Scope 3 in Year Two of AASB S2: What First-Wave Reporters Deferred, and What Your Assurer Will Test
Most first-wave AASB S2 reporters deferred Scope 3. In year two that option is gone, and the hard part is evidence: dockets that say N32, concrete ordered in cubic metres against per-tonne factors, subcontracts that count the same material twice. Here is what an assurer will sample, and what to fix first.
RSM looked at roughly 30 to 40 of the first AASB S2 disclosures and found that only around one-third included Scope 3 emissions. The rest deferred, citing data maturity and assurance readiness. That was allowed, and it's what the first-year relief is for.
There is no relief in year two. AASB S2 Scope 3 reporting in year two is where the deferral runs out, and plenty of Australian sustainability teams are about to find out how much work they put off. Multiplying a quantity by a factor is easy. Finding a defensible quantity, in the right unit, counted once, with a document behind it that an assurer can pull, is where the time goes.
So our view is that the Scope 3 bottleneck is evidence, and most of that evidence already sits in documents you hold. It isn't waiting in the supplier survey you sent in March.
For the basics of which categories apply and when, see Scope 3 reporting is mandatory. This post picks up after that: what your assurance provider will do with the number.
What the first wave put off
Section 296D of the Corporations Act requires Scope 3 in the climate statements, with a carve-out for the first financial year an entity has to prepare a sustainability report. AASB S2 carries matching transition relief in Appendix C. From the second year, Scope 3 is in.
The timing depends on your year end. Group 1 started with periods beginning on or after 1 January 2025. A June-balancing Group 1 entity did its first year in FY26 and is inside its Scope 3 year now, in FY27. A December balancer is in calendar 2026 and will lodge its first Scope 3 numbers in early 2027. Group 2 started on 1 July 2026, so for a June balancer its Scope 3 year is FY28, running 1 July 2027 to 30 June 2028. The data that report draws on starts accumulating in nine months.
PwC's AASB S2 unpacked analysis says much the same. Year one leaned heavily on the Scope 3 transitional relief. Year two needs Scope 3 built assurance-ready from the outset, with documented methodologies, formalised controls and a transparent basis for what was included, excluded and estimated, or restatements follow.
ASIC's early observations on sustainability reporting, published 6 May 2026, didn't single out Scope 3. But one line lands hardest there: disclosure of "judgements, assumptions and areas of measurement uncertainty" should be "clear, effective and proximate". Scope 3 is mostly judgements and assumptions. If year one tucked them into an appendix, year two can't.
The rules aren't about to soften either. Treasury's consultation on improving the efficiency of climate-related financial disclosures closed on 2 October 2026 and explicitly wasn't seeking views on changing Scope 3 reporting requirements. Its options on assurance and value-chain guidance are proposals, not law. Plan for FY27 and FY28 as the standard reads today (our Treasury consultation explainer has the detail).
Why waiting for suppliers is the wrong plan
The default year-two plan at most organisations is a supplier survey. Questionnaire to the top 50 suppliers by spend, then wait, chase and collate.
Sphera's 2025 Scope 3 report, a survey of 315 sustainability professionals, found 79% name supplier data availability as a top challenge for accurate Scope 3. And 62% of those already reporting Scope 3 cite internal data quality as a major barrier. For an Australian Group 2 entity, a survey-first plan bets the FY28 disclosure on the step most reporters say they can't control.
Treasury's consultation paper went the same way. It rejected a standardised supplier questionnaire because it would encourage data requests "when they are not necessary", noting that AASB S2 already permits secondary estimates. Supplier engagement still matters for your largest categories, and we've covered running it properly in supplier engagement for Scope 3. We just wouldn't put it on the critical path.
A construction business already holds plenty for Categories 1, 4 and 5 without asking anyone: concrete delivery dockets, reo schedules, subcontractor progress claims, freight consignment notes, waste transfer notes, and the EPDs suppliers attached to their tenders. Property, manufacturing and logistics have their equivalents. That paperwork carries physical quantities (tonnes, cubic metres, lineal metres, kilometres), which is activity data nobody has extracted yet.
Fuel dockets are the trivial end of that pile. One fuel, one unit, one NGA factor, often a supplier CSV, and it's Scope 1 anyway. The documents that cause trouble are the untidy ones.
Where year-two Scope 3 breaks
We've read a lot of construction supply documents while building extraction for them. The failure modes repeat, and almost none of them are arithmetic.
Start with the docket that says N32. A concrete delivery docket (formally the identification certificate under AS 1379) gives you grade, slump, aggregate size and volume. N32 means normal-class concrete with a 32 MPa characteristic strength at 28 days. That's a strength specification and says nothing about carbon, which depends on cement content, how much fly ash or slag replaces it, the plant and the product line. A registered Boral Tasmania EPD puts its Hobart normal-class 32 MPa mix at 340 kg CO2-e per cubic metre (A1 to A3). A Holcim project-specific N32 mix in Sydney is declared at 181. Different state, plant and mix design, so it isn't like for like. Still, the same three characters on a docket can sit on footprints almost twice apart.
Then unit chaos. Concrete is ordered and invoiced in cubic metres, while many factors, especially generic ones, are per tonne. Bridge the two with a density (the Holcim EPD declares 2,383 kg per cubic metre) and 400 cubic metres becomes roughly 950 tonnes. Apply a per-tonne factor straight to the cubic metres and that line is understated about 2.4 times. Reinforcing steel has the mirror problem: schedules quote lineal metres by bar size and the factor wants mass. An N16 bar runs at roughly 1.58 kg per metre, so 4,800 lineal metres is about 7.6 tonnes. Any line where the document's unit doesn't match the factor's unit is a likely restatement.
Double counting is the third. Supply-and-install subcontracts are the classic trap: the concrete arrives as a docket recorded under Category 1, then again inside the concreter's progress claim, run through a spend-based services factor. NGER reporters face a second version. The CER's contracts and leasing guideline sets out that a subcontractor's activities at a facility under your operational control can be your Scope 1, so the subbie's diesel can't sit in your Scope 3 as well. Joint ventures add a third layer, which we've written up in Scope 3 double counting in construction.
The fourth should worry a CFO most, because spend-based accounting can't see your best decision. Suppose a project switches 1,000 cubic metres of 32 MPa concrete from a standard mix to a lower-carbon blend at the same contract rate. Using the Boral Hobart figures from that EPD:
| Method | Standard 32 MPa mix | Lower-carbon 32 MPa mix (Envisia) | Reported change |
|---|---|---|---|
| Activity-based (EPD, A1 to A3) | 340 t CO2-e | 300 t CO2-e | 40 t lower (about 12%) |
| Spend-based (same dollars, same EEIO factor) | Same figure | Same figure | Nil |
The scenario is hypothetical (we've assumed equal pricing to isolate the effect). The mechanism is real. Spend-based accounting maps dollars to a sector average, so if the dollars don't move, the reported emissions don't either. Procurement makes the better choice and the disclosure shows nothing. More on that in spend-based emissions accounting.
Last is the sampled line. Say an assurer picks a Category 1 record of 1,180 t CO2-e against "Concrete supply, Stage 2" and asks for the source. You need the invoice, the dockets behind it, the quantity, any unit conversion, the factor, which EPD and which version, and whether that EPD was valid for the period. The validity check catches people out. The Boral Tasmania EPD above was valid until 1 May 2026, so a factor that was fine in a year-one model can expire before the year-two period ends.
What your assurer will test under ASSA 5010
The phasing is less uniform than people assume. Under paragraph 9 of ASSA 5010, an entity's Year 1 is any annual period commencing in its group's first-year window, which for Group 1 runs from 1 January 2025 to 30 June 2026. Year 1 limited assurance covers governance, strategy, Scope 1 and Scope 2. From Year 2, the limited assurance review extends to all disclosures, Scope 3 included. Reasonable assurance over all disclosures applies for periods starting on or after 1 July 2030 under the Corporations Act, unless Treasury's proposals are legislated.
We're not certain how one edge case plays out, so ask your auditor rather than relying on our reading. A December-balancing Group 1 entity has two periods commencing inside that window, calendar 2025 and calendar 2026. On a plain reading, its first Scope 3 year may still fall within Year 1 assurance scope, leaving Scope 3 outside the mandatory review for that one report. June balancers don't get that gap. We'd build as if it will be reviewed regardless, because the following year it will be.
Limited assurance is inquiry and analytical procedures with some testing, less than a full audit. The questions are predictable.
Methodology comes first. AASB S2 paragraph 29(a)(iii) requires you to disclose your measurement approach, inputs and assumptions, and why you chose them, and 29(a)(vi) requires you to say which Scope 3 categories are included. The assurer will check that the written methodology matches what the spreadsheet or system does. The AASB's GHG educational material from August 2025 also says you can't limit Scope 3 to the GHG Protocol's minimum boundaries, so a thin category screen will draw questions.
Next is the data hierarchy. Paragraph B40 lists four characteristics to prioritise when choosing Scope 3 inputs: direct measurement, data from specific activities in your value chain, timely data that reflects the jurisdiction and technology, and verified data. The standard doesn't demand primary data everywhere. It asks you to prefer better inputs where they're reasonably available and to disclose how far you used them. If you used spend-based for a category where invoices carry physical quantities, expect to explain why.
Judgements and uncertainty come next, which is ASIC's point again: which factor, which allocation, what you assumed about missing data, and which GWP set. NGER works on AR5 global warming potentials. AASB S2 measures on AR6, and the AASB S2025-1 jurisdictional relief that lets NGER-covered emissions stay on AR5 is built around what NGER covers, which isn't your Scope 3. EPDs and spend-based tables carry their own GWP basis, so the methodology note should record what sits under each factor.
Then consistency and traceability. Year two is the first time anyone can compare against a prior period, so a methodology change between years needs to be identified, explained and, if material, restated. Every sampled number also needs a chain back to a document. If that chain runs through one analyst's memory, expect a finding.
A year-two checklist that isn't a supplier survey
Start with a fresh category screen. Run all 15 categories against your business model, as paragraph B32 asks, and write down why each excluded one is out. Year one's screen was probably rushed and may have leaned on a deferral you no longer have.
Then set a materiality view per category, ranked by estimated tonnes rather than spend. Accept spend-based as a reasonable estimate for the long tail. Office supplies and software licences don't need activity data.
Pick two or three categories to move from spend to activity this year. For construction that's usually Category 1 concrete and steel, then Category 4 freight or Category 5 waste. Choose the ones where quantities already appear on documents you receive. The effort pays back fastest there, and those are the categories where spend-based can't reflect procurement decisions.
Fix evidence retention before you fix factors. Every Scope 3 record should link to its source document, with the factor, its version and any conversion recorded alongside. NGER reporters already keep records for five years from the end of the reporting year, and holding Scope 3 evidence to the same standard costs little extra.
Finally, write a restatement policy before you need one. Decide your significance threshold (5% is widely used, though the GHG Protocol leaves the level to you) and what triggers a recalculation: a methodology change, a factor correction, a boundary change, a discovered duplicate. We've drafted the wording in how to write an emissions restatement note.
If a consultant is leading your engagement, much of this is a call they'll make with you, and this is roughly the order they'll want it in. Handing them a clean evidence base frees their time for judgement work instead of data entry.
We don't think anyone has fully solved Scope 3 for large, fragmented supply chains. A head contractor with several hundred active subcontracts will still have categories where the best available input is spend-based, and nominated-supplier arrangements still blur who owns the material. For year two, aim for a defensible method, applied consistently, with the judgements written down.
Where Carbonly fits
We built Carbonly around the documents, because that's where the year-two problem sits. The document engine reads supplier invoices, delivery dockets, consignment notes and other procurement paperwork across eight file formats and extracts physical quantities first. It keeps each line's own wording, including any spray or spread rate, as evidence next to the extracted quantity. Extracted lines go through five-tier material matching against NGA factors or a construction EPD factor library, with provenance flagged on each factor.
Lines with low match confidence, or with a unit the system can't convert against the matched factor, are flagged for review. We designed it that way on purpose, because the cubic-metre-to-tonne call is a judgement a person should confirm. Spend-based records are marked as lower data quality, so the disclosure reflects the input hierarchy. Every emission record keeps its source document, the Auditor Workspace exports an evidence pack for an ASSA 5010 review, and period locking makes a correction land as a restatement with a reason attached.
The software won't decide your boundary, set your materiality threshold or sign off your methodology. It supports the disclosure, and your preparer, board and assurance provider remain responsible for it.
This week, pull the biggest Category 1 line from your year-one or draft year-two model and try to walk it back to a physical document in under ten minutes: quantity, unit, conversion, factor, factor version. If you can't, that's probably the line your assurer finds first, and it's where to start. To see that walk-back on your own documents, book a 30-minute call.
FAQ
Is Scope 3 mandatory in the second year of AASB S2 reporting?
Yes. The Corporations Act (s296D) and the transition relief in AASB S2 Appendix C only excuse Scope 3 in the first financial year an entity prepares a sustainability report. From the second year it must be disclosed, along with the categories included and the measurement approach.
When does Group 2 have to report Scope 3?
Group 2's first reporting year began on 1 July 2026, so for a June balancer Scope 3 is required in FY28, the year running 1 July 2027 to 30 June 2028. Data for that year starts accumulating from July 2027, so evidence capture needs to be working before then.
Will my Scope 3 number be assured?
Under ASSA 5010, limited assurance extends to all disclosures, including Scope 3, from an entity's Year 2. Reasonable assurance over all disclosures applies for periods starting on or after 1 July 2030, unless legislation changes after Treasury's 2026 consultation. Some December-balancing Group 1 entities may have their first Scope 3 year fall inside the Year 1 window, so confirm the scope with your auditor.
Can I keep using spend-based estimates in year two?
Yes. AASB S2 permits secondary estimates, and spend-based is reasonable for low-impact categories. Paragraph B40 still asks you to prioritise better inputs where reasonably available, and you must disclose the extent you used them. Spend-based also can't show reductions from buying lower-carbon materials at the same price.
Does modified liability protect our Scope 3 disclosures?
Partly, for a transitional period. The modified liability regime restricts private action over Scope 3 statements for three years from the start of the regime on 1 January 2025, while ASIC keeps its powers. ASIC RG 280 has the detail. Treat it as a buffer for estimation uncertainty and keep documenting. How far it reaches into Group 2's first Scope 3 year depends on the statutory drafting, so get legal advice rather than assume cover.
Which GWP values apply to Scope 3?
NGER uses AR5. AASB S2 measures emissions on AR6, and the AASB S2025-1 jurisdictional relief is tied to emissions covered by NGER, which doesn't include Scope 3. Your methodology note should record the GWP basis under each Scope 3 factor, including EPDs and spend-based tables.
Related Reading:
- Concrete: The Hardest Line Item in Australian Scope 3 - Choosing a defensible factor when the docket only gives you a grade
- ASSA 5010 in Practice: The 8 Evidence Tests Your Auditor Will Run - What to test before your assurance provider does
- How to Collect Scope 3 Data from Your Suppliers (Without Losing Them) - For the categories where you do need supplier input
- Browse the Scope 3 hub and the AASB S2 hub