Carbon Accounting for Food & Beverage Manufacturing in Australia

Woolworths Group's Scope 3 emissions are 23 times larger than their Scope 1 and 2 combined — 96% of their total carbon footprint sits in the supply chain. And they've had their targets validated by SBTi: a 40% reduction in forest, land and agriculture emissions by FY2033. If you're a food or beverage manufacturer supplying Woolworths, Coles, or any major Australian retailer, those aren't their targets. They're yours.

Carbon accounting for food and beverage manufacturing in Australia is about to become a non-negotiable cost of doing business. Not because the government said so (though they did — ASRS Group 2 mandatory climate reporting kicked in from July 2026). Because your customers are going to stop buying from suppliers who can't produce emissions data. And in a sector where more than 70% of emissions sit in Scope 3 — agricultural inputs, packaging, transport — the complexity of actually measuring this stuff is genuinely brutal.

We've spent a lot of time talking to food manufacturers about their carbon accounting problems. The pattern is always the same: they know they need to do it, they've got a rough sense that their gas bills and cold storage are significant, but when they try to put real numbers against their emissions profile, they hit a wall. Refrigerant records are incomplete. Supplier data is nonexistent. And nobody on staff knows the difference between a location-based and market-based Scope 2 factor.

This is the guide we wish someone had written for them.

The Emissions Profile You're Actually Dealing With

Food and beverage manufacturing doesn't have a single dominant emission source the way mining has diesel or construction has embodied carbon. You've got emissions everywhere — gas boilers for process heat, electricity for refrigeration and cold storage, refrigerant leaks from ageing cooling systems, diesel for distribution fleets, and then the massive Scope 3 iceberg underneath it all.

Oxford Economics estimated that the Australian food and grocery sector generates roughly 50 million tonnes of CO2-e per year in Scope 3 emissions alone — from suppliers, transport, and packaging. That's a staggering number. And it's the emissions your customers will increasingly ask you to quantify.

But let's start with what you can actually control and measure directly.

Scope 1 in food and beverage manufacturing typically comes from three sources: natural gas combustion (boilers, ovens, dryers, pasteurisers), diesel for on-site forklifts and delivery vehicles, and fugitive refrigerant emissions from cold storage and process cooling. Under the NGA Factors 2025, natural gas has a Scope 1 emission factor of 51.53 kg CO2-e per gigajoule. A mid-size dairy processor running gas-fired pasteurisers and CIP (clean-in-place) hot water systems might consume 40,000 GJ of natural gas annually. That's roughly 2,061 tonnes of CO2-e — just from gas. Add a fleet of 20 refrigerated delivery trucks and your Scope 1 starts pushing toward NGER reporting territory fast.

Scope 2 is your purchased electricity. In food and beverage, electricity consumption is disproportionately driven by refrigeration. In meat processing, more than 65% of electricity demand goes to refrigeration — blast freezers operating at minus 40 degrees Celsius, cold storage rooms, chillers on processing lines. A large meatworks in Victoria consuming 15,000 MWh per year generates approximately 11,700 tonnes of CO2-e at the Victorian grid factor of 0.78 kg CO2-e per kWh. That same facility in Tasmania? About 3,000 tonnes. The state you operate in matters enormously, and we cover the full state-by-state NGA Factors breakdown here.

Scope 3 is the monster. We'll come back to it.

Refrigerants: The Scope 1 Source Nobody Tracks Properly

Here's something that consistently surprises food manufacturers we talk to. A single kilogram of R-404A refrigerant — the stuff that's been the workhorse of commercial cold storage for decades — has a global warming potential of 4,728. One kilogram leaked into the atmosphere is equivalent to nearly 4.73 tonnes of CO2-e.

Australia holds approximately 55,000 tonnes of refrigerants in operating systems nationally, equivalent to about 100 million tonnes of CO2 if fully released. In 2022, refrigerant emissions across the built environment contributed roughly 9 million tonnes of CO2-e — about 7% of total built environment emissions. And a lot of that sits in food manufacturing and retail cold chains.

