TABLE 3

ANALYSIS OF THE CONSERVATIVENESS OF ASSUMPTIONS AND
    PARAMETERS USED IN SOME OF THE CONTAMINANT
                 TRANSPORT MODELS



Most            Least
conservative    conservative



end of ranqe



Parameter or assumption       end of range


One-dimensional Vertical Model

  Unsaturated zone thickness 8m

  Till properties              heterogeneous

  Degree of cap membrane
  failure                      100% failure

  Duration of cap membrane
  integrity                    0 yrs.



Two-dimensional Vertical Model

  Longitudinal dispersivity
  ( A ,)

  Dispersivity ratio
  (  -IL/ o,.T   )



  Amount of cap drainage
  assumed to recharge at
  MWDF boundaries

  Inclusion of dilution
  beyond MWDF boundaries



5m



              Value used
              for results
Value used    shown in
by D'ADD/IT   DEIS



13m           8m             8m

homogeneous   homogeneous    homogeneous


100% failure  1007 failure   100% failure


9,000 yrs.    9,000 yrs.     9,000 yrs.



60m



1,000





0%

Model
ignores



5





100%

Model
includes



60m


50





100%

Model
ignores



60m


5,20,50 & 1,0





100%

Model
ignores



              The DEIS has chosen to parrot all of the non-conservative assumptions
              of D'Appolonia/IT (Table 3) except that on the dispersivity
ratio.
              For that parameter, model results for all the tested values
are shown
              (Figs. 3-6 and 3-7). However, the end result of these assumptions
is
              that the vertical distributions of contaminations shown in
the DEIS
              are almost the best possible case.   Moreover, the DEIS grossly
    t285)     underestimates the rates of delivery of contaminations to the
              compliance boundary and strongly misrepresents the distribution
of
              concentrations of sulfate at any time.

         g.   Finally, the contaminant transport models have not been used
to
              predict the full range of possible contaminant distributions.
              Instead they have been used to show what in my opinion is about
the
              smallest impact possible. They have failed to show the worst
              possible case or anything close to it.   Consequently they
provide a
    _____    totally inadequate base from which to predict possible ground
water
              impacts. The DEIS's proclamation that there will be no significant
              ground water contamination is without foundation. Its failure
to
              address contingencies in the event of ground water contamination
is
              unjustified.

         The contaminant transport models used in the DEIS have not been
shown by
     DNR, by D'Appolonia/IT or by the DEIS to have any ability to correctly
predict
     contaminant movement at Crandon. They apparently have been used in an
attempt
     to capitalize upon policymakers' and regulators' enthrallment with computer
results, and perhaps even to overwhelm the non-technical public with computer
(287'predicted chemical patterns. The ploy has been successful in the former
\.../instance because the results are presented as fact in the DEIS rather
than
     highly uncertain, even erroneous, guesses. The writers of the DEIS have
     indiscriminatly accepted the transport model results and have used them
to
     proclaim no ground water contamination impacts of significance will
occur.
     That point remains absolutely unproven.

         The DEIS makes the point that agreement between the IT and DNR/USGS
models
     indicates the validity of the models. Nothing could be further from
truth.
     First, DNR/USGS have withdrawn their initial modelled results (the ones
quoted
     in the DEIS) because of subsequently discovered errors. Secondly, both
-- DNR/USGS and IT used the same basic framework for modelling (similar
(288)boundaries and assumptions; similar lack of calibration or verification).
The
     correspondence of their results merely shows that two separate models
fed the
     same data will yield similar results. It is proof that two mathematical
     schemes will solve the same equations and get the same result--a verification
     that the mathematical solution techniques are good, not a verification
that
     the models resemble reality.

     C. Other comments on impacts

         1.   The statement made on p. 112 (DEIS) that "In the final
cover remained
               intact, any percolating water would probably not contain sufficient



-. I --..     -_7 -I I #1 11. I- -. -



a