Chapter 01 · Foundations

Methodology

This chapter documents the thermodynamic basis, the calculation model, the data sources, the feasibility logic, the verification procedure and the standards alignment behind Ardor reports. The methodology follows CEN/CENELEC EN 50600-4-6 (Energy Reuse Factor) and Commission Delegated Regulation (EU) 2024/1364, and is coordinated with the waste heat platform operated by the BfEE.

At a glance

Waste heat (MWh/yr) = IT load (MW) × PUE × 8,760 h × η_capture. The capture coefficient η_capture is assigned by cooling type (air, mixed, liquid) and may be overridden with measured operational data. The recoverable share is tested against a proximity-based offtaker and classified as feasible, borderline or not feasible. The output is expressed as the ERF metric under EN 50600-4-6 and as the reporting figure under Annex II of Commission Delegated Regulation (EU) 2024/1364.

1. Thermodynamic basis

A data centre converts nearly all of its electrical energy input into heat. The IT load generates waste heat in the server hall, the cooling plant, the uninterruptible power supply and the ancillary systems generate heat at their respective operating levels. The total electrical power draw P_tot equals the total heat dissipation on an annual-average basis. Differences between input and dissipation are in the single-digit percent range and are absorbed into the measurement uncertainty.

The heat usable for external offtakers is strictly smaller than the heat dissipated. It becomes usable only when it crosses the cooling system boundary in the sense of EN 50600-4-6 and is delivered into a heat-accepting infrastructure: a district heating network, an industrial process, a greenhouse cluster, a residential heating system. The difference between dissipated heat and externally usable heat is the recovery share. It depends on the cooling technology, the delivery temperature, the spatial proximity of an offtaker and the temporal alignment of supply and demand.

2. Calculation model

2.1 Core equation

Waste heat (MWh/yr) = IT load (MW) × PUE × 8,760 h × ηcapture

IT load is the annual-average active electrical power drawn by the IT payload, in megawatts. It differs from installed nameplate capacity by the actual utilisation factor and should be taken from the annual mean of measured active power where available.

PUE (Power Usage Effectiveness) is the ratio of total facility energy to IT energy. A PUE of 1.0 is the theoretical optimum; typical values are between 1.3 and 1.6. § 11 EnEfG requires new-build data centres in Germany to meet PUE ≤ 1.2 from July 2026.

8,760 h is the number of hours in a year. Below full availability, the effective hours reduce proportionally. Ardor uses 8,760 h as a default and permits the operator to substitute a measured or contractual availability (typically 7,500 to 8,500 h).

η_capture is the recovery share of waste heat. It is assigned by cooling type and combines effects of temperature level, coupling losses and the practically achievable extraction rate. Default values are conservative and fall within the ranges reported by Pehnt et al. (2010) and Schlomann et al. (2020).

2.2 Default values for η_capture by cooling type

Cooling type	Temperature window	η_capture literature	η_capture Ardor
Air cooling	25 – 40 °C	30 – 50 %	50 %
Mixed cooling (air plus liquid)	40 – 60 °C	50 – 70 %	60 %
Liquid cooling (direct-to-chip)	60 – 80 °C	70 – 90 %	80 %

Ardor default values lie within the literature ranges. Operators may override η_capture in the intake assistant on the basis of their own ERF measurement; in that case the ERF metric is taken directly from the measurement and the default is not used.

2.3 Worked example: 10 MW, air cooling, PUE 1.4

Total power     = 10 MW × 1.4                     = 14 MW
Annual energy   = 14 MW × 8,760 h                 = 122,640 MWh/yr
Recoverable     = 122,640 MWh × 0.50               = 61,320 MWh/yr
Households (≈)  = 61,320 MWh ÷ 20 MWh/household    = 3,066

The household conversion of 20 MWh per household and year follows the Bitkom and AGFW convention for average German residential heat demand. The figure is illustrative only; it does not enter any regulatory submission.

3. ERF and standards alignment

The Energy Reuse Factor under CEN/CENELEC EN 50600-4-6 is the reference indicator for data centre waste heat utilisation. The numerator is the externally used thermal energy over the reporting period; the denominator is the total facility energy input. The measurement boundary is the interface between the cooling system and the heat-accepting infrastructure.

ERF = Ereused,ext ÷ Efacility,total

Annex II of Commission Delegated Regulation (EU) 2024/1364 uses ERF directly as a reporting figure. EU-wide reporting is submitted annually by 15 May through the national competent authority. Ardor produces an ERF-conformant output per site and can export it in the machine format defined by the Delegated Regulation.

4. Feasibility logic

Recoverable waste heat is an upper bound. Regulatory assessments additionally require a judgement on whether the heat can plausibly be delivered to an offtaker in the foreseeable term. Ardor classifies each site into one of three categories:

Category	Condition	Interpretation
feasible	Distance to network centroid ≤ 3 km or inside documented network coverage	Delivery realistic at accepted temperature; usable as upper bound of ERF.
borderline	Distance ≤ documented network coverage but > 3 km, or delivery temperature < 40 °C	Delivery technically possible, economics to verify; to be examined in the cost-benefit analysis.
not feasible	Distance > network coverage and no industrial single-offtaker identified	Exemption under § 17 EnEfG, Art. L.236-2 Code de l'énergie or Art. 26 EED potentially available; recorded as a documented negative finding.

The distance check is performed on the geocoded site address against the relevant national district heating dataset. Network coverage per operator corresponds to the service area polygon or a documented radius recorded in the dataset. In the absence of network coverage, Ardor performs a second check for industrial single-offtakers (greenhouse clusters, chemicals, food processing, paper and board) up to 5 km away and evaluates temperature compatibility.

