In August 2026 the ABC reported on community concern about a proposed 520 megawatt battery energy storage facility at Uriarra Station in the Australian Capital Territory, around two kilometres from Uriarra Village and roughly 30 kilometres west of Canberra. Residents raised visual impact, lighting, bushfire and noise. One resident, who works from home, described his concern as hearing a constant background hum.

The proponent’s response was that design and layouts were still being finalised and that the project would comply with relevant government regulations for day and night-time noise. At that stage of a project, that is a reasonable and accurate thing to say.

It is also worth noticing what happened in that exchange, because it happens on a great many projects. A resident described a specific experience, continuous, at home, all day and all night. The response described a regulatory outcome. Those are different questions, and the gap between them is where a lot of avoidable conflict on renewable energy infrastructure begins.

Noise is the concern that does not fade

Battery projects attract several categories of community concern, and they behave differently over time. Visual impact is contested at the outset and then largely habituated. Fire risk is probabilistic, technically complex, and argued between experts. Traffic is concentrated in construction and then ends. Land use change is decided at approval and not revisited.

Noise is different in three ways that matter. It is continuous, so it is present every night for the life of the asset rather than during a phase. It is attributable, because a resident can point at the source and it was not there before. And it is verifiable, because anyone can measure it, and the measurement can be compared against a number in a condition of consent.

That combination makes noise the most durable grievance available to a community that is unhappy with a project, and the most falsifiable commitment available to a proponent that wants to be believed. Both of those points argue for taking it seriously early.

Low background levels are the whole problem

Battery projects tend to be built where land is available and a grid connection exists, which frequently means rural and peri urban locations. Those are the quietest places in Australia at night.

The NSW Noise Policy for Industry derives the intrusiveness noise level as the rating background level plus 5dB. It sets minimum assumed rating background levels of 35dB(A) during the day and 30dB(A) in the evening and at night, which produce minimum project intrusiveness noise levels of 40dB(A) during the day and 35dB(A) in the evening and at night. Those minimums exist because measured rural background levels can fall below them, and the policy does not require a project to chase a background that low.

The practical consequence is that a rural battery site is often working against a night-time trigger of 35dB(A) while an equivalent facility in an established industrial area might have ten decibels more to work with. The equipment is identical. The margin is not.

This is also why background monitoring is worth doing early and doing properly. A short survey in the wrong season, at the wrong location, or during an atypical period produces a background level that the project will then be held to for decades. Monitoring long enough to characterise seasonal and weekly variation, at locations that represent the receivers who will actually complain, is one of the cheapest risk reduction measures available on a battery project.

Compliance and acceptability are not the same question

A project can satisfy every applicable criterion and still be experienced as intrusive.

This is not a failure of the frameworks. Environmental noise criteria are designed to limit the incidence of impact across a population, not to guarantee that no individual notices a change. A level based criterion is also a poor description of what a resident actually reports, which is usually about character and persistence rather than magnitude. A constant hum at 33dB(A) is a different lived experience from intermittent traffic at 45dB(A), and the first can generate complaints while the second does not.

Australian frameworks recognise part of this through modifying factor corrections for tonality, low frequency content and intermittency, and battery sites are strong candidates for a tonality correction because transformers and power conversion equipment produce tonal content by design. But the corrections are a blunt acknowledgement of a subtle effect, and they differ substantially between states. Research presented to the Australian Acoustical Society found more than 12dB of variation at low frequency between the level at which a tone attracts a penalty in Western Australia and the level at which it does in New South Wales.

The useful position for a proponent is therefore not that the project complies. It is that the project has understood what it will sound like, has said so specifically, and has committed to something that can be checked. AcousTech has argued elsewhere that acoustic design and acoustic compliance are not the same thing, and the distinction is nowhere sharper than in a community setting.

Six things credible acoustic planning can contribute

Acoustics does not create social licence. It can supply evidence that a project has measured what it is asking a community to accept, which is one component of being believed.

Background monitoring done early and honestly. Enough duration, representative locations, disclosed methodology, and a stated view on how variable the background is. A project that publishes its background data before it publishes its predictions is in a stronger position than one that does the reverse.

Site selection and setback. This decision moves the acoustic outcome more than any other on a battery project, and it is made before an acoustic consultant is usually appointed. Distance, intervening terrain and the orientation of the site relative to dwellings do more than any treatment applied later.

Source selection and layout. Which equipment carries tonal content, and where it sits relative to the boundary, is a design decision rather than a mitigation one. Placing transformers and power conversion equipment where site geometry provides screening, and keeping identical repeated units away from the nearest receiver, costs nothing at layout stage.

Specific night-time commitments. A commitment that can be tested is worth more than an assurance that cannot. A stated predicted level at named receivers, under a defined operating scenario, at a defined time of night, is a specific commitment. Compliance with applicable regulations is not.

Transparency about uncertainty. Every prediction carries uncertainty, and pretending otherwise is corrosive when the first measured result differs from the model. Stating the assumptions, including where supplier data was substituted from a similar unit, is more defensible than a single confident number that later needs explaining.

Verification and a response pathway. Post commissioning measurement at the same locations under comparable conditions, published, with a stated process for what happens if a complaint is made. A project that has already said what it will do about a problem is in a very different position from one improvising after the first phone call.

What acoustic planning cannot do

It cannot substitute for the other concerns. A community worried about bushfire, water, land use or visual change will not be reassured by a noise report, and offering one in response reads as evasion.

It cannot make a poorly sited project acceptable. Where a facility is close to dwellings with a very low background, treatment can bring a design within criteria but is unlikely to make it inaudible, and it is more honest to say so.

And it cannot be delivered late. Acoustic commitments made after layouts, equipment selections and grid connection arrangements are fixed are constrained to whatever those decisions left available.

Requirements vary, and conditions of consent vary more

Australia does not have a single environmental noise framework. New South Wales applies the Noise Policy for Industry. Victoria assesses commercial, industrial and trade premises under EPA Publication 1826.4. Queensland works from the Environmental Protection (Noise) Policy 2019. Western Australia applies the Environmental Protection (Noise) Regulations 1997. The Australian Capital Territory, South Australia, Tasmania and the Northern Territory each have their own arrangements.

Beyond the general framework, individual approvals frequently impose conditions that are more specific or more stringent than the underlying policy, including limits at named receivers, restrictions on particular operating modes, monitoring obligations and complaint response requirements. A developer working across several states should assume that the acoustic commitments made on one project do not transfer to the next.

Where mitigation is required, the treatment has to work with the equipment rather than against it. Battery cooling plant needs airflow, so screening and enclosure have to preserve it, which is the domain of products such as Sonic Series acoustic louvres and barrier walls built from Sonic System acoustic modular panels. That is a design question, not a rescue operation, and it belongs in the layout rather than in a condition of consent.

The argument worth making

Social licence is not won by an acoustic assessment. It is won or lost across the whole way a project deals with the people who will live next to it, over years, and noise is one strand of that.

What acoustic planning contributes is a specific and unusually testable form of credibility. A project that measured the background properly, said what it would sound like at named houses on a still night, explained where its data was uncertain, then measured afterwards and published the result, has done something that most infrastructure proposals do not do. It has made a claim that can be checked, and then invited the check.

For a proponent, that translates into four things worth committing to before the layout is fixed: a background survey long enough to characterise seasonal variation, predicted levels stated at named receivers under a defined night-time scenario, disclosed assumptions where supplier data was substituted, and post commissioning measurement at the same locations with the results published.