> ## Content Index
> Fetch the complete content index at: https://www.aixenergy.io/llms.txt
> Use this file to discover other available public pages before exploring further.

# Bring Your Own Capacity, Take Two: What PJM's November 16 Refiling Has to Fix
- URL: https://www.aixenergy.io/byoc-take-two-pjm-refiling/
- Published: 2026-11-14T03:23:20.000Z
- Updated: 2026-11-14T03:23:20.000Z
- Description: FERC kept PJM's large-load registry and sent the rest back. A second try at bring-your-own capacity needs deliverability, ordinary ELCC accreditation and a curtailment path that runs through the states.
- Author: Brandon N. Owens
- Tags: Policy & Regulation

Picture the meeting. On one side of the table sits a utility that has been told by its grid operator that the region will not have enough capacity for the 2027/2028 delivery year. On the other sits a data center developer with land, a power purchase agreement half negotiated, and a board that wants to know when the first hall energizes. Both of them were waiting for a rule that would let the developer bring its own new supply and skip the queue for scarce existing capacity. On October 9 the Federal Energy Regulatory Commission sent most of that rule back.

The order, in docket ER26-3515, kept two pieces of what PJM Interconnection filed on August 13 and rejected the rest. PJM, which runs the wholesale grid and capacity market from New Jersey to Illinois, now says it will make a set of new filings on or about November 16.[1](#fn1)[2](#fn2) Whether a revised bring-your-own arrangement rides with them is not yet clear. What is clear, from the order itself, is what a second attempt would have to get right.

## What survived and what did not

FERC accepted PJM’s definition of a Large Load, meaning end-use load with a cumulative peak of 50 megawatts or more at a single site, and the Large Load Registry that goes with it, both effective October 12\. It rejected everything else: the Interim Resource Adequacy Service, known as IRAS; the requirement that load-serving entities serving new large loads designate new capacity at least equal to those loads’ peak; the rules for Bring Your Own New Capacity, or BYONC; and the plan to take uncovered new large load out of the capacity auction’s demand curve from 2029/2030.[3](#fn3)

The commission did not treat these as separate failures. It found the rejected pieces “inextricably linked,” with the new-capacity obligation serving as the foundation for both IRAS and BYONC. Only the definition and the registry could stand on their own.[4](#fn4)

It would be easy to call the registry the consolation prize. That undersells it. FERC found it would let PJM “better forecast load growth and to plan and manage its transmission system,” and that its value stands “separate and apart from” IRAS.[5](#fn5) For the first time, load-serving entities must tell PJM where their big customers are, how large they are and how fast they ramp. The 50 MW threshold counts affiliated facilities within a mile as one site, which closes the obvious trick of building several 49 MW halls next door to each other.[6](#fn6) Every serious version of bring-your-own capacity will depend on that data, because you cannot credit a load for the supply it brings until you know where the load is.

## Why curtailment by default failed

IRAS was the stick behind the carrot. A new large load whose utility did not bring enough new capacity would be “subject to load reduction requests” when supply ran short in real time.[7](#fn7) PJM described this as a wholesale arrangement: it would tell electric distributors how many megawatts to shed, and state regulators would decide which customers.

FERC’s problem was that the tariff PJM actually filed did not say that. The commission found that several provisions “appear to place obligations directly on Eligible New Large Loads,” which are mostly retail customers, that others appeared to dictate which retail customers would be curtailed, and that the compensation scheme “appears to establish a retail rate.”[8](#fn8) Those are state matters under the Federal Power Act. FERC went out of its way to say that the version described in PJM’s transmittal letter “may be within the Commission’s jurisdiction.” The words in the tariff were the issue, not the idea.[9](#fn9)

There was a reliability problem too. NERC Reliability Standard EOP-011-4 permits involuntary curtailment of firm load only after an Energy Emergency Alert level 3\. IRAS would have cut new large loads much earlier in an emergency.[10](#fn10) Commissioner David Rosner, concurring, pressed the point harder. PJM’s design, he wrote, “flips this paradigm” by forcing IRAS load off before the demand response customers who are paid to volunteer. He was not soft on data centers either: “homes, hospitals, and schools should never be curtailed before data centers.” His objection was to sequence, and to a market that might signal plenty of capacity while thousands of megawatts were being shed on an involuntary basis.[11](#fn11)

For the deal table, the lesson is that a curtailment backstop can survive in some form, but it has to work through state retail authority, and it has to sit behind voluntary demand response rather than in front of it. FERC also pointed to a voluntary non-firm transmission service, of the kind raised in its separate large-load show-cause proceeding, as “a potential avenue” to some of the same ends.[12](#fn12)

## What BYONC needs the second time

Two technical objections in the order matter most to anyone pricing a bring-your-own arrangement.

