Cut HVAC Runtime by 35% Without Comfort Loss

Most buildings run longer and harder than they need to. The trick is to reduce runtime without creating hot/cold complaints. Here’s a playbook we use to cut HVAC runtime by ~35% while keeping (or improving) comfort.

The 35% formula

• Tighten schedules to match true occupancy and use adaptive start/stop.
• Widen deadbands and apply intelligent set‑back/setup outside core hours.
• Align ventilation with people using CO₂‑driven demand control and economiser logic.
• Tune sequences and resets (SAT, duct static, CHW/HW temps) to avoid over‑delivery.
• Fix the silent killers: simultaneous heat/cool, overrides, stuck dampers/valves.

Scheduling that fits reality

Align plant operation with actual occupancy, not the imagined 8–6. Use badge data, PIR counts, or booking calendars to set core hours and apply lean wings/zone schedules.

• Pre‑start only what’s needed for the first occupied zones.
• Staggered shutdown by floor/wing to avoid late‑evening drift.
• Holiday/weekend calendars baked into the schedule source of truth.

Adaptive start/stop

Let the building’s thermal behaviour do the work. Adaptive logic learns warm‑up/cool‑down times per season and starts plant just in time, then stops early while zones coast within comfort.

• Use outside air temp and recent ramp rates to predict start time.
• Stop plant when the coast window (time to drift to limit) exceeds remaining hours.

Smart setpoints and deadbands

A too‑tight deadband forces short cycling and simultaneous heat/cool. Relax it slightly in occupied hours, and widen out of hours.

• Occupied: 21–24°C cool, 20–22°C heat (example), humidity 40–60%.
• Unoccupied: setup/setback +2–4°C with alarms if drifting too far.

Ventilation aligned to people

Deliver fresh air when people are present, not as a fixed percentage.

• DCV using CO₂/TVOC/occupancy for AHUs and large RTUs.
• Economiser logic to favour free cooling within humidity limits.
• Min‑OA trims during warm‑up/cool‑down to avoid conditioning empty air.

Sequence tuning and resets

Right‑size delivery instead of blasting full power then throttling.

• Supply air temp (SAT) reset with load to reduce compressor time.
• Duct static reset to the most open VAV, saving fan kWh.
• CHW/HW temp resets with outside conditions to optimise plant COP.
• Pump/VFD optimisation via differential pressure trims.

Find and fix the silent killers

These issues quietly extend runtime and waste energy:

• Simultaneous heat/cool from bad sequences or failed valves.
• Permanent overrides that defeat schedules.
• Stuck dampers/actuators spotted via command vs. response mismatch.
• Biased sensors (NTC drift) causing false calling.
• Short cycling due to tight deadbands and aggressive PID.

Comfort safeguarding

Prove no comfort loss with transparent KPIs.

• % hours within target band per zone and building.
• Complaint log correlated with setpoint deviations.
• Thermal comfort using operative (radiant + air) when sensors allow.

Comfort and savings verification

Track zone conditions and apply weather‑normalised M&V so finance trusts the numbers.

• Before/after runtime and kWh with confidence bands.
• Attribution by measure (schedules, DCV, resets).
• Evidence packs for auditors and stakeholders.

“We cut weekday runtime by 32% and weekends by 58%—comfort scores went up.”

Case study: 35% runtime cut in a 10,000 m² HQ

• Week 1–2: schedule audit, adaptive start, weekend modes.
• Week 3: DCV enabled, economiser fixed, SAT/duct static resets.
• Week 4–6: fix overrides, deadband tuning, stuck valve replaced.

Outcome: −35% runtime, −18% kWh HVAC, zero comfort penalties.

30‑day runtime reduction plan

1. Map schedules vs. occupancy; implement quick trims.
2. Enable adaptive start/stop and widen deadbands.
3. Turn on DCV and validate economiser logic.
4. Apply resets (SAT, static, CHW/HW); monitor responses.
5. Fix faults surfaced by alerts; export M&V report.

Related reading

• The Real Cost of Ignoring HVAC Optimisation
• Integrating BMS, Meters and IoT for Smarter Energy Management
• Five Quick Wins for Multi‑Site Energy Managers
• Anomaly Alerts Ranked by Cost Impact

Talk to an optimisation expert