Reactive Loads, Real Money: Diagnosing Power Factor Charges on UK Business Bills

Most business energy reviews start and end with the unit rate. That is understandable: pence per kilowatt-hour is visible, comparable and easy to put in a board pack. But for many UK sites, the more revealing number sits beside it on the invoice: maximum demand in kVA, and the reactive power charge that follows poor power factor.

Power factor is the ratio of useful power (kW) to apparent power (kVA). A site with a power factor of 0.80 draws 125 kVA to do 100 kW of useful work. The same site corrected to 0.98 would draw roughly 102 kVA. The kilowatt-hours might barely change. The capacity requirement, current flows, losses and network charges do.

Why identical kWh can produce different bills

Imagine two manufacturing sites with the same annual consumption and the same production pattern. Site A runs at a power factor of 0.97; Site B at 0.78. If both peak at 200 kW, Site A needs about 206 kVA, while Site B needs about 256 kVA. Where a supplier or distribution charge is based on kVA maximum demand, Site B is paying for capacity it is not using productively.

The gap widens when reactive power is charged separately. Many half-hourly supply arrangements include a reactive power charge once kVArh exceeds a defined proportion of kWh - commonly around 33%, equivalent to a power factor of about 0.95. A site can be comfortably within its kWh budget and still trigger that charge because motors, pumps, refrigeration, compressors, fans and uninterruptible power supplies are drawing reactive current.

The result is not a billing error. It is a pricing signal. Two sites with similar production can face different bills because one has managed its reactive load and the other has not.

Reading the evidence in half-hourly data

A half-hourly meter is the most useful diagnostic tool available. Pull at least twelve months of data and compare kWh with kVAh for each half-hour. Power factor is kWh divided by kVAh. Do not rely on a single monthly average; averages hide the peaks and the quiet periods that drive cost.

Look for three patterns. First, a power factor that falls sharply at low load. That often points to fixed reactive load - lightly loaded motors, old lighting ballasts, idling pumps or a fixed capacitor bank that is too large for the current profile. Second, a power factor that deteriorates as load rises. That suggests variable reactive demand from motors or drives and may favour automatic compensation. Third, a power factor that is stable but persistently low. That is usually a sign of an underlying design issue rather than a single faulty item.

Then reconcile the data with the invoice. Check for lines described as reactive power, kVArh, kVA maximum demand, capacity charge or availability charge. Compare the charging period with production records. A month with low output but high kVA demand is a warning sign: something is drawing reactive current even when the site is not busy.

Correction options and their trade-offs

Capacitor banks remain the most common correction method. Fixed banks suit stable loads; automatic stepped banks switch capacitors in and out as demand changes. They are generally the lowest-capital option, but they need a proper survey. If the site has significant harmonic distortion - common with variable-speed drives, rectifiers and modern lighting - standard capacitors can resonate with the supply and create more problems than they solve. Detuned capacitor banks or active filters may be required.

Active filters are more flexible. They can correct power factor dynamically and reduce harmonic distortion, which makes them attractive for sites with fast-changing loads. The trade-off is capital cost, complexity and maintenance. They are rarely the first answer for a simple motor-dominated site, but they can be the right answer where harmonics and power factor are tangled together.

A word of caution: overcorrection is a real risk. Leading power factor at light load can trigger penalties or operational issues. Most designs target a power factor of around 0.95 to 0.98 rather than 1.00. Correction should be sized against the actual load profile, not a nameplate total.

Calculating a realistic payback

The payback calculation should be built from the contract, not from a generic savings table. Start with the charging basis. If capacity is priced in kVA, calculate the kVA reduction at the site's peak. If reactive power is charged on kVArh above a threshold, calculate the chargeable kVArh avoided. Then add an estimate for reduced losses in cables and transformers, which is real but often modest.

For example, a site peaking at 250 kVA and running at a power factor of 0.82 might correct to 0.97, reducing peak demand to roughly 211 kVA. At a capacity charge of £90 per kVA per year, that is about £3,500 saved. Add £1,200 for avoided reactive charges and losses, and the annual benefit is £4,700. If the installed cost is £12,000, including survey, switchgear, installation and commissioning, the simple payback is around 2.6 years. That may be attractive; it may not be, depending on the remaining contract term and the site's operating life.

Use net capex, not gross quote, and include maintenance. If the supplier charges capacity in kW rather than kVA, the case may rest mainly on reactive charges and losses. That is not a reason to dismiss it, but it is a reason to model it honestly.

A practical sequence

Begin with data, not capital. Extract twelve months of half-hourly kWh and kVAh, audit the invoice for every demand-related line, and commission a power quality survey before specifying equipment. Model the saving against the actual tariff, then ask for a fixed-price installation quote with a clear scope. If the payback is marginal, consider a phased correction or a smaller automatic bank rather than a site-wide solution.

Power factor correction is not a substitute for switching off idling plant or repairing oversized motors. It is a targeted fix for a measurable cost. The businesses that benefit most are those that treat kVA as a managed cost rather than an unavoidable line on the bill.