Fluorinated Gas Inventory Emissions — Methodology and Calculation Approach
A facility leaks four different fluorinated gases from four different systems — refrigerant from the chillers, SF₆ from the switchgear, NF₃ from the cleaning process. Each is a few kilograms a year. Each carries thousands of times the warming power of CO₂.
The arithmetic is trivial. The discipline is not: every gas must land on the same global warming potential basis before it is summed, or the total is wrong in ways a spot check will not reveal.
An F-gas inventory aggregates the leaked mass of every fluorinated gas — HFCs, PFCs, SF₆, NF₃ — into one Scope 1 line in CO₂e. For each gas, multiply kilograms leaked by its global warming potential, then sum. The one rule that governs the result: apply a single GWP basis (AR6 for current corporate reporting) to every gas, and source blend GWPs from the refrigerant data, not the per-gas table.
This methodology covers the aggregation of a fluorinated-gas inventory under the EU F-Gas Regulation 2024/573, the GHG Protocol and IPCC AR6. It assumes you already know how much of each gas leaked — that upstream step belongs to the refrigerant leakage mass-balance methodology — and focuses on the operation that turns those leaked masses into one defensible Scope 1 figure: the per-gas conversion to CO₂e, the consistent-basis rule that holds the total together, the AR5/AR6 decision, and the blend-GWP trap that silently corrupts inventories. Two worked columns, computed on both bases, make the basis choice concrete.
This methodology is the aggregation layer of the fluorinated-gas cluster. It takes the GWP concept from the glossary, the gas list and reporting obligations from the EU F-Gas Regulation, and the published GWP values from the data layer, and tells you how to combine multiple gases from multiple systems into one consistent Scope 1 total. If you already have leaked masses by gas, jump to the core operation or the worked example.
What an F-Gas Inventory Aggregates
Fluorinated gases are synthetic gases used as refrigerants, electrical insulators, cleaning agents and propellants. They share two properties that make them an accounting category of their own: extremely high global warming potential — hundreds to tens of thousands of times CO₂ — and release through leakage rather than combustion. An F-gas inventory pulls together every such release across an organisation into a single Scope 1 fugitive line.
Four families dominate corporate inventories:
- HFCs (hydrofluorocarbons) — refrigerants in air conditioning, chillers and heat pumps. The largest F-gas source for most organisations.
- PFCs (perfluorocarbons) — niche refrigerants and electronics process gases; very long atmospheric lifetimes.
- SF₆ (sulphur hexafluoride) — the insulating gas in high-voltage electrical switchgear. The highest GWP of any gas in routine industrial use.
- NF₃ (nitrogen trifluoride) — an electronics manufacturing and cleaning gas, shared with the semiconductor process inventory.
The aggregator’s job is narrow and specific: take the leaked mass of each gas, convert each to CO₂-equivalent at a consistent GWP basis, and sum. It does not estimate how much leaked — that is a separate, prior step.
Where the Leaked Masses Come From
This methodology starts from a known leaked mass per gas. Establishing that mass — the activity data — is the subject of a separate methodology, and the boundary between the two is deliberate: each is a single source of truth for its own step.
If you do not yet have leaked masses by gas, start at the mass-balance methodology and return here with the kilograms. Everything below assumes that input is in hand.
The Four Gas Families
Each family spans a wide GWP range, and the gas mix on a given site determines which dominates the total. Representative current (AR6) global warming potentials are shown for orientation; the live reference values are in the GWP table.
Refrigerants in chillers, split systems, heat pumps and refrigeration. Often the largest F-gas source. Includes single-component gases (R-134a, R-32) and blends (R-404A, R-410A, R-407C) whose GWP is a mass-weighted mix of components.
AR6 GWP-100 range: R-32 ≈ 771 to R-404A ≈ 4,728 · legacy R-23 ≈ 14,600
Perfluorocarbons — used in some refrigeration, heat-transfer fluids and electronics manufacturing. Very long atmospheric lifetimes and high GWP. Overlaps with the semiconductor process inventory.
AR6 GWP-100: PFC-14 (CF₄) ≈ 7,380, rising across the homologous series
Sulphur hexafluoride — the insulating and arc-quenching gas in high-voltage electrical switchgear and some process equipment. The highest GWP of any gas in routine industrial use, so even a kilogram of leakage is material.
