grain size class considered for the capping model), one for the rate of change of the tip of 

 the cloud expanding on the bottom, and one for the dynamic formation of the sediment cioud 

 as a slice of a half and ellipsoid. The initial conditions for the convective descent phase 

 were estimated from barge dimensions, while conditions for the dynamic collapse are 

 estimated from values at the last step of the connective descent phase. 



Based on the considerations described above the capping model was designed to allow 

 input of the important parameters available to the user. The input screens of the model list 

 the parameters needed by the user to operate the model. These include: 



material properties; 



material volume; 



in situ bulk density; 



radius of operations; 



physical oceanographic parameters. 



The geotechnical properties of the dredged material are usually supplied by the 

 permittee along with the estimated volume of material to be dredged and the approximate 

 size of the individual scow. The in situ bulk density of the sediment to be dredged is 

 required to convert the volume of the material to mass to insure the conservation of mass all 

 the way to the disposal point. The in situ densities can range from 1300-1600 kg/m 3 . The 

 radius of operations can be estimated by the user based on past performance by the dredging 

 contractors, whether a taut-moored buoy is in place and whether the barge and scow are 

 actually stopped at the buoy prior to disposal. The physical oceanographic parameters are 

 very general descriptions of the disposal site location. The ambient water density is assumed 

 to be constant throughout the entire water column. Because the model assumes that no 

 density gradient occurs to prevent the dredged material from reaching the bottom, this value 

 has little effect on the results of the model. An average sigma-t value for Long Island Sound 

 is around 20. The depth of the site is very important to these results, however, because it 

 controls the time required for the material to reach the bottom and, subsequently, this time 

 period affects the size of the material "cloud" upon impact with the bottom. The mean 

 bottom current does not significantly affect the results of the model at the depths normally 

 encountered in New England because the descent time is so short. Any offset of the material 

 during descent due to currents would be small. 



Another consideration for the present model was that the impact of multiple dumps of 

 material should be estimated. In order to accomplish this, a scheme for randomly placing the 

 barge within the user- input radius of operations was developed. The capping model 

 incorporates two mathematical random number generators. The first generates a uniform 

 random variate within the range 0.0 to 1.0. The second random variate generator uses the 

 first and produces a normally distributed variate with zero mean and unit internal clock to 

 initiate the random number generators. Two different algorithms for calculating the position 

 were run to determine which was most representative of actual conditions. The first 

 generated positions whereby any point within the radius of operations is a likely as any other 

 (Figure la), while the second distributed positions uniformly along a radius from the center 

 operation (Figure lb). These algorithms were run for a large number (e.g. 2500) of 



