- 43 - 

 X. THE RELIABILITY OF THE MODEL TESTS 



The final figure of merit for all the observations repre- 

 sented in the model tests can be estimated only very roughly. 

 The model itself is rated at 7 on a scale of 10. The equilibrium 

 distribution of bleed water in the model is estimated at the same 

 figure. Hence the reliability of the bleed water tracer tests 

 must be about 5 on a scale of 10. This figure of merit should be 

 interpreted to mean that the contours of bleed water distribution 

 may be overextended with respect to the actual occurrence in the 

 prototype by a factor of 2 or they may be underextended by a 

 factor of 1/2. The kinematic density scale chosen in connection 

 with simulated bleed water discharge caused the bleed water to 

 sink unnaturally and inhibit the mixing processes that would 

 otherwise dilute the discharge. With the condition of insuffi- 

 cient mixing applying so heavily to the first possibility and to 

 the estimate of the figure of merit itself, underextension of 

 the dilution contours of bleed water distribution is improbable. 



