Now that you are an expert OSI developer, use the OSI to help make your release decisions. With the rain in the forecast, the uncontrolled spillway could activate at Reservoir A, Reservoir B is also close to the top of Phase III, and Reservoir C is just below Phase III. Uncontrolled local flows downstream of the projects are forecasted to send a 5,000 cfs pulse of water downstream three to four days from now. This happens to also be the approximate travel time from Reservoir A to J110.

Review HEC-ResSim Plots

Review HEC-ResSim plots for each reservoir and control point.

  1. Right-click on each location and select Plot for reservoirs and Junction-Plot for each control point.

    1. Tip: under View Select Live Display… this will keep your plots up-to-date as you make changes to releases.

      How would you operate the reservoirs based on water on the ground (no future precipitation)? How might the forecasted precipitation impact your release decisions? Which reservoir will you give the highest priority for releases and why?

      If the forecast was full of warm sunny days, you would balance releases from the reservoirs to get the reservoirs back to target (aka multipurpose pool) in the next 10 days if possible. Reservoir A and B share J070 and J080, so space at these control points are shared. Reservoir A has more volume in the flood pool, so it might get a larger percentage of available space at the control points. All three reservoirs share space at J110, so priority goes Reservoir A - Reservoir B - Reservoir C due to size, if they are all at the same percentage occupied of their respective flood pools. If the Phase III target at J070 is 7,000 cfs and you have 1,000 cfs in local flows, that leaves 6,000 cfs to release from your reservoirs. You might allocate 3,600 cfs to Reservoir A and 2,400 cfs to Reservoir B. Reservoir C can run its max Phase II release, because even when you add the local flows and releases from Reservoir A and B there is plenty of room at J110.

      Homes and businesses have been flooded out recently and may be in the process of cleanup. To help manage travel time from Reservoir A we might release a smaller amount from the reservoir and give more priority to reservoirs closer to the control points. The idea being if we see the rain developing in the basin we can reduce releases to low flow and still remain under our target at the control points. For example: 7,000 cfs target minus 5,000 cfs local flows equals 2,000 cfs worth of room from releases from the reservoirs. The 4,000 cfs reduction from Reservoir B starts to reduce flow at J070 in six hours, and the Reservoir A reduction reached J070 in 24 hours. The project office will cut releases if they get over an inch of rain per hour, even if the water management office doesn't call. Releases from Reservoir C will be reduced once there is rain in the gages.

Use the OSI to Adjust Releases

Now you will use the OSI to change the release schedule based on your understanding of the system.

  1. Open the OSI and start overriding the HEC-ResSim calculated releases as needed.

  2. Start with Reservoir C, it is the furthest downstream and the closest to J110.

    If the forecasted rain appears, when will you need to cut back releases? Does it look good?

    You will have to cut back releases on the afternoon of June 4th for Reservoir C. Review the model results by looking at the HEC-ResSim plot for Reservoir C. The model says you can cut back at 0600 UTC (1:00 AM local time). Are you going to be very popular with the project staff if you ask them to do this? Maybe they can cut back releases before they leave for the night. You can play with the timing of releases in the OSI to estimate when you need to cut back.

  3. The HEC-ResSim rules are giving priority to Reservoir B based on the J070 Control Point, but Reservoir A is about to activate the earthen open channel spillway.

  4. In the OSI, from the Reservoirs tab (or Res A tab) scroll down to the forecast start time and enter 2,000 cfs at time 01 JUN 2019 1300. Highlight that cell, hold Shift, and select a time on 07 JUN 2019. The cells from June 1st to June 7th should have turned blue.

  5. Right-click in the blue area, select Fill then Repeat Fill and OK.

  6. Select Compute.

  7. Review the HEC-ResSim Reservoir plots for Reservoirs A and B. Also look at the plot for Control Point J070. You may need to run Compute in the Action tab of the CAVI if the plots don't update.

