of 100 feet (prototype) in diameter. Fischer and Hanamura (1975) have 

 shovm that the effect of the roughness strips on momentiam exchange is 

 considerably greater than that of the boundary shear in the model. Model 

 test results showed that the roughness strips dominate the velocity dis- 

 tributions in the model, and the dispersion coefficient is thus a function 

 of the roughness strips. They conclude that agreement of transverse mix- 

 ing between model and prototype is possible through a proper combination 

 of strip widths and velocities, but that such agreement should be investi- 

 gated in each case. 



Near-field dispersion of heated discharges is dominated by momentum 

 entrainment in the immediate vicinity of the discharge where inertia of 

 the jet is more important than density differences. Since the three- 

 dimensional turbulence structure of the jet cannot be distorted, near- 

 field heat dispersion cannot be directly reproduced in a distorted-scale 

 model if vertical exchange is important. 



In the far field, heat dispersion is governed by convective spread of 

 the plume over the surface of the receiving waters, mass transport of the 

 plume by ambient currents, diffusion and dispersion due to turbulence in 

 the receiving waters, and surface heat exchange. The steady-state form 

 of the equation governing conservation of heat in an advective, turbulent- 

 flow field is (Stolzenbach, 1971; Zitta and Douglas, 1975) 



u^ + v^+Wg^ = 3^^E^g^y-Hg^^Eyg^j+3^(E,3^; (3-17) 



ax^ ai 

 ay ay 



where 



u = velocity in the x direction 



T = temperature 



V = velocity in the y direction 



w = velocity in the z direction 



E = dispersion coefficient 



By inspectional analysis, it can be determined that 



E„ = 



\ - \{\), - (H'"M, ''-''' 



where it has been previously determined that the appropriate velocity 



scale for a distorted-scale model is v = (L 1 ■^^^. 



r ^ v-'x 



56 



