PCLR Origin Story Part 2: Speed to Power for AI Loads in ERCOT

How to get ERCOT comfortable with "managing" controllable load in exchange for earlier "connection" - NPRR1188 as PGRR134 Foundation

Update - June 18, 2026: The Public Utility Commission of Texas approved the provisional controllable load resources solution as part of PGRR145 and accompanying NPRR1325, which broadly establish the transitional "Batch Zero Process," which replaces individual high-load studies with a coordinated, system-wide batch evaluation framework. PGRR134 was withdrawn in May 2026 in light of the market construct getting adopted into formal market rules. The withdrawal comments from Luminary Strategies provide a history of the proceeding.

---

Original Article:

• PGRR134 was filed as “Interconnection Studies Reform for Dispatchable Loads.”

• NPRR1188 is PGRR134’s implementation vehicle. It was passed in Nov. 2024 as “Implement Nodal Dispatch and Energy Settlement for Controllable Load Resources.”

On Monday March 30, 2026, ERCOT will supply public copies of the controllable load resource framework in which both documents will be addressed; formal rule proposals will be available April 8, 2026.

PGRR134 was filed as “Interconnection Studies Reform for Dispatchable Loads.”

The nice part is, hardly any reform is happening to the interconnection studies themselves - those reforms, are actually the purview of standards and rules set by the Public Utility Commission of Texas, for the utilities and for ERCOT, and they are now the “batch load” rules being set up under ERCOT PGRR145 and PUCT Project 58481 et seq. In the few redlines on PGRR134 that Luminary Strategies, LLC filed subsequent to the original posting, we have made this aspect easier to grasp.

• Original Filing: https://www.ercot.com/files/docs/2025/11/01/134PGRR-01-Interconnection-Studies-Reform-for-Dispatachable-Loads-110125.docx

• Original Process Mechanics: https://www.ercot.com/files/docs/2025/11/03/134PGRR-03-Agentic-Infrastructure-Comments-110325.docx

• Revised Filing: https://www.ercot.com/files/docs/2025/11/14/134PGRR-09-Joint-Commenters-Comments-111425.docx

• Bonus - No Doom Loops for Speed to Power, Please: https://www.ercot.com/files/docs/2025/11/17/134PGRR-11-Luminary-Strategies-Comments-111725.docx

What is happening, is reform to the order of process operations entailing an energization “go” date for commissioned load on a piece of real estate in Texas, prior to the date that the full firm service date allocated to the complete expected load ramp. The “connect and manage” aesthetic is preserved, such that managing constraints that bind on the load (which is itself dependent on the algorithm mechanics and heuristics utilized by ERCOT) in this setup are fully on the load - not on ERCOT to solve for. The “firm planning” and financial pay-your-way discipline is also preserved: loads seeking to energize on an as-available basis are still being studied as firm loads, and the rules of load commissioning apply equally.

There is plenty of system mechanics reform to get through, however. In other words, there is still the implementation of NPRR1188 to get through. Though NPRR1188 is a fully baked blueprint for ERCOT to accept Controllable Load Resources dispatched and priced (Nodal-CLRs) so they are truly under the grid operator’s control (bankable “management” in exchange for earlier “connect”), it means ERCOT still needs to implement market operations updates such that Nodal-CLRs are dispatched at a nodal shift factor and settled for their energy consumption at a nodal price. There are system changes that need to happen; the ERCOT Impact Analysis defines the impact percent of the rule changes across systems:

· Market Operation Systems 63%

· Settlements & Billing Systems 22%

· Credit 6%

· Energy Management Systems 2%

· Data Management & Analytic Systems 2%

· Credit Management Systems (CMM) 1%

· Resource Integration and Ongoing Operations (RIOO) 1%

· Integration Systems 1%

· Channel Management Systems 1%

· ERCOT Website and MIS Systems 1%

· EPS Metering 1%

In explanatory comments Luminary Strategies, LLC wrote for the October Large Load Working Group meeting, I wrote:

“NPRR1188 is a successful policy change that moves away from pricing what can and will be gigawatt-sized ERCOT loads from a prior zonal construct which used an inefficient, and generally problematic approach of apply zonal shift factors and zonal pricing to loads in SCED. Those attributes did not correctly price incentives in the market for withdrawal levels of a CLR and zonal dispatch factors misrepresented the impact of the withdrawal levels of these Resources on transmission constraints. In the case of very large loads, the impact of zonal shift factors and pricing would be egregious for the system: distorting the correct allocation of congestion costs and system operations which could be effectuated by small movements in these large loads’ consumption behavior at individual pricing nodes on the system.”

Well yes. Even if you understood little of the prior paragraph, you would understand the idea of gargantuan loads moving around on the system on a zonal basis, when in fact, each is exepcted to be so huge, it is a “zone” into and unto itself. Nodal dispatch (location-specific) for loads has to exist for the simple sake of safety and reliability in managing load-side operations and balancing the grid every day.

