SWMM5 - DW Inflow Anamoly again

This is an actual network.

I ran the attached with zero Courant Factor as well as with 0.75. My hand calcs show inflow to be 118 m3 instead of 147 m3 produced by SWMM5 in either case. Inflow to a node is based on 0.011 LPS ADWF per tenement. My time pattern sums to 1.

I am trying to match the pump starts in a 24 hr period. Actual is 63. Programme yields 89. If I can get the inflow right, maybe there will be better correlation.

Can some one have a quick look at this please?

Regards.
Tilake.
Tilake.Weerasekara@GRIFFITH.NSW.GOV.AU

Tilake,

There are a couple of problems with the data set you are trying to run with SWMM 5 Beta A. First, as posted on the SWMM web page, Beta A has a bug which fails to convert DW inflow from user units to CFS for internal processing. Second, your 24 hourly pattern factors (which vary from about 0.02 to 0.06) should each be multiplied by 24 so that the daily average DW inflow can be recovered during the simulation. (The sum of the factors should equal the number of factors, not 1.0, to properly describe a repeating variation around a mean value). Finally, the input file you supplied used a '/' as the date separator instead of a '-'. This caused the SWMM 5 GUI to skip over your specified start/end dates and use the current date instead. (We might take a more liberal approach to reading dates in the future, allowing both types of separators.)

When I ran your revised data through Beta B (which fixes the units bug) I got 64 pump starts and a total DW inflow of 124.5 m3.

Lewis Rossman
U.S Environmental Protection Agency
rossman.lewis@epa.gov

Hi Lewis

Thanks again for going through my file.

Regarding the date separator, I do not know how it jumps from - to /. I posted this error in the web about two weeks ago. Beta B looks much better. The error is only about 5.5%. Is this order of error for Inflow, acceptable by the way?

In the mean time I changed the Time Pattern (still summing to 1 so I have a handle on the changes) and the Control Rules so that I will get the desired pump starts on an hourly basis. (Input file PSG9-3 txt.txt attached). I applied the same time pattern but summing to 24 (basically multiplied the previous time pattern figures by 24) and ran the model. I got a DW inflow 24 times the previous inflow. (3531 m3 compared to 147). (The Input file PSG9-3-Time Pattern Sum 24.txt is attached).

I examined the inflow reports of the run ( with Time Pattern sum =1) for two nodes - 660 B & 664 A in a Excel sheet (attached). It shows as if the Program attempts converting the Time Pattern so that its sum = 24. Only in my case it sums approximately to 28. (28.36 for node 660 B & 28.31 for the node 664 A). It looks to me the Average Flows are processed in L/h by the programme, and the Time Pattern is therefore multiplied by 24.

The basis may be as follows.

Say Qav = Average Flow in L/h and Qj is the average flow in the j th
hour. (j=1,24) again in L/h.

Then daily flow (LPD) = 24*Qav = sum of Qj (j=1,24) - L per day

Qj = fj * Qav, fj being the Time Pattern Factor for the j th hour.

.'. Daily Flow = 24 * Qav = Sum (Qj) = Sum ( fj*Qav) = Qav * Sum ( f j
=1 , 24.

.'. 24 = Sum ( fj) , j=1,24.

If the above holds, the error in inflow may be in the process that converts Time Pattern to sum to 24.

Regards.

Tilake.
Tilake.Weerasekara@GRIFFITH.NSW.GOV.AU

Lewis

I ran the file in Beta B and got the same volume as you did. How can I produce a Pump Status Report like in Beta A?


Regards.

Tilake.
Tilake.Weerasekara@GRIFFITH.NSW.GOV.AU

Tilake,

We dropped the listing of each and every change in status or setting of pumps and regulators in the SWMM 5 Status Report because for long-term simulations it could add up to thousands of lines of text. We are thinking about making it a user-selectable option in the next beta release.

Lewis Rossman
Environmental Scientist
U.S Environmental Protection Agency
rossman.lewis@epa.gov