Published: March 24, 2025
By Danielle Powers and Mark Karl
Key Takeaways:
- While capacity markets have historically played an important role in supporting grid reliability, their foundational design no longer reflects the realities of today’s power grid. These markets were built around a generation fleet dominated by dispatchable resources, yet the current mix increasingly includes intermittent renewables and resources shaped by policy objectives.
- Some of the core assumptions embedded in capacity market structures, such as the idea that all capacity resources contribute equally to reliability or that investment will consistently respond to price signals, have been called into question. Recent experiences with price volatility, shifting accreditation standards, and out-of-market interventions have raised concerns about whether these markets are sending the right signals to support long-term investment in reliability.
- These challenges have prompted a broader conversation about whether the existing capacity market framework is sufficient on its own, or whether alternative models should be considered. Options such as bilateral contracting, centralized procurement, or new reliability products that reflect the specific attributes needed to operate the grid may offer more tailored solutions for today’s changing landscape.
For decades, capacity markets have played a central role in the design of wholesale power markets in the United States, particularly in regions such as PJM, ISO-New England, and NYISO. These markets were originally established to help ensure grid reliability by securing adequate generation to meet peak demand. However, as resource portfolios evolve, renewable energy grows in prominence, and policy priorities shift, questions have emerged about whether existing capacity market structures remain well-suited to today’s energy landscape. These developments have prompted discussion around whether incremental adjustments are sufficient, or if more substantial reforms may be necessary.
The Origins and Evolution of Capacity Markets
Capacity markets were introduced in the late 1990s and early 2000s as part of deregulation efforts in the energy sector. Their purpose was to address the missing money problem, which is the gap between the revenue that generating resources need to cover their fixed costs and the revenue they actually earn in the energy and reserves markets. Capacity markets were introduced as a solution to this problem, providing an additional stream of revenue to ensure sufficient investment in generation capacity and long-term grid reliability.
The fundamental assumption was that all power plants contributed equally to system reliability, that all megawatts were created equal, and that a simple auction-based market could incentivize investment in new resources as older ones retired. However, this model was built around a system dominated by traditional, dispatchable power plants—coal, natural gas, and nuclear generators—which provided energy and essential grid services such as voltage control, frequency stability, and inertia. These plants could be counted on to supply power and gird services whenever demand required it.
Another foundational element of capacity market design was that new generation entry would come as a result of a market signal for the needed capacity. The rapid expansion of renewable resources, along with the policies designed to incent and, in some cases, subsidize these resources, has upended these foundational assumptions. In addition, the price volatility and rule instability that result from constant “tweaking” of the design in an attempt to address shortcomings as they arise makes it difficult to support substantial investment.
Why Capacity Markets No Longer Work
Why are capacity markets increasingly seen as unsustainable in their current form? The reasons are simple: the foundational assumptions on which capacity markets were created no longer hold true.
- Mismatch Between Capacity Markets and Modern Energy Resources
The original design of capacity markets assumed that all qualified capacity megawatts were functionally equivalent. This is no longer the case. The modern energy mix includes a growing share of intermittent renewables like wind and solar, which do not always generate electricity when needed. Capacity markets have attempted to adapt through mechanisms like the Effective Load Carrying Capability (ELCC) rating process, performance incentives, and fuel supply requirements, but these changes are incremental fixes that fail to address the full reliability need and fail to address the larger issue: capacity markets are designed for a power grid and a supply resource mix that no longer exists. - Distortion from Public Policy Interventions
The rise of state-level clean energy mandates and direct subsidies for renewables has further complicated capacity markets. Many new renewable projects are entering the market not because of price signals, but because they receive out-of-market financial support to achieve specific policy goals. All else being equal, this artificially suppresses capacity prices, making it even harder for traditional generators to remain viable. As a result, necessary resources are being pushed toward retirement, even when they are still essential for reliability. Capacity markets were never designed to accommodate these policy-driven shifts, and they have proven ineffective at integrating them into the broader reliability framework. - Failure to Account for Essential Grid Services
Traditional power plants provided a “bundle” of reliability attributes beyond just megawatts of capacity. They offered fuel security, , frequency regulation, and fast-ramping capabilities. Although different resource technologies provided different quantities of these attributes, for the most part, they provided the full “bundle.” New capacity resources, particularly renewables, do not inherently provide all these same services, yet capacity markets still treat them as interchangeable with traditional generators. This has led to reliability gaps, forcing grid operators like PJM to intervene with out-of-market payments to keep critical plants from shutting down. If grid operators must frequently override market outcomes to ensure reliability, it is a clear indication that the market is failing. - Increasing Market Volatility and Inefficiencies
Capacity market prices have become increasingly unstable, fluctuating from near-zero levels in oversupplied years to dramatic spikes when retirements accelerate. The most recent PJM Base Residual Auction saw prices jump nearly tenfold, largely due to resource retirements and new constraints placed on capacity accreditation. Such volatility discourages long-term investment in new generation, as developers cannot count on stable revenue streams. This instability undermines the very purpose of capacity markets, which is to provide financial certainty for generators and ensure long-term resource adequacy. - Inability to Adapt to Rapid Changes in Demand
Since capacity markets were first introduced, electricity demand in the U.S. has grown modestly overall. From the early 2000s to the mid-2010s, total electricity consumption remained relatively flat, influenced by improvements in energy efficiency, a shift toward a more service-based economy, and the decline of energy-intensive manufacturing. As a result, capacity markets provided sufficient incentive for the construction of new generation resources. However, the demand for electricity from data centers is expected to grow significantly in the coming years due to the rapid expansion of cloud computing, artificial intelligence (AI), cryptocurrency mining, and the electrification of the economy. According to the NERC 2024 Long-Term Reliability Assessment, summer peak demand for the U.S. is expected to grow by 132 GW over the next 10 years, significantly greater than the 80 GW projected in the 2023 assessment. Given the substantial challenges faced in recent years in meeting even modest load growth, it is extremely unlikely the current capacity construct and markets will be capable of delivering the resources needed in time to meet the projected increase.
What Comes Next? Alternatives to Capacity Markets
To borrow a phrase from FERC Chair Mark Christie, “we have a rendezvous with reality”. It is time to move beyond incremental adjustments to capacity markets and begin exploring alternative approaches to ensuring grid reliability. We can’t afford to continue to put reliability at risk when “baseload” retirements are happening faster than dispatchable generation can be added. As Chair Christie recently stated in comments made at CERAWeek when stressing the need for dispatchable resources to maintain grid reliability, “we’re simply not ready to run a grid where we don’t have dispatchable resources”. How can the current capacity market design incent dispatchable gas-fired resources critical to ensuring reliability when these resources might operate for a handful of peak hours during the year?
There are market designs, such as those used in MISO and ERCOT, that provide useful models. MISO relies on load-serving entities (LSEs) to demonstrate sufficient resource adequacy through bilateral contracts and self-supply options. ERCOT operates an “energy-only” market, where real-time prices reflect scarcity conditions and encourage investment in new capacity when needed.
Another viable approach is the creation of a centralized procurement agency—such as a state or regional power authority—that would oversee long-term reliability contracts. It is important to recognize that the creation of such an agency need not represent the abandonment of wholesale electricity markets. Certainly, the energy and reserves markets need not change, and can continue to provide the same efficiency benefits they do today.
The power authority need not own or operate supply resources either. Such an entity could competitively procure the right mix of resources on a contract basis from independent owners and operators to balance dispatchability, fuel security, reliability, affordability, and policy goals, rather than relying on an outdated market mechanism that no longer serves its intended purpose. The procurement process could also allow for self-supply by load serving entities, utilities, or municipal systems and would provide a more stable revenue stream to facilitate lower cost financing.
Conclusion
The electricity system is undergoing a fundamental transformation, and capacity markets are failing to keep pace. Designed for a different era, they no longer align with the realities of modern energy markets, technological advancements, and policy objectives. Instead of continuing to modify an outdated system, policymakers and grid operators should move toward market structures that better reflect today’s energy needs. Whether through direct procurement, LSE-led resource planning, or new reliability products that disaggregate the attributes currently assumed in the current capacity product, the time has come to move beyond capacity markets and embrace a model that ensures a reliable, cost-effective, and sustainable energy future.
Links to Cited Sources:
2024 Long-Term Reliability Assessment. North American Electric Reliability Corporation
“US Grid Must Embrace Natural Gas in ‘Rendezvous with Reality’: FERC Chair.” Upstreamonline.com
All views expressed by the authors are solely the authors’ current views and do not reflect the views of Concentric Energy Advisors, Inc., its affiliates, subsidiaries, related companies, or clients. The authors’ views are based upon information the authors consider reliable at the time of publication. However, neither Concentric Energy Advisors, Inc., nor its affiliates, subsidiaries, and related companies warrant the information’s completeness or accuracy, and it should not be relied upon as such.
