Utility-Scale Solar EPC: Field Production Tracking

Why solar production tracking is different

Most commercial construction has variety — different rooms, different systems, different conditions from one area to the next. A solar field has almost none of that. A 200 MW site might have 40,000 piles, 8,000 racks, and 500,000 modules, and the design intent is that every one of them takes roughly the same labor. That sameness is a gift if you track it and a trap if you do not, because a crew that is 10% slow on pile driving does not look slow on any single day — it looks slow only when you compare cumulative output to the plan curve.

The other wrinkle is that solar EPC work is almost entirely self-perform or thinly subbed, with tight fixed-price or unit-price contracts. There is rarely a change-order safety net the way there is on a lump-sum building job. The margin is built at the bid in dollars per watt, and it gets protected or destroyed in the field, unit by unit, day by day.

The production units that matter

A handful of activities carry almost the whole labor budget on a typical fixed-tilt or single-axis-tracker site:

• •Pile driving: piles per crew per day, tracked against soil classification since refusal and pre-drilling change the rate dramatically.
• •Racking/torque tube assembly: rows or bays completed per crew per day, including torque-spec QC on fasteners.
• •Module installation: modules landed per crew per day, typically the highest-count, most repetitive activity on the site.
• •DC wiring and combiner terminations: strings terminated and megger-tested per day, since this is where electrical QC gates production.
• •AC collection and trenching: linear feet of trench or cable pulled per crew per day.

Building the daily report around units, not hours

A daily report that only captures "Crew A: 8 hours, pile driving" tells you nothing about whether the job is on pace. The report needs a quantity field tied to the activity: piles driven today, cumulative to date, and percent complete against the total count for that block. Divide today's hours by today's units and you get labor units per pile (or per rack, or per module) — the same productivity-factor math used across the trades, just applied to solar-specific units of work.

Cost-coding this to the WBS matters just as much as it does on a building job. A typical solar cost-code structure runs by system and activity: 03-100 civil/grading, 03-200 pile driving, 03-300 racking, 03-400 module install, 03-500 DC wiring, 03-600 AC collection, 03-700 inverters/transformers, 03-800 commissioning. Each crew's daily hours and units get coded to the matching line so the job cost report shows dollars-per-unit-installed alongside dollars-per-hour, which is the number that actually predicts whether the block finishes on budget.

Field PM's <a href="/features/daily-reports">cost-coded daily reports</a> handle this the same way they handle linear-foot and square-foot trades: a quantity-installed field per cost code, rolled up automatically into a productivity factor per crew per day, so a slipping pile crew shows up on the dashboard within a day or two instead of at the monthly cost report.

A worked example: catching a pile-crew slip early

Say the pile package is bid at 1,600 piles for a crew, at a planned rate of 40 piles/day (10 hours/day, 4 piles/hour, roughly 0.25 labor units per pile). Week one the crew logs 38, 41, 39, 40, 37 — right on plan, no alarm. Week three, after hitting a rockier section, they log 31, 29, 33, 30, 28 — a rate of about 30/day, or 0.33 labor units per pile, a 32% productivity hit.

On an hours-only report, that week just shows "50 hours, pile crew" and looks normal. On a units-tracked report, the productivity factor jumps from 1.0 to 1.32 immediately, and the forecast recalculates: at 30/day instead of 40/day, the remaining 1,200 piles take 40 days instead of 30 — a 10-day slip on that one activity alone, caught in week three instead of at 90% complete. That is enough lead time to add a second rig, adjust the crew mix, or renegotiate soil conditions with the owner while there is still schedule float to work with.

Weather, curtailment, and the real drivers of variance

Solar sites are exposed sites — wind holds on crane and module-handling work, rain days that shut down grading and trenching, and heat that limits productive hours in southern and desert regions are normal, not exceptions. A production-tracking system needs a weather/delay field on the daily report so a low-count day tied to a wind hold does not get blamed on the crew, and so the delay log builds the record needed if a schedule extension claim ever has to be made.

The other real driver is soil and site variability across a large footprint — pile refusal rates, obstructions, and grading quality differ block to block even on one project. Tracking production by block or area, not just by crew, shows whether a slow week is a crew problem or a ground-conditions problem, which changes what you do about it.