How long should a noise assessment take?

This a common and understandable question that people looking for a noise assessment often ask - how long will it take.

Eight hour daily averages in noise assessment and measurement time

The Noise Regs talk about an employer having a ‘reasonable estimate’ of the daily noise exposure levels for employees and the standard way to calculate that is over an eight hour working day. But, that doesn’t mean a noise assessment has to last for eight hours.

To take two scenarios:

• Bob works on a press, doing the same thing roughly every 30 seconds. If we measure for a few minutes the noise reading will settle at say 91 dB(A). From that point on it doesn’t matter if you measure for five minutes or for his entire shift, the result will be exactly the same.

• Steve works on the press for three hours and then on a robotic welder for five hours. If you measure the press and the welder separately you will get a noise reading for each. Then some maths can add the two exposures together, including their respective exposure times, and give you an eight hour daily average figure.

In both, you don’t need to measure for eight hours to get a daily noise exposure figure. Once a reading has settled for a job, measuring for longer doesn’t change it at all. Measuring like this is not a short-cut, it is the normal way of assessing noise exposures and the HSE include a process for doing it in L108.

This is also important when looking at things like overtime or longer shifts where the exposure time is different to the standard eight hours. If people occasionally do a couple hours of overtime you don’t then need to do a ten hour measurement, you can work out a ten hour exposure from the noise level measured for the job itself.

Number of noise meters makes a big difference.

Back when I started doing noise assessments, most noise assessments were done with one noise meter whereas now I am using up to fifteen noise meters at once. That means I can gather the same data now in an hour which would have taken two days to do in the past.

For smaller companies I can often measure the entire site at once rather than bits at a time.

Machine running times influence the noise assessment

One of the biggest influences on a noise assessment duration is whether the machines are actually running as a lot of sites have machines which only run at certain times of day. That is no problem for a noise survey and I would just make sure I am there for it as needed, if that means waiting around so be it.

Size of site or numbers of people are less important than you may think

What is crucial for a noise assessment is not really the number of people or square footage of a site but the variation in the work.

If you have a factory with 250 employees then often most do the same jobs on the same machines which have a regular cycle, so from a noise assessment perspective it is quite simple. Conversely, if you have a joinery workshop with 20 employees then that can be complex as the variation in their work from day to day can be considerable.

Sample measuring in noise assessments

Noise assessments use sample measuring techniques, which means if 15 people all do the same job on the same machines in the same way, you don’t measure every one of them. Practically what that means is for those 15 you maybe do two longer-term wearable measurements, and while they are running measure all 15 individually using the hand-held meter. Those shorter duration noise measurements confirm that all machines are indeed giving the same noise exposure and that the two wearables are genuinely representative for everyone. Conversely, if the hand-held is showing some differences you up the number of wearable noise measurements to make sure any variations are caught.

As an example of what I mean - a noise assessment I did in Barnsley in 2024. Two types of machines present and about 30 of each. For one type of machine the hand-held showed they were all indeed the same noise-wise so the three wearables I used on them were good to go for everyone. For the other type of machine the hand-held showed four of the seemingly similar-looking machines were actually a little different (they had slightly different placings of a motor compared to the others) so I made sure wearable results were done specifically for those machines as well as covering the rest of them.

I am not the type to pad noise assessment invoices

In May 2025 I did a noise assessment for a client who had used one of the big boys for noise assessments before. The other people charged something in the region of £2k and insisted it was a two and half to three day job. Cobblers. It turned out they were using quite a small number of noise meters so must have spent 90% of their time waiting around before swapping them to a new location.

I charged a third of that and as I have a lot more noise meters and the client’s machines ran consistently all day every day, I did the measurements inside a day and still obtained more data than the previous lot did and provided a better client report.

When he got my noise assessment report the HSE manager there replied “This noise report is fantastic, we will certainly be using you again in the future [rather than the previous supplier], very knowledgable as well”.

The best way to explain it is that I am being paid to provide an accurate noise assessment, not for a fixed time on a site. I will be there as long as is needed to get an accurate noise assessment but will not be kicking my heels for hours just to make it look longer and get the invoice bigger.

I would rather do two sites in two days than dawdle around one not doing a lot except looking busy and getting the invoice higher. I could have invested in three or four noise meters and then drag jobs out to cover multiple days rather than spending over £20k of my own money on a lot of noise meters, but that would make my working life exceedingly tedious and rather pointless and I’m too old and grumpy for that.

Its a win-win, the client gets a bob-on accurate noise assessment but for a cracking price as it hasn’t been dragged out needlessly, and I still charge the same over two days, just that it is one day on one site and another day elsewhere, which keeps work interesting.