United State;-     O" c • r So a Waste

Agency       O" ce 3; Waste Programs L—C'cs.
                         Seote-ce- ' 593
Mixed Fund in.
Evaluation Report:
The Potential Costs
of Orphan Shares


           ORPHAN SHARES
              Prepared by:

Office of Solid Waste and Emergency Response
    Office of Waste Programs Enforcement
            September 1993

                    MIXK1) ITNDIM,  I \ VJ I \li«>^ RK!'t)RT:
                          THE POTENTIAI (OS IS i VI
       To obtain the current views of organizations and individuals (including PRPs)
 pertaining to mixed funding, the Agency retained a contractor, Kndispulc.   Kndispute
 noted some problems and perceptions pertaining to mixed funding settlements, such as
 the perceived incompatibility of mixed funding with EPA's "enforcement first" approach.
 Endispute aiso noted that PRPs  believe that the procedures and documentation
 requirements for submitting a  claim against the Fund  (in a  preauthorized agreement)
 are too burdensome.  In summary, the report found that mixed funding could be used
 more frequently at sites with, for example, a significant orphan share.

       Under Initiative 4 of the Superfund Administrative Improvements, Final Report
 (June 23, 1993), the Agency agreed to conduct a two-part evaluation of mixed funding.
 This report is the first part of  the evaluation. The second part of the evaluation  will
 explore options for streamlining the mixed funding decision making process and  options
 for streamlining the application and documentation requirements in a preauthorization
 mixed funding agreement.  In addition, the Agency will be piloting several mixed  funding
 settlement  projects in FY94.


      A.     Collecting Data from Ongoing RD/RA Negotiations

      OWPE developed a regional survey to obtain orphan  share information from on-
 going RD/RA negotiations. We chose this group of sites because it provides the Agency
 with  the most current information on orphan shares.

      OWPE identified 78 sites from the May/93 CERCLIS report where Regions are
 conducting ongoing RD/RA negotiations with PRPs.  Some of the negotiations started
 in FY92, others in FY93. We collected remedy and enforcement data. The remedy cost
 figures identified were Record of Decision (ROD) cost  figures not necessarily site cost
 figures,  as a particular RD/RA negotiation may only address one operable unit at a site.
 OWPE made follow-up telephone calls to the Regions to perform quality assurance. We
 present  the data in this report  in aggregate form to preserve the confidentiality of the
enforcement negotiations and internal deliberations on case strategy.

      B.     Orphan Share

      In this report  we define the site "orphan share"  as the portion of liability
 allocated to PRPs who are  non-viable (i.e., defunct, bankrupt, or insolvent) and the
 portion  of waste at a site which is unallocable (i.e., the volume of hazardous substances
    3 "Fact Finding Regarding Mixed Funding and Special Covenants Under CERCLAT
prepared by Michael D. Young, Endispute, dated March 1993.

 which EPA is unable to allocate to a known generator or transporter).5  In Section IV.
 we present cost implications of funding the total orphan share 'both non-viable and
 unallocable waste shares), as well as separate cost breakdowns for funding non-viable or
 unallocable waste shares.

       In the survey we asked  the Regions to identify the number of PRPs at a site, then
 separated the PRP group identified for each site between owner/operators and
 generator/transporter PRPs.  We then  asked whether any of the PRPs were non-viable.
 When a Region identified a PRP as non-viable, we asked them to assign a percentage
 share of responsibility to that  party.5 We asked Regions to identify whether there were
 unallocable waste volumes at the site.  Where a Region answered yes, we asked for the
 basis of the determination (e.g., Did you use volumetric or anecdotal df»*a?). Where a
 Region identified unallocable wastes, we asked them to assign a percentage to th
       We also found it necessan. to make a pidiminan Decision on allocating
                               t.           *          •*
 responsibility between the owner/operator and generator/transporter PRPs at a site.
 We did this to reduce the possibility of overestimaunu the site orphan share (i.e., a
 100% owner/cperator orphan share is  not the site orphan  share) and ensure consistent
 between Regions for producing  national cost figures.  EPVs "Interim Guidelines for
 Preparing Nonbinding Preliminary Allocations of Responsibility," OSWER Directive
 9839.1 (May 28,  1987)(NBAR guidance) provides factors,  but does not specify
 (numerically) the allocation of responsibility between owner/operators and
 generator/transporters. Actual decisions are made on a siie-by-site basis. For purposes
 of this report, we decided  to allocate responsibility evenly between owner/operators and
 generator/transporters: owner/operators are 50% liable and generator/transporters are
 50% liable. We chose these numbers for two reasons; (1) it ensures no Agency bias as to
 the respective cuIpabilL  between different r*RP groups, and (2) we understand
 anecdotaily that  PRPs  use this  breakdown in performing allocations.  For example, if a
 Region said that non-viable owner/operators were responsible for 50% of the
 owner/operator share, we  converted that 50% to the "site" orphan  share, which would be
 25% (50% of 50%).  Regions identified  sites with both non-viable owner/operators and
 non-viable generator/transporters.  In  those situations the "site" non-viable orphan
 share is a combination of these components.