The first is deliverability. PJM already sets reliability requirements for Locational Deliverability Areas, meaning capacity must be inside a constrained area or deliverable into it. The BYONC proposal ignored that. A new large load in one zone could have counted a new plant on the far side of a transmission constraint and escaped IRAS anyway, which FERC said “could exacerbate the resource adequacy shortfall in the New Large Load’s zone.” The commission suggested PJM consider conditioning BYONC on a finding that the new resource is deliverable to the area where the load sits.[13](#fn13)

For a utility this is a reasonable ask, and probably the one most likely to reappear. For a developer it changes the economics. A cheap gas plant two hundred miles away may no longer count; a more expensive one inside the zone, or storage on the right side of the constraint, might. Deals already signed on the assumption of region-wide credit will need a second look.

The second objection is accreditation. PJM measures each resource’s contribution to reliability through Effective Load Carrying Capability, or ELCC, and those ratings change as the fleet changes. PJM proposed to lock a BYONC resource’s accredited value for the life of the arrangement, which gives the data center certainty. FERC found that PJM had not shown the benefit outweighs the risk that locked values drift above what the resources can actually deliver, creating “a new resource adequacy risk for the entire PJM Region.” It noted that PJM’s own long-term forecast shows ELCC ratings falling for every resource type.[14](#fn14)

There is a deal here, but it is not a lock. One workable design would accredit BYONC capacity the way PJM accredits everything else, a principle PJM’s own November response matrix repeats for generators selling into the capacity market, and then let the contract between utility and customer allocate the risk of later derating.[15](#fn15) A developer could overbuild against a cushion, buy replacement capacity, or accept a step-down schedule. That keeps the market whole and moves the risk to the parties who chose to take it.

FERC raised smaller points that still bite. A 49 MW customer that adds a single megawatt would become a 50 MW New Large Load with obligations on the entire 50 MW, and the proposal did not account for existing behind-the-meter netting rules or for commercial deals already signed with load ramping after June 1, 2027.[16](#fn16)

## The states now hold the data

The order hands the states two things at once. One is information. Registry data goes to state commissions, transmission owners, electric distributors, load-serving entities and PJM’s market monitor under existing confidentiality rules, with only aggregates made public. State consumer advocates asked for access and did not get it.[17](#fn17) PJM says it will also use the registry to feed a large-load dashboard on its website.[18](#fn18)

The other is a job. FERC wrote that states “have a crucial role to play” in “designating which retail customers must be curtailed when the system is stressed,” and it “strongly” encouraged them to keep at it.[19](#fn19) Mary Yang, a former PJM counsel, put the outcome as a setback for bring-your-own deals but not a dealbreaker, with the real momentum now in state capitals.[20](#fn20) That seems right. A Virginia or Pennsylvania commission that can see every 50 MW load in its territory is better placed than PJM to write a large-load tariff, a curtailment priority list or a self-supply option that a federal regulator would not touch.

The arithmetic underneath has not moved. PJM’s 2028/2029 auction cleared about 6,831 MW short of its regional reliability requirement at the price cap, according to PJM’s filing as summarized by The Duck Curve.[21](#fn21)

## What to ask for on November 16

PJM’s matrix describes several Section 205 filings around November 16, including new firm and non-firm contract demand transmission services for large loads and an expanded interim network service, with the “precise number and structure” still subject to change.[22](#fn22) Anyone negotiating a large-load supply deal in PJM should read them with a short list of questions in hand.

Does any curtailment obligation run through electric distributors and state retail authority, in words a commission could not mistake for a direct federal order to a retail customer? Does it wait until voluntary demand response has been used, and stay consistent with EOP-011-4? Is self-supplied capacity credited only where it can be delivered to the load’s area? Is that capacity accredited under ordinary ELCC, with any protection against derating left to contracts rather than written into the market? Are existing agreements and small incremental additions given a clean transition? And is there a voluntary non-firm path for loads that would rather accept interruption than wait for capacity?

A filing that answers those questions would give the utility a way to serve new load without exporting risk to its other customers, and give the developer something a lender can underwrite. A filing that does not will probably meet the same fate as the first one, and the meeting at the table will start over.

## Notes

## Bibliography

- Data on Data Centers. “FERC Rejected PJM’s Data Center Curtailment Plan. What’s Next for BYOC?” October 13, 2026\. [https://dataondatacenters.com/resource/ferc-rejected-pjm-s-data-center-curtailment-plan-what-s-next-for-byoc-2](https://dataondatacenters.com/resource/ferc-rejected-pjm-s-data-center-curtailment-plan-what-s-next-for-byoc-2?ref=aixenergy.io).
- Duck Curve, The. “FERC Rejects PJM’s Large-Load Curtailment Plan, Keeps the Load Registry.” October 10, 2026, updated November 3, 2026\. [https://theduckcurve.com/posts/ferc-rejects-pjm-iras-keeps-large-load-registry](https://theduckcurve.com/posts/ferc-rejects-pjm-iras-keeps-large-load-registry?ref=aixenergy.io).
- Federal Energy Regulatory Commission. *PJM Interconnection, L.L.C.*, Order Accepting Tariff Revisions in Part and Rejecting Tariff Revisions in Part. 197 FERC ¶ 61,019\. Docket No. ER26-3515-000\. October 9, 2026\. [https://www.pjm.com/pjmfiles/directory/etariff/FercOrders/9207/20261009-er26-3515-000.pdf](https://www.pjm.com/pjmfiles/directory/etariff/FercOrders/9207/20261009-er26-3515-000.pdf?ref=aixenergy.io).
- PJM Interconnection. “PJM’s Response Matrix to the Large Load Show Cause Order Opportunities.” Members Committee special session, November 5, 2026\. [https://www.pjm.com/-/media/DotCom/committees-groups/committees/mc/2026/20261105-special/item-01---2-large-load-compliance-matrix.pdf](https://www.pjm.com/-/media/DotCom/committees-groups/committees/mc/2026/20261105-special/item-01---2-large-load-compliance-matrix.pdf?ref=aixenergy.io).

---

1. Federal Energy Regulatory Commission, *PJM Interconnection, L.L.C.*, Order Accepting Tariff Revisions in Part and Rejecting Tariff Revisions in Part, 197 FERC ¶ 61,019, Docket No. ER26-3515-000 (October 9, 2026), P 1 and P 3\. PJM states it will be short of capacity to meet the PJM Region Reliability Requirement beginning in the 2027/2028 Delivery Year.[↩︎](#fnref1)
2. PJM Interconnection, “PJM’s Response Matrix to the Large Load Show Cause Order Opportunities,” prepared for public posting October 29, 2026, posted with Members Committee special session materials, November 5, 2026, 1, 3–9\. The matrix responds to FERC’s June 18 large-load show-cause order; it does not state that a revised IRAS or BYONC proposal will be included.[↩︎](#fnref2)
3. FERC, Order, 197 FERC ¶ 61,019, P 1, P 2, P 13.[↩︎](#fnref3)
4. Ibid., P 15.[↩︎](#fnref4)
5. Ibid., P 48.[↩︎](#fnref5)
6. Ibid., P 27; The Duck Curve, “FERC Rejects PJM’s Large-Load Curtailment Plan, Keeps the Load Registry,” October 10, 2026, updated November 3, 2026.[↩︎](#fnref6)
7. FERC, Order, P 12.[↩︎](#fnref7)
8. Ibid., P 77.[↩︎](#fnref8)
9. Ibid., P 81 (page 36).[↩︎](#fnref9)
10. Ibid., P 107; Rosner concurrence, n. 1 (IRAS triggered at EEA 1).[↩︎](#fnref10)
11. Ibid., Rosner, Commissioner, concurring, PP 1–5.[↩︎](#fnref11)
12. Ibid., P 111.[↩︎](#fnref12)
13. Ibid., P 109.[↩︎](#fnref13)
14. Ibid., P 105 and n. 245.[↩︎](#fnref14)
15. PJM, Response Matrix, 11 (capacity sold into the market “will be accredited as any other capacity resource”).[↩︎](#fnref15)
16. FERC, Order, P 110.[↩︎](#fnref16)
17. Ibid., P 49.[↩︎](#fnref17)
18. PJM, Response Matrix, 7.[↩︎](#fnref18)
19. FERC, Order, P 95.[↩︎](#fnref19)
20. Data on Data Centers, “FERC Rejected PJM’s Data Center Curtailment Plan. What’s Next for BYOC?,” October 13, 2026\. Paraphrase of the published summary.[↩︎](#fnref20)
21. The Duck Curve, “FERC Rejects PJM’s Large-Load Curtailment Plan.” Figure attributed there to PJM’s ER26-3515 filing; not stated in the FERC order.[↩︎](#fnref21)
22. PJM Interconnection, “PJM’s Response Matrix to the Large Load Show Cause Order Opportunities,” prepared for public posting October 29, 2026, posted with Members Committee special session materials, November 5, 2026, 1, 3–9\. The matrix responds to FERC’s June 18 large-load show-cause order; it does not state that a revised IRAS or BYONC proposal will be included.[↩︎](#fnref22)