United States
Environmental Protection
Agency
 Air and Energy Engineering
 Research Laboratory
 Research Triangle Park NC 27711
Research and Development
 EPA/600/S7-85/036   Nov. 1985
Project  Summary
Framework  for  Uncertainty
Analysis of the NAPAP
Emissions  Inventory
Carmen Benkovitz
  The objective of this project  is to
develop a methodologies framework to
assess the uncertainties associated with
the emissions values as presented in the
National Acid Precipitation Assessment
Program (NAPAP) emissions inventory
and to implement a prototype computer
system to  estimate the uncertainties
associated with the base NAPAP emis-
sions values.  Previous projects that
addressed the assignment of quantita-
tive values to the uncertainties in emis-
sions data were reviewed, and a statis-
tical methodology was developed that
allows the computation of more accu-
rate uncertainty values. Auxiliary data
required to carry out the computations
were identified, and the prototype
computer system to implement the
calculations has been designed and is
being implemented.
  This Project Summary was developed
by EPA's Air and Energy Engineering
Research Laboratory, Research Triangle
Park. NC, to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title (see Project Report ordering infor-
mation at back).


Introduction
  The NAPAP emissions inventory was
based on the National Emissions  Data
System (NEDS) currently operated by the
Office of Air Quality Planning and Stand-
ards (OAQPS) of EPA (National Air Data
Branch, 1983). NEDS provided the basic
data from which all other levels of aggre-
gation or disaggregation will be calcu-
lated. The basic  NEDS data are statistical
averaged parameters which  allow the
calculation of yearly emissions of the five
criteria pollutants (particulates, SO,, NO,,
hydrocarbons, and CO) on an individual
source/process basis for point sources
and on a county level for area sources.
Current plans call for the application of
spatial, temporal, and species disaggre-
gation algorithms which will be based on
disaggregation factors (or modifications
thereof) developed  for the Northeast
Corridor Regional  Modeling Project
(NECRMP). Higher levels of aggregation
will be calculated as sums of the NEDS
data.
  Calculation of  the uncertainty of the
emissions values will be based on the
statistical formulas expressing the vari-
ance of a function based on the expected
values and variances of the parameters
used to calculate the function. Application
of these techniques  will start with the
algorithms used to calculate the yearly
emissions values and will be extended to
include currently known algorithms for
spatial and temporal aggregation and
spatial, temporal, and species disaggre-
gation, as applicable.

Theoretical Background
  The statistical formulas to be used are
those expressing the variance of a func-
tion based on the expected values and the
variances of the parameters used to
calculate the function. For N independent
parameters and a function of the form x =
N
I  UK, the variance is given by:
k=1
    N
V(x)=I V(UK)                  (1)
    k=1
                      N
For a function of the formx= n  Uk, the
                      k=1

-------
variance is given by:

     N
V(x)= TT   {[EdJJP
                        N
                      - TT   [E(U,)]*(2)
where E( } represents the expected value
and V( ) represents the variance  of a
population. The sample mean and the
sample variance can be used as estima-
tors of the population expected value and
variance; equations (1) and (2) are exact
for these estimators as they are for the
population moments.

Review of Previous Work
  Several previous projects have address-
ed the problem of assigning, in statistical
terms, quantitative values to the errors in
emissions data. Final  reports from the
following projects were reviewed.
  1.   Weighted Sensitivity Analysis of
      Emissions Data, project conducted
      by IBM Corporation under contract
      to the Office of Air Quality Planning
      and Standards (OAQPS), EPA, July
      1973.
  2.   Source  Inventory and  Emission
      Factor Analysis  (SIEFA)  project
      conducted by PEDCo - Environmen-
      tal Specialists, Inc., under contract
      to OAQPS, September 1974.
  3.   Emissions Inventory for  the SURE
      Region, project conducted by GCA
      Corporation under contract to the
      Electric Power Research Institute
      (EPRI), April 1981.
  4.   Emissions,  Costs and Engineering
      Assessment, Work Group 3B, US-
      Canada  Memorandum  of Intent
      (MOI) on Transboundary Air Pollu-
      tion, June 1982.
  5.   Preliminary Evaluation  of Acidic
      Deposition Assessment Uncertain-
      ties, project conducted by Argonne
      National Laboratory (ANL) under
      contract to the U.S. Department of
      Energy, November 1982 (prelimi-
      nary report).

  All  these projects have based  their
calculations on the statistical formulas
for  error  propagation as derived  for
"small" values of the errors; i.e., the
Taylor series expansion included only the
first derivatives of the function. Some of
the error values used in subsequent
calculations have been as high as 80-
90% of the mean. For errors of this size,
the exact equations derived in this project
represent more accurate solutions.
Implementation of Prototype
Computer System
  The final task of this project addresses
the design  and implementation of the
basic  framework of computer software
needed to calculate uncertainties associ-
ated with  yearly emissions values for
both point and area sources.  The con-
ceptual  design is  independent of the
software system used to support the
NAPAP inventory; the design and imple-
mentation of the software modules will
allow portability between computer sys-
tems  and will be as independent  as
possible of the current NAPAP inventory
software system. Figure 1 is a schematic
of this design.

Conclusion
  Detailed quantification of the uncer-
tainties associated with emissions values
in the scale required by the  NAPAP
researchers  has never been attempted
before.  In this project, methodologies
have been developed to allow quantifica-
tion of these uncertainties, and computer
software has been developed to perform
these calculations.
                                           NAPAP Emissions
                                           Inventory
                                           Data Files
                                           Uncertainty
                                           Data File
                                                                                          Emissions Reports
                                                                                           —emissions values
                                                                                           —uncertainty values
                                        Figure 1.    Implementation based on the current NAPAP emissions inventory system.

-------
     Carmen Benkovitz is with Brookhaven National Laboratory, Upton, NY 11973.
     J. David Mobley is the EPA Project Officer (see below).
     The complete report, entitled "Framework for Uncertainty Analysis of the NAPAP
       Emissions Inventory," (Order No. PB 86-112 570/AS; Cost: $9.95,  subject to
       change) will be available only from:
             National Technical Information Service
             5285 Port Royal Road
             Springfield. VA22161
             Telephone: 703-487-4650
     The EPA Project Officer can be contacted at:
             Air and Energy Engineering Research Laboratory
             U.S. Environmental Protection Agency
             Research Triangle Park, NC 27711
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300

EPA/6
-------