AR6 GWP-100: SF₆ ≈ 25,200 (AR5: 23,500)
Nitrogen trifluoride — a chamber-cleaning and etch gas in electronics and display manufacturing. Where a site both manufactures electronics and reports F-gas leakage, coordinate with the semiconductor etch-gases methodology to avoid double counting.
AR6 GWP-100: NF₃ ≈ 17,400 (AR5: 16,100)
The Core Operation: Mass to CO₂-Equivalent
The calculation per gas is a single multiplication; the inventory is the sum of those products. What makes it a methodology rather than a sum is the discipline around the GWP value — which basis, and which source for each gas.
[gc_gwp gas="…"]. Blends and all refrigerants: [gc_factor key="refrigerants.r_…"].
The single rule that governs an F-gas total is that every GWP in the sum comes from the same basis. An inventory that converts SF₆ at its AR5 value and the HFCs at AR6 is not on either basis — it is a meaningless hybrid. Pick AR6 (the current corporate-reporting default) or AR5 (for legacy comparison), apply it to every line, and disclose which. The next section is entirely about getting this right, because the failure mode is invisible on a casual check.
The AR5 / AR6 Basis Decision
GWP values are revised with each IPCC assessment report. The current corporate-reporting basis is AR6 (2021); the previous basis, still seen in older inventories and some regulatory contexts, is AR5 (2014). For most F-gases the two differ by several percent — enough to matter in a total, and enough to make a mixed-basis inventory wrong.
5.1 Why the basis matters
Switching basis moves the single-component gases materially: SF₆ from 23,500 (AR5) to 25,200 (AR6), NF₃ from 16,100 to 17,400, HFC-134a from 1,430 to 1,526, PFC-14 from 6,630 to 7,380. On a high-SF₆ or high-NF₃ site, a basis change shifts the whole inventory by close to ten percent. That is why the basis must be a deliberate, disclosed choice — not an accident of which table someone happened to use.
5.2 The blend trap — and a live inconsistency to watch
Two blends commonly seen in HFC inventories, R-404A and R-410A, appear in two different places in the data with two different values — and this is the single most dangerous thing on this page.
| Gas | Type | Refrigerant data — AR6 | Refrigerant data — AR5 | Per-gas table — both | Use |
|---|---|---|---|---|---|
| HFC-134a | Pure | 1,526 | 1,430 | 1,526 / 1,430 | Either — they agree |
| HFC-32 | Pure | 771 | 677 | 771 / 677 | Either — they agree |
| R-404A | Blend | 4,728 | 3,922 | 3,922 / 3,922 | Refrigerant data only |
| R-410A | Blend | 2,256 | 1,923 | 2,088 / 2,088 | Refrigerant data only |
For pure gases the two data families agree exactly. For the two blends they do not: the per-gas table carries a single legacy figure in both basis slots (R-404A 3,922; R-410A 2,088), while the refrigerant data correctly recomputes the blend GWP on AR6 component values (R-404A 4,728; R-410A 2,256) and carries a distinct AR5 figure. The per-gas blend values are neither a correct AR6 figure nor properly basis-differentiated.
Take blend GWPs — R-404A, R-410A, R-407C and any other mixture — from the refrigerant dataset (4728 for AR6, its AR5 alternate for AR5), never from the per-gas GWP table. Sourcing R-410A from the per-gas table prints 2,088 where the refrigerant data shows 2,256 — a 7.5% understatement on that line that reads as a data bug, not a rounding difference. Pure gases (SF₆, NF₃, HFC-134a, HFC-32) may come from either family; blends must come from the refrigerant data.
5.3 Which basis to report
Use AR6 for all current corporate reporting — the GHG Protocol, CSRD/ESRS E1, CDP and SBTi all expect AR6 GWP-100. Use AR5 only to reconcile against a historical inventory built on that basis, or where a specific regulatory regime still mandates it. Whichever you choose, state it in the inventory and hold it across every gas and every scope. The worked example shows the same site on both bases so the magnitude of the choice is explicit.
Governing Standards
Four reference layers govern F-gas accounting and reporting. They nest from the regulatory obligation, through the inventory boundary, to the GWP science.
| Layer | Standard | What it supplies |
|---|---|---|
| Regulatory obligation | EU F-Gas Regulation 2024/573 | The gas list, leak-checking and record-keeping duties, the HFC phase-down quota schedule, and bans on high-GWP gases in new equipment. |
| Phase-down treaty | Kigali Amendment to the Montreal Protocol | The international HFC phase-down schedule that drives the shift to low-GWP HFO substitutes — the policy context shrinking HFC inventories by mandate. |
| Inventory boundary | GHG Protocol Corporate Standard | Organisational and operational boundaries; the Scope 1 classification of fugitive F-gas leakage; the requirement to report it on one consistent GWP basis. |
| GWP science | IPCC AR6 (AR5 for historical comparison) | The 100-year global warming potentials used to convert each gas to CO₂-equivalent, on both the current AR6 basis and the prior AR5 basis. |
GWP Reference Table
Representative current values for the gases most often found in corporate F-gas inventories, surfaced live. Refrigerant values come from the refrigerant dataset, which carries the AR6 figure as its value and the AR5 figure as an alternate on the same row; pure-gas values are confirmed against the per-gas GWP table, with which they agree. Blends are sourced exclusively from the refrigerant dataset for the reasons set out in §5.
| Gas | Type | AR6 GWP-100 | AR5 GWP-100 | Source key |
|---|---|---|---|---|
| R-32 (HFC-32) | Pure HFC | 771 | 677 | refrigerants.r_32 |
| R-134a (HFC-134a) | Pure HFC | 1530 | 1300 | refrigerants.r_134a |
| R-410A | HFC blend | 2256 | 1923 | refrigerants.r_410a |
| R-404A | HFC blend | 4728 | 3943 | refrigerants.r_404a |
| R-23 (HFC-23) | Pure HFC | 14600 | 12400 | refrigerants.r_23 |
| R-1234yf (HFO) | Low-GWP HFO | 0.501 | <1 | refrigerants.r_1234yf |
| SF₆ | Pure | 25200 | 23500 | gwp.SF6 |
| NF₃ | Pure | 17400 | 16100 | gwp.NF3 |
| PFC-14 (CF₄) | Pure PFC | 7380 | 6630 | gwp.PFC_CF4 |
Refrigerant values: refrigerant dataset (IPCC AR6 GWP-100, with AR5 alternates), aligned to the EU F-Gas Regulation 2024/573 gas list. Pure-gas values cross-checked against the per-gas GWP table. Low-GWP HFO figures (R-1234yf ≈ 0.5) show the magnitude of the Kigali-driven substitution. Full live values render on the published page.
Worked Example
A single site leaking four gases, computed on both bases so the basis choice is explicit. Per-gas GWPs render live (drift-proof); only the leaked masses are fixed, and the products are labelled illustrative at the snapshot MasterBrain version. The blends are sourced from the refrigerant dataset, the pure gases from the per-gas table — exactly the sourcing rule from §5.
Leaked masses for the reporting year (from the mass-balance methodology):
R-410A (comfort cooling)
45.0 kg · R-134a (legacy chillers) 12.0 kg · SF₆ (HV switchgear) 3.5 kg · NF₃ (on-site electronics) 8.0 kgGWPs live @ MB v2025.74. Blends from refrigerant data; pure gases from per-gas table.
| Gas | kg leaked | AR6 GWP | AR6 tCO₂e | AR5 GWP | AR5 tCO₂e |
|---|---|---|---|---|---|
| R-410A BLEND | 45.0 | 2256 | 101.520 | 1923 | 86.535 |
| R-134a | 12.0 | 1530 | 18.312 | 1300 | 17.160 |
| SF₆ | 3.5 | 25200 | 88.200 | 23500 | 82.250 |
| NF₃ | 8.0 | 17400 | 139.200 | 16100 | 128.800 |
| Total F-gas (Scope 1) | — | — | 347.232 | — | 314.745 |
Per-gas CO₂e is leaked mass × GWP ÷ 1,000 (kg → tonnes). On the AR6 basis MultiGasCo reports 347.232 tCO₂e; on AR5, 314.745 tCO₂e — a difference of 32.487 tCO₂e, or about 10.3%, driven almost entirely by the SF₆ and NF₃ lines where the two bases diverge most. Note the structure of the result: NF₃ at 8 kg contributes more than R-134a at 12 kg, because its GWP is more than ten times higher. In an F-gas inventory the gas mix, not the leaked mass, sets the ranking.
