How many met masts does a complex-terrain project actually need

There's no lookup table for this. Anyone who quotes you a flat mast count per megawatt hasn't sited a project on terrain complex enough to make the question worth asking. What drives the count is how many distinct flow regimes sit across your array, independent of megawatts, turbine count, or project acreage.

On a flat agricultural site, one mast with a representativeness radius of several kilometers can cover the whole layout and a lender will sign off on it. Put that same single mast on a ridge with a saddle, a leeward bowl, and a second parallel ridge half a kilometer over, and you've got three different wind regimes reporting through one sensor. The mast still reports correctly for where it's standing. It simply doesn't speak for ground it was never on.

What drives the count

Three things push the number up, and none of them are "how big is the project":

Terrain complexity index (RIX). Sites with RIX values above zero, meaning the terrain exceeds the slope where linear flow models like WAsP start to break down, need more measurement points because the model can't be trusted to extrapolate wind speed from one point to the next ridge.

Elevation change across the array. A couple hundred meters of relief between your northernmost and southernmost turbine strings usually means separate flow regimes, separate masts.

Flow separation and recirculation zones. Leeward slopes, bowls, and anywhere you'd expect turbulent wake off a ridge crest are places a single upstream mast will systematically misread. If your layout has turbines sitting in or near a recirculation zone, that cluster needs its own measurement, full stop.

Some resource teams follow a simple rule of thumb: one mast per terrain feature that could plausibly produce its own wind regime. Turbine count doesn't enter into it. A 40-turbine project spread across two parallel ridges with a valley between them might need three masts. A 15-turbine project tucked onto one clean ridgeline with consistent exposure might get by on two, sometimes even one if the RIX comes back low and uniform.

Mast spacing in complex terrain

Spacing follows the same logic as count. In flat terrain, representativeness radius is mostly about distance. In complex terrain, it's about similarity of exposure, and two points 300 meters apart can have completely different wind climates if one sits on a convex ridge crest and the other sits just over the lee side.

That's the trap teams fall into when they space masts the way they would on flat ground: evenly, by distance along the planned string. A mast every 2 km might look reasonable on a site plan and still miss the one saddle where wind shear flips the rotor-equivalent wind speed. Spacing decisions in complex terrain should follow terrain breaks, not a grid.

This is also where a lot of resource teams end up adding a mast after the fact, once SCADA data from year one shows an energy yield gap the pre-construction model didn't predict. Cheaper to catch that gap before met masts go up than after.

Screening before you commit a crew

Most of this terrain analysis, slope, aspect, surface roughness, ridge geometry, used to mean sending someone out with a clinometer and a notebook or waiting on a topo survey. A fair amount of it can now be done from a desk, working off stereo very-high-resolution imagery to pull slope angle, surface roughness, and obstacle height across the whole candidate area before anyone drives out. Screening the ridgelines that way first tends to cut the list of "maybe we need a mast here" sites down to the ones worth the mobilization cost.

Count terrain features, not turbines. A ridge with two flow regimes needs two masts no matter how small the project is; a project with twenty turbines on one uniform slope might get by on a couple.

If you're trying to figure out which ridgelines on your candidate site are worth that kind of scrutiny before the mast order goes in, that's the gap Wind Farm Siting fills.