A marsh village in western France voted to roof its flooding graveyard with a 1.3 megawatt solar canopy and told the world, but the summer it promised for first power went by over a year ago

A scatter of low islands in a peat marsh north of the Loire estuary, in western France.

The commune sits barely above sea level, and its graveyard fills with water every winter.

Draining it would mean fighting the wetland forever, so the mayor proposed a roof instead.

In a commune of roughly 4,000 people that works out near a third of a kilowatt for every resident, which is a community scale plant rather than a regional one.

Above the graves would stand about 2 acres of steel trusses carrying panels.

The announcement went around the world.

The switch on date came and went.

Why peat decides the shape of anything built on it

Peat is saturated by nature, and saturated ground does not resist a point load the way firm soil does.

A standard ground mounted array drives piles into the earth at close spacing, which in waterlogged peat means either very long piles or none that hold.

The alternative is to spread the load. Widely spaced trusses put their weight onto footings sized for soft ground, and the structure carries the span between them.

That choice has a consequence overhead. A canopy on long spans catches Atlantic wind as a sail, and horizontal force on a sail cannot be shed downward the way a low rack sheds it.

So the structure gets heavier and more expensive precisely because the ground underneath is soft.

The marsh sets the spacing, and the spacing sets the price.

The roof came before the generator

This is the part that inverts the usual story, and it is why the project drew attention at all.

The structure was going to be built whether or not it made electricity, because the purpose was to keep rain off the graves.

A canopy also collects water rather than repelling it, and the plan channels that runoff toward a neighboring sports complex for irrigation.

Once a roof of that size was going up anyway, the marginal cost of laying panels across it was small compared with building a solar farm from nothing.

Most arrays make a case for their own structure. This one arrived with the structure already justified.

The panels are the second reason for a thing already needed.

What was actually specified and what it was to cost

The design calls for 5,000 photovoltaic panels across a solar canopy of about 2 acres, rated together at 1.3 megawatts.

In a commune of roughly 4,000 people that works out near a third of a kilowatt for every resident, which is a community scale plant rather than a regional one.

The municipality was to finance the whole thing at about 3.35 million euros, drawn from capital gains tax receipts rather than borrowed.

That is close to 2.77 dollars a watt, higher than flat ground mounted solar and defensible only once the drainage function is counted alongside it.

Households joining as consumers pay a one time entry of about five euros and were projected to save between 150 and 250 euros a year.

The roof pays part of itself before any panel earns.

Sharing one array across a thousand meters

A single 1.3 megawatt source supplying an entire village is harder than it sounds, because standard shared supply arrangements in France assume a handful of participants.

This design needs more than a thousand connections drawing from one plant at the same time, which is an unusual number for that framework.

The answer was metering. Readings every half hour from both the canopy and each consumer go to the grid operator, and an algorithm allocates production proportionally before anyone is billed.

Nothing about that is technically novel. It is novel at this participant count, in a commune where the plant and the customers sit inside the same few square miles.

Municipal water bodies take the gentler version of these problems, as an Ohio city found floating 3,120 panels on its drinking supply.

The panel count, the capacity and the financing are described by a solar title.

One plant, a thousand meters, half hourly.

What nobody has said since

Here is the uncomfortable part, and it is the reason to write about this now rather than when it was announced.

The plan dates from early 2024 and named the following summer for first full delivery. That was more than a year ago.

No operator statement confirming the canopy was completed or energized appears in the public record, and the French trade press that covered the announcement has not returned to it.

Silence is not failure. Permits, procurement and a prototype section that went up separately can all absorb a year without anything going wrong.

But a project this widely reported should have a completion notice by now, and the absence of one is itself the current state of the story.

Panels age into a disposal problem eventually too, which is why recycling rates matter as fleets mature.

The costs, the savings and the original timetable are set out by an energy outlet.

The graveyard still floods, and the best available answer to whether the roof exists is nobody has said either way.