United States        Prevention, Pesticides,     EPA 749-R-94-001
             Environmental Protection    And Toxic Substances     July 1994
             Agency           (7407)
&EPA      Expansion Strategy For The TRI
             Information Management System
                                         Printed on Recycled Paper

-------
                            Contents
Executive Summary 	  ii
Introduction  	 l
The Current TRI Workload  	 1
TRI Expansion Strategy  	 2
The TRI Information Management System 	 4
Capacity of the TRI Information Processing System 	 8
Information Processing Cost of TRI Expansion  	  12
Conclusions	12
                          List of Tables
Table 1   TRI Expansion Schedule	3
Table 2   Cost of TRI Information Processing
          for Reporting Year 1991	8
                          List of Figures
Figure 1  Required % Electronic Submissions
          to Reach System Capacity  	  11
Figure 2  TRI Data Entry Cost at System Capacity	13
Figure 3  TRI Information Processing Cost
          at System Capacity	14


-------
                      Executive Summary
     In 1986 Congress passed the Emergency Planning and Community
Right-to-Know Act  (EPCRA) to provide citizens with the "right to
know" about toxic  chemicals in their communities.  By July 1 of
each year, facilities must report to EPA their releases of toxic
chemicals to the environment if the facilities are in the
manufacturing sector, have ten or more full-time employees, and
manufacture, process, or import more than certain amounts of
listed chemicals and categories.  As required by EPCRA, EPA makes
this data publicly available in a database called the Toxics
Release Inventory  (TRI).

     Congress and  the public have expressed a desire to expand
TRI reporting to cover additional chemicals and facilities.  EPA
plans to expand TRI reporting requirements in two phases.  In
phase I, EPA is adding approximately 350 chemicals to the list
and considering establishing an alternative reporting threshold
for facilities with small volume releases.  In phase II,  EPA will
require facilities in certain non-manufacturing industries to
submit reports. EPA must ensure that it has the capacity to
process the expanded number of annual forms.

     EPA has created the TRI Information Management System to
process TRI forms.  The system consists of EPA performance goals
and EPA resources.  The performance goals ensure accuracy of
data, timeliness of release of the data, and availability of the
data to the public.  EPA resources consist of the processing
facility, the processing system, EPA staff, and the budget to pay
for facilities, equipment and contractor services.  All
processing is done at the EPCRA Reporting Center (ERC)  in
Ballston, Virginia.

     EPA defines the capacity of the TRI Information Management
System as the number of data elements that can be processed
during a reporting cycle using available EPA resources while
meeting goals for accuracy,  timeliness and availability.   A data
element is a single unit of information on the reporting form
(e.g.,  a facility name, a release number).

     For reporting year 1991,  EPA processed 9.6 million data
elements.  Phase I chemical expansion will add about 3.6 million
data elements. If EPA implements the small volume release
provision, about 1.4 million data elements will be eliminated.
With the small volume release provision, phase I expansion will
increase the number of data elements EPA must process by 2.2
million, or 23 percent over 1991.

     The number of data elements that will be added in Phase II
facility expansion depends on final decisions about which non-
manufacturing industrial categories should be included.

     EPA has determined that the processing capacity of the
current TRI Information System is about 20 million data elements
per reporting cycle.  This is based on an analysis of operations
during the 1991 reporting cycle and implementation of
enhancements such as document imaging and the addition of extra

                              -  ii  -

-------
 Expansion Strategy for the TRI Information Management System


processing workstations and shifts.  Phase I expansion along with
the small volume provision will result in about 11.8 million data
elements per year, well within the processing capacity of the
existing TRI Information Management System.

     The processing capacity of the system is highly dependent on
the number of forms that are electronically submitted. The
processing capacity with no electronic submissions, i.e., all
forms are submitted on paper, is about 12 million data elements.
All forms will have to be electronically submitted to reach the
maximum capacity of about 20 million data elements.

