Forecast ObjectPS-0003

PJM AEP-zone large-load forecast cohort

This is a screened PJM forecast object, not a project state or realized-load record.

Forecast — no project C-stateAggregate forecast cohort

Executive evidence ceiling

What a decision maker may safely take from this record

Permitted reading

Forecast method and vintage only; no project state, deliverable capacity or realization rate.

Not established

Any implied project progress or conversion-chain position.

Owner attention

Project-level mapping and realization

Geography
Ohio/PJM AEP zone
Object subtype
Regional Load Forecast
AI connection
Data-center/large-load category; AI share not established
Recorded figure
See 2026 forecast and FERC Figure 7
Unit basis
Forecast MW by year/zone
Countable operating capacity
NO

Conversion-chain position

C0–C21, without implied progression

Handling modeNO PROJECT STATE

Explicitly represented in this record Not established by this record

  1. C0–C6

    Entry and permission

    1. C0 Concept or rumor
    2. C1 Public announcement
    3. C2 Site identified
    4. C3 Land option or negotiation
    5. C4 Land controlled
    6. C5 Zoning or land-use application
    7. C6 Zoning or land-use approval

  2. C7–C10

    Service-model fork

    1. C7 Utility service or interconnection request
    2. C8 Study or review accepted
    3. C9 Tariff or service pathway identified
    4. C10 Agreement executed or obligation recorded

  3. C11–C17

    Build and operation

    1. C11 Network, generation, or facility upgrades funded
    2. C12 Construction started
    3. C13 Major electrical and cooling equipment delivered
    4. C14 Grid or onsite supply energized
    5. C15 Facility commissioned
    6. C16 IT capacity installed
    7. C17 Compute cluster operating

  4. C18–C20

    Dark Zone

    1. C18 Contracted customer use observed
    2. C19 Utilization or revenue effect observed
    3. C20 Sustained economic compute capacity

  5. C21

    Changed status

    1. C21 Delayed, cancelled, downsized, superseded, or repurposed

Recorded position: Forecast — no project C-state

PJM screened forecast adjustment

No numbered C-state is assigned by this record. This preserves a forecast, framework, mixed, unknown or otherwise non-project boundary rather than forcing a false position.

State boundary

Scope-specific assertions

Forecast — no project C-stateAggregate forecast cohort

No forced project state

PJM screened forecast adjustment

Control rule

Any implied project progress or conversion-chain position.

Four-register control

Figures stay in the register that produced them

No cross-register conversion is performed without an explicit source-stated bridge.

  1. R01Load

    Not normalized

    See 2026 forecast and FERC Figure 7 is retained as a non-normalized or non-load figure; no MW conversion is performed.

    No load quantity is imputed from money, supply, accelerator count or project state.

  2. R02Supply

    No bridge inferred

    No supply-register quantity is normalized for this record.

    Project state is not treated as proof of deliverable, firm or operating supply.

  3. R03Money

    No bridge inferred

    No separate money-register metric is normalized for this record.

    Capex, contract value, backlog, RPO, revenue and utilization are not converted into capacity or operating state.

  4. R04AI-share

    Not established

    Data-center/large-load category; AI share not established

    Total data-center, campus or portfolio measures remain broader than AI unless the source provides an explicit AI-share boundary.

Aggregation policy: Not summable in this edition. No total, rank or realization percentage is generated.

AI-share boundary

Broader infrastructure does not become AI-only by implication

Data-center/large-load category; AI share not established

01

Broader measured scope

Campus, data-center, portfolio, utility, accelerator or financial quantity.

Explicit source bridge

Named AI workload or customer scope · denominator · period · allocation method.

02

AI-attributable subset

Reportable only to the extent directly defined by the source.

AI-share not established

Without the bridge, retain the broader label and stop the attribution.

Control rule: GPU count, company narrative, total campus MW or portfolio capex does not supply an AI-share denominator on its own.

Atomic evidence stack

Claims retain their own ceiling, locator and sources

6 linked claims at the current cutoff.

CLM-0016 · forecast

PJM’s 2026 load forecast process screens large-load adjustment requests rather than automatically including the full requested amount.

Claim ceiling
Forecast-inclusion method, not validation of each included project.
Evidence
Tier A · HIGH
Locator
SRC-0029 methodology; SRC-0030 criteria; SRC-0031 slides 30–31

CLM-0017 · interpretation

Official 2026 records show that large-load forecasts can be materially revised downward after updated screening or assumptions, supporting the forecast-inflation rival as a live hypothesis.

Claim ceiling
Forecast-revision mechanism; no claim that all pipelines are inflated.
Evidence
Tier A · HIGH that revisions occur; MEDIUM on broader generalization
Locator
SRC-0012 pp. 9–10, Figure 7; PJM forecast and accuracy materials

CLM-0022 · implementation

Public records identify flexibility and conditional service as possible adaptation mechanisms, but current evidence does not justify treating all AI data-center load as flexible.

Claim ceiling
Mechanism availability/hypothesis; no universal technical capability or customer commitment.
Evidence
Tier A · MEDIUM
Locator
FERC State of Markets Figure 8; SPP ER26-1323; Virginia conference/testimony

CLM-0025 · interpretation

Available national and regional public records use incompatible vintages, geographic boundaries, and state definitions; quantitative consolidation requires raw-value preservation and explicit normalization.

