Sectors
Energy operations for industrial sites
How industrial and light-manufacturing sites can use their half-hourly data, capacity position and load flexibility to cut energy cost structurally.
Last updated
Industrial sites are the most measurable energy users in the economy and among the least managed. Almost every site of scale is settled half-hourly, meaning years of interval data already exist for it, describing exactly when it consumes and what that consumption should cost. In most businesses nobody has ever analysed that data. Energy is bought at renewal through a tender or a broker, the rate is filed, and the structure underneath, which is where an industrial site’s real money sits, goes unexamined for another contract term.
The structure matters more in industry than anywhere else because industrial load has shape and size. Time-banded network charges, capacity charges and wholesale price patterns interact strongly with production schedules, shift patterns and process loads. Two sites with the same annual consumption and the same unit rate can have materially different fair costs. A tender cannot see any of this. The half-hourly data sees all of it.
The levers beyond the unit rate
Four structural levers recur across industrial sites. Tariff structure: whether flat, time-of-use or pass-through pricing fits the production pattern, and what the flat-rate risk premium is costing if the shape is favourable. Network charges: consumption sitting in expensive distribution bands that scheduling could move. Capacity: agreed capacity that drifted from actual maximum demand years ago, in either costly direction. Baseload: out-of-hours consumption that no process explains. Each is invisible on a bill and obvious in the interval data.
One site with a large continuous load and a 930 kVA agreed capacity shows the pattern well. The summary bill reads as a single blended rate, but the half-hourly data tells a different story. DUoS red-band charges concentrate in the weekday 16:00 to 19:00 window, at 1.764 p/kWh against 0.011 p/kWh in the green band, roughly 160 times the price for the same unit of electricity. A daily capacity charge of 9.26 p/kVA applies to the agreed capacity whether the site uses it or not. Neither line is something a manager would ever question on a monthly bill. Both are obvious the moment the interval data is read against the network operator’s time bands.
Flexibility is an asset; treat it like one
Any load that can move in time has financial value, and industrial sites hold more of it than they think: compressors, cold storage, thermal loads, batch processes. Against time-of-use pricing or day-ahead pass-through, that flexibility becomes revenue-relevant, and storage can mechanise it. Market developments continue to widen what demand-side flexibility can earn. The sequence is fixed though: measure the flexibility first, monetise it second. Assets bought before the load is understood are how oversized batteries happen.
Three tariff structures are where flexibility actually gets paid. Network charges are banded by time of day, and shifting load out of the Red band, and away from the handful of winter peak periods that set Triad and Capacity Market charges, converts a fixed cost into a variable one the site can manage down. A day-ahead pass-through tariff goes further: it exposes the site to the real half-hourly wholesale price, so flexible load earns its keep every day of the year, not just on the days that set network charges. Beyond the site’s own tariff, National Grid ESO’s flexibility markets and the wider Balancing Mechanism pay directly for load that can move or assets that can respond, a revenue line that exists in principle but has to be sized against real, metered flexibility before it is worth entering.
Each of these routes rewards a different kind of flexibility, some day-ahead, some near-real-time, and a site rarely has enough of one type to chase every route at once. Valuing the flexibility correctly means matching what the load can actually do to the structure that pays for that specific behaviour, not defaulting to whichever structure a supplier happens to offer.
Generation on real numbers
Industrial roofs and land make solar an obvious question, and daytime process load often makes the self-consumption case strong. The discipline is the same as everywhere: size from the metered load profile, secure the G99 connection position early, and keep the model independent of the party selling the hardware.
An industrial site’s half-hourly history is an audit waiting to be read. Request a benchmark at /benchmark or book a review at /book.
Frequently asked questions
We already went out to tender. Isn't that energy dealt with?
A tender addresses one variable: the commodity rate. For an industrial site, the larger levers often sit elsewhere: tariff structure against your production pattern, time-banded network charges, agreed capacity versus actual maximum demand, and whether flexible loads run in the wrong periods. A site can win its tender and still carry substantial avoidable cost, because procurement alone never touches the structure.
What is an agreed capacity review and why does it matter?
Half-hourly sites hold an agreed capacity with the network operator and pay for it whether used or not, with penalties for exceeding it. Production changes over the years, but capacity rarely gets revisited, so sites end up paying for headroom they no longer need or courting excess charges they do not see coming. Comparing agreed capacity against actual maximum demand from the half-hourly data is one of the cleanest checks available.
Does load shifting mean disrupting production?
Rarely. The candidates are the loads with natural slack: compressed air, refrigeration and cold storage, water heating, charging, and any batch process with timing freedom. Shifting these out of peak price and network bands leaves the production schedule untouched. Where genuine flexibility exists at scale, storage and market-facing tariff structures extend the same logic further.