How we did this

Methodology

Three rankings, three federal-or-SEC sources, no guesswork. Here's exactly where each number comes from, what we did to it, and — just as important — what it doesn't prove. If we got something wrong, tell us and we'll fix it in public.

Reliability — who keeps the lights on

Source: U.S. EIA Form EIA-861, Reliability file, 2024. What we use: SAIDI (average minutes a customer is without power per year) and SAIFI (average number of interruptions).

• We rank on the storms-excluded ("without major event days") SAIDI — the normal-day operations a utility actually controls. We show the including-storms figure beside it, but don't rank on it, so a utility isn't branded "worst" for sitting in a hurricane corridor.
• We flag each utility's reporting standard (IEEE 1366 vs. "Other"). "Other"-standard figures may not be perfectly comparable to IEEE ones; treat close gaps across standards as approximate, not exact ranks.
• FirstEnergy reports four former Pennsylvania utilities (Met-Ed, Penelec, Penn Power, West Penn) as one combined company, so they appear as a single, customer-weighted row.
• It's a single year (2024). One bad storm season can move a utility; we'd rather show a snapshot honestly than imply a permanent verdict.

Disconnections — who shuts off the most customers

Source: U.S. EIA Form EIA-112, Residential Utility Disconnections Survey, 2024 — the first year this was collected. What we use: residential disconnections for nonpayment.

• We rank on the rate per 100 residential customers, not the raw count — raw counts mostly just track which utility is biggest. We show the raw household count too, because each one is a real family.
• The rate is confounded by things outside a utility's control: local poverty and energy burden, climate, the strength of a state's winter-moratorium and medical-protection rules, and assistance-program funding. A high rate is a real outcome, not proof of intent — and it doesn't capture payment plans or how fast people were reconnected.
• Two utilities we track are named but not ranked: PECO Energy (reported only 6 of 12 months — we don't annualize a partial year) and Rockland Electric/RECO (didn't report). We don't drop them silently.
• 2024 is the survey's first year, so completeness is still maturing.

Profit — where your bill money goes

Source: SEC Form 10-K filings, fiscal year 2024 (net income attributable to the parent), via EDGAR.

• Net income is reported only at the parent-company level. A parent owns many utilities across several states, so we never present a parent's profit as a single utility's. We list which of the utilities we track roll up to each parent.
• The authorized return on equity (~9.5–10.5%) is a regulator-set target on the utility's grid investment — not a guarantee, and not the same as net income. Utilities can earn above or below it; we describe it as an industry norm, never as a specific utility's number.
• We do not rank individual utilities by profit, because the data to do that fairly doesn't exist publicly.

What these rankings don't prove

They rank reported outcomes, not effort or intent. A utility in a storm-prone or low-income service area can run a tight operation and still land near the bottom. We rank what happened to customers and cite exactly where each number came from — we don't tell you who to be angry at.

Dates & reuse

"Last reviewed" (or "verified") next to a figure means we checked it against the cited source on that date. It's a freshness marker, not a guarantee the number is still current — rates and cases change, so confirm with the utility or commission before acting.

The underlying figures come from public-domain U.S. government data (EIA, SEC EDGAR) and from utility and state filings — facts that aren't subject to copyright and are free for anyone to reuse. Our original compilation, charts, and written analysis are offered under CC BY 4.0 — reuse them with attribution to RateWatchdog.