Following the Guilherme's request, I would like add a more specific one. We are trying to simulate the different effects of detention tanks on combined sewer overflows. We tried to check the Italian normative criteria through simulations and using a number of tanks lay out. Now we are comparing simulations with synthetic and real hyetographs.

Key parameters are those for the catchment wash off (in SWMM: Arra and Wash), which can significantly vary and for which we failed in finding any table, correlation, test or anything that can help the designer in deciding the correct value to be assigned.

I would be grateful should someone give advices.

Advising input parameters is problematic, absent detailed descriptions of the specific modeling application, design/analysis environment and model objectives. Available SWMM documentation is a useful starting point, even though it probably does not meet the desired handbook standard for water quality simulation. Further than that, I usually recommend that modelers choose their own input parameters and check the resulting computed response against known, appropriate observed data. In N America, APWA (Sartor and Boyd), U of Alabama (Bob Pitt) and NURP have published useful information. Careful sensitivity and error analysis will help establish uncertainty bounds for the unknowable input parameters.

I guess recommending default or broadly appropriate parameters might not be in anyone's professional interests.