Author: Vyacheslav Yurdyk, Quality Engineer, LK Energy Group.
Over 1,000 units of equipment accepted at the factory (quality control, QC) and 200 electrical installation sites — under technical supervision.

Short answer (for those in a hurry)

The cheapest bid for building a solar power plant is often cheaper not because the contractor is more efficient — but because they saved money on what you won't see at commissioning: an undersized transformer, an undersized cable cross-section, cheaper protection. The plant will be commissioned, the acceptance certificate will be signed — and then for years it won't be able to deliver full power exactly when the panels are producing the most. In plain terms: you paid for that share of electricity (it's within the rated capacity, and the panels do generate it), but because of a weak link it has to be curtailed — and it simply never reaches the grid. That lost revenue is gone for good and no one will give it back.

Below is a generalized example from our service practice, explained in plain language, and an 11-point checklist of where budgets get cut, and how to check for it in the project before signing. The breakdown is based on our experience servicing industrial solar power plants in Odesa region and across Ukraine.

The paradox of the cheap plant: savings at the front end mean you underdeliver at the output

An investor enters a solar power plant project with simple logic: there are several bids, and the cheapest one wins. This seems rational — a megawatt of capacity looks the same across all of them. The problem is that rated capacity on paper and the plant's actual output are two different things, and the difference is created precisely by decisions a non-specialist won't see.

Let me state the main point up front, because it's not obvious: when a plant component is undersized, it heats up — but the real money is lost not from the heating itself (that's a small part), but because the risk of overheating forces the plant to be curtailed — and part of the energy the panels are ready to deliver at the sunniest moment simply never reaches the grid. We'll show this with an example below.

We see these stories firsthand. Among the sites we take on for maintenance, there are plants with mistakes built in back at the construction stage by a different contractor. Savings at the start turned into a plant that doesn't reach full power at peak for the owner — and has stayed that way for years. Let's break down the two most telling types of such mistakes using a conditional example, then give the full checklist.

Example 1. A transformer “right at the limit” instead of a transformer “with margin”

In plain terms. The transformer is the component through which all of the plant's energy exits into the grid. It's the “neck” of your plant: no matter how powerful the body (panels and inverters) is, everything outside passes only through this neck. If the neck is narrow, it physically cannot pass as much as the panels produce at the sunniest moment.

Take a plant with a capacity of around 1 MW. Such projects often get a 1000 kVA transformer, following the logic “1 MW ≈ 1000 kVA, it all adds up.” Such plants usually have tier-1 modules (LONGi, JA Solar, Trina Solar) and Huawei or Sungrow inverters — but all of their power exits to the grid only through the transformer: if the “neck” is narrow, even the best equipment underdelivers.

Why this is your money. A transformer sized “right at the limit” means that at midday peak, when the panels deliver their maximum, the neck fills right up to capacity — and the component starts operating at its limit and heating up. The heating itself is a small loss. The real problem is different: to keep the transformer from overheating and failing, the plant has to be curtailed. The inverters get a command to “hold back full power” — cutting off exactly the peak of generation, the largest volume of kilowatt-hours, during the sunniest hours. The result: the panels are ready to deliver a full megawatt, you paid for that megawatt when you bought a 1 MW plant — but it doesn't reach the grid, because the neck won't let it through. And this repeats every day at every sunny peak, for years.

Analogy. Imagine you bought a truck expecting to haul a ton per trip, but you were given an engine that can only pull 800 kg. It can technically move a ton — but at its limit, overheating. To keep the engine from burning out, the driver is forced to underload the truck every time: 200 kg of cargo stays behind at the warehouse on every run. The cargo exists, you paid for it — but it doesn't move. Over a year of such runs, a mountain of unshipped goods piles up at the warehouse. In exactly the same way, an undersized transformer forces you to leave part of the “cargo” — your electricity — undelivered to the grid every day. The scale in this example is illustrative, but the direction holds: part of the output stays unsold.

Engineering detail (for those who want to verify). A competent engineer does not load a transformer to 100% of its rating — they build in a margin of roughly 20%, meaning a 1 MW plant gets a 1250 kVA transformer, so that in normal operation the transformer runs at ~80% and has thermal headroom. This is standard reliability practice; its logic is reflected in the loading guide for power transformers, IEC 60076-7. The margin isn't there “just in case” for its own sake: it provides thermal headroom for hot days and peaks, without which the insulation's service life burns out faster and the plant has to be curtailed.

Example 2. Transformer protection “for show”

In plain terms. A transformer needs a “fuse” that instantly disconnects it if a fault occurs inside (a short circuit). The quality of this protection determines whether an expensive component survives a fault or burns out. This is another place where contractors cut costs — installing simpler, cheaper protection that doesn't always trip correctly in time.

The second typical place to cut costs is protection of the transformer on the 10 kV side. Instead of a circuit breaker (today, typically vacuum) with relay protection, a cheaper combination is installed: a load-break switch plus fuses.

Why this is your money. The difference between these two solutions is the difference between a “smart guard” and a “simple fuse.” A load-break switch can only switch normal operating current on and off — it is not designed to interrupt fault currents from a short circuit; for that it relies entirely on the fuses. If the fuses are selected incorrectly, or the fault inside the transformer is such that the fuse doesn't “catch” it in time, the transformer will suffer more damage than it would have with full relay protection. One unlucky event, and you're either repairing or replacing the transformer, while the plant sits idle and earns nothing.

Engineering detail (for those who want to verify). Relay protection of a transformer is required by PUE (Electrical Installation Rules), Section 3, Chapter 3.2, and it can only act through a circuit breaker, not through a fuse. Protection by high-voltage fuses (type PKT) is applicable only for low-power transformers — under established engineering practice, roughly up to 630 kVA. So for a powerful 1 MW plant transformer, “lightweight” fuse protection with a load-break switch is selected outside the standard.

How much this costs the owner in money

The exact figure has to be calculated for the specific site — it all depends on how badly the component is undersized and how many sun-hours the site gets. But the logic is clear even without exact numbers.

Curtailment cuts off the peak of generation — the largest volume of kilowatt-hours the plant could have delivered during the sunniest hours. Under a fixed tariff or a PPA, the full value of that energy is lost. And on the day-ahead market (DAM), this “cheap” midday energy is exactly the resource that should be charging battery energy storage systems (BESS) for an evening sale at the price peak.

If the plant has no storage, this surplus simply has nowhere to go — the energy is not generated at all. In other words, you don't just lose the current sale, you also deprive yourself of the ability to balance the grid and earn extra revenue on evening prices. On top of that, the shortfall isn't a one-time event — it's a daily effect over the plant's 20–25 years of service.

That's exactly why this “hidden” flaw is worth quantifying before you buy the plant, while you can still negotiate or walk away.

Checklist: 11 points where a contractor cuts the solar power plant budget

Points worth checking in the project and the commercial proposal before signing the contract.

What we've seen firsthand:

Transformer rating “right at the limit”: Check the ratio of plant capacity to transformer rating. If the transformer is loaded to ~100% at peak, that's a mistake — there should be a margin (~80% loading). Why it's money: otherwise the plant will have to be curtailed at peak.
Transformer protection by fuses instead of a circuit breaker: A powerful 6/10 kV transformer (above ~630 kVA) needs a circuit breaker (vacuum type) with relay protection. Why it's money: in a fault, weak protection won't save an asset worth millions.
Undersized cable cross-section: A thinner cable is cheaper — but causes greater losses and voltage drop. Why it's money: part of the energy heats the cable instead of reaching the grid.
Cutting corners on grounding: A grounding loop built to the bare minimum. Why it's money: risk to people and to expensive equipment during a fault or lightning strike.
Cheap support structures: Cheapened steel structures / piles under the modules. Why it's money: tilted modules underdeliver generation, and in the worst case the structure fails.

What's common in the industry (also worth checking):

Weak surge protection (SPD): Surge protection is standardized under the IEC 61643 series. Why it's money: without proper SPD, a single serious surge can knock out the most expensive electronics on the plant — the inverters.
Simplified lightning protection: A large open-field site without proper protection. Why it's money: a direct strike means repairs at your expense and downtime.
No monitoring / dispatch system: Why it's money: in a proper monitoring system, curtailment shows up as a clipped generation curve — a flat “plateau” instead of a proper solar bell curve. If such a plateau isn't part of the design (a standard inverter limitation from DC/AC oversizing), your plant is physically cutting off your money right now. Without a dispatch system, a site can run in that plateau mode for months and you won't even know it. Ideally, network analyzers are installed at every substation (transformer station) plus one for the plant overall — only by comparing their readings can you see that every part of the plant is delivering full power, rather than one part quietly underdelivering.
Cutting costs on inverters: no-name brand or a single “all-in-one” unit: Cost-cutting via a no-name brand instead of tier-1 — or one large inverter instead of several smaller ones. Why it's money: a no-name unit degrades faster and without warranty; and when the whole array is tied to one large inverter, its failure shuts down a large chunk of the plant at once (for comparison: one 333 kW inverter versus three 110 kW units — losing one of three costs you a third, losing the single large one stops everything) — and a spare part for a cheap brand can take months to arrive from China.
Warranty “on paper” and formal commissioning tests: A warranty only works if there's someone in Ukraine to honor it. Why it's money: a “warranty” nobody will actually fulfill is worth zero.
No maintenance after commissioning: Who will service the plant for 20–25 years, and for how much — panel washing, monitoring, fast fault repair. Why it's money: without service, panels get dirty and produce less, and when something breaks, the plant sits idle and earns nothing while you scramble to find someone to fix it.

What if the plant is already built?

If you're reading this as the owner of a plant that's already underdelivering — there's good news: most of these mistakes are fixable. An undersized transformer can be replaced with a larger one, protection can be brought up to standard, and SPD and monitoring can be added.

A technical audit of the plant answers the question of whether it's worthwhile. We assess the equipment's actual performance, estimate roughly how much output you're losing to curtailment and losses, and show exactly what's worth fixing first — so that the investment in the fix pays for itself through recovered generation.

How to tell a trustworthy contractor apart

A contractor's integrity is visible as early as the proposal stage — in how much detail they show for the engineering decisions, not just the final price.

Signs of a contractor who isn't hiding cost-cutting:

there's a single-line diagram with a specific protection type and ratings;
there's an equipment specification with manufacturers and models;
there's a loss calculation and justification for cable cross-sections and transformer rating;
a margin is built into the transformer sizing, along with full protection, SPD, and monitoring;
the contractor is willing to calmly explain every decision.

The cheapest bid with a “bare” price and none of these details is almost always a bid where the savings are already baked in where you won't see them — until the day the plant starts underdelivering.

Frequently asked questions

What transformer is needed for a 1 MW solar power plant?

For a plant of around 1 MW, a 1250 kVA transformer is used, not a 1000 kVA one. A margin of ~20% provides thermal headroom: at midday peak, the transformer runs at ~80% of its rating and doesn't overheat. The loading logic is described in IEC 60076-7. A transformer sized “right at the limit” forces the plant to be curtailed at peak.

Why is protecting a transformer with fuses dangerous?

A load-break switch with fuses is not designed to interrupt short-circuit currents. For a powerful 10 kV transformer (above ~630 kVA), the PUE (Section 3, Chapter 3.2) requires relay protection through a vacuum circuit breaker. Cutting costs here puts an asset worth millions at risk.

What does plant “curtailment” mean?

It's a forced limitation of generation when a weak link (transformer, cable) can't pass full power. The inverters get a command to clip the peak of generation — on a monitoring chart this looks like a flat “plateau” instead of a solar bell curve. The energy simply never reaches the grid or the storage systems.

Can an already-built solar power plant that's underdelivering be fixed?

Yes, most mistakes are fixable. An undersized transformer is replaced with a larger one, and protection is brought up to standard. A technical audit of the plant calculates how much output is being lost to curtailment and shows whether the fix will pay for itself through recovered generation.

How much does an undersized solar power plant component cost the owner?

The exact figure is calculated per site. Curtailment clips the largest volume of kilowatt-hours during peak hours. It deprives you of the ability to store energy in BESS for evening sale and creates daily losses throughout all 20–25 years of the asset's service life.

An independent look at your project or plant

If you're at the stage of choosing a contractor, you can send us your project or commercial proposal for a free preliminary technical review. If the plant is already operating and you suspect it's underproducing, we'll carry out a technical audit of the site.

The preliminary project review is informational, provided at the company's discretion, and is not an offer, a guarantee of identifying all risks, or a recommendation regarding any specific contractor. A full assessment is only possible based on the results of an inspection.

LK Energy builds turnkey solar power plants and takes on maintenance of operating sites — including those that came to their owners with problems. We see these mistakes from the inside, because we're the ones who end up fixing them.

Send your project or commercial proposal for a free preliminary review →

Related equipment: 4.95 MW solar power plant with an energy storage system.

A Cheap Solar Power Plant Bid: Where the Contractor Cuts the Budget — and Why the Plant Underdelivers Power at Peak for Years