Carbon Reporting Across a 200-Building Property...

A sustainability team at a listed property trust told us they spent 14 weeks collecting utility data for their annual NGER submission. Fourteen weeks. Not because the calculations were hard. Because they had 213 buildings across five states, each with its own electricity retailer, its own gas provider, and its own billing cycle. Some buildings had one meter. Some had eleven. The total was north of 640 utility accounts, and every quarter, roughly 500 bills needed to be opened, read, and entered into a spreadsheet that no auditor could follow.

That's property portfolio carbon reporting in Australia at the ASX200 scale. And it's about to get significantly harder.

ASRS Group 2 reporting kicked in for financial years starting 1 July 2026. That captures every NGER-registered corporation not already in Group 1, plus asset owners with $5 billion or more in assets under management. Most listed REITs crossed that threshold years ago. But NGER only required Scope 1 and Scope 2 for assets under operational control. AASB S2 now demands Scope 3 from the second reporting period, and for a property company, that means tenant energy, embodied carbon in fit-outs, and waste from every building in the portfolio. The data burden roughly triples.

We built Carbonly to deal with exactly this kind of volume. Our team spent 18 years building enterprise data platforms at BHP, Rio Tinto, and Senex Energy before starting this company. The pattern was always the same: smart people wasting months on data collection instead of doing anything useful with the results. Property portfolios are the most extreme version of that problem we've seen.

640 utility accounts and the quarterly data avalanche

The arithmetic is straightforward and the result is brutal. A 200-building portfolio will typically have two to four utility accounts per building: base building electricity, gas (if connected), water, and sometimes a separate common area or car park meter. That's 400 to 800 individual accounts.

Each account generates a bill every month or quarter. At quarterly billing, you're looking at 1,600 to 3,200 bills per year. Monthly billing pushes that toward 5,000 to 10,000 documents annually. And those bills don't arrive in a neat, standardised format. They come from AGL, Origin, EnergyAustralia, Alinta, Momentum, Powershop, local water authorities, and a handful of smaller retailers. Each one formats their PDF differently. Each one puts the consumption figure in a different spot.

Now multiply the extraction problem. For each bill, someone needs to find the total kWh (or MJ for gas, or kilolitres for water), confirm the billing period, match it to the right building and meter, and enter it into the right cell. At five minutes per bill, 3,200 bills per year equals 267 hours. That's nearly seven full work weeks of pure data entry. At $110 per hour loaded cost for a sustainability analyst, you're spending roughly $29,000 a year on typing numbers from PDFs into spreadsheets.

And that's just the first-pass entry. It doesn't count the time spent chasing missing bills, correcting retailer portal downloads that export in the wrong date format, or reconciling the gap when a building's electricity account changes hands mid-lease because the tenant switched providers.

The ANAO found that 72% of 545 NGER reports contained errors, with 17% classified as significant. In a 200-building portfolio, you're aggregating hundreds of small data points into a corporate total. One misread digit on a shopping centre electricity bill, one gas account accidentally coded to the wrong state, one meter that dropped off the tracking sheet when a building was sold mid-year. Each error is small. Together, they can shift your reported Scope 2 by thousands of tonnes.

The state factor problem: same building, four times the emissions

Here's the number that catches most people the first time they see it.

A commercial office consuming 1,000,000 kWh per year in Melbourne generates 780 tonnes of CO2-e from Scope 2 electricity. The exact same building with the exact same consumption in Hobart generates 200 tonnes. Victoria's location-based grid emission factor under the NGA Factors 2025 is 0.78 kg CO2-e/kWh. Tasmania's is 0.20.

That's not a small difference. It's nearly four to one.

For a single-state operator, this doesn't matter much. You use one factor and move on. But an ASX200 REIT with assets in every mainland capital plus Hobart is applying a different emission factor to every building depending on which state it sits in. And the factors vary significantly across the grid:

State/Territory	Location-Based Factor (kg CO2-e/kWh)	Scope 3 Transmission (kg CO2-e/kWh)
NSW & ACT	0.64	0.03
Victoria	0.78	0.09
Queensland	0.67	0.09
South Australia	0.22	0.04
WA (SWIS)	0.50	0.06
Tasmania	0.20	0.03

Source: DCCEEW NGA Factors 2025, Table 1

This creates a strange strategic dynamic. A REIT selling a Victorian asset and buying a South Australian one of the same size and tenancy reduces its portfolio Scope 2 by roughly 72% on that building without changing a single light globe. The carbon isn't gone. The SA grid is just cleaner. But the effect on reported numbers is dramatic, and under AASB S2 paragraph 29(a)(v), location-based Scope 2 is the mandatory disclosure method.

