Capacity potential for onshore wind

Where can you build wind turbines?

Many analyses of the potential for onshore wind work by stacking exclusions on top of one another, and finding the area that's left. A density assumption is then applied to the km² which remains (generally between 10–20 MW/km²). But this approach will systematically underestimate the amount of onshore wind which can be built.

To make the problem tractable, capacity analyses are generally run on grids of a country, either square or hexagonal. Land can be excluded for geotechnical reasons (too low wind speed, or too steep a slope), or siting ones. Some siting decisions are hard limits: turbines cannot be built too close to roads or railways, or on sites of natural interest such as SSSIs or Ramsar sites. Others are softer: you might be able to safely build a wind farm 400 m from a residence, but a developer might only target sites at a further setback as there are likely to be fewer planning-related objections.

The animation below will explain where this analysis often goes wrong.

Animated visualisation — scroll to run
Landscape—
Turbine may stand—
Density estimate—
Red inside redline—
Sited—
Capacity—
Still farmed—
A turbine may stand here It may not Still farmed
01 / THE GROUND

Sixty-four square kilometres

The map shows a sample area for consideration. A river passes through the middle of the zone, and a highway down one side. There are clusters of buildings, a few forests, and, next to the river, a small nature reserve.

Nothing is ruled out yet. Now we'll apply the constraints one at a time. Watch the area where turbines might be placed (displayed in the table) gradually fall.

02 / WATER

The river goes first

You cannot place a turbine in a watercourse. The river and its banks turn red, and the channel removes 6.1 km² from consideration.

03 / WOODLAND

Then the forests

Three separate blocks of commercial forestry. Felling to build is possible but may not be worthwhile. Wind turbines also cannot be situated close to forests, as they will interfere with the wind. Removing the forests takes out a further 5.9 km².

04 / DWELLINGS

The clusters of buildings

The buildings themselves occupy very little land. But minimum distances from buildings push the exclusion out around them, and 4.1 km² goes.

What setback constraint you assume greatly impacts the exclusion here.

05 / THE ROAD

The single biggest reduction

The carriageway is one tile wide. Its verge exclusion sets aside a tile on each side, and together they take 8.9 km² — more than the river, or all the forestry.

06 / DESIGNATION

And the sanctuary

The last constraint is the smallest: 1.0 km² of protected wet ground nature reserve to the west.

07 / WHAT SURVIVES

Everything still standing turns green

37.9 km² — 59% of where we started. All of the grid cells remaining could hold a wind turbine.

This is what every land-availability study produces. What happens next is where it goes wrong.

08 / ONE PARCEL

Now take a single site

Developers will focus on individual parcels of land, generally leased from the local landowner. Our parcel here is 17.6 km² with the river straight through it, a block of forestry, and the buildings making a hole in the middle.

Of that parcel, 8.1 km² is green. The developer could try to just lease those green cells, but the wind farm may interfere with other uses of the land. They may also wish to repurpose the buildings for use in the wind farm.

09 / THE REDLINE

The boundary that becomes the denominator

Every planning application draws a red line around its site. Inside the line: 17.6 km². Green: 8.1 km². The remaining 9.5 km², more than half, is river, forestry and setback that no turbine will ever stand on.

10 / THE SHORTCUT

Multiply by a density

An existing 150 MW wind farm sits inside a 21 km² redline, giving 7.1 MW/km². Studies often take this density, and multiply the green area by it to predict the capacity.

But that 21 km² was a redline just like the one in the figure, and it was never all buildable. The density figure has unbuildable ground baked into its denominator, so you have removed the river and the trees once from the map, then again inside the multiplier.

11 / PLACE THEM INSTEAD

Site the turbines and count

Drop turbines onto green ground, keeping ~900 m between machines (about six rotor diameters) so they don't steal each other's wind.

Watch the tally overtake the estimate.

12 / THE REASON

A turbine only needs 1 tile

Each ring is ground that must stay clear of other turbines. Twelve of the fifteen spill across the river, over the forestry, into the setback.

A turbine has to stand on green. Its exclusion zone does not.

Density estimate—
Actually sited—
Understated by—
13 / LIMITATIONS

It doesn't all run one way

Some machines end up on scraps of green barely wider than a foundation. This may be technically possible, but it won't be economically viable to build a single turbine on a scrap of land in the middle of nowhere.

This can be accounted for by removing the smaller parcels of land from the analysis.

14 / WHAT THE LAND STILL DOES

How much of the land is used?

Fifteen machines now stand on fifteen of the parcel's forty-one green tiles. The other twenty-six — 5.1 km², nearly two thirds of the buildable ground — carry no turbine at all, and could be used for something else, such as agriculture.

And the picture is generous to the turbines. Each is drawn owning a whole 444 m tile, twenty hectares of it. The foundation, access road and crane pad will take up a lot less space in reality.

The land which wind farms "take up" can easily be used for other purposes. The wind farm does restrict what you can use the land for (houses etc. will not be viable), but agriculture and many kinds of recreation are still possible.