Regulation — energy-specific

FERC Order 890 — transmission-service transparency + comparability

Order 890 (2007) reformed the pro forma Open Access Transmission Tariff to require greater transparency and comparability in transmission-service decisions. When those decisions are AI-influenced, the transparency obligation attaches to the AI's inputs, model version, and output — not just to the human-visible dispatch signal.

What the order actually says

"… the Commission requires transmission providers to calculate Available Transfer Capability … using a consistent methodology … and to post the methodology, inputs, and results on the OASIS … to ensure that transmission service is provided on a not-unduly-discriminatory basis." FERC Order 890 — Available Transfer Capability + OASIS posting requirements (paraphrased)

"… transmission planning must be conducted in an open, transparent, and coordinated manner … with sufficient documentation of assumptions, inputs, and results to permit meaningful stakeholder participation and review." Order 890 — transmission planning transparency (paraphrased)

What this means in plain English

Two obligations that AI-driven grid-load-forecasting directly bears on:

Methodology transparency. If AI/ML is embedded in the ATC calculation or transmission planning process, the model, inputs, and outputs are part of the methodology that has to be documented and made reviewable.
Not-unduly-discriminatory. If the AI systematically under-serves a market zone (or one class of transmission customer), the discrimination bar is lowered dramatically for a FERC complaint.

What triggers the exposure in the sample

Zone E3 (high-renewable-penetration region) forecast intervals under-routed to emergency-reserve for 45+ days. If the operator's transmission planning or ATC methodology relies on the AI grid-load-forecaster's output, Order 890 transparency attaches to the model + inputs + output for the affected forecast intervals. Without model-version pin + input snapshot + decision hash, the transparency obligation is not being met defensibly.

What the $499 Snapshot shows against this rule

Model version pinned per forecast interval — methodology-transparency record
Per-zone distributional-shape analysis — the not-unduly-discriminatory evidence
Retention pipeline distinct from production model — the reviewable-documentation record
Signed independent-verifier declaration — the record for FERC stakeholder-review response

See the signed independent-verifier declaration →

How does this help me?

Order 890 stakeholder complaints are a common entry point into FERC Section 206 investigations. Independent-verifier evidence changes the opening stance of any such matter.

Read: Order 890 -- what transparency evidence saves you if a stakeholder files a complaint →

$499 Snapshot. 3 business days.

Methodology + inputs + outputs retention evidence for your operator's AI grid-load-forecaster.

Buy $499

Snapshot credit applies to Baseline ($2,500) or Enterprise Attestation ($35-55K) upgrade within 30 days.