We've talked to sustainability teams that track what they call "factor-adjusted intensity" to separate genuine efficiency gains from portfolio reweighting across states. That's the right instinct. Without it, you can't tell whether your emissions went down because your buildings got more efficient or because you bought a Brisbane asset and sold a Melbourne one.

Carbonly applies state-specific NGA emission factors automatically based on each building's registered address. When a building changes hands or a new acquisition enters the portfolio, it picks up the correct factor without anyone remembering to update a lookup table. For a 200-building portfolio where assets trade every quarter, that eliminates one of the most common sources of Scope 2 calculation error.

Base building versus tenant: the boundary that breaks everything

This is where property carbon accounting diverges from every other industry. The same physical building has at least two different emissions boundaries depending on which framework you're reporting under. Sometimes three.

NABERS Energy (base building) covers the energy consumed by central services: HVAC plant, lifts, common area lighting, fire systems, car park ventilation. It's the rating that matters commercially. Under the Commercial Building Disclosure program, any office building of 1,000 sqm or more must disclose a NABERS rating at point of sale or lease. That rating drives rental premiums, tenant decisions, and increasingly, institutional investor allocation.

NGER uses operational control. For a landlord operating the base building services, the NGER boundary typically captures base building energy. But where the landlord pays for tenant electricity under a gross lease and recovers costs, that energy falls under the landlord's operational control too. The NGER boundary shifts with lease structure, not with physical metering.

AASB S2 then adds the third layer. Scope 1 and 2 follow the reporting entity boundary (broadly, operational control for most property companies). But from the second reporting period, material Scope 3 is mandatory. Tenant energy is Category 13: downstream leased assets. For a commercial office portfolio, tenant consumption is typically 40% to 60% of whole-building energy. That's material by any reasonable definition.

So a single REIT maintaining 200 buildings needs to track, for each building: base building energy (for NABERS), operationally controlled energy (for NGER), and total building energy including tenant consumption (for AASB S2 Scope 3). Three different boundary definitions applied to the same meters.

Getting this wrong has real consequences. If you report tenant energy in your NGER Scope 2 when you don't have operational control over it, you're overstating. If you exclude tenant energy from your AASB S2 Scope 3 because "that's the tenant's problem," you're understating, and your assurance provider will flag it. We covered the boundary mechanics in detail in our property managers guide, but the scale of a 200-building portfolio turns what's a manageable complexity at 20 buildings into a systematic data architecture problem.

The tenant data gap nobody has solved

We need to be honest about this. Tenant energy data is the single hardest data collection problem in property carbon accounting. And at portfolio scale, it's exponentially worse.

When a tenant holds their own electricity account, the landlord doesn't see the bills. The data flows directly between the tenant and their retailer. Unless the lease includes a data-sharing clause, the landlord has no legal right to request it. Pre-2015 leases almost never include such clauses. Even modern green leases with data-sharing provisions rarely include enforcement mechanisms.

For a 200-building REIT, this means hundreds of individual tenants who may or may not respond to data requests. We've heard from portfolio sustainability teams that get response rates of 25% to 35% when they send quarterly data requests. Some tenants send the wrong period. Some send dollar amounts instead of kWh. Some just don't reply.

AASB S2 doesn't give you a pass on this. If Category 13 is material (and for commercial property, it almost certainly is), you need to disclose it. The modified liability provisions under the Corporations Act protect Scope 3 disclosures from private litigation for the first three years, but ASIC can still act if the methodology is unreasonable. "We asked and they didn't reply" is not a methodology.

So you estimate. The GHG Protocol allows it. Floor area proportioning works for many buildings: if you know base building consumption and total building consumption (from the incoming supply meter), the difference is tenant load. Where you only have base building data, NABERS benchmarks by building type and star rating provide a reasonable estimate of whole-building energy intensity. Apply the building's gross lettable area, and you have an estimated tenant figure.

It's not perfect. We're not sure any approach to tenant Scope 3 data at portfolio scale is genuinely accurate yet. The property industry needs standardised data-sharing clauses in every commercial lease, and we're probably five to seven years from that being universal. In the meantime, the best approach is: measure where you can, estimate transparently where you can't, and document your methodology so your auditor can follow the logic.

Refrigerant leakage: the Scope 1 that sustainability teams overlook

Most property carbon accounting discussions focus on electricity. But for a 200-building portfolio, refrigerant leakage from HVAC systems can be a material Scope 1 source that barely anyone tracks properly.

Commercial buildings use centralised chiller systems, split systems, and sometimes a mix of both. The dominant refrigerants in Australian commercial HVAC are R-410A (global warming potential of 2,088 under AR5) and, in older systems still being phased out, R-22 (GWP of 1,810). One kilogram of R-410A leaked equals 2.09 tonnes of CO2-e. That's not trivial.

The DCCEEW's Cold Hard Facts report estimates 6.9 megatonnes of CO2-e in total annual direct refrigerant emissions across Australia's built environment. For commercial HVAC specifically, leak rates have improved from around 20% annually in 2000 to under 5% today. But even at 3% to 4% leakage on a large commercial building with 300 to 500 kg of refrigerant charge, that's 9 to 20 kg lost per year. At R-410A's GWP, that's 19 to 42 tonnes of CO2-e per building per year.

Multiply that across 200 buildings, and portfolio-level refrigerant emissions could sit between 3,800 and 8,400 tonnes CO2-e annually. For context, that's equivalent to the Scope 2 emissions from 10 to 20 medium-sized office buildings in Victoria. It's not a rounding error.

The data collection problem is practical, not technical. Refrigerant top-ups are recorded on maintenance job cards by HVAC contractors. Those job cards are often handwritten, filed in a plant room binder, or entered into a facilities management system that nobody in the sustainability team has access to. For a REIT with 200 buildings maintained by dozens of different contractors, getting consistent refrigerant data requires a deliberate data collection process. You need the refrigerant type, the quantity topped up, and the system involved. And you need it from every contractor, every time.

We won't pretend automated utility bill processing solves this one. It doesn't. Refrigerant tracking requires a different data pathway: structured maintenance reporting from HVAC contractors, ideally feeding into the same emissions platform as your utility data so it all consolidates into a single Scope 1 + 2 + 3 view. Carbonly's Incidents module and manual data entry for refrigerant top-ups handle this, but the upstream data quality depends entirely on how well your FM contracts specify reporting requirements.

For REITs managing the R-22 phase-out (HCFC imports banned from 1 January 2030 under the Ozone Protection and Synthetic Greenhouse Gas Management Act), tracking which buildings still run R-22 systems and planning the transition to lower-GWP alternatives like R-32 (GWP 675) or R-1234ze (GWP 7) is both an emissions reduction opportunity and a capital planning exercise.

NABERS and Green Star: where carbon data becomes asset value

For listed property trusts, carbon data isn't just a compliance exercise. It directly affects asset values.

A NABERS Energy rating of 5 stars or above commands rental premiums of 8% to 12% over comparable unrated buildings, according to multiple property industry surveys. In a portfolio of 200 commercial buildings with an average annual rent roll of $3 million to $5 million per building, improving the portfolio average from 4 stars to 5 stars translates to tens of millions in additional rental income. The data that feeds a NABERS assessment is the same utility consumption data you're collecting for NGER and ASRS.

Green Star certification drives a similar premium. The GBCA reports that Green Star buildings achieve 66% lower carbon emissions, 51% less potable water consumption, and produce 96% less waste going to landfill compared to average buildings. Green Star Buildings v1.1, with registration required from 1 May 2026, requires a minimum 10% reduction in embodied carbon for all rated buildings, with the trajectory targeting 40% by 2030.

For institutional investors, GRESB scores increasingly determine capital allocation to property funds. GRESB uses building-level energy and emissions data as core inputs. A REIT that can't provide granular, verified data at the building level loses points. A REIT that automates data collection and can demonstrate year-on-year improvement in data completeness gains points. The difference in GRESB score between a fund with patchy self-reported data and one with automated, auditable data can shift rankings enough to influence institutional mandate reviews.

This is the part that makes the business case for automation in property carbon accounting. The compliance cost of NGER and ASRS reporting is real but manageable. The revenue impact of NABERS ratings, Green Star, and GRESB scores on a 200-building portfolio is an order of magnitude larger. Both depend on the same underlying data: accurate, building-level utility consumption processed with the correct state-specific emission factors.

What a working system looks like at 200 buildings

Here's what we think this should actually look like. Not the theory. The operational workflow.

Each building in the portfolio gets a dedicated intake channel in Carbonly. That could be a unique email address per project (building23@inbox.carbonly.ai) where facilities managers forward bills directly, or it could be an OneDrive sync folder where retailer portal exports land automatically. Either way, bills arrive and get processed without anyone sorting them manually.

Carbonly's AI Document Engine reads the bill, extracts consumption, billing period, meter reference, and site address, then matches it against the building's expected utility accounts. If Building 47 in Brisbane usually shows 120,000 kWh per quarter and this bill reads 12,000, the Anomaly Detection module flags it before it enters the calculations. Someone reviews it. Was it a genuinely low quarter, or did the retailer misprint a digit? That 10x discrepancy would silently become a 74-tonne Scope 2 error (at Queensland's factor of 0.67) if nobody caught it.

The emission calculation happens automatically with state-specific NGA Factors. Victorian buildings get 0.78. South Australian buildings get 0.22. No lookup tables. No manual factor assignment. And because AASB S2 requires location-based Scope 2 as the primary disclosure but allows market-based as supplementary, Carbonly calculates both. If the REIT has GreenPower contracts or LGC surrenders for specific buildings, the market-based figure reflects those instruments while the location-based figure stays unchanged.

Reporting splits happen at the boundary layer. The same consumption data feeds into NABERS assessments (base building boundary), NGER submissions (operational control boundary), and AASB S2 disclosures (reporting entity plus material Scope 3). Three outputs from one data set. The Scheduled Reports module delivers building-by-building summaries to asset managers monthly, portfolio-level executive summaries to the board quarterly, and full NGER and ASRS data packs to the compliance team annually.

And every extraction, every factor applied, every boundary assignment is logged with a full audit trail. Under ASSA 5010, your assurance provider needs to trace any reported emissions figure back to the source document. For 200 buildings, that's thousands of links between reported numbers and original utility bills. A spreadsheet can't hold that chain together. An audit trail built into the calculation engine can.

What we haven't solved yet

We don't want to oversell this. A 200-building property portfolio has carbon accounting problems that no software fully automates today.

Tenant data gaps remain genuinely unsolved at scale. Until every commercial lease in Australia includes a standardised energy data-sharing clause, Scope 3 Category 13 for property companies will rely partly on estimation. We handle the estimation methodologies and document the assumptions, but the underlying data quality improves building by building, lease renewal by lease renewal. It's slow.

Refrigerant data depends on HVAC contractor reporting discipline. We can receive and process the data, but we can't force a contractor to fill out a structured form instead of scrawling "topped up 3kg R-410A" on a job card. This is an FM contract specification problem, not a software problem.

Mixed-use assets with retail, office, and residential components create boundary headaches that require case-by-case judgment. Is the apartment building in the same trust as the office tower? Who has operational control over the retail podium? These questions don't have one-size-fits-all answers under either NGER or AASB S2.

And NABERS data still needs an accredited assessor. We provide the consumption data in the right format, but the rating process requires independent human expertise. That's by design. The assessor's independence is what makes the rating credible.

The actual next step

If you manage carbon reporting for a property portfolio north of 50 buildings and your process still involves downloading bills from retailer portals, renaming files, and manually entering consumption into a spreadsheet, you're spending six figures a year on data entry and producing numbers your auditor will struggle to verify.

The math on automating this is not close. At 200 buildings, the time savings alone justify the investment before you factor in error reduction, audit trail requirements, and the commercial value of accurate NABERS and GRESB data.

Contact us at hello@carbonly.ai. We price per project, so scaling from 50 buildings to 200 doesn't require a new contract negotiation.

Related Reading: