March 9, 2023

March 9, 2023

March 9, 2023

Unpacking Winter Storm Elliot’s grid emergencies

Unpacking Winter Storm Elliot’s grid emergencies

Unpacking Winter Storm Elliot’s grid emergencies

Trevor McManamon, Director of Market Development

Trevor McManamon, Director of Market Development

Trevor McManamon, Director of Market Development

Trevor McManamon, Director of Market Development

Trevor McManamon, Director of Market Development

Trevor McManamon, Director of Market Development

Winter Storm Elliot wreaked havoc this past Christmas Eve - not only on holiday travel, but also on electric grid stability across the eastern half of the United States. Two of the largest U.S. grid operators, ISO-New England (ISO-NE), which covers the six New England states, and PJM RTO, which spans 15 states plus Washington D.C., had emergency events on December 24th.


Including the day prior, this was PJM’s first system-wide emergency event since its implementation of more stringent penalties for underperformance during the tightest grid conditions, known as Capacity Performance. This was also ISO-NE’s second ever system-wide emergency event since implementing its own version, Pay-for-Performance. There were some similarities, though also some differences, in the circumstances that led to these two events. Digging into the causes of these two power system emergencies can help shed light on how the grid can be better prepared to maintain stability during future extreme weather events.


PJM: 4:25 am to 10:00 pm, December 24th


Despite starting the day prepared to deliver approximately 156 GW of capacity against a peak load forecast of approximately 127 GW, PJM experienced shortfalls that triggered an emergency event.


So what happened? Actual demand ended up over 10% higher than expected. In addition, about 46 GW of expected capacity ended up being unavailable. Over 90% of this capacity made PJM aware of its unavailability with less than an hour’s notice. Nearly half of this 46 GW was unavailable due to “plant equipment,” with another 10 GW attributing their unavailability to either the cold temperatures themselves, an inability to start or unit trips.


While those are not necessarily all caused by the cold temperatures, it’s reasonable to expect a large number of those were. An additional 10-15 GW was unavailable due to lack of fuel supply, which is a common occurrence during cold snaps in which natural gas is prioritized for heating purposes before power plants. The remaining two GW was unavailable due to emissions restrictions. A full report from PJM on the event is expected in April.


During the emergency event, capacity resources, which range from traditional fossil fuel-generating facilities to nuclear power to wind and solar, as well as battery storage, demand response and energy efficiency resources, were subject to penalties if they did not deliver on their commitments. Underperforming resources pay penalties into a pool, from which overperforming resources are compensated with bonus payments. 


The initial estimate of this pool is that $1-2 billion will be changing hands, although ratepayers won’t have to foot any of the bill. To put the magnitude of the penalties into perspective, a resource that was offline during the 18 hours of the event on the 24th would have to pay more than its entire year’s worth of capacity revenue in penalties.


ISO-NE: 4:30 pm – 7 pm, December 24th


Notably, ISO-NE’s emergency event wasn’t triggered by excessive peak demand. ISO-NE’s peak demand on December 24 was about 17.5 GW, and unlike PJM, ISO-NE’s day-ahead forecast for the peak was only off by 0.14%. ISO-NE saw peak demand over the summer reach a high of nearly 25 GW on July 20, yet had no issues maintaining reliability without calling an emergency event. ISO-NE had 28 GW of capacity committed for the season, which provides a sufficient buffer above the 17.5 GW peak demand.


Although lower-than-expected imports from nearby regions contributed to the problem, the largest cause of the emergency event was roughly two GW of capacity that unexpectedly went offline, some of which reported issues due to the cold weather.


So why was an emergency event called when several power plants totaling two GW of capacity had an unplanned outage? Nearly eight GW of committed capacity was unavailable to contribute to make up the difference due to their slow startup times. Having fast, flexible resources at the system operator’s disposal is crucial. As we saw in PJM, there is precedent for much more than 5-10% of a region’s capacity to unexpectedly have issues, in which case having additional fast responding resources available could be essential to the lights staying on.



Lowering the risk of emergencies


ISOs could utilize two potential mechanisms to reduce the likelihood of these situations materializing again in even worse circumstances:


More accurate capacity credits for fossil fuel resources that have challenges operating in extreme cold.


Due to their dispatchability during most circumstances, fossil fuel generators have traditionally received little, if any, reduction to the capacity they’re allowed to sell into markets, while renewable resources typically receive significant reductions.


