NYSEG’s system-wide interruption frequency in 2025 was 1.29 — meaning the average customer on its system could expect about 1.29 sustained outages of five minutes or more during the year. That figure exceeded the company’s annual threshold of 1.20, the seventh consecutive year it has done so, and triggered a $3.5 million negative revenue adjustment paid by shareholders rather than ratepayers, according to the Department of Public Service’s 2025 Electric Reliability Performance Report filed in Case 26-E-0017.
That number is not a count of how often the lights flicker on a particular block near the Chenango. It is an average across a service territory that covers roughly 40 percent of New York State, and it is calculated under rules — storm exclusions, cause codes, division boundaries — that determine what gets counted and what does not. The practical task for a Binghamton resident, landlord, or neighborhood group is not to memorize the headline index. It is to learn which pages of which filings answer the question you actually have, and to ask for the division-level and circuit-level numbers that sit behind the corporate average.
Two metrics, two different questions
The state judges utility reliability primarily on two indices, defined in the report’s own terms. SAIFI — the System Average Interruption Frequency Index — is “the average number of times that a customer is interrupted for five minutes or more during a year.” CAIDI — the Customer Average Interruption Duration Index — is “the average interruption duration time in hours for those customers that experience an interruption during the year.”
Read those definitions closely and a gap appears. SAIFI is a per-customer average across an entire reporting area. If one feeder in a neighborhood goes out six times and the rest of the division goes out once, the division average can still look respectable. CAIDI only counts customers who were actually interrupted, so it tells you how long the lights were out for the people who lost them, not how many people lost them. Neither index, as published at the corporate level, tells you what happened on your street.
Next step: when you see a SAIFI or CAIDI figure quoted, ask what geography it covers. A corporate number, a division number, and a circuit number are three different claims.
The storm exclusion changes what the scorecard measures
The Commission’s reliability targets “exclude the impacts of outages caused by major storms,” the 2025 report states. Staff reviews data both including and excluding severe weather, on the reasoning that some interruptions are within a utility’s control — equipment failures, tree contacts — and some are not. The report is explicit that the exclusion is a balance, not a claim that storms do not matter to customers.
This is where a resident’s memory and a utility’s scorecard can diverge without either being wrong. A year with fewer qualifying storms can produce a better-looking target metric while the storms that do arrive still dominate what people remember. The 2025 report notes that the state experienced 25 major storm events that year, 17 fewer than in 2024, and that customers affected by major storms fell about 35 percent.
Next step: when a filing gives both including-storm and excluding-storm figures, read both. The excluding-storm number is the one tied to penalties; the including-storm number is closer to the customer experience.
What the 2025 filing says about NYSEG
Beyond the frequency miss, the report identifies the dominant cause. “Tree contacts continued to be the single largest contributor to system interruptions for NYSEG, accounting for 48 percent of interruptions and affecting approximately 580,000 customers.” Staff attributes part of the pressure to unreclaimed circuit miles encroaching on overhead lines and to the need to remove hazard trees outside the right-of-way.
The regulatory response described in the filing is a vegetation-management program. As part of the 2023 NYSEG Rate Order, the Commission authorized an expanded distribution vegetation management budget to continue the company’s Reclamation and Danger Tree Programs and to perform system-wide routine trimming on a six-year cycle. NYSEG has been providing monthly progress updates since that order. The report is candid that reliability “continues to suffer” despite those efforts, and Staff says it will keep monitoring vegetation progress and key capital expenditures.
What the filing does not establish is completion. It describes an authorized program and a monitoring regime, not verified results on any particular circuit. That distinction matters if you are trying to argue that a specific line has or has not been trimmed.
Next step: if tree contact is the stated cause of most NYSEG interruptions, the productive question for a neighborhood group is not “why did we lose power” but “where does our circuit sit in the six-year trim cycle, and when was it last reclaimed?” That is a question the company’s monthly vegetation updates and annual reliability report are designed to answer.
The baseline: this is not a new metric
NYSEG’s annual reliability reports show the same indices tracked over time, and they show that missing the target is a recent pattern rather than a permanent condition. The company’s 2010 Annual Reliability Report, filed under Case 02-E-1240, states that NYSEG met its PSC CAIDI and SAIFI targets that year, with targets of 2.08 and 1.20. The 2012 report likewise states the company met its targets, with corporate SAIFI of 0.98 and CAIDI of 2.00 excluding major storms, against the same 1.20 and 2.08 thresholds.
Two cautions before anyone cites those numbers in a meeting. First, the 1.20 threshold has stayed constant while the reported result has moved from 0.98 in 2012 to 1.29 in 2025 — but the territory, the reporting rules, and the underlying system have all changed over thirteen years, so a direct comparison is suggestive, not conclusive. Second, the older reports are corporate summaries; they do not by themselves prove anything about Binghamton specifically.
Cause codes are where policy becomes arithmetic
Utilities group interruption data into ten cause-code categories, defined in regulation, that delineate why the power went out: major storms, tree contacts, overloads, operating errors, equipment failures, accidents, prearranged interruptions, customer equipment, lightning, and unknown. The 2025 report explains the logic plainly: if a circuit is prone to lightning-caused interruptions, the utility could install arrestors; the cause data is what directs capital and maintenance.