On February 20, 2025, the Federal Energy Regulatory Commission (“FERC”) issued a highly anticipated order under Section 206 of the Federal Power Act addressing concerns related to large loads co-located at generating facilities within the PJM Interconnection. The growing interest in co-location arrangements, particularly involving data centers and industrial facilities, has raised questions about how interconnected generators should serve these co-located loads when they are physically connected to an existing or planned generator on the generator side of the point of interconnection. These arrangements have introduced issues around potential cross-subsidization, cost shifting, grid reliability, resource adequacy, and jurisdictional boundaries.
In this show-cause order (“Order”), FERC found PJM’s Tariff to be potentially unjust, unreasonable, unduly discriminatory, or preferential for lacking explicit provisions on co-location arrangements. The Order highlighted several key issues:
1. Jurisdictional Debate:
Co-located arrangements introduce jurisdictional questions. Some stakeholders have argued that FERC’s jurisdiction should be limited to interstate wholesale transactions and that states should retain control over retail sales and behind-the-meter arrangements. Others argue that load served directly by a generator is analogous to behind-the-meter generation and is exempt from FERC oversight. PJM and others maintain that co-located loads still benefit from grid services and should thus fall under FERC’s oversight when those services affect wholesale rates and grid reliability.
2. Cost Allocation and Grid Services:
A significant concern is whether co-located loads can fully isolate from the electric grid and avoid paying their share of costs for transmission services and for ancillary services from PJM. PJM and its market monitor have argued that co-located loads should be treated like other grid-connected loads and should pay for network services, ancillary services, and capacity. Other stakeholders have countered that since co-located loads can fully isolate and not draw power from the grid, they should not incur transmission service charges.
3. Reliability and Resource Adequacy:
Several parties have highlighted potential risks that co-located loads might impose on grid stability, particularly when large loads bypass the traditional planning process. For example, sudden shifts in demand or the loss of a co-located generator could compromise grid stability. PJM emphasized that the rapid growth of such loads could strain existing capacity reserves and suggested that planning frameworks need adjustments to incorporate these arrangements effectively. However, proponents of co-located load arrangements have argued that such configurations can offer benefits like reducing grid congestion, easing interconnection backlogs, and energizing data centers more quickly.
In the Order, FERC directed PJM and the Transmission Owners to provide justifications for the current tariff or suggest changes within 30 days. These justifications must address concerns related to jurisdiction, cost allocation, reliability, and potential discriminatory practices. FERC requested answers to approximately 40 questions related to jurisdictional principles, the type of transmission service used under various configurations, cost allocation, and the impacts on the wholesale market and ancillary services.
Concentric Energy Advisors’ Wholesale Energy Markets practice helps utilities, independent power producers, and government entities shape and understand wholesale electric market design and operational issues. Contact Danielle Powers, Chief Executive Officer, at dpowers@ceadvisors.com or 508.263.6219 to learn more about our services.
— All views expressed by the author are solely the author’s current views and do not reflect the views of Concentric Energy Advisors, Inc., its affiliates, subsidiaries, related companies, or clients. The author’s views are based upon information the author considers reliable at the time of publication. However, neither Concentric Energy Advisors, Inc., nor its affiliates, subsidiaries, and related companies warrant the information’s completeness or accuracy, and it should not be relied upon as such.
Published: July 2, 2024
By: Concentric Staff Writer
A proposal for a special arrangement whereby a planned Amazon data center would be directly supplied by a co-located nuclear power plant in Pennsylvania sent rapid ripples across the industry this month. Other users of the transmission system are sounding the alarm over its possible effects on other customers and the precedent it could set, supported by Concentric Energy Advisors (Concentric).
There has been a quick line-up of parties filing to intervene in reaction to the proposal filed on June 3 with the Federal Energy Regulatory Commission by the PJM Interconnection LLC (PJM), the grid operator of the mid-Atlantic region including 13 states and the District of Columbia. The deal concerns a data center campus formerly owned by Talen Energy that the company sold to Amazon Web Services for $650 million earlier this year and sits next to the 2,514-megawatt (MW) Susquehanna Steam Electric Station, which Talen Energy also owns. Amazon plans to develop a large data center at the site to be powered by the close-by nuclear plant, which would be the largest such installation in U.S. history.
The proposal filed by and for PJM as transmission provider, Susquehanna Nuclear LLC as interconnection customer, and PPL Electric Utilities Corporation would amend the existing interconnection service agreement (ISA) to raise from 300 megawatts to 480 MW the amount of co-located load from the data center and make other revisions and changes [Docket No. ER24-2172].
