
are conducted to support decisions, RCRA staff generally works fro~ a
limited base of monitoring data. In one case, evaluations by the two
programs of adjacent sites, one a CERCLA site and the other a °C?i
landfill, produced very different conclusions. Based on the mininal.
monitoring required under RCRA, there appeared to be no movement of
ground water. THP investigation at the nearby CKRC.',' site deterrin^
ground water flowed upward, almost vertically. The accuracy of the
more extensive CERCLA investigation was confirmed when the barrels in
the now-close-! RC15.-. landfill began to float.
RCRA also relies on owners and operators to provide data, which can
d^lsy site assessr.er!-. RCRt. staff must know and be able to define
specifically what dat.^ are required accurately to ensure thev can
obtain the infDrnatior. t'iev nee."1 to accurately characterize the site.
-------
-------
_1 O_
IN SUMMARY
THERE IS SUPPORT FOR CONSIDERING RISK IN RCRA
- While current application of risk in RCRA decisions is limited,
support exists in Headquarters, Regions, and states for increasing
the role risk plays in the program.
BUT FIELD CONCE?.i:S MUST BE RKCOGNIZEO
In order .to successfully move the RCRA, procrran to a nore risk-base,-!
approach, the underlying concerns expressed by Headquarters, Regions,
and Spates about hov risk-base^ decisionmakino will he implemented
nust be recoanized and addressed:
WHAT rio«s "a move to risk-based decisionmaking" really mean in RC
-- WHE~.E in the program will risk be incoroorated?
HOW will risk be factored into decisions, and how will the move to
risk-based decisionrtaking he accomplished?
WHE'i will a risk-base', approach he put into place?
-------
-------
-20-
RECOMMFNDATIONS
SET THE RECORD STRAIGHT
While it will take time to provide specific answers to questions the
field has about where and how risk factors will be incorporated in RCRA, OSW
needs to outline in general what principles will be followed in moving toward
a more risk-based approach.
DEFINE WHAT "RISK-BASED DECISIONMAKING" MEANS IN RCRA AND
CLARIFY OSW'S SHORT-TERM STRATEGY
The principles we suggest below address underlying concerns about
how, where, and when risk-based decisionmakinq will be applied and
seem to us consistent with OSWs philosophy:
Risk factors will be used in combination with technology
considerations and not replace them in decisionmaking.
- Risk assessment can be quantitative or qualitative,
depending on the specific decision.
Considorv,lion of risk in decisions will take place within
existing regulations, and will n~t result in broad chanqes
to the. current program, framework.
Incorporation of risk will accompany, not precede, development
of tools arid capability.
USE UPCOMING OPPORTUNITIES TO SEND THIS MESSAGE TO HO, REGIONS,
AND STATES
It is important that OSW clarify its approach as soon as possible,
using upconincr m^-tinrrs, speeches, and other contacts with HO, Reaion=,
and States to communicate its intentions.
It is e-j = lly inps-srtait that HO i" t*l':inq with the Regions, an-!
Regional personnel in their contacts with the States, understand
and reinforce this ness-3-re.
-------
-------
-21-
APPLY IN LIMITED AREAS NOW
RECOMMENDATIONS
We recommend that OSW continue to work primarily within the current
regulations in the short term, focusing on areas where consideration of risk
is most productive and most needed. Based on our findings, these areas are:
CLEAN UP DECISIONS IN CORRECTIVE ACTION AND CLOSURES
The consideration of risk information, in combination with technology
factors, is important in determining how much corrective action to
undertake, because flexibility from the background standard is needed
at some sites.
- While regulatory change might be required, there is support for
applying risk factors in evaluating the extent of clean up required
at closure and in determining whether a facility has clean-closed.
PRIORITY-SETTING '
OSW needs to ensure that Regions and States systematically identify
and organize information on all facilities to ensure their priorities
reflect risks as accurately as possible.
while recognizing the importance of statutory deadlines, OSW should
allov Regions and States as much flexibility as possible to use risk
as the basis for det^rminina which facilities are addresse-i first.
EXPANDING TO OTHER PROGRAM AREAS
Ask the question, "is it worth it?" In evaluating whether and how
risk should be used in future regulatory areas, OSW must recognize
the time, data, and skilled resources required to make risk-based
decisions. Consider the benefits of factoring in risk in light of
the cost.-i of implementation.
For exar.ple, ask whether the costs of over- or under-recialatinn are
minimal; if they arp, a risk-based approach, requirino site-snecif i
risk assessment, may not be warranted.
Ask whether public resistance to the use of risk in the program
decision might be too great.
Consider whether the burden on the field of implementing a
provision can be minimized. For example, can parts of the risk
aflsessrii-rr.t be standardized? Can uncertainty be tolerate or the
decision baseo on qualitative assessment? Can risk factors be inco
porated into regaiatlon, rat'->=- than reauirincr site-specific
jir'.d appl ic=> i ; or. of ri. "••'•;?
-------
-------
-22-
RECOMMENDATIONS
BUILD PERFORMANCEAND CREDIBILITY INACLS
Decisions on ACL petitions are risk-based and therefore provide a good
test of EPA's ability to evaluate risk and act responsibly on that evaluation.
progress to date on ACLs has been slow. Regions and States lack capability
and confidence in their ability to act on petitions, although new policy and
technical guidance has been recently been developed that will clarify how the
ACL decision should be made. To bolster performance on ACLs, we support OSWs
efforts to:
GET HQ INVOLVED IN INITIAL ANALYSES AND DECISIONS
Provide active HQ support on early petitions at both the analytic and
decisionmaking stages. While HQ PAT teams have been involved in some
petitions, HQ needs more extensive involvement during the learning
phase.
- .While Regions and States are developing capability, HO can undertake
some o.f the analysis and can participate in decisions to helo ensure
consistency across Regions. A better understanding of "the real
world" — the nature of the petitions that have been received and the
issues and questions that arise — can also improve the guidance and
technical assistance that HQ provides to the field on ACLs.
MAKE SOME GOOD DECISIONS: APPROVALS AND DENIAL?
A critical step in improving performance is to make sone ACL decision?
to demonstrate that the provision will he applied resoonsihlv an3
without compromising environmental protection. It is important to
deny as well as approve petitions to show that EPA will apolv the
provision stringently and not delay action indefinitely if a petitioner
has failed to make his case.
Given the natural reluctance of the Regions to "be first," wo'g
encouragement and participation will help produce action on ACL
petitions.
LOOK FOR SHORT-CUTS IN DECISION FRAMEWORK AND PROCESS
OSW, with field assistance, should examine the review and decision
process for ACL petitions to identify way? to minimize the burden on
the field and to simplify the assessment required at each site.
~or e.o-.,-!•», it may be possible to standardize steps in the process
by prescribing the types of models that can be used or the assumptions
the.* are incorporated into models. A screening process based on an
assessment of a few critical indicator constituents conld weed out
petitions e.a.rly that are not goo^ candidates; if the assessment finds
-.hat corrective a-iion will be needed for these constituents, th-s
petition, can he reject-?'; without reauiring a detailed assessment of
=U1 con*t: f.ue-it.^ fl t tho sit-*.
-------
-------
-23-
RECOMMENOATIONS
BUILD CAPABILITY OF THE FIELD
It is vital that OSW continue and expand efforts to improve the skills
and capability in the Regions and states to assess site-specific risks accurately
and make decisions that reflect risks. Activities to build capability must
reflect the somewhat different needs of RCRA managers,-staff, and technical
experts, as well as the specific decisions they face, to ensure that risk
assessment/ risk management, and communication of risks can be done well.
PROVIDE TRAINING AND TECHNICAL ASSISTANCE
OSW has developed a general training course, using a corrective action
case study, in the "art" of risk assessment and risk management.
Such training can be useful in providing a general understanding of
these tools and ensuring everyone speaks a common language. Training
courses should be tailored to reflect the different needs of staff,
managers, and technical experts.
- Equally valuable, however, is training on specific program decisions,
because the decision criteria and the type of risk assessment required*
varies for each decision. For example, training explaining the new
ACL and corrective action guidance can improve performance as well as
risk assessment and decisionnaking capability by emphasizing and
explaining the risk components in these decisions.
Technical assistance and training are also needed in specific areas
of the risk assessment process. One area mentioned repeatedly bv the
field is predictive modeling, where help is needed to determine what.
site data is needed aric which, models are appropriate for certain
sites. Another area is the interpretation of chemical data and the
appropriate use of tool« like the Integrated Risk Information Syste-.
(IRIS), which will provide hazard assessment information.
Activities that encourage the transfer of information and experience
anong Regions and States can also help build capability and improve
consistency. For example, OSW could develop a clearinghouse that
provides examples of individual analyses and decisions done in the
Regions as experience with risk assessment and risk-based decisions
grows.
?.-?gional RCRA staff should be encouraged to participate -lore active!-.-
in Regional-wide risk projects and risk coordination groups. Many
Regions have newly-formed groups of risk experts to coordinate and
support Regional activities relate^ to risk. Projects like the stu-^v
of capability needs in the Regions led by Region IV can assist Reaions
and HJ in improving ris'< assessment skills. To date, RCRA oarticinatior
in these efforts has been minimal, but greater involvement can be a
useful means to improve capability.
-------
-------
-24-
RECOMMHMDATIOMS
BUILD CAPABILITY OF THE FIELD (cont)
SUPPORT HIRING OF REGIONAL TOXICOLOGISTS
Training and assistance, however, cannot substitute for in-house
expertise. Regions and States need aualified tpxicologists to interpret
toxicity data where standards do not exist, to respond to needs for
immediate guidance when unanticipated situations arise, to serve as
expert witnesses to give credibility to decisions in contacts with
the public, and to assist the States. RCRA will increasingly need
more toxicology expertise than is currently available in the Regions.
DEVELOP EFFECTIVE RISK COMMUNICATION APPROACHES
OSW needs to undertake activities to support EPA's ability to communi-
cate with the public about risk. A systematic evaluation of current
experience, for example in CERCLA or the States, might identify
approaches that have worked, as well as pitfalls to avoid, and help
create a. more effective communication program in RCRA.
Providing HO experts, or perhaps expert witnesses from outside the
Agency, to assist the- Regions and States in public hearings can lend
credibility to decisions. Hiring specialists in risk communication
to help Regions formulate their strategy for public hearings an*, to
train Regional staff might also improve our ability to communicate
effectively vith the public about risk.
FUt;D RESEARCH EFFORT? TO IMPROVE SCIENTIFIC AND DATA BASE
It is critical to improve the base of knowledge -if EPA is to imorove
the quality and certainty of risk assessments. OSW needs to ensure
its research strategy is tailored to suoport the risk-based aporoach.
Regions and States identified a number of areas where risk assessment
science needs to be enhance;4.. Areas for improvement include the dat^
base on chemical effects, effects of mixtures, chemical fate of
constituents in anaerobic environments, models to determine and
pre.:i?t no'.'tw.-r.i nf wss-.-es in qround water, evaluation of contaminant*!
in soils, and environmental effects.
Equally i-=p-trtar.t is ensuring the field has access to th«s dat? an^
information that exists. Regions and, to a greater extent, States
h^v-r por;r dC-'-s- I--. e-:L^r.iiT t-.:i!-oloav data on chemical". Tran?e-
of r-3se«rc>. r.f.-
\\~. reports also need? to be done more efficiently,
-------
-------
-25- . .
RECOMMENDATIONS
DEVELOP TOOLS TO SIMPLIFY SITE-SPECIFIC DECISIONS
Given the time and resources required to assess risk and make risk-based
decisions, simplifying each decision as much as possible at each site will
ease the burden of implementation on Regions and' States.
O UNDERTAKE EVALUATION OF CHEMICAL HAZARDS FOR THE FIELD
- VJhile the decision criteria or risk assessment process may vary for
different decisions, the interpretation of toxicity data and decisions
about effects associated with concentrations of a chemical should
not. To minimize the need for each Region and state to undertake its
own evaluation, with potentially inconsistent outcomes and duplicative
use of resources, HQ should undertake the evaluation of individual
chemicals at the national level as much as possible.
The Integrated Risk Information System (IRIS) under development in
ORD is a vital first step in getting national guidance on chemical
toxicity to the field. However, while the workgroup supporting IRIS
has reached conclusions on nearly 100 chemicals, the schedule for
getting the information to the field has met with delays. OSW needs
to actively support and track IRIS' progress to ensure that problems
are resolved and the effort receives the priority attention it deserves'.
IRIS, however, is intended only as a first step in assessing chemicals;
it will not have the leqal weight of national standards. National
standards would ensure consistent decisions on chemical hazards and
ease decisionnaking. As a first step, OSW should quickly resolve how
the new MCLs being developed as required by the Safe Drinkinq Water
net will apply ro RCD.-. .
o DEVELOP MODELS, STANDARD ASSUMPTIONS AMD SCREENING PROCESSED
- 0-~"-',' needs to examine each decision that will require a site-soecif ic
assessment for ways to simplify and standardize the assessnent, an-i
thus reduce the burden on the field. The possibilities for standard-
isation and for tools to aid implementation will varv for each decision.
For example, ic nay be possible to develop standard models, like
used in delisting, that would guide the decision based on only minimal
site-specific data. The development of screening processes for
waiver provisions could allow the field to reject "frivolous" petitions
early on after mini-rial analysis and thus concentrate its efforts or,
t'r.e best candidates.
Standardisation nay increase consistency, but tends to trade off
accuracy , producin-j answers that are not appropriate in all situation^.
