- Author
- Byron Raal, CAS Founder-Editor About the author
- Checked against
- ISO 8573-1AS 4343
- Date last checked
- 21 August 2026
Looking for a compressed air provider? Tell us what you need. There is no cost to enquire.
Any proposed introduction is discussed with you first. We do not sell equipment.
Compressed air runs the parts of your brewery that move: pneumatic valves and actuators, the filler, seamer and labeller, keg washers, air knives drying cans, and sterile filtered air injected into cooled wort so the yeast gets its oxygen. It’s not the gas for jobs where oxygen pickup would wreck the beer. That’s CO2 and nitrogen. It’s also the least visible cost in the building: there’s no compressed air line on your power bill. In the reference breakdown of brewery energy, compressed air is around 10 per cent of a brewery’s electricity use, and the US Department of Energy says leaks can sometimes waste 20 to 30 per cent of a compressor’s output. On the small single-shift brewery worked through below, that is roughly $2,300 a year of electricity turned into hissing. For energy, work in this order: leak survey, pressure set-point, then storage and control. In parallel, identify every point where compressed air contacts product or the inside of a package and address any food-safety risk immediately. Do not defer product-contact air quality behind energy work.
Where breweries actually use compressed air
Air does the mechanical work. Gas does the chemistry. Get that boundary right and the rest of this page follows.
Compressed air, typically: valve actuation across the brewhouse, cellar and CIP skid, since actuated butterfly valves are usually air driven; packaging line motion, meaning fillers, seamers, cappers, labellers and reject arms; keg washers; air knives drying cans before labelling; pneumatic conveying of spent grain; and wort aeration. That last one is genuine product contact: cooled wort needs dissolved oxygen before pitching, and plenty of breweries deliver it with sterile filtered air rather than bottled oxygen.
CO2 or nitrogen, typically: purging tanks, kegs, cans and bottles before fill; counter pressure filling and carbonation; tank blanketing on bright beer; and pushing finished beer where you cannot afford oxygen pickup.
Anyone telling you air can do the CO2 jobs has never tasted oxidised beer. Post fermentation, oxygen is the enemy. Pre fermentation, it is a nutrient. Same element, opposite job, specified separately. Plenty of small breweries never treat air as a utility at all, because CO2 arrives on a pallet with an invoice and air arrives out of a shed with none. One of those two turns up in your accounts every month. The other one hides inside the power bill, which is exactly why it goes unmanaged.
The cost problem, and why nobody sees it
Two numbers set the scale, and both need honest framing.
First, the share. The ENERGY STAR guide for breweries, published by Lawrence Berkeley National Laboratory, puts compressed air at around 10 per cent of brewery electrical energy, alongside refrigeration at 30 to 40 per cent and packaging at 15 to 35 per cent. That is electricity, not total site energy. Breweries burn a lot of gas boiling wort, so air is a smaller slice of the whole energy pie than of the power bill. Anyone quoting “10 per cent of a brewery’s energy” without that distinction has not read the figure they are repeating.
Second, the leaks. The US Department of Energy figure, the one the brewery literature leans on, is that leaks can sometimes waste 20 to 30 per cent of a compressor’s output. Note what that is a percentage of: the air the compressor makes, not the money you spend. You will also see 30 to 40 per cent quoted around the sector. We went looking for its primary source and could not find one, so we do not repeat it. We are not calling it wrong either, because we cannot show you that. What we can source is 20 to 30 per cent, and that is what every figure on this page is built on.
Two honesty notes. A leak rate is a flow percentage, not automatically a bill percentage: repairs only cut power if your controls respond by unloading, slowing or shutting down a compressor. And nominal motor kW is not what your meter sees, because package input includes cooling fan, drive and control losses.
