A^u = area below upper confidence limit equation (Cube equa- 

 tion) 



A 1CU = area below lower confidence limit equation (Cube equa- 

 tion) 



105 . The measure of error e represents the area between the con- 

 fidence limits of a particular curve normalized by the area below the curve. 

 This analysis is, perhaps, oversophisticated since the number of points on 

 which to base the calculations is very limited. The confidence bands are 

 wide, most likely resulting in overestimates in the measure of error. 



106. Average sand- trapping efficiencies and standard deviations for on- 

 bed and off -bed nozzles are presented in Table 10. On-bed sand- trapping 

 efficiencies and corresponding standard deviations were calculated for the 

 DUCK85, SUPERDUCK, and C nozzles as 0.13 ± 0.50, 0.68 ± 0.51, and 1.00 (by 

 definition), respectively. Off -bed sand-trapping efficiencies and standard 

 deviations for each nozzle were assumed to be equivalent to values for mid- 

 flow hydraulic efficiencies in this flow range. On-bed sand- trapping effi- 

 ciencies for the H-S nozzle were estimated from Figure 30 at 1,000 percent. 

 Sand fluxes obtained with a particular nozzle type should be divided by either 

 the near -bottom or midflow sand -trapping efficiency (depending on the nozzle 

 elevation) to obtain an estimate of the true sand flux. 



Table 10 

 Sand -Trapping Efficiencies and Standard Deviations 



Nozzle Type Near-Bottom E s Midflow E s 



DUCK85 0.13+/-0.50 0.92+/-0.024 



SUPERDUCK 0.68+/-0.51 1.02+/-0.013 



CUBE 1.00 (by definition) 1.00+/-0.015 



H-S 10.00 



107. If the basin sampler had been used as the measure of ambient sand 

 transport, the on-bed sand- trapping efficiencies and corresponding standard 

 deviations for the DUCK85, SUPERDUCK, and C nozzles would be 0.19 ± 0.77, 1.02 

 ± 0.83, and 1.50 ± 1.03, respectively. 



77 



