01Establish the load before the projectInterval consumption across a full operating cycle, plus generator start hours where a generator runs for load. The profile, not the monthly total, is the design input. Everything later depends on it.
02Survey the constraints that decide buildabilityRoof or land area with structure and access, transformer capacity, protection settings, the interconnection point the utility will accept, and every obstruction the site intends to add later.
03Read the yield figure properlyAsk what soiling, temperature, downtime and degradation were assumed, and who is accountable if the plant does not reach the stated figure. A yield without its assumptions is a marketing number.
04Separate the price from the excluded scopeInterconnection works, protection upgrades, structural strengthening and utility approvals are the usual exclusions. Compare tenders on the same scope boundary or the cheapest is only the least complete.
05Decide where storage changes the caseStorage earns its place by shifting generation into the hours that need it, shaving a demand peak, carrying critical load, or cutting generator hours. If none of those is true, storage is an expense without a mechanism.
06Define completion in measurable termsCapacity, performance ratio, protection verification, documentation and operator training. A completion test that cannot be measured cannot be enforced, and retention against it is worthless.
07Plan the twenty-five years, not the twoSpares strategy, monitoring, review cadence and who responds when a string drops out. The asset outlives the warranty by two decades, and the operating plan is decided at design stage.
08Interrogate the engineering, not the presentationAsk who signs the design, which engineer is accountable at commissioning, and what happened on the last project of comparable scale and constraint. The answers describe the organisation you are actually hiring.
