When power prices change by the hour, your building can respond

Why timing now matters

Electricity is cheapest and cleanest when wind and solar are plentiful, and most expensive when the grid is stretched. More and more tariffs pass that through. Spain's regulated tariff sets a new price every hour, many business contracts in Europe and the UK are time-of-use or indexed to wholesale prices, and US commercial bills often include charges for the single highest peak of the month.

The EU recognises this too. Its Smart Readiness Indicator, an optional scheme for rating buildings, scores how well a building can "adapt to signals from the grid" alongside efficiency and comfort.

How a building shifts its load

A building stores heat and cold in its structure, its water and, in a store, its stock. That gives it room to move:

• Pre-cool or pre-heat before an expensive period, then coast through it.
• Let temperatures drift slightly, within an agreed comfort band, during a short peak.
• Move flexible loads such as heat pumps, hot water and EV charging into cheaper hours.
• Use the cold you've stored. Cold rooms and freezers can pull down a little further before a peak and ride through it, always inside food safety limits.

None of this needs new equipment if the controls are sound. It's the same work as any tune-up: rules, schedules and setpoints, with price and grid signals added as inputs.

Fix the basics first

Grid response only works on a building that already behaves. If heating and cooling are fighting, sensors are reading wrong or overrides are everywhere, adding price signals just adds complexity. And because responding to the grid means another connection into your controls, that connection has to be secure from day one.

That's why we treat it as part of performance, not a separate product: secure the controls, fix the rules, then let the building respond, and keep checking that comfort and safety limits hold.

The same idea at a larger scale

Data centres are the extreme case. New sites often wait years for a grid connection, because networks are planned for a few hours of peak demand a year. A 2025 Duke University study estimated that the US grid could absorb about 76 GW of new load if those loads could reduce their demand for just 0.25% of the year, about 22 hours, in events averaging under two hours.

Doing that well means cooling, power and computing schedules working together, which today are usually run by different teams with different tools. It's the same principle as a building that pre-cools before a peak, with far more at stake, and the same work we do for data centres: connect the layers, fix what's there, and keep checking. See data centre services.