Return to Neptune's GoldSim page
This example includes some typical operational constraints, such as the need for a constant pool (water surface) elevation in some reservoirs, the mandatory satisfaction of minimum downstream flow demands (for fish populations, in this case), and the desire to provide extra electrical power at peak times of the day. The model allows the user to experiment with a few operating strategies and evaluate the results in terms of power generation, downstream flows, and pool elevations.
The model incorporates many features present in customized hydroelectric power programs, such as tabulated functions of power output as a function of turbine flow and reservoir volume as a function of pool elevation. Other more sophisticated features have been omitted for the sake of simplicity, but could easily be incorporated into the model. Wedge storage in the reservoir, for example, could be added. This was assumed to be zero in this case since the reservoirs are relatively small.
The basic peaking flow operations incorporate three adjacent reservoirs: Cochrane is upstream, and fills during off-peak hours, saving the bulk of the water for running through its turbines during peak demand. In doing so, the pool elevation of Cochrane can vary by as much as 25 feet in a day. Ryan, the next reservoir down, receives this peaking flow and simply routes it through its turbines, maintaining a level pool. In effect, then, Ryan's generating capacity is synchronized exactly with that of Cochrane, and the two powerhouses run almost as one. Downstream of Ryan is Morony, which fills during peaking operations, receiving the substantial flows from Cochrane and Ryan. Morony, therefore, also experiences extremes in pool elevation. Being the last reservoir in the project, Morony is charged with meeting downstream water flow demands. These demands are generally not satisfied by the storage in Morony alone, so Rainbow Reservoir, above Cochrane, shares its storage capacity, routing flows through Cochrane and Ryan as necessary to meet downstream demand.
All these functions are incoporated into the GoldSim Model, using Reservoir elements and associated tables of volume/elevation and power output as a function of turbine flow.
Return to index
The principal user interface for the model is the Operations Control Panel, shown at the right. Here, the user can select from a limited set of options and can see the results of various operations on power generation and on the satisfaction of downstream flow demands.
Power generation is shown as the power from each dam as well as total output. Overall performance metrics are average power output and total power generated for the duration of the simulation.
The downstream flow is shown, as well as any deficit below the minimum demand. This deficit is integrated to sow the cumulative deficit, and the maximum deficit at any time is also recorded. These reults serve as indicators of project performance from the point of view of downstream demand.
More detailed graphs of interesting model results are also provided.
The graph above shows the degree to which downstream
water demands are met for a representative season.
At the beginning, the upstream flow is
a bit higher than the downstream demand, allowing for effective
peaking operations, and resulting in periodic flows out of Morony
Reservoir. During mid-May (and again in June and December),
flows increase to the point where dams are spilling water that
excees turbine capacity, and the whole operation is essentially
run-of-river. During these times, power generation is full-on.
In early August, there are a couple of times when the upstream
inflow of the Missouri River is less than the downstream demand.
Since these low flow periods are of short duration, the project
can nevertheless satisfy downstream demand with reservoir storage.
In late August and into October, however, the upstream inflows
are less than downstream demand for an extended period, and the
relatively small storage capacity of the system is soon depleted.
In this case, there is no option but to fail to meet downstream
demand.
A graph of power generation through time shows the periodicity of peaking power generation:
The power generation shown here spans only a month, so that we may see the peaking generation in detail. In general, there is a daily periodic fluctuation as the system attempts to respond to power demands during peak hours. In the first week, peak power demands can be met for only part of the desired hours, but as river flow increases during the second week, longer hours of peak power generation are possible. By the third week, there is enough water available to run all powerhouses full tilt round the clock. In the remaining time, a standard peaking pattern is established.
A log of version upgrades:
Here are links to the full model and software needed to run it:
Here are links to the Player version of the model and software needed to run it:
- John Tauxe
Send comments regarding these pages to
John Tauxe
Last modified: 23 October 2007