How to do a lux survey of a workplace

Before you start: what are you measuring against?

Decide this first, because it sets the plane, the grid and the figures the survey is compared with. There are three usual answers.

• HSE's figures in HSG38. HSG38, Lighting at work, is HSE guidance (second edition, 1997). For each kind of work it gives two figures: an average for the area as a whole, and a minimum at any single point. You choose the row by the work done in the space, not by what the room is called. HSG38 deals with lighting for safety. For comfort and performance it refers you to industry lighting guidance.
• A figure your client names. The specification or brief may state its own levels. Write down which document the figure came from and its revision, so a reader can find it.
• No figure at all. Sometimes the job is to record the levels as they are, with no judgement. That is a legitimate survey, and the record should say so.

The legal duty is in regulation 8 of the Workplace (Health, Safety and Welfare) Regulations 1992. It requires suitable and sufficient lighting and gives no figures. HSG38 is HSE's guidance on that duty. A reading can be at or above HSG38's figure, or below it. Neither of those makes a workplace lawful or unlawful.

HSG38's table, and what the two columns mean

The instrument

Use a calibrated lux meter. A phone's light sensor is not a calibrated instrument, and its readings are not survey readings.

• Note the meter's make, model and serial number.
• Note its calibration certificate: the reference, the date, and when the next calibration is due. If it has lapsed, say so on the record rather than leave it out.
• Some meters have a setting for the type of light source, such as LED or fluorescent. Choose the one that matches the lamps on site, and record which you used.
• Follow the maker's instructions for switching on, removing the sensor cap, and any zero adjustment.

Which meters connect to Luxwork over Bluetooth, and how any other calibrated meter is used by typing its readings in, is on the meters page.

Conditions before the first reading

A reading only means something if the lighting was in a known state when it was taken. Check these before you start, and record each one as something you confirmed.

1. Lamps run up. Some lamps take a while to reach full output after switching on. Let them get there first.
2. Daylight and stray light kept out. If you are surveying the electric lighting, daylight will add to every reading. Survey after dark, mask the windows, or record that daylight was present.
3. Every fitting working. Walk the area and look. If some fittings are out, you can still survey. The record should say how many were out, because the readings describe the lighting as you found it.
4. New or existing installation. Design figures are usually maintained values: the level the lighting should still give at the end of its maintenance cycle, after lamps age and fittings get dirty. A new installation would normally read above that figure on the day. Record which kind you surveyed.

The measurement plane

The plane is the surface the readings describe: usually the floor, or a working plane at the height where the work is done, such as a desk top. Hold the sensor flat, facing up, at that height for every reading.

The requirement you are measuring against should say which plane it means. If it doesn't, choose one and write it down. A reading at desk height and a reading at the floor are different measurements, and a reader can't compare them if the plane isn't stated.

The grid

A survey uses a grid because readings at evenly spaced points describe the whole area. A few readings taken where it looks dark, or where it looks fine, describe only those spots.

• How many points. Best practice sets the number of points from the size and shape of the area. Bigger or more awkward areas need more. The height of the fittings above the plane matters too, which is why two rooms with the same floor area can need different numbers of readings.
• The edges. Best practice usually keeps the grid a short distance in from the walls. Readings right against a wall behave differently from those in the open floor.
• Walk order. Walk the grid in a fixed order, for example row by row, turning at each end. Then every reading has a known position and none gets missed.
• Points you can't reach. A rack or a machine may sit on a grid point. Record the point and the reason rather than leaving a silent gap. A reader needs to know the result rests on fewer points than planned.
• Long, thin spaces. Corridors and platforms can end up with very few points across their width. Look at the plan before you start and check it covers the space you actually need to know about.

Luxwork works out the grid from the method you choose. It records how the grid was arrived at, including anything it had to assume, such as a mounting height that was not measured on site.

Keeping out of your own light

Your body blocks light and reflects it. If you stand between the sensor and a fitting, the reading drops. If you lean over the sensor in a light-coloured jacket, it can rise. Either way the number describes you as well as the lighting.

• Stand to the side of the sensor, away from the main fittings, and keep the same stance at every point.
• A sensor on a lead lets you hold it on the plane while you stand back.
• With a meter that sends its readings over Bluetooth, you can set the meter down on the point and step away. The reading arrives on the phone.
• Give the reading a moment to settle before you record it.

Average, minimum and uniformity

Three figures come out of the readings.

For example, on the specimen report used across this site, the lowest of 36 readings is 64.9 lux and the average is 222.6 lux. Uniformity is 64.9 ÷ 222.6, which is 0.29.

The average alone can hide a dark patch. HSE makes this point in HSG38: an area can have a good average and still have spots that are too dark to be safe. That is why HSG38 gives a minimum as well as an average, and why a survey should report both.

What to record

Someone who wasn't there should be able to check the survey from the record alone. As a minimum, write down:

• The site, level and area, and the drawing you worked from, if there was one.
• What the survey was measured against, and where that figure came from.
• The instrument, its serial number, its calibration certificate and due date, and the light-source setting.
• The measurement plane.
• The grid, how it was set, and anything assumed.
• The conditions you confirmed before the first reading.
• Every reading, with its position and time, and whether it was typed in or came from the meter.
• Any points not measured, and why.
• For each figure: whether the survey came in at or above it, or below it.
• Who did the survey, and when.

What a lighting survey report should contain goes through the same list on a specimen report.