    How did the Reservoir A pool respond? How did the change in Reservoir A releases change releases from Reservoir B? Looking at J070, if we do get the forecasted precipitation, when will we need to reduce Reservoirs A and B to low flow? Review the reservoir plots for Reservoirs A and B. Also look at the plot for Control Point J070.

    The pool elevation peaks at 1056.35 feet. This is 0.55 feet below the previously calculated crest.

    HEC-ResSim will reduce the releases from Reservoir B to account for the 2,000 cfs release from Reservoir A.

    Reduce releases from Reservoir A by the morning of June 3rd if possible, and reduce releases from Reservoir B the evening of June 3rd (or at the latest midnight on the 3rd).

  8. Now that we have a rough plan for Reservoirs A and C, modify releases from Reservoir B. 

  9. In the OSI, from the Reservoirs tab (or Res B tab) scroll down to our forecast start time and enter 4,000 cfs at time 01 JUN 2019 1300. Highlight the cell, hold Shift, and select a time on 07 JUN 2019. The cells from June 1st to June 7th should have turned blue.

  10. Right-click in the blue area, select Fill then Repeat Fill and OK.

  11. Select Compute.

  12. Review the Reservoir HEC-ResSim plots for Reservoirs A and B. Also look at the plot for Control Point J070.

    How did the Reservoir B pool respond to your release changes? If we get the forecasted rain, how far over our Phase III target of 7,000 cfs will we go at J070? If the travel time from Reservoirs A and B are 24 hours and 6 hours respectively, when would you cut back to low flows?

    Decreasing releases from the reservoir from 6,000 cfs to 4,000 cfs slows the drop in pool elevation.

    If we continued releasing at the current rates and received the forecasted rain, the flows at J070 will go over target by approximately 3,700 cfs.

    Reduce releases from Reservoir A by the morning of June 3rd if possible, and reduce releases from Reservoir B the evening of June 3rd (or at the latest midnight on the 3rd).

  13. What if we enter 1,000,000 cfs as an override in the OSI for Reservoir A? Let's find out.

  14. In the OSI, scroll down to our forecast start time and enter 1,000,000 cfs at time 01Jun2019, 1300. Highlight that cell, hold Shift and select a time on 07 JUN 2019. The cells from June 1st to June 7th should have turned blue. Right click in the blue area, select Fill then Repeat Fill and OK. Select Compute.

  15. Review the HEC-ResSim Plot for Reservoir A as well as the Release Decision Report.

    What was the maximum release from Reservoir A? Did it actually go to 1,000,000 cfs? If not, why was the release constrained?

    The maximum release is 6,165 cfs from Reservoir A.

    No, the release did not increase to 1,000,000 cfs, because the physical constraints of the outlet structure captured in the rating curves restricted releases from the reservoir.

  16. Fill out the following table from your final release plan.


    Reservoir A

    Reservoir B

    Reservoir C

    Peak Reservoir Elevation




    Peak Reservoir Outflow




    Reservoir Elevation at End of Forecast




    Flood control Pool Phase at End of Forecast




Now that you have run both forecast alternatives, you should notice that the release overrides are associated to each forecast alternative (independent of each other). Your release overrides in one HEC-ResSim forecast alternative are not seen by subsequent HEC-ResSim forecast alternatives, even though they are the same HEC-ResSim model alternative. However, this is not true for other physical model elements in HEC-ResSim. If you change any of the model elements, such as capacity curves, rating curves, routing parameters, operating zones, rules, operational rule sets, etc., the changes will carry over to subsequent forecast alternatives that use the same HEC-ResSim model alternative. So if you make changes to model parameters that you want to be saved and seen by other subsequent future forecasts, than a Save to Base needs to be performed. Also, if you make changes and cannot undo them, you should perform a Replace from Base to copy the base model into your active forecast.

Prior Task(s):

Task 1: Watershed Orientation

Task 2: Setting up the OSI

Current Task:

Task 3: Using the OSI to Make Reservoir Release Decisions

Next Task(s):