Energy flexing: measures, activations and how flexibility reaches the market

How Properate turns a building's technical systems into tradable flexibility: defining measures, the activation lifecycle, how delivery is measured, and how capacity reaches the market.

Energy flexing — dynamisk energibruk — makes a building's ability to shift consumption into something sellable. Ventilation, heating, cooling and electrical load can all tolerate a short nudge without anyone noticing, and the power system will pay for that nudge because it helps keep the grid in balance.

Properate turns that into two objects: measures, which describe what a building can do, and activations, which are the moments it actually does it.

Measures

A measure is one lever on one system — for example "reduce supply air temperature on AHU 3". Each measure records:

Potential — how much power the lever is worth, in kW. This can go in both directions: reducing consumption, or increasing it. Increasing is genuinely useful, because the grid sometimes needs load added rather than removed.

Maximum duration — how long the effect can be sustained.

Ramp-up time — how long it takes to reach the full effect.

Buffer — the time before a power change becomes measurable. This is a measurement property, not a control one, and it is why the buffer is excluded from the paid window.

Recovery time — how long until the setpoint is back to its pre-activation state.

Negative impact on the building — a severity score, shown as low, medium or high. Internally this is a 1–100 priority: 1–33 low, 34–66 medium, 67–100 high. It is what stops a cheap kW from being taken out of a system where the comfort cost is high.

Priority and system, so measures can be ranked and traced to the equipment they act on.

Availability

A measure is only usable once it is marked Available (Klargjort). This is the deliberate arming step: defining a measure describes a possibility, marking it available offers it up. Anything not marked available is documentation, not capacity.

The activation lifecycle

An activation has a start, an end and an activation rate — a percentage, so a measure can be called on partially rather than all-or-nothing. It moves through six states, each showing when the next one begins:

Starts — scheduled, not yet running.

Ramp up — moving towards full effect.

In progress — the paid window. The Time period shown on an activation is this window only; ramp-up, buffer and ramp-down sit outside it.

Buffer — waiting for the power change to become measurable.

Ramp down — returning towards normal.

Finished.

When an activation runs, Properate writes target setpoints and keeps a record of the original values, so the building is restored to exactly where it was rather than to a guess. Use Preview and the Timeline to see what an activation will do before committing to it.

How delivery is measured

Getting paid depends on proving the change happened, so the measures view shows consumption several ways at once:

Current usage — right now.

15-minute average — the settlement resolution the market works in.

Baseline and baseline potential — what consumption would have been without the activation, which is what delivery is judged against.

Activated versus total potential — what was actually called on against what was offered.

Online HAN ports — how many meter connections are currently reporting.

That last one deserves attention: historical usage can read low for periods when HAN ports were inactive. A baseline computed across such a period understates real consumption, which in turn understates delivery. Check HAN port coverage before drawing conclusions from a disappointing activation.

How flexibility reaches the market

Individual buildings are too small to bid on their own, so capacity is aggregated. Each building is registered as a flexibility asset — with its meter, its grid operator (DSO) and its location — and assets are collected into asset groups. Activations arrive at group level.

Properate reports each asset's available potential upstream, where it is pooled with capacity from elsewhere and offered into the market. When the market calls, the activation comes back to Properate automatically and the setpoint changes follow. Assets carry a status of enabled, opt-out or disabled, so a building can be held out of trading without deleting its configuration.

The markets involved are the system operator's reserve markets and local flexibility markets run by grid operators. Which specific products a given building's capacity ends up in is determined by the aggregator, not configured in Properate.

Where to find it

Flex automation now lives in its own standalone app, opened from Properate in a new tab. The previous module remains reachable as the legacy Flex module if you need it.

Good to know

Setpoints can only be driven where the underlying timeseries is marked writable — a measure pointing at a read-only setpoint will show nothing to control.

Flex automations are built on the same calculation-flow engine as virtual sensors and cloud automations, so they share frequency, execution window, plan/exceptions and health reporting. See the Virtual sensors article.

Errors surface as notifications on the flow itself rather than as building alarms, so check the last-run timestamp if activations stop behaving.