Carbon Accounting for Water Utilities in Australia

Melbourne Water's Western Treatment Plant generates over 86,000 MWh of renewable electricity per year from captured biogas. In the same breath, the sewage treatment process that produces that biogas is also Melbourne Water's single largest source of greenhouse gas emissions — roughly 85% of its total footprint. That's the paradox at the centre of carbon accounting for water utilities in Australia. Your biggest emission source and your best decarbonisation opportunity are the same thing.

The Australian water sector contributes an estimated 2-3% of national carbon emissions, according to Aurecon's technical analysis for the Australian Water Association. That sounds small until you realise it's roughly a million tonnes of CO2-e per year from Victoria's water corporations alone. Melbourne Water accounts for about 408,000 tonnes of that — 51% of the entire Victorian water sector total. These aren't small operations flying under the radar. Sydney Water, Melbourne Water, SA Water, Water Corporation in WA, and the various Queensland urban water entities are all NGER reporters, many of them sitting well above the 50 kt corporate threshold.

And the emissions profile is nothing like a standard commercial building or even a manufacturing plant. Water utilities deal with fugitive methane from anaerobic wastewater treatment, nitrous oxide from biological nitrogen removal, massive electricity loads for pumping and desalination, chemical supply chains, fleet vehicles, and biosolids that keep emitting long after they leave the treatment plant. If you're responsible for reporting these emissions — whether under NGER, ASRS Group 2, or as part of a state government's own net zero target — you need to understand why this sector is genuinely difficult to account for.

Scope 1: The methane and nitrous oxide problem

Scope 1 emissions from water utilities are dominated by two gases that most carbon accounting teams outside the sector barely think about: methane (CH4) and nitrous oxide (N2O). Both come from the biological processes that treat wastewater, and both have warming potentials that make CO2 look mild by comparison.

Under the NGA Factors 2025, methane has a global warming potential of 28 (AR5 values, which is what NGER uses). Nitrous oxide sits at 265. So a relatively small quantity of either gas, measured in tonnes, translates into a substantial CO2-equivalent figure.

Here's where it gets messy. The NGER framework provides default emission factors for wastewater treatment based on treatment type — Table 17 in the NGA Factors workbook. But these defaults are population-based estimates using standard per-capita organic loading and nitrogen excretion assumptions. They work tolerably for a standard secondary treatment plant handling domestic sewage. They fall apart when your plant handles significant industrial loads, when you're running anaerobic lagoons alongside activated sludge, or when your treatment configuration doesn't match the neat categories in the guideline.

Melbourne Water ran an innovation competition specifically targeting Scope 1 reductions because they found that Scope 2 (electricity) could be addressed through renewable energy procurement, but the methane and N2O from sewage treatment had no straightforward fix. Their Eastern Treatment Plant's sludge drying pans alone account for about 43,000 tonnes CO2-e per year. That's a single emission source at a single plant generating more than many entire NGER-registered corporations.

The N2O measurement challenge is particularly frustrating. Australia's NGER default factor is 0.5% of influent total nitrogen, but actual measurements at treatment plants in Australia and internationally have found emission factors ranging from near zero to 1.8% — a variation that could change a utility's reported N2O emissions by a factor of three or more. The IPCC's 2019 refinement guidelines increased the default to 1.6% of influent TN, which is substantially higher than what NGER currently uses. We're not sure yet whether the CER will move to align with the updated IPCC factor, but if they do, water utilities could see their reported Scope 1 emissions jump significantly without any actual change in operations.

And then there's the biogas itself. Anaerobic digestion produces biogas — roughly 60-70% methane — which is either flared, used for on-site power generation, or (in a handful of Australian plants) upgraded to biomethane and injected into the gas network. Sydney Water and Jemena completed Australia's first biomethane-to-gas-grid injection at the Malabar wastewater treatment plant in June 2023, producing around 95 terajoules of renewable gas per year. When that biogas is captured and combusted, it reduces greenhouse gas emissions by over 90% compared to letting the methane vent to atmosphere. But the accounting treatment matters: captured and combusted biogas still generates CO2 (biogenic, so reported differently), and any fugitive leaks from digesters, gas lines, or lagoon covers before the biogas reaches combustion are pure methane emissions you need to quantify.

