NGER Report: What Section 19 Actually Demands
A Section 19 NGER report needs facility-level emissions by gas, method, scope and activity, plus energy production and consumption, with every number traceable to a bill for five years. Here is what the report contains, where the errors come from and what software cannot do.
Beach Energy is a useful cautionary example. The Clean Energy Regulator accepted an enforceable undertaking from the company in July 2025 for "inadvertent misstatements" across multiple reporting periods. According to the CER's announcement, the undertaking requires three years of externally commissioned reasonable assurance audits, at the company's cost, plus an external consultant to review its control systems.
The CER described the misstatements as inadvertent. Good intentions do not prevent errors.
That's the problem with NGER reports. They look like a simple annual compliance exercise. Fill in some emissions numbers, hit submit on the EERS portal by 31 October, move on. But the actual data structure that Section 19 of the NGER Act demands is dense, specific, and unforgiving. And if you're assembling it from spreadsheets and PDF invoices, as many reporters still do, you're gambling that your manual process catches every error before the Regulator's data analytics tools do.
We built Carbonly because we'd spent nearly two decades watching this exact failure mode play out inside mining and energy companies. The report isn't the hard part. Assembling the data that feeds it is.
What a Section 19 NGER Report Actually Contains
Most people who've never had to prepare an NGER report think it's a summary. Total Scope 1, total Scope 2, total energy, done. It isn't.
Under Section 19 of the NGER Act, a controlling corporation's annual report must be submitted electronically through the Emissions and Energy Reporting System (EERS) and contain facility-level data across several distinct sections. Here's what the CER actually expects, mapped to the structure their system requires.
Section A: Registration details. Your ABN, the name of the controlling corporation, the registered contact person, and corporate contact information. Sounds simple. But ABN validation alone trips people up - the CER uses the modulo-89 check digit algorithm (the same one the ATO uses) and invalid ABNs are rejected outright.
Section B: Corporate group structure. Every entity in your corporate group, their relationships, and which ones hold operational control over which facilities. If you've done an acquisition or disposal during the reporting year, section 19 requires you to account for the portion of the year each entity held control. This isn't a static org chart. It changes.
Section C: Facility information. Each facility needs its state or territory, ANZSIC code, grid connection region (NEM, SWIS, or DKIS), and activity type. Grid region matters because it determines which state-based emission factors apply to your Scope 2 electricity consumption. Get the mapping wrong and your entire Scope 2 figure is wrong.
Section D: Greenhouse gas emissions. This is where it gets genuinely complicated. You don't just report total CO2-equivalent. The EERS requires a breakdown by individual greenhouse gas - CO2, CH4, N2O, HFCs, PFCs, SF6, and NF3 - each in tonnes. On top of that, you report by estimation method (Method 1 through 4, depending on the emission source), by scope (Scope 1 and Scope 2 separately), and by activity type (fuel combustion, fugitive emissions, industrial processes, waste). That's a multi-dimensional matrix for every facility.
Then there's energy. Production and consumption, broken down according to the fuel and energy commodity classifications in Schedule 1 of the NGER Regulations. Electricity consumed. Natural gas consumed. Diesel. LPG. Each in the correct units.
When you add it all up, a single facility with a few emission sources can require dozens of individual data points, each of which needs to trace back to a source document and a calculation methodology. A corporate group with ten facilities? You're looking at hundreds of data points, all needing to be internally consistent and compliant with the NGA Factors for the correct reporting year.
Where the Errors Actually Come From
The ANAO's performance audit of the NGER scheme found that 72% of 545 reports examined contained errors, with 17% containing significant errors. The most common problems were gaps in own-use electricity data, missing or incorrect emission sources, errors in facility aggregates, problems with energy production figures, and omitted corporate entities or facilities.
That audit examined early NGER reporting years, so you could argue data quality has improved since then. But the Beach Energy enforceable undertaking in 2025 suggests the systemic problem hasn't gone away - it's just that the CER now has better tools to catch it. Their 2025-26 compliance priorities explicitly mention using "advanced data analysis tools" to identify high-risk reporters and targeted audit programs.
So where do the errors actually originate? In spreadsheet-based processes, they tend to come from five places.
Wrong emission factors. Someone used last year's NGA Factors instead of the current year. Or they used the national average electricity factor (0.62 kg CO2-e/kWh in the 2025 edition) instead of the state-based factor. Victoria is 0.78. Tasmania is 0.20. That's a 290% difference. For a facility consuming 5,000 MWh in Victoria, using the national average instead of the state factor understates emissions by 800 tonnes CO2-e. Per facility. Per year.