Claim ceiling
Methodological conclusion; no merged national project total in Pass 01.
Evidence
Tier A · HIGH
Locator
Methods and vintages identified in each source

CLM-0032 · interpretation

FERC staff reported more than 50 GW of U.S. data-center capacity in service at the end of 2025 and a 24% CAGR since 2020, but the public report relies on a proprietary Yes Energy database whose project-level state definitions are not fully auditable in the public record.

Claim ceiling
External source-defined benchmark; not additive to the Pass 02 project ledger and not a C14–C20 audit.
Evidence
Tier A · HIGH that FERC reported the values; MEDIUM-LOW for state reconciliation
Locator
PDF p. 9 (report p. 7), Figure 5 and note 14
Linked sources
SRC-0012

CLM-0033 · interpretation

TWh/year, equivalent average GW, source-defined facility capacity, system peak GW, requested load, design capacity, MW-IT, GPU count, and announced capex are non-interchangeable metrics and are held in separate comparable groups.

Claim ceiling
Normalization rule; no conversion beyond explicit formula and assumptions.
Evidence
Tier A + CONTROLLING INPUT · HIGH
Locator
Operating Prompt §§7–8; Seed Charter §§3,6; quantitative source locators in Pass 02 ledgers
Linked sources
SRC-0001internal recordSRC-0002internal recordSRC-0023SRC-0012SRC-0018

Record stewardship

What is known, missing and capable of changing the record

Contrary evidence4 linked tests

Review rival propositions before strengthening the claim.

Negative findings4 recorded gaps

These searches bound what the public record did not establish.

Reopen triggers4 active rules

A qualifying evidence delta requires review and rebuild.

Contrary evidence and rival tests

CTR-0001 · Forecast inflation

Requested and announced pipelines can materially exceed screened forecasts or realized load.

Evidence result:PJM differentiates utility demand asks from preliminary/final included adjustments; FERC reported a 15% lower 2032 AEP-zone forecast versus the prior vintage after large-load adjustments.

Ceiling effect:No pipeline-to-realization conversion; preserve vintage and screening.

CTR-0003 · Flexible-load response

Some computational loads may curtail, shift, or accept conditional service, reducing some system constraints.

Evidence result:FERC identifies flexible crypto-mining behavior as illustrative; SPP filed conditional HILL service; Virginia convened a flexibility record. Transferability to AI loads is unproven.

Ceiling effect:Possible mechanism only; not universal capability or solved reliability risk.

CTR-0008 · Policy and data uncertainty

Evolving rules, changing forecast vintages, and inconsistent definitions may be a principal planning and evidentiary cost.

Evidence result:FERC, NERC, PJM, ERCOT, and state records are contemporaneously changing and use different units/states.

Ceiling effect:Every current claim requires cutoff and reopen trigger.

CTR-0013 · Registry and methodology opacity

A source-defined national in-service estimate may not reconcile to a public C0–C20 project audit.

Evidence result:FERC’s >50 GW value rests on proprietary Yes Energy data; no unified official project-state registry was located.

Ceiling effect:Benchmark cannot be added to or used to validate every project row.

Public-record gaps

NEG-0001

Question checked:Is there a unified official U.S. registry of AI/data-center projects with state-normalized capacity?

UNKNOWN — PUBLIC EVIDENCE NOT LOCATED; no unified registry identified.

Implication:Build project crosswalk; no national pipeline total.

NEG-0006

Question checked:Is there a standardized national public record for commissioning and utilization?

UNKNOWN — PUBLIC EVIDENCE NOT LOCATED.

Implication:Require project-specific C15–C20 evidence.

NEG-0008

Question checked:Can computational-load flexibility be generalized from crypto mining to AI data centers?

No sufficient public evidence located for a universal generalization.

Implication:Case-specific, measured claim only.

What could change this record

  • OIR-0008 · New official demand forecast, project pipeline, operational-data accuracy report, or material project cancellation/energization is published.

    Required update: Add vintage rather than overwrite; reconcile project IDs; update state and double-counting register.

  • OIR-0009 · A public field demonstration or binding service agreement provides measured AI/computational-load flexibility evidence.

    Required update: Capture workload, baseline, response speed/duration, customer constraints, sample, and transferability limits.

  • OIR-0017 · A new official national or regional demand forecast, AEO/LBNL update, or method revision changes scope, horizon, scenario range, or unit bridge.

    Required update: Add new vintage without deleting prior; reconcile scope, method, denominator and supersession; rerun normalization.

  • OIR-0018 · A reproducible public methodology or project-level dataset becomes available for the FERC/Yes Energy in-service capacity series.

    Required update: Inspect definitions, coverage, duplicates and state thresholds; reconcile with C0–C20 and revise benchmark ceiling.

Source register

Trace the public record

3 linked sources; rights and locators remain attached.

Source records

  1. SRC-0012 — 2025 State of the Markets Report

    Federal Energy Regulatory Commission, Office of Enforcement · 2026-03-26 · Tier A

    Locator: PDF pp. 9–10 (report pp. 7–8), Figure 5 and note 14; other market-review figures as separately cited

    PUBLIC OFFICIAL RECORD — CITE/SUMMARIZE

  2. SRC-0029 — 2026 PJM Load Forecast Report

    PJM Interconnection · 2026-01-14 · Tier A

    Locator: Executive summary; large-load adjustment methodology; zone tables

    PUBLIC OFFICIAL RECORD — CITE/SUMMARIZE

  3. SRC-0032 — Load Forecast Accuracy Report — May 2026

    PJM Interconnection · 2026-05-29 · Tier A

    Locator: Overview p. 3; operational-data comparisons

    PUBLIC OFFICIAL RECORD — CITE/SUMMARIZE