NTIS PB93-963318
                                   OERR 9375.5-19
                                   April 1939
THE COST OF REMEDIAL ACTIONS
 (CORA) MODEL: OVERVIEW AND
            APPLICATIONS
                     by

                Richard K. Biggs
                   U.S. EPA
               Washington, D.C. 20460

                  Kevin  Klink
                  CH2M fflLL
               CorvaUis, Oregon 97330

                 Jacque Crenca
                  CH2M HILL
               Reston, Virginia 22090
       As submitted for proceedings ofHAZMACON 89
               Santa Clara, California
                  April 1989

-------
      THE COST OF REMEDIAL ACTIONS (CORA) MODEL: OVERVIEW AND APPLICATIONS
        Richard K. Biggs                      Kevin Klink                       Jacque Crenca
           U.S. EPA                         CH2M HILL                        CH2M HILL
     Washington, D.C. 20460              Corvallis, Oregon 97330               Reston, Virginia 22090

ABSTRACT

      The Cost of Remedial Actions (CORA) model estimates site-specific remedial action costs for hazardous
waste sites. The model is microcomputer-based and has two components: an expert system to recommend a range of
remedial technologies, and a cost system. The expert system interacts with the user and develops ranges of recom-
mended remedial action technologies. The cost system contains algorithms capable of developing order-of-magni-
tude cost estimates for 40 demonstrated technologies.

      The CORA model has been used successfully  in a number of different applications. The model was used for
the U.S. EPA for the outyear Super fund remedial action budgeting for FY 1989 and FY 1990, and will be used for
the upcoming FY 1991 budgeting. The model was also used to develop U.S. Navy Installation Program budgets for
FY 1989,1990, and 1991.

BACKGROUND

      Capital and operation and maintenance (O&M) cost estimates are required for Superfund remedial actions at
sites in the U.S. EPA's Superfund Comprehensive Accomplishments Plan (SCAP). These estimates are used to
manage current activities and develop outyear budgets.

       During the early years of the Superfund program (1981-83), little historical data existed for developing cost
estimates for the wide variety of conditions found at Superfund sites. Therefore, the program instead relied on
"average" pricing factors to develop budgets. However, the subjective nature of these pricing factors and the absence
of studies to confirm the factors were considered weaknesses in the program.

       In mid* 1983, the U.S. EPA commissioned a study to attempt to quantitatively define pricing factors for
remedial actions. Because of scant historical construction cost information, a modeling approach for developing
pricing factors was selected. Information was obtained about site conditions at a small sample of Superfund sites and
a set of written decision rules was used to select remedies. The sites in the study were segregated into site types
(e.g., landfills, drum sites, etc.), and the costs of the remedies were estimated using a unit-pricing approach. The
resultant estimated costs were averaged and extrapolated to include the 546 sites on the NPL. The estimates *ere
then aggregated to arrive at  an average cost of construction for an NPL site.

       As the 1985 update of the FY 1986 budget approached, the U.S. EPA sought to develop site-specific budget-
ing. The U.S. EPA attempted to disaggregate  the 1985 results for use on individual sites. Efforts to refine these
estimates pointed out the need for a more accurate pricing technique for individual sites.

       Since budgets are developed 18 months prior to the SCAP operating  year, the U.S. EPA realized it needed a
method to estimate remedial action costs in the prefeasibility stage of analysis. This method was to incorporate

       • A reproducible and consistent method of applying the remedy selection guidance

       • A straightforward method of developing site-specific order-of-magnitude cost estimates

-------
       The CORA model was developed in response to these needs and is now being used to estimate the cost of
outyear Superfund remedial actions for specific sites. Cost estimates are aggregated from the CORA model results
and a variety of other sources, including U.S. EPA Records of Decision (RODs) and feasibility studies (FSs), to
form the U.S. EPA regional and overall outyear budgets.

