Carbon Accounting for Healthcare and Hospitals in Australia

One bottle of desflurane — the 240 mL bottle sitting in the anaesthetics storeroom right now — has the same carbon footprint as driving a car 3,539 kilometres. One bottle of sevoflurane? 196 km. Same clinical outcome. Eighteen times the emissions difference.

And until NSW removed desflurane from its public hospital formulary in March 2024, most hospitals weren't tracking this at all. Western Australia delisted it in August 2023. But private hospitals across the country are still buying the stuff, with no reporting requirement and no plan to stop.

That's a small window into a much bigger problem. Healthcare sustainability reporting in Australia is years behind other sectors — despite healthcare contributing an estimated 23.5 Mt CO2-e per year, or 5.3% of national emissions, according to the Australian Centre for Disease Control's 2021-22 baseline estimates. The Lancet Planetary Health put it higher at 7% using an economic input-output lifecycle approach. Either way, we're talking about an emissions footprint roughly equivalent to the entire state of South Australia. And most of it is untracked.

Why Hospitals Are Different from Every Other Building

We've built carbon accounting systems for commercial property, construction sites, and warehouses. Hospitals are a different beast entirely.

A typical office building uses electricity for lighting, HVAC, and some IT equipment. It runs roughly 10-12 hours a day, five days a week. A hospital runs every hour of every day with no downtime. Ever. The HVAC systems alone are enormous because you can't just set the thermostat to 22 degrees and walk away — you've got operating theatres requiring precise temperature and humidity control, negative-pressure isolation rooms, pharmaceutical storage at controlled temperatures, and MRI machines that need dedicated cooling systems running constantly.

Australian public hospitals consumed 4,122 GWh of energy in 2018-19, according to research published in the Medical Journal of Australia. That's enough to power 630,000 average homes. Electricity made up 61% of that (2,504 GWh), natural gas another 35% (1,436 GWh), and renewables just 2.3%.

Read that again. 2.3% renewable energy across 693 public hospitals. Queensland accounted for 76% of that small figure through a single government purchasing agreement. Most states were near zero.

But electricity is only part of the story — and honestly, it's the easier part to account for. The sources that make hospital carbon accounting genuinely hard are the ones most sustainability teams haven't even started measuring.

Scope 1: The Emissions Nobody's Counting

When most people think about hospital emissions, they think electricity bills. Scope 2. That's the wrong place to start.

Hospital Scope 1 emissions are unusual, significant, and frequently missed. Here are the big ones.

Anaesthetic gases are the headline act. Desflurane has a 100-year global warming potential (GWP) of 2,540 — meaning each kilogram released into the atmosphere has the warming effect of 2,540 kg of CO2. Sevoflurane sits at 130-144. Nitrous oxide is 273. These gases are vented directly into the atmosphere through the hospital's waste anaesthetic gas disposal system. They don't get filtered. They don't get captured. They go straight up.

A 2025 study in the Medical Journal of Australia tracked anaesthetic gas emissions from 2002 to 2022 and found total emissions of 21,802 t CO2-e in 2022 — down from a peak around 2012-2013 when desflurane accounted for 88% of anaesthetic emissions. By 2022, desflurane's share had dropped to 68%, largely because NSW and WA pulled it from formularies. But that's still the majority. And it's still going to atmosphere.

The calculation is dead simple: number of bottles purchased x volume x density x GWP = total CO2-e. You don't need a flow meter. You need a procurement record.

Backup diesel generators are another Scope 1 source that every hospital has and few account for properly. Hospitals are required to maintain backup power — it's a life-safety obligation. Those generators run on diesel, and they get tested regularly. A large tertiary hospital might test generators weekly, burning through thousands of litres per year even without a single power outage. Each litre of diesel produces approximately 2.7 kg CO2-e (using NGA Factors for diesel oil combustion).

Natural gas boilers for heating, sterilisation, and hot water make up the largest Scope 1 source by volume for most hospitals. With 35% of total hospital energy coming from gas, this isn't trivial.

Fleet vehicles — patient transport, pathology couriers, between-campus shuttles — are Scope 1 if the hospital owns or leases them. Ambulance services have found that over 60% of their total Scope 1 and 2 emissions come from the vehicle fleet alone.

