The Second Meter Rule: Proving Voltage Optimisation Savings Before You Buy

A voltage optimisation unit is, in essence, a transformer with a fixed or selectable tap that holds site voltage below the incoming supply level. The principle is sound: resistive loads draw less power at lower voltage, and the arithmetic is well understood. The difficulty is not the physics but the number attached to it. Savings quotations of 5–15% are routinely presented to UK businesses as though they were a property of the product rather than of the site, the load mix and the metering arrangement it sits behind.

For a finance director signing off a five-figure capital sum, the useful question is not whether voltage optimisation can save energy. It is whether it will save energy here, at this site, on this supply, in a quantity that survives contact with an independent meter.

Why the Physics Is Correct and the Percentage Rarely Is

Lower voltage reduces consumption in loads that behave resistively or draw current in proportion to voltage. Legacy lighting with magnetic ballasts, resistive heating elements and older fixed-speed motors fall into this group. Many modern loads do not. Switched-mode power supplies, LED drivers, inverter-driven chillers and variable speed drives are largely constant-power devices: as voltage falls, current rises to compensate, and kWh consumption at the meter barely moves.

On a typical UK commercial or light industrial site, constant-power loads now account for a substantial share of consumption. A building that has replaced its lighting, installed inverter drives and added IT load is a very different proposition from the site on which the original 15% claim was validated. Before any quotation is discussed, establish the load inventory and estimate the split between voltage-sensitive and constant-power equipment.

Begin With the Incoming Supply, Not the Sales Deck

The first piece of evidence is the voltage actually arriving at the point of common coupling. Log it at one to ten minute resolution for a minimum of four weeks, including weekends and any shutdown periods, and compare the distribution against the BS EN 60038 tolerance band of 230V +10%/-6%.

Two practical points frequently go unexamined. First, ask the distribution network operator whether the supply transformer tap has already been adjusted; a site that has been optimised at the substation has less headroom to recover. Second, note the season. Supply voltage tends to run higher in lighter-load periods, so a trial conducted in spring or summer will typically flatter the saving. A baseline that cannot be compared with a winter period is a weak foundation for a payback calculation.

The Metering Arrangement Decides Whether the Claim Is Provable

Voltage optimisation savings must be demonstrated at the supplier's meter, not at a downstream sub-meter. Several structural issues follow from that:

• Position in the topology. If the unit is installed downstream of the half-hourly meter, savings appear on the bill. If it sits upstream of a landlord supply that is recharged to tenants, the consumption reduction accrues to the party paying the supplier, while the recharged sub-meters show the change. Clarify who owns the benefit before the capital is committed.
• Measurement uncertainty. A 1.5% saving sits within the combined error of many class 1 current transformer and meter combinations. Independent verification generally requires a reference meter of class 0.5S or better, installed for the trial period.
• Existing power factor correction. Capacitor banks and the optimiser interact. Where power factor correction is present, the effect of the unit on reactive power must be measured rather than assumed.

Separating kWh From kVA

Voltage reduction lowers kWh in voltage-sensitive loads, but current rises in constant-power loads, so apparent power can move in either direction. That matters because a proportion of a business energy bill is not energy at all. Capacity and availability charges under distribution use of system tariffs, and any reactive power charge levied on the site, respond to kVA rather than kWh.

A vendor may present a reduction in peak kVA as an opportunity to reduce agreed capacity. That is a legitimate benefit, but only where the reduction is measured during the charging window the tariff actually uses, and only where the supplier or distributor will accept a capacity variation at the point in the year when it is permitted. Treat capacity release as a separate line in the business case, evidenced separately, never as a supporting assumption inside an energy-saving figure.

Interactions That Quietly Erode the Case

• Electric cooling. Reduced voltage raises running hours in compressor motors working to hold a setpoint. In air-conditioned buildings the summer net saving can be materially smaller than the headline figure, and occasionally negative.
• Thermostatically controlled heating. Savings may appear as longer operating hours rather than lower consumption, particularly where controls maintain a fixed internal temperature.
• Harmonics. Electronically controlled optimisers can amplify existing distortion. Measure total harmonic distortion before and after, and check that the unit's specification is compatible with the site's existing equipment.
• Refrigeration. Compressor efficiency curves are not linear with voltage; a trial is the only reliable test.

Claims That Should Prompt a Trial

An independent metering period costing a few hundred pounds is trivial against a five-figure installation, and is the appropriate response to any of the following: savings presented as a percentage of spend rather than of kWh; evidence drawn from a single day or from vendor-supplied readings; a guaranteed figure offered without an agreed measurement plan; assertions that the technology suits all loads; resistance to an independent measurement and verification protocol; quotations that exclude switchgear, bypass arrangements, installation or maintenance; and any proposal that leaves unclear who owns the asset at lease end.

A Workable Trial Protocol

Agree the method in writing before signature. A defensible approach follows four steps.

1. Baseline. Four weeks minimum, at the meter, with weather and occupancy data captured alongside.
2. Post-installation. Four to six weeks, ideally crossing a seasonal boundary, with the unit in bypass for part of the period where a bypass is fitted. An A/B/A sequence is the strongest available design.
3. Normalisation. Adjust for heating and cooling degree days and for production or occupancy volume, so the comparison is like for like.
4. Reporting. Express the result in kWh and in pounds at the contracted unit rate, with the Climate Change Levy shown separately, and state the measurement uncertainty alongside the saving.

The IPMVP Option B framework, retrofit isolation with continuous measurement, is well suited to this exercise and is widely understood by independent verifiers.

Building a Business Case That Holds

Model the payback on verified first-year kWh savings, present the capacity and reactive power effects as a distinct and clearly evidenced line, and run the arithmetic at two, six and ten per cent rather than at the vendor's single figure. Include maintenance, warranty terms and a realistic asset life, and consider whether electrification of heat or the addition of on-site generation would strand the unit partway through that life.

The second meter is not a bureaucratic add-on to the purchase. It is the purchase decision. A supplier confident in the product should welcome it; one that does not is telling you something useful about the number.