Subcontractor Emissions: The Scope 3 Data Gap in Australian Construction

Australian construction subcontracts $116 billion in work annually. Under AASB S2, those subcontractor emissions are your Scope 3 to report - but most subcontractors can't tell you their carbon footprint. Here's what data you can actually extract from procurement documents you already have.

Carbonly Team April 3, 2026 14 min read
Scope 3Construction EmissionsSubcontractor ReportingASRSNGERSupply ChainCarbon Accounting
Subcontractor Emissions: The Scope 3 Data Gap in Australian Construction

The ABS Construction Industry Survey for 2023-24 revealed something that should worry every sustainability manager at a large Australian builder. Construction services businesses earned $116.6 billion from subcontracting arrangements - more than the $71.4 billion they earned from primary contracting. Infrastructure Australia's 2025 Market Capacity Report puts subcontracting at 41% of infrastructure construction value. On any given commercial or infrastructure project, somewhere between 60% and 80% of the work is done by someone other than the head contractor.

That's not news to anyone who works in construction. But here's the part that is: under AASB S2, those subcontractor activities sit squarely in your Scope 3 Category 1 - purchased goods and services. You're required to estimate and disclose those emissions. And the vast majority of your subcontractors - the electricians, the formworkers, the concreters, the steel fixers, the plumbers, the earthmovers - have never measured a gram of CO2-e in their lives.

This isn't a minor data quality issue. For most Australian construction companies, subcontractor emissions represent the single largest gap in their greenhouse gas inventory.

The Scale of What You Don't Know

Australia has roughly 463,000 construction businesses. Of those, the ABS Construction Services subdivision alone employs 883,000 people across building installation, completion, and land development services. The overwhelming majority are SMEs with fewer than 20 employees. Only 961 corporations report under NGER. So when a Tier 1 or Tier 2 builder engages 30 to 60 subcontractors on a major project, the odds of any of them having a carbon inventory are close to zero.

But the emissions are real. A concreting subcontractor running two agitator trucks and a concrete pump consumes diesel. A steel-fixing crew's crane burns fuel every hour it operates. Formwork subcontractors use timber, which carries embodied carbon. Earthworks subcontractors move tens of thousands of litres of diesel per month through excavators, graders, and dump trucks. We covered the diesel calculation for construction in detail - at roughly 2.7 kg CO2-e per litre (NGA Factors 2025), a single earthworks subcontractor on a road project can generate 500 to 1,000 tonnes of CO2-e in a year without even trying.

Add it all up across every trade on every project, and you're looking at a Scope 3 figure that could dwarf your own direct emissions. Research consistently shows Scope 3 represents 70-90% of a construction company's total greenhouse gas footprint. Most of that comes from purchased materials (concrete, steel, aluminium - which we've covered in our embodied carbon guide) and subcontracted services. And right now, the data behind those subcontracted services is mostly a guess.

What AASB S2 Actually Requires

There's a common misconception that Scope 3 is optional or that construction companies can simply say "data not available" and move on. That's wrong. Under AASB S2, Scope 3 disclosure is mandatory from an entity's second reporting year. For Group 1 entities, that's already here. For Group 2 entities (reporting from financial years starting 1 July 2026), Scope 3 kicks in from year two. Many construction companies fall into Group 2 either through size thresholds or because their NGER reporting status pulls them in automatically.

The standard doesn't require you to have perfect data. It does require you to have a methodology, disclose your approach, and improve over time.

Paragraph B40 of AASB S2 establishes the Scope 3 measurement framework. It requires entities to prioritise inputs and assumptions that use timely data faithfully representing the jurisdiction, technology, and specific activities in the value chain. In practice, this creates a hierarchy: supplier-specific data is best, activity-based estimation using physical quantities comes next, and spend-based estimation using economic input-output factors is the fallback.

Paragraph 29(a)(iii) requires you to disclose the measurement approach, inputs, and assumptions used. Your auditor isn't just looking at the number - they want to see documented reasoning for why you used a particular method for each category, what data sources you relied on, and what the limitations are.

The modified liability protections that Parliament extended in December 2025 provide some relief. Scope 3 disclosures made in good faith are protected from private litigation during the transition period (through to 2028). But that protection doesn't mean you can skip the work. ASIC can still request corrective disclosures. And your auditor still needs evidence.

So the question isn't whether you need subcontractor emissions data. It's how you get it when your subcontractors can't give it to you directly.

Spend-Based Is Where You Start. It's Not Where You Stay.

Most construction companies reporting Scope 3 for the first time will default to spend-based estimation. You take your subcontractor expenditure, classify it by ANZSIC industry code, and multiply by economic input-output (EEIO) emission factors from datasets like the Australian Spend-Based Emission Factor database (available via Zenodo and the IELab).

It works. It's accepted by the GHG Protocol. It'll get you through your first reporting year.

But spend-based estimation has a fundamental accuracy problem in construction. Research on spend-based versus activity-based methods for construction projects shows deviations averaging around 106% - meaning the spend-based figure can be roughly double the activity-based result in some cases, or half in others. That's not a rounding error. That's a coin flip on whether your Scope 3 number is materially right.

The reasons are structural. EEIO factors are industry averages based on economic models that don't reflect what a specific subcontractor actually did on your specific project. A $2 million earthworks subcontract on a greenfield site in western Sydney has a completely different emissions profile from a $2 million fit-out subcontract in the Melbourne CBD. One burns thousands of litres of diesel. The other is mostly labour and some materials. But if they're both classified under the same ANZSIC code, the spend-based method treats them identically.

There's another problem that we don't think gets enough attention: inflation distortion. EEIO factors are typically calculated from base-year economic data. If construction costs have inflated 20-25% since the base year (and in Australia, they have), your spend figure is higher than the economic activity it actually represents. That means you're systematically overestimating emissions - sometimes by 20% or more - before you even get to the factor accuracy issue.

Spend-based gives you coverage. But it doesn't give you accuracy, it doesn't give you the ability to compare subcontractors on environmental performance, and it won't satisfy an auditor doing a reasonable assurance engagement. You need to move toward activity data. The question is how.

The Data Hiding in Your Procurement Documents

Here's the thesis of this article, and it's one we feel strongly about: you already have far more emissions-relevant data from your subcontractors than you think. It's sitting in procurement documents that flow through your systems every day. You're just not extracting it.

Think about what a subcontractor claim or payment application contains. On a well-run project, a subcontractor's monthly progress claim includes a breakdown of work completed, often with quantities. A concreting subcontractor's claim will list cubic metres poured by element - footings, slabs, walls, columns. A steel-fixing claim will list tonnes of reinforcement placed. An earthworks claim will list cubic metres of cut and fill, and sometimes fuel consumed if the contract has a rise-and-fall fuel clause.

Purchase orders and variation documents contain material specifications and quantities. Delivery dockets - which come attached to invoices or filed separately by site admin - record exactly what arrived on site: cubic metres of concrete by grade, tonnes of steel by bar size, aggregate by the truckload.

Equipment hire invoices often include operating hours and sometimes fuel consumption. A 30-tonne excavator hired for 160 hours at an average consumption of 25 litres per hour is 4,000 litres of diesel - roughly 10.8 tonnes of CO2-e. That data is on the invoice. It's just nobody's reading it for emissions purposes.

Even a bare-bones subcontractor tax invoice tells you something. The description of works, the quantities where they're listed, the materials referenced. It's not supplier-specific emissions data. But it's physical activity data - and that's a massive step up from spend-based estimation.

The problem is extraction. These documents come in hundreds of different formats. PDFs, scanned images, Excel sheets, photos of handwritten dockets. A Tier 1 builder running eight concurrent projects might process 3,000+ subcontractor documents per quarter. Nobody has the time to read every one and pull out the emissions-relevant fields.

This is where AI document processing earns its keep. Carbonly's document engine reads subcontractor invoices, claims, delivery dockets, and purchase orders in any format - PDF, scanned image, Excel, Word, even photos. It doesn't need a template for each subcontractor's format. It understands the document layout, finds the quantities and descriptions, and extracts the data fields that matter for emissions calculation: material types, quantities, units, equipment hours, fuel volumes.

Our 5-tier material matching system then maps those extracted materials to the correct emission factors. When a concreting subcontractor's delivery docket says "N32 with 30% fly ash, 6m3" - the system matches that specific concrete grade to the right factor, not a generic "concrete" average. The matching improves over time through Carbonly's AI learning loop: every confirmed match trains the system on that subcontractor's terminology, so the next claim from the same sub gets matched faster and more accurately.

We won't pretend this solves everything. Some subcontractor invoices are just a dollar amount and a vague description - "$45,000 for plumbing rough-in, progress claim 3 of 5." There's nothing to extract from that except spend data. And handwritten dockets from site deliveries are sometimes genuinely illegible. We're not going to claim 100% extraction accuracy because that wouldn't be honest. But even getting activity data from 40-50% of your subcontractor documents - the ones that do contain quantities and material descriptions - shifts a significant chunk of your Scope 3 inventory from spend-based to activity-based estimation. That's a meaningful improvement in data quality, and it happens without asking a single subcontractor to change anything about how they work.

The NGER Boundary Question

There's a second layer to subcontractor emissions that many builders miss, and it's got nothing to do with Scope 3.

Under section 11 of the NGER Act, operational control of a facility determines who reports. If you're the head contractor with operational control of a construction site - you set the environmental management plans, you control site access, you manage the safety systems - then activities happening on that site under your control are your facility's emissions. That includes a subcontractor's excavator burning diesel on your site.

This means some subcontractor emissions aren't Scope 3 at all. They're Scope 1 - your direct emissions from your controlled facility. And they count toward your NGER thresholds. The facility-level threshold is 25 kt CO2-e or 100 TJ of energy. The corporate group threshold is 50 kt CO2-e or 200 TJ. A major infrastructure project running for three years with heavy earthworks can breach those thresholds, particularly when you aggregate subcontractor fuel consumption across the site.

The Clean Energy Regulator's guidance on defining facilities makes this explicit: when estimating emissions and energy data, you must consider all activities that take place at the facility under your control, including activities of contractors and sub-contractors. You can't exclude subcontractor fuel consumption from your NGER facility report just because the sub owns the machine.

We've written about NGER thresholds and facility definitions in more detail elsewhere. The practical takeaway for construction: your subcontractor diesel consumed on your operationally controlled sites is your NGER Scope 1 emission to report. Your subcontractor's business operations off your site - their depot electricity, their workshop gas consumption, their own fleet fuel between jobs - that's your Scope 3 Category 1 under AASB S2. Different reporting frameworks, different scopes, same subcontractor. The boundary matters.

And here's a wrinkle worth flagging: NGER uses AR5 Global Warming Potential values, while AASB S2 requires AR6. The AASB S2025-1 amendment (December 2025) provides jurisdictional relief allowing NGER reporters to use AR5 GWPs for the NGER-covered portions of their AASB S2 disclosure. But for the Scope 3 portions not covered by NGER - which includes most subcontractor off-site emissions - you need AR6. That means maintaining two sets of GWP factors for what is, in many cases, the same subcontractor's activities. It's annoying. But it's the requirement.

Beyond Subcontractors: The Full Scope 3 Picture

Subcontractor emissions are the biggest data gap, but they're not the only Scope 3 challenge for construction. The full Category 1 picture includes purchased materials that you procure directly - concrete, steel, aluminium, timber, glass, insulation, and dozens more. We covered the material diversity problem for Tier 1 builders in detail, including the 50+ material types and unit conversion headaches.

But construction's Scope 3 extends well beyond Category 1.

Category 4 - Upstream transportation and distribution. Every material delivery to your site has a transport emission. The concrete agitator doing 15 km from the batching plant, the steel delivery from 200 km away, the imported aluminium curtain wall from China. These show up on freight invoices and delivery dockets, and they're often overlooked because they're mixed in with material costs.

Category 5 - Waste generated in operations. Construction and demolition waste is a significant emission source. A commercial building generates roughly 20 to 40 kg of waste per square metre during construction. Some of that goes to landfill (methane emissions). Some gets recycled (processing emissions). Your waste contractor's manifests have the tonnes and waste stream classification needed for this calculation.

Category 6 - Business travel. Flights between offices and project sites. The sustainability team travelling to do site audits. Project managers driving to client meetings. Not typically large for construction, but not zero either - especially for national builders running projects in multiple states. We covered the calculation methodology for business travel separately.

Category 7 - Employee commuting. Hard to measure, but getting attention from auditors. A construction company with 2,000 site workers driving to projects in outer suburbs has a meaningfully different commuting profile from an office-based company.

Category 11 - Use of sold products. This one matters for developers. The buildings you design and sell will consume energy for 50+ years. If the building is your product, its operational emissions are your Scope 3 Category 11. The embodied carbon discussion intersects here, because the whole-of-life carbon profile is increasingly what green building ratings and government tenders are asking for.

The point is that Scope 3 for construction is genuinely wide. And subcontractor emissions sit at the centre of it because subcontractors both consume fuel and materials on your behalf and generate waste from your project. Their activities touch Categories 1, 4, and 5 simultaneously.

From Data Gap to Data System

We're not going to pretend the subcontractor emissions problem is solved. It isn't. Not by us, not by anyone. The underlying issue is structural: Australia's construction industry runs on subcontracting, and subcontractors overwhelmingly lack the capability, incentive, or systems to report their emissions.

What we can do - and what we've built Carbonly to do - is extract the maximum possible value from documents that already exist. Subcontractor claims, invoices, delivery dockets, purchase orders, equipment hire records. These aren't new data sources. They're procurement documents your accounts payable team processes every month. The emissions-relevant data inside them - material quantities, fuel volumes, equipment hours, concrete grades, steel tonnages - just needs to be read, extracted, matched to the right emission factors, and tracked with an audit trail that your NGER reporting and AASB S2 assurance process can rely on.

The progression looks like this. Year one: spend-based estimation for most subcontractors, with activity data extracted from the subset of documents that contain quantities. Year two: more targeted procurement, requiring key subcontractors to provide basic activity data (fuel consumption, major material volumes) as a contract condition - something that's becoming common in government projects and green building-rated work. Year three: supplier-specific data from your largest subcontractors, supported by tools that make it easy for them to participate. We wrote about how to collect Scope 3 data from suppliers without destroying the relationship - the same principles apply to subcontractors.

We're honest about what's still hard. Subcontractor invoices that bundle labour and materials into a single lump sum give you almost nothing to work with beyond spend data. Fuel allocation on shared sites - where three subcontractors fill from the same bowser - is a reconciliation problem that no software can fully automate without some process change at the site level. And Scope 3 Category 11 (use of sold products) for developers requires assumptions about building operational performance over decades that we genuinely don't know how to make with high confidence.

But the gap between "we have no idea what our subcontractor emissions are" and "we have activity-based estimates for our top 20 subcontractors and spend-based for the rest, documented with source data" - that gap is closable. And for an ASRS Group 2 builder preparing for their second reporting year, closing it is the difference between a Scope 3 disclosure that an auditor can work with and one that gets flagged.

Start with procurement. Your subcontractors won't fill out a carbon survey. But they'll keep sending you invoices.

Related Reading: