instability during high flows in a small reservoir

High flow events (>1000 cms) are causing model instability in a small, narrow reservoir (~6 miles long; <8000 ac-ft). I've reduced the maximum timestep and also spread the flows out over a larger time period in order to prevent the model from bombing. Inflows are "distributed evenly from top to bottom" and all daily flows are interpolated to sub-daily values so there are no sudden changes in flows. Any other suggestions to overcome the instabilities?