SUMMARY DESCRIPTION OF THE CORA MODEL

       The CORA model includes two distinct microcomputer-based subsystems. One subsystem is an expert
system for selecting a range of reasonable remedial action technologies from among 40 such technologies in the
system. The other is a cost system with cost modules for all 40 remedial action technologies in the expert system.
The cost system is used to develop order-of-magnitude cost estimates for site remedial scenarios. The cost and
expert subsystems operate independendy of each other.

       The system is not intended to incorporate all of the many technologies that would be necessary to address
every type of site; the goal instead was to address the majority of sites. "Outliers" include sites with radioactive
waste and mining sites. Figure 1 lists the 40 technologies now in the expert system and cost modules. Four auxiliary
cost modules (site preparation, site administration, health and safety, and contingencies and allowances) interface
with these 40 primary technology cost modules to generate scenario-specific site costs for remediation. Each
technology was selected based on its frequency of use for hazardous waste remediation, and the ability to define a
scope range and develop cost estimates for it. Some emerging technologies (such as in situ  vitrification or UV-
ozonation) were not included in the model because of scope and cost uncertainties. However, the CORA framework
allows  for expansions, and other technologies will be considered for addition during annual updates of the model.

EXPERT SYSTEM

       There are two components to an expert system: an "inference engine," which contains general problem-
solving knowledge, and one or more programs ("knowledge bases") that contain the "domain knowledge" (specific
know ledge about a particular problem area).
                  /
       The expert system portion of CORA was developed using the Level 5 Expert System shell version 1.0. This
functions as the inference engine, processing the compiled knowledge bases, making queries to the  user, executing
external programs, and evaluating the rules of the knowledge bases to establish conclusions and recommendations.

       The CORA knowledge bases consist of approximately 670 decision rules for applying 40 fairly well-
developed technologies at Superfund sites. The decision rules reflect both engineering expertise and approaches
drawn  from hazardous waste projects and policy issues. Also included in the decision rules  are questions regarding
interpretation of the language of the Superfund Amendments  and Reauthorization Act (SARA) and  Hazardous and
Solid Waste Amendments (HSWA). The expert system analyzes a site by focusing on separate user defined contami-
nated areas. The user responds to system-selected questions for each waste type within a contaminated area. For j
particular set of user answers corresponding to a certain contaminated area, the expert system recommends a ran go
of potentially implementable and applicable remedial action technologies. These technologies can be combined by
the user to form one or more remedial action alternatives. The user can change his or her answer to  particular
questions to explore a range of outcomes.

COSt SYSTEM

       The CORA cost system was developed using dBASE III Plus software. Ninety separate programs *
-------
cost modules and a system designed to organize the cost estimates by site, operable unit, and alternative scenario.
The following approach was used for developing each cost module:

       •   Key parameter range limits were assumed (e.g., treatment systems for gioundwater extraction were
          limited to 2,000 gpm per unit)

       •   Conceptual designs were developed for each technology

       •   Detailed cost line items were defined within specified range limits for each technology

       •   Microcomputer cost spreadsheets were created for each technology, with relationships allowing individual
          cost line items to vary over defined design ranges

       •   Sensitivity analyses were performed to identify key cost variables

       •   Cost algorithms were developed based on the key variables

       •   Cost modules were developed with the key variables as user inputs and with some default values where
          users may not initially have a site-specific value

       The cost system is organized by site, operable unit, scenario, and technology. The system first asks the user
to either select an existing site that is in the site cost data base, or designate a new site. The user then designates
operable units and scenarios to be considered. Technologies for the scenarios may be based on recommendations
from the CORA expert system or other technology screening and alternative development methods. The user then
runs the cost modules and inputs the required costing parameters. The cost system calculates capital and first-year
O&M cost estimates for each technology selected. Individual technology cost runs are stored under scenarios named
by the user. The user can select combinations of previous cost runs from the site-summary cost scenario menu and
generate ranges of overall site costs for different alternatives.