Worked example: a 2,000 hL brewery, before and after a leak program
Assumptions. Output 2,000 hL a year, about 200,000 litres, one packaging line. Compressor 15 kW nominal motor rating, an illustrative proxy for electrical input. Hours 2,000 a year, single shift. Electricity $0.30 per kWh, our illustrative 2025-26 rate, not a national benchmark; yours might read $0.22 or $0.38, so substitute it. Specific power 6.5 kW per cubic metre per minute of free air. Grid factor 0.62 kg CO2-e per kWh.
Step 1. Baseline and capacity. 15 kW x 2,000 hours x $0.30 = $9,000 a year to run the compressor. Its output is 15 divided by 6.5 = 2.31 m3/min, which is 38.5 litres per second of free air.
Step 2. What the survey finds. Take an illustrative register: one bad coupling and a few pinholes. Say it comes back with one 3 mm equivalent leak and three 1 mm leaks at 7 bar gauge. On our leak model a 3 mm leak flows 7.31 L/s for 2.85 kW; a 1 mm leak flows 0.81 L/s for 0.32 kW.
Leak flow: 7.31 + (3 x 0.81) = 9.74 L/s. Leak power: 2.85 + (3 x 0.32) = 3.81 kW. Read that carefully: the 9.74 L/s leak load corresponds to about 3.81 kW at the assumed full-load specific power, and that is not automatically 3.81 kW of metered saving. What you actually avoid depends on how the compressor responds, whether that is load/unload, start/stop, modulation or a VSD. Share of output: 9.74 divided by 38.5 = 25 per cent, which sits inside the 20 to 30 per cent band DOE describes. That is the sanity check. It also tells you this is a leaky site rather than an average one, so read what follows as what a bad result looks like, not as a forecast for your brewery.
Step 3. What it costs, if the controls respond. For illustration only: if your compressor and its control system turn the avoided leak flow into a proportional cut in package input at 6.5 kW per cubic metre per minute, then 3.81 kW x 2,000 hours = 7,620 kWh, at $0.30 that is $2,286, call it $2,300 a year, and at 0.62 kg per kWh about 4.7 tonnes of CO2-e with no beer attached to it. On a real load/unload, modulating or VSD system, work the saving out from logged loaded and unloaded power or the actual part-load curve instead.
Step 4. What repair recovers. Do not model a leak program down to zero. Leaks regenerate and some sit behind machines you cannot shut down. A commonly used target is holding leaks under 10 per cent of air produced. Under the same illustrative proportional-power assumption, a residual leak flow at 10 per cent of 38.5 L/s is 3.85 L/s, corresponding to 1.50 kW, 3,000 kWh and $900 a year. Against that illustrative starting case, $2,286 less $900 = about $1,400 a year recovered, and roughly 2.9 tonnes of CO2-e. For a real site, replace these figures with the saving calculated from your compressor’s measured or control-specific power response.
Step 5. Payback, as a range and not a promise. Our own published figures put a professional ultrasonic survey anywhere from $1,500 to $8,000 depending on site size and scope, and a single-compressor brewery sits at the bottom of that. Add fittings and call the program $1,800 to $3,500. Under the illustrative proportional-power assumption above, $1,800 to $3,500 against about $1,400 a year gives roughly 16 to 30 months. Actual payback must use the measured or control-specific kW reduction, not this illustration. Not the “pays back in a quarter” line vendors use, because you only run 2,000 hours. If your annual leak-on hours double and the same control-response assumption still holds, the illustrative saving doubles to about $2,800 a year and simple payback roughly halves.
The nearly free lever: every bar of unnecessary pressure costs roughly 6 to 7 per cent of compressor energy and makes every leak leak faster. If your brewery is one of the many running 7.5 bar because one machine allegedly needs it, dropping to 6.5 bar where nothing genuinely needs the higher figure is worth $540 to $630 a year against that $9,000 baseline, and it costs you a commissioning visit. Run your own numbers in the leak cost calculator.
Air quality where air touches product
This is where pages written by equipment sellers go wrong.