The problem for carbon accounting isn't that refrigerants are complicated to calculate. The formula is simple: mass leaked (in tonnes) multiplied by GWP. The problem is that most food manufacturers don't know how much they're leaking.

Older industrial cold storage systems — the kind you find in meat processing plants, dairy facilities, and large bakeries — can lose 15-25% of their refrigerant charge annually. A system holding 500 kg of R-404A leaking at 20% loses 100 kg per year. That's 473 tonnes of CO2-e from a single system. Many food processing sites run multiple refrigeration circuits. The fugitive emissions can rival or exceed the gas boiler emissions that everybody focuses on.

Under NGER, you're required to report HFC emissions if the system holds 100 kg or more of refrigerant with a GWP above 1,000. But here's the catch — there's no mandatory leakage register. You estimate it from top-up records. If your maintenance contractor tops up the system and the invoice just says "R-404A recharge — $3,200," without specifying kilograms added, you've got a data gap that'll show up in an audit. We see this constantly.

We're not going to pretend there's an easy automated fix for this one. Refrigerant tracking requires discipline at the maintenance level — recording kilograms added, system by system, every single service visit. Software can calculate the emissions once you've got the mass. But garbage in, garbage out.

DCCEEW is proposing GWP limits for new commercial refrigeration equipment — 1,500 for most applications, 2,500 for equipment cooling below minus 20 degrees. That'll push the industry toward lower-GWP alternatives like R-449A or CO2 (R-744) systems over time. But if you're operating existing R-404A systems, your Scope 1 obligation is here right now.

Gas-Fired Process Heat: The Emission Source That's Hardest to Eliminate

Natural gas is everywhere in food and beverage manufacturing. Bread ovens. Boilers for steam generation in dairy processing. Dryers in cereal and snack production. Kill-floor hot water systems in abattoirs. Pasteurisation. Sterilisation. Spray drying for milk powder.

According to the Department of Industry, gas accounts for 37% of manufacturing energy consumption in Australia. In food processing specifically, process heat is the single largest energy use — and most of it comes from burning natural gas.

The calculation itself is simple. Under NGER Method 1 (the most commonly used), you take your gas consumption in gigajoules (which your retailer gas bill provides, or you can convert from cubic metres using the higher heating value) and multiply by 51.53 kg CO2-e per GJ. A brewery consuming 25,000 GJ of natural gas per year — typical for a mid-size operation running kettles, mash tuns, and bottle pasteurisers — generates about 1,288 tonnes of CO2-e from gas alone.

The tricky part isn't the maths. It's figuring out whether your gas consumption is heading up or down, and what to do about it. High-temperature heat pumps can now produce steam at temperatures relevant to food processing, but the capital cost is significant and the payback depends heavily on your electricity price and state-based grid factor. Electrifying process heat in Victoria (0.78 kg CO2-e/kWh) produces a very different emissions outcome than doing it in South Australia (0.22 kg CO2-e/kWh). We've seen manufacturers model the switch and find they'd actually increase Scope 2 emissions by electrifying in coal-heavy grids — which is a real "be careful what you wish for" moment.

If you're interested in how state grid factors affect these decisions, our Scope 2 electricity calculation guide walks through the maths in detail.

Do You Hit NGER Thresholds? Probably.

The NGER reporting thresholds catch more food manufacturers than people expect. The corporate group threshold is 50 kt CO2-e or 200 TJ of energy. The individual facility threshold is 25 kt CO2-e or 100 TJ.

Let's run the numbers on a hypothetical. A dairy processor with two facilities — one in Victoria, one in NSW — consuming a combined 12,000 MWh of electricity, 50,000 GJ of natural gas, 200,000 litres of diesel, and running cold storage with 1,500 kg of R-404A at a 20% leak rate.

Source	Calculation	CO2-e (tonnes)
Natural gas	50,000 GJ x 51.53 kg/GJ	2,577
Electricity (VIC, 8,000 MWh)	8,000 x 0.78	6,240
Electricity (NSW, 4,000 MWh)	4,000 x 0.64	2,560
Diesel fleet	200,000 L x 2.7 kg/L	540
Refrigerant leakage	300 kg x 4,728 GWP	1,418
Total Scope 1 + 2		13,335

That's 13,335 tonnes. Below the 50 kt emissions threshold — but run the energy calculation. The electricity alone is 43,200 GJ (12,000 MWh x 3.6). Add 50,000 GJ of gas and 7,720 GJ of diesel (200 kL x 38.6). That's just over 100,900 GJ — more than halfway to the 200 TJ corporate group threshold. Scale up to a company running four or five sites and you'll blow past it.

The ANAO found that 72% of 545 NGER reports contained errors, with 17% having significant errors. Food manufacturers running multiple sites with mixed fuel types and refrigerants are exactly the kind of reporters where errors accumulate. Our NGER thresholds guide covers the facility boundary rules in detail — and if you're a multi-site food business, the facility aggregation question is critical.

And here's the kicker that catches people off guard: NGER reporters get automatically pulled into ASRS Group 2 mandatory climate reporting via the registration pathway. So even if you thought ASRS was only for big public companies, if your energy consumption triggers NGER, you may be reporting under both regimes.

Scope 3: Where 70% of Your Emissions Actually Sit

This is the part where carbon accounting for food and beverage manufacturing gets genuinely hard. And where we need to be honest about what's possible and what isn't — at least right now.

Oxford Economics and industry bodies consistently put Scope 3 at more than 70% of total food sector emissions. For some product categories, it's closer to 90%. The biggest sources are agricultural inputs (Category 1: purchased goods and services), upstream transport (Category 4), packaging materials (Category 1 again), and downstream distribution (Category 9).

If you're a meat processor, your purchased cattle carry enteric methane emissions that dwarf everything happening inside your facility. If you're a dairy manufacturer, the on-farm emissions from feed production, manure management, and livestock alone can represent 70% of the carbon intensity of every litre of milk you process. A bread manufacturer's wheat inputs carry fertiliser emissions — nitrous oxide from synthetic nitrogen has a GWP of 265 under AR5 (298 under AR6).

We've written extensively about how to collect Scope 3 data from suppliers and the practical reality is that most agricultural suppliers can't give you activity-based emissions data. Not yet. Spend-based estimation using industry-average emission factors is where most food manufacturers start. It's imprecise — BCG research suggests spend-based methods can carry 30-40% error margins — but under the ASRS modified liability provisions for Scope 3, it's an acceptable starting point for early reporting years.

The pressure is real though. Woolworths has SBTi-validated targets requiring a 40% reduction in FLAG (forest, land and agriculture) Scope 3 emissions by FY2033, and a 55% reduction in energy and industrial Scope 3 by the same date. Coles is under similar pressure. These retailers will systematically push carbon data requirements down to their suppliers. If you manufacture food or beverages and sell through major Australian grocery channels, you will be asked for product-level carbon data. Not "sometime." Soon.

This is exactly why product-level life cycle assessment — LCA — matters for this industry more than almost any other.

Product Carbon Footprints: The Coming Requirement

Here's where the conversation shifts from "how do I comply with NGER" to "how do I keep my shelf space."

The EU's Packaging and Packaging Waste Regulation (PPWR) entered into force in 2025 with progressive milestones through 2035. Product carbon footprint declarations are becoming standard for goods entering European markets. In Australia, with the introduction of mandatory climate disclosures under the Corporations Act, large corporations are now expecting supply chain partners to provide product carbon footprints (PCFs) annually to support their own reporting.

For a food or beverage manufacturer, a PCF means tracing emissions from raw material sourcing through processing, packaging, distribution, retail, and end-of-life. That's a full cradle-to-grave (or cradle-to-gate, depending on the scope boundary) life cycle assessment.