However, given the struggles that these generators have experienced during times when the grid is most stressed, there should be meaningful reductions in the capacity that these generators are allowed to sell. There are ongoing discussions in both ISO-NE and PJM related to the future of their capacity markets, and one key focus is properly evaluating the true contributions to reliability of different resource types. The events of December 24, 2022 should be top of mind for the ISOs and stakeholders when determining appropriate capacity credits for fossil fuel generators.


Revisit the level of ancillary services products that ISOs procure.


While ancillary services span a wide array of uses, the most commonly utilized products, Regulation and Reserves, provide balance to the grid, ensuring that supply and demand stay near equilibrium and that supply is available to quickly replace any generation that is suddenly lost. Total reserve requirements are currently in the range of 5-10 GW in PJM and 2-3 GW in ISO-NE, which won’t always be sufficient to cover the magnitude of last minute, unexpected outages that occur during extreme cold weather.


Although PJM does increase its reserves requirement when it identifies a risk of natural gas supply shortages, it could benefit from including a factor related to increased risk of power plant equipment failure during extreme cold temperatures. This could be triggered by PJM’s existing Cold Weather Alert. ISO-NE has no triggers for additional reserves requirements during cold weather events, so they could benefit from a standardized methodology to increase reserves requirements during those periods.


Thankfully, Winter Storm Elliot didn’t lead to rolling blackouts being enacted in these regions - but it was a close call. Other areas in the U.S., including the Tennessee Valley Authority and Duke Energy territory in North and South Carolina, were not as fortunate and did succumb to rolling blackouts due to a shortage of generating capacity. Continuing to scale up the participation of more flexible resources in energy markets - like demand-side energy assets - will be crucial in order to provide sustainable, reliable support to the grid both during emergency events and on a day-to-day basis.

Winter Storm Elliot wreaked havoc this past Christmas Eve - not only on holiday travel, but also on electric grid stability across the eastern half of the United States. Two of the largest U.S. grid operators, ISO-New England (ISO-NE), which covers the six New England states, and PJM RTO, which spans 15 states plus Washington D.C., had emergency events on December 24th.


Including the day prior, this was PJM’s first system-wide emergency event since its implementation of more stringent penalties for underperformance during the tightest grid conditions, known as Capacity Performance. This was also ISO-NE’s second ever system-wide emergency event since implementing its own version, Pay-for-Performance. There were some similarities, though also some differences, in the circumstances that led to these two events. Digging into the causes of these two power system emergencies can help shed light on how the grid can be better prepared to maintain stability during future extreme weather events.


PJM: 4:25 am to 10:00 pm, December 24th


Despite starting the day prepared to deliver approximately 156 GW of capacity against a peak load forecast of approximately 127 GW, PJM experienced shortfalls that triggered an emergency event.


So what happened? Actual demand ended up over 10% higher than expected. In addition, about 46 GW of expected capacity ended up being unavailable. Over 90% of this capacity made PJM aware of its unavailability with less than an hour’s notice. Nearly half of this 46 GW was unavailable due to “plant equipment,” with another 10 GW attributing their unavailability to either the cold temperatures themselves, an inability to start or unit trips.


While those are not necessarily all caused by the cold temperatures, it’s reasonable to expect a large number of those were. An additional 10-15 GW was unavailable due to lack of fuel supply, which is a common occurrence during cold snaps in which natural gas is prioritized for heating purposes before power plants. The remaining two GW was unavailable due to emissions restrictions. A full report from PJM on the event is expected in April.


During the emergency event, capacity resources, which range from traditional fossil fuel-generating facilities to nuclear power to wind and solar, as well as battery storage, demand response and energy efficiency resources, were subject to penalties if they did not deliver on their commitments. Underperforming resources pay penalties into a pool, from which overperforming resources are compensated with bonus payments. 


The initial estimate of this pool is that $1-2 billion will be changing hands, although ratepayers won’t have to foot any of the bill. To put the magnitude of the penalties into perspective, a resource that was offline during the 18 hours of the event on the 24th would have to pay more than its entire year’s worth of capacity revenue in penalties.


ISO-NE: 4:30 pm – 7 pm, December 24th


Notably, ISO-NE’s emergency event wasn’t triggered by excessive peak demand. ISO-NE’s peak demand on December 24 was about 17.5 GW, and unlike PJM, ISO-NE’s day-ahead forecast for the peak was only off by 0.14%. ISO-NE saw peak demand over the summer reach a high of nearly 25 GW on July 20, yet had no issues maintaining reliability without calling an emergency event. ISO-NE had 28 GW of capacity committed for the season, which provides a sufficient buffer above the 17.5 GW peak demand.


Although lower-than-expected imports from nearby regions contributed to the problem, the largest cause of the emergency event was roughly two GW of capacity that unexpectedly went offline, some of which reported issues due to the cold weather.


So why was an emergency event called when several power plants totaling two GW of capacity had an unplanned outage? Nearly eight GW of committed capacity was unavailable to contribute to make up the difference due to their slow startup times. Having fast, flexible resources at the system operator’s disposal is crucial. As we saw in PJM, there is precedent for much more than 5-10% of a region’s capacity to unexpectedly have issues, in which case having additional fast responding resources available could be essential to the lights staying on.



Lowering the risk of emergencies


ISOs could utilize two potential mechanisms to reduce the likelihood of these situations materializing again in even worse circumstances:


More accurate capacity credits for fossil fuel resources that have challenges operating in extreme cold.


Due to their dispatchability during most circumstances, fossil fuel generators have traditionally received little, if any, reduction to the capacity they’re allowed to sell into markets, while renewable resources typically receive significant reductions.


However, given the struggles that these generators have experienced during times when the grid is most stressed, there should be meaningful reductions in the capacity that these generators are allowed to sell. There are ongoing discussions in both ISO-NE and PJM related to the future of their capacity markets, and one key focus is properly evaluating the true contributions to reliability of different resource types. The events of December 24, 2022 should be top of mind for the ISOs and stakeholders when determining appropriate capacity credits for fossil fuel generators.


Revisit the level of ancillary services products that ISOs procure.


While ancillary services span a wide array of uses, the most commonly utilized products, Regulation and Reserves, provide balance to the grid, ensuring that supply and demand stay near equilibrium and that supply is available to quickly replace any generation that is suddenly lost. Total reserve requirements are currently in the range of 5-10 GW in PJM and 2-3 GW in ISO-NE, which won’t always be sufficient to cover the magnitude of last minute, unexpected outages that occur during extreme cold weather.


Although PJM does increase its reserves requirement when it identifies a risk of natural gas supply shortages, it could benefit from including a factor related to increased risk of power plant equipment failure during extreme cold temperatures. This could be triggered by PJM’s existing Cold Weather Alert. ISO-NE has no triggers for additional reserves requirements during cold weather events, so they could benefit from a standardized methodology to increase reserves requirements during those periods.


Thankfully, Winter Storm Elliot didn’t lead to rolling blackouts being enacted in these regions - but it was a close call. Other areas in the U.S., including the Tennessee Valley Authority and Duke Energy territory in North and South Carolina, were not as fortunate and did succumb to rolling blackouts due to a shortage of generating capacity. Continuing to scale up the participation of more flexible resources in energy markets - like demand-side energy assets - will be crucial in order to provide sustainable, reliable support to the grid both during emergency events and on a day-to-day basis.

Winter Storm Elliot wreaked havoc this past Christmas Eve - not only on holiday travel, but also on electric grid stability across the eastern half of the United States. Two of the largest U.S. grid operators, ISO-New England (ISO-NE), which covers the six New England states, and PJM RTO, which spans 15 states plus Washington D.C., had emergency events on December 24th.


Including the day prior, this was PJM’s first system-wide emergency event since its implementation of more stringent penalties for underperformance during the tightest grid conditions, known as Capacity Performance. This was also ISO-NE’s second ever system-wide emergency event since implementing its own version, Pay-for-Performance. There were some similarities, though also some differences, in the circumstances that led to these two events. Digging into the causes of these two power system emergencies can help shed light on how the grid can be better prepared to maintain stability during future extreme weather events.


PJM: 4:25 am to 10:00 pm, December 24th


Despite starting the day prepared to deliver approximately 156 GW of capacity against a peak load forecast of approximately 127 GW, PJM experienced shortfalls that triggered an emergency event.


So what happened? Actual demand ended up over 10% higher than expected. In addition, about 46 GW of expected capacity ended up being unavailable. Over 90% of this capacity made PJM aware of its unavailability with less than an hour’s notice. Nearly half of this 46 GW was unavailable due to “plant equipment,” with another 10 GW attributing their unavailability to either the cold temperatures themselves, an inability to start or unit trips.


