Abstract

In its effort to meet regulatory requirements outlined in the United States Environmental Protection Agency’s (USEPA) Combined Sewer Overflow (CSO) Control Policy, the City of Philadelphia characterizes its combined sewer system (CSS) using the USEPA’s storm water management model (SWMM). Philadelphia’s CSS is divided into three separate drainage districts each ending at its own wastewater treatment plant. The first hydraulic models of the CSS were developed in 1994, using the U.S. Army Corps of Engineers STORM and Camp Dresser McKee’s NetSTORM. During 1995-1998, the city expanded these models in SWMM4 to include all major trunks, interceptors, and regulating structures. Due to the complexity of these models (~8 000 junctions and conduits) and resulting computational expense (model run times of 14 h to16 h), the models underwent a simplification process (~4 000 junctions and conduits) to achieve model run times on the order of 2 h to 4 h. This simplification process was performed cautiously to avoid losing any of the hydraulic characteristics of the systems. Planning level estimates included in the city’s long term control plan update (LTCPU) submitted in fall 2009 were generated using the SWMM4 simplified models. In late 2009, the city began converting its simplified SWMM4 models to SWMM5. At the time of writing (April 2011), the conversion process is 95% complete.

We note here some of the major advantages of SWMM5 over its predecessors:

- dynamic memory allocation-there are no limits on the number of model elements;

- better solution techniques-Saint-Venant equations (for dynamic waves) are solved by a successive approximation technique that results in faster convergence;

- better simulation of orifices and weirs through use of the classical orifice and weir equations instead of using equivalent pipe approximations;

- a variable time step option for simulations;

- a pipe lengthening option-helps with convergence issues in short pipes; and

- better simulation of force mains.

However, SWMM5 has some disadvantages:

- the engine has bugs (that are being addressed);

- elements like bridges have not yet been included;

- the output format makes post processing more cumbersome; and

- without care, result files can be extremely large and difficult to post-process.

As part of the SWMM4 to SWMM5 model conversion, it was necessary to update the post processing tools originally developed in SAS. SWMM4’s output format made it simple to import the data into SAS or Excel in a single datastep as a space delimited or fixed width format file. Conversely, SWMM5’s output format is stacked (i.e. the time series output for each model element is written below the previous one) and must be read in sequentially to be post processed. For smaller datasets, sequential reads are not too time consuming; but for larger datasets, sequential reads are inefficient.

In this chapter, we describe several of the post processing tools that were created or updated and modified as part of the SWMM4 to SWMM5 model conversion:

- a SWMM5 output reformatting tool;

- a hydrograph comparison tool;

- a percent capture calculation tool; and

- a calibration tool.

This paper is only available in PDF Format:

  View full text PDF