Evidence, anonymised

The practice is an NHS GP surgery. Like most sites its size, it treated energy as a bill to pay rather than a system to run. When the practice decided to invest in solar and a battery, it did something unusual. It modelled the decision before buying anything.

That order matters. Most sites buy the hardware first and discover the economics later. Vester ran the sequence the other way. Fix the tariff, size the solar to the load, then add storage. Each step informed the next.

We started with the tariff, because the cheapest saving needs no hardware at all. The practice sat on a legacy non-half-hourly meter. It restricted which tariffs the site could reach and hid the shape of demand entirely. An immediate price correction took the unit rate from 42p/kWh to 27p/kWh, over £7,000 saved before a single panel went up. We then upgraded the meter to half-hourly settlement. That opened access to time-of-use tariffs and, more to the point, gave us the data to design the solar and battery accurately.

Twelve months of half-hourly demand showed how much solar the site could genuinely use. We sized the array to that load, not to the available roof. Oversized solar exports its surplus cheaply and pays back slowly. A system matched to consumption keeps more of its generation on site, where it is worth most. We then ran a structured competitive process across pre-qualified installers, comparing panel and inverter specification, battery compatibility, performance modelling, warranty and baseline pricing. That process saved over £10,000 of capital on a £50,000 project.

The battery came last, and only because the data justified it. It stores daytime solar for use later, discharges during peak-price periods, and charges from the grid when power is cheapest. Tariff, solar and storage stopped being three separate purchases. They became one decision.

The result: annual electricity cost fell from £22,566 to £6,694, a reduction of 70%. The tariff correction contributed £7,553, the solar £6,677, the battery £1,641. Across gas and electricity together, the practice is on track to save over £18,000 in the first year.

“Vester helped us take control of our energy. We’re on track to save over £18,000 in the first year, which we’ve been able to put back into service provision.” The practice manager.

None of this required us to sell the practice anything physical. Vester earns nothing on the panels or the battery, so the modelling answers one question only. What does this site actually need. That independence is the point. A hardware seller sizes to the roof. An operator sizes to the load.

A multi-site commercial operator asked a simple question. Are we paying what we should be? It is a hard question to answer, because the portfolio sat on a bundled flexible rate. That single number folded wholesale cost, shaping, imbalance risk and supplier margin together. Nobody could benchmark it, because there was nothing to compare it against.

So we took it apart. Vester priced the portfolio’s actual half-hourly load against a day-ahead pass-through structure, where the rate moves with the disclosed market price plus a stated fee. Moving to it cut commodity cost by £86,779 a year, a reduction of 12.0%. No capital was spent. Nothing changed in how the sites ran.

The saving did not come from a keener rate. It came from a better structural fit. A bundled figure charges every site the same shape, whether or not that shape matches how the site consumes. Once the structure moved to a rate that tracks the market, the portfolio’s own demand pattern did the rest. The sites were already efficient. They were simply priced as though they were average.

Then we went further, using the same half-hourly data to model battery storage. Storage shifts consumption away from peak-price periods and holds cheaper power for when the sites need it. Tariff and battery, modelled on one dataset, reached a combined payback of around 3.3 years.

The lesson holds beyond this portfolio. A flexible rate is not automatically a competitive rate. It is often just an opaque one. When a business cannot see the components of its price, it cannot know whether that price is fair, and the market has no reason to volunteer the answer. Measuring the load first is what makes the market visible.