A chiller plant that used more energy at night than during the day

Starting point

The operator was asked to approve a plant upgrade. Without a baseline set beforehand, any savings claim afterwards is a subtraction of this year from last, dominated by whatever else changed. The plant’s own meters and the site’s occupancy schedule were the data available.

Analysis

The baseline models the plant’s energy use from a weather driver and an occupancy schedule, over measured intervals. Intervals the meter did not report are excluded and counted. They are not interpolated and not zero-filled.

Why it stayed invisible

1. Consumption was shown against time, never against the schedule.
2. Totals were reviewed monthly, and at monthly grain an out-of-hours problem sits inside every data point.
3. Nothing alarmed. The plant ran within every configured limit, and no limit describes running when nobody needs it.

Finding

76.4% of the plant’s measured energy use fell outside the building’s occupied hours. Night-time use was 158% of day-time use: the plant used more energy when the building was empty than when it was occupied.

Some cooling outside occupied hours can be legitimate, so this is a finding to investigate, not a verdict on cause.

Action and outcome

The operator now has the plant’s operating pattern fixed before the upgrade, so the upgrade’s effect can be measured against it. A schedule review is the first question the finding raises. No savings are claimed.

Client anonymised pending publication sign-off. The two percentages are measured values from the engagement.