While those are not necessarily all caused by the cold temperatures, it’s reasonable to expect a large number of those were. An additional 10-15 GW was unavailable due to lack of fuel supply, which is a common occurrence during cold snaps in which natural gas is prioritized for heating purposes before power plants. The remaining two GW was unavailable due to emissions restrictions. A full report from PJM on the event is expected in April.


During the emergency event, capacity resources, which range from traditional fossil fuel-generating facilities to nuclear power to wind and solar, as well as battery storage, demand response and energy efficiency resources, were subject to penalties if they did not deliver on their commitments. Underperforming resources pay penalties into a pool, from which overperforming resources are compensated with bonus payments. 


The initial estimate of this pool is that $1-2 billion will be changing hands, although ratepayers won’t have to foot any of the bill. To put the magnitude of the penalties into perspective, a resource that was offline during the 18 hours of the event on the 24th would have to pay more than its entire year’s worth of capacity revenue in penalties.


ISO-NE: 4:30 pm – 7 pm, December 24th


Notably, ISO-NE’s emergency event wasn’t triggered by excessive peak demand. ISO-NE’s peak demand on December 24 was about 17.5 GW, and unlike PJM, ISO-NE’s day-ahead forecast for the peak was only off by 0.14%. ISO-NE saw peak demand over the summer reach a high of nearly 25 GW on July 20, yet had no issues maintaining reliability without calling an emergency event. ISO-NE had 28 GW of capacity committed for the season, which provides a sufficient buffer above the 17.5 GW peak demand.


Although lower-than-expected imports from nearby regions contributed to the problem, the largest cause of the emergency event was roughly two GW of capacity that unexpectedly went offline, some of which reported issues due to the cold weather.


So why was an emergency event called when several power plants totaling two GW of capacity had an unplanned outage? Nearly eight GW of committed capacity was unavailable to contribute to make up the difference due to their slow startup times. Having fast, flexible resources at the system operator’s disposal is crucial. As we saw in PJM, there is precedent for much more than 5-10% of a region’s capacity to unexpectedly have issues, in which case having additional fast responding resources available could be essential to the lights staying on.



Lowering the risk of emergencies


ISOs could utilize two potential mechanisms to reduce the likelihood of these situations materializing again in even worse circumstances:


More accurate capacity credits for fossil fuel resources that have challenges operating in extreme cold.


Due to their dispatchability during most circumstances, fossil fuel generators have traditionally received little, if any, reduction to the capacity they’re allowed to sell into markets, while renewable resources typically receive significant reductions.


However, given the struggles that these generators have experienced during times when the grid is most stressed, there should be meaningful reductions in the capacity that these generators are allowed to sell. There are ongoing discussions in both ISO-NE and PJM related to the future of their capacity markets, and one key focus is properly evaluating the true contributions to reliability of different resource types. The events of December 24, 2022 should be top of mind for the ISOs and stakeholders when determining appropriate capacity credits for fossil fuel generators.


Revisit the level of ancillary services products that ISOs procure.


While ancillary services span a wide array of uses, the most commonly utilized products, Regulation and Reserves, provide balance to the grid, ensuring that supply and demand stay near equilibrium and that supply is available to quickly replace any generation that is suddenly lost. Total reserve requirements are currently in the range of 5-10 GW in PJM and 2-3 GW in ISO-NE, which won’t always be sufficient to cover the magnitude of last minute, unexpected outages that occur during extreme cold weather.


Although PJM does increase its reserves requirement when it identifies a risk of natural gas supply shortages, it could benefit from including a factor related to increased risk of power plant equipment failure during extreme cold temperatures. This could be triggered by PJM’s existing Cold Weather Alert. ISO-NE has no triggers for additional reserves requirements during cold weather events, so they could benefit from a standardized methodology to increase reserves requirements during those periods.


Thankfully, Winter Storm Elliot didn’t lead to rolling blackouts being enacted in these regions - but it was a close call. Other areas in the U.S., including the Tennessee Valley Authority and Duke Energy territory in North and South Carolina, were not as fortunate and did succumb to rolling blackouts due to a shortage of generating capacity. Continuing to scale up the participation of more flexible resources in energy markets - like demand-side energy assets - will be crucial in order to provide sustainable, reliable support to the grid both during emergency events and on a day-to-day basis.

Get started with grid services

Access energy markets