Maine spent more than three and a half years building a process to make its two big utilities plan the electric grid in public. The first plans arrived last winter, and on July 31, 2026 the Public Utilities Commission closed the cycle by finding both of them adequate.

The result was not close and not much contested. What is worth reading is the record underneath it: what regulators asked the utilities to add, an unresolved argument about the forecast the plans were built on, and a list of questions everyone agreed to take up in the next cycle instead. No news coverage of the decision appears to exist.

What an integrated grid plan is, and why Maine has one

Utilities have always planned their systems. What they have not always done is show their work.

In 2022 the Legislature passed An Act Regarding Utility Accountability and Grid Planning for Maine’s Clean Energy Future, now codified at 35-A M.R.S. §3147. It requires the Public Utilities Commission — the three-member state body that regulates electric utilities and sets their rates — to start a proceeding every five years naming the priorities its investor-owned utilities must address, and then requires those utilities to file an integrated grid plan within 18 months. An integrated grid plan is a public document in which a utility sets out what its grid will need, what it proposes to build, and why it chose that over the alternatives.

Maine has two such utilities: Central Maine Power, serving most of southern and central Maine, and Versant Power, serving the north and east.

The Commission set three priorities in July 2024: reliability and resilience improvements; better data quality for distribution system planning; and promoting flexible management of customers’ own energy resources — rooftop solar, batteries, EV charging, heat pumps. Two principles sat over the top of all three: keep it affordable, and help the state hit its climate targets.

Then the Commission deliberately stepped back. Its own 2024 order explains why: many of these investments return later as requests for money in rate cases, so the Commission and its staff need “appropriate distance in the planning process to ensure impartiality” when that happens.

Staff asked both utilities for the same addition

CMP filed on December 12, 2025, Versant on January 12, 2026. Commission staff reviewed both and, in memos issued in March, raised two areas of concern about each.

One was common to both companies. Reliability and resilience was the first of the three priorities, and staff found the plans did not analyse investments in it. CMP’s plan, staff wrote, “did not include any analysis of reliability or resilience investments”; it was organised around capacity constraints instead, meaning places where the grid is running out of room. Versant named reliability as a priority in its executive summary and elsewhere, but did not include detailed analysis of the capital projects meant to deliver it; there too, “the vast majority of the capital projects” were capacity-driven.

Staff were explicit that this was a question rather than a verdict: their memos noted that concerns “may or may not rise to the level of a deficiency,” and that a utility addressing a requirement differently from what the Commission or a stakeholder expected does not by itself make a plan deficient.

Staff invited both utilities to supplement their filings. Both did. CMP’s response argued that “the Company does not view this as a deficiency of its IGP” — its case being that capacity headroom is reliability, since a grid with slack absorbs storms and equipment failures better — and then supplied the reliability analysis anyway, across six attachments.

The other staff concern to CMP was about cost. The Commission wants the utilities moving toward time series analysis — modelling the grid hour by hour across a year, rather than at a handful of worst-case moments, which is the only way to see what rooftop solar and EV charging actually do to a circuit. CMP had not said what that would cost. Its supplement put the figure at $54.38 million between 2026 and 2030. It also conceded something more interesting: it does not yet have clear standards for how to read time series results, and developing them is a prerequisite to using the method properly. Getting there, CMP said, would require changing its interconnection procedures — the rules governing who may plug into the grid and on what terms, including when they can be switched off.

The forecast the plans were built on

One question ran underneath the whole review and was not settled: the forecast.

The Commission required both utilities to plan against the 2024 CELT forecast — ISO-New England’s annual projection of how much electricity the region will need. ISO-NE is the independent operator that runs the six-state grid and its wholesale markets, and its forecast is the regional baseline everyone builds from.

By the time the plans were reviewed, numerous commenters said the 2024 CELT had overestimated adoption of electric vehicles and heat pumps, and that planning against it risked overbuilding the grid, which ratepayers would fund. The Office of the Public Advocate, the state office that represents utility customers, put it plainly: when the Commission issued its order, “rapid adoption of EVs was more plausible.”

The Efficiency Maine Trust, the quasi-state body that runs Maine’s efficiency and heat-pump programs, and the Department of Energy Resources, the governor’s energy office, both suggested the Commission hire an independent expert able to adjust the regional forecast to Maine’s actual data, or build a new one.

Versant had already gone partway there. Rather than rely on the top-down regional forecast, it built its own bottom-up projection from individual feeders — the local circuits that run down streets. Asked to compare, Versant showed its method finds more need than the regional forecast does: higher peaks, more overloaded devices and lines, more voltage problems. Its explanation is that a regional forecast captures the moment the whole system peaks together, while local peaks happen at different times in different places, so the regional number “will tend to not reliably capture distribution-level constraints.”

Staff’s objection was not that the method was wrong but that it was opaque: Versant modelled two boundary cases where the Commission’s template required six seasonal snapshots, which left stakeholders unable to trace which need came from where. The energy office commended the approach and urged its wider adoption. The Trust noted the regional forecast is imperfect but at least transparent.

That is a genuine, unresolved disagreement about how to size Maine’s grid, and it is downstream of a question this site tracks directly: how fast are heat pumps and EVs actually being adopted?

Nothing was found deficient, and most parties agreed

Eleven parties filed comments, including the Public Advocate, the Efficiency Maine Trust, the energy office, a coalition of Acadia Center, Conservation Law Foundation, the Natural Resources Council of Maine and the Union of Concerned Scientists, the Maine Chamber of Commerce, the Island Institute, and a sitting legislator.

Two commissioned technical reports are more careful than their titles suggest. A consultancy, Dynamic Grid, filed one called Deficiencies in Central Maine Power’s Integrated Grid Plan that states at the outset it “does not advocate for any particular outcome” and does not recommend approval or rejection. A review filed by the environmental coalition concluded CMP’s plan is “broadly consistent with good utility practice.”

Staff issued a recommended decision on July 2. Exceptions came on July 17 from CMP, the Public Advocate and the Trust — and all three supported the outcome. On July 31 the Commission, unanimously, found that both plans meet the statute and its 2024 order. Concerns raised by staff and commenters, it held, “do not constitute deficiencies in the grid plans but instead provide areas for clarification and improvement for future integrated grid plans.”

The reasoning is worth stating because it will govern the next cycle too: the fact that a utility addressed a requirement differently than the Commission or a stakeholder would have does not, by itself, make the plan deficient.

What got deferred, and to where

Most parties told the Commission no further process was needed on these plans, and that attention should turn to the next round. The Commission agreed. What it did with the specific asks is the useful part:

  • Lessons-learned workshops before the next cycle. The Trust argued in its exceptions that one workshop was inadequate given what stakeholders had invested. The recommended decision had said “a workshop.” The final order says a series of workshops — a small, traceable instance of an intervenor changing the result.
  • CMP’s reliability cost-benefit model, its reliability targets, and its value of lost load — already contested in CMP’s pending rate case, Docket 2026-00043.
  • Ambient adjusted ratings, which let a line carry more power when the weather is cool: judged possibly premature, because ISO-New England has not settled on a method yet.
  • Energy storage, including when a utility may own batteries: live in two other inquiries.
  • Non-wires alternatives — meeting a grid need with batteries, efficiency or demand management instead of building: the Commission reviewed this process for the Legislature, statutory changes were enacted this session, and it expects to open another proceeding.

What to watch

The next planning proceeding commences in November 2027 — the five years in the statute is how often the cycle restarts, not how long anyone waits. Between now and then:

  • The workshops, which is where the reliability-analysis gap, the assessment metrics and the forecast question are supposed to get resolved. Whether a series of workshops produces sharper requirements for 2027 is the test of whether “areas for improvement” meant anything.
  • CMP’s rate case, Docket 2026-00043, where the reliability model deferred here gets argued with money attached.
  • The forecast. If EV and heat-pump adoption keeps running below the 2024 projection, the case for a Maine-specific forecast strengthens — and so does the risk that plans built on the old one authorise more construction than the state needs.

The documents behind this post sit in MPUC Docket No. 2022-00322. The Commission’s case management system serves files through a session rather than at fixed web addresses, so a document cannot be linked directly — but the filing pages can, and opening one starts your own session. The July 31 order is item 154 and the July 2 recommended decision is item 150; each page lists its documents, and clicking one opens it. The whole docket record is here.

Both the order and the recommended decision were read in full for this post. The underlying grid plans, the public comments and the exceptions were not.

Maine now has a public record of what its utilities intend to build, and a schedule for arguing about it. Whether that changes what actually gets built is a question for 2027 — or for the rate cases that arrive sooner.