     There is significant potential to increase the number of
electronic submissions.  For reporting year 1991, 12 percent of
all forms were electronically submitted.  This increased to 35
percent for reporting year 1992.  EPA believes that with
voluntary efforts, 60-70 percent of all forms could be
electronically submitted, resulting in a system capacity of 17-18
million data elements per reporting cycle.  Depending on the
number of electronic submissions, EPA could add up to eight
million data elements  (almost 80,000 forms) from phase II
facility expansion using the existing information management
system.

     Because of the importance of electronic submissions to TRI
processing capacity, EPA is pursuing several approaches to
increase electronic submissions.  EPA is encouraging companies
that submit the large numbers of forms to submit them
electronically.  EPA will begin an evaluation and pilot of
electronic data interchange  (EDI) during fiscal year 1994.

     This paper does not address capacity beyond 20 million data
elements per reporting cycle.  Exceeding 20 million data elements
would require substantial revision of the information management
                             - Ill -

-------
                  Expansion Strategy for the
            TRI Information Management System
Introduction

     In October of 1986,  Congress passed the  Superfund Amendments
and Reauthorization Act (SARA).   Title III, Section  313 of  SARA,
also known as the "Emergency Planning and Community  Right-to-Know
Act" (EPCRA), is based on the premise that citizens  have a  "right
to know" about toxic chemicals in their communities.

     EPCRA requires certain businesses to submit  annual reports
on the amounts of toxic chemicals their facilities released into
the environment or are transferred away from  their facilities to
other sites.   EPCRA also mandates that an inventory  containing
this data be  established and made publicly available through a
computer database and other means.   The Toxics  Release Inventory
(TRI) was established by EPA to meet these legislative
requirements.

     The TRI  has been very successful in informing citizens about
toxic chemical releases.   There is a great deal of interest on
the part of Congress and the public in increasing the utility of
the TRI.  EPA is considering various options  for  expansion,  and
Congress is considering new legislation that  will expand EPCRA's
coverage.

     EPA spends considerable resources processing TRI submissions
and making the information publicly available.  Any  expansion of
reporting requirements must be carefully managed  to  ensure  that
accurate data can continue to be made available to the public in
a timely manner.

     EPA must ensure that the TRI Information Management System
has adequate  capacity to process TRI forms.   EPA  defines the
capacity of the TRI Information Management System as the number
of data elements that can be processed during a reporting cycle
using EPA resources while meeting goals for accuracy, timeliness,
and availability.

The Current TRI Workload

     Section 313 of EPCRA mandates facilities to  report if,
during the calendar year,  they meet the following conditions:

     •    Are in the manufacturing sector (SIC  codes 20-39); and

     •    Have ten or more full-time employees; and

     •    Manufacture, import, or process 25,000  pounds or more
          of  a  listed chemicals, or use 10,000 pounds or  more of
          a listed chemical.  There are currently about 350
          chemicals and chemical categories on  the list.
                              - 1 -

-------
 Expansion Strategy for the TRI Information Management System


     By July I of each year, owners or operators of facilities
that meet these criteria must submit reports to EPA for the
previous calendar year.  Reports are submitted on the Toxic
Chemical Release Inventory Reporting Form, also known as Form R.

     For the 1991 reporting year, EPA processed about 93,000
forms  (facilities submit one form per chemical).  Of these, about
12 percent were electronically submitted; the remaining 88
percent were submitted on paper.  The 93,000 forms represent
about 9.6 million data elements.1

     The number of forms per year has remained fairly consistent
since 1987, the first reporting year.  Prior to the 1991
reporting year, EPA entered 64 data elements per form.  In 1991,
with the addition of pollution prevention information required by
the Pollution Prevention Act of 1990, the number of data elements
entered from Form R increased to 103.
TRI Expansion Strategy

     EPA is proposing to expand TRI reporting in two phases.  In
phase I, EPA will focus on adding chemicals to the chemical list.
In phase II, EPA will focus on adding non-manufacturing
industries to the facility list.  EPA has chosen to focus on
chemicals first because screening has already been accomplished
on many candidates and there is a well established process for
chemical list modification.  Facility expansion will occur later
because it requires substantial analysis and justification.

