Search
January 21, 2026
Stakeholder Forum | Opinion
ERCOT Generation Netting Isn’t Yet Investment Grade for Renewable Firmed Data Centers
Where Protocol 10.3.2.3 Helps, Where it Stops and What ‘Netting Plus’ Needs to Standardize

Listen to this Story Listen to this story

© RTO Insider 
ERCOT is absorbing a wave of large, price‑sensitive load, especially data centers, faster than the market rules were built to 'productize,' writes Alexandre Alonso Carpintero in an opinion piece.

Alexandre Alonso Carpintero |

By Alexandre Alonso Carpintero

ERCOT is absorbing a wave of large, price‑sensitive load, especially data centers, faster than the market rules were built to “productize.”

ERCOT planning materials show about 226 GW of large loads seeking interconnection as of Nov. 18, 2025 (up from 63 GW in December 2024), with about 225 new large‑load requests submitted in 2025 and about 73% of the queue attributed to data centers. If the finance path for renewable‑firmed supply is uncertain, the default “under-writeable” answer becomes on‑site gas.

What Generation Netting Really is

Generation netting for ERCOT‑polled settlement (EPS) meters (Protocol 10.3.2.3) is a settlement boundary rule: Under specific electrical configurations and metering constraints, ERCOT may settle a paired generator and load on a net basis. The protocol is intentionally restrictive (“generation netting is not allowed except under” defined conditions) and depends on site topology (e.g., common switchyard concepts, EPS metering points and limits on alternate grid connections). Netting can reduce settled energy volume. It does not convert a complex behind‑the‑meter campus into a financeable product. (See Aurora Research Report.)

Why it Fails the ‘Investment‑grade’ Test

Credit committees don’t finance “net MWh.” They finance the residual risk stack, especially correlated tail risk. Even with netting, a renewable‑firmed data center typically retains:

    • Scarcity price tail on backup imports. ERCOT’s systemwide offer cap (HCAP/SWOC) remains $5,000/MWh (with a low cap framework that can apply under certain conditions).
    • Congestion/basis risk (nodal price separation between where supply is produced and where load settles).
    • Operational/curtailment risk: the usable “firming” value of renewables plus storage can degrade precisely when the grid is stressed (telemetry/dispatch constraints, emergency operating modes or required load shedding).
    • Administrative/process risk: eligibility, metering design and true‑ups can become bespoke legal/settlement work, hard to replicate across multiple campuses.

Residual risk stack after Generation Netting | Alexandre Alonso Carpintero

SB 6 Adds Layer of Uncertainty

Texas SB 6 (effective June 20, 2025) added PURA §39.169, requiring system‑impact review of certain net‑metering arrangements involving new large loads and stand‑alone generation. ERCOT’s market notice M‑B090225‑01 implements interim procedures, publishes a list of stand‑alone generation resources (as of Sept. 1, 2025), and states the process may change or be pre-empted by forthcoming Public Utility Commission of Texas rules.

ERCOT’s large‑load interconnection Q&A further notes that some arrangements involving existing “stand‑alone” resources require approval through the net metering review process before the load can be energized. “Interim and subject to change” is not bankable language when you’re trying to finance repeatable, gigawatt‑scale campuses.

A Simplified 300-MW Hybrid Example (Wind + Solar + BESS)

Assume a 300-MW flat-load campus behind EPS metering with 300 MW of wind plus 300 MW of solar plus a 100-MW, 400-MWh (four‑hour) battery. Netting can reduce settled imports across many hours. The financing problem is the tail.

Illustrative stress case: 20 scarcity hours per year when renewables are low and the battery is depleted or held for contingency. If the campus must import 100 MW during those hours and real‑time prices clear at $5,000/MWh, the annual cost is: 20 h × 100 MW × $5,000/MWh = $10 million.

That volatility is correlated with grid stress and uptime risk. The easiest way to cap both is a 300-MW on‑site gas plant, hence gas becoming the “insurance policy” for load growth.

What ‘Netting Plus’ Should Standardize

ERCOT does not need to copy another RTO. It needs standardized pathways that turn behind‑the‑meter engineering into predictable settlement plus performance rules:

    • Campus netting: standardized netting across a defined private network footprint (multiple meters/feeders under common control) with clear telemetry and true‑up rules.
    • Measurable firmness: a standardized add‑on (e.g., a performance obligation or ancillary‑service bundle) that lets large load pair renewables with qualifying firming (storage, fast response, contracted curtailment) and get settleable credit.
    • Clear hybrid “serve‑load‑first” rules: reduce ambiguity for storage charging/discharging, exports and when the site is acting as load vs generation.
    • Transparent backup settlement: make residual grid exposure bounded and hedge-able rather than a surprise.

Protocol 10.3.2.3 is a starting point. “Netting plus” is what makes renewable‑firmed data centers financeable at scale.

Alexandre Alonso Carpintero works on market design and commercial structures for large loads, including data centers.

CommentaryERCOT