ISO 8573-1 does not tell you what class to run. Your process risk does. The standard is the language you use to specify and audit purity across three contaminants, particles, water and oil, each with its own numbered class. It is not a regulation and it mandates no class for food or beverage contact. In Australia the legal baseline comes from the FSANZ Food Standards Code as applied and enforced in your own state or territory. A HACCP or other food-safety plan may add site-specific controls, and a jurisdiction, a customer or a certification scheme may require one, but it is not a universal national requirement for every brewery. ISO 8573-1 gives you the language for specifying and verifying particles, water and oil at a nominated point. It is not by itself proof that product-contact air satisfies all Food Standards Code or microbiological controls; those need process-specific food-safety evidence and, where relevant, microbiological testing or sterile filtration.
Specify per point of use, not per plant. Air actuating a valve outside a tank has a different risk profile from air bubbled through cooled wort. Blanket “food grade air everywhere” specs are how small breweries buy desiccant drying they do not need while leaving the points that genuinely touch product on unfiltered air.
Know what it does not cover. ISO 8573-1 classifies particles, water and oil. Microbiological contamination sits outside the class system entirely, so a system can be fully compliant on paper and still carry a microbial risk. Point of use sterile filtration is a separate decision from your ISO class. And ask where the commissioning sample gets taken: at the outlet, or at the compressor discharge? A supplier who cannot answer that is quoting the machine, not your air.
One compliance note. An air receiver is pressure equipment, and your obligations depend on its hazard level under AS 4343 and on your jurisdiction. AS 4343 is prescribed through the schedules of the harmonised WHS Regulations, with registration triggers tied to hazard level, and Victoria runs its own OHS regime that must be checked on its own terms. Work it out for your vessel, in your state, not off a forum.
Batch demand, and why brewery compressors are usually the wrong size
Brewery demand is not flat. There is a brew day, a packaging day, a CIP cycle, and a lot of hours where the only load is a few actuators holding position. Then packaging starts and demand can triple for hours. Fixed speed compressors handle that badly: the machine cycles between loaded and unloaded, and an unloaded compressor still draws a meaningful fraction of full load power while delivering nothing. On a batch profile that part load penalty can be the biggest audit line after leaks, which is a reason to measure it rather than assume it.
Three moves, in order of cost. Storage before capital: more receiver volume, sized against the burst you actually have rather than the compressor nameplate, absorbs packaging peaks and stops short cycling. Size from the excess burst flow, the shortfall between demand and what the compressor delivers during the burst, because sizing on total peak buys steel you do not need. Sequencing and set points: with two machines, make them cooperate rather than fight. Plenty of breweries running two have never had the strategy reviewed, and telling a sequencing fault from a straight capacity shortfall takes load state and power logged on each machine, not one system total. A VSD, once you have the data: how much your demand actually varies is the major screening input, and a VSD earns its premium where the swing is real. What you actually save also depends on the full load and part load efficiency curves of the machines you are comparing, your pressure settings and control band, and how your existing compressors are sequenced. Any brochure quoting a fixed saving without seeing your logged data is selling, not engineering.
Grants and incentives
There is no single national compressed air grant in Australia. Support runs through separate federal and state schemes that open and close faster than most breweries plan capital. As at August 2026, NSW pays incentives through the Energy Savings Scheme, Victoria covers custom industrial projects through the Victorian Energy Upgrades project based pathway, and the ACCU Scheme names compressed air systems on its Industrial Equipment Upgrades method, paying in tradeable carbon credits rather than cash.
What matters more than the scheme list is the evidence, and evidence requirements are method-specific. A measured demand or energy profile, a leak register and a costed upgrade case are all useful on a compressed-air project, and an independent audit produces them, but a dollar figure against each individual leak is not a universal ESS, VEU or ACCU requirement. Scope the evidence to the applicable method before work starts. The ACCU Industrial Equipment Upgrades method requires the relevant recommendation to be less than two years old and to have existed before project registration, not merely before physical work starts. Check the current status on our grants map before you commit spend.
Pre-audit checklist for a brewery
Do these before anyone quotes you. They cost nothing and they change the conversation.
