February 2, 1994



MEMORANDUM

SUBJECT:  Wind Tunnel Report for Determining Equivalent
          Building Height Determinations for the Cape Industries
          Facility of Wilmington, North Carolina

FROM:     Dean A. Wilson, Meteorologist
          Source Receptor Analysis Branch (MD-14)

TO:       Brenda Johnson, Regional Modeling Contact
          Region IV


     In response to your request, the Model Clearinghouse has completed its
review of your position that equivalent building heights, as determined
from a wind tunnel study, may be used in Industrial Source Complex (ISC2)
modeling of the subject source.  In principle, we agree with your position
that such a wind tunnel study can be accomplished.  However, we are not
convinced that the Company's consultant, CPP, has demonstrated and
documented, within an acceptable level of uncertainty, that their wind
tunnel derived building dimensions are appropriate.

     As you are aware, we asked Dr. William Snyder, Chief of the
Environmental Protection Agency's (EPA's) Fluid Modeling Facility, to
review the technical details of the wind tunnel study performed by CPP. 
Dr. Snyder's comments, as well as our request memorandum to him, are
attached.  It is clear from his comments that he had a very hard time
forming a conclusive opinion on the study.  Furthermore, Dr. Snyder does
not believe that the wind tunnel study should be designed around the
requirements of Good Engineering Practice (GEP).  

     We had a followup discussion with Dr. Snyder in order to better
understand some of his comments.  In that conversation, we mutually agreed
that in principle it is possible to design a wind tunnel study to derive a
set of pseudo building parameters for input to the ISC2 model.  However,
given the results presented, and setting aside the many problems with
errors and inconsistencies, we are not convinced that the CPP study has
enough data points in every case (five points along each downwind
direction) to derive the appropriate pseudo parameters with an acceptable
degree of uncertainty.  

     A more complete analysis and write-up might be convincing that the
wind tunnel study that CPP conducted is acceptable and that the building
dimensions so derived can be used in ISC2.  If the Company chooses to do
this, they should insure that their write-up is "educational" enough that
people such as ourselves who are not intimately familiar with fluid
modeling, as Dr Snyder is, can understand what was done and be convinced
that it is accurate and appropriate.  Information should be provided that
clearly define the criteria employed to insure that the wind tunnel
simulations are appropriate and meaningful.  Are five locations sufficient
to define the variation of concentration values as a function of downwind
distance, including the magnitude and location of the maximum concentration
value?  As an illustration of this point, note that for one set of data, 
Dr. Snyder was able to fit several possible curves to the five data points,
showing a possible significant uncertainty in the location and magnitude of
the maximum concentration.  The uncertainties resulting from the inherent
limitations of such simulations should be explained, so that one can
understand the basis for the criteria employed on defining "acceptance." 
For instance, if repeatability in wind tunnel simulations is typically
within plus and minus 10% (they should provide a basis appropriate to their
experience), this should be reflected in the definition of the criteria
used to determine when the equivalent building dimensions have been
determined.  Having defined the "acceptance criteria," the statistical
results should be clearly available showing that such criteria were met. 
For instance, where is the assessment exploring whether bias and scatter
between the pseudo building results and the actual simulation results vary
as a function of distance downwind?  

     Insuring not only that the various technical aspects are adequately
addressed, but that the summary of the results is clear and complete, is
important not only so that we can be convinced that what was done is
technically supportable but so that we can further communicate the
information to managers and to the general public in such a way that they
can also make an informed judgment on the worth of the proposal.  This
means that the report will need to contain not only the scientific details
but also considerable explanatory material such that these scientific
details can be translated into more commonly understood language.

     In summary, we agree with your position that in principle the fluid
modeling can be used to derive ISC2 input parameters.  However, we are not
confident that the study for Cape Industries has derived parameters that
can be confidently used.

     If you have any questions, please contact me.

Attachments

cc:  J. Irwin
     D. Neeley
     F. Schiermeier
     W. Snyder