Medical gas systems running piped nitrous oxide also leak. Not dramatically, but persistently — and at a GWP of 273, even small leaks matter over a full year.

Here's an honest admission: we're still not sure how most hospitals should handle metered-dose inhaler (MDI) propellant emissions. The HFC propellants in puffers (HFA-134a has a GWP of 1,430) are technically released on hospital premises when administered to patients, but they're prescribed pharmaceuticals, not operational emissions. The NHS counts them. Most Australian hospitals don't count them anywhere. It's a grey area we haven't seen anyone resolve cleanly.

Scope 2: Massive, But at Least Measurable

Electricity is the biggest single emissions source for most Australian hospitals, and it's also the most straightforward to calculate. The formula is the same as any commercial building: kWh consumed multiplied by the relevant state-based emission factor from the NGA Factors workbook.

But the scale is different. A 500-bed tertiary hospital can easily consume 40-60 GWh per year. At Victoria's emission factor of 0.78 kg CO2-e/kWh, that's 31,200-46,800 tonnes of CO2-e from electricity alone. A hospital in Tasmania running the same load would report 8,000-12,000 t CO2-e because the grid factor there is 0.20. Same hospital, same consumption, four times the emissions difference depending on which state you're in.

That state-level variation matters enormously for health services that operate across jurisdictions. A state health department running 30 hospitals across metro and regional sites can't use a single factor for the lot. Each site needs the correct grid emission factor for its location and connection point.

The other thing that catches hospital sustainability teams: the sheer number of meters. A large hospital campus might have 15-30 electricity meters across different buildings, substations, and separately metered tenancies (like the cafe, the car park, or a co-located research facility). Pulling consumption data from that many meters, across quarterly billing cycles from potentially different retailers, and reconciling it against NMI numbers — that's where the real time sink is. Not the maths. The data wrangling.

Scope 3: Where 70% of the Footprint Actually Sits

Here's the uncomfortable truth that most hospital sustainability reports dodge. Between 50% and 75% of healthcare's carbon footprint comes from Scope 3 — the supply chain. Pharmaceuticals. Medical devices. Surgical consumables. Food services. Linen. Waste disposal. IT equipment. Construction and maintenance.

The OECD found that supply chain procurement (manufacturing, packaging, and transport of goods) is the dominant source of healthcare emissions globally. In England's NHS data, medical equipment contributed 13.1% and pharmaceuticals 12.1% of total Scope 3 emissions.

For an Australian hospital, the biggest Scope 3 categories are typically:

Pharmaceuticals — drug manufacturing is energy-intensive, often offshore, and emissions data from suppliers ranges from excellent to non-existent
Medical devices and consumables — single-use surgical items, PPE, implants, prosthetics — the supply chain is global and opaque
Food services — patient meals, staff cafeterias, food waste
Clinical waste treatment — Australia generates approximately 42,000 tonnes of clinical waste per year, with high-temperature incineration producing around 1,074 kg CO2-e per tonne
Patient and visitor travel — for a major metro hospital, this can be substantial
Capital works — hospital construction and renovation projects carry large embodied carbon

We won't pretend Scope 3 is easy for hospitals. It isn't. The pharmaceutical supply chain alone crosses dozens of countries and hundreds of intermediaries. Getting product-level emissions data from a medical device manufacturer in Germany or a generic drug supplier in India is a multi-year project, not a spreadsheet exercise.

But ASRS mandatory reporting requires Scope 3 from the second reporting period. For Group 1 health entities (those meeting the $500M+ revenue or $1B+ gross assets test), that clock is already ticking. And here's the thing — several state health departments and large private hospital groups absolutely clear the NGER thresholds, which means they're also automatically pulled into ASRS Group 2 from July 2026.

So if you're a sustainability officer at a state health service thinking Scope 3 is a problem for later: it isn't.

Where to Actually Start

We've talked to enough hospital administrators to know that the gap between "we should be doing this" and "here's our emissions inventory" feels enormous. It doesn't have to be.

Start with what you can count right now. Electricity and gas bills give you Scope 2 and most of your Scope 1 natural gas emissions. If you're a multi-site health service, you probably have 200-500 utility bills per quarter across your portfolio. That's your foundation. Get the data out of the bills, apply the right NGA emission factors by state, and you've got 40-50% of your direct footprint mapped.

Then go to procurement records for anaesthetic gases. Your pharmacy department tracks every bottle purchased. Multiply by the published GWP values: desflurane at 2,540, sevoflurane at 130, nitrous oxide at 273. This doesn't require fancy equipment. It requires someone asking the pharmacy for a report.

Diesel fuel records for generators and fleet come from your facilities management team or fleet manager. Litres consumed multiplied by the NGA emission factor for diesel gets you there.

For Scope 3, start with spend-based estimates. You won't get supplier-specific data in year one. Use your procurement spend by category, apply the relevant spend-based emission factors from the NGA Factors workbook or a reputable database, and acknowledge the uncertainty. An 80%-accurate Scope 3 estimate based on spend data is infinitely more useful than a perfectly precise zero because you haven't started yet.

This isn't a two-year consulting project. A mid-size hospital (200-400 beds) with decent financial records and utility data should be able to produce a credible first emissions inventory — Scope 1, 2, and a spend-based Scope 3 — within 8-12 weeks. We've seen it take less when the data's already digital and someone actually owns the project.

The Regulatory Pressure Is Real and Increasing

Australia's first National Health and Climate Strategy, launched at COP28 in December 2023, set a vision for a "net zero health system." The Australian Centre for Disease Control is now publishing baseline emissions estimates. Three governance groups were established in 2024 to drive implementation.

But the actual mandatory reporting requirements don't come from the health strategy. They come from NGER and ASRS. Any hospital or health service that crosses the NGER thresholds — 25 kt CO2-e at the facility level or 50 kt CO2-e at the corporate group level — is already required to report to the Clean Energy Regulator by 31 October each year. And given that a single large hospital can produce 30,000-50,000 tonnes of CO2-e from electricity and gas alone, plenty of health services are well above these thresholds.

The ANAO found that 72% of NGER reports contained errors. In healthcare, where the data sources are more complex than most sectors — anaesthetic gas procurement, piped medical gas systems, diesel generators, multi-site electricity metering — the error risk is higher, not lower.

This is where Carbonly's 18-module platform fits the healthcare picture. The AI Document Processing engine handles eight document formats (PDF, CSV, multi-sheet Excel, Word, PPT, RTF, images, and scanned documents) with 5-tier material matching and confidence scoring — which matters when you're pulling consumption data from 200-500 utility bills per quarter across a multi-campus health service, each arriving in a different format from a different retailer.

But document processing is just the starting point. The multi-facility project tracking is purpose-built for organisations like state health services that operate 30+ hospital campuses, each with its own meters, gas systems, and fleet. Each campus runs as its own facility within Carbonly, rolling up into a single corporate-level NGER return. You see your emissions at the building level, the campus level, and the health service level — all from the same data set.

The Anomaly Detection module uses AI-powered pattern recognition with five rule types and a full investigation workflow. When a hospital campus shows an energy spike — maybe an MRI chiller failed and the backup ran 24/7 for a fortnight, or a boiler system malfunctioned — the system flags the anomaly before it silently inflates your quarterly numbers. For hospitals where 24/7 operations mean consumption patterns are usually stable, anomalies are easier to detect and more important to catch.

The Incident Management module tracks environmental incidents with a structured investigation workflow — medical gas leaks from piped nitrous oxide systems, refrigerant releases from HVAC maintenance, diesel spills at generator facilities. Each incident links directly to the affected emission records, so your NGER return reflects what actually happened, not what your spreadsheet assumed.

And the Scheduled Reports module delivers automated reports — NGER, GHG Protocol, Custom, or Executive Summary format — on a recurring schedule to the people who need them. Monthly emissions summaries to the hospital board, quarterly NGER progress reports to the CFO, annual reports to the state health department — configured once and delivered without someone remembering to run the numbers and email a PDF.

But we also know that no tool solves the harder problem: getting a hospital's leadership to treat emissions reporting with the same rigour they apply to clinical data.

That's the real gap. Not technology. Culture.

If you're at a hospital or health service that hasn't started its emissions inventory, the move is simple. Pull your electricity bills, your gas bills, and your anaesthetic gas purchasing records for the last financial year. That gives you enough to see where you stand against the NGER thresholds — and enough to know whether you've been ignoring a reporting obligation you already have.

Related Reading: