Factory solar projects fail in a characteristic way: they begin with the roof. Someone measures the available area, divides by a panel, multiplies by a wattage, and calls the result a plan. The roof is the easy measurement, which is exactly why it misleads. A factory is not a surface; it is a load with a schedule. These notes set out the planning order that respects that — machinery first, roof second, and price a distant third. Where a concrete configuration is mentioned, it is the 200KW factory configuration, used as a worked example of the format.
Before any array is discussed, the factory's electricity draw should exist as a picture: consumption across the day, across the week, across the production year. Bills alone flatten this picture into monthly totals, and monthly totals conceal the only thing that matters — the shape. Two factories with identical bills can need entirely different systems: one running a single daytime shift in step with the sun, the other running furnaces through the night that solar generation will never touch directly.
Gather the curve from metering or from the production schedule translated into loads. Mark the peaks and ask what causes them. A solar configuration designed against the real curve can be defended figure by figure; one designed against the bill can only be defended by hope. Suppliers who ask for the curve before quoting are telling you something good about themselves.
Steady draw is the polite part of a factory's appetite. Motors, compressors and welders introduce the impolite part: starting surges that briefly multiply a machine's running draw. A configuration — above all its inverter side — that carries the steady load and faints at a motor start will be remembered by the production manager in terms unsuitable for a brochure. Put the largest motors and their starting behaviour on the planning table explicitly.
This is also where honesty about simultaneity pays. Not every machine starts at once, and planning for the theoretical worst instant oversizes the system; planning for the average undersizes it. The realistic busy hour — the one the shift supervisor recognises — is the design point to agree, in writing, between buyer and supplier.
Factory systems are bought for different masters. Some exist to shave daytime grid consumption, and their success is measured on the bill. Some exist to protect production — or a chosen part of it — against grid failure, and their success is measured in outages nobody noticed. Some are asked to do both, which is legitimate, provided the budget understands that the second purpose adds storage and circuit work the first does not need.
Projects that refuse this decision drift. The array grows to satisfy the bill argument while the storage stays sized for the brochure, and the first serious outage delivers the verdict. State the purpose in the project file, in one sentence, before configurations are compared. Every later argument becomes shorter.
When the roof finally gets its turn, treat it as engineering, not geometry. Construction, age, load-bearing capacity and existing plant all constrain the array, and the mounting specification must answer for local wind and snow. The 200KW factory configuration states structural figures for its mounting as part of its specification; that is the completeness to expect, because an array is a long-lived structure bolted to a working building, and its failures are structural before they are electrical.
Leave access in the layout. Panels need cleaning, inspection and eventual replacement, and a roof packed edge to edge with modules and no paths will tax its owner for decades. A slightly smaller array that can be maintained beats a maximal one that cannot, over any honest accounting period.
Whether a factory system carries storage, and how much, should be decided by production, not by fashion. Storage earns its place when it protects processes whose interruption costs real money — a line that cannot stop mid-cycle, cold storage that cannot warm, controls that cannot drop. It earns less when it merely shifts energy between hours at a price the tariff does not justify. That configuration offers its storage in two forms, gel and lithium, which is the right shape for the conversation: the choice between them is a project decision about duty, footprint and handling, not a slogan.
Decide the protected list the way a household should — in daylight, in writing, before the outage. For a factory the list is written in processes rather than appliances, and the commissioning test is the same: cut the supply, watch what carries, and write down what happened.
A factory project closes the way it opened: with documents. The final configuration, complete with models and quantities; the load curve it answers; the purpose sentence; the structural specification; the documents for the named models — certificates available upon request, asked for against this configuration, at this stage, while answers can still change the order. Before ordering, confirm the minimum order quantity and lead time for your configuration in the same written thread.
None of this is bureaucracy for its own sake. A factory system is a capital asset assembled from crates on a working site, and the written file is the only version of the project that every party — buyer, supplier, installer, and the engineer who inherits the system in five years — will ever share. Write it like it matters, because it is the project.
Not every factory project should be built at full scale in one pass. Phasing — installing a first stage against one part of the load curve, learning from a production cycle of real generation data, then expanding — suits factories whose demand is growing or whose roof will be rebuilt in stages. Phasing is not indecision; it is the project's respect for measured results. The first stage of a well-phased project is designed so the second stage extends it rather than replacing it, which is a design requirement to state at the very beginning.
What phasing must never become is a way to avoid the planning notes above. A phase still needs its load curve, its purpose sentence and its structural specification; it simply draws them around a smaller boundary. Factories that phase the hardware while completing the planning get the best of both orders: full knowledge first, capital spent as the results arrive.
Before ordering, confirm the minimum order quantity and lead time for your configuration, and ask which documents apply to the exact system you are quoted.
Send an inquiry