United States
              Environmental Protection
              Agency
                                  Air and Energy Engineering
                                  Research Laboratory
                                  Research Triangle Park, NC 27711
              Research and Development
                                  EPA/600/SR-94/012  March 1994
EPA      Project  Summary
              Comparison  of  the  1985  NAPAP
              Emissions Inventory with  the
              1985  EPA Trends  Estimate for
              Industrial  SO2  Sources

              David Zimmerman and Rebecca Battye
  Section 406 of the 1990 Clean Air Act
Amendments (CAAA) requires that the
Administrator of the Environmental Pro-
tection Agency (EPA) transmit a report
to Congress containing a  national  in-
ventory of annual sulfur dioxide (SO2)
emissions from industrial sources, as
well as emission projections for the
next 20 years, not later than 1995. The
requirement stems from the 5.6 million
ton emissions cap on SO2 from indus-
trial sources that is contained in Title
IV of the  CAAA  and  based on esti-
mated 1985 emissions from the National
Acid Precipitation Assessment Program
(NAPAP).  This report  presents analy-
ses  of 1985  industrial SO2 emissions
from two available data sources: the
NAPAP inventory and the EPA Trends
report.  These analyses conclude that
the two data sources estimate compa-
rable emissions in the aggregate, but
estimates  for specific  categories and
for processes within those categories
vary widely.  The Trends  method is
limited to source categories that emit
10,000 tonnes per year of SO2. In gen-
eral, the Trends method overestimates
emissions from these  source catego-
ries due primarily to the  absence of
SO2 control  efficiency assumptions.
Overestimation of emissions  in the
Trends data set is offset by the  inclu-
sion of additional source categories in
the NAPAP inventory with the final ag-
gregate estimates within less than 10%
of each other. (NOTE: Trends method-
ology is being changed for 1993, using
the 1985 NAPAP  emissions inventory
as a base.)
                                                 This Project Summary was developed
                                                by EPA's Air and Energy Engineering
                                                Research Laboratory,  Research Tri-
                                                angle Park, NC, to announce key find-
                                                ings of the research project that is fully
                                                documented in a separate report of the
                                                same title (see Project Report ordering
                                                information at back).

                                                Introduction
                                                 Section 406 of the 1990 Clean Air Act
                                                Amendments  (CAAA)  requires that the
                                                Administrator of the Environmental Pro-
                                                tection Agency (EPA) transmit a report to
                                                Congress containing "a national inventory
                                                of annual sulfur dioxide  emissions from
                                                industrial sources not later than January
                                                1, 1995 for all years for which data are
                                                available, as well as the likely trend  in
                                                SO2 emissions over the following twenty
                                                year period (1995 to 2015)." The CAAA
                                                also establishes an emissions cap of 5.6
                                                million tons  per year from industrial
                                                sources;  this annual cap  is  equivalent  to
                                                the 1985 industrial SO2 estimate from the
                                                National  Acid Precipitation Assessment
                                                Program  (NAPAP) inventory. To provide
                                                the analysis mandated by Congress, the
                                                1985 baseline data must first be  exam-
                                                ined to identify strengths and weaknesses
                                                in the available emission and supporting
                                                data.  The purpose of the  overall study
                                                was to understand the similarity and dif-
                                                ferences  between existing data sets and
                                                determine which data are suitable to serve
                                                as a baseline for the SO2 emission projec-
                                                tions. This paper presents the initial analy-
                                                sis of two major sources of industrial data
                                                currently available: the 1985  NAPAP emis-
                                                sion inventory and the 1985 national emis-

-------
sion estimates,  referred to as the Trends
emission estimates.

Overview of the Data Sets
  The  1985 NAPAP emission inventory
effort supported acid deposition research,
including atmospheric modeling, through
comprehensive, detailed source emission
estimates provided  by  local and  state
agencies.  It is a bottom-up inventory and
a 1985 snapshot.  The SO2 emission data
for significant (>100 tons per year) sources
were systematically quality assured, with
greater effort expended on larger sources.
The inventory included a unique confirma-
tion step, allowing individual  plants emit-
ting at least 2500 tons per year to  review
their emission  estimates prior to finaliza-
tion.
  Prior to 1993, EPA prepared an annual
emissions Trends report representing both
current  and historic emissions (1940  to
present).   Industrial  emission estimates
were derived from national, published ac-
tivity data  and  standard, process-level
emission factors;  historic estimates were
altered based on  the most recent activity
data and emission factors to better repre-
sent the most  current understanding  of
emission processes.   For the industrial
sector,  activity data were obtained  prima-
rily  from the Department of Energy, En-
ergy Information Administration; the De-
partment of the Interior, Bureau of  Mines;
and  the  Department of Commerce, Bu-
reau of Census.  It is  essentially  a top-
down approach  designed to  follow cat-
egory emissions through time and  not a
true inventory.  It presents a consistently
derived  national emission estimate at the
emission category level (e.g., industrial oil
combustion) rather than  the source (e.g.,
boiler) level.
  Due to findings discussed here, as well
as other factors, the Trends methodology
has been revised as of 1993.  References
to Trends in this  paper will no longer be
valid for years 1985 and  beyond, effective
with the 1993 edition of the Trends  report.
The reader  is cautioned that comments
herein on the Trends report are valid only
for editions  prior to 1993.

Methodology
  The 1985 NAPAP and Trends industrial
emission estimates were directly  com-
pared.  The 1985 NAPAP-published emis-
sion estimates as well  as the annual U.S.
point source file were used as the basis
for the analysis. The 1992 report National
Air  Pollutant Emission Estimates,   1900-
1991, and  the  background   procedures
(Procedures Document  for the Develop-
ment of National, Regional and Prelimi-
nary Air Pollutant Emissions Trends Re-
port) formed the basis of the analysis of
the published  Trends estimates.   (Note
that  the Trends estimates are updated
annually; the  Trends procedures them-
selves have also been revised in  1993.
This analysis uses the Trends procedures
used through 1992,  but does not include
the new Trends procedures that will be
reflected in the forthcoming 1993 report.)
This study of industrial SO2 emissions veri-
fied emission estimates and activity (i.e.,
throughput) where possible based on  the
original  source materials.
  The comparison of NAPAP and Trends
analyzed the derivation  of individual in-
dustrial  category estimates.  Such analy-
ses were complicated by several factors:

   NAPAP is comprehensive and  in-
    cludes all reported industrial emission
    categories;  Trends  is limited to cat-
    egories thought to emit at least 10,000
    tonness SO2 (uncontrolled) annually.


   NAPAP is source and plant specific;
    Trends is  national and category spe-
    cific.  No opportunity exists to match
    individual  data  values between  the
    inventories;  in fact,  category defini-
    tions differ between  the two invento-
    ries.
   The  1985  NAPAP  inventory is a
    single-year inventory and is not  up-
    dated; Trends adjusts historical emis-
    sion  estimates  based  on  the most
    current information.

Results of the Comparison
  This research led to a highly detailed
view of the two sets of emission estimates
on a category basis, principally relying on
emission  and activity (throughput) data.
The  authors attempted to reproduce  the
1985 Trends emissions  estimates and
noted any irregularities between the cal-
culated and  published data. The analy-
ses proved complex, especially when dis-
aggregating  data to  create comparable
categories between  NAPAP and Trends
data sets, and raised a number of ques-
tions.
  Table 1 summarizes the differences be-
tween the NAPAP and Trends estimates.
Overall, the two 1985 estimates (NAPAP
and Trends) compare favorably:  NAPAP
estimates 5.6 million tons SO2 (as reflected
in  the CAAA),  and Trends  estimates  6.0
million  tons.  However, Table  1  shows
distinct variability in the category-level es-
timates between the two  data sets.  The
major industrial SO2 emission  producing
categories, industrial combustion, non-fer-
rous metals, iron and steel, petroleum re-
fineries, oil and natural gas production,
pulp and paper, and  cement manufactur-
ing  are discussed in detail in the report.

Summary
  As these two data sets are reviewed
and  compared,  the genesis and purpose
of each methodology must be considered.
The 1985 NAPAP inventory was a com-
prehensive emissions inventory designed
to reflect actual conditions in 1985, with
great emphasis on the gathering and qual-
ity assurance of the emission estimate at
the plant and process level.  It was as-
sembled mainly at the state and local level
by professionals familiar with the individual
sources and with significant review by the
major sources.  The Trends methodology
was designed to provide  a longitudinal
view of the direction and probable magni-
tude of emission changes.  It  was as-
sembled  from standard  emission factors
and  national  activity  data in  most cases
and  was  not intended  to reflect  actual
conditions for individual plants  or source
types at any  given time. A de minimus
level of 10,000 tonnes SO2 (uncontrolled)
was established for inclusion of a source
category, so Trends is not intended to be
comprehensive.   The Trends methodol-
ogy  reviewed and presented here  is cur-
rently in transition; subsequent Trends re-
ports will reflect a fundamental shift in the
estimation approach.
  The two 1985  estimates,  Trends and
NAPAP,  provide  reasonable agreement,
within less than 10%, when viewed  as an
aggregate industrial SO2 estimate.  There
is greater divergence when the data sets
are compared at the category and subcat-
egory levels.  Because the data sets were
developed from  different data sources and
with different  purposes, disagreement at
disaggregated category levels  primarily
reflects  the inherent  differences  in  the
methodologies.
  The overall agreement appears  to be
due to a balance between Trends overes-
timates where subcategories overlap be-
tween Trends and NAPAP and the inclu-
sion of  more categories and more  indi-
vidual emission points in NAPAP.   The
systematic Trends overestimate,  relative
to the NAPAP inventory, is due in  part to
the absence of SO2 control efficiency esti-
mates in the methodologies.  SO2 control
technologies  have been applied to most
of the large  industrial SO2  categories
through the promulgation of New Source
Performance Standards (NSPS), issuance
of operating  permits, and  New  Source

-------
Table 1. Magnitude Differences Between 1985 Trends and NAPAP SO2 Emission Estimates
Source Category *
Coal"
Oilc
Natural Gas d
Wood
Miscellaneous Fuel
Other Fuel Combustion Emissions
Reported through NAPAP
1  Copper
1 Lead and Zinc
2 Lead
1 Aluminum
Other Primary and Secondary
Metals Emissions Reported through NAPAP
Iron and Steel
Iron and Steel Foundries
Oil and Natural Gas Production
Pulp and Paper
Cement
Glass
Lime
Sulfuric Acid
Carbon Black
Petroleum Refineries
Other Industrial Process Emissions
Reported through NAPAP
Total
Trends
(tons)
1,840,000
540,000
0
10,000
80,000


650,000
240,000
30,000
70,000


360,000

160,000
250,000
620,000
30,000
30,000
210,000
10,000
830,000


5,960,000
NAPAP
(tons)
1,721,000
713,000
33,000
42,000
14,000

74,000
655,000
106,000
21,000
58,000

42,000
204,000
16,000
332,00
130,000
291,000
23,000
32,000
217,000
28,000
640,000

220,000
5,612,000
Delta
(tons)
119,000
-173,000
-33,000
-32,000
66,000

-74,000
-5,000
134,000
9,000
12,000

-42,000
156,000
-16,000
-172,000
120,000
329,000
7,000
-2,000
-7,000
-18,000
190,000

-220,000
348,000
Delta
(percent)
6.9
-24.3
-100.0
-76.2
471.4

-100.0
-0.8
126.4
42.9
20.7

-100.0
76.5
-100.0
-51.8
92.3
113.1
30.4
-6.3
-3.2
-64.3
29.7

-100.0
6.2
"  Except where noted, the emissions for a source category represent process level emissions only and do not include emissions from the combustion of fuel.
b  Excludes bitumnous coal and lignite consumed at cement and lime manufacturing facilities.
0  Excludes both distillate and residual oil consumed at cement plants and petroleum refineries and residual oil consumed at iron and steel mills.
d  Excludes natural gas consumed in cement manufacturing, petroleum refining, the iron and steel industry, glass manufacture, and at crude petroleum and natural gas production
  facilities
Review Permits. The strengths and weak-
nesses revealed in  this review are pre-
sented below.

NAPAP Data Set
   The 1985 NAPAP inventory still rep-
    resents the  most comprehensive and
    accurate emissions estimates for 1985
    because of  its rigorous quality assur-
    ance of emissions and bottom-up na-
    ture.  The inventory accounts for indi-
    vidual  source operating characteris-
    tics, controls and emission factors.
   Activity data in the 1985 NAPAP  in-
    ventory were not subject to the same
    standard of quality assurance or com-
    pleteness.   Some  data were unre-
    ported due  to  confidentiality  restric-
    tions,  and  activity data  for small
    sources  (i.e., <100 tons  per year)
    passed only the grossest quality as-
    surance checks.   There  are  known
    reporting problems among  miscella-
    neous fuels and other categories. The
    accuracy and representativeness of
    activity data in the NAPAP  inventory
    are best evaluated source by source;
    category-level summaries are unreli-
    able without adjustments.
    It is still possible to locate question-
    able data values in the 1985 NAPAP
    emission inventory when examined on
    a source by source basis, especially
    for smaller emitters.

Trends Data Set
    Some industrial emission  categories,
    notably iron and steel foundries, are
    missing from the Trends method. As
    such,  the inventory  is  not  compre-
hensive,  although  few  major  gaps
were found under close scrutiny.
As  a  top-down approach,  broad as-
sumptions of emission  factors  and
controls are used across a category.
Frequently,  estimates make  no ad-
justment for controls.  Accommodat-
ing  individual  source operating  char-
acteristics, including emission factors,
is impossible.
The underlying industrial activity data
are largely reliable and  probably far
superior to the corresponding NAPAP
estimates at the category level.   Any
method for 1995 and beyond should
take advantage of these independent
data sources.
Based on the  Trends documentation,
the actual Trends execution contains
minor to moderate  errors  in  calcula-

-------
    tion of activity and emissions.  Trends
    could also benefit from  recently  re-
    vised standard  emission  factors and
    updated sources of activity data.

Conclusion
  The 1985 NAPAP SO2 emission esti-
mates remain  the most reliable emission
estimate  baseline because that was one
of the intentions  of the  inventory effort.
Reliable emission projections  of the 1985
data,  especially to 1995 and beyond, are
difficult because the activity  data in the
NAPAP inventory are incomplete  due,  in
part, to confidentiality  and other  restric-
tions.   Because it  is a  1985 inventory,
emission factors and activity  data do not
reflect current or future operations for these
source categories.
  The Trends methodology  reflected  in
this review is  poorly suited to provide a
baseline estimate for industrial emissions,
fundamentally  due to  its  design and pri-
mary  objectives.  Its strengths lie in the
identification and use  of the underlying
industrial activity  data  which  would pro-
vide a firm foundation for year-to-year pro-
jection of specific industrial activities and
their baseline emissions.
  Results using the new Trends  method
could not be reviewed in the scope of this
research. Due to findings discussed here,
as well as other factors, the Trends meth-
odology  has been  revised  as of 1993.
References to Trends  in this paper will no
longer be valid for years 1985 and be-
yond, effective  with the 1993  edition  of
the Trends  report.  The  reader  is  cau-
tioned that comments herein on the Trends
report are valid only for editions  prior  to
1993.
  Several additional sources of data that
may provide more recent data and fill data
gaps for an analysis of industrial SO2 emis-
sions were identified:  Information Collec-
tion Requests obtained in support of the
Maximum Achievable Control Technology
(MACT) standards developed  under Title
III of the  CAAA, SO2 state implementation
plan (SIP) inventories,  and ozone/carbon
monoxide SIP inventories.  Of these, the
MACT data are the  most promising be-
cause  many of the  significant industrial
SO2 source categories are under MACT
development; the data collected from the
sources have  been targeted  for and will
therefore   be  accessible   from  the
Aerometric Information Retrieval System;
and  major  sources, production data, and
control equipment information can be iden-
tified from the  data collected.
  Finally, the 1985 NAPAP data were ana-
lyzed to determine whether individual
source tracking would be a feasible option
to improve the reliability of the emission
estimates and the projections through time.
Such an option would improve  and up-
date the baseline estimate  and provide
current information on processes and con-
trol equipment. It could also be used with
the industrial activity data available from a
variety of sources to  project  the remain-
der of the inventory.  Figure 1 shows that
relatively few sources (about 130) account
for approximately 50% of the total indus-
trial  SO2 emissions.  To capture 80% of
the  SO2 emissions,  approximately 500
sources would need to be inventoried.
         5,000,000  _
         4,000,000  -
  CO
   
-------
   D. Zimmerman andR. Battye are with EC/R, Inc., University Tower, Suite 404, 3101
     Petty Road, Durham, NC 27707.
   Charles C. Masser is the EPA Project Officer (see below).
   The complete report, entitled "Comparison of the 1985NAPAP Emissions Inventory
     with the 1985 EPA Trends Estimate for Industrial SO2 Sources," (Order No.
     PB94-152220/AS; Cost: $36.50, subject to change) will be available only from:
           National Technical Information Service
           5285 Port Royal Road
           Springfield, VA 22161
           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 Researchlnformation
Cincinnati, OH 45268

Official Business
Penalty for Private Use $300
     BULK RATE
POSTAGE & FEES PAID
         EPA
   PERMIT NO. G-35
EPA/600/SR-94/012

-------