Renergia
EN RO
← All case studies
Banking February 2026

Converting abnormal HVAC operation into measured savings

At a second commercial banking space, Renergia's AI analysis detected abnormal HVAC operation outside the required schedule. The client accepted the automation recommendation — and the measured, weather-adjusted result was a 60.5% reduction in out-of-hours consumption.

2,526 kWh measured savings Observed in the 19–28 February window, adjusted for weather
€808 measured savings Financial result of the observed ten-day window
60.5% observed reduction Of total load in the 19:00–07:00 window
38,142 kWh seasonal potential Orientative extrapolation for November–March · €12,205

Context

The case covers a real, anonymized commercial banking space with no on-site photovoltaic production, so every kilowatt-hour saved directly reduces imported energy. The analysis focused on HVAC operation and out-of-schedule consumption in the cold-season profile.

Renergia used total load, HVAC load, and hourly outdoor temperature from February 2026 monitored data. The automation was implemented on 18 February, and direct savings were measured over the confirmed low-occupancy window from 19 to 28 February.

18 Feb 2026

Automation implemented

19–28 Feb

Measured savings window

No PV

Savings reduce imported energy directly

What Renergia found

Renergia's AI-based automated analysis compared the real hourly consumption profile with the expected technical reference profile, adjusted for outdoor temperature — and detected heating operation continuing well outside the required schedule.

19:00–07:00 The window with the strongest opportunity, when the building should operate close to its technical minimum.

The disaggregation algorithm isolated the HVAC component from the total load, making visible the consumption most likely driven by heating operation outside the required schedule.

How the issue was identified

The analysis separates the effect of weather from the effect of HVAC automation: colder hours are accounted for before attributing any savings to the measure.

01 Total load, HVAC load + outdoor temperature
02 Compared to temperature-adjusted reference
03 Out-of-schedule heating operation detected
04 AI disaggregation isolates HVAC component
05 HVAC automation implemented 18 Feb

What we did

Renergia proposed HVAC automation through the platform. The client accepted the recommendation and the measure was implemented from 18 February 2026, targeting the 19:00–07:00 low-occupancy window.

The direct savings calculation deliberately excludes two operational windows:

Pre-heating

The morning interval before the space opens

After office

The transition period after normal hours

These periods may still be operationally justified, so they were not counted as direct savings.

Expected vs. real

In the observed 19:00–07:00 window, the temperature-adjusted comparison shows the measured effect of the automation.

Expected without automation 4,172 kWh

Technical reference consumption expected in the observed window under the recorded outdoor temperatures.

Real after automation 1,646 kWh

Measured consumption after implementation — 2,526 kWh saved in ten days.

Measured results

The observed window delivers a measured result; the seasonal figures are an orientative extrapolation based only on the February profile, without invented temperatures for other months.

Observed window (19–28 Feb) 2,526 kWh · €808

Total load, 19:00–07:00, adjusted for weather. Avoided emissions: 0.44 t CO₂.

HVAC component 795 kWh · €254

Technical confirmation of the measure — an 88.6% HVAC reduction in the main window. 0.14 t CO₂.

Full cold season (Nov–Mar) 38,142 kWh · €12,205

Orientative extrapolation for the five-month period. 6.60 t CO₂.

Average cold-season month 7,628 kWh · €2,441

Monthly scenario derived from the February profile.

Financial values use an exchange rate of 5 RON = 1 EUR.

Why it matters

Renergia converted raw energy data into a measurable operational result. The AI-based analysis detected abnormal HVAC consumption, the disaggregation algorithm isolated the HVAC component, and the implemented automation reduced unnecessary out-of-schedule operation.

Because the savings are measured against a temperature-adjusted reference, the result reflects the automation itself — not a milder week of weather.

Want results like these in your buildings?