OpenData Wiki

Dataset documentation and review before publication

User Tools

Site Tools


This report is being reviewed before publication. It will appear here once the review is complete.
datasets:tdc:5c838461-ad82-5ce1-ae17-4b14915023ca

This is an old revision of the document!


Table of Contents

Life Cycle Inventory of Carbon fiber, PAN-based, wet spinning, UHM, sized — Europe (RER)

Unit-process dataset, cradle-to-gate, 1 kg of UHM PAN-based carbon fiber — documentation for external critical review

Item Value
Dataset name (BAFU naming rules, DQR §4.7.1) Carbon fiber, PAN-based, wet spinning, UHM (RER)
Dataset UUID / version carbon-fiber-pan-based-wet-spinning-uhm-rer-4611b8 / draft (tdc)
Reference product, unit Carbon fiber, PAN-based, wet spinning, UHM, sized, kg
Location (DQR location code) RER — Europe
Reference year / time period 2024-2026 / 2024-2026
Technology level current average technology (as described in the sources, Section 3)
Background database BAFU 2026 v1 (Memgraph graph bafuquery-db); no other background database
LCIA method EF v3.1 (European Commission, Joint Research Centre 2022) (Brightway 2 / bw2calc (Mutel 2017))
Status ready for external review
Prepared by tdc pipeline (datasetcreator agent), documentation and final check by tdc-doc

1 Introduction

Conversion of wet-spun PAN precursor fiber to ultra-high modulus (UHM) PAN-based carbon fiber: oxidation/stabilization (~250-300 C), low-temperature carbonization (~350-400 C), high-temperature carbonization, graphitisation (~2500-3000 C), surface treatment and sizing. From precursor fiber of the region to sized carbon fiber. Functional unit: 1 kg carbon fiber. Reference: BAFU 2026 (Memgraph).

This report documents the dataset according to ISO 14040 (ISO 14040:2006 Environmental Management — Life Cycle Assessment — Principles and Framework (Incl. Amd 1:2020) 2006) and ISO 14044 (ISO 14044:2006 Environmental Management — Life Cycle Assessment — Requirements and Guidelines (Incl. Amd 1:2017, Amd 2:2020) 2006) and to the BAFU LCI Data Quality Requirements v2.1 (Frischknecht 2023). It is written for the external critical reviewer (ISO 14044 clause 6, ISO/TS 14071 (ISO/TS 14071:2014 Environmental Management — Life Cycle Assessment — Critical Review Processes and Reviewer Competencies 2014)): all calculation steps are reproducible from the files listed in Annex G, all input values are traceable to a source (reference list and verbatim excerpts in Annex D) or to a stated assumption (Section 4.6).

2 Goal and scope definition (ISO 14044, 4.2)

2.1 Goal of the study

  • Intended application: background dataset for LCA studies (BAFU 2026 database extension)
  • Reasons for carrying out the study: missing dataset in the background database
  • Intended audience: LCA practitioners, external critical reviewer
  • Comparative assertions intended to be disclosed to the public: no

2.2 Functional unit and reference flow

The functional unit is 1 kg of UHM PAN-based carbon fiber of product at the plant gate. Reference flow: 1 kg of UHM PAN-based carbon fiber of Carbon fiber, PAN-based, wet spinning, UHM, sized at the plant gate. Product definition: main product: 1 kg sized UHM PAN-based carbon fiber (98.7% core fiber + 1.3% epoxy sizing) [see product exchange].

2.3 Product system and system boundary

Cradle-to-gate: raw-material and energy supply (BAFU datasets), transport to the plant, the process steps of Section 3, emissions and waste treatment.

Product system and system boundary (unit processes, flows crossing the boundary).

Product system and system boundary (unit processes, flows crossing the boundary).

Life cycle stage Included Notes / source
Fiber stabilization, carbonization, electrolysis and washing yes 0.9769 kg per kg product
Exhaust gas treatment 1 yes 35.99 kg per kg product
Exhaust gas treatment 2 (regenerative thermal oxidizer) yes 10.53 kg per kg product
Fiber drying and sizing yes 0.9928 kg per kg product
Graphitisation (UHM grade) yes 1 kg per kg product

2.4 Cut-off criteria

No intentional cut-off; missing items are listed as data gaps. Flows excluded by cut-off are listed with their estimated mass/energy share in Section 4.6.

2.5 Allocation procedures

No co-product allocation (single reference product).

2.6 LCIA method and impact categories

EF v3.1 (European Commission, Joint Research Centre 2022), all impact categories, characterisation only. Normalisation and weighting are not applied (optional elements, ISO 14044 4.4.3); no weighting is used for any comparison.

2.7 Data requirements and data quality requirements (ISO 14044 4.2.3.6)

Requirement Target Achieved (see Section 4.5)
Time-related coverage 2020–2026 data 2024-2026
Geographical coverage Europe 2 geographical proxy dataset(s): Heat, natural gas, at heat plant: ‘Heat, natural gas, at heat plant’ used at ENTSO-E (no RER version; Wastewater treatment, class 3: ‘Treatment, sewage, to wastewater treatment, class 3’ used at CH (no
Technology coverage current industrial technology Section 3
Precision lognormal uncertainty and pedigree on every flow (Frischknecht 2023) lognormal on every exchange; Monte Carlo GWP100 95 % interval 56.6–69.5
Completeness all inventory classes (raw materials, catalyst, neutralisation, water, energy, emissions to air/water/soil, by-products, infrastructure, transport) all classes present
Representativeness market-weighted route and feedstock Section 3 (market and routes)
Consistency one modelling approach for all regions/grades family and regional checks: ok
Reproducibility every value computed by a tool or cited all amounts reproduced by tdc-doc
Sources of data ≥ 3 peer-reviewed articles (Crossref-verified), ≥ 1 patent, ≥ 1 grey document 16 references, 9 with DOI, 3 patent(s)
Uncertainty of information Monte Carlo on the pedigree distributions Section 6.5

2.8 Assumptions and limitations

  • none recorded

2.9 Type of critical review

External critical review by an independent expert (ISO/TS 14071) — pending.

3 Product, market and technology

3.1 Product definition and grades

main product: 1 kg sized UHM PAN-based carbon fiber (98.7% core fiber + 1.3% epoxy sizing)

3.2 Market, production volumes and production routes

3.2.0.1 Production routes and shares

PAN-based carbon fiber dominates the market (>90% of all high-performance CF). Pitch-based and rayon-based CFs together cover <10%. (Source: Prenzel2023 — “With a significant share (more than 90%) of today’s high-performance CFs being made from polyacrylonitrile (PAN)”)

The ultra-high-modulus (UHM) grade (>500 GPa) requires a graphitisation step at 2,500–3,000 °C in addition to the standard stabilization and carbonization. Graphitised fibers cover a marginal market share relative to HT/IM grades. (Source: Huang2009 — “In terms of final mechanical properties, carbon fibers can be roughly classified into ultra high modulus (>500 GPa)”; Prenzel2023 — “since this type of fiber only covers a marginal market share, it will not be considered further in this paper”)

3.2.0.2 Raw materials and origins
Material Global origin share (2023) Source
PAN precursor fiber Asia-Pacific ~63% Prenzel2023: “the Asia-Pacific region is the most relevant provider of all produced PAN worldwide (forecast 2023: 63% of global production)”
Acrylonitrile monomer Asia ~70%, North America ~18%, Western Europe ~7% Guo2026 / market data
Sizing (epoxy resin) Asia-Pacific dominates BCC Research / Huang2009
Nitrogen (inert gas) Produced on-site from air JapanCFMA_LCI_2022: “All nitrogen is purchased nitrogen”
3.2.0.3 Production volumes
3.2.0.3.1 Carbon fiber (PAN-based, all grades)
Region / world Volume Year Source
World (total CF consumption) 34,200 t/yr 2010 Huang2009 Table 1
Japan (PAN-based CF) 17,288 t/yr FY2017 JapanCFMA_LCI_2022: “Thus 17,288 tons of Japanese CF are obtained”
World (estimated, 2024, all grades) ~140,000 t/yr (≈ 140 kt) 2024 BCC Research 2024; Prenzel2023 Figure 3
3.2.0.3.2 PAN precursor fiber

The precursor requirement for UHM grade (45% mass yield) is ≈ 2.2 kg per 1 kg CF. At ~140 kt/yr world CF production, total PAN precursor demand ≈ 310 kt/yr.

Region CF production share Source
Asia-Pacific (dominated by China, Japan, Korea) ~60–70% Prenzel2023; BCC Research
North America ~15–20% Prenzel2023 Figure 3
Europe (RER) ~10–15% Prenzel2023 Figure 3
3.2.0.4 Main producers by region
  • RER (Europe): SGL Carbon (Germany), Toray Europe (Germany/France), Teijin Europe
  • RNA (North America): Hexcel (USA), Toray (USA), Cytec (USA)
  • RAS (Asia): Toray Industries (Japan), Teijin (Japan), Mitsubishi Chemical (Japan), Zhongfu Shenying (China), HS Hyosung (Korea)
3.2.0.5 UHM grade specifics

UHM carbon fiber (>500 GPa) is a niche grade used mainly in aerospace, hydraulic cylinders, and pressure vessels. Production volumes are in the hundreds to low thousands of tonnes per year globally, far less than the HT/IM grades. The graphitisation step (2,500–3,000 °C, argon atmosphere, several hours) is the defining additional process step for this grade. (Source: US5266294A — “Carbon fiber products having a modulus of more than 500 GPa”; Sha2018 — “graphitization is normally carried out at 2000–3000 °C”)

3.2.0.6 Alternative routes (sensitivity)

The main alternative to the standard wet-spun UHM route is pitch-based UHM fiber (higher modulus but lower strength, different precursor). This dataset models the PAN-based UHM route, which is the most widely used for high-modulus aerospace applications. The pitch-based alternative is noted here but not modeled.

Item Value Year Source

3.3 Process description

Conversion of wet-spun PAN precursor fiber to ultra-high modulus (UHM) PAN-based carbon fiber: oxidation/stabilization (~250-300 C), low-temperature carbonization (~350-400 C), high-temperature carbonization, graphitisation (~2500-3000 C), surface treatment and sizing. From precursor fiber of the region to sized carbon fiber. Functional unit: 1 kg carbon fiber. Reference: BAFU 2026 (Memgraph). Processing steps as recorded in the unit process:

Step Conditions (T, p, time) Main inputs/outputs Source
1. Precursor fiber intake and line loading Wet-spun PAN precursor fiber (from the precursor dataset of the region, 1 kg per 1/(yield) kg of carbon fiber) is fed to the stabilization line: fiber unwinding, inspection and tensioning onto the continuous line. []
2. Oxidation / stabilization Continuous oven at about 250-300 C in air, tensioned: the PAN polymer cross-links and becomes infusible; non-carbon elements (H, N, O) are partly released to the off-gas. Off-gas from this stage goes to exhaust gas treatment. []
3. Low-temperature carbonization Inert atmosphere (N2) furnace, about 350-500 C: nitrogen and residual oxygen are eliminated, the fiber skeleton becomes carbonaceous; off-gas (HCN, CO, NH3) to exhaust gas treatment. []
4. High-temperature carbonization Inert (N2) furnaces, temperature ramped up to about 2000 C: decarburation, turbostratic carbon forms; off-gas to exhaust gas treatment. []
5. Graphitisation UHM-specific step: inert (Ar) furnaces at about 2500-3000 C with high tension: crystalline size grows, pseudo-graphitic structure -> ultra-high modulus; off-gas to exhaust gas treatment. []
6. Exhaust gas treatment Thermal oxidiser (RTO/DFTO) of the combined off-gas (HCN, CO, NH3, H2O): destruction by combustion with support natural gas; residual flue gas (CO2, N2, H2O, minor NOx) to air. []
7. Surface treatment Anodic etching in acidic electrolyte (e.g. nitric/organic acid bath) under electric current: functional oxygen groups on the fiber surface; electrolyte consumed, regenerated or discarded. []
8. Washing and drying Fiber washed to remove residual electrolyte; the wash water goes to wastewater; fiber dried (hot air). []
9. Sizing Sizing agent (epoxy-compatible) applied as aqueous solution, a few wt% on the fiber; water evaporated. []
10. Winding and packaging Sized fiber wound on creels/packs, quality control, packaging as the finished UHM carbon fiber product. []

4 Life cycle inventory analysis (ISO 14044, 4.3)

4.1 Data collection

Two-step literature search: (1) the user’s embedded reference library (local vector search, — queries, 0 excerpts retained), then (2) a web and database round (scholarly indexes, patents, agency and industry documents) to complete the process description. Each reference was verified (DOI against Crossref; patents against the patent office record; other documents fetched and the quoted passage found in the text). Reference statistics: 16 references, 9 Crossref-verified articles, 3 patents, 4 grey documents.

4.2 Calculation procedures

4.2.1 Stoichiometry and mass/element balances

Stoichiometric model of the dataset (spec.json), recomputed by tdc-doc: balances pass. Balances computed by the tdc stoichiometry model (Annex B):

Balance In (kg) Out (kg) Residual (%)
C — — -0.761
H — — -0.799
N — — -0.034
O — — -0.722
mass — — -0.3

4.2.2 Energy model

Heat and electricity are computed per consumer from process physics with the tdc energy model (Annex A); each parameter is cited or declared as an assumption.

Consumer Energy form Duty / power Time Energy per kg Source
[Fiber stabilization, carbonization, electrolysis and washing] Fiber sensible heat, stabilization (25 to 300 C) heat (electric) — — 1.183 MJ [Stabilization 200-300 C in air, 1-2 h residence (Prenzel2023 sec 3.1.1). Precursor mass 2.8683 kg per kg of core fiber = 2.82244 kg per kg p]
[Fiber stabilization, carbonization, electrolysis and washing] Stabilization oven air heating (recirculating oxidizer air) heat (electric) — — 36.37 MJ [‘the air is heated in an electric heater to a predetermined temperature and heats the processed fibers due to convective heat exchange’ (IJA]
[Fiber stabilization, carbonization, electrolysis and washing] Fiber sensible heat, low-temperature carbonization (300 to 1000 C) heat (electric) — — 1.807 MJ [Stabilized fiber 2.5815 kg per kg core (2.5402 kg per kg product / 0.984; grade-common 10% stabilization mass loss, precursor 2.82244; DATA ]
[Fiber stabilization, carbonization, electrolysis and washing] Fiber sensible heat, high-temperature carbonization (1000 to 1800 C) heat (electric) — — 1.033 MJ [Mid-route fiber 1.2908 kg per kg core (1.2701 kg per kg product / 0.984; grade-common 50% LT-carbonisation loss; DATA GAP - same harmonizing]
[Fiber stabilization, carbonization, electrolysis and washing] Carbonisation + surface-treatment stage powers: N2 purge heating, wall losses, electrolysis, drives (sourced RSC stage kW residual) heat (electric) — — 28.97 MJ [RSC_ESI_2016 Table 1 sourced stage powers. SHARED grade-common physics of the cf-pan-ws-v2 family (same ~29 MJ as IM/HM-rer).]
[Fiber stabilization, carbonization, electrolysis and washing] Stabilization furnace NG burner (on-site gas assist, 0.22 kg NG per kg product) heat (fuel) — — 6.695 MJ [Sourced on-site NG for the stabilization oxidation furnace (carrier fuel). Per user decision 4 the on-site fuel is carried by the energy mod]
[Fiber stabilization, carbonization, electrolysis and washing] Inert gas make-up compression (N2, slight overpressure of the carbonisation atmospheres) electricity 0.4109 1 0.4109 kWh [All nitrogen purchased externally (JapanCFMA_LCI_2022 sec 5-3-6: ‘All nitrogen was purchased externally’); inert N2 for carbonisation (Prenz]
[Fiber stabilization, carbonization, electrolysis and washing] Fiber line tensioner drives, furnace drive trains, electrolysis and washing drives electricity 3.378 1 3.378 kWh [Continuous line under constant tension (Prenzel2023 sec 3.1.1) with drive trains per furnace; 3.04 kW per kg/h of core (3 kW per kg/h produc]
[Exhaust gas treatment 1] First off-gas treatment (quench/wash/scrubber) specific energy electricity (utility) — — 0.023 kWh [First exhaust-gas treatment of the continuous CF line (before the RTO); BAFU dataset ‘Carbon fiber production, exhaust gas treatment 1’ has ]
[Exhaust gas treatment 2 (regenerative thermal oxidizer)] Regenerative thermal oxidizer NG fuel (on-site, 2.06 kg NG per kg product) heat (fuel) — — 5.908 MJ [Sourced on-site RTO fuel (carrier fuel). Per user decision 4 carried by the energy model fuel_supply as BAFU heat per MJ. This plus the step]
[Exhaust gas treatment 2 (regenerative thermal oxidizer)] RTO fans and blower (off-gas and oxidising-air circulation) electricity (utility) — — 0.0262 kWh [RTO fans and post-treatment blower of the regenerative thermal oxidizer; the sourced RSC_ESI_2016 abatement system (25 kW at 12.5 kg/h) incl]
[Fiber drying and sizing] Sizing-bath water heating (technical drying, recovered closed loop) heat (electric) — — 1.2 MJ [The sizing-bath water is a recovered closed loop: ’the water evaporates in technical drying, is recovered and fed back; the coating remains ]
[Fiber drying and sizing] Drying air handling + winding drives (sourced RSC stage powers) electricity 1.2 1 1.2 kWh [RSC_ESI_2016 Table 1 sourced stage powers (measured, 100 t/a pilot line) - independent of the Benitez2021 benchmark.]
[Graphitisation (UHM grade)] Graphitisation furnace electric power, UHM grade (graphite-block/induction furnace at 2500-3000 C in Ar) electricity 35 1 35 kWh [YongXian2026 sourced specific power range (the defining energy sink that makes UHM need more than HM). No UHM graphitisation furnace dimensi]

