You are here

Home › Forums › W2 Forum › Hydrodynamics

Run-time error

Mon, 02/16/2004 - 3:15pm — bdcequal
have found the answer to![addsig]
Forums: 
Hydrodynamics
Wed, 02/18/2004 - 10:48pm — bdcequal

Hello there, I'm having trouble running the CE-QUAL-W2 model for a low flow river. The following is the persisting error message that I'm getting. ---------------------------------------------------------------------------------------------------------- Computational warning at Julian day = 36.000 at segment 8 timestep = 30.000 water surface deviation [Z] = 1.391 m layer thickness = -1.141 m Negative surface layer thickness in segment 8 time step reduced 90% to 3.000 s on day 36.000 at iteration 0 Computational warning at Julian day = 36.000 timestep = 0.025 sec ------------------------------------------------------------------------------------------------------ I tried decreasing the DLTMAX and DLTF as was suggested in the manual but couldn't solve it. However, I can read the timeseries output (tsr_1.opt) but with the column for flow (Q) all the way down zero. What other adjustments should I consider to get it right? your help is very much appreciated, BD[addsig]
Thu, 02/19/2004 - 6:54am — Tom

This is not an error message. It is a warning letting you know that the current settings for DLTMIN, DLTMAX, and DLTF result in a potential problem, and the code is attempting to recover by reducing the timestep and going back through the water surface elevation solution. This is in the w2.wrn file. You need to be concerned only if the w2.err file states that you end up with a negative surface layer thickness even at the minimum timestep. In this case, the model stops running and you have to either decrease DLTMAX, DLTMIN, DLTF, or a combination of them. Sometimes, the model will not be stable at any combination of the preceeding. In that case, you have to look further into adjusting bathymetry or slope. There is an art to getting the model to run stabily for sloping river sections. We will be trying to improve the ease of applying the model to a sloping river in the future, but there is a point at which you just "have to have a feel" for what needs to be done.[addsig]
Thu, 02/19/2004 - 11:36am — bdcequal

Thanks Tom for your swift answer. Its not clear to me yet why the flow (Q) column is zero in the timeseries file (tsr_1.opt) when I have inflows to the system. Is that may be related to the computational warning? Your help is very much appreciated, BD[addsig]
Thu, 02/19/2004 - 11:46am — Tom

No, the zero Q has nothing to do with the warning. I believe you have to make certain that the cell you are outputting information to the time series file always contains water in it. Try entering a positive value (say 0.1) for KTSR. This ensures that information is output for the cell 0.1m down from the water surface. Let me know if you still get zeros.[addsig]
Sun, 02/22/2004 - 12:41am — bdcequal

Thanks Tom - I believe by "KTSR" you mean "ETSR". I've changed "ETSR" from "zero" to all the way to "two" step-by-step and the flow (Q) is still zero. thanks again, BD[addsig]
Mon, 02/23/2004 - 8:14am — Tom

Send me the files and I will look at it. Don't zip them up - out e-mail program is currently filtering out all zip files.[addsig]