Before and after: measuring the savings from a lighting upgrade

Saying "the bill went down" is not enough after an LED upgrade. How to record a proper baseline, make a fair comparison and document the result.

A decision to upgrade lighting usually rests on one promise: energy use will fall. A few months after the project, management asks the obvious question: "So how much did we actually save?" To answer convincingly, measurement has to start well before installation. Reconstructing the old situation from memory afterwards almost never produces an accurate comparison.

Why the overall bill is not enough

The simplest approach is to compare the building's monthly electricity bill before and after. But dozens of factors besides lighting affect that number: air conditioning, heating, new equipment, changes in headcount, public holidays, a hot or cold season. If lighting is only part of total consumption, its change easily disappears in that noise. Sometimes the bill even rises despite real savings, because a new chiller or server room came online in the same period.

A reliable assessment therefore needs to see lighting consumption on its own.

Record the baseline

Before the upgrade, collect the following over at least several weeks:

• energy consumption of the lighting circuits, ideally with a dedicated sub-meter;
• the number and type of fittings and the actual power of each;
• how many hours a day the lights are on, zone by zone;
• illuminance at key workplaces, measured with a lux meter;
• the operating pattern: working days, shifts, closed days.

An important note on power: the rating printed on an old lamp does not always reflect real consumption. In fluorescent fittings the ballast also draws power, so measurements should be taken on the circuit itself. With a clamp meter, read the active power in watts, because simply multiplying current by voltage can differ from real consumption due to power factor.

Recording burning hours is more reliable than estimating them. A logging meter on the circuit or a simple light-and-time logger will do the job. It also helps to photograph each zone in its old state: if someone later argues that "this room used to be brighter", photos and a lux table are the most objective evidence.

Make the comparison fair

In the "after" period, repeat the same measurements under the same conditions. The comparison only means something if:

• the measurement periods are comparable, for example with equal numbers of working days;
• consumption is normalised per hour or per working day if operating hours have changed;
• the light level in the new system has not dropped below the old one.

That last condition is often forgotten. If the new lamps give noticeably less light, lower consumption is not a saving but simply less light. Lux measurements are therefore as important as energy measurements: the result should read "the same or better light for less energy".

In international practice these issues are systematised in methodological documents such as IPMVP (International Performance Measurement and Verification Protocol). The underlying idea is simple: establish the baseline, adjust for changing conditions, document the outcome.

A simple worked example

Suppose a warehouse lighting circuit runs 20 working days a month, 10 hours a day, and its measured average power before the upgrade is 4 kilowatts. That is 800 kilowatt-hours a month. If the new system, under the same operating pattern, measures 1.8 kilowatts, consumption drops to 360 kilowatt-hours a month, a difference of 440 kilowatt-hours. If motion sensors were installed at the same time, record how much operating hours fell as well, so that their effect can be seen separately.

These figures only illustrate the logic of the calculation. Each site's result comes from its own real measurements.

Typical mistakes and indirect effects

The most common errors are:

• calculating old consumption from nominal lamp ratings but new consumption from measurements;
• comparing summer months with winter months, although burning hours depend on the season;
• switching off some fittings during the upgrade and presenting that as an "LED saving" too;
• ignoring old lamps that had already failed, which artificially lowers the baseline.

It is also useful to record indirect effects. An LED system releases less heat into the room, which in an air-conditioned building can reduce cooling load. Less frequent lamp replacement lightens the maintenance workload. Reporting these effects separately, rather than mixing them into the lighting savings, gives a more honest picture.

Keeping the results in a simple table — baseline, post-upgrade state, adjustments, illuminance — also creates a reference for future projects. If the technical data of Zəfər İşığı lamps made at Zaferoğlu Elektrik's Ganja plant are used in the calculation, they should be taken from the product information.