Concentric is among the commenters; on June 24 filing an affidavit from Chairman of the Board, John Reed and Chief Executive Officer, Danielle Powers. Concentric, drawing from its decades of experience in utility regulation, filed the affidavit in support of a protest of PJM’s filing that was submitted to FERC by Exelon Corporation (Exelon) and American Electric Power Service Corporation (AEP).
“The significance of this case lies in its potential to set far-reaching precedents for how similar situations will be handled in the future,” the Concentric filing says. “The sheer scale of the Co-Located Load presents unique challenges and complexities that have not been encountered before on such a magnitude.”
Ms. Powers, in an interview, said that the proposed amendment to the ISA provided little detail on the costs to other customers. It is this lack of detail and impact on customers that are so important to understand and why FERC must set this matter for hearing, she said.
“We need to understand what the issues are and what you are requiring of both the generator and the co-located load,” Ms. Powers said of the amended ISA. “Since the load is located behind the generator, there are many unanswered questions around how much and how the generator offers its capacity and energy into the PJM wholesale markets, and what the co-located load will or should pay under the PJM Open Access Transmission Tariff.” She also said that PJM and market participants have been involved in lengthy discussions on how to deal with co-located load and have been unable to come to a consensus. A filing places these unresolved issues in front of the FERC, she said.
The Concentric affidavit says there are substantial implications for the case as it could “fundamentally impact the regulatory landscape, influencing how regulators address cost allocation and rate design.” If the agreement results in significant avoided costs it could lead to other similar arrangements, leading to widespread cost-allocation issues and leaving unresolved questions of cost responsibility for using the electric grid, the filing says. The cost shift could be up to $140 million per year and the avoided transmission component makes up approximately 98 percent of the avoided costs, Concentric said.
In the protest filed by Exelon and AEP, the two companies said the matter must be set for hearing because of many unresolved facts and that it includes “by the filing’s own admission, an ISA that establishes novel configuration.” If FERC does not set the matter for hearing, it should reject the ISA amendment because it amounts to an “end run” around PJM’s stakeholder process and violates PJM’s tariff by creating a new type of load, the protest says.
“The Parties’ non-conforming ISA must be set for hearing because it raises more questions than it answers,” Exelon and AEP said. “Given the scant information provided in the transmittal, absent further factual development, the Commission will be unable to make an informed decision whether to accept the ISA and parties to the proceeding will be denied necessary notice and opportunity to raise informed protests before the Commission.”
There are huge financial consequences around the filing as there are likely to be many other similar situations, the protest says, and in the absence of other precedent, it is reasonable to think that other parties could take a similar approach.
“The number of expected, non-conforming ISAs that the filing anticipates could have a profound effect on the market,” the protest says. “Should large quantities of load rush to co-locate with generation on terms that bear even a resemblance to the ISA at issue here, PJM capacity markets will have steadily decreasing volume as the capacity resources flee to serve load that uses and benefits from—but does not pay for—the transmission system and the ancillary services that keep the system running.”
But Talen Energy (Talen) fired back in the public arena, on June 27 issuing a press release characterizing the proposal as a new way to deal with rising data center demand. Powering this new category will require both metered and behind-the-meter solutions, the company said.
“Exelon and AEP’s protest of the Susquehanna ISA is a misguided attempt to stifle this innovation by interfering with an ISA amendment agreed to and supported by all impacted parties – which Exelon and AEP decidedly are not,” the press release says. “The factual recitations in the protest are demonstrably false. The legal positions are demonstrably infirm.”
Nearly all of the issues raised by Exelon and AEP are not even subject to FERC oversight, Talen argued, because transmission is not implicated, and Talen has a right to contract with Amazon for long-term, committed power. It also said that PPL agrees that Talen has the right to sell power directly to Amazon, and the filing is supported by PJM, it said.
The proposal to FERC by PJM and others involved in the co-located data center/power plant project raises many new questions regarding what is being recognized as a new frontier in energy infrastructure development. As of June 29, there were 33 motions to intervene filed in the docket.
—
All views expressed by the author are solely the author’s current views and do not reflect the views of Concentric Energy Advisors, Inc., its affiliates, subsidiaries, related companies, or clients. The author’s views are based upon information the author considers reliable at the time of publication. However, neither Concentric Energy Advisors, Inc., nor its affiliates, subsidiaries, and related companies warrant the information’s completeness or accuracy, and it should not be relied upon as such.