     For phase I, EPA has identified about 350 chemicals that
meet the statutory criteria for addition to the list and are
produced in sufficient quantities to justify annual reporting.
These include ozone depleting chemicals  (HCFCs),  RCRA chemicals,
pesticide active ingredients and about 150 other chemicals drawn
from Right-to-Know More legislation and other Federal statutes
such as the Clean Air Act and Clean Water Act.  Federal
facilities are also being added under phase I.
     1     A data element is a single unit of information reported
on  Form R.  (e.g.,   facility  address,  number of  pounds  of  the
chemical  released  to  the  air)   that  is  entered  into the  TRI
Information  Management  System.    The  number of  data  elements
processed in a  reporting cycle,  is  calculated by multiplying the
average number of data elements entered per form by the number of
forms  received in  the  reporting cycle.   There  were  103  data
elements entered  into  the  system for each  1991  form,  and 93,000
forms submitted (including prior year revisions)  resulting in the
processing of  9.6 million data  elements for  the  1991 reporting
cycle.

                              - 2 -

-------
 Expansion Strategy for the TRI Information Management System
     EPA is considering an alternative threshold for small volume
releases.  Under this provision, EPA would establish a higher
threshold for a category of facilities that have releases below a
certain volume level.  Facilities that fall below this release
volume would not be required to submit a full TRI report.  Only a
shorter certification statement would be required.

     EPA estimates that phase I expansion, which will be
implemented in reporting years 1994 and 1995, will add about
35,000 new forms in each reporting cycle, resulting in an
additional 3.6 million data elements.

     EPA estimates the small volume provision might be
implemented in reporting year 1995 and could replace about 20,000
full reports with shorter certification statements.  This would
eliminate about 1.4 million data elements per year.

     For phase II, which will be implemented after reporting year
1995, EPA will add reporting by non-manufacturing facilities that
contribute significant release volumes. Adding these facilities
will provide a more complete picture of the sources responsible
for chemical releases in a community.  The number of new forms
will be based on final decisions about which non-manufacturing
should be added.
Table 1
TRI Expansion Schedule
Phase
Reporting Year
Chemical Expansion



Facility Expansion


Small Volume Provision

Number of Forms
Full Forms
Cert. Stmts.
Total Forms
Number of Million Data
Elements
Full Forms
Cert. Stmts.
Total Forms
I
1993










93,000

93,000


9.6

9.6
1994
RCRA U
List,
HCFC

Federal
Fac.




100,000

100,000


10.3

• 10.3
1995
RTK More
Lists,
Pesticide
AI



Cert.
Stmts.

108,000
20,000
128,000


11.1
0.7
11.8
II
1996




Non-
Mfgr.
Fac.



To be
Deter-
mined


To be
Deter-
mined
                              - 3 -

-------
 Expansion Strategy for the TRI Information Management System


The TRI Information Management System

     EPA has established a TRI information management system to
meet its responsibilities under EPCRA.  The system consists of
two components:  EPA performance goals, and EPA resources.  The
performance goals ensure that accurate, timely information is
made available to the public.  The resources provide the means to
accomplish the goals.  Both the performance goals and the
resources combine to define the capacity of the TRI Information
Management System to process TRI information.

EPA Performance Goals

     EPA has established three performance goals:

     •    Accuracy,

     •    Timeliness, and

     •    Availability.


     Accuracy

          To be of use to the wide variety of users envisioned in
     the EPCRA legislation, the information in the TRI database
     must be accurate.  The accuracy goal consists of two sub-
     goals.

          The first accuracy sub-goal ensures that the TRI
     database accurately reflects what is reported by facilities.
     EPA has defined this component as achieving near 100 percent
     data entry accuracy for certain key data fields,
     particularly those fields that characterize chemical
     releases, and over 98 percent data entry accuracy for all
     fields.   EPA has consistently achieved this goal.  A 1991
     GAO investigation of data quality on 11 important fields
     found 99.7 percent accuracy for 1988 reporting year data.
     EPA's own audit of about three percent of reporting year
     1990 forms found a data entry accuracy rate of 99.5 percent
     for release value fields.

