About this document. This guidance was prepared in 2026 as a good-faith effort to document how the Higg platform calculated the FEM 2018 energy, greenhouse gas, water, waste and refrigerant values. FEM 2018 predates the calculation model maintained by Cascale today: the platform applied its own conversion factors and emission factors and wrote the results back into the assessment as answers, where data consumers see them next to the facility’s own answers. What follows is reconstructed from the FEM 2018 question and answer definitions, from those stored results across 9,949 assessments, and from the release history. It may not accurately describe all platform behaviour, in particular edge cases that leave no trace in the data. Statements derived from the data rather than from a specification are marked observed, with the number of filings behind them.

Scope

This document describes the energy, greenhouse gas and refrigerant values the Higg platform calculated for a FEM 2018 assessment and reported back to the facility, its verifier and data consumers. They are the FEM’s own simplified footprint, not a GHG Protocol inventory, and they are neither purely location-based nor purely market-based:

• One totalGHGemissions figure combines fuel combustion, refrigerants, purchased electricity, steam and chilled water; the only split published is renewable versus non-renewable. There is no scope split and no location/market pair.
• Purchased electricity carries a single country grid factor (location-based). Energy Attribute Certificates were not collected.
• Biogenic CO2 from biomass and biodiesel is inside the totals, at full combustion factors.
• Purchased steam and chilled water are counted at a fixed fuel-oil-type factor with no supplier data; refrigerant purchased or added is treated as emitted; wastewater emissions are not calculated at all in FEM 2018.
• Factors are single CO2e values per source on an IPCC AR4/AR5 basis (FEM 2023 moved to AR6); there is no gas-by-gas breakdown and no process emissions.

The Higg Assessment Model recomputes FEM 2018 answers on a GHG Protocol aligned basis with current factors; those results differ from the platform values described here by design and are documented in the HAM guidance.

Energy Data

Facilities first flagged each energy source they use (yes/no, one answer per source; coal, biomass, natural gas and fuel oil have a further flag per type), then reported a quantity and a unit for each source they track. Energy is reported for the site as a whole: there is no split by facility type, domestic use or vehicle use in FEM 2018.

The platform’s MJ result for each source is stored as an answer under the ref_id in the last column; for coal, biomass, natural gas and fuel oil the platform also stored a group total (ensourcecoaltotal, ensourcebiomassgentotal, ensourcebiomasswoodtotal, ensourcenaturalgastotal, ensourcefueloiltotal).

Conversion to MJ

The platform converted each reported quantity to MJ in its own code; no specification of the conversions was published. They are recovered here from the stored totals (total ÷ quantity, deterministic within each unit across all filings). Energy units use the standard factors below; mass and volume units imply the energy densities in the next table. Electricity, purchased steam, chilled water and the onsite and offsite renewables share the energy-and-volume unit set, which includes volume units: a volume unit chosen for those sources converts to 0 MJ (observed in every such filing), so that energy is missing from the facility’s totals.

Energy units (all sources):

Mass units convert to kilograms and volume units to litres with the standard factors (g 0.001, lb 0.4535, oz 0.0283, ton (metric) 1,000, ton (short) 907.185; m³ 1,000, gal 3.7854, cu ft 28.3168, cu yard 764.554, CCF 2,831.68).

Energy densities implied by the stored totals (median of total ÷ quantity per unit, converted to the base unit; observed, with the number of filings that reported the source):

Notable in this table: the coal types carry distinct densities, the biomass types share one density of 20.33 MJ per kg regardless of type, and purchased steam and chilled water have no density because they only convert from energy units. The platform grouped biomass into two totals (observed): ensourcebiomasswoodtotal covers wood briquette, sawdust and slivers of wood, and ensourcebiomassgentotal covers every other biomass type; the group totals, not the typed values, feed the emissions totals.

Conversion to kg CO2e

Each source’s emissions are its MJ total multiplied by a single factor held by the platform: ensource[source]totalghg = ensource[source]total × factor. For coal, natural gas and fuel oil the platform also stored a group emissions total that is the sum of the typed values (coal: in 97% of filings), so a facility’s effective coal factor depends on the mix of coal types it reported.

Factors applied (kg CO2e per MJ) — observed as the median of totalghg ÷ total; unless noted, every filing shows the same factor:

Purchased electricity is multiplied by a country grid factor (ensourceelectricpurchtotalgwp, kg CO2e per MJ, selected by the facility’s sitecountry answer), licensed from IEA and updated by the platform between cadences. The factors the platform applied are not preserved outside the assessments themselves. Comparing the factor stored with each filing against the FEM 2018 country factors now held in Cascale’s shared reference data (which were re-derived for the Higg Assessment Model), the platform’s factor is on median 7% higher, and only 17% of filings are within 5% of the shared value, so the shared data cannot be used to reproduce FEM 2018 platform emissions.

The renewable sources (solarphoto, wind, geotherm, hydro, microhydro) carry a factor of 0. Purchased steam and chilled water carry a fixed fuel-oil-type factor of 0.0629; the facility could not report a supplier factor or a generation mix in FEM 2018.

Refrigerants

Refrigerant use is reported in the Air Emissions section in up to five entries (airsourcerefrig1 … airsourcerefrig5): the refrigerant (airsourcerefrig[n]), the quantity added or used (airsourcerefrig[n]quant) and its unit (airsourcerefrig[n]unit: g, kg, lb, oz, ton (metric), ton (short)). The platform converted the quantity to kilograms (airsourcerefrig[n]total), looked up the refrigerant’s global warming potential (airsourcerefrig[n]totalgwp) and multiplied (airsourcerefrig[n]totalghg); totalRefrigerantEmissions is the sum of the five entries. Observed to equal the sum of its components in 100% of 918 filings.

The GWP values below are the ones the platform stored with FEM 2018 filings (IPCC AR4 basis, blends as marketed). Only refrigerants that at least one facility reported can be recovered this way; 293 of the 344 refrigerants offered in FEM 2018 were never reported and their platform values are unknown (the FEM 2023 Energy Source Changes document lists FEM 2022 values for those that changed in FEM 2023, several of which were 0).

Totals

The platform stored these aggregates with each assessment. Their composition is observed: the share of filings in which the stored total equals the sum of the listed components within 0.5%.

ensourcetotal — total MJ across all energy sources. Observed to equal the sum of its components in 96% of 7,703 filings.

totalNonRenewableEnergyEmissions — kg CO2e from the non-renewable sources: coal, naturalgas, natgaslpg, natgaspropane, diesel, fueloil, petrol, steampurch, chilledwater, electricpurch. Observed to equal the sum of its components in 97% of 7,668 filings.

totalRenewableEmissions — kg CO2e from the renewable sources including biomass and biodiesel: biomassgen, biomasswood, biomassbiodiesel, solarphoto, wind, geotherm, hydro, microhydro. Observed to equal the sum of its components in 99% of 633 filings.

totalRefrigerantEmissions — kg CO2e from the five refrigerant entries.

totalNonRenewableEmissions — totalNonRenewableEnergyEmissions plus totalRefrigerantEmissions. Observed to equal the sum of its components in 100% of 7,690 filings.

totalGHGemissions — totalNonRenewableEmissions plus totalRenewableEmissions. Observed to equal the sum of its components in 100% of 7,697 filings.

totalNormalizedGHGemissions — totalGHGemissions divided by the annual production volume sipfacilityannualprodvolquant (in the unit the facility chose in sipfacilityannualprodvolunits). Observed to equal the sum of its components in 99% of 7,604 filings.

The remaining few percent of filings whose totals do not add up were, as far as the data shows, assessments whose totals were not recomputed after a late edit; the platform did not publish a rule for them.

Unit Answer Codes

Unit answers in FEM 2018 are numeric codes. Which set a code belongs to depends on the source, as given in the source table above.

Energy and volume set (electricity, steam, chilled water, the renewable sources, diesel, petrol, the natural gas group and the fuel oil group):

Mass and energy set (the coal and biomass groups):

Refrigerant mass set (airsourcerefrig[n]unit):

Cadence Notes

FEM 2017, FEM 2018 and FEM 2019 share one design: the same questions, ref_ids, answer options and units. The differences between them are the country grid emission factors, which the platform updated each year, and behaviour that is only visible in the data:

• FEM 2017 steam, chilled water and electricity quantities occur only in energy units; from FEM 2018 facilities also chose volume units for these sources and the platform converted those entries to 0 MJ (observed: every such filing totals 0), so that energy vanished from the totals.
• Answer values in these cadences are numeric codes ("1", "8"), not the labels used from FEM 2020. The same code means different units for different sources: 1 is kWh for electricity but g for coal. The tables under Unit answer codes give the mapping for each unit set, and any consumer must resolve a unit answer through the set that applies to that source.

Not Calculated in FEM 2018

No wastewater treatment emissions, no vehicle energy, no domestic/production split, no facility-type split, no energy or emissions normalised per facility type, and no market-based figures were calculated. Baseline and target answers (enbaseline…, en…target…) were converted to MJ with the same densities but do not enter any total.