14 Oct 2026

How to rank a multi-site portfolio before a PPA bid

Several industrial buildings with solar panels on their flat roofs, seen from a drone

A PPA bid deadline doesn't care how many sites are in the portfolio. It cares which ones you can commit capacity numbers for, and you usually have weeks, not months, to turn a spreadsheet of addresses into a shortlist worth sending a site team to. The teams that make the deadline stopped ranking by square metres first.

Why square metres get the ranking wrong

Raw roof area is the number everyone pulls first because it's the easiest to get: pull a footprint, measure the polygon, done. It's also the number that ranks sites wrong more often than it ranks them right.

A 4,000 m² roof with a north-facing pitch, three HVAC clusters, and a skylight run down the middle might carry less installable capacity than a 2,800 m² flat roof with clean orientation and one rooftop unit tucked in a corner. If your shortlist is built on area alone, the first site bids high and the second gets dropped before anyone looks twice. That's the opposite of what a PPA bid needs. Buyers and lenders want kWp they can underwrite, not square footage they have to discount later.

What belongs in the ranking

For a portfolio ranking that holds up once a site team walks the roof, three things matter more than total area:

Usable area. What's left after you subtract parapets, access paths, fire lanes, and anything else code or the facility manager won't let you build over.

Orientation. Not just "south-facing good, north-facing bad." Pitch, azimuth, and how much of the usable area sits in a favorable orientation band.

Obstruction. HVAC units, vents, skylights, existing equipment, and the shadow each one throws across the array for the hours that matter.

Run those three against usable area and you get installable kWp per site, which is the number a bid needs. It's also the number that lets you rank forty sites against each other on the same basis, instead of comparing forty different site visit reports written by different people on different days.

Building the shortlist before anyone drives to a site

The old way to get this ranking was a desk screen against aerial imagery, followed by a site-visit shortlist to confirm the top candidates before anyone touched a design. It works, but it's slow, and it front-loads the site visits that are expensive to schedule across a scattered portfolio.

A faster pass is to run every roof and parking canopy in the portfolio through one VHR capture and rank them on installable kWp before you decide which ones earn a site visit at all. That's the approach behind Rooftop Solar Suitability: one pass across the portfolio, ranked by what's buildable, not by raw footprint. It decides which sites are worth walking, before the walk-down gets scheduled.

For a bid with a real deadline, that ordering matters. If you're ranking sixty sites and can only staff site visits for fifteen, those fifteen should be the ones the data already flags as strong, ahead of whichever sites happen to be closest to the regional office or whichever ones a junior analyst eyeballed first off a satellite thumbnail.

When to re-run the ranking

Roofs change. A tenant adds HVAC capacity, a canopy gets extended, a neighboring structure goes up and throws new shade across a parking lot. A ranking built on last year's imagery can quietly go stale, especially for a portfolio you're re-bidding on a rolling basis rather than once.

An annual VHR pass is enough to catch that drift for most portfolios, and that's the cadence a PPA bid ranking actually needs: current enough that the top of the list hasn't moved since the last time someone photographed the roof.

If your portfolio is big enough that ranking by hand before every bid cycle is eating the time you'd rather spend on the sites that clear the bar, it's worth seeing what a single annual pass across the whole portfolio turns up.

Start a pilot

← Back to the blog