Dispersion Calcs using SCREEN3

[Bill3752]

I am attempting to determine the stack height for a vent required to maintain the concentration of contaminant (ammonia) below the LDHL during a release. The stack will be installed above the roof a building. The key is that maintenance will sometimes be required on a platform installed above roof level.

I am using EPA's Screen3 program, but am confused about the best approach. In one case I set up with a the roof = terrain level (~80') and a receptor (the maintenance man) = 98', then determined the required height of the stack. Is it appropriate to assign the terrain level = roof level, since the flare discharge is above the roof line?

I also ran the calculations assuming that the roof height = ground level of the flare (and no terrain change), and maintaining the same height difference between the receptor and roof. Frankly, I assumed I would see the results as in the first case - but the required stack height in the second case was much lower.

In a third trial, I assumed the roof did not exist (i.e. no terrain). The results were nearly identical to Case 2.

Anyone have experience using SCREEN3? If so, any input would be appreciated.

 

[georgeverghese]

I've run these Gaussian dispersion calcs in my younger days, but with a Company program. From your description, it looks like the roof level is of no significance in this simulation - all you've got is the release point, the true ground level, and the x,y,z coordinates of the receptor (and of course the windspeed and wind stability class). I dont have an explanation for the differing results in Cases 1-3. You've got to have a simulation input which you believe most closely reflects the actual spatial details, from an Audit point of view also.