Questioning the quality and accuracy of noise assessments of battery energy storage systems

As those consultants working in the renewables industry will know there has been a rapid rise in applications for battery energy storage systems (BESS) over the past few years, either as stand-alone or attached to solar or turbine developments. Noise consultant, Dick Bowdler, has reviewed about a dozen noise assessments for these systems and, in this letter, voices his concern over their quality and accuracy.

There are three problems: the poor quality of manufacturers’ information, the technical accuracies of the calculations and the variety of assessment methods. I will outline the issues in a bit more detail in the hope that it will lead to improved quality and accuracy of assessments.

Taking manufacturers’ information first, it is perhaps no surprise that with new plant being developed at a fast rate, manufacturers do not produce good quality data until a range of models has been established. However, it is as much in the developer’s interest as anyone else to get good data because it reduces the commercial risk and it is also very much in the consultant’s interest to get it as accurate as possible for obvious reasons. Noise data are almost all of low quality (and hence, high uncertainty) or simply absent. Most of the data that are available are presented as 1m (or some other distance) from the equipment. For example, the sound pressure level of a transformer 9m by 6m by 6m is quoted as 68 dB at 2m. This really tells us very little. Is that the level at the one noisy part of the transformer, or the average level at 2m all over the transformer, or is it the average of a few spot measurements? Some manufacturers provide no noise data at all and hardly any provide frequency spectrum data. Others provide data under an NDA so it becomes more difficult to verify that it is valid.

On my second point on the accuracy of the calculations, in the transformer example above, the consultant simply converted the 68 dB at 2m to 82 dB SWL. That treats it as a point source with hemispherical propagation. Of course it is extremely unlikely that it is a point source as it would only be so if all the sound from the transformer came from a single point 2m from the microphone and none from anywhere else on the transformer. If the sound was equally emitted from all over the transformer at 68 dB at 2m, the SWL would be about 94 dB. Unfortunately, using point source calculations where it is clearly not a point source is still common and will most likely under-estimate the noise at sensitive receptors very significantly.

On my final point of the assessment method, nearly all of the assessments say how wonderful SoundPLAN or CADNA is, as if their use means the result must be correct. Some assessments discuss uncertainty in the calculations of ISO 9613 but not one of the assessments that I have seen discusses the uncertainty of the input data. Hardly any assessment put a comprehensive list of the inputs.

The guidelines for assessment are certainly inadequate and this is apparent from the methods put forward. In one case, a specific noise of 44 dB with background level of 26 dB in the evening is low impact and minor adverse. In another a specific noise of 31 dB compared with 25 dB background is rated as high impact. Almost everyone focuses on BS 4142 but most also mention or use one or more of BS 8233, WHO Enger, WHO night guidelines, NR curves. In fact, in some instances, the reader might be forgiven for thinking the assessor had gone through all the guidelines in turn to find one that worked. Some sort of coordinated agreement is needed to create an element of standardisation. Maybe BS 4142 is the correct approach but there is certainly widespread agreement that, at low noise levels, an absolute level is more realistic. How this latter point is dealt with is the crucial matter.

I’ve avoided introducing my personal view so far, but I would like to make one comment, that any use of BS 8233 is not appropriate. Since its origin as Chapter III of CP3 in 1972 it has always been about the design of sound insulation of buildings. It is nothing to do with industrial developments affecting existing houses. The Scope of BS 8233:2014 says it ‘… does not provide guidance on assessing the effects of changes in the external noise levels to occupants of an existing building.’

To summarise:
Assessments should be transparent. Where data is inadequate consultants should attempt, or ask their clients to attempt, to get better data. Failing that the inadequacy of the data should be clearly described. Make sure your calculations are right. Failing any agreed assessment method, each consultant needs to consider what their own view is and, of course, use the same method on all similar projects. The assessment method should be clearly defined at the start of the assessment.