Scope 2: Pumping is everything

If Scope 1 is about biology, Scope 2 is about physics. Moving water uphill — from reservoirs to treatment plants to distribution networks — takes enormous amounts of electricity. Pumping typically accounts for 70-80% of a water utility's total electricity consumption.

This is where state-based emission factors create real reporting complexity. A pumping station in Victoria running at the same load as one in South Australia will produce wildly different Scope 2 emissions because the grid factors are so different. Under the NGA Factors 2025, Victoria's location-based emission factor is 0.78 kg CO2-e/kWh. South Australia's is 0.22 kg CO2-e/kWh. That's a 3.5x difference for an identical electricity load.

For a water utility operating across state boundaries — and several do, particularly in border regions or where bulk water supply crosses jurisdictions — you can't use a single emission factor across the portfolio. Every pump station, every treatment plant, every office needs to be mapped to the correct state grid.

Desalination makes this worse. Reverse osmosis desalination plants consume 3.5 to 4.1 kWh per cubic metre of water produced, compared to 0.5 to 1.5 kWh for conventional surface water treatment and distribution. Perth's Kwinana desalination plant runs at about 4.1 kWh per cubic metre. Melbourne's experience is instructive: annual per-capita energy use for water supply increased by 859% — from 15.7 to 150.6 kWh per person per year — after adding desalination capacity. When a drought hits and desalination ramps up, a utility's Scope 2 emissions can spike dramatically in a single reporting year.

The good news is that Scope 2 is the more tractable problem. Most large Australian water utilities have moved aggressively on renewable energy procurement. SA Water's Bolivar wastewater treatment plant hit 112% energy self-sufficiency in July 2022 — generating 3,099 MWh of renewable energy in a single month through biogas combustion and solar, with excess exported to the grid. Their average self-sufficiency at Bolivar climbed from 85% to 95% over twelve months. SA Water's broader investment in more than 500,000 solar panels across 35 sites is projected to cut 89,000 tonnes of CO2-e annually. Melbourne Water has committed to sourcing 100% of its electricity from renewables by end of 2025, with a net zero by 2030 target.

But here's the accounting wrinkle. Under NGER, power purchase agreements and renewable energy certificates affect your Scope 2 reporting through the voluntary market-based method, introduced from 2023-24. Under AASB S2 paragraph 29(a)(v), you must report location-based Scope 2 emissions, and you may additionally report market-based. So even if a utility buys 100% renewable electricity, their location-based Scope 2 number won't change. That catches some sustainability managers off guard when they're preparing dual-framework reports.

The NGER reporting specifics that trip water utilities up

Water utilities that hit the NGER thresholds — 50 kt CO2-e corporate or 25 kt CO2-e facility — have specific methodological requirements that go beyond simple electricity and gas bill tallying.

Wastewater emissions must be calculated using the NGER (Measurement) Determination methods, which require you to estimate methane generation based on the chemical oxygen demand (COD) or biochemical oxygen demand (BOD) of the wastewater, multiplied by a methane correction factor that varies by treatment type. An anaerobic deep lagoon has a different correction factor than an aerobic activated sludge system. A plant that uses both — and many do, in series — needs to account for each treatment stage separately.

The calculation formula is essentially: Emissions (CO2-e) = Population x EF (from Table 17), for the default method. But any large utility doing this properly will use higher-tier methods that rely on actual influent load data, measured BOD/COD concentrations, and process-specific methane correction factors. The difference between the default population-based method and actual measurement can be 30-50%, which means your Scope 1 number could be wildly inaccurate on the default approach.