However, all decisions need not be perfect. For example, in some
cases, the cost'? to the facility of a wrong decision may not he
significant, thus justifying the use of a standard assessnent that
errs on the si-ie of over-protertion. If th« ranqe of actions ic
limited, far exa-ple when orl1/ a *>-•• clean un technoloaies are nossiv>3'1,
t; •; a:.dly-sis of risk fcr enc'ri constituent or.lv has to be sufficiently
n to choo=A correr tl v ^nonu tho<=«? t^ohnc.locrv ootions.
-------
-------
-26-
RECOMMENOATIOMS
TAKE IT ONE STE_P__AT__A TIME;
RECOGNIZE THIS IS A LONG TERM EFFORT
There are many difficulties to overcome in moving the RCRA program to a
more risk-based approach. It is important that OSW continues its inititatives
to respond to the problems and needs that Regions and States face in makinq
risk-based decisions.
O START NOW TO ADDRESS THE PROBLEMS OF IMPLEMENTATION
OSW can take immediate actions to improve field capability, resolve
policy and technical issues, and develop tools to simplify risk-based
decisions. A short-term strategy might include initiatives to provide
general and program-specific training in risk concepts, improve
access to data and research results, ensure that the research strategy
supports the risk-based initiative, and support decisionmakinq on ACLs.
Recognize, however, that the development of field skills and a regulatory
framework for a nore risk-based approach in RCRA is a continuing an3
long-term effort. Like any new and complicated endeavor, it will
take time to build the necessary base of experience and capability.
The shift in program emphasis must be done in a way that minimizes *
disruption of ongoing activities and priorities.
O CHARGEr RISK ^WORKGROUP WITH FORMULATING SHORT-TERM INITIATIVES ANO
LONG-TERM STRATEGY
The mechanise to develop a detailed agenda of actions to address the
problems oc implementation is already in place. Use the workaroun
set up under the Office of Policy Planning and Information to follow
through on recommendations and develop a detailed strategy that will
enable the Regions and states to "make decisions based on risk.
Include representatives from both HO and Regions on the workaroun so
the strategy reflects the experience and ideas of the field. Test
the workgroup's recommendations and approach in sessions with Reaional
and st^itft personnel t-> ensure they are realistic.
MAKE IT SOMEONE'S JOB
Assign one office or individual responsibility for coordinating an?.
tracking risk activities in RCRA. Responsibilities could include
ensuring that the activities laid out in the risk strateqv are carriej
out, ti-a.~;-:ir.,: progress anl trainin-r in assistance tools like IRIS^
and working with those developing regulatory guidance and assistance
to the fiel-i to ensure risk components are highlighted.
Make the office an,; its responsibilities known to the Regions, States
r.nd H ;. • "iv a forial p<-.;.':-, for risk activities, 'it car. areatlv i^nrov.?
r:oranu:ii-7ation. between HO an-i fieln personnel by relaying information
t ri.y'-: />-tiviti«v; ani resnomii rv to Regional and H^ auestioris,
-------
-------
V
V?"
-27-
AFTERWORD
The findings and recommendations described in this report are based on
information collected^from July through September, 1986. Since that time,
some of the policies._tha£ affect risk-based decisions have changed. OSW has
also begun a-number of?-activities and made progress on several fronts to
address implementation need's. For example:
»
•^- • - •—%.
o OSW lias revised the.ACL guidance, which is in Agency-wide review and
will soon become final. The new guidance significantly changes current
ACL policy. For example, a facility must clean up all off-s~ite ^contam-
ination, with a few limited exceptions; the use of "institutional controls,'
or the purchase of land to extend its boundary to avoid clean-up, is no
longer allowed.
o The first decision granting an ACL, in draft at the time of the study,
has been issued in a final permit. ,
o OSW has made significant progress in getting guidance to the field on
the hazards associated with constituents. An Agency-wide workgroup has
completed an evaluation of approximately 200 chemicals; the assessments
for nearly 100 of these chemicals have been reviewed within the Agencv
and will soon be available to the field through the IRIS system.
o OSW is also working to improve the data base on chemical constituents.
It is working with the Office of Toxic Substances to require testina
of approximately 70 chemicals to better characterize the' chemical fate
and health effects associated with these constituent?. This test data
will support the relisting effort and will imorove the certaintv of risk
assessment?.
o OSW's Office of Policy, Planning, and Information has been assigned
responsibility for coordinating risk activities1 within OSW and undertakina
new inititatives to support the. field. OPPI staff has begun to talk
with Regional personnel to identify needs and share information.
These and other support activities respond to th
-------
U.S. Environmental Protection Agencf.
Library, Room 2404 PM-211-A J
401 M Street, S.W. . .»
DC 20460 .;;jf
"
*" *"" *"
------- |