August 7, 1992


MEMORANDUM

SUBJECT:  Healy Clean Coal Visibility Analysis

FROM:     Jawad S. Touma, Meteorologist
          Techniques Evaluation Section, SRAB (MD-14)

TO:       Robert B. Wilson, Regional Meteorologist
          Region X


     As you requested in your memorandum dated June 10, 1992, I
have reviewed the report titled "Permit Application for Air
Quality Program - Prevention of Significant Deterioration (PSD)
and Air Quality Control - Permit to Operate, Healy Clean Coal
Project," dated April 1992, and have some comments regarding the
visibility analysis sections.

     An analysis of the effect of a new source on the visibility
in a Class I area is outlined in the document Workbook for Plume
Visual Impact Screening and Analysis, EPA-450/4-88-015.  If
screening criteria calculated with the VISCREEN model are
exceeded, a more detailed plume visual impact analysis to
ascertain the magnitude, frequency, location, and timing of plume
visual impacts would be required.  Such detailed analysis is
called level-3 analysis, and is carried out by more sophisticated
plume visibility models such as the PLUVUE-II model.  The
Environmental Protection Agency (EPA) had funded an effort that
led to the development of this model after it was determined that
the predecessor model, PLUVUE or PLUVUE-I, had several technical
shortcomings.  Since the introduction of the PLUVUE-II model,
users have brought several coding errors to EPA's attention.  As
a result, work has been underway to fix these coding errors, as
well as, test for other problems; completion is expected at the
end of FY-92.

     In the Healy Project study, the applicant concluded that the
PLUVUE-I model has a significant advantage in execution speed,
and that the additional technical shortcomings can be fixed and
overcome so that the results simulate as closely as possible
those that would have been obtained if the model had performed as
the authors of the code had intended.  Many errors are reported
in Section 5.2, but it is not entirely clear to me whether all of
them were fixed.  Among the errors identified and fixed, is one
dealing with a numerical overflow in the vertical dispersion
parameter (sigma z).  The fix consisted of substituting the
stability class B sigma z parameter for Class A.  It is not clear
how many hours were influenced by this change, nor what effect it
might have on the impact analysis; however, this change is
arbitrary and there is no technical basis for it. 

     In addition to fixes to the code, the report discusses
modifications to input data in Section 5.3.3.  Among these are
using the model for a portion of the data (using monitored data
for the other portion), excluding the F stability class from
model input, and decreasing the stability class by one unit to
account for the effects of complex terrain.  These changes are
also arbitrary and contrary to EPA modeling guidance.  The
effects on modeled concentrations of some of the other changes
mentioned in this section (e.g., adding 30 minutes to the time,
substituting mean for maximum ozone data or using 75 percent
relative humidity) is unclear. 

     In conclusion, then, the analysis went beyond the stated
objective of fixing PLUVUE-I coding errors to simulate the
PLUVUE-II model to include many technical changes as well.  It is
not clear to me what all the changes are, although some seem to
be very arbitrary.  Nor is it possible to tell how the sum total
of these changes have affected the visibility impact analysis for
this study.  It would have been useful to report the performance
of this model with all of its changes using an independent data
set.

     If I can be of any further assistance to you, please let me
know.

cc: D. Wilson