Cause attribution is therefore not bookkeeping. An interruption coded as a tree contact counts against the vegetation program and feeds the frequency metric. One coded as a major storm is excluded from the target calculation. The report notes that utilities must submit detailed monthly interruption data and that Staff maintains the database back to 1989. It also notes that cause codes can be contested: RG&E, a sister Avangrid utility, filed a petition in February 2026 asking the Commission to exclude certain 2025 outages from its SAIFI calculation, citing causes including ash tree decay and emerald ash borer infestation. The report declines to address the merits of that petition.
Next step: when reading any reliability table, look at the cause-code breakdown, not just the total. A division whose interruptions are mostly tree contacts is telling you something different from one whose interruptions are mostly equipment failures, and the remedy — and the page of the rate case to read — is different too.
Where the Binghamton numbers actually live
NYSEG organizes its reporting by operating division, and Binghamton is one of them. The 2012 report’s table of contents lists a Binghamton division section alongside Auburn, Brewster, Elmira, and the rest, and it also includes a worst-performing-circuit analysis and a circuit CAIDI/SAIFI analysis. That structure is the practical map for anyone who wants local numbers.
The sequence to follow:
- Start with the most recent annual reliability report filed with the Department of Public Service. Find the division section for Binghamton. That gives you division-level SAIFI and CAIDI, typically with a multi-year history.
- Turn to the circuit performance section. This is where individual feeders are listed and ranked. The worst-performing-circuit analysis is the closest thing in the public filing to a feeder-by-feeder view.
- Pull three to five years of reports, not one. A single year’s division average can move for reasons that have nothing to do with your neighborhood. A trend across reports is harder to explain away.
- Cross-reference the cause codes. If your circuit appears in the worst-performing list, check whether its interruptions are attributed to tree contacts, equipment, or storms. That tells you which part of the rate case — vegetation budget, capital plan, or storm response — is the relevant lever.
- Ask for the specific circuit. If you know which feeder serves your block — it is often printed on the pole or available from the utility on request — ask for that circuit’s SAIFI and CAIDI history directly. The filed reports may not name it, but the underlying data exists at that level.
One honest limitation: the utility’s own outage-history web page was not retrievable when this piece was prepared, so this guide relies on the regulatory filings rather than the company’s public outage map archive. The filings are the more durable record anyway, because they are submitted under penalty of the Commission’s review and they carry the cause codes and division breakdowns that a live outage map does not.
What a negative revenue adjustment does and does not do
The $3.5 million figure attached to NYSEG’s 2025 frequency miss is a negative revenue adjustment, and the report is precise about who pays it: “Negative revenue adjustments are paid by shareholders and not by ratepayers.” It is a penalty on the company’s return, not a refund to customers and not a repair schedule. It signals regulatory dissatisfaction. It does not, by itself, shorten anyone’s outage.
That is why the penalty figure should always be read alongside the commitments in the same rate case — the vegetation budget, the capital program, the corrective action plans. The 2025 report notes that utilities must file annual reliability analyses including “analyses of worst-performing feeders” and, where needed, corrective action plans. Those plans are where a penalty becomes a work order, or does not.
Next step: if you are preparing public comment or a letter, cite the penalty and the cause code together, then ask what specific corrective action is attached to the worst-performing feeder in the Binghamton division. A penalty without a corresponding plan is a number; a penalty with a plan is a lever.
What the filings cannot tell you
Three limits are worth stating plainly, because overreading these documents is its own kind of error.
First, the published indices are aggregates. Even the division-level and circuit-level tables describe populations of customers, not addresses. A street that goes dark repeatedly can be invisible in a division average, and a division that looks stable can contain one feeder that is deteriorating.
Second, the storm exclusion means the target metric is deliberately not a full account of the customer experience. The report says as much: data including major storms “reflects the overall customer experience during a year.” If your concern is resilience during the events that cause the longest outages, the excluding-storm number is the wrong number to cite.
Third, the filings do not connect reliability to the other systems that shape risk in Binghamton’s urban core — the floodplain, the combined sewer, the levee system, the local grid configuration. Those connections may matter, but they are not established in these documents, and this piece does not assert them. What the filings do establish is narrower and still useful: a system-wide frequency problem, a dominant cause, a regulatory penalty, and a reporting structure that reaches down to the division and circuit level if you know where to look.
Frequently asked questions
Does the 1.29 figure mean my power will go out 1.29 times this year? No. It is a system-wide average for 2025, not a forecast and not a count for any one location. Your experience depends on the circuit serving your address.
Why do storms seem to be left out of the numbers? The Commission’s reliability targets exclude major storms so that utilities are judged on interruptions more within their control, such as equipment failures and tree contacts. The report also publishes data including storms, which is closer to the customer experience.
Where can I find Binghamton-specific reliability numbers? In the division section of NYSEG’s annual reliability reports filed with the Department of Public Service, and in the circuit performance and worst-performing-circuit sections of the same reports. The 2012 report’s table of contents shows the structure.
Does the $3.5 million penalty come back to me? No. Negative revenue adjustments are paid by shareholders, not ratepayers, per the 2025 report. They are a regulatory consequence, not a refund.
What is the single most useful thing to ask for? The SAIFI and CAIDI history for the specific circuit serving your address, plus the worst-performing-circuit list for the Binghamton division across several recent annual reports. That combination answers the question a system average cannot.