Wrong units. Natural gas billed in megajoules versus gigajoules. Fuel in litres versus kilolitres. Energy content factors applied to the wrong unit base. These sound like trivial mistakes but they produce order-of-magnitude errors. Enter a gas bill as 45,000 GJ when it was actually 45,000 MJ and you are out by a factor of 1,000. In a spreadsheet, nobody catches that unless they're specifically looking for it.
Wrong facility allocation. A corporate group runs two manufacturing sites in different states. A shared diesel account serves both. The fuel gets attributed 100% to one facility because that's where the invoice is addressed. The other facility reports zero Scope 1 diesel emissions. The corporate total might be right. The facility-level breakdown isn't. And the EERS needs facility-level data.
Transcription errors. Someone reads a utility bill, types a number into a spreadsheet, and gets a digit wrong. It happens constantly with manual data entry. A quarterly electricity bill for 847,293 kWh becomes 874,293 kWh. That's a 27,000 kWh difference - about 17 tonnes CO2-e in NSW. Small enough to not look obviously wrong. Large enough to matter under assurance.
Missing data. A facility's September gas bill didn't arrive before the October deadline. Nobody noticed. The Q4 figure is zero, which drags down the annual total. Or a refrigerant top-up got recorded in the maintenance system but never made it into the emissions spreadsheet. These gaps are invisible until an auditor spots the discontinuity.
Every one of these errors becomes a compliance risk. Section 19 carries a civil penalty of up to 2,000 penalty units for failure to provide a compliant report - that's $728,000 at the $364 penalty unit that applies from 1 July 2026. The CER can also issue infringement notices ranging from 12 penalty units ($4,368) up to one-fifth of the maximum court-imposed penalty. And they publish a list of late reporters publicly. Your company's name on that list is a reputational problem that no amount of internal PR can fix.
What the Report Generator Hands You to Review
Here's what we think an automated NGER report generator should do, and where Carbonly fits today.
The starting point is a readiness check that runs against your data well before September: is your ABN configured, are all facilities set up with their ANZSIC codes and state assignments, does emissions and energy data actually exist for the reporting period, and where does each facility and the corporate group sit against the threshold checks (25,000 tCO2-e or 100 TJ at facility level, 50,000 tCO2-e or 200 TJ at corporate group level). Carbonly computes NGER facility and corporate-group threshold status from the ledger, so you can see that position in July, not October. It does not send alerts as you approach a threshold; someone has to look.
Report generation should then build each section from data already in the system: registration details, group structure, facility details including state, ANZSIC code and grid region, and the Section D breakdown by gas, method, scope and activity type, aggregated facility by facility.
The emission factors should come from the NGA Factors workbook for the correct reporting year, with state-based electricity grid factors. In Carbonly, a Victorian electricity bill is calculated with the VIC grid factor (0.78 kg CO2-e/kWh in NGA 2025), and each record stores the factor version used, the calculation and the link to the original document in the audit trail.
Energy production and consumption should be calculated alongside emissions from the same source documents, mapped to the Schedule 1 classifications.
You then review the output, check anything flagged as uncertain, and lodge it through the CER portal yourself.
Carbonly does not submit directly to the EERS. That's deliberate. The report is your company's legal submission. A human needs to review it before it goes out the door.
The Audit Trail Problem That Nobody Talks About
Here's something that doesn't make it into most NGER compliance conversations: the Clean Energy Regulator requires you to keep records for five years from the end of the reporting year. That means your 2025-26 data needs to be retrievable until at least June 2032.
Not just the final numbers. The source documents. The calculations. The methodology decisions.
When an auditor (or the CER itself) asks "show me the source document for this facility's Q3 Scope 1 diesel figure," you need to produce the original fuel invoice, the extraction of the relevant data points (litres, energy content, billing period), the emission factor applied (69.9 kg CO2-e/GJ from NGA Factors 2025 for diesel), and the final calculation. In the same session. Not "we'll get back to you in a week."
This is where spreadsheet-based processes completely fall apart. The spreadsheet might survive five years. But the 200 utility bills, 15 fuel invoices, 4 gas statements, and 3 refrigerant top-up receipts that fed into it? They're in someone's email. Or on a shared drive that's been migrated twice. Or in a filing cabinet at a site that's been decommissioned.
In Carbonly, every emission record links back to its source document - the original PDF or invoice that our AI document engine processed. You open a record and see the chain: source document, extracted data, matched emission factor, calculated result. That chain doesn't degrade over time. It doesn't depend on Sarah from sustainability still being at the company. It's structural.
We're genuinely not sure any regulator has tested this level of traceability at scale across hundreds of facilities yet. But the direction of travel is clear. The CER's enforcement approach is getting more sophisticated, not less. Beach Energy was an early warning. Building the evidence chain now is cheaper than rebuilding it under an enforceable undertaking later.
What Software Can't Do (An Honest List)
We'd be greenwashing our own product if we claimed software solves everything about NGER reporting. It doesn't.
Higher-tier methods need measurement data. If your facility uses Method 2, 3, or 4 for certain emission sources - which larger facilities often must - those methods require specific measurement equipment data. Gas chromatography results. Stack monitoring data. Continuous emission monitoring systems. Software can ingest and calculate from this data, but it can't generate it. You still need the instrumentation.
Fugitive emissions often require manual input. Refrigerant leaks, coal seam gas emissions, equipment venting - these sources don't come with neat invoices. Someone has to record the data. A refrigerant top-up log. A gas drainage measurement. Software can't observe that a technician added 15 kg of R-410A to an air conditioning unit last Tuesday. But once that data is entered, it can apply the correct GWP (1,924 for R-410A under AR5) and slot it into the right facility, the right scope, the right gas category.
Custom industrial processes may need manual factor entry. If you're running a cement kiln or an aluminium smelter with process-specific emission factors, those factors might not be in any standard library. Software can hold a custom factor, but someone with process engineering knowledge needs to determine the right one.
The system doesn't submit to the CER. You prepare and lodge the report yourself through EERS Online Services. This is a conscious choice. NGER submissions carry legal weight - a civil penalty provision sits behind them. A human should be reviewing the final output before it's submitted under their name.
The AR5/AR6 GWP mismatch is still manual work. NGER uses AR5 Global Warming Potential values. AASB S2 (the new mandatory climate reporting standard) requires AR6. If you're an NGER reporter who's also been pulled into ASRS Group 2, you'll need to reconcile the two. Carbonly calculates and stores on AR5 only, which is what the NGER submission needs and what AASB S2025-1 lets your NGER-covered portion carry into the AASB S2 disclosure. Restating anything outside that relief on AR6 happens outside the platform, and either way you need to understand which number goes where.
The 31 October Reality Check
Every year, the same thing happens. Companies start thinking about their NGER report in August. They realise they're missing three months of utility data from one facility. They find out the contractor who managed a remote site changed fuel suppliers and nobody updated the records. They discover that a facility that was supposed to be aggregated actually exceeds the 25 kt threshold individually and needs its own facility-level report.
Then it's September. Thirty days to go. The sustainability manager is working weekends. The finance team is annoyed because they need the same data for the annual report. Everyone wishes they'd started earlier.
Don't plan on an extension to the NGER deadline. The statutory date is 31 October every year; see the CER's compliance update for its current position. When 31 October falls on a weekend the deadline moves to the next business day, so 2025-26 reports are due by midnight AEDT on Monday 2 November 2026. Penalties apply for late submission. Your name goes on the late reporter list. Our NGER October 2026 deadline guide covers the September sign-off and the October submission steps.
The 2025-26 reporting year brings additional complexity. The legislative amendments introduce market-based reporting for biomethane and hydrogen consumption, updated flared gas emission factors, and a requirement for consistent market-based method application across all facilities in a corporate group. If you weren't tracking location-based and market-based Scope 2 separately before, you'll need to start, and companies using spreadsheets may need to restructure their data collection.
And here's the part that should really focus your attention: if you're a registered NGER reporter that lodges financial reports under Chapter 2M, you're in scope for ASRS mandatory climate reporting (Group 2, or Group 1 above the publication threshold) from financial years starting 1 July 2026 at the latest. Your NGER data becomes the foundation for your climate-related financial disclosures. Getting it right isn't just about NGER compliance anymore - it's about not having to do the work twice.
Start With the Data, Not the Report
The report is the output. The data is the problem.
If your NGER report for the 2025-26 year is due by 2 November 2026 (31 October falls on a Saturday) and you haven't automated data extraction from utility bills, you're already behind. Every month that passes without a systematic approach to capturing electricity consumption, gas usage, fuel volumes, and refrigerant top-ups is another month of data you'll be scrambling to reconstruct later.
Feed the documents in now. Let the system build the audit trail as you go. When October comes, you spend your time reviewing the numbers - not building them.
That's what we think NGER reporting software should actually do. Not give you a template to fill in. Give you a report that's already filled in, backed by evidence, with every number traceable to a source document.
The CER is watching more closely than ever. Make sure the numbers are right. If you want to see your Section 19 data built from the source documents, email hello@carbonly.ai.
Related reading: