Racked tables vs. installed modules: reading a solar site from above

A construction monitoring report lands on your desk saying a site is at 70% mechanical completion. You pull up the latest satellite pass to sanity-check it before the draw gets funded, and the field is full of rows. Dark, uniform, reflective. Looks like panels. But at the resolution most people are working with, a fully racked table with no modules on it can read almost the same as a table that's done.

That gap between "racking is up" and "modules are mounted" is where a lot of percent-complete disputes start. Here's what separates the two in imagery, and why it's worth getting right before you sign off on a milestone.

What bare racking looks like from above

An empty table, torque tube and piers set, module clips installed but nothing racked yet, shows up as a thin linear structure. From a VHR pass at 0.3 m or so, you're looking at narrow rails and shadow lines rather than a solid surface. The ground between rows is often still visible through the frame, especially on single-axis trackers before the panels close the gaps. Color is whatever the steel and the underlying soil or gravel happen to be, usually gray or tan, not the deep blue-black of a module face.

Shadows are the other tell. A bare table casts a thin, broken shadow pattern because light passes between individual rail members. A table with modules mounted casts one continuous shadow the width of the row, because the panel surface is unbroken.

What installed modules look like

Mounted modules read as a continuous, uniform rectangle: consistent width down the full run of the table, consistent dark color, and in a lot of passes you'll pick up glint where the sun angle hits the glass at the right incidence. That glint is actually one of the more reliable cues at sub-meter resolution. Bare racking has no glass to reflect, so it never produces that bright, almost white flare that a module array throws back on a clear day.

Row spacing also tightens visually once modules are on. Racked-but-empty tables tend to look "gappy," you can often make out the piers as individual points along the rail. Once panels are mounted, the row reads as a solid block with clean edges, and that edge becomes countable: a table, not a scaffold.

Why this distinction matters on a draw schedule

An EPC's percent-complete figure is often based on man-hours logged or linear feet of racking erected, not on modules torqued down and wired. Racking crews frequently run ahead of the stringing crews behind them, especially when module deliveries are delayed or a change order holds up a block of the site. That means "mechanical complete" can be true for structural steel while the modules themselves are still in boxes.

For a lender or asset owner tying a draw to an installed-module count or a percent-of-interconnection-capacity threshold, that gap is the whole ballgame. A visual count that distinguishes racked tables from mounted modules gives you the as-built number against the schedule, independent of whatever the EPC's field report says. It's also a decent early flag for a stalled procurement line: a site with plenty of racking and no module growth quarter over quarter usually points to a supply problem upstream.

If you're trying to settle a milestone dispute or just want a number you didn't have to take on faith, this is the specific problem Solar Panel Counting is built around: a counted inventory of modules, tables and inverters from VHR imagery each quarter, so the draw gets verified against what's standing in the field.

Next time a completion report feels ahead of where it should be, pull the imagery and look for the glint, not just the rows.