Emission Factor Databases: Which Ones Australian Companies Actually Need

We plugged the same electricity consumption data into three different emission factor databases last quarter and got three materially different answers. Not 5% different. Thirty percent different. The factor source you pick for a single line item can shift your reported emissions by more than some companies emit in an entire category.

That's the thing nobody warns you about when you start carbon accounting in Australia. The NGA Factors workbook tells you exactly what to use for Scope 1 and 2. It's published by DCCEEW, mandated for NGER, and updated annually. But the moment you step into Scope 3 — purchased goods, business travel, capital goods, upstream transport — you're staring at a dozen international databases with different methodologies, different currencies, different base years, and wildly different numbers for what is supposed to be the same activity.

This isn't a theoretical problem. ASRS disclosures under AASB S2 require you to document your emission factor sources. Auditors will check them. And with limited assurance requirements tightening for Group 1 entities through 2026, "we found a number on the internet" doesn't cut it anymore.

Here's the reality: no single emission factor database covers everything an Australian business needs. You'll use at least two. Probably three or four. The question is which ones, and when.

NGA Factors: The Non-Negotiable Starting Point

The National Greenhouse Accounts Factors workbook is an Excel file. Published annually by DCCEEW, it contains the emission factors that every NGER reporter is legally required to use and every ASRS reporter will default to for Scope 1 and 2 disclosures.

The 2025 edition — which applies to the 2025-26 NGER reporting year — covers stationary energy (natural gas, diesel, LPG, coal), transport fuels, fugitive emissions, industrial processes, and electricity grid factors by state. The electricity factors in Table 1 (location-based) are what most companies need first.

The state-by-state variation is enormous. Victoria's grid sits at 0.78 kg CO2-e per kWh. Tasmania is 0.20. South Australia is 0.22. If you're running multi-site operations across states — and plenty of property managers, retailers, and construction companies are — using the national average of 0.62 instead of state-specific factors is a reportable error. We've written about this in detail in our NGA Factors deep dive.

NGA is free. It's authoritative. It's the only emission factor database that matters for NGER compliance. But it has two significant gaps. First, it uses AR5 Global Warming Potential values (methane = 28), while AASB S2 requires AR6 values (methane = 27.9 for fossil, 27.0 for non-fossil). This sounds minor until you're reporting fugitive methane from landfills or gas operations and your NGER and ASRS numbers need to reconcile. Second, and more importantly, NGA has essentially nothing for Scope 3 spend-based calculations. It provides activity-based factors — kWh, GJ, litres — not dollar-denominated factors. Once you need to estimate emissions from purchased goods and services using procurement data, NGA can't help you.

Cost: Free. Download from DCCEEW. Coverage: Scope 1 and 2 (Australian-specific). Limited Scope 3 (only transmission and distribution losses for electricity). Methodology: Activity-based. GWP: AR5. Update frequency: Annual.

EXIOBASE: The Scope 3 Workhorse (With Caveats)

When Australian companies first tackle spend-based Scope 3 emissions, they almost always end up at EXIOBASE. It's the most widely used environmentally extended input-output (EEIO) model globally, and for good reason — it covers 44 countries plus five rest-of-world regions representing roughly 95% of global GDP, with around 200 product categories.

The latest versions (3.9 and 3.10) incorporate updated supply-and-use table data through 2022 and revamped energy account methodologies. EXIOBASE factors are expressed in kg CO2-e per euro, at basic prices (excluding transport, tax, and trade margins). That "basic prices" detail matters. If you're applying EXIOBASE factors to your total invoice amounts including GST and freight, you'll overestimate.

EXIOBASE is a reasonable choice for Scope 3 Category 1 (purchased goods and services) screening — figuring out which procurement categories are your big emitters so you know where to focus data collection. It's also useful for categories 2 (capital goods) and 4 (upstream transport) when you only have spend data. But it has real limitations.