Both columns are correct; they differ only in basis. This is precisely why the basis must be disclosed: a reader cannot tell 347 from 315 tCO₂e apart without knowing which IPCC assessment the GWPs came from. Report the AR6 figure for current corporate disclosure, state the basis explicitly, and keep every other gas and scope in the inventory on the same basis.
Aggregation Across Sources and Scopes
A real organisation leaks F-gases from several systems, and the aggregator combines them — but only those that belong on the same line. Three boundary rules keep the aggregation clean.
- All operational fugitive F-gas leakage is one Scope 1 line. Refrigerant leakage from cooling, SF₆ from switchgear, NF₃ from on-site processes — all are direct fugitive releases from owned or controlled equipment, so they aggregate into a single Scope 1 F-gas total. The gas families differ; the scope and the accounting do not.
- Process F-gases may belong to a process inventory, not here. Where a site manufactures electronics or semiconductors, the NF₃, SF₆ and PFCs consumed in the process are accounted under the semiconductor etch-gases methodology, which uses a process-specific utilisation-and-abatement method rather than a leakage mass. Decide once which inventory each gas stream belongs to, and do not let a gas appear in both.
- F-gas in sold products is downstream, not Scope 1. Refrigerant charged into a product that leaks during its use phase at the customer is a Scope 3 (use of sold products) emission for the manufacturer, not a Scope 1 fugitive release. Only gas leaking from the organisation’s own equipment is aggregated here.
Rolling the F-gas line up into a full Scope 1 + 2 + 3 corporate total is the job of the GHG inventory aggregator — this page produces the single F-gas figure that feeds it.
Edge Cases
- Blends versus components. A blend such as R-410A can be reported either with its blend GWP applied to the total blend mass, or by decomposing it into component gases (R-32 + R-125) and applying each component GWP. Use one approach consistently; do not mix a blend figure for one charge with a component decomposition for another. The blend GWP from the refrigerant dataset already reflects the mass-weighted component mix.
- Reclaimed and recycled gas. Gas reclaimed from equipment and reused on site is not a fresh purchase and does not itself create an emission — only the fraction that leaks does. The mass-balance methodology handles reclaim on the activity-data side; the aggregator simply converts whatever leaked mass it is given.
- End-of-life recovery. Gas recovered from decommissioned equipment and sent for destruction is not emitted and is not counted. Gas vented at end of life is emitted in full and must be included in the leaked mass. The distinction — like the flaring/venting distinction for methane — turns on whether the gas was destroyed or released.
- Banked gas in equipment. The charge sitting inside operating equipment is a bank, not an emission; it becomes an emission only as it leaks or is vented. Do not report installed charge as emissions — only the leaked fraction.
- Legacy ODS refrigerants. CFCs and HCFCs (R-12, R-22) still in older equipment are ozone-depleting substances controlled under the Montreal Protocol, and their direct CO₂e is often reported for completeness even though they sit outside the Kyoto basket. Report them where required, with their own GWPs, and label them clearly as ODS.
- Unknown or mixed charges. Where the exact refrigerant in old equipment is unknown, identify it from nameplate data or service records before assigning a GWP. Guessing a low-GWP gas where a high-GWP blend is present can understate a line by thousands of times the mass — the highest-leverage uncertainty in the whole inventory.
The Kigali Phase-Down Context
HFC inventories are shrinking by mandate. The Kigali Amendment to the Montreal Protocol commits parties to a stepwise phase-down of HFC consumption, implemented in the EU through the F-Gas Regulation’s declining quota and equipment bans. The trajectory is a falling cap on the CO₂-equivalent of HFCs that may be placed on the market, pushing the market toward low-GWP HFO substitutes such as R-1234yf (GWP ≈ 0.5) and R-1234ze.
The schedule below shows the characteristic shape of the phase-down — a declining quota against a baseline — which is why a well-managed F-gas inventory should trend downward year on year as high-GWP blends are replaced.
| Point | % of HFC baseline (CO2e quota) |
|---|---|
| 2015 | 100.0 % of HFC baseline (CO2e quota) |
| 2018 | 93.0 % of HFC baseline (CO2e quota) |
| 2019 | 63.0 % of HFC baseline (CO2e quota) |
| 2021 | 45.0 % of HFC baseline (CO2e quota) |
| 2024 | 31.0 % of HFC baseline (CO2e quota) |
| 2027 | 24.0 % of HFC baseline (CO2e quota) |
| 2030 | 21.0 % of HFC baseline (CO2e quota) |
| 2036 | 15.0 % of HFC baseline (CO2e quota) |
For inventory practice this means two things: the gas mix shifts over time toward low-GWP HFOs, dropping the inventory’s CO₂e even at constant leaked mass; and reporting must keep pace with substitutions, because a chiller recharged from R-410A (GWP ≈ 2,256) to R-32 (GWP ≈ 771) or an HFO changes that line’s CO₂e by a factor of three or more. Track the refrigerant actually in each system, not the one it was commissioned with.
Uncertainty and Data Quality
For an F-gas inventory the GWP values are fixed and certain — they are published constants. Essentially all the uncertainty lives on the activity-data side (how much leaked) and in gas identification (which gas it was), both of which sit upstream of this aggregator.
| Data quality | Leaked-mass basis | Typical use |
|---|---|---|
| Highest | Mass-balance from purchase, sales and inventory records, with gas identified per system | Material refrigerant banks; assured inventories |
| Mid | Screening method — installed charge × default annual leak rate by equipment type | Distributed small equipment; first inventories |
| Lowest | Estimated charge and assumed gas where records are absent | Legacy equipment pending survey — flag for improvement |
Because GWPs span three orders of magnitude — from an HFO at about 0.5 to SF₆ at 25,200 — misidentifying the gas in a system dwarfs any error in the leaked mass. Confirming that a system contains R-410A rather than an assumed R-134a, or SF₆ rather than a lower-GWP alternative, changes the line by thousands of kg CO₂e per kg leaked. Prioritise positive gas identification on the largest charges before refining leak-rate estimates.
Verification and Assurance
- Basis consistency. The first check: every gas in the F-gas total — and every other line in the wider inventory — is on one declared GWP basis. A mixed AR5/AR6 inventory fails on sight.
- Blend GWP sourcing. Verifiers confirm blend GWPs (R-404A, R-410A, R-407C) were taken from the refrigerant dataset, not a generic per-gas table that may carry an undifferentiated legacy value. An R-410A line at 2,088 rather than 2,256 is a flag.
- Gas identification. Each charge’s gas is substantiated against nameplate or service records, not assumed. Unidentified charges are flagged, not silently assigned a convenient GWP.
- Boundary integrity. No gas appears in both the F-gas line and a process inventory (semiconductor) or a downstream Scope 3 line. Recovered-and-destroyed end-of-life gas is excluded; vented gas is included.
- Leaked-mass provenance. The kilograms feeding this aggregator trace back to the mass-balance or screening method, with the method disclosed per the mass-balance methodology.
The audit framework follows the GHG Protocol Corporate Standard for boundaries and the basis-consistency rule, with the EU F-Gas Regulation 2024/573 supplying the gas list and record-keeping expectations.
Error Traps with Magnitudes
Each error below produces a specific, quantifiable distortion. Magnitudes use the MultiGasCo example (347.232 tCO₂e AR6 total) to make validation and disclosure risk concrete.
| Error | What happens | Magnitude (MultiGasCo) | How to avoid |
|---|---|---|---|
| Source a blend from the per-gas table | R-410A taken at the per-gas table’s 2,088 instead of the refrigerant data’s 2,256. | −7.56 tCO₂e on the R-410A line (−7.45%) 45 kg × 2,088 = 93.96 vs 101.52 tCO₂e. R-404A is worse: −17% per line (3,922 vs 4,728). |
Source every blend from the refrigerant dataset. The per-gas table carries an undifferentiated legacy blend value. |
| Mix GWP bases within the total | Some gases converted at AR5, others at AR6 — e.g. SF₆ left on AR5 inside an AR6 inventory. | −5.95 tCO₂e and an invalid basis SF₆ at 23,500 not 25,200 gives 341.28 vs 347.23 — but the real failure is that the total is on no coherent basis at all. |
Apply one basis to every gas and every scope. Disclose which. Never mix AR5 and AR6 in a total. |
| Misidentify the gas | A high-GWP gas assigned a low-GWP value (e.g. an SF₆ charge logged as a refrigerant). | Up to thousands of × per kg GWPs span ~0.5 (HFO) to 25,200 (SF₆). Misidentification dwarfs any leaked-mass error. |
Confirm each gas from nameplate or service records before assigning a GWP. Flag unknowns. |
| Report installed charge as emissions | The gas bank inside operating equipment counted as if it had leaked. | Overstates by the un-leaked bank A 45 kg R-410A charge reported whole is 101.52 tCO₂e instead of the few kg that actually leaked. |
Count only the leaked fraction. Installed charge is a bank, not an emission. |
| Count end-of-life recovered gas | Gas recovered and destroyed at decommissioning included as emitted. | Overstates by the recovered mass Mirror error: vented end-of-life gas omitted understates by the same. |
Destroyed gas is excluded; vented gas is included. The test is destruction, not removal. |
| Double-count a process gas | NF₃ or SF₆ appearing in both the F-gas line and the semiconductor process inventory. | Doubles the affected line NF₃ at 139.2 tCO₂e counted twice adds a phantom 139.2 tCO₂e. |
Assign each gas stream to one inventory only. Coordinate with the process methodology. |
Methodology Metadata — for GHG Inventory Documentation
Copy into your GHG inventory methodology statement for inventory-transparency compliance. Adjust the basis and gas-list lines to match your inventory.
Frequently Asked Questions
An F-gas inventory is the Scope 1 total of all fluorinated gases an organisation leaks — HFCs from refrigeration and air conditioning, PFCs from niche and electronics uses, SF₆ from electrical switchgear, and NF₃ from electronics processes. They are aggregated as their own category because they share two defining features: very high global warming potential, from hundreds to over twenty-five thousand times CO₂, and release through leakage rather than combustion. Each gas is converted from leaked mass to CO₂-equivalent using its own GWP, then summed into one line on a single consistent basis.
Use AR6 for all current corporate reporting — the GHG Protocol, CSRD/ESRS E1, CDP and SBTi all expect AR6 GWP-100. Use AR5 only to reconcile against a historical inventory built on that basis or where a specific regulation still mandates it. The choice is material: switching basis moves SF₆ from 23,500 to 25,200 and NF₃ from 16,100 to 17,400, shifting a high-SF₆ inventory by close to ten percent. Whichever basis you use, apply it to every gas and every scope, and state it in the inventory — a reader cannot interpret the total without knowing the basis.
Because R-410A is a blend, and the two data sources treat it differently. The refrigerant dataset recomputes the blend’s GWP from its AR6 component values, giving 2,256 on the AR6 basis and 1,923 on AR5. A generic per-gas GWP table may carry a single legacy figure (2,088) in both basis slots, which is neither a correct AR6 value nor properly basis-differentiated. Always source blend GWPs — R-410A, R-404A, R-407C and others — from the refrigerant dataset, never the per-gas table. For pure gases such as SF₆ or HFC-134a the two sources agree and either can be used.
They are two halves of one calculation. The refrigerant leakage mass-balance methodology establishes how much of each gas leaked — through the screening method (installed charge × annual leak rate) or the full mass-balance from purchase, sales and inventory records. This aggregator takes those leaked masses and turns them into one CO₂-equivalent total, applying GWPs on a consistent basis and summing across gases and sources. Start at the leakage methodology if you need leaked masses; use this page once you have them.
No. The charge inside operating equipment is a bank, not an emission — it becomes an emission only as it leaks or is vented. Reporting the full installed charge as emissions would overstate the inventory enormously; a 45 kg R-410A charge is over a hundred tonnes CO₂-equivalent, against the few kilograms that actually leak in a year. Count only the leaked fraction. The same logic applies at end of life: gas recovered and destroyed is not emitted, while gas vented to atmosphere is emitted in full.
It depends on the use. NF₃ and SF₆ leaking from owned equipment — switchgear, on-site systems — belong in this Scope 1 F-gas inventory. But NF₃, SF₆ and PFCs consumed as process gases in semiconductor or electronics manufacturing are accounted under the semiconductor etch-gases methodology, which uses a process-specific utilisation-and-abatement method rather than a leakage mass. The rule is to assign each gas stream to exactly one inventory: decide whether a given stream is equipment leakage or process consumption, and never let it appear in both.