          EPA uses several processes to maintain this high level
     of data accuracy.  All release numbers are reviewed after
     data entry.  In addition, three percent of all forms are
     subject to verification of data entry of all data elements.
     All submitters receive a printout of their release and
     transfer data after entry into the system and are given an
     opportunity to verify its accuracy.

          The second accuracy sub-goal ensures that facilities
     accurately report their releases.   This is accomplished
     through guidance materials,  a technical support hotline,
     facility inspections and, if necessary, enforcement actions.


                              - 4 -

-------
 Expansion Strategy for the TRI Information Management System
     Identification of non-submitters is a high priority.   EPA
     sends submitters a notice of technical error (NOTE)  when it
     identifies errors on the reporting form.   EPA provides
     technical and compliance support to submitters to correct
     errors.    EPA sends to states reports of  major facility
     releases and gives states an opportunity  to identify errors.
     EPA,  along with states,  conduct a considerable amount of
     outreach and training.   EPA provides a technical hotline to
     ensure that facilities have every opportunity to submit
     accurate information.

     Timeliness

          Data must be available to users within a reasonable
     time  after it is submitted to EPA.  EPA has defined the
     timeliness goal as releasing the data to  the public in an
     on-line database within nine months of the July 1 reporting
     deadline.  Meeting this goal requires that EPA balance the
     need  for accurate data with the public's  desire to have it
     as quickly as possible.

          EPA achieved the nine month goal for the first time
     with  reporting year 1991 processing.  EPA processed
     reporting year 1991 forms between September 1992 and April
     1993.2  Despite  the  shorter  time available and the increased
     number of data elements (from 6.0 million for 1990 to 9.6
     million for 1991), EPA was able to achieve its goal to
     complete data processing by the beginning of April,  1993.
     The public release of the 1991 data occurred during May,
     1993.  EPA is on schedule to release 1992 data in April,
     1994.

     Availability

          The availability performance goal ensures that EPA
     meets, its statutory mandate to make TRI data available to
     the public on-line and through other means.  EPA recognizes
     that  the public consists of many different types of users
     with  varying needs.  The primary mechanism for making TRI
     data  available is on-line through the National Library of
     Medicine  (NLM).  This is an easily accessible database with
     a nominal charge for on-line usage.  EPA  also makes TRI data
     available on-line through RTK Net, a database used mainly by
     environmental and public interest groups.  States and other
     government users can access the TRI data  on EPA's mainframe
     2     Data processing  would normally  have started  in  July.
However,  the 1991 reporting form contained new data elements added
by the Pollution Prevention Act.  Approval of the revised FY 1991
reporting form was  delayed. EPA therefore  exercised enforcement
discretion  and  allowed  companies   to  submit  1991  reports  by
September 1 instead  of the  statutory July 1  deadline.    Data
processing for the 1991 forms therefore started in September.

                              - 5 -

-------
 Expansion Strategy for the TRI Information Management System


     computer.  EPA also makes data available through other means
     such as a written reports, magnetic tape, computer
     diskettes, and CD-ROM.

          EPA has a number of projects underway to assess the
     public's use of the information and to find new ways to make
     it accessible.  EPA is investigating ways to link TRI data
     with other kinds of environmental data to enhance data
     usefulness.

EPA Resources

     EPA resources consist of four components:

     •    Budget,

     •    Staff,

     •    Processing Facility, and

     •    Processing System.

     Budget and Staff

          The EPA budget is the dollars available to pay for
     facilities, equipment, and contractor services.  For
     Reporting Year 1991,  EPA spent about $5.6 million for TRI
     information processing.   This required 16.6 EPA full time
     equivalent (FTE)  staff.

     Processing Facility

          The processing facility is the site at which TRI forms
     are processed and stored.  The facility must be large enough
     to accommodate the contractor-run operation.  This includes
     room to receive and store forms,  and room for data entry
     workstations.  In October, 1992 EPA moved the EPCRA
     Reporting Center (ERC) to a new,  larger facility in
     Ballston, Virginia to accommodate expansion from the PPA and
     to prepare for future expansion.

     Processing System

          The processing system includes procedures and equipment
     to process and quality assure TRI data,  make the data
     available, and archive forms.  It also includes computer
     hardware and software.  The current processing system
     includes a local area network (LAN) at the EPCRA Reporting
     Center (ERC)  and the  EPA mainframe at Research Triangle
     Park,  North Carolina.
                              - 6 -

-------
 Expansion Strategy for the TRI Information Management System


Information Processing

     The information processing budget consists of six components
needed to prepare the data for use.  These are:

•    ERC Fixed Costs:  .Cost of rent for the EPCRA Reporting
     Center, contractor management overhead, EPA program
     management overhead, and form storage.

•    Data Entry:  Cost to receive forms and enter them into the
     computer system.  This cost is highly dependent on the
     number of forms, the number of data elements per form and
     whether the form is submitted on paper or electronically.
     It costs EPA much less to enter an electronic form (i.e.,
     one submitted on diskette) than it does to enter a paper
     form.

•    Data Quality:  Cost to quality assure the data.  This
     includes reviewing forms for accuracy, generating notices of
     noncompliance (NONs) and notices of technical error  (NOTEs),
     data reconciliation, providing submitters with an
     opportunity to review the accuracy of their reports after
     the data is entered, and ensuring that a submitter's forms
     can be linked across multiple reporting years.  This cost is
     also highly dependent on the number of forms and data
     elements per form.  Forms submitted electronically generally
     require less quality assurance effort than paper forms.

•    Magnetic Media:   EPA provides submitters, free of charge, a
     computer program that runs on IBM compatible computers.
     This program allows a company to submit its TRI reports on a
     diskette, i.e.,  electronically.  This budget component
     includes the cost to produce the program, provide user
     support, and provide support to some vendors who
     independently produce their own versions of the reporting
     program.

•    Computer Systems:  Cost to maintain computer systems to
     process and store the data.  Data is originally entered on a
     local area network  (LAN) at the ERC.  Data entry
     verification and some quality assurance is done before the
     data is uploaded to EPA's mainframe.  Additional quality
     assurance is done on the mainframe.  The mainframe contains
     the official version of the TRI known as TRIS.  It is
     accessible by EPA and states.  This category includes the
     cost to maintain the mainframe-and LAN hardware and
     software, and to modify the software to accommodate changes
     in reporting requirements.

•    Data Analysis:  Cost to develop tools to use TRI data,
     analyze data to support EPA needs, and prepare data for use
     by others.
                              - 7 -

-------
 Expansion Strategy  for the TRI  Information Management  System
     Table 2 summarizes the cost of TRI information processing
for Reporting Year 1991.
Table 2
Cost of TRI Information Processing for
Reporting Year 1991
Category
ERG Fixed Costs
Data Entry
Data Quality
Magnetic Media
Computer Systems
Data Analysis
Total
FTEs
3.0
3.6
3.3
1.1
2.8
2.8
16.6
Dollars
(Thousands)
1,650.0
2,159.0
785.0
180.0
540.0
280.0
5,594.0
Capacity of the TRI Information Processing System
Efficiency Evaluations and Improvements

     The TRI Information Management System began operation in
1988 to process.forms for reporting year 1987, the initial
reporting year.  Recognizing the need to make the operation as
efficient as possible, EPA contracted with a management
consulting firm to evaluate EPCRA Reporting Center operations. In
July, 1990,  EPA received the final consultant report.  The report
contained numerous recommendations, the most important of which
were:

     •    Combine two separate contracts (one for systems
          development and another for forms processing) into a
          single .contract to improve management and operational
          efficiency.

     •    Develop and distribute of a software package to allow
          companies to submit TRI data on microcomputer diskette.

     These recommendations,  along with others, were implemented.

     In August, 1992, the same consultant completed a follow-up
study which examined ways to expand the system to handle more
                              - 8 -

-------
  Expansion  Strategy  for  the  TRI  Information Management  System


 forms.  The report concluded that  EPA would have  to fundamentally
 redesign  the information management  system if  it  became necessary
 to  expand TRI beyond 14-15 million data  elements  per reporting
 cycle.

 EPA did not accept the consultant's  conclusions about the  limit
"on  the number of  forms that  could  be processed during a reporting
 cycle.  In  September, 1992 EPA staff conducted a  comprehensive
 review of system  operations  with the goal  of identifying ways to
 increase  system efficiency.  Based  on that  review,  EPA
 implemented the following actions  for processing  reporting year
 1991  forms:

      •    Expansion  of the night shift for data entry allowing
          about 30 percent more  data to  be entered.

      •    Institution of preventive  maintenance versus  reactive
          maintenance of computer  systems  resulting in  no  major
          down  time,  allowing full operation over all shifts.

      •    Expansion  of the LAN by  adding another  file server,
          allowing simultaneous  data entry and data
          reconciliation operations  to be  conducted.

      •    Enhancement of the LAN data entry software to increase
          the efficiency of  data entry operators.

      •    Establishment  of a computer system to track all  forms
          by using barcodes,  resulting in  substantially increased.
          efficiency by  reducing the waiting time for forms
          needed  during  data reconciliation.

      Experience gained while processing  1991 forms,  which  was
 started in  September, 1992 and was completed in April,  1993,   has
 confirmed that  implementation- of the above measures substantially
 improved  reporting center operations.  For reporting year  1991,
 EPA faced a 60  percent increase  in data  (from  6.0 million  for
 1990  to 9.6 million  for  1991)  and  a  two  month  shorter time period
 to  process  the  data.  Despite the  shorter  time available and  the
 increased number  of  data elements, EPA was able to achieve its
 goal  to complete  data processing by  the  beginning of April, 1993.
 About 60  percent  more data was processed in 20 percent  less time
 than  in previous  years.

      Additionally, in the Fall of  1992,  EPA received a
 preliminary report from  the  EPCRA  Reporting Center contractor
 suggesting  that the  LAN  used by  the  ERG  for data  entry  could  be
 expanded  beyond its  current  capacity.  Further analysis and the
 experience  gained while  processing the 1991 reporting year forms
 has proven  that the  LAN  can  be expanded  further while maintaining
 efficient operations.
                               -  9  -

-------
 Expansion Strategy for the TRI Information Management System


Analysis of the Processing Capacity of the Existing TRI
Information Processing System

     Reporting Year 1991 is the base year for the capacity
analysis.  For that year, the ERC operated at full capacity and
processed a total of 9.6 million data elements  (MDE).

     EPA has identified several additional areas where the
capacity of the existing system can be increased.  These are:

     •    General efficiencies which include minor improvements
          such as changes in document handling procedures.

     •    The implementation of a document imaging system,
          scheduled for July, 1994.  This will reduce delays in
          processing due to unavailability of files for data
          entry and quality assurance.

     •    Additional workstations for data entry and
          reconciliation and an expanded shift during peak
          processing periods.

     EPA has analyzed the potential increase in capacity from
these improvements and the impact of electronic submissions on
processing capacity.  Based on this analysis EPA has developed a
model of the processing capacity of the existing TRI information
management system as a function of the percentage of electronic
submissions.  The model estimates that the system capacity ranges
from 12 MDE with no electronic submissions to 20 MDE when all
forms are electronically submitted.

     The capacity of the existing TRI information management
system is highly dependent on the number of forms electronically
submitted.  The capacity of the system increases as the
percentage of electronic forms increases.  Figure 1 shows the
percentage of electronic forms required to achieve various levels
of system capacity.  The maximum capacity of the existing system
with the proposed improvements (20 MDE)  can be achieved if all
forms are electronically submitted.