The product categories are broad. "Chemicals and chemical products" is one bucket. That means industrial solvents and pharmaceutical ingredients get the same factor. The model also reflects average economic structures — it doesn't know whether your specific supplier runs a coal-fired factory or a solar-powered one. And the currency denomination creates a practical headache for Australian companies: you need to convert AUD spend to EUR, matching the factor vintage year's exchange rate, before applying the factor. Get that conversion wrong and your numbers shift by 10-20% before methodology even enters the picture.

We've seen EXIOBASE produce estimates that differ from process-based life cycle assessment (LCA) data by a factor of two or more for specific product categories. It's a screening tool, not a precision instrument.

Cost: Free for academic/research use. Commercial access typically via aggregators like Climatiq. Coverage: 44 countries, 200 product categories, Scope 3 (spend-based). Methodology: Environmentally extended input-output (EEIO). EUR-denominated. Update frequency: Periodic (versions 3.9/3.10 current).

Climatiq: The Aggregator's Aggregator

Climatiq takes a different approach. Instead of being a primary source, it aggregates emission factors from 120+ datasets — including EXIOBASE, ecoinvent, IEA, BEIS/DESNZ, EPA, and others — into a single API with 920,000+ factors across 300+ regions.

For software platforms and companies building automated carbon accounting workflows, Climatiq's API-first design is its main draw. You send an activity description and get back a factor and calculation, without needing to maintain your own database of factors from multiple sources.

The free Community tier gives you 250 API calls per month — enough for testing, not for production. The paid tiers (Data and Calculate) require contacting sales for pricing, but they include access to premium datasets like ecoinvent at additional cost, plus features like audit trails and GLEC-compliant freight calculations.

The challenge with Climatiq for Australian companies is specificity. When you search for an emission factor, you might get a UK BEIS factor, a US EPA factor, or an EXIOBASE factor — and which one the API returns depends on the matching logic. For NGER compliance, you can't use Climatiq factors; you need the NGA numbers directly. For ASRS Scope 3, Climatiq can be useful, but you need to document exactly which underlying source each factor came from because your auditor will ask.

Cost: Free tier (250 calls/month, non-commercial). Paid tiers: contact sales. Coverage: 920,000+ factors, 300+ regions, all scopes. Methodology: Aggregated from 120+ sources. Mixed (activity-based and spend-based). Update frequency: Continuous.

BEIS/DESNZ: The UK Proxy That Everyone Uses

The UK Government Greenhouse Gas Conversion Factors — historically published by BEIS, now by DESNZ (Department for Energy Security and Net Zero) — are the most widely borrowed Scope 3 factors in the English-speaking world. The 2025 edition dropped in June 2025 with updated well-to-tank factors for bioenergy and a 14.5% reduction in UK electricity carbon intensity.

Australian companies use DESNZ factors as a proxy for activities where no Australian-specific factor exists. Business travel (flight emission factors by cabin class and distance band), hotel stays, waste disposal by type, and freight transport by mode — these are all covered in DESNZ with a level of granularity that NGA doesn't provide.

The factors are GBP-denominated for spend-based items and activity-based for physical quantities (kg CO2-e per passenger-km, per tonne-km, per night). The activity-based factors are directly applicable regardless of currency — a passenger-kilometre is a passenger-kilometre whether you're in London or Brisbane. But the spend-based factors need currency conversion and carry the same inflation and exchange rate sensitivity as EXIOBASE.

We use DESNZ business travel factors in Carbonly as a default when clients don't have airline-specific data. They're well-documented, annually updated, and defensible under assurance. But they reflect UK energy mixes and supply chains, not Australian ones. For most Scope 3 categories, the difference is small enough to be acceptable under modified liability protections. For energy-intensive categories, it can matter.

Cost: Free. Published on GOV.UK. Coverage: All scopes, strong on transport, waste, energy, supply chain. UK-centric. Methodology: Mix of activity-based and spend-based. GBP-denominated for monetary factors. Update frequency: Annual (June).

IEA: The Global Electricity Benchmark

The International Energy Agency publishes country-level electricity emission factors for 190+ countries, making it the go-to source for companies with international operations needing Scope 2 or Scope 3 factors outside Australia.

If your business has offices in Singapore, manufacturing in Vietnam, and a data centre in Ireland, you need electricity emission factors for each of those countries. NGA only covers Australia. IEA fills that gap. It also provides upstream (life cycle) emission factors for electricity generation — useful for Scope 3 Category 3 (fuel- and energy-related activities not included in Scope 1 or 2).

The 2025 edition covers CO2 factors from electricity generation for world countries from 1990 to 2023, with provisional 2024 data for OECD members. Data is released each September.

There's a catch. IEA data isn't free. A single-user license for the Emissions Factors database costs approximately EUR 640 (around AUD 1,050). Multi-user and subscription options cost more. And the licensing terms are restrictive — you can't embed the raw data into client-facing tools or shared models without a developer license. For companies that only operate in Australia, IEA is unnecessary; NGA is better and free. But if you have any international Scope 2 or 3 exposure, IEA is hard to avoid.

Cost: ~EUR 640 single user (~AUD 1,050). Multi-user and developer licenses higher. Coverage: 190+ countries, electricity/energy factors only. Methodology: Activity-based. CO2 per kWh. Update frequency: Annual (September).

IPCC: The Defensible Last Resort

The IPCC Emission Factor Database (EFDB) provides Tier 1 default factors — global averages derived from the 2006 IPCC Guidelines and their 2019 Refinement. These are the broadest, least specific factors available, covering everything from stationary combustion to land use change.

You'd use IPCC Tier 1 factors when you genuinely have no country-specific or industry-specific alternative. Reporting emissions from an unusual fuel type? IPCC probably has a default. Estimating process emissions from a niche industrial activity in a country where IEA doesn't have data? IPCC again.

For Australian companies, IPCC factors are rarely the right choice for Scope 1 and 2 — NGA is always more specific. But for certain Scope 3 categories, particularly where you're estimating emissions from suppliers in developing countries with limited national inventory data, IPCC defaults are defensible and well-documented. Auditors understand them. They won't give you the most accurate number, but they'll give you a number that nobody can reasonably challenge as methodologically unsound.

Cost: Free. Available via the GHG Protocol website and IGES. Coverage: Global, all sectors, all scopes. Tier 1 defaults. Methodology: Activity-based. Global averages. Update frequency: Tied to IPCC assessment cycles (irregular).

Ecoinvent: Precise, Expensive, and Probably Not For You

Ecoinvent is the gold standard for process-based life cycle inventory data. With over 21,000 datasets covering the life cycle of products and services, it provides the most granular emission factors available — specific to process, geography, and technology. If you need to know the embodied emissions of 1 kg of hot-rolled steel produced in Germany versus cold-rolled steel from China, ecoinvent can tell you.

It's the backbone of serious LCA and embodied carbon work, and many carbon accounting platforms (including Carbonly's LCA module) use ecoinvent data for product-level carbon footprinting.

But the pricing changed in April 2025 to an annual subscription model. A single-user license starts at around EUR 2,180/year (approximately AUD 3,600). Enterprise and developer licenses are significantly more. If you're an SME doing corporate carbon accounting for NGER and ASRS compliance, ecoinvent is overkill. You don't need process-level LCA data to report your electricity, gas, and fleet emissions. You need it when you're doing product carbon footprints, EPD development, or detailed Scope 3 Category 1 analysis beyond screening level.

We're honest about this: most of the companies we work with don't need a direct ecoinvent license. They benefit from ecoinvent data indirectly — through platforms like Carbonly that have developer licenses and expose the relevant factors through our Material Library — without paying thousands per year for direct access to a database they'd use for 2% of their reporting.

Cost: From ~EUR 2,180/year single user (~AUD 3,600). Enterprise higher. Coverage: 21,000+ datasets, global, process-level. All scopes. Methodology: Process-based LCA. Most granular available. Update frequency: Regular (v3.11 current).

EPiC and NABERS: Australian Construction-Specific

For construction and property companies doing embodied carbon calculations, two Australian databases fill a gap that NGA and EXIOBASE can't touch.

The Environmental Performance in Construction (EPiC) database, developed by the University of Melbourne, covers 250+ construction materials with Australian-specific embodied energy, water, and greenhouse gas emission coefficients. It's open-access, peer-reviewed, and uses a hybrid LCA approach that combines process data with input-output data to minimise truncation error. If you need the embodied carbon of Australian-produced concrete, structural steel, or aluminium framing, EPiC is your first stop.

NABERS launched a separate National Emission Factors Database in late 2024, specifically designed to support the NABERS Embodied Carbon rating tool. It provides building product emission factors where product-specific Environmental Product Declarations (EPDs) aren't available — acting as a fallback default database for the NABERS methodology.

Both databases are free. Neither covers operational emissions — that's still NGA territory. But for the growing number of companies being asked about embodied carbon under NCC 2025 voluntary provisions or Green Star Buildings v1.1 (mandatory from May 2026 with a 10% minimum reduction requirement), these are the databases that matter.

FootprintLab: The Australian EEIO Alternative

FootprintLab deserves a mention as the most Australian-specific spend-based emission factor source available. Built on data from the Industrial Ecology Virtual Laboratory (IELab) at UNSW and the University of Sydney, it provides 344 emission factors mapped to Australian economic sectors using ABS input-output classifications.

The data already underpins Australia's Climate Active accreditation program and has been selected by the United Nations for SDG monitoring. It's peer-reviewed and published in academic journals. The ISAPC classification system maps directly to how the ABS categorises Australian economic activity, which means you're matching your procurement spend to factors built from Australian industry structure — not trying to shoehorn AUD spend into a European input-output model.

For Australian companies doing Scope 3 screening, FootprintLab's factors are arguably more appropriate than EXIOBASE for domestic procurement. You're using factors derived from the same economy your supply chain operates in. International procurement still needs EXIOBASE or similar.

Access requires contacting FootprintLab directly; it's not a free download like NGA or IPCC.

The Comparison That Actually Matters

Here's the thing nobody wants to say plainly: which emission factor database you choose can change your reported emissions by 30-50% for the same underlying activity.

Criterion	NGA	EXIOBASE	DESNZ (BEIS)	IEA	IPCC	Ecoinvent
Best for	Scope 1 & 2 (AU)	Scope 3 spend screening	Scope 3 transport/waste	International electricity	Last resort defaults	LCA / product footprints
Cost	Free	Free/via aggregator	Free	~AUD 1,050+	Free	~AUD 3,600+/yr
Australian-specific	Yes	Partially (AU is 1 of 49 regions)	No (UK)	Yes (AU included)	No (global averages)	Partially
Currency	N/A (activity-based)	EUR	GBP / activity-based	N/A (activity-based)	N/A (activity-based)	N/A (process-based)
NGER compliant	Yes (required)	No	No	No	No	No

That table tells you the structure. But it doesn't capture the methodological tension that trips people up in practice.

Consider a manufacturing company trying to estimate Scope 3 Category 1 emissions for steel purchases. Using EXIOBASE spend-based factors, you'd multiply your total steel spend (converted to EUR) by the "basic iron and steel" sector factor. Using ecoinvent, you'd identify the specific steel product (hot-rolled coil, cold-rolled sheet, galvanised), the production geography, and apply a process-based factor per tonne. The EXIOBASE number might come in at 0.4 kg CO2-e per EUR. The ecoinvent number, converted back to a per-dollar basis, might be half that — or double it — depending on which steel product and which producer.

Both numbers are defensible. Both follow recognised methodologies. And they can disagree by a factor of two. That's not a bug in the data. It's the reality of emission factors at different levels of specificity.

What Auditors Actually Check

Under AASB S2, there's no mandated emission factor database for Scope 3. The standard requires you to disclose the measurement approach, inputs, and assumptions used — including emission factor sources. What auditors check, especially under limited assurance requirements rolling out for Group 1 entities, is whether your factor choices are:

Documented. Every factor needs a source citation — database name, version, vintage year, and the specific factor ID or table reference.

Appropriate. A UK DESNZ factor for business travel is defensible. A UK DESNZ factor for Australian grid electricity is not (use NGA).

Consistent. If you used EXIOBASE for Scope 3 screening last year, switching to ecoinvent this year without disclosing the change and its impact will raise questions.

Current. Using 2020-vintage factors for 2026 reporting when a 2025 edition exists is a finding waiting to happen. NGA factors lag by approximately one year — the 2025 edition reflects 2023-24 grid data. That lag is acceptable because everyone uses the same vintage. Using a factor from three editions ago is not.

The modified liability protections under Treasury Laws Amendment (Financial Market Infrastructure and Other Measures) Act 2024 provide some cover for Scope 3 through December 2027, specifically recognising that Scope 3 measurement involves inherent uncertainty. But "modified liability" doesn't mean "no liability." If your factor selection is clearly inappropriate or undocumented, those protections won't save you.

How We Handle This in Carbonly

We built Carbonly's Material Library to pull from all of these sources because we had to. No single database gives an Australian business everything it needs.

The library includes 139+ pre-loaded NGA factors for Scope 1 and 2, plus factors from EXIOBASE, DESNZ, IEA, IPCC, ecoinvent, EPiC, and FootprintLab for Scope 3 and embodied carbon calculations. Each factor carries its source, vintage year, methodology type, and unit — so when an auditor asks where a number came from, the answer is already in the system.

The harder problem is matching. When a user uploads a utility bill or invoice, the system needs to figure out which emission factor applies. Is this a "natural gas" bill or a "town gas" bill? Is the diesel "automotive diesel" or "industrial diesel"? Is the steel "structural sections" or "reinforcing bar"? The wrong match means the wrong factor, which means the wrong emissions number.

We solve this with a 5-tier matching approach: direct match against known materials, alias matching (because "unleaded petrol" and "ULP" and "regular gasoline" are the same thing), AI-based context matching (reading the invoice to understand what was actually purchased), fuzzy string matching, and LLM-based fallback for genuinely ambiguous items. The system learns from corrections — when a user overrides a match, that correction feeds back into the matching logic for future documents.

For currency-denominated factors, the platform handles automatic conversion across AUD, USD, EUR, GBP, and NZD. When an EXIOBASE factor is in EUR and an invoice is in AUD, the system applies the conversion using the factor vintage year rate — not today's rate, not the invoice date rate. It's a detail that matters for consistency and audit trail integrity.

We're not going to pretend this solves every problem. Scope 3 factor selection still requires human judgement. When EXIOBASE and ecoinvent give you meaningfully different answers for the same category, someone has to decide which is more appropriate for your specific situation and document why. The software can flag the discrepancy. It can't make the judgement call for you.

The Practical Decision Tree

Here's how we'd approach emission factor selection for an Australian company starting carbon accounting today.

For Scope 1 and 2 under NGER: NGA Factors. Full stop. No alternatives. Use the current edition (2025 for 2025-26 reporting). Apply state-based electricity factors, not the national average. If you're also reporting under AASB S2, note the AR5/AR6 GWP difference and reconcile.

For Scope 3 screening (figuring out which categories matter): EXIOBASE or FootprintLab spend-based factors. FootprintLab for domestic procurement, EXIOBASE for international. Accept the 30-40% error margin. This is a prioritisation exercise, not a final number.

For Scope 3 reporting (going beyond screening): Move to activity-based factors where possible. DESNZ for travel and transport. IEA for international electricity. Supplier-specific data for your top 10-20 procurement categories. The GHG Protocol hierarchy is clear: supplier-specific beats hybrid beats average-data beats spend-based.

For embodied carbon and product footprints: EPiC for Australian construction materials. Ecoinvent if you're doing formal LCA. EPD Australasia for product-specific declarations.

For everything else: IPCC Tier 1 defaults as the last resort. Defensible, documented, and better than a gap in your inventory.

Start with NGA. Layer in EXIOBASE or FootprintLab for Scope 3 screening. Upgrade to activity-based factors as your data improves. Document every factor choice. That's the path from a first-year ASRS disclosure to an assurance-ready emissions inventory — and it doesn't require buying a EUR 3,600 ecoinvent license to get there.

Related Reading: