An emissions figure should be reconciled like revenue

For most of the past fifteen years an Australian energy company's emissions figures have lived in two worlds. Facility emissions from power stations were measured and reported to government under Australia's National Greenhouse and Energy Reporting scheme, a regime with defined methods and real penalties. Everything else, including the far larger emissions from the gas and electricity that retailers sell, appeared in voluntary sustainability reports, prepared by capable people in spreadsheets, with methods that shifted as the voluntary frameworks did.

That arrangement has ended. Australia's mandatory climate-related financial disclosures took effect for the largest companies for financial years beginning in 2025. For Wirrandel, whose year ends in June, the first climate statement covers the year to June 2026. It sits in the annual report, the directors are responsible for it as they are for the financial statements, and the assurance our auditor provides over it will widen year by year. My argument is simple. An emissions figure that directors sign should be produced under the same controls as the revenue figure they sign beside it, and for a retailer the two should be built from the same data.

Two ledgers that should agree

A gentailer bills its customers for every kilowatt-hour of electricity and every megajoule of gas it sells. Those volumes are reconciled to meter data, to market settlement and to cash every month, under controls that internal and external auditors test. The emissions from the gas our customers burn are, to a close approximation, those volumes multiplied by an emission factor. The emissions behind the electricity we buy from the market and resell are, likewise, purchased volumes multiplied by factors for the grid in each state.

In other words, the largest emissions figures a retailer reports are arithmetic on numbers it already controls. Yet when we began preparing for the mandatory regime in 2024, our sustainability spreadsheets did not take their volumes from billing. They took them from a quarterly extract that predated several system changes, and then applied adjustments that only two people fully understood.

We built the emissions ledger on our enterprise data platform instead. Customer emissions are calculated account by account from billed volumes, using emission factors chosen and documented by our sustainability team. Each month the ledger reconciles to billing, and any difference above 0.2 per cent must be explained before the month closes. Changes to factors or methods pass through the same change control as a revenue rule, and the outputs are signed off by the same chain of people.

Building it was humbling. Reconciliation showed that customers who switched retailers part-way through a billing cycle, and accounts carrying both estimated and actual reads, had been counted twice under the spreadsheet method. Our 2023–24 voluntary report had overstated customer gas emissions by 3.8 per cent. The error was honest and modest, and it ran in the conservative direction. We corrected it publicly all the same, because a company that asks readers to trust its first mandatory statement should not leave a known error in its last voluntary one.

Most of our emissions are in our customers' kitchens

Wirrandel's own gas peaking stations run for a few hundred hours a year and produce a small share of our footprint. The largest single category, about 3.2 million tonnes a year, is the gas our customers burn in their heaters, stoves and hot-water systems, followed by the generation behind the electricity we purchase and resell. In the greenhouse gas accounting framework on which Australia's climate standard draws, these are scope 3 emissions: real and reportable, but produced outside our own operations.

There is an argument that scope 3 figures are inherently estimates and do not merit revenue-grade controls. For many companies that is true; a manufacturer estimating the emissions of its suppliers' suppliers is working from averages. A retailer is not. We know from the bill how much gas each customer used. The uncertainty lies in the emission factor, not in the volume, and a board is entitled to expect that the part we can measure is measured.

There is a strategic reason to care as well. A retailer that wants to help households move from gas to electric appliances needs to know, home by home, where the gas goes. The same ledger that supports our climate statement now shows which suburbs burn the most gas for space heating, and that is where our heat-pump offers began. A spend-based estimate would have told us nothing of the kind.

The approach had costs, and the colleagues who argued against it were right to raise them. Our head of sustainability preferred specialist carbon software that her team knew and could have running in four months; we took eleven, and her team handed control of the volumes to mine. Our CFO pointed out that the transitional arrangements allowed some disclosures to be phased in, and asked why we would spend A$4.3 million doing more than the first year required. My answer was that we would need these controls within three years regardless, and that building them later, under deadline, would cost more.

What a board should ask before it signs

Directors are now accountable for emissions numbers in a way they were not before. They need not become carbon accountants, but they can ask questions that test whether the figures are under control.

Ask where the volumes come from. If the answer is anything other than the systems that produce revenue, ask why.

Ask whether the emissions ledger reconciles to billing each month, and at what tolerance. A figure reconciled once a year is reconciled too late to correct.

Ask who owns which part. In our arrangement the sustainability team owns methods and factors, and my team owns volumes and controls. That division was uncomfortable, and it took a facilitated workshop and two attempts at the process maps to settle, but it means every number has an owner who can explain it.

And ask what the assurance provider has said about lineage, not only about the result. Assurance will deepen over the coming years, moving from limited towards reasonable. A board that hears about lineage problems in the first year, while they are cheap to fix, is far better placed than one that discovers them under the stricter standard.

I do not think emissions reporting should become an exercise in false precision. Emission factors carry genuine uncertainty, and our statement says so. But the volumes beneath them are among the most carefully measured numbers in our business, and we should report them that way.