- Read the compressor nameplate: kW, model and year.
- Log actual running hours for a fortnight covering one brew day and one packaging run. Guessing is where every bad business case starts.
- Find your pressure set point, and the highest pressure any single item genuinely requires. If those two are far apart, you have found money.
- Do an off-production leak check with every legitimate air user isolated or accounted for. On a start/stop or load/unload machine, estimate the leak load from compressor cycling; for a pressure-decay test you need known system volume, pressure drop and elapsed time. A simple overnight fall in header pressure on its own is not proof of leakage.
- List every point where air can contact wort, beer or the inside of a package. That list drives your air quality spec, not a brochure.
- Find the marginal rate on your electricity bill, not the average, and ask whether the person quoting you sells the equipment they are recommending.
Take that to an independent assessor and you get a scoped audit, not a sales visit.
Frequently asked questions
How much of a brewery’s energy does compressed air use?
The ENERGY STAR brewery guide from Lawrence Berkeley National Laboratory puts compressed air at around 10 per cent of a brewery’s electrical energy, with refrigeration at 30 to 40 per cent and packaging at 15 to 35 per cent. That is a share of electricity, not of total site energy, and the underlying data is a UK kegging brewery survey published in 2000, so treat it as indicative rather than as current Australian data.
Can I use compressed air instead of CO2 in my brewery?
Not for the jobs CO2 does. CO2 and nitrogen handle anything where oxygen pickup damages beer: purging tanks, kegs and cans before fill, counter pressure filling, carbonation, blanketing and pushing finished beer. Compressed air runs the mechanical side, plus wort aeration before pitching, which is the clearest product contact use for air.
How much do compressed air leaks cost a small brewery?
On a 15 kW compressor running 2,000 hours a year at 30 cents per kWh, a 25 per cent leak load is about $2,300 a year of wasted electricity and roughly 4.7 tonnes of CO2-e. That conversion assumes your controls turn the avoided leak flow into a proportional cut in power. On that assumption, getting to a 10 per cent leak load recovers around $1,400 a year against a program of roughly $1,800 to $3,500, so payback runs about 16 to 30 months. On a real machine, work the saving out from logged package input and the control-specific relationship between power and delivered air. On load/unload or start/stop, use the loaded, unloaded and stopped periods; on modulating or VSD, use the measured part-load power response. That is a leaky site, not an average one, so use it as a worked method rather than a forecast.
What ISO 8573-1 class does brewery air need to be?
There is no single required class, and any supplier who quotes one without asking about your process is guessing. ISO 8573-1 classifies purity across particles, water and oil but mandates no class for food or beverage production. In Australia the legal baseline is the FSANZ Food Standards Code as applied and enforced in your state or territory. A HACCP or other food-safety plan may add site-specific controls and may be required by a jurisdiction, a customer or a certification scheme, but it is not a universal national requirement for every brewery. Microbiological contamination sits outside the ISO class system entirely.
Is a variable speed drive compressor worth it for a brewery?
Often, because brewery demand is lumpy, but not before you have the logged data. How much your demand actually varies is the major screening input: a VSD removes most of the energy a fixed speed machine wastes idling unloaded between packaging runs, and almost nothing on a site running steady near full load. What you actually save also depends on the efficiency curves of the machines you are comparing, your pressure settings and control band, and how your existing compressors are sequenced. Fix leaks, set point and storage first; they cost a fraction of a compressor.
Get an independent compressed-air assessment for your brewery
CAS does not sell, install or service compressors. Describe your brewery, operating pattern and the issue you need assessed. We can review your brief and discuss a suitable Australian provider if one is available. We ask for your written permission before an introduction and explain any referral payment arrangement. There is no cost to enquire.
Tell us your annual output in hL, compressor kW, operating hours and postcode. We can review your requirements and discuss a provider if a suitable one is available. We ask for written permission before an introduction.
Form not loading? Email us directly at byron@compressedairsolutions.com.au with your enquiry.