Methodology

Emissions are calculated to the GHG Protocol Corporate Standard. Reports are prepared with reference to the GRI Standards: GRI 1: Foundation 2021, GRI 2: General Disclosures 2021 and GRI 3: Material Topics 2021, plus a topic standard for each topic the business decides is material. The topic standards carry their own years — GRI 305: Emissions 2016, GRI 302: Energy 2016, GRI 303: Water and Effluents 2018, GRI 306: Waste 2020 — and the report lists the year of every standard it uses. GRI 305 does not set out how to calculate emissions; it points to the GHG Protocol, so the two fit together rather than compete.

Greenhouse gases other than carbon dioxide are converted using the IPCC AR5 100-year global warming potentials, and each factor records the basis it uses. Scope 2 is location-based: it uses the emissions of the grid the site is connected to, not any contract for cleaner power. Quantities and emissions are stored as exact decimals and rounded only when they are shown.

The factor registry holds 58 factors. 45 of them have been read off the publisher’s own document by a person and match it. The rest are listed by name further down, with what is wrong with each. Country-specific coverage: Malaysia, Singapore. Each factor records what it converts, the country and region it applies to, the year it was published for, who published it, where to find it, how good a match it is, and whether a person has checked it. Every factor is listed, one by one, at /factors.

The registry also holds factors for Thailand, which is not offered as a choice yet. Loading a country’s factors is the easy half of supporting it. The other half is working out how its regions differ, deciding which activities must never borrow another country’s figure, and having somebody check each number against the publication it came from. Until that is done, the country does not appear in the list.

Counted from the factors themselves each time this page is built, so the table cannot fall behind what the calculation uses.

The quantity is first put into the unit the factor expects. If the two units cannot be reconciled, the calculation stops rather than guessing.

Then a factor is chosen, by looking for the closest match to the activity, the country, the region and the year. If there is no factor for that exact year, the nearest earlier year is used; if there is nothing for the region, the national figure is used; if there is nothing national, a worldwide default is used. How close the match was is recorded on the entry, and using a factor from a different year to the one being reported puts a warning on the report.

The answer is then stored together with a copy of the factor used — its value, its quality and how it was chosen. This is what keeps an old figure checkable. If a publisher revises a number next year, a report issued before the revision still adds up, because the entry holds the figure that was actually used rather than a pointer to today’s. Stored figures are never quietly rewritten. Recalculating is something somebody has to ask for, and it is logged.

For electricity, waste and water, the figure depends on how the country does it, not just on how much is used. Singapore burns its rubbish. Malaysia buries it, which gives off methane for years afterwards, and how much escapes depends on whether the site captures the gas. Using one country’s number for the other is not a rough answer, it is the wrong answer. So for these activities there is no borrowing from another country and no worldwide average:

• Electricity from the grid
• Food waste
• Gas leaked upstream of the power station
• General waste
• Mains water
• Plastic waste
• Waste sent for recycling

Where there is no factor for the country in question, no figure is produced, and the report records the gap along with the reason for it. This carries through to the questions asked in the first place: a Malaysian business is not asked about waste on the quick footprint, and the field does not appear on the collection form either.

Chilled water bought from a district cooling plant is the one thing this service asks about and then does not turn into a figure. It asks because leaving the question out would make a business with no district cooling look identical to one whose district cooling was ignored. The quantity entered is printed on the report, so a reader can see the amount that was set aside and why.

No Singapore or Malaysian authority publishes an emission factor for purchased cooling. SEFR, the national registry, covers grid electricity; it does not cover cooling bought from a district plant.

The GHG Protocol Scope 2 Guidance does explain how to work one out: take the electricity factor and divide it by how efficiently the chiller plant runs. The method is sound. It is the second number we do not have — efficiency belongs to one particular plant, and we do not know which plant chills which building. A guessed figure could be out by more than the whole total it was added to, which is worse than saying plainly that it is missing.

Where a cooling supplier states its own emission factor, this stops being a gap. A business on a district cooling contract is entitled to ask for one, and the figure can then be entered as the supplier’s number rather than our estimate.

Every report states that it is prepared with reference to the GRI Standards. It never states that it is prepared in accordance with them, and there is no setting that changes this.

To report in accordance, an organisation has to answer all thirty of the GRI 2 general disclosures. For any it leaves out, it has to give one of the four reasons GRI allows and explain why that reason applies. This service asks for eleven of the thirty and has nowhere to record those explanations. So an in-accordance claim from here would not be one GRI recognises.

Reporting with reference has three requirements, set out in GRI 1: Foundation 2021. First, the report has to carry a GRI content index: which version of GRI 1 was used, the full title and year of every standard the answers come from, each answer with its number and title, and which part of the report to look in. Second, the statement of use goes on the front, word for word as GRI writes it. The third is for the organisation to do itself: once the report is published, GRI must be told that the Standards were used. That is done through GRI’s Report Registration System, and the report page says so at the point the document is produced.

The report screen also shows what is still blank. That is a list of what a fuller report would want, not a pass or a fail, and it does not change what may be claimed. Two things do stop a report being produced at all. One is an emission factor we do not stand behind, listed below. The other is an account ZYT has not activated. The factor check is run again on our servers when the document is generated, so it cannot be got round from the browser.

A blank is reported as a blank, with the reason for it. No figure is ever estimated to make a total look complete — GRI 305-5 stays empty in a first report because there is no base year to compare against, and that is the correct outcome rather than a gap to fill.

All of it, in one place. A limitation found in review that we did not disclose here would be good reason to doubt everything else on this page.

• No assurance. The figures are the business’s own and nobody has audited them. Getting them audited is a separate job for a qualified provider. Most GRI reports are self-declared like this, so it is a normal position — but it is not assurance, and this service never suggests otherwise.
• Location-based Scope 2 only. Electricity is counted using the emissions of the local grid. A green tariff, a power purchase agreement or a holding of renewable energy certificates does not show up in the figure. Where contracts like those can be bought — and they can in Singapore — the GHG Protocol expects a second set of figures reflecting them, and this service does not produce one.
• No base-year comparison. There is no year-on-year comparison yet, so the disclosure about reducing emissions (GRI 305-5) is blank in every report, with the reason printed next to it.
• GRI is not notified by this service. Notifying GRI is one of the three requirements for reporting with reference to the GRI Standards, and it is the one nobody else can do for the organisation. It is done after the report is published, through GRI's Report Registration System, or by writing to reportregistration@globalreporting.org. Until it is sent, the claim printed on the report is not complete. The report page says so at the point the document is produced.
• No GRI Sector Standards. GRI publishes extra standards for some industries — oil and gas, coal, agriculture — and expects them to be used where one exists. None is applied here, so the topics a report treats as important are decided without one.
• Energy is converted at published averages, not measured. GRI 302-1 asks for all the energy consumed by the organisation. Metered electricity and gas are counted, and so is the energy in the fuels burned — petrol, diesel, LPG, fuel oil and charcoal — converted at the net calorific values published in the IPCC’s 2006 Guidelines, using the fuel densities published by the UK government in its annual conversion factors. Both are averages for the fuel, not measurements of the particular batch delivered to a particular site, and a real consignment varies around them by a few per cent. That is the normal basis for a self-declared inventory and it is stated here rather than implied.
• No emissions breakdown by gas. GRI 305-1 asks which gases are counted. The factors used are combined figures covering carbon dioxide, methane and nitrous oxide, plus whichever refrigerant was reported, and the publishers do not break them apart — so neither can we, and the report cannot show emissions gas by gas. One place this bites: Singapore’s electricity authority publishes its grid figure as carbon dioxide only, and it is stored here alongside the combined ones. On a gas-fired grid the difference is small, but it is a known imprecision rather than an exact match. Carbon dioxide from burning plant material is also not calculated. Charcoal is collected and its factor deliberately leaves that part out, so where charcoal is reported it is missing from the totals and is not shown separately either, which the GHG Protocol would normally want.
• Inputs are not verified. We cannot see the meters, the bills or the sites. Where a figure was entered wrongly, the result is wrong, however carefully it was calculated.
• Some Scope 3 lines use spend-based proxies. Some purchases are converted from money spent rather than from anything physical. That means a price rise raises the calculated emissions while nothing has actually changed. These are labelled as estimates wherever they appear, and they are a poor thing to base a reduction claim on.

13 of the 58 factors are not good enough to put in a report, for two different reasons. All fourteen are still collected: the fields exist, and what a business enters against them appears in its own totals, so it can see roughly where it stands. An inventory that uses any of them cannot produce a report at all. The request is refused, and the refusal names the factor that stopped it.

They could have been hidden instead. A figure a business cannot see is a figure it cannot plan against, and the honest form of a rough number is a rough number labelled as one.

Averages, and there is nothing to check them against. These convert money spent, or a count of devices, into emissions using an industry average. No authority publishes the figure we would be checking, because there is no single right answer — two laptops of the same price can differ by a factor of three. They will not become verified by someone looking harder. They improve when a supplier states its own figure, which is then used instead.

• Laptops, per device — all countries
• Desktop computers, per device — all countries
• Servers, per device — all countries
• Phones, per device — all countries
• Tablets, per device — all countries
• Monitors, per device — all countries
• Printers and multi-function devices, per device — all countries
• Paper and cardboard packaging, by spend — Singapore
• Plastic packaging, by spend — Singapore
• Mixed packaging, by spend — Singapore
• Cleaning and facility services, by spend — Singapore
• Professional and business services, by spend — Singapore
• General purchased goods, by spend — Singapore

In September 2026 we checked our Malaysian electricity factors against the Energy Commission’s published Grid Emission Factor and found both of them wrong: Peninsular Malaysia 23% too low, Sarawak 31% too high. They were replaced with the published numbers — three regions across 2022, 2023 and 2024, nine factors, each one read off the publication.

No report had gone out with them, because a factor nobody has checked stops a report being produced at all. That is the safeguard doing its job. It is written up here rather than fixed quietly.

Everything above comes from these four. They are linked so that our reading of them can be checked against the documents themselves.

• GRI 1: Foundation 2021 · Global Reporting Initiative
• GHG Protocol Corporate Accounting and Reporting Standard · Greenhouse Gas Protocol
• GHG Protocol Scope 2 Guidance · Greenhouse Gas Protocol
• IPCC AR5 Working Group I, Chapter 8 — global warming potentials (Table 8.A.1) · Intergovernmental Panel on Climate Change

The emission factors have their own publishers, one per factor, listed at /factors.

The answers section covers the underlying concepts — what an emission factor is, why Singapore and Malaysia are not interchangeable, and what a customer or a regulator will actually accept.