Irradiance is table stakes. A site that clears the initial screen for a utility-scale project, whatever region it sits in, almost always has adequate sun. The differences that separate a project that gets built and makes money from one that quietly dies in year two of development live somewhere else. They live in the grid, the land, and the permit.
Why sun is a weak filter
The core problem with ranking sites on irradiance is that solar resource varies gently across most regions people are actually developing in. Move fifty miles and the annual sun a site receives might shift by a percent or two. That is a real, bankable difference at the margin, worth capturing when everything else is equal. It is not a difference large enough to justify choosing a site on that basis alone, because the other factors in play do not vary gently. They vary in steps, and some of those steps are a cliff.
A site either has spare capacity on the nearby grid or it does not, and if it does not, the difference is not a percentage, it is years. Land either sits flat and cheap or it does not, and the gap between those two conditions can be the entire economic case for the project. A site either sits clear of wetlands and protected habitat or it does not, and if it does not, the project may not exist at all, regardless of how much sun falls on it. None of those are tiebreaker-sized differences. They are gate-sized differences, and gates should be checked before tiebreakers.
Gate one: grid capacity
The first and largest gate is the grid, and it is not the same question as grid proximity. A site can sit a few hundred yards from a transmission line and still have nothing to offer a developer, because the line, or more precisely the substation and the interconnection queue behind it, has no spare capacity left to accept new generation. Being close to infrastructure and being able to use it are two different facts, and only the second one matters.
This is where the irradiance-first instinct does the most damage, because a beautiful, sunny, flat parcel next to a congested substation is not a good site. It is an option on a good site, contingent on a grid study and an interconnection queue that, in many markets today, runs years long. A less sunny site sitting next to a substation with real headroom can be delivering electricity, and revenue, while the perfect site is still waiting for its interconnection agreement. Time value of money alone can make the weaker-sun site the better investment.
Being close to the grid and being able to use it are two different facts. Only the second one matters.
Gate two: land cost and slope
The second gate is what the land itself does to the economics. Irradiance sets how much energy a given array will make. Land cost and slope set how expensive it is to build that array and how much of its output survives the build.
Cheap, flat land is forgiving. It keeps grading and racking costs low, it simplifies drainage and access roads, and it lets a project with merely adequate sun still return a strong result, because so little of the revenue gets consumed by the cost of the ground underneath it. Expensive or steep land is unforgiving in the opposite direction. Grading a hillside, managing drainage on a slope, and paying a premium for the parcel itself can eat deeply into a project’s economics, and a site with genuinely excellent sun can still fail to pencil once those costs are honestly counted. Irradiance describes the top line. Land cost and slope describe how much of that top line survives to the bottom one.
Gate three: permitting risk
The third gate is the one that kills projects the sun never gets a vote on. Wetlands, floodplains, protected or endangered species habitat, cultural resources, and local zoning opposition have ended more solar projects than weak irradiance ever has, and they tend to do it after a developer has already spent real money on land options, surveys, and early engineering. A site can have excellent sun, an available substation, and cheap flat land, and still never become a plant, because it sits on the wrong side of an environmental or political line that nobody checked early enough.
The frustrating part is that most of this risk is visible before a single site visit. Wetland maps, floodplain maps, species habitat data, and local zoning ordinances are public information in most jurisdictions. Screening against them costs a desktop review, not a construction budget, and it is exactly the kind of check that a sun-first mentality skips, because a bright irradiance number feels like progress and a permitting search feels like paperwork.
The order that actually works
Put the three gates in front of the tiebreaker and the correct screening order falls out on its own. Start with grid capacity, because a site with no path to interconnect is not a site, it is a hope. Then land cost and slope, because they determine whether the project is economically viable on the land it would actually occupy. Then permitting risk, because it determines whether the project is legally and politically viable at all. Only once a candidate has cleared those three does irradiance earn a vote, and at that point it is doing exactly the job it is suited for: choosing between several viable sites, not selecting one unproven site out of many.
This is not an argument that sun does not matter. Among a shortlist of sites that all have grid capacity, affordable flat land, and a clean permitting path, the sunnier one is genuinely the better choice, all else equal. The argument is about sequence. Irradiance is a fine way to break a tie. It is a poor way to build a shortlist, because it says nothing about whether a site can actually be interconnected, financed, or permitted at all.
The reframe
A solar resource map is the easiest data to pull and the most tempting one to lead with, because it produces a clean number and a satisfying color gradient. But the map does not know whether the substation nearby has room, whether the land beneath the brightest pixel is flat or steep, or whether a protected wetland sits under the same bright pixel. Those answers live in different datasets entirely, and they are the ones that decide whether a project gets built.
So the discipline worth building into a site selection process is resisting the pull of the sunniest pixel on the map until the grid, the land, and the permit have all already said yes. Sun is real, and sun is valuable. It is just not first.
