Storm water calculations and continuity - STORMCAD 3

[CPENG78]

Hello everyone,
I came across an interesting scenerio with a co-worker's STORMCAD model that she had worked on. In analyzing the output reports for the model, we noticed that at a few locations (a junction in this case) the sum of the Q's in is slighlty more than the Q going out. This structure is defined as a junction box only, not a weir or a detention structure, just a junction box.

Can someone explain to me, why this is the case? By continuity you would not create or destroy mass so Q in should equal Q out. Any losses due to bend, entry, structure losses would affect the energy and hydraulic grade lines only, but not the Q. Am I thinking about this correctly?

The model is made up of three pipe lines (ranging in size from 6 to 12 inches) joining at the junction box in question and a single pipe line going out. The time of concentration is 6 minutes for all 3 upstream points. So the initial time of concentration of 6 minutes plus the section time for each individual pipe should more or less be about the same at this junction since all three pipe lines are not much different in length. Does StormCAD obtain flow rates based on the time of concentration and disregards continuity? Please help.

 

[psmart]

Conservation of mass means that the outflow VOLUME must equal the sum of all the inflow volumes. But there is no law that says the peaks must add up.

The peak flows will add directly only if they occur at exactly the same time. If the peaks are offset (due to different Tc, travel time, or detention) they will not add directly, and the combined peak may be slightly lower than expected.


Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[jeremyk55]

the intesity applied to the sum of the three areas should correspond to the highest tc of the three areas.

 

[CPENG78]

Psmart - I understnad your point with regards to conservation of mass, makes sense to me.

Now I see what you mean by the offset regardless of what it is. What bothers me is that the pipes in my system are like I said more or less the same length and the initial time of concentration is set to 6 minutes for all three upstream points. Only differences are slight differences in pipe sizes and slopes which for the size of my sistem should not make much of a difference. The major difference is the size of the areas (But like I said Tc initial ends up calculated to be 6 minutes for all)

The difference between the sum of my flow rates in and out of the structure is 0.13 cfs (sum of inflow is 1.24 and with outflow of 1.11) A difference of about 10%. In a toral separate and different model I obtained such a difference in the order of 2 to 3%. While I understand what you explained to me, am I overreacting to such a fluctuation between two different models. I would think that if all variables are about the same some answer in the order of less than 5% would be acceptable. Am I incorrect? Thanks.

 

[RWF7437]

http://www.stormh2o.com/july-august-2006/urban-hydrology-methods.aspx

 

[Drew08]

It seems a quick check would be to make all the "more or less the same" parameters the same, and see what happens. If you still get different inflows, then something in the model is set up wrong.

 

[psmart]

Hydrographs are normally added one point at a time, and there should be a way you can verify this addition. Unless there is storage (detention), the outflow at each time step should be equal to the sum of the inflows.

Also note that reported peak flow may be interpolated. If so, it may differ from the tabulated values and contribute to the discrepancy.

Sorry I can't be more specific, but I don't have as much experience with StormCAD as some other programs!

Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[cvg]

typically, hydrologists round their calculated flows to the nearest 1, 5, 10 or even 100 cfs. To expect that you can accurately determine runoff to the nearest 10th or 100th of a cfs using a computer or by hand is simply wishfull thinking.

 

[psmart]

While the absolute accuracy of these flow rates is certainly debatable, the hydrograph addition is basic math, and should always be correct. Any apparent discrepancies are good reason for further investigation.


Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[CPENG78]

cvg,
Yes I agree, you must round to nearest 1, 5, 10, etc but where you round its all in the order of magintude of the calculations you are carrying out. I would not be concerned about a 10th or a 100th if my order of inputs and outputs were somewhere in 100 cfs or even 10 cfs but I'm talking about about 1 cfs where the difference in error is 10%. Yes, there is a fine line between wishful thinking and design practicality, you are not going to callout a 17.5" trench just because your calculations say so. The contractor will look at it and laugh. Like psmart says, accuracy is debateable but basic math is basic math.

Drew08,
I do have this gut feeling that something may be wrong with the way the model was set up, I will attempt your recommendation and see what happens.

Thank you to everyone for your input, its great insight, more suggestions are welcomed while I try the different reoomendations suggested.

 

[gbam]

Have you checked the travel time in each system? This can potentially offset your peaks enough to lower your outflow once all hydrographs are overlapped. Also, the travel time is based on velocity which will vary for different flow rates, so a larger catchment (larger flow rate) would have a reduction in travel time due to higher velocity, if the pipes were equal. Just another thought.

 

[cvg]

to answer your initial question, check the following values which are set in StormCad which may affect the accuracy of your results:

Discharge convergence test
hydraulic grade convergence test
maximum network traversals
number of flow profile steps
maximum design passes

Your model runs an interative algorithm and depending on how these variables are set, you may get more or less "accurate" results at the expense of more time required to run the model...