Time controls

Hi all, Hoping someone on here can help me with this. I am modeling a small reservior in Kansas (drains ~13 mi^2). I have all my .npt files set up and can run the pre-processor with no errors. However, when I run w2_cvf it crashes at day 1. I get an error message saying... Computional warning at day 275.042 at segment 35 timestep = 5.0 water surface elevation [Z] = 1.680 m layer thickness = -0.68 m Negative surface thickness in segment 35 Timestep reduced 90% to 0.500 s on day 275.042 at itteration 0 Computational warning at Julian day 275.042 timestep = 0.090 sec My bathymetry is setup with 36 segments and 17 layers that are 1 meter thick. Inflow is through branch 1. Outflow is over a spillway at segment 35. Any thoughts as to what is causing this? I suspect it has somthing to do with my time controls (DTLMIN, DTLMAX) It also may be my spillway, I derived my coefficents using observed water surface elevation in the reservior and gaged flow just below the spillway and fit the data using the power law method. I set the upper withdrawal layer to 2 but not sure what the lower withdrawal layer should be set to. In addition, flow into the reservoir does go to zero at times during the summer, however at my starting time there is inflow. Thanks, Lee