By Michael Goggin, Vice President at Grid Strategies LLC
A new report by the Department of Energy and its fossil energy laboratory (National Energy Technology Laboratory, or NETL) overstates the contribution of coal power plants and understates the contribution of renewable resources to keeping the lights on during the January 2018 “Bomb Cyclone” cold snap event across much of the Mid-Atlantic and Northeast. In reality, coal power plants experienced a failure rate comparable to other energy sources during the event, while renewable energy output was well above average.
The report employs a flawed metric of resilience that does not indicate the performance of different types of generators, but instead simply finds which energy sources are the most expensive. As the Federal Energy Regulatory Commission weighs proposals to compensate resources for their contributions to electricity system resilience, this report should serve as a warning against the danger of creating new resilience metrics. Studies of resilience should focus on actual performance in providing needed grid services to customers, a metric at which renewable resources excelled while coal plants often failed during the recent event.
More importantly, the Bomb Cyclone event shows that the focus should be on the power lines and equipment that deliver electricity to customers, not the power plants that supply the generation. While no customers lost power during the Bomb Cyclone because of generation shortages, the sole focus of the NETL report, hundreds of thousands did lose power when electricity distribution lines and equipment failed in the extreme weather. This confirms previous analysis that over 96% of customer outage hours result from severe weather, while only 1 in 10,000 are related to generation supply issues.
The report shows coal is expensive, not resilient
The report’s sole metric of resilience is comparing the electricity output of different energy sources during the Bomb Cyclone event versus the preceding 26 days of December 2017. Coal generation increased more than gas generation during the Bomb Cyclone event, but that was only because most coal generation had been idle or running at low output during December 2017 as coal was more expensive than natural gas at that time.
The increased coal output is not a metric of resilience, but rather just a testament to the poor economics of coal generation that so much capacity had been idle or only partially utilized. While coal generation increased 36% during the Bomb Cyclone event relative to December versus a 14% increase for gas, the report notes that oil and dual fueled generation increased by over 1000%. Of course, that is only because oil was infrequently used for electricity generation in December because it costs an order of magnitude more than natural gas.
Identifying energy sources that increased output during a time of high prices will only find the resources that are the most expensive, not those that are most resilient. However, this is the sole metric employed by the NETL report, which explains that “we examine resilience afforded by each source of power generation by assessing the incremental daily average gigawatt hours during the [Bomb Cyclone] event above those of a typical winter day.” (page 4) According to that definition of resilience, an even more expensive energy source than oil would have appeared even more resilient than coal and oil because its utilization would have increased by infinity percent during a period of high prices.
The report claims that coal’s increased output is an indicator of its dispatchability and flexibility. In fact, oil and natural gas (and even solar and wind) generators far outperform coal and nuclear generators in their capability to provide dispatchability and flexibility, as noted by DOE’s August 2017 Staff Report (page 86). The NETL report does note that nuclear plants did not significantly increase their output during the Bomb Cyclone event, despite the fact that nuclear plants meet the report’s criteria of being “baseload” and having onsite fuel, while the oil generators that did increase their output do not meet those criteria. That confusion clearly shows that none of those metrics (increased utilization during times of high demand, having onsite fuel, or being “baseload”) are a meaningful measure of resilience.
A more accurate way to assess the performance of resources is their rate of failure during the Bomb Cyclone event. The following table from a report on the Bomb Cyclone (page 21) by PJM, the grid operator for the Mid-Atlantic and Great Lakes region, shows that in many cases coal plants experienced higher failure rates than natural gas and other resources, particularly during the early part of the event.
PJM generator forced outage rate as a percent of committed capacity during Bomb Cyclone
Coal generators were no more reliable than other resources because equipment failures, not fuel supply issues, were the primary cause of generator outages during the Bomb Cyclone, as shown in the chart below (from page 19 of PJM’s report).
Coal plant equipment failures also accounted for a large share of the failures during the 2014 Polar Vortex event and a similar cold snap that hit Texas in 2011, as shown in the charts below.
When Hurricane Harvey hit Texas last August, having large stocks of onsite fuel became a liability for several coal plants that were forced to operate at reduced output or switch to other fuels when their coal piles were inundated by extreme rainfall and flooding. NERC’s event report (page 17) noted that four coal plants were de-rated due to wet coal.
Renewable output greatly exceeded expectations during the Bomb Cyclone
The NETL report also incorrectly states that renewable energy output was below average during the Bomb Cyclone event because it arbitrarily compares renewables’ performance against a 26 day period in December that is not representative of typical performance. In reality, renewable energy output was well above average during the Bomb Cyclone event, with wind output higher than both average winter and average annual output across the Northeast throughout the event. More importantly, renewable output was several times higher than the level that grid operators plan for and compensate renewable resources for under their capacity markets.
During the most challenging periods of the Bomb Cyclone, in PJM wind output was more than 40 percent above average, while in New England wind output was more than twice its normal level. Wind’s output was three times greater than what PJM plans for and compensates wind for in its capacity market. A chart in NETL’s report does note that renewables provided an above-average 11 percent of on-peak generation during the Bomb Cyclone event across the grid operators it examined (page 22).
A true examination of resilience would assess actual performance in keeping the lights on for customers. Such an effort should focus on the transmission and distribution system failures that cause the vast majority of customer outages. Such an analysis would also include a range of threats to the power system. For example, coal and nuclear plants are vulnerable to other weather disturbances, such as flooding, drought, high temperatures, and rail congestion, while renewable resources like wind and solar PV are resilient to most of those events as they are not dependent on deliveries of fuel or cooling water.
