Evaluating the Impact of the 2017 Solar Eclipse on U.S. Western Interconnection Operations

Abstract

With support from the U.S. Department of Energy Solar Energy Technologies Office, the National Renewable Energy Laboratory (NREL) partnered with Peak Reliability to evaluate the impact of the August 21, 2017, total solar eclipse on reliability and electric grid operations in the Western Electricity Coordinating Council (WECC) territory. As a North American Electric Reliability Corporation Reliability Coordinator, Peak Reliability manages the grid reliability across all or parts of the14 western United States, British Columbia, and the northern region of Baja California, Mexico. Peak Reliability used inputs from the study to augment its own analysis for preparing transmission operating plans for the day of the eclipse. Peak Reliability also provided NREL with data from across its footprint to uncover important insights into the impact of the eclipse from the penetration of photovoltaics (PV) along its path. The 2017 total solar eclipse came and went without causing any issues to the operation of the North American electric power system. This report presents the results of NREL's studies performed before and after the August 21 event. In a preevent analysis, NREL researchers focused on the spatial and temporal profile of the eclipse and how the output of PV sites across WECC would be affected. Although the size and location of utility-scale PV (UPV) sites are known and their outputs are monitored, distributed PV (DPV) (e.g., rooftop solar) poses a greater challenge for grid operations because of its lack of visibility. The estimated installed capacity of UPV and DPV across the WECC region is 15,800 MW (without concentrating solar power plants) and 9,200 MW, respectively. NREL used databases developed through previous efforts funded by the U.S. Department of Energy to determine how much DPV was present at the substation level and then estimated the impact of the eclipse on its output. This geospatial analysis assumes a typical meteorological day from NREL's typical meteorological year database. NREL researchers estimated that during the peak of the event, the loss of PV would be around 5.2 GW, with UPV loss at 4 GW and the rest from DPV. The second part of the analysis was a simulation of production costs. Using the results from the geospatial study, the economic dispatch and unit commitment strategies were evaluated to see how the other generators would be redispatched. The analysis showed that the loss of PV generation can be easily compensated by the existing generation fleet, with the natural gas fleet picking up the majority of the slack because of its flexibility. The change in system production costs were minimal. The third part of the analysis consisted of transient simulations to assess the stability and reliability of grid operations. No reliability concerns were identified, even for large system disturbance events. The system also exhibited high damping capabilities to mitigate inter-area oscillations.

Renewable Energy Management System: Optimum Design and Hourly Dispatch

“…Applied studies on actual RE data conclude that suppressing the drastic effects of RES on the power system requires large amounts of load-following reserves [1][2][3][4][5]. For example, Sørensen et al [6] highlight that the most extreme drop in a wind farm's output within a 1-minute interval is 6% of the plant's rating.…”

“…is defined as the HA deviation δ t , and described in (5). This deviation might be inevitable, and may also be tolerable within a certain limit.…”

“…The hourly dispatch problem is described by (6) - (14), in addition to (1) - (5). This problem is solved for the full season length of 2184 hours.…”

“…The net power output of the HREP is sold to the grid, as described by (7). Variables with undetermined sign are decomposed to positive and negative parts in (1), (5), and (8). The BESS SoC is updated in every period by (10), where η BESS is the BESS round-trip efficiency.…”

Case study on the effects of partial solar eclipse on distributed PV systems and management areas

“…On the basis of weather prediction and the projected loss of generation, many other conventional generation sources and energy storage are dispatched to compensate for the loss in the generation and the change in load during the eclipse [6]. This in‐depth pre‐event analysis is, however, largely restricted to utility‐scale PV systems, because there is little operational visibility on distributed PV such as rooftop solar and small commercial systems at offices, universities, and buildings [7]. Although not of concern now, the dynamics will change when the penetration level (the volume of PV‐connected relative to the amount of load) of such distributed PV systems increases drastically, wherein grid operations will be significantly impacted [8, 9].…”