 |
Background of Model
Qualitative model description
WinDS minimizes systemwide costs of meeting electric loads, reserve requirements, and emission constraints by building and operating new generators and transmission in 26 two-year periods from 2000 to 2050. The primary outputs of WinDS are the amount of capacity and generation of each type of prime mover—coal, gas combined cycle, gas combustion turbine, nuclear, wind, etc.—in each year of each 2-year period. Figure 1 shows an example of WinDS capacity estimates for the United States for different generation technologies over the next 50 years.
This section also includes information on the linear program formulation, qualitative details on transmission, and qualitative details on wind intermittency.
 |
| Figure 1. Base Case WinDS Capacity Estimates |
While WinDS includes all major generator types, it was designed primarily to address the market issues of greatest significance to wind—transmission and intermittency. The WinDS model examines these issues primarily by using a much higher level of geographic disaggregation than other models. As Figure 2 represents, WinDS uses 358 different regions in the continental United States. These 358 wind supply regions are then grouped into three levels of large regional groupings—the power control areas (PCAs), North American Electric Reliability Council (NERC) regions, and national interconnect regions. The WinDS regions were selected using the following rules and criteria:
- Build up from counties (so that electric load can be determined for each wind supply/demand region based on county population).
- Do not cross state boundaries (so that state-level policies can be modeled).
- Conform to PCAs as much as possible (to better capture the competition between wind and other generators).
- Separate major windy areas from load centers (so that the distance from a wind resource to a load center can be well approximated).
- Conform to NERC region/subregion boundaries (so that the results are appropriate for use by integrating models that use the NERC regions/subregions).
- Conform to the three major interconnects within the U.S. grid system (to limit capacity and energy transmission exchanges between the interconnects).
 |
| Figure 2. Regions Within WinDS |
Much of the data inputs to WinDS are tied to these regions and derived from a detailed GIS model/database of the wind resource, transmission grid, and existing plant data. The geographic disaggregation of wind resources allows WinDS to calculate transmission distances, as well as the benefits of dispersed wind farms supplying power to a demand region.
As shown in Figure 3, WinDS disaggregates the wind resource into five classes ranging from Class 3 (5.4 meters/second at 10 meters above ground) to Class 7 (>7.0 m/s). WinDS also includes offshore wind resources and distinguishes between shallow and deep offshore wind turbines. Shallow-water turbines are assumed to have lower initial costs than deep offshore turbines, because they employ a solid tower with an ocean bottom pier; while deep-water turbines are assumed to be mounted on floating platforms tethered to the ocean floor.
 |
| Figure 3. Wind Resources in WinDS |
These different classes and types of wind have different costs and performance characteristics. Generally, the higher wind class sites (i.e. Class 7) are the preferred sites. However, Figure 4 shows that, at any given point in time, the wind turbines installed will be at a mix of sites with different wind resource classifications. This occurs because, in selecting the installation sites, WinDS considers not only the resource quality, but also includes factors such as transmission availability, costs, and losses; correlation of the wind output with neighboring sites; environmental exclusions; site slope; and population density.
 |
| Figure 4. Wind Capacity Results by Type and Class |
WinDS is also disaggregated over time, not only with the 26 two-year periods between 2000 and 2050, but also within each year. Each year is divided into four seasons with each day of each season divided into four diurnal time slices. These 16 time slices during each year allow WinDS to capture the intricacies of meeting peak electric loads both with conventional sources and intermittent wind generators.
WinDS models the major conventional electricity generators, including:
- pulverized coal
- integrated gasification combined-cycle coal
- existing unscrubbed coal boilers
- existing scrubbed coal boilers
- natural gas combined cycle
- natural gas combustion turbines
- nuclear
- hydroelectricity
These technologies are characterized in WinDS from one year to the next by their:
- equipment lifetime (years)
- capital cost ($/MW)
- fixed and variable operating costs ($/MWh)
- fuel costs ($/Mbtu)
- heat rate (Mbtu/MWh)
- escalation in operating costs and heat rates with plant aging (%/year)
- construction period (years)
- financing costs (nominal interest rate, loan period, debt fraction, debt-service-coverage ratio)
- tax credits (investment or production)
- minimum turndown ratio (%)
- quick-start capability and cost (%, $/MW)
- operating reserve capability
- planned and unplanned outage rates (%).
WinDS is a national electric capacity expansion model, not a general equilibrium model. To assess the potential of wind energy under any given scenario, the model requires that the scenario be exogenously specified in terms of fuel costs and electric loads for each NERC region over the 50-year time horizon of WinDS.
While the focus of WinDS is on wind energy technologies, the model includes details on other generation technologies. For example, there are four types of coal-fired power plants within WinDS—existing boilers without SO2 scrubbers, existing with scrubbers, new advanced pulverized coal plants, and new integrated-gasification combined cycle plants. These plants can burn either high-sulfur or, for a cost premium, low-sulfur coal. Generation by coal plants is restricted to be base and intermediate load with cost penalties (representing ramping/spinning costs) if power production during peak load periods exceeds production in shoulder-peak hours. New coal plants are assumed to be able to provide more spinning reserve capability than older units. Combined-cycle natural-gas plants are considered to be able to provide some operating/spinning reserve and quick-start capability. Nuclear is considered to be base load. Hydroelectricity is not allowed to increase in capacity, due to resource and environmental limitations. Hydro is also energy-constrained, due to water resource limitations. Hydro is assumed to be able to provide both quick-start capability and operating/spinning reserve.
Retirements of conventional generation can be modeled either through exogenous specification of planned retirements (currently used for nuclear, hydro, and oil/gas steam plants), economic retirements, or as a fraction of remaining capacity each period. All retiring wind turbines are assumed to be refurbished or replaced immediately, because the site is already developed with transmission access—and the cost of wind energy technology is only getting cheaper, while the fuel cost of conventional generation is generally assumed to continue to climb. Similarly, any storage at the wind site is assumed to be replaced immediately upon retirement. On the other hand, grid storage retires automatically when its assumed lifetime has elapsed.
WinDS tracks emissions from both generators and storage technologies of carbon, sulfur dioxide, nitrogen oxides, and mercury. Caps can be imposed at the national level on any of these emissions. Alternatively, a carbon tax can be imposed with a ramp up to the maximum tax level over time.
Supporting Documents
|
|