      We assigned the unallocable waste share  to the generator/transporter portion of
 liability at a site. This is  consistent with our NBAR guidance and reduces the
 possibility that we are overestimating the size of the "site"  orphan share.  There were a
 number of sites where Regions identified unallocable wastes and non-viable PRPs. For
 example, if a Region told OWPE that non-viable generator/transporters were
 responsible for a 50% share of responsibility and 25% of the wastes at the site were
 unallocable, the "site" orphan share would be 37.5% (50% of 50% and  25% of 50%).
When estimating costs, if the combination  of non-viable generator/transporter and
 unallocable waste shares exceeded the  50% of liability allocated to the
generator/transporter category, we then adjusted the amount to the maximum of 50%.
The pie chart below illustrates OWPE's definition of orphan share.
               Liability Apportionment Assumptions
               of Orphan Share Analysis*
               1  50% ol liability is apportioned to owner/operators and 50% to generator/transporters
               2  Any unallocable waste is apportioned to generator/tranporters
 IV.    ANALYSIS 01 COSTS TO THI  1 !.  M)

       OWPE identified 78 sites ssiih undoing KU/H\ iv.•ii-.f.i.uions.  At 52 of 78 sites
 (67*7 of sites- u Region  identified an orphan -hart.

       A.     Annual Mixed Funding Cost EMimotes for Orphan Shares

       To establish the orphan share cost, we multiplied the ROD cost figure by the
 percentage of the orphan share.  We then added up  the costs for each site.  Based on
 this  survey, the annual cost to completely mixed fund orphan shares is approximately
 $420 million  dollars.  The annual cost to completely mixed  fund non-viable shares is
 approximately $270 million dollars. The annual cost to completely mixed fund
 unallocable shares is approximately $150 million dollars. Thus, the annual cost to
 mixed fund orphan shares ranges from $150 to $420 million.

      The Regions were able to  identify percentages of the  orphan share at 35 of the 52
 sites. To develop cost shares for the other  17 sites, OWPE  used the average percentage
 for that component of the orphan share identified (whether it was non-%iacle
owner/operator, non-viable generator/transporter or unallocable waste), then multiplied
 that  number  by the ROD cost estimate.  For example, if the estimated remedy cost was
$30 million, and the Region identified an owner/operator as non-viable but could  not
assign a percentage of liability to that party, we assigned the average non-viable
owner/operator share (17.3%) to estimate costs.7  In this example, the site cost to  fund
the non-viable owner/operator share would be $5.19 million ($30 million X 17.3%).  Set
forth below is the annual cost to completely mixed fund orphan shares, as well  as a cost
breakdown for non-viable PRPs  and unallocable waste.
                   Annual Mixed Funding Cost Estimates for Orphan Shares
                   (Rounded lo me nearest JlO Million)
    7  See Section IV.C. for a summary of the average percentages for non-viable
owner/operators, non-viable generator/transporters and unallocable wastes.


        B.     Frequeno Of the Orphan Share Issue

        There were 52 out of 78 sites \vhere a Region identified at least one component ol
 the orphan share (either non-\inhle owner/operators, non-viable generator/transporters
 or unallocable waste).  There were 36 sites out of 78 where a Region identified an
 owner/operator as non-viable. There were 18 sites  out of 78 where a Region identified a
 generator/transporter as non-viable. There were 22 sites  out of 78 where Regions
 identified unallocable wastes. The  graph  below  shows both  the overall frequency and
 individual frequency where a  Region raised an orphan share issue. [Note:  There were a
 number of sites (19) where Regions identified two or more of the components of the
 orphan share at  the same  site. To  avoid double counting, readers should not add the
 individual components together.]
            Orphan Shares
            Frequency of orphan share issue among the 78 sites surveyed.
of Tola
70 -
60 -
50 -
30 -
20 '
10 -
i SJ

B 67%
• •••':"-';'.'.'.. . . . .-.':•• •

' .