Dick Bowdler

Response to: Questioning the quality and accuracy of noise assessments of battery energy storage systems

I read with interest the letter from Mr Bowdler regarding the prevalence of battery energy storage system (BESS) planning applications in recent times (page 59 of this issue) and respond as a member of the IOA’s Renewable Energy Advisory Group (REAG). Please note this is my personal response and does not represent the REAG as a whole.

Within the REAG we have a sub-group, which has been formed to focus on producing a good practice guide (the Guide) for the assessment of noise from BESS and solar developments.

The formation of the sub-group was presented in a wider paper regarding the REAG at the Acoustics 2024 conference, which was subsequently summarised in Acoustics Bulletin Nov/Dec 2024. The Guide is still in its preliminary stages but the group has already expressed similar concerns to Mr Bowdler and we hope that when it is published it will help to normalise the approach to noise assessments for such developments. In particular, we share Mr Bowdler’s concerns over the use of manufacturer supplied noise level data, both in the quality of the data that is published and in the various approaches that consultants attempt to use it.

One of the most significant problems we perceive, is that of a lack of knowledge about how a BESS site (and similarly related solar plant) actually operates, both in terms of layout design, and in respect of operating parameters and related noise output. In terms of layout, this includes: what plant is required, how many batteries and inverters do you need for a particular capacity and how does that vary by manufacturer, what is the difference between a two-hour storage facility and a four-hour storage facility? Maybe most critically, how can we be sure that appropriate candidate plant is being specified and how does that affect the required layout?

In terms of operation and noise, this is particularly complex; when do cooling fans start to operate, is cooling based on load or ambient temperature (or both), is more noise generated during charge or discharge cycles, and often the most difficult question, what do I do with this poorly presented source noise level data? This lack of knowledge is not just present at the acoustics consultant level, but also within councils (many of whom treat BESS the same as substations), and even for the developers themselves. Accordingly, the Guide will seek to provide guidance, not just through purely acoustics assessment terms, but also by introducing the reader to what makes up a typical BESS (and solar) layout, how the noise output might vary depending on plant specifications, layout and operation, and what mitigation options are available (and realistic).

I hope that the Guide will also encourage manufacturers to ‘up their game’ when it comes to providing source noise level data and encouragingly, Over the past 12 months or so, I have seen noise being taken more seriously by some manufacturers. At TNEI, we have undertaken ISO 3744 measurements for several plant suppliers, who now have 1/3 octave sound power level data for each façade and top of their battery and/or inverter units, operating at various different fan capacities and similar, and that is the type of data that the Guide will seek to encourage plant manufacturers to provide to consultants and developers. The fact is, however, that currently many manufacturers do provide sub-standard noise level data, so the Guide will need to demonstrate how to manage this data, how to consider uncertainty (both for the noise data and the operational status of the plant), and how best to use this within a noise model.

The Guide should also encourage manufacturers to provide more detailed noise and operational data, penalising, for example, the use of noise data without accompanying operational data, by assuming that all plant is always operating at 100% capacity (and maximum noise level output). I am sure Mr Bowdler will remember a time when wind turbine noise level data was provided in a multitude of formats and of varying degrees of quality, though nowadays, thanks to standardised measurement procedures and in no small part to the publication of the IOA GPG 1, the way that wind turbine noise level data is considered within assessments in the UK is generally agreed upon, and a common approach to modelling has been adopted. I do hope that we will get some way there soon for BESS plant and similar.

With regards to the use of BS 8233 and other standards and guidelines that sometimes seem to be adopted inappropriately, I am not sure that is isolated to just BESS noise assessments and I personally share similar concerns to how and where these ‘guideline levels’ are used more generally for industrial and commercial noise assessments. I expect the proposed changes to BS 8233, if adopted, will help to alleviate this issue somewhat. We now have enough people within the sub-group to produce the Guide, however, I have invited Mr Bowdler to join the wider peer review team and if other IOA members are also interested in being part of the peer review, then please feel free to get in touch.

Jim Singleton
jim.singleton@tneigroup.com

1. A good practice guide to the application of ETSU-R-97 for the assessment and rating of wind turbine noise