Sound Judgement: Assembling an Acoustic Case File for UK Hydrogen Refuelling Depots
Photo: Arthur C Harris , CC BY-SA 2.0, via Wikimedia Commons
Hydrogen refuelling sites are usually discussed in the language of emissions, safety and grid capacity. Noise is lower on the agenda until a site moves close to homes, a retail park or a school. At that point, planning committees and environmental health officers tend to ask a more practical question: can the operator prove what the site will sound like, and what it actually sounds like once it is running?
In 2026, a manufacturer datasheet is not enough. A credible acoustic case file should connect design assumptions to commissioning measurements and then to an operational routine. The aim is not to produce a glossy document for the planning file. It is to create an auditable record that a site manager can use when a complaint arrives, when plant is replaced, or when a local authority asks for evidence that planning conditions are being met.
Why the boundary matters more than the brochure
Hydrogen refuelling depots generate noise from several sources: compressors, chillers, pumps, inverters, transformers, dispensers, purge vents and delivery tankers. Battery-electric depots bring their own profile, particularly inverter and charger cooling fans. A sound power level quoted for a single enclosure says little about the combined effect at the nearest sensitive receptor, especially when several units cycle on and off.
Planning conditions often set limits at agreed receptor positions, restrict operating hours, and require a complaints procedure. Some authorities will ask for an assessment against BS 4142, while others will apply local planning guidance. The case file should follow the method the authority has specified and explain any deviation. It should also be written for two audiences: the planner assessing the application and the operator who will inherit the condition.
The baseline survey: fixing the starting point
A baseline survey should establish the existing acoustic environment before the development changes it, or before a new phase of operation begins. That means measurements at agreed receptor positions, covering day, evening and night where relevant. The record should include instrument calibration certificates, microphone heights, distances, ground conditions, weather and the operating states of nearby plant.
Background noise is not a single number. It varies with traffic, weather and time of day. A useful baseline distinguishes residual noise from specific sources and records uncertainty rather than hiding it. If the site is already operating, the baseline should be clearly labelled as pre-change or as an existing-environment survey. Ambiguity here is one of the quickest ways to undermine an otherwise sound pack.
Tonality: the component that attracts complaints
Compressors and inverters are often tonal. A simple A-weighted level can understate the disturbance caused by a discrete hum, whine or harmonic. The case file should include narrowband or third-octave data where tonality is plausible, and should record whether a tonal adjustment is warranted under the chosen assessment method.
Tonality is not a fixed property. It can change with load, enclosure condition, maintenance and the way equipment is mounted. Commissioning measurements should therefore be taken across representative operating modes, not only at full load. Anti-vibration mounts, acoustic enclosures and attenuators should be documented as installed, with photographs and commissioning records. If a compressor is later replaced or re-rated, the acoustic evidence should be revisited.
Night-time purge cycles and the 03:00 question
Purge and vent cycles deserve particular attention. They can be short, intermittent and relatively loud, which makes them poorly represented by an average level. Night-time limits are usually tighter, and a single event at an unsociable hour can generate a complaint even when the overall Leq looks acceptable.
The pack should log the expected frequency, duration and level of purge cycles, and state whether they are automatic or manual. Where possible, control-system timestamps should be correlated with acoustic measurements so that events can be identified rather than inferred. If night purging is unavoidable, the evidence should show what mitigation is in place and how the cycle has been scheduled. Averages without event counts are weak evidence for this kind of source.
Deliveries, dispatch and the timetable as a control
Tanker deliveries, yard movements, reversing alarms, coupling operations and pressure releases all form part of the site's acoustic signature. Delivery windows are often written into planning conditions, so the case file should map the declared schedule to the noise assessment. If out-of-hours deliveries are expected, they should be assessed explicitly rather than left to operational discretion.
Operational controls matter as much as plant specification. Routeing, idling restrictions, barrier placement and staff briefing can change the effect at the boundary. For battery-electric depots, charging schedules and cooling fan activity should be treated in the same way. Monitoring should compare actual activity with the declared schedule, because a planning condition that is not observable in practice is difficult to defend.
From condition to routine: the monitoring plan
A post-commissioning survey is the bridge between prediction and operation. It should be carried out once plant has settled into normal service, using the same receptor positions as the baseline where possible. After that, the pack should set out a routine: periodic checks, event-triggered measurements after complaints or major maintenance, and an annual review of the evidence.
The routine should name a responsible person, specify calibration and data-retention arrangements, and include a complaint log with response times. Raw data, weather notes and maintenance records should be kept together so that a later reviewer can reconstruct what happened. The point is not to create an unmanageable paperwork burden. It is to make sure that the site can answer a question with evidence rather than anecdote.
What a defensible pack contains
A practical case file will usually include a site plan and receptor map; an equipment schedule with sound power levels; the baseline survey; prediction assumptions; commissioning measurements; a tonality assessment; purge-cycle and delivery logs; the monitoring plan; the complaints procedure; calibration certificates; and a version-control record. Each item should be dated and tied to a named site or phase.
Common weaknesses
Recurring weaknesses include relying on manufacturer data as if it were site evidence, taking baseline measurements in unrepresentative weather, ignoring tonality, averaging intermittent purge events, and failing to connect planning conditions to day-to-day operations. A pack that is never updated after commissioning is also vulnerable, particularly where equipment is replaced or operating hours change.
Noise evidence is not merely a planning hurdle. For hydrogen and battery-electric depots entering mixed-use areas, it is part of operational assurance. The strongest case file is the one that lets a site manager demonstrate, with dated measurements and clear records, that the site is behaving as promised.