How Dcycle calculates purchase and CapEx emissions

Purchased goods and services (Scope 3, Category 1) and capital goods (Category 2) are usually the largest and least intuitive part of a corporate inventory. You upload a spend file, and a number comes out. This article explains exactly how that number is produced, so you can read it — and challenge it — with confidence.

The two methods

Every purchase line in Dcycle is calculated one of two ways, and you choose which when you upload:

• Standard (spend-based) — you upload the amount, Dcycle assigns the emission factor. The factor comes from EXIOBASE, an environmentally extended multi-regional input-output model, expressed as kgCO2e per euro of spend for a given economic sector, country and year.
• Custom (supplier-specific) — you upload physical quantities (kg, units, m³, tonnes) and your own emission factor, typically supplied by the vendor. EXIOBASE is bypassed entirely.

Both methods can coexist in the same inventory: use custom factors for the suppliers who give you real data, and spend-based for the rest.

The spend-based chain, step by step

This is the part that surprises people, so it is worth reading closely. In the standard method the emission factor is not derived from the product description on the invoice line. It is derived from the supplier.

supplier name + country
   → matched against Dcycle's official supplier registry (DUNS-based)
   → the supplier's activity code (US SIC 1987, or CNAE 2009 mapped onto it)
   → the corresponding EXIOBASE product / sector
   → monetary factor for that sector × country × year
   → emissions = amount in EUR × factor

Each link in that chain matters:

1. Supplier matching. Dcycle looks for an exact text match on supplier name + country, first among the suppliers already created in your organization, then against the official registry. If it matches, the supplier inherits that company’s real activity code.

2. Activity code. The primary classification is US SIC (1987 edition); CNAE 2009 codes are mapped onto it. This code is what determines the sector — and you can correct it at any time in Data > Categories > Purchases and CapEx > Supplier management, which triggers a recalculation. See Fix a supplier’s activity code.

3. No match, no factor. If the name doesn’t match anything and no activity code is supplied, the supplier is created with status Under review and the line stays without emissions until you assign a code. An unassigned supplier is a visible gap, not a silent zero.

4. Currency. Non-euro amounts are converted to EUR using the exchange rate at the purchase date before the factor is applied.

5. The factor itself. EXIOBASE monetary factors in Dcycle cover 200 products across 49 regions and 60 sectors, as an annual series from 2015 to 2022 — 9,800 product-country combinations and 78,400 factor rows. The factor applied is the one for the country of purchase and the year of the purchase date.

Years after 2022. The EXIOBASE series currently ends in 2022. Purchases dated 2023 or later carry the 2022 factor for their sector. A 2024 line and a 2022 line in the same sector will therefore show the same intensity — that is expected, not a bug.

Why the same spend can give very different numbers

Because the factor is a sector average per euro, the sector assignment dominates the result. These are real Spanish factors for 2022, as applied in production:

EXIOBASE sector (Spain, 2022)	kgCO2e per €
Air transport services	1.1484
Fabricated metal products, except machinery and equipment	0.4922
Other electrical machinery and apparatus	0.4313
Other machinery and equipment	0.3687
Office machinery and computers	0.1831
Construction work	0.1285
Other services	0.1204
Other business services	0.1180
Post and telecommunication services	0.1023
Retail trade services	0.1019
Computer and related services	0.0819
Public administration and defence services	0.0685

One million euros of spend becomes 68.5 tCO2e as public administration services, or 492 tCO2e as fabricated metal products — a factor of seven, from the same invoice total. This is the single biggest driver of accuracy in a spend-based inventory, and it is why supplier classification deserves more attention than the spend figures themselves.

Factors also move year to year, because they are the ratio of a sector’s emissions to its monetary output. Construction work in Spain, as loaded in Dcycle:

Year	kgCO2e per €
2015	0.3312
2016	0.2913
2017	0.2817
2018	0.2392
2019	0.2483
2020	0.2581
2021	0.2706
2022	0.1285

Part of that decline is real decarbonisation; part is prices rising faster than emissions, which mechanically lowers the intensity per euro. Keep it in mind before reading a spend-based reduction as an operational win.

Capital goods (Category 2)

Capital goods run through the same engine, with two differences:

• The line is flagged as CapEx rather than OpEx, so it reports under Category 2 instead of Category 1.
• The full acquisition value is accounted for in the year of acquisition. It is not amortised across the asset’s useful life, in line with the GHG Protocol treatment of capital goods.

That means a year with a large investment will show a step change in Category 2. A €5.6M plant investment in a single year is a genuine spike in the inventory, not an upload error — and it will often outweigh every operational category combined.

How to sanity-check a spend-based upload

Four checks catch almost every problem:

1. Look at the spread of sectors. If dozens of lines share one identical implied intensity, they share one supplier and therefore one sector. That is a red flag whenever the lines describe genuinely different things.
2. Divide emissions by amount, line by line. The result is the applied factor. Compare it against the table above — an IT purchase priced at a construction intensity is visible in one glance.
3. Check the largest lines first. Spend-based results are dominated by a handful of big-ticket rows; the misclassification that matters is nearly always in the top five.
4. Review suppliers Under review. Those lines are contributing zero. Assigning the code is what brings them into the inventory.

What the spend-based method is, and is not

The GHG Protocol treats spend-based calculation as a screening method: good enough to size a category and find where the emissions sit, not to demonstrate a reduction. Its structural limits:

• The factor is a national sector average, not your supplier’s performance. A best-in-class vendor and the worst in its sector receive the same intensity.
• Sector granularity is coarse — 200 products for the whole economy. “Other business services” covers a lot of very different activities.
• Price, not physics. Paying more for the same item raises the calculated emissions.
• Reductions are hard to prove. Cutting spend lowers the number without necessarily lowering emissions.

This is why the GHG Protocol prefers supplier-specific data, and why moving your largest suppliers to custom factors is the highest-value improvement you can make to Categories 1 and 2. Start with the vendors that account for most of your spend, and see Bulk upload supplier emission factors.

Key takeaways

• The spend-based factor comes from the supplier’s activity code, not the invoice description.
• Emissions = amount in EUR × EXIOBASE factor for that sector, country and year.
• Factors run 2015–2022; later years reuse the 2022 value.
• Sector assignment can move a result by 7x. Review suppliers before questioning the spend.
• CapEx counts in full in the year of acquisition, with no amortisation.
• Spend-based is a screening method. Supplier-specific factors are the way out.