(29e) Dispatching Electricity Generation to Increase Water Availability in Times of Drought: A Feasibility Study in Texas

Over the past several decades, the frequency and intensity of droughts has increased throughout the United States, and this trend is predicted to continue.  The use of water as a coolant fluid in thermoelectric power generation accounts for 40% of freshwater withdrawals and 3% of consumption in the United States, which leaves these facilities vulnerable during times of decreased water availability.  Thus, drought can stress both the water and electricity generation resources within an affected region.

            This work seeks to characterize the feasibility, cost, and air quality implications of using the grid as a virtual water pipeline to “deliver” increased water availability in regions with high drought using a case study of eastern Texas during the summer of 2006.  This idea involves shifting power generation from parts of the grid which are under extreme drought conditions to areas with relatively more water availability, changing the magnitude and spatial extent of consumptions and withdrawals of freshwater resources for power plant cooling.  This approach could be rapidly implemented using existing power plants and transmission lines and could respond to spatial changes in the location of drought during other periods.  Changing the spatial location of power generation based on water availability will impact the cost of electricity generation by shifting generation to facilities which are lower on the dispatch order (and thus more expensive to operate) and may affect regional air quality by changing the location and magnitude of air quality pollutant precursors such as nitrogen oxides (NO_x) and sulfur dioxide (SO₂), which are related to regional ozone and fine particulate matter (PM) concentrations.  Texas offers an interesting test bed for the interconnections of water and electricity resources since the state has its own self-contained electric grid (ERCOT) that has a fuel-mix that is a reasonable approximation of electricity resources within the United States as whole.  In addition, Texas has 23 self-contained river basins that drain to the sea, reducing complications in modeling associated with downstream water users from other states.  Also, projected population growth within the state of Texas will likely add further the strain on the electric grid and water resources in the next few decades.

            This work is the first to demonstrate that ERCOT has the existing capacity to switch ~10% of its base case generation from a highly-drought stricken area to other parts of the grid.  For the June 2006 episode, water-withdrawing electricity generation was re-dispatched away from areas classified as being under extreme and exceptional drought by the U.S. Drought Monitor, which were in South Texas during this period.  While drought is a complex hydrological phenomenon and water planning decisions within this framework would likely be more complex than assumed in this work, the case study offers insight into the resiliency of the grid to respond to localized drought conditions.  In addition, scenarios in which water consumption in severe drought regions was also limited were examined to limit the exportation of water shortages to borderline drought regions.  The June 2006 episode is important for air quality planning in Texas as it was the period chosen for the State Implementation Plan (SIP) filed with the Environmental Protection Agency in response to non-attainment of the 8-hour ozone standards in the Dallas-Fort Worth area.

            Coupled models of ERCOT power generation (Power World Simulator), power plant water withdrawals and consumption, and regional air quality (Comprehensive Air Quality Model with Extensions, CAMX) were utilized in this work.  Factors for water consumption and use as well as air quality precursor emissions were developed for individual power plants within ERCOT on a per kWh basis so that these parameters changed with generation at the facility.  A base case model that minimized the total cost of electricity generation in ERCOT without water-based constraints was utilized in order to compare the impacts of four water-consumption based scenarios.  The first scenario involved turning off all water-withdrawing electricity generation in extreme and exceptional drought regions in South Texas and re-dispatching electricity generation from that region to other power plants on a least cost basis.  Other scenarios in which water consumption in severe drought regions (the next most intense US Drought Monitor classification) was constrained to be equal to the base case, to be reduced by 5%, and to be reduced by 10% were also modeled in the work.

            The cost of the drought-based re-dispatching scenarios was compared to the base case operation of the grid to determine the financial impacts of this strategy to increase water availability.  The grid-based re-dispatching modeling indicated that water consumption for thermoelectric power generation in South Texas could be forgone during the 2006 episode, creating enough water-availability in the region for the personal use of 360,000 Texans.  The cost of this strategy was compared to the installation and use of air cooling technologies at the subset of power plants in extreme and exceptional drought regions as well as to a pipeline project in Texas.  The use of drought-based grid dispatching was found to be generally more expensive than either of these strategies but had the advantage of being able to respond to seasonal changes in the location of the most severe drought within ERCOT.  Future work and costing strategies could potentially improve the economics of grid re-dispatching.  In addition, drought-based grid re-dispatching could be used as a parallel strategy while more long-term strategies, such as pipelines or air-cooling, were implemented.

            Drought-based re-dispatching of power generation also changed the spatial location of power plant emissions of ozone and PM precursors within Texas in addition to changing patterns of water use within the state.  Impacts on 8-hour ozone and average PM concentration, which are the metrics related to federal air quality standards, were generally modeled to be small.  The areas with the greatest net air quality benefits were generally down-wind from the South Texas drought area from which power generation was shifted.  The locations with the greatest air quality dis-benefit were near coal-fired facilities with increased power generation during the episode.