Potential for Electronic Submissions

     There is significant potential for increasing the number of
TRI forms electronically submitted. The number of electronic
forms increased from 12 percent in Reporting Year 1991 to 35
percent in Reporting Year 1992.  In Reporting Year 1992,  about
1,000 parent companies (out of about 22,000 companies) accounted
for almost 40 percent of the electronic forms.  If these 1,000
companies had submitted all their forms electronically, the
overall percentage of electronic submissions would have increased
from 35 percent: to about 50 percent.

     EPA's strategy on electronic submissions is to make a
special outreach effort to the companies that submitted the most


                              - 10  -

-------
                   Figure 1
  Required  % Electronic Submissions
         To Reach System Capacity
  100
o
'co

E  80-
_Q
D
(/)

•1  60 H
O
•M
O

rr?  4OH
  20-
D
C7

-------
 Expansion Strategy  for the TRI  Information Management  System


paper forms in Reporting Year 1992 to encourage  them  to
electronically submit their Reporting Year 1993  forms.   In
addition, EPA is working with trade associations,  such  as the
Chemical Manufacturers Association, to develop programs to
encourage their members to submit electronically.  EPA  has
already enhanced it's electronic reporting package to make  it
more user-friendly and plans to  expand user support for Reporting
Year 1993.  EPA will begin an evaluation and pilot of electronic
data interchange  (EDI) during fiscal year 1994.

     EPA's believes  that with voluntary efforts, 60-70  percent of
all forms could be electronically submitted.  This would make the
processing capacity  17-18 million data elements per reporting
cycle.  To increase  to 20 million data elements would require
electronic reporting from all submitters.

Information Processing Cost of TRI Expansion

     EPA has developed a model for the six categories of cost for
information processing (ERC fixed cost, data entry, data quality,
magnetic media, computer systems, and data analysis).   The model
identifies fixed costs and incremental costs that vary  with the
number of forms to be processed.

     The major saving from electronic submissions is  in data
entry.  There is some saving in data quality because  forms
electronically submitted tend to require less review  and
correction.

     It costs EPA about $25.00 to enter a paper form  and about
$7.00 to enter an electronic form.  As the number of  electronic
forms increases,  the system capacity increases and the  total data
entry cost decreases.  This is illustrated in Figure  2.

     Figure 3 shows the relationship between information system
capacity and total processing cost.  Although the data  entry cost
decreases as the number of electronic forms increases,  other
processing costs increase because of the larger number  of forms.
Once entered,  there is little difference in the subsequent
processing cost of paper and electronic forms.  However, the
total cost of information processing remains almost constant
because data entry is a large portion of the total processing
cost.

Conclusions

     The processing capacity of the system is highly dependent on
the number of forms that are electronically submitted. The
capacity with no electronic submissions,  i.e., all forms are
                              -  12  -

-------
                           Figure 2
                    TRI Data Entry Cost
                      At System Capacity
U)
I
        0
         12
13    14    15    16    17    18    19
  System Capacity (Million Data Elements)
20

-------
                      Figure 3
        TRI Information Processing Cost
                 At System Capacity
  8
  6-
0)
§5
  3

-------
 Expansion Strategy for the TRI Information Management System


submitted on paper, is 12 million data elements.  To reach the
maximum capacity of about 20 million data elements requires that
all forms be electronically submitted.  EPA believes that with
voluntary efforts, 60-70 percent of forms will be electronically
submitted, making the capacity of the existing TRI information
processing system 16-18 million data elements per reporting
cycle.

     EPA has enough capacity to process the increased number of
forms from phase I chemical expansion.  EPA can also add new
forms (up to about 80,000) from phase II facility expansion using
the existing information management system.

     Because electronic forms cost less than paper forms to enter
in the system, the total information processing cost remains
almost constant as the system capacity increases.  EPA can
process up to about 20 million data elements per reporting cycle
with about the about same budget as required for 12 million data
elements because of the larger number of electronic submissions.

     This paper does not address capacity beyond 20 million data
elements per reporting cycle.  Achieving these capacities would
require substantial expansion or revision of the information
management system.
                              -  15 -

-------