       Most cost modules provide the user with base-case default values for some parameters. The user may use the
default values, or input known or estimated site-specific information. Example default parameters include:

       •   For a multilayered RCRA cap—thicknesses for seven different cap layers

       •   For a soil bentooite slurry walk—percent bentonite required for slurry, percent slurry loss due to waste
          and seepage
                                       \
       •   For onsite incineration—percentage ash and moisture content, depending on user-selected waste form:
          kiln and afterburner temperature, depending on user information on waste constituents

       •   For air stripping—volatile organic compound (VOQ specific effluent concentrations for discharge to
          surface water

    .   •   For soil vapor extraction (SVE)—default radius of influence for SVE extraction wells, depending on
          user-selected type .of soil

       If the user selects default values, he or she can easily edit the input value and update the cost esumax
site-specific information is available.

       Some of the CORA cost modules contain powerful built-in modeling capabilities, including those tor
             j
       •   Groundwater extraction. If the number of extraction wells is not known, CORA will estimate
          based on the  following factors: aquifer storativity, hydraulic conductivity, aquifer thickness.
          contamination, depth of wells, and desired time for cleanup.

-------
      •  Soil excavation. CORA allows for sequencing of excavation activities at sites where multiple lifts may be
         taken and where there may be loss of productivity due to analytical turnaround times between lifts. The
         excavation productivity modeling algorithms include considerations for anticipated depth of contaminated
         zone, contaminated zone excavation layer thicknesses, area to excavate, and levels of worker health and
         safety protection.

      •  Onsite incineration. CORA runs through more than 70 material and energy balance equations to deter-
         mine case-specific waste feed rates, auxiliary fuel, power requirements, and makeup water for a rotary
         kiln incinerator.

      •  Air stripping. CORA sizes air stripping systems (tower diameter, packing height, blower horsepower, air
         flow rate) based on user inputs for influent flow rate and specific VOC influent and desired effluent
         concentrations.

EXAMPLE OUTPUT

      Figure 2 shows an actual CORA expert system input summary and corresponding output. The summary is
for a Superfund site where a city well field supplying potable water was found to be contaminated with VOCs,
principally chlorinated solvents. The contamination source was found to be a solvent wholesaling company. VOC
contamination in soils at the site was found to be as high as  1,000 pans per million (ppm). Well field groundwater
concentrations were found to be contaminated at. levels of up to 18 ppm.

      Figures 3 and 4 show an actual CORA cosi sysittn site cosi summary based on a U.S. EPA ROD-seleciea
remedy. The proposed site remedy includes:

      •  A 66-acre clay cap

      •  A 9,000-foot-long and 60-foot-deep soil bentonite slurry wall

      •  An active landfill gas collection system and flaring for landfill gas from a 100-acre area

      •  A 120-gpm groundwater extraction system with subsequent air stripping, metals precipitation and sludge
         dewatering, and pressure discharge to the local publicly owned treatment works (POTW)

EXPERIENCE

      Version 1.0 of the CORA model was completed in April 1987, and Version 2.1 in June 1988. The U.S. EPA
contracted with an outside consultant to conduct a validation study of the model The study (Performance Evalu-
ation of CORA Model, ICF, January 1989) included a review of the decision rules and expert system operation and
recommendations. The study also ran the CORA cost system for 12 Superfund sites to compare the results w i in
existing design, bid, or construction costs. Of the 12 sites, 10 of the 12 Version 1.0 CORA estimates and all 1 2
Version 2.1 estimates were within the system design cost range based on comparison with the U.S. EPA design. t»d.
or construction costs. The study concluded that the expert system  "is  a useful tool for EPA budget estimates . " u *: s
sound logic, and develops reasonable recommendations.
     .  In May 1987, the CORA model was used to develop cost estimates for 97 U.S. EPA Superfund sites i
be FY 1989 remedial action candidates. For each site, CH2M HILL team members worked one-on-one wuh I S
EPA regional project managers (RPMs) and ccmplcxd CORA expert system and cost system runs.

       Results for sites in the pre-FS stage were combined with cost information from FSs and RODs to de^i^p j-.
FY 1989 budget A number of analyses have been conducted on the FY 1989 site costs, and the results ha ve nc i pea
the U.S. EPA shape the selection of remedy processes under the Superfund Amendments and Reauthorizauon  A. :
The model was also used in April  1988 to develop costs for the FY 1990 budget, and will be used in Apnl  ;-^  o
develop costs for the FY 1991 budget


As submitted for proceedings of HAZMACON 89
Santa Clara, California, Aoril 1989

-------
       The CORA model was also applied during the summer of 1988 to 661 Navy installation restoration program
sites. As with the U.S. EPA costing exercise, each team member worked one on one with the Naval Facilities
Engineering Command (NAVFAC) engineer responsible for the site. The costs were used by Navy personnel to
estimate Defense Environmental Restoration Act funding for fiscal year 1989,1990, and 1991. In addition, CORA is
being used to develop remedial action strategics and estimate total Department of Defense wide remediation costs.

       The CORA model has also been used for RCRA regulatory support. For the RCRA Location Standards Rule,
the model was used to analyze remediation costs for six site types in differing hydrogeologic, ecological, and
geographic settings to support the regulatory impact analysis. A total of 30 corrective action alternatives were
identified and costed.

       The CORA model has also been used to screen technologies, develop alternatives, and estimate initial
remediation costs for several other sites. To date, more than ISO copies of the model have been distributed to federal
and state agencies, foreign governments, environmental consultants, and industries.

FUTURE APPLICATIONS

       The CORA expert and cost systems were both designed to allow revision and expansion. U.S. EPA funds
have been appropriated for continued maintenance, enhancements, and incorporation of user feedback to reflect
current regulatory policies, uemonstraied technologic,;, ani coat considerations.

       Use of the CORA model is expected to continue to  expand. Future applications include:

       •   Use by EPA regions to develop fiscal outyear Superfund remediation budgets and to perform initial site-
          specific remediation scoping

       •   Use by the U.S. Navy and other federal agencies to estimate outyear and total programmatic remediation
          budgets /

       •   Use to anticipate cost effects for Regulatory Impact Analyses of new environmental regulations

       •   Use by states for total program  and site-specific  remediation budgeting and scoping

       •   Potential use by U.S. EPA, states, industries, and environmental professionals in technology screeniMf.
          scoping, and budgeting of RCRA Corrective Actions and Facility Closures

       •   Use by environmental consultants for pre-FS (and some FS) technology screening, scoping, and budget
          estimating

       The CORA model has proven to be a powerful tool for scoping potential costs of hazardous waste remedia-
tion, even during the initial stages of site investigations. Early awareness of potential site alternatives and i le jnup
costs can help expedite site remediation by focusing site investigations, studies, and designs on site-specific  priori-
ties.
  t tuhmnltmA fnr nmf**Ai* at nf H^7MA.CCiN ftQ

-------
                                          RGURE1
                                  CORA COST MODULES
   CONTAINMENT
101  Soil  Cap
102 Asphalt Cap
103 Multilayered RCRA Cap
105 Soil/Ben tonite Slurry Wall
106 Surface Controls
   REMOVAL
201 Soil Excavation
202 Sediment Excavation/Dredging
203 Pumping Con'tained  Wastes
204 Drum Removal
205 Active Landfill Gas Collection
206 Groundwater Extraction
   AUMLUARY
I
Health and Safety
Site Preparation
Site. Administration
Contingencies and Allowances
                301 Onsite Incineration
                302 Offsite Incineration
                303 Soil Flushing
                304 In-Situ Biaremediation
                305 Soil Vapor Extraction
                306 Flaring
                307 Air Stripping
                308 Vapor Phase Ccrbon
                309 Activated Carbon
                310 Activated Sludge
                311 Metals Precipitation
                312 Ion Exchange
                313 Pressure Filtration
                314 Residential  Activated
                    Carbon Units
                315 Offsite RCRA Treatment
                    and Recycling
                316 Solidification
                                                    DISPOSAL
401 Offsite RCRA Landfill
Onsite RCRA. Landfill
   402 Below Grade
   403 Above Grade
404 Offsite Solid Wests
    Landfill
 405 Discharge to POT*
 406 Discharge to Surface
    Water
 407 Water  Reinjection
 408 Water  Infiltration
501 Transportation
502 Municipal Water SwOO*r
503 Groundwater Mon
-------
                                     Figure 2
    Examples of CORA Expert System Remedial Technology Selection
                                     OBAFT
                                                            GATE:   03/03/39
                                                            TIME:   09:2V:44
                           CORA EXPERT SYSTEM
RUN: EXAMPLE TBEATMENT RUN
RON BY: lit
SITE: HAZMACON EXAMPLE
CONTAMINATED AREA: SOILS AT SOLVENT PLANT « CITY VBLLFIELD  AQUIFER

WASTE TYPE: HOT SPOTS (UNSATURATED MTL AROUND LEAKY. TANKS OR  DRUMS)
INPUT
Raiponce type: Treataenta
Soil description: Medium '«and
Soil contaeinant: Volatile organic coapounje
Situ eonditiona could threaten:  True
Exposed to eroaion: Fall*


RECOMMENDATIONS FOR HOT SPOTS (UNSATURATED KTt AROUND LEAKY TANKS OR DRUMS)
GENERAL
e 50* Sit* *ecef* rentrieeioaa
o 503 Croundwatar •onitorinc

IN SITU SOILS TREATMENTS
o 305 Soil vapor extraction for VOCa
Either
o 306 Plarln*- for VOC*
Or
o 308 Vapor pha«« carbon for VOC«
WASTE TYPE: CONTAMINATED SATURATED SOILS (C800NDUATEB>
 INPUT
SaCuraCod son* doacripeloa: Karit li»««eon«
R««pon*« action: Activ* roitoraeioa
Sourc*  in *aeurat«d ton«i F«laa>
USER RESPONStS FOI ooneaainaewi laturaead
Liquid plMM eontasinaat*: VOCt in water (olucion
Diichari* option*: Oiaohars* to turfao* water
Iap«r**abl*. (tratu* •xiie«: Fat«a> •
OoMaeie water aupeljr it contaminated: True
Peraeaenc ale water cupply: Fait*
Exceed oaneer riik: True

RECOMHWBATIONS FOB CONTAMINATED SATURATED SOILS ( CROUNDVATER )
 CENSHAL
 o  503 Crounduater eonitorIng
 o  502 Muaioipal water aupplf

 CROUNDUATER  SXTRACTION
 o  206 Croundvater extract Ion

 DISCHARGE
 o  406 Oitcharie to aurfae* water

 WATER TREATMENT of water froe extracted croundwater
 p  307 Air  itrippin* for VOC«
 o  Evaluate need for 308 vapor phaie carbon or 306 flarinc for VOCa
 NOTE: Nufl«*ra uMd wNIt •RMomnendaltonr (t.g.. "504 ttt aoctss rMtrtcMons*) convsppnd to CORA
       con system modutt nuinbera (SM Rgur* 1).

-------
                                  Four* 3
        CORA Cost System Summary for EPA-Selected Remedy
                              (Capital Costs)
                               3IATT ••••••••••             OATl:
                                                          TIHI:
                    CAPITAL COST OWLCPWWT


  SITE  NADS: HAMACCN EXAHPU StTT — TOXIC UHOPtU.     illCtON: 10
 OPftASU UNIT:   SNTIRE SITS
 ISTIHATIO STABT: HID FT IW
 ROM IT:                PHONC

 INDIVIDUAL TfCHNOLOCT COSTS
 1C WAI 10: ROD RIHIOT

   SOIL CAP                     .                           *.400.000
   SOtL/lfKTONITX StORRT VALL                               3.100.000
   SUV Ml VATIt OCVIUION^COLUCTIOH                           24.000
   POMP INC CONTAINU VA5TU                                   4SO.OOO
   ACTtVt UNOPILL CAS COUICTIOM                           1.400.000
   CtOUNOVATIt tXTIACTION                                     120.000
   ruime                                                   uo.ooo
   All SniPPlNC                                              44,000
   UTAU PUCIPITATIOH                                     t.fOO.OOO
   OPPSITt ICtA THATWMT 6 UCTCltHC                               0
   TIANSPOnATIOH TO OPTStTt ICtA TUAWWT C UCTCllllC              0
   orrstn iou UMWIU.        -                                  o
   TlAISKITATtOli TO OPTSlTt ICtA LAjnTILL                          0
   OISCHAKS TO POTV                                          ff.OOO
   CtOOWUATlt HONITOtlNe                                      SV.OOO
   stn ACCXSS usntertoMs                                  IM.OOO

