How to Reduce Peak Demand in Critical Facilities

A single 15 or 30-minute interval can materially affect an electricity bill for an entire month. For data centres, industrial sites and major commercial facilities, learning how to reduce peak demand is therefore not simply an energy-efficiency exercise. It is an electrical design and operational discipline that must protect continuity of supply while preventing avoidable demand spikes.

Peak demand is usually driven by coincident loads: equipment that starts, recovers or operates at the same time. Chillers restarting after an outage, multiple air-handling units responding to high ambient temperatures, battery charger recovery, process motors, electric heating and IT load growth can combine into a short but expensive maximum-demand event. The right response depends on the tariff, the facility load profile and the operational consequences of deferring or reshaping a load.

Understand what the meter is measuring

Before selecting equipment or setting a demand target, confirm how the site is billed. Network tariffs may assess maximum demand in kW, kVA or, in some cases, both. The averaging period, seasonal windows, time-of-use rules and ratchet provisions all matter. A site that reduces a five-minute spike may see no financial benefit if the tariff records the highest 30-minute average.

This distinction is particularly important where poor power factor or harmonic distortion is present. Real power, measured in kW, represents useful energy consumed by the load. Apparent power, measured in kVA, also reflects reactive and distortion-related current. If billing is based on kVA, improving power factor can reduce recorded demand without changing the productive work performed by the facility.

Demand data should be reviewed at an interval fine enough to identify the event that creates the monthly maximum. Utility interval data is a useful starting point, but it may need to be supplemented with switchboard metering, power-quality analysis and equipment operating logs. The aim is to identify not only the peak value, but also the combination of loads that caused it.

Establish a credible peak-demand baseline

A practical study begins with at least 12 months of interval demand data where available, including the highest-demand day in each season. Overlay this with weather data, production records, maintenance activity, generator tests, UPS battery recharge periods and building management system alarms. Peaks often have an operational explanation that is invisible in a monthly bill.

At the main switchboard and major distribution boards, measure phase current, voltage, kW, kVA, power factor, total harmonic distortion and maximum-demand intervals. Phase imbalance also deserves attention. One heavily loaded phase can constrain transformer, cable or switchboard capacity before aggregate site demand appears excessive.

For critical facilities, classify loads into three groups: non-deferrable, deferrable and controllable. Life-safety services, critical IT loads and essential process controls are generally non-deferrable. Thermal storage charging, selected HVAC stages, electric vehicle charging, non-critical pumping and some process preparation loads may be scheduled or curtailed for short periods. The classification must be developed with operations teams, not imposed solely from an electrical drawing.

Reduce coincidence before adding capacity

The lowest-risk demand reduction is often better sequencing. Many facilities accumulate independent control systems over time, allowing equipment to start simultaneously after a voltage disturbance, planned shutdown or loss of communications. Staggered restart sequences, start permissives and time delays can prevent several large loads from appearing in the same billing interval.

Motor starting requires particular care. Direct-on-line starts can create high inrush current and a local voltage dip, while repeated starts across multiple motors can compound the event. Soft starters and variable-speed drives can reduce starting stress where appropriate, but their effect on billed maximum demand depends on the tariff averaging interval and the motor duty cycle. They should be assessed alongside harmonic performance and the capability of the upstream supply.

HVAC is frequently a major source of peak demand in commercial and data-centre environments. Resetting temperature setpoints within approved operating limits, sequencing chiller stages, pre-cooling thermal mass before a tariff peak window and preventing simultaneous compressor restart can all flatten demand. The acceptable operating envelope should be clearly documented, particularly where temperature and humidity affect IT equipment, product quality or worker safety.

Use controls to manage demand in real time

Demand management is most effective when it uses measured electrical conditions rather than static schedules alone. A demand controller can monitor site kW or kVA, compare the projected interval average with a defined threshold, and shed or defer pre-approved loads in stages. The objective is not indiscriminate load shedding. It is to make small, controlled interventions early enough to avoid crossing the target.

A staged strategy should specify the load priority, available reduction, maximum curtailment duration, recovery logic and operator override for every controllable circuit. It must also avoid rebound peaks. For example, releasing all deferred loads at the end of a demand window can create the next maximum-demand event. Controlled restoration is as important as controlled shedding.

Real-time monitoring and automated alerts allow facilities teams to investigate abnormal demand before it becomes routine. Advanced analytics can identify recurring relationships, such as a particular production line operating during chiller recovery, or UPS battery charging coinciding with morning building start-up. This provides a stronger basis for operating changes than relying on nominal connected load calculations.

Address power factor and power quality separately

Power factor correction may be an effective peak-demand measure when a tariff includes kVA demand or reactive-power charges. However, capacitor banks should not be specified from a simple site-wide power-factor snapshot. Loads with variable-speed drives, rectifiers, UPS systems and other power electronic equipment may introduce harmonic currents. In those conditions, conventional capacitor correction can interact with system impedance and create resonance or overstress components.

A power-quality assessment should therefore review harmonic spectrum, transformer loading, voltage distortion, existing correction equipment and expected future load changes. Detuned or harmonic-filtered correction solutions may be required, but the correct arrangement is site-specific.

Voltage regulation equipment has a different role. Automatic voltage regulators, voltage stabilisers and magnetic induction stabilisers are intended to maintain voltage supplied to sensitive or variable loads within an engineered operating range. They are not, by themselves, peak-demand reduction devices. In fact, a stable supply may enable equipment to operate consistently and can expose the true load rather than suppressing it through undervoltage.

Their value in a demand strategy is indirect but important. Stable voltage supports predictable motor, control and electronic equipment performance, helping avoid nuisance trips, uncontrolled restart events and inefficient operation caused by sustained voltage variation. AC power line conditioners may similarly support sensitive loads where electrical noise, transients or disturbance-related supply quality is a concern. Selection should follow measured site conditions and coordination with protection, earthing and bypass arrangements.

Consider storage and backup systems carefully

Battery energy storage can reduce demand by discharging during defined peak intervals, while standby generation may reduce grid import where permitted by site approvals, fuel arrangements and environmental requirements. Both approaches require more than a simple capacity calculation. The system must have sufficient power output for the target reduction, sufficient energy for the demand interval, a reliable control interface and a recharge profile that does not simply shift the peak to another period.

UPS systems require similar consideration. In a critical facility, the UPS is primarily a continuity asset, not a routine demand-management resource. Its battery autonomy, redundancy configuration, bypass arrangements and recharge behaviour must remain aligned with the site resilience strategy. Using critical backup capacity to chase marginal demand savings can create unacceptable operational risk.

Voltage drop compensators, variable transformers and frequency converters also have legitimate application roles where load voltage, test conditions or supply frequency must be controlled. Their use should be based on the requirements of the connected equipment, not assumed to reduce peak demand. Any change that affects load voltage or speed can alter process output, motor torque, thermal performance and ultimately site demand.

How to reduce peak demand without compromising resilience

The best projects treat demand reduction as a controlled operating constraint rather than a one-off equipment purchase. Start with tariff analysis and high-resolution measurement, then model the likely reduction from sequencing, controls, power-factor improvement and, where justified, storage. Validate the model during representative operating conditions and retain enough margin for extreme weather, production changes and contingency operation.

For Australian and New Zealand critical infrastructure projects, electrical consultants and facilities teams should also consider supply constraints, transformer thermal capacity, fault levels, generator coordination and planned expansion. A demand target that looks attractive on a spreadsheet may be inappropriate if it forces critical plant to operate too close to capacity or reduces recovery capability after an outage.

Ashley-Edison Asia voltage regulation and power-conditioning technologies can form part of an electrical system where voltage instability or power-quality issues are affecting equipment performance or reliability. OzGreen Energy provides access to Ashley-Edison Asia's voltage-control and power-protection solutions and can assist customers and project teams with product information, application considerations and technical support when evaluating suitable equipment for their requirements.

The most durable result is a facility that knows which loads create its peak, can respond before the billing interval closes, and maintains the voltage quality and continuity required for critical operations while doing so.