Here’s how energy markets across the U.S. are reshaping their capacity frameworks

Every energy market in the U.S. is currently facing the perfect storm of reliability risk. Demand for electricity is increasing, supply for electricity is decreasing due to the retirement of thermal generation and the grid is grappling with an accelerated integration of intermittent resources - all coupled with more extreme weather and more frequent climate disasters. 

Recent grid emergency events - including the 2020 and 2022 heat waves in California, Winter Storm Uri in Texas and Winter Storm Elliot on the East Coast - have been forcing mechanisms for Independent system operators (ISOs) and regional transmission organizations (RTOs) across North America to thoroughly reexamine their capacity frameworks. We’re starting to see grid operators reshape the reliability models that underpin major resource adequacy decisions, the constructs for how resources are accredited and the testing and penalty guardrails that incentivize resources to reliably show up when needed. 

Here are some of the most significant trends we currently see developing in capacity markets across the country:

Grid operators are rethinking reliability metrics. 

Historically, the norm across energy markets has been to use Loss of Load Expectation (LOLE) as the primary reliability metric. LOLE is the expected number of times in one or more years where supply will be unable to meet demand and a firm load shed is required. 

However, LOLE only looks at the frequency of emergency events; it doesn’t take into account event duration, magnitude or correlated contingencies (the probability that a lot of bad stuff will happen at once). During Winter Storm Elliott, lower availability of imports, increased demand due to the holiday season, higher than expected thermal outages, and infrastructure failures to transport liquified natural gas due to the unexpected freezing temperature all culminated in ISO-NE calling a Capacity Scarcity Event on December 24th, 2022. 

Given the struggles all ISO/RTOs have faced in maintaining reliability in the wake of more frequent extreme weather events, it’s clear that additional reliability metrics are required. As a result, markets such ERCOT and PJM are expanding their focus to include different reliability measurements that capture the breadth and depth of reliability risk. One example is Expected Unserved Energy (EUE), which measures how much energy will not be served in a given time period due to multiple events happening simultaneously, such as weather, generation outages or increased load. 

As grid operators incorporate more inclusive reliability metrics into their planning, we can expect to see changes in how all resource classes - including demand response - are valued. 

A more comprehensive view of accreditation values

Accreditation values determine how much capacity a resource class can sell in the capacity market. For example, an accreditation value of 80% means that for a resource with 10 MW of qualifying capacity, it will only be able to sell and get credit for 8 MW. 

NYISO and ISO-NE are both undergoing separate initiatives to redesign their capacity accreditation processes, with the goal to more accurately assess a resource’s ability to meet resource adequacy needs and to respond during emergency situations. The ISOs are looking at a more comprehensive list of factors to influence accreditation values, including historical performance, start-up times, natural gas constraints, fuel availability, correlated outages and sensitivities to different resource mixes on the system. 

It’s too early to tell what the new accreditation values will be for each resource class, but given recent reliability issues it’s safe to say that the accreditation process is becoming more stringent. 

Weatherization and infrastructure hardening are no longer nice-to-haves, but need-to-haves for increasingly winter-peaking systems.

Some narratives out there allude to the idea that recent grid failures are the result of an increasing penetration of intermittent resources, such as solar and wind, and that renewables are less reliable. However, during Winter Storm Elliot, only 3% of PJM’s installed capacity consisted of renewable resources - the rest was largely traditional dispatchable generation. And in January 2019, citizens in Newport, Rhode Island didn’t have access to heat or hot water for close to a week because a significant portion of the natural gas distribution system was shut down due to low pressure in the system. 

These events are leading markets to reexamine requirements for the weatherization of thermal plants or infrastructure hardening for natural gas delivery. PJM’s capacity market redesign proposal, for example, includes minimum winterization requirements that exceed the North American Electric Reliability Corporation (NERC) minimum requirements. If a resource is found to not be in compliance, it would not be able to participate in the winter commitment period. 

Weather-related impacts are also making their way into reliability models that will shape resource adequacy needs and accreditation value of resources. In its initial Reliability Standard draft, ERCOT proposes to reflect weather-induced thermal outages and fuel limitation outages during Winter Storm Uri. 

If done correctly, the additional weatherization requirements for resources or inclusion of weather impacts in reliability models should result in a decrease of accreditation values for fossil fuel generation. Increased weatherization requirements would also increase the cost to generators that have to undergo these changes, making demand response a cheaper and more reliable resource in comparison.

A shift towards seasonality 

As more intermittent and weather-sensitive resources are integrated into energy markets, annualized views on market planning are no longer sufficient. Rather, a more granular view on capacity planning is needed. 

Both ISO-NE and PJM are considering moving towards seasonal capacity markets, similar to those in NYISO and IESO. A seasonal capacity market would allow resources to take on three potential obligations: a winter obligation only, a summer obligation only or a winter and summer obligation combined. 

Seasonal constructs are more in line with how intermittent resources and weather-sensitive demand response operate because they allow resources to provide shaped offers to the market during different seasons. Solar, for example, can offer more capacity in the summer than in the winter - a seasonal construct wouldn’t penalize solar for not being able to perform as well during the winter season. Similarly, for weather-sensitive demand response, a seasonal obligation will more accurately reflect the operational characteristics of these resources and not discount them for their lack of load during the winter as an annual approach currently does. 

Some believe that an even more granular view, such as hourly, is needed to ensure reliability throughout all hours of the day. California is doing just that with a new “Slice-of-Day” construct that will require load-serving entities to procure enough resource adequacy to meet load during every hour of the day.   

__

These confluence of changes in capacity markets across the U.S. make it clear: traditional models won’t work for the grid of the future. In the face of increased threats to grid reliability due to climate change, grid operators are rethinking how they measure reliability, how they value different types of energy resources and how they structure capacity programs to meet the needs of the grid throughout the year. As capacity markets continue to evolve, distributed energy resources will have a critical role to play in the transition to a decarbonized grid.