               S01TOTAL                        .            U.000,000

 SITC COSTS
 S1TI PUPAIATtOH                                                  0
 SITE AMUMISTIATtON                                         510.000

 CIMIAL COMtTtOM
 STAIT-flF COSTS                                              1*0.000

              coMXTtocrroM SOTOTAL                       it, 000,000


 110 COHTIIICWCIU                                         1.100.000
 scope eonriiieaciu                                       t.foo.ooo

              CONSTtOCTtOR TOTAL                          2*. 000.000

 PCniTTtM AMO LtCAL COSTS                                  VOO.OOO
 sntteu outtNe coiisnocrtoH                               i.too.ooo

              TOTAL SIT! CAPITAL COST                     2«. 000. 000


 nans
"* All  s««l« «r» rxm<«< t»  cw« •l«ntfie««t figure*.

••• Ttl«
    Til* f(«»l c*t(* vill
    e«*4i «!•••.  A* • r«*wtt.  ch« fi«al  yr*J«** a«*«« will *«ry

-------
                                  Figura 4
        CORA Cost System Summary for EPA-Selected Remedy
                  (Operation and Maintenance Costs)
                    "*"	OlAfT ••••••••••             BATt.  OJ/07/J9
                                                          TIM:  |7:)»:47

            OPERATION AND SA1KTINANCS COST OCVCLOPNCHT


  SITS  NAM: HAZHACON IlAflfU SITS — TOXIC UNOrtU.     IICIOH:  10


  OPUAIU QUIT:   tNTIRS  SITS
  ISTtNATSO STAIT:  HID  FT  I»M
  RUN IT:               PHONI HOtUIB:

  INDIVIDUAL TSCHNOLOCT COSTS
  SCIMAIIO: 100 RM10T

   SOIL CA?                                                   1.2.000
   SOIL/ICNTONITS SUIIT VAU                                  31.500
   SUVACI UATSI OIVfUIOM/COUXCTION                            1.400
   PWIHC C3XTAI9ICO VJtfTU                                         0
   ACTIVS unortu CAS COUICTION                             jjo.ooo
   etoumuATS* smAcrioM                                      57.000
   rtAIIHC                      .                               4,400
   All STUPPINC                                              10.000
   NSTAU PUCIPITATION                                       142.200
   orrsin iou TIIATHSMT c ISCTCUNC                           M,:OO
   TIAMSTCtTATtOH TO 0PMITS Kit TUATMSNT C IfCTClIHC           4.700
   orrstn KM LANOTIU,                                       30.000
   TIAMKITATIOH TO CfTSITS KU UNOriU                      If .000
   OlSOUtCS TO POTV                                           71.000
   CMONOUATSt NONiTOIINe                                      42.000
   sin ACCSSJ lunicrioNS                                    77.000

               SOITOTAL                                     1.200.000

 SITS COSTS
 SITS PtmiATIOM                                                  0

 CSMSIAL COMITtOM

                     \

 INKnUMCS AM PtMlT IlISUAt                                  1*0.000

              SUBTOTAL                                     1,400.000

 lOTtttCT COSTS
 AOH1MISTIATIOM                                              210.000
 covrmesKiu                                               210.000
              TOTAL SITS OCR COST                         l.MO.OOO


 Ncrrss


••• All «••€• *r*
•••  Th« <*•« ••«iaata« «h»v« *r« b«*«4 ••  eh« d«t* input «•
            n4 e**t •!«•'>">•• 4«««l«pa4  ftr «*4 •• »ct\i«L  tit*. 4««t«n 
-------