Most Australian carbon accounting platforms don't offer LCA at the product level. They'll track your facility-level emissions, maybe split by scope, and generate a corporate report. But they won't tell you that your 500ml oat milk carton carries 0.4 kg CO2-e while your 1L full-cream milk bottle carries 1.3 kg CO2-e. That's the granularity your retail customers are heading toward. And it's the granularity the European market already expects.

We built LCA into Carbonly's platform specifically because we kept hearing food manufacturers ask for it. The module maps emissions from raw material inputs through your processing steps to finished product, using the same AI-powered document extraction that handles utility bills and invoices — but applied to supplier data, freight records, and packaging specs. It's not perfect for every product category (we're still figuring out the best approach for complex multi-ingredient products with 30+ inputs sourced from different suppliers with wildly different farming practices). But it gives manufacturers a defensible starting point for product-level carbon claims — which is what your customers and your auditors actually need.

The Safeguard Mechanism Catches Big Food Processors Too

When people think Safeguard Mechanism, they think mines and gas plants. But the scheme applies to any facility exceeding 100 kt CO2-e in Scope 1 emissions per year. Large sugar mills — burning thousands of tonnes of bagasse — can get close. Major meatworks with massive refrigeration loads and gas-fired rendering operations aren't far off. And the baseline decline rate of 4.9% per year means facilities that were comfortable five years ago may be staring down an excess emissions position soon.

If your facility is a Safeguard facility, you need to either reduce emissions or purchase Australian Carbon Credit Units (ACCUs) to cover the gap. With the secondary market hovering around $35-40 per ACCU and the cost containment price at $82.68 for 2025-26, the financial incentive to actually measure and reduce — rather than just buy your way out — is significant. We cover the Safeguard Mechanism changes in detail here.

Even if you're below the 100 kt threshold, the direction of travel matters. The government has signalled that coverage may expand. And your NGER data — which flows directly to the Clean Energy Regulator — is the basis for determining whether you're captured.

What to Actually Do About It

We could write another 3,000 words on this topic and still not cover every nuance of food manufacturing emissions. There are entire sub-problems we've barely touched — wastewater treatment emissions, food waste decomposition in landfill (Category 5 Scope 3), employee commuting for shift workers across multiple sites, and the boundary question of contract manufacturing (are they your Scope 1 or your Scope 3?).

But here's what we'd tell a food or beverage manufacturer starting from scratch.

Start with your gas and electricity bills. This is your Scope 1 and 2. It's the most defensible, most auditable, most immediately useful data you have. Get 24 months of bills, apply the correct NGA Factors for your state, and calculate your baseline. If you're running spreadsheets for this, it'll take a few weeks. Carbonly can do it in about an afternoon — our AI reads any bill format and matches it to the right emission factor automatically.

Audit your refrigerant records. Call your refrigeration maintenance contractors and ask for service records showing kilograms of refrigerant added per system, per service visit, for the past two years. If they can't provide that, you've got a data quality problem you need to fix before next reporting season. This is a manual process. There's no shortcut.

Map your Scope 3 by spend. Take your top 20 suppliers by annual spend. That'll cover 60-80% of your purchased goods emissions. Use spend-based emission factors as a starting point (the GHG Protocol's EEIO factors or Australian-specific factors from DCCEEW). Yes, it's imprecise. It's still infinitely better than a blank cell in your ASRS disclosure.

Have the product carbon footprint conversation now. If you supply Woolworths, Coles, Aldi, or any major retailer, ask your key account manager what emissions data they'll need from you in 12 months. Don't wait to be surprised.

The food and beverage sector contributes over $170 billion to Australia's economy and employs more than 290,000 people. It's also sitting on one of the most complex emissions profiles of any manufacturing sector. Getting carbon accounting right isn't a nice-to-have any more — it's the cost of staying in business.

If you're a food or beverage manufacturer trying to work out where your emissions actually come from, talk to our team. We built Carbonly to handle exactly this kind of complexity — multi-site energy data, refrigerant tracking, and product-level LCA — without the six-figure consulting bill.

Related Reading: