Operating Constraints and Other Assumptions

Pacific Northwest hydro project data and operating constraints, reflecting Pacific Northwest Coordination Agreement data submittals, originated from the Bonneville Power Administration (BPA). These data were modified by the River Management Joint Operating Committee (RMJOC) to incorporate climate change projections for natural river flows from 2020 to 2049, for three separate downscaled General Circulation Model (GCM) scenarios.

Also, for each climate change hydro data set, all flow-dependent operating constraints were adjusted to account for the different sets of natural flows. These include flood control, refill and critical rule curves and biological opinion bypass spill requirements. Biological opinion bypass spill requirements reflect the 125-percent Flex Spill Agreement operation (summarized in the table below). 

The number of regulated projects in the RMJOC studies is also different from usual BPA hydro studies. The following projects were removed from the regulation:

Cushman 1, Cushman 2, Alder, LaGrande, Ross, Diablo, Gorge, Upper Baker, Lower Baker, Timothy, Oak Grove, North Fork, Faraday, River Mill, Swift 1, Swift 2, Yale, Merwin, Packwood Lake, Mossyrock, and Mayfield

The aggregate monthly generation for these 21 projects is estimated by summing the median monthly generation over the last 30 historical water years for each project. This fixed monthly generation is added to the fixed monthly hydro independent generation, which is estimated by using the median monthly hydro independent generation over the last 30 historical water years.    

For climate change studies run with the classic GENESYS (which includes BPA’s monthly hydro regulator), monthly hydro independent generation is fixed (i.e., does not change with water year) and includes the estimated generation from the 21 removed projects. However, for climate change studies run with the redeveloped GENESYS, the 21 projects are part of the regulation, which includes climate change natural flows for those projects. 

Because of the literally thousands of monthly operating constraints and complicated operating procedures related to the Columbia River Treaty, the Council chose to use resulting end-of-period reservoir contents from BPA’s monthly hydro model (HYDSIM) as a proxy. End-of-period reservoir contents from the classic GENESYS model (which includes HYDSIM) were interpolated into end-of-week target contents for the redeveloped GENESYS. The hypothesis is that replicating the monthly hydro operation from a model that includes all constraints and operating procedures ensures, to a high degree, that all monthly operating constraints will be satisfied.  

The classic GENESYS does not simulate plant-specific hourly hydro operations, instead it only simulates aggregate (regional) hourly hydro generation. It does this by using sustained-peaking capacity curves, which determine the amount of aggregate hydro capacity that is available over various peak periods of the day, as a function of available monthly energy. The sustained-peaking capacity curves are created by the Council’s Trapezoidal model, a plant-specific hourly optimization model that maximizes peaking capacity for a subset of regulated hydro projects that includes only significant hourly operating constraints at key projects.

The redeveloped GENESYS is a multi-stage optimization model that explicitly simulates project-specific hourly hydro generation and does not require the sustained-peaking capacity approximations. But it does require a complete set of project-specific operating constraints and their associated priorities (applied via penalties in the optimization). Because of the nature of optimization programs, adjusting constraint penalties to yield a reasonable simulation of real-life operations generally requires a great deal of time and effort. The Council has devoted the last couple of years working with stakeholders to fine tune the hourly simulation. A description of key operating constraints and their associated penalties is documented in detail within the model itself^[1]. 

To better represent hourly hydro operations in the redeveloped GENESYS for all branches of the river system, it was also necessary to convert several run-of-river projects in the HYDSIM data to storage projects in the redeveloped GENESYS. A complete list of regulated hydro projects for both the HYDSIM (classic GENESYS) and redeveloped GENESYS models can be found here. Because of this reclassification, the set of regulated projects in the redeveloped GENESYS has 4.8 million acre-feet more useable storage, an increase of over 18 percent (from 26 million acre-feet in HYDSIM to 30.8 million acre-feet in the redeveloped GENESYS). This additional storage should not affect annual average hydro generation, but it does increase hydro flexibility (or the within-day generation range). Information for generating resources, energy efficiency and demand response programs can be found in the supporting material sections “Integrating Generating Resources Data,” “Integrating Energy Efficiency Data,” and “Integrating Demand Response Data,” respectively. Assumptions regarding the availability of in-region and out-of-region market supplies for both the classic and redeveloped GENESYS models is summarized in the table below.

Market Supply Assumptions

Biological Opinion 125% Flex Spill Requirements

[1] Individual penalties and interpretation are available upon request.

[2] All market supply in the new GENESYS is limited by a net imports limit, transmission availability, market fundamentals.

[3] The week and day-ahead market stages replace the purchase ahead functionality in new GENESYS.