:•• i'life SiB*: i: •'':
.-:->:0:0: '•:•• :

'WY. .«„.... .....

££•: .-•' ' : ' ." '* ' "'"'' °' -: ' ;^:o:->::: '.
liiiii l
Total Sites Non-viable (ton viable Unallocable Waste
wth Orphan Shares Owner/Operatore Generator/Transporters 132 Sites)
(52 S«BS) (36 Sues) 118 Sues)
      C.     Average Size of the Orphan Share

      Where a Region identified the percentage of orphan share, OWPE then totalled
the amounts to derive statistical data.  Of the 52 sites identified with an orphan share,
Regions identified percentages of the orphan shares at 35 sites.  The average size of the
orphan share was 26.7%.  Of the 36 sites where a Region identified an owner/opei aior
as non-viable, they were able to assign a percentage at 20 of those sites.  The average
non-viable owner/operator share was 17.3%.  Of the 18 sites where a Region identified a
generator/transporter as non-viable, they were able to assign a percentage at 13 of
sites, The average non-viable generator/transporter share was 16.5%.  Of the 22 sites

 where Regions identified  unallocable Bastes. ihc\  assigned a percentage at  15 of those
 sites.  The average unallocable waste share was 26.9rc.  The graph below shows the mean
 (average), median and range of the orphan share for the total sites in the survey and b>
 individual component. [Note: There were a number of sites (10) where  Regions
 identified and assigned percentages to two  or more of the components of the orphan
 share.  To avoid double counting, readers should  not add the individual components

           Average Size of Orphan Share (by percentage)
           Total Liability ^ _ _                               '00%
           Average SUB Orcnan Si e (Overall)
               Mean ___^^__^^_
               Median ___^^^____ 25%
               Range  ^___^_____________^^________^_ 05.35%

           Non-viable Owner/Operaiors
               Mean ____— ^^_ >7.3%
               Median ___^_ 12.5%
               Range  ___^^^__^.^_^_^^^ 05-40%

           Non-viable Generator/Transporters
               Mean — — — —  16.5%
               Median _ 5%
               Range         ••                 0.5-50%
           Unallocable Shares
               Mean __^_^___^^^_ 269%
               Median _Mn.««__wa__ 25.0%
               Range  ____^^_— ^_________— 0.15-50%
       PRPs now perform the RD/RA at over 70% of sites, which represents
approximately $1.14 billion dollars of commitments to perform work per year - that is,
money the Agency will not have to spend for cleanup.  To put the PRP RD/RA numbers
in perspective, the Superfund appropriation for FY93 is approximately $1.57 bi'lion
    8 The FY94 Administration Budget for Superfund proposes $1.49 billion for the
Superfund Program, approximately a 9.4% budget reduction.

        In situations \shere I'RPs do not perform the RD/RA ui little less than jn't ol
 sites) the Agency \vill obligate funds to perform the uork.  From KY89-*>2, the Vi»enc>
 averaged $474.0 million dollars per \ear for fund-financed RD/RA obligations.  For
 FY93, \ve estimate obligations of approximate!} $447 0 million dollars for fund-financed
 RD/RA cleanups. The graph  below sets forth these numbers relative to funding of
 orphan  shares.
            Comparison of Orphan Share Funding
            to Other Program Costs
                of Dollars
                5 *-
               • :
               05 •
                     $1 57 B.nion
                                           S474 Million
                    Total Supertund    Average Value      Average
                     Appropriation      of PRP     EPA Obligation to*  Qfpnan Share Cost
                       FY93     RCVRA Commitments  Funa-leaa RCVRA
                                 FV90-92       FY89-92

       This report does not constitute a rulemaking by the Agency and may not be relied
upon to create a specific right or benefit, substantive or  procedural, enforceable at law
or in equity by any person.

       This document does not represent a decision by the Agency to mixed fund orphan
shares, and may not be used for any purpose in negotiations with PRPs.