5. Data sources

The data base is structured by country and follows the publicly accessible district heating and industrial offtaker registers of the respective Member States. Third-party primary sources (BAFA, BfEE platform, France Chaleur, RVO) are imported without modification. Derived attributes (temperature grade, coverage radius) are versioned with provenance and date.

Country	District heating	Regulation & reporting	Ardor status
Germany	AGFW heat map, Mainova, municipal utilities	BAFA / BfEE waste heat platform	Live (32 sites)
France	France Chaleur, Fonds Chaleur (ADEME), SNCU, Idex, Engie Solutions, Dalkia, Veolia	DGEC, Code de l'énergie L.236-2, Décret 2025-1382	Port in preparation (15 sites)
Netherlands	Ennatuurlijk, Vattenfall NL, HVC, Stadsverwarming Purmerend, Eneco	RVO, Omgevingswet, Erkende Maatregelenlijst	Port in preparation (20 sites)
Ireland	SEAI District Heating, Dublin DHS	CRU, SEAI, EirGrid/ESB Networks	Groundwork (primarily exemption dossier)
Italy	AIRU, A2A, Iren, Hera	MASE, GSE (decreto legislativo in consultation)	Seed data (6 sites)
Spain	ADHAC, Ecoenergies Barcelona, Veolia España	MITECO, Real Decreto in preparation	Seed data (6 sites)
Nordics (SE/DK/FI)	Stockholm Exergi, HOFOR, Helen, Fortum	Energimyndigheten, Energistyrelsen, Energiavirasto	Seed data (24 sites, SE/NO/FI)

Non-EU sites in Norway and Switzerland are held under bilateral recognition. Their reporting does not count toward EED obligations but is carried on the platform for market visibility.

6. Verification and uncertainty

Report quality depends on the measurement basis. Ardor distinguishes three verification tiers and records the tier used on every output.

Tier	Basis	Uncertainty band
V1 estimate	publicly available site data, operator statement without measurement record	± 20 %
V2 operator-confirmed	operator confirms IT load, PUE and cooling from internal operating data	± 10 %
V3 measured	ERF derived directly from a submitted measurement record or from a BAFA submission	± 5 %

V3 is the target tier for every report submitted to an authority. V1 is used only for public sample reports or for marketplace listings where the operator explicitly consents to the publication of a preliminary estimate.

Independent verification is available on request through a body accredited under ISO/IEC 17065 or through the competent national authority. Measurement records and verification reports are retained under version control for at least ten years and disclosed to the competent authority on request.

7. Link to the Waste Heat Certificate (WHC)

The methodology of this chapter is the calculation basis of the proposed Waste Heat Certificate (WHC). One WHC corresponds to one megawatt-hour of externally used waste heat verified under EN 50600-4-6. Issuance follows the close of the measurement period and an independent verification. The serial number is unique, the certificate is transferable and its retirement is terminal. Registry architecture is described in the position paper Waste Heat Certificates.

8. Scientific foundation

The methodology rests on two foundational works at Fraunhofer ISI and ifeu Heidelberg. Pehnt et al. (2010, ifeu Heidelberg) defined the national inventory methodology for industrial waste heat, including the temperature and utilisation classification. Schlomann et al. (2020, Fraunhofer ISI) updated the methodology and set the quantitative frame for the BfEE Plattform für Abwärme. The parameter bands published there form the basis for the η_capturedefaults used here.

For data centres specifically, EN 50600-4-6 defines the measurement boundary and the ERF calculation. Ardor combines the site-level methodology, the standard-level measurement boundary and the reporting formats of Commission Delegated Regulation 2024/1364 into a single operational workflow.

9. Disclaimer

Disclaimer

Reports produced by Ardor are generated with automated methods. The operator is responsible for the accuracy of inputs and for compliant submission to the competent authority. Ardor accepts no liability for the use of its reports in regulatory filings. All methods, data sources and assumptions are documented in every report and can be audited. Methodological feedback is welcomed at /en/contact.

10. References

[01]Directive (EU) 2023/1791 of the European Parliament and of the Council of 13 September 2023 on energy efficiency (recast).
[02]Commission Delegated Regulation (EU) 2024/1364 of 14 March 2024 supplementing Directive (EU) 2023/1791.
[03]Directive (EU) 2023/2413 of the European Parliament and of the Council of 18 October 2023 (RED III).
[04]Energieeffizienzgesetz (EnEfG), Federal Law Gazette I, 18 November 2023.
[05]Loi n° 2025-391 du 30 avril 2025 (DDADUE), Journal Officiel de la République française.
[06]Décret n° 2025-1382 du 29 décembre 2025, Journal Officiel de la République française.
[07]CEN/CENELEC EN 50600-4-6: Information technology, data centre facilities and infrastructures, Part 4-6: Energy Reuse Factor.
[08]Schlomann, B., Rohde, C. and Eichhammer, W. (2020). Evaluation of the German waste heat potential. Fraunhofer ISI.
[09]Pehnt, M., Bödeker, J., Arens, M., Jochem, E. and Idrissova, F. (2010). Use of Industrial Waste Heat. ifeu Heidelberg and Fraunhofer ISI.
[10]Ardor Institute (2026). Toward a Market for Waste Heat Certificates. Discussion Paper No. 1.
[11]Ardor Institute (2026). Waste Heat Certificates as a European Market Instrument. Discussion Paper No. 2.

Last updated: April 2026. Document version 1.0. A methodology review by BfEE Referat 511 has been requested; feedback will be incorporated in the next revision.