I further wrote, that “NPRR1188 effectuates critical operations changes that ERCOT must fully implement in advance of operating large loads which elect to establish voluntary participation in ERCOT SCED, managing obligations as a CLR through their Qualified Scheduling Entity’s compliance with ERCOT Resource requirements. Interconnecting large load entities (ILLEs) which choose this status must be assigned a Resource Node Settlement Point and must comply with SCED instructions such that OUTL status can only be telemetered to ERCOT if the CLR “is truly outaged and is consuming zero MWh.” Now that ERCOT has created an incentive for more loads to contribute to the system’s reliability management, the incentive for loads to participate as CLRs must come full circle to afford the market to capitalize on the benefits of this solution for congestion management, system operations transparency, reliability management, and by necessity to effectuate these outcomes, large load energization studies.” And this full circle, is PGRR134.

Beginning March 30, 2026, stakeholders should expect progress in several areas that the PGRR134 proceeding was unable to address in October-November 2025. This delay stemmed from ERCOT’s December 2025 directive to redesign the entire large load study process into a new load batching framework. Now, in the context of the batch load study PGRR145, and the CLR design-specific materials ERCOT will soon post, the work done to date by all involved parties will begin to shine through. In the appendix to this piece, check out my own running list of issues we’ve worked on, in the quiet candlelight hours of December 2025 to March 2026.

The explanatory paper of September 2025 makes particular effort to expand on the inclusive and universal nature of Loads in SCED as a nodal dispatch and settlement solution for ERCOT, long overdue. (Ask the right ERCOTian, and he or she will rightfully complain that “we should have done this a decade ago.”) In this paper, I write:

“Private solutions may include, but are not limited to, onsite gas turbines, various forms of energy storage, geothermal power, dispatachable IT and balance-of-plant infrastructure, compute workload, and software-enabled load-shifting protocols. By its nature, the proposed solution contemplates a technology-neutral approach inviting CLRs to voluntarily participate in grid-supportive energization solutions. Importantly, the approach is not about compelling loads themselves to participate as non-firm: rather, the solution creates private firming incentives which can be reflected in technology stacks, co-location PPAs, and innovative structure in QSE portfolios to privatize firming risk in exchange for earlier load energization and resolution of TSP queue constraints which are a detriment to all loads (of any size) seeking firm interconnection. In this approach, the dispatachability of the load itself behind a CLR election remains a “dispatch of last resort.”

On Monday March 30, flex load fans should expect ERCOT materials to include a first cut at a “just in time” approach to mitigating constraints on nodal-CLRs. Explaining the universe (and I do mean, to everyone who asks me the wrong question, which goes something like “how many exact curtailment hours will each load be responsible for”) that a connect-and-manage load solution will be a "just in time and last in line" approach for bid-to-buy mitigation, will get a lot easier for me. Thank goodness.

Relatedly, the fun of NPRR1188 implementation: the confirmed go-live implementation target for NPRR1188 is Q4 2026. So I guess, we could have a “Nodal CLRs” RIOO moat for ERCOT to set up variables and assumptions before any of the + 75 MW large loads sign an interim FEA committing to energize as a Nodal CLR.

• NPRR1188 was packaged together with NPRR1244 (Clarification of Controllable Load Resource Primary Frequency Response Responsibilities) into a single master project internally dubbed “Controllable Load Enhancements.”

• ERCOT officially started the project labor and impact analysis work in December 2025.

• The Finish Line: The project is currently active and “in flight” with a firm system implementation target of Q4 2026/Q1 2027.

A.S.F.

Candlelight Hours: Nodal CLRs To-Do List

1) Process / sequencing

• Need for a defined contractually binding process (facilitated by Load electing CLR status for energization) prior to receipt of approval to energize

• When a CLR can begin consuming vs. planned/chosen behavior to remain CLR or not, after full firm upgrade completion

• What happens if firm service = zero at initial energization of nodal CLR (needing a numerical offset for that load to show “load is consuming above LPC” and mark offsets against that number as a response in SCED).

2) Qualification + commissioning

• Establish CLR qualification timeline and requirements

• Need for telemetry, testing, and performance validation (e.g., curtailment tests)

3) Planning vs. operations boundary

• Confirm that CLRs are treated as firm in long-term planning vs. non-firm operationally

• How to handle loads exiting CLR status

4) Operational dispatch & control

• How SCED interacts with CLRs:

  • Dispatch obligations beyond LLIS constraints (future constraints too)

  • What happens if CLR fails to resolve a constraint → curtailment?

• Use of:

  • VDIs vs. offer mitigation vs. automated controls (minimize any/all manual operator interventions)

  • Disconnect switches / breaker schemes for enforceability (already happening for SB 6 compliance prior to a load energization)

5) Risk allocation

• Explicitly making curtailment a private risk borne by the load

• Clarifying consequences if load cannot perform as CLR

7) Queue equity / market impacts

• Ensuring CLRs do not harm firm queue timelines

• Addressing perceived cost allocation risk

8) Technical constraints outside SCED gate Load energization

• Handling non-thermal constraints (voltage, stability) that CLR cannot solve (lift this from PGRR134 language)

• Ensuring required special projects still get built and utility clears their completion before energization

9) Definitions / eligibility

• What counts as a CLR vs.:

  • load + behind-the-meter resources

  • co-located generation / storage configurations

• Whether storage / co-located assets which are NOT commonly metered can be used for compliance

[For all the investment analysts reading this, sorry, bad news - all good load flex solutions are location-specific, power study-specific, and pre-planned system upgrade-specific.]