NREL: Wind Deployment System Model (WinDS): Background of Model

Qualitative details on transmission

WinDS considers the availability of capacity on existing transmission lines, the cost of accessing and using those lines, and the cost of building new transmission lines dedicated to wind generation when existing lines are not available. To determine how much wind can access existing transmission lines and the cost of building a line from the wind site to the grid, we use a Geographic Information System (GIS). The GIS is used to develop a four-step supply curve for each class of wind in each wind supply region that presents the amount of wind capacity that can access the grid at each of four different costs (the supply curve is formed of discrete steps, with each step represented by a different variable within the linear program.)

The costs increase with increasing distance from the wind resource to an existing transmission line that has adequate remaining capacity available to accommodate the wind generation. Although the lines are usually carrying generation from other sources, at any given instant, they may or may not have the capacity to transmit additional power from new wind generators. It is practically impossible at the national level to assess the capacity available at any given time on each line in the country. Thus, WinDS requires that the user input the fraction of the capacity of each line that will be available for wind; the default fraction is set at 20% for all lines. This transmission availability constraint severely limits the amount of wind that can be transmitted on existing lines, well below that found in previous studies (Parsons, B.; Wan, Y. H. (1995). U.S. Wind Reserves Accessible to Transmission Lines. IEEE Power Engineering Review. Vol. 15(9), September 1995; pp. 5-6; NREL Report No. 21239) that required only that the wind resource be within 20 miles of an existing transmission line.

In addition to the cost of building a line from the wind site to the grid, WinDS also allows the user to input a cost for the use of the grid. That cost can be based on the distance the power is transmitted or on the number of power control areas that the wind transmission must pass through ("postage stamp").

WinDS also verifies that the existing transmission lines crossing the border of a supply/demand region have enough capacity to carry the wind generation into and out of the region. In addition, all generation (including that from wind and conventional generators) is constrained from flowing between any two PCAs in each time slice by the capacity of lines that connect the two PCAs. WinDS does not account for loop-flows, contingencies, etc. that could further restrict transmission on existing lines.

While new transmission lines dedicated to wind are not constrained by the remaining transmission capacity available, they do have additional cost. For lines built to serve wind sites, the entire cost of constructing and maintaining a new line is attributed to wind. This means that the lines are used only when the wind is blowing at the wind resource site, and their costs must be amortized over that intermittent power. The costs of new transmission lines can vary significantly based on terrain, congestion, labor costs, etc. Currently, WinDS assumes a single cost for new lines expressed in $/MW-mile, which is increased for rough terrain and population congestion. In the future, we anticipate modifying WinDS to vary the new transmission line cost per mile with the length of the transmission line and the amount of wind capacity potentially available within the wind supply region.

New transmission lines dedicated to wind can be built either between wind supply/demand regions as described above or within a region. Dedicated wind transmission lines built within a region are assumed to transport the wind generation directly from the wind site to a load center within the region, bypassing the transmission grid and connecting to the distribution system within the load center. As with the construction of lines connecting wind to the grid described above, the GIS is used to develop supply curves for each wind class in each wind supply region for the cost of building these intraregional transmission lines directly to load centers.

New transmission lines are also built in WinDS to transmit power from one PCA to another. These lines can be accessed by either conventional systems or wind power. WinDS builds these lines when it is cost-effective and there is a need for more transmission capacity between the PCAs in one or more of the 16 time slices in each year; or when needed to ensure capacity reserve margins are met in the different PCA, NERC, or interconnection regions.

Transmission losses are modeled in WinDS as a linear function of the distance the power is transmitted. These losses apply to the transmission of both wind and conventional generation, and are currently specified in terms of the fraction of power lost per MWh-mile. In the future, for the transmission of wind generation, we anticipate modifying WinDS to vary the fraction lost per MWh-mile with the length of the wind transmission line and with a rough estimate of the size of the transmission line. (The size of the transmission line cannot be a direct function of the amount of wind capacity to be transmitted, as this would introduce a nonlinearity into the LP formulation. Instead, we anticipate that before the LP is optimized for each period, a transmission line size will be selected based on the wind resource remaining in each region for each wind class.)