Infiltration/Inflow in Runoff

Does anyone know of the use of the SWMM RUNOFF Block to simulate I/I in a Sanitary Sewer Sytem? I am looking for any documentation or information that supports this application.

The Runoff Block contains an I/I routine based on a unit hydrograph approach. Although intended for combined sewers, there is no reason it can't work for sanitary as well. Just don't put any surface runoff into the pipe. This routine was contributed by CDM based on their Detroit work and is documented in the RUNOFF.DOC file. Basically, the rainfall data are run through a pre-processor in which a simple loss is calculated, to simulate infiltration into the soil. Then the infiltrated water enters pipes as I/I based on a combination of three triangular unit hydrographs, to simulate quick, intermediate, and slow response. Data are entered in F3, F4 and H5 lines.

Alternatively, use the Transport Block or Extran Block to input constant or varying hydrographs.

Another approach used successfully by a contributor was:

1. Use Runoff Block
2. Input observed rainfall
3. Calibrate percent imperviousness and/or area to
observed hydrographs (typical values should be 2.5-10%
of the original values, i.e., only 2.5-10% of the
separate area, as compared to 100% of a combined area,
is contributing wet weather flow.
4. Run Runoff block with calibrated parameters and
observed or design storms to generate I/I response
from separate areas.
5. Limitation: need calibration rainfall and flow data.

Another contribution:

I have also used the method described above with success. For a fast response of extraneous flows (via direct hydraulic connections, i.e. inflow through cross-connections, etc), Uzair's method effectively treats RUNOFF's %Impervious parameter as %Directly-Connected-Combined- Area. However, for a slow response (i.e. infiltration through ******* or broken pipes/manholes, offset or open joints, etc) I've found it better to input a constant average infiltration rate directly in EXTRAN (or as a line hydrograph - triangular or whatever, depending on the duration and antecedent conditions). Not only does this slow, infiltration-like response depend on the geometry and hydraulic capacity of defects in the sanitary sewer, but it also depends to a large extent on the location of the sewer in relation to the groundwater level. This is a whole new ball game, since the design storm to be used to predict the 5-yr or 10-yr inflow response might not be appropriate for the infiltration response (since groundwater/saturated bedding materials are involved).

To drag this along further, we had also witnessed an in-between, not- slow but not-very-fast-either response (we called it a weeping-tile response) that was best modelled using the groundwater cards of RUNOFF (the H2, H3, and H4 lines, where some of the water that has infiltrated through the surface is allowed to be collected by your sewer system). Beware of instability, although, maybe it's because I was using SWMM4.05 at the time.

And another:

I have modeled several complex combined and separated sewers as well as combinations of these. If it is an existing collection system, it is best to determine dry weather infiltration during the calibration of the model by utilizing collected flow data. However if it is a new sewer system that you are modeling, sanitary flows are best estimated by the development of the drainage area, i.e. population, industry, etc.

Wet weather inflow for an existing system can be accounted for during calibration by adjusting the drainage area of the separated subcatchments to account for that fraction of the total area that contributes wet weather flow. From experience, I have found that for well constructed sewers, approximately 5% of the total drainage area produces wet weather inflow.

The really tricky part is rain induced infiltration. This is the sustained increased flow that often occurs following a rain event due to percolation of rainfall into the soil followed by infiltration into the collection system. Even for existing systems with flow monitoring data I have been unable to develop a consistent approach for simulating this behavior. It occurs in both sanitary and combined sewers. If anyone has a viable technique for simulating this sewer response, I would love to hear it.

And, finally, a word of caution is given for the above contribution:

CAUTION:

..............Wet weather inflow for an existing system can be accounted for during calibration by adjusting the drainage area of the separated subcatchments to account for that fraction of the total area that contributes wet weather flow............

This approach ignores the continuity balance in the total volumes. If you are interested in flooding, treatment volumes, etc. you should not ignore the totals. This is the approach which is universally used in the UK and by the people who use Hydroworks. You to can use this approach but you should also be aware of the limitations associated with it.

I find that not knowing where all the flows are going to or coming from can create unique problems all of its own. One needs to be aware of the total volumes and maintain continuity or find you have just made a big assumption that it is not important. This is particularly true when a study/design is nearing an end and there are just a few little questions that need to be answered.

Been there, done that, don't want to do it again. Keep your eyes open and ask questions.