There's also a technical mismatch that water utility reporting teams need to watch. NGER uses AR5 global warming potential values (28 for CH4, 265 for N2O). AASB S2 requires AR6 values (27.9 for CH4, 273 for N2O). For most companies the difference is rounding error, but for water utilities where methane and N2O are the dominant Scope 1 gases, the divergence is more noticeable. You'll need separate calculations — or at minimum, a conversion layer — if you're reporting under both frameworks.

The ANAO performance audit found that 72% of 545 NGER reports contained errors, with 17% having significant errors. Water utilities, with their multiple treatment processes and site-specific measurement methods, are exactly the type of complex reporter where errors compound. A wrong methane correction factor applied to the wrong treatment stage at one plant can swing the corporate total by thousands of tonnes.

Scope 3: Chemicals, pipes, and the hot water your customers use

WSAA's 2024 guide on Scope 3 emissions management for the water sector identified chemicals and embodied carbon in infrastructure as the two biggest Scope 3 sources most utilities undercount. Chlorine, aluminium sulfate, polymer flocculants, sodium hypochlorite — the chemical supply chain for water treatment is substantial, and the emissions embodied in manufacturing and transporting those chemicals add up.

Then there's the infrastructure itself. Thousands of kilometres of pipes, most of them steel, ductile iron, PVC, or concrete. The embodied carbon in a water distribution network is enormous but rarely calculated because the assets were laid decades ago and the data doesn't exist in any usable format. We're honest about this: Scope 3 for water utilities is still early days, and very few utilities publicly report on it in any detail.

There's one Scope 3 category that's uniquely large for water: end-use energy for heating water. Category 11 (use of sold products) is technically where this sits, and for a water utility that supplies residential customers, the emissions from gas and electricity used to heat that water at the household level dwarf the utility's own Scope 1 and 2 combined. Whether utilities choose to report this — and how they calculate it — is still a matter of some debate within the sector. WSAA's guidance acknowledges the challenge but stops short of prescribing a single methodology.

Biogas: The emission that's also an asset

We touched on this earlier, but biogas deserves its own discussion because it's the most unique feature of water utility emissions — and the biggest abatement opportunity.

Melbourne Water's Western Treatment Plant increased its biogas electricity generation from 70,000 to over 86,000 MWh per year after commissioning a second power station. That's enough to power about 8,600 homes. SA Water's Bolivar plant processes 150 million litres of sewage daily and regularly exceeds 95% energy self-sufficiency from biogas and solar.

The Malabar biomethane project with Sydney Water and Jemena takes it further — upgrading biogas to pipeline-quality biomethane and injecting it into the gas distribution network. Production capacity is 95 TJ per year, equivalent to the annual gas consumption of roughly 6,300 homes. This earned Australia's first GreenPower Gas certification.

But the carbon accounting is tricky. You need to accurately measure total biogas produced, biogas captured versus fugitive losses, energy content of the captured gas, and emissions from combustion (including the CO2, residual CH4, and N2O from the engine or flare). A utility that over-reports biogas capture efficiency and under-reports fugitive losses will under-state its Scope 1 emissions. There's genuine measurement uncertainty here — fugitive losses from lagoon covers, pipe joints, and pressure relief valves are difficult to quantify with precision. Melbourne Water's real greenhouse gas emissions measurement program acknowledges this gap explicitly.

For NGER reporting, biogenic CO2 from biogas combustion is reported separately and doesn't count towards your Scope 1 total. But the uncombusted methane — anything that escapes before it hits the flare or engine — absolutely does count. And that's the number most utilities find hardest to pin down.

The data collection reality

A typical large water utility operates dozens to hundreds of sites: treatment plants, pump stations, reservoirs, offices, depots, laboratories. Each pump station has its own electricity meter (sometimes several). Each treatment plant has process instrumentation generating operational data. Fleet vehicles have fuel cards. Chemical deliveries come with invoices from multiple suppliers.

The bill volume alone is substantial. A utility with 200 pump stations, 10 treatment plants, 5 offices, and a fleet of 300 vehicles might process 3,000 to 4,000 utility invoices per year — before you count chemical supply invoices, waste disposal records, and fleet fuel transactions.

When we built Carbonly's document processing pipeline, water utilities were one of the use cases that pushed us to support every input format — PDF invoices, CSV meter data exports, Excel workbooks with 30 tabs of pump station data, even scanned paper records from older facilities. The anomaly detection module turned out to be particularly useful for pump stations: a sudden spike in electricity consumption at a remote pumping station often signals a mechanical issue (a failing pump draws more current), and catching that early has both maintenance and emissions implications. If a pump station's kWh consumption doubles for three months because of a worn impeller, your Scope 2 number for that site doubles too — and if nobody notices, it goes into your NGER return as if that's the normal baseline.

We should be honest about the limits here. Some water utility data — particularly process emissions from wastewater treatment — can't be extracted from invoices or meter readings. It requires process data: influent flow rates, BOD/COD concentrations, nitrogen loading, sludge production volumes, biogas flow and composition. That data lives in SCADA systems and laboratory information management systems, not in PDFs. Pulling it into a carbon accounting platform requires integration work that goes beyond document processing. Our team's background in industrial data platforms — SCADA, OSI PI historian, and similar systems from years working with BHP and Rio Tinto — means we understand the integration challenge, but we won't pretend it's solved with a drag-and-drop upload.

Where to start if you're a water utility

If you're an environment or sustainability manager at a water corporation or council-owned utility, the first thing to do is separate the tractable from the hard.

Scope 2 from electricity is tractable. You have the bills. You know the NGA emission factors for your state. Map every metered connection to a site, apply the correct state-based factor, and you've got a defensible Scope 2 number. If you're operating in multiple states, don't average the factors — calculate site by site. This is the part of your NGER return (and your ASRS disclosure) where errors are least forgivable because the data and methodology are well-established.

Fleet emissions are tractable too. Fuel card data, mapped to vehicle type and fuel type, with NGA Factors applied. Not glamorous work, but it's standard Scope 1 reporting.

Wastewater treatment emissions are harder, and that's okay. Start with the default NGER method using population-based estimates and Table 17 factors, but document your treatment configuration clearly. If your plant has a non-standard process train — say, a Membrane Bioreactor instead of conventional activated sludge — flag it. The default factors may not represent your actual emissions accurately, and you'll want to build the case for moving to a higher-tier method over time.

Biogas accounting requires instrumentation. If you don't have gas flow meters on your digesters and flares, the numbers you report are estimates with wide uncertainty bands. Installing proper metering is the single highest-value investment you can make for emissions accuracy — it improves both your Scope 1 reporting and your understanding of biogas energy recovery potential.

Then do the NGER threshold check. Many mid-size water utilities sit right around the 50 kt CO2-e corporate threshold, and whether you're above or below depends on how accurately you've estimated your Scope 1 wastewater emissions. Get that number wrong in the conservative direction and you might fail to register when you should. Get it wrong in the other direction and you're over-reporting by a margin that an auditor will question.

If you're a smaller, council-owned water utility that doesn't hit the NGER threshold, check whether your parent council falls into ASRS Group 3. From July 2027, smaller reporting entities face mandatory climate disclosure too, and the utility operations within a council group will be some of the most material emission sources to disclose.

The sector is moving fast. Victorian water corporations have legislated net zero targets for 2030 and 2050. Sydney Water is investing in PPAs and biomethane. SA Water is already approaching 100% renewable energy across some of its largest facilities. The utilities that started measuring properly three years ago are now making informed abatement decisions. The ones still guessing are about to find out their numbers don't hold up under assurance.

Get the measurement right first. The abatement strategy follows from the data — not the other way around.

If your water utility is drowning in pump station electricity bills and treatment plant data, Carbonly.ai can help. Our platform reads any document format, matches emission factors automatically using NGA Factors 2025, and flags consumption anomalies across hundreds of sites. Book a demo and bring your messiest data — that's where we do our best work.

Related Reading: