UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
          Office of Air Quality Planning and Standards
          Research Triangle Park, North Carolina 27711

                           OCT 19 1992





MEMORANDUM

SUBJECT:  Clarification of Prevention of Significant
          Deterioration (PSD) Guidance for Modeling Class I
          Area Impacts

FROM:     John S. Seitz, Director
          Office of Air Quality Planning and Standards (MD-10)

TO:       See Addressees

     A question has arisen concerning the appropriate modeling
range for completing the required Class I area analysis under the
prevention of significant deterioration (PSD) permit program. 
Specifically, the issue is whether the Class I area analysis should
be limited to the impacts of sources that are located within 100
kilometers of a Class I area.  As described in detail below, the
Agency's position is that generally a 100 kilometer range is an
acceptable modeling domain.  However, impacts from large sources
located at distances greater than 100 kilometers need to be
considered when such impacts reasonably could affect the outcome of
the Class I analysis.

     In order to implement the PSD program, some States have
adopted a policy which limits long-range transport modeling to a
fixed distance such as 100 kilometers.  Several recently-issued PSD
permits have been challenged in part because of the imposition of
a 100-kilometer modeling limitation which allegedly has resulted in
certain major source impacts being excluded from Class I area air
quality analyses. Petitioners have argued that had such impacts
been considered, the Class I analysis could have had different
results.

     Historically, the Environmental Protection Agency (EPA)
guidance for modeling air quality impacts under the PSD program has
tended to focus more on the requirements for a Class II modeling
analysis.  Such guidance has provided that applicants need not
model beyond the point of significant impact of the source or 50
kilometers (the normal useful range of EPA-approved Gaussian plume
models).  PSD policy guidance has been less specific for Class I
area analyses, although the importance of modeling beyond the
accepted range of a Class II analysis has been stated.



                                2

     In a 1979 EPA memo1, Regional Offices were advised to provide
notice to the Federal Land Manager of any proposed PSD sources that
would locate within 100 kilometers of a Class I area.  The memo
further indicated that certain sources (i. e. , very large sources)
located at distances greater than 100 kilometers could affect air
quality related values and, therefore, should be included in the
notification process on a case-by-case basis.  Other EPA guidance
concerning the evaluation of source impacts on Class I areas, while
nonspecific, generally indicated that the special protection
afforded to Class I areas under the Clean Air Act warranted the
consideration of any "reasonably expected impacts" regardless of
the 50 kilometers limitation applied to Class II area analyses2.

     Trajectory-based long-range transport models are available for
estimating air pollutant concentrations beyond the 50-kilometer
range typically associated with Gaussian plume models used in Class
II area analyses.  Representative model evaluations of long-range
transport models3 have generally shown that, while the models tend
to overestimate the maximum values, most values are within a factor
of two of measured concentrations.  Also, long-range transport
model performance is strongly coupled to the quality of the
meteorological data, and the best performance is obtained when the
application of the model is tailored to the location of the
analysis.



     1March 19, 1979 memorandum from David G. Hawkins, Assistant
Administrator for Air, Noise, and Radiation, to Regional
Administrators.

     2See the following references:  43 FR 26380, June 19, 1978,
p. 26398.

     U.S. Environmental Protection Agency, (1980): "Ambient
Monitoring Guidelines for Prevention of Significant Deterioration
(PSD) , " EPA-450/4-80-012.  Office of Air Quality Planning and
Standards, Research Triangle Park, NC, p. A-7.

     U.S. Environmental Protection Agency, (1986): "Guideline on
Air Quality Models (Revised)." EPA-450/2-78-027R.  Office of Air
Quality Planning and Standards, Research Triangle Park, NC, p. 7-8.

     3See the following references:  U.S. Environmental Protection
Agency, (1986): "Evaluation of Short-Term Long-Range Transport
Models.  Volume I. Analysis Procedures and Results."
EPA-450/4-86-016a.  Office of Air Quality Planning and Standards,
Research Triangle Park, North Carolina.
     U.S. Environmental Protection Agency, (1989): "Evaluation and
Sensitivity Analyses Results of the MESOPUFF II Model with CAPTEX
Measurements." EPA-600/3-89-056.  Office of Research and
Development, Research Triangle Park, North Carolina.

                                3
     These and similar studies suggest there are no technical
constraints on the use of long-range transport air quality models
to evaluate impacts for transport distances in the 100-200
kilometer range.  However, until a sufficient reservoir of
expertise and experience develops, long-range transport models will
require expert assistance in order to provide useful information
for assessing impacts.  Also, for these transport distances, the
selection of sources for PSD modeling involves a judgement of
whether these distant sources are germane to the assessment of
Class I area impacts.  Thus, the selection of sources and the
adaptation of the long-range transport modeling to accommodate
these sources for a specific situation are interconnected and must
be done on a case-by-case basis.

     There are acceptable examples of situations where States have
used long-range transport models on a case-by-case basis to carry
out Class I air quality analyses.  The State of North Dakota has
used MSPUFF (an adaptation of the original MESOPUFF model) for
analyses involving sources located at distances of 70 to 140
kilometers from a Class I area.  The State of Maryland used LONGZ
and VALLEY (Gaussian plume dispersion models) to perform a
screening analysis involving sources up to 130 kilometers from a
Class I area in West Virginia.  Also, the State of Florida recently
used MESOPUFF II to assess the impact of sources up to 110
kilometers from a Class I area.

     In conclusion, the use of long-range transport models is
recommended for PSD Class I area modeling situations so long as a
case-by-case analysis is done.  Routinely, major source emissions
within 100 kilometers of a Class I area should be considered when
assessing pollutant impacts on that Class I area.  However,
circumstances may warrant consideration of other sources (initially
using various screening techniques) which are located more than 100
kilometers from a Class I area if there is reason to believe that
such sources could affect the air quality in the Class I area.  If
a long-range transport model is going to be used for a Class I
area, the source inventory, modeling procedures and long-range
model selected for use should be determined on a case-by-case basis
in consultation with the appropriate EPA Regional Office and
Federal Land Manager.

Addressees:
Director, Air, Pesticides, and Toxics Management Division,
Regions I, IV, and VI
Director, Air and Waste Management Division, Region II 
Director, Air Management Division, Region III 
Director, Air and Radiation Division, Region V
Director, Air and Toxics Division, Region VII, VIII, IX  and  X





                                4

cc:  J. Calcagni (MD-15)
     D. Devoe (ANR-443)
     G. Foote (LE-132A)
     W. Laxton (MD-14)
     T. Kiernan (ANR-443)
     J. Rasnic (EN-341W)
     M. Shapiro (ANR-443)
     New Source Review Contact, Regions I-X
     Regional Modeling Contact, Regions I-X