Totals per kg: heat 68.77 MJ, electricity 63.53 kWh, primary energy — MJ.

4.2.3 Emissions

Direct emissions: Water, off-gas vapor (to air) 1.424 kg; Nitrogen, off-gas (to air) 0.7378 kg; Nitrogen oxides, fossil (to air) 0.00012 kg; Hydrogen cyanide (to air) 0.00191 kg; Carbon monoxide, fossil (to air) 0.04726 kg; Ammonia (to air) 0.001357 kg; Methane (to air) 0.001444 kg; Ethylene (to air) 0.001366 kg; Ethane (to air) 0.001314 kg; Carbon dioxide, fossil (to air) 3.348 kg; Nitrogen, liquid, at plant (to air) 8.822 kg.

4.2.4 Transport

Statistical transport from the provenance of each raw material to the plant region (Annex C): Transport, freight, lorry, 16t-32t gross weight, fleet average 0.00208 tkm

4.2.5 Infrastructure

Chemical plant, organics 8e-10 p

4.3 Inventory table

Per 1 kg of UHM PAN-based carbon fiber of product. GSD² = squared geometric standard deviation of the lognormal distribution; pedigree scores (reliability, completeness, temporal, geographical, technological) per BAFU DQR v2.1 (Frischknecht 2023).

Exchange Direction Amount Unit BAFU dataset / elementary flow GSD² Pedigree Source
Transport, freight, lorry, 16t-32t gross weight, fleet average input 0.00208 tkm Transport, freight, lorry, 16t-32t gross weight, fleet average [RER] 3.05 (4,2,2,2,1) [—]
Chemical plant, organics input 8e-10 p Chemical plant, organics [RER] 3.25 (3,4,5,2,1) [—]
Electricity, medium voltage, RER, at grid input 63.529 kWh Electricity, medium voltage, production RER, at grid [RER] 1.14 (2,3,3,1,1) [—]
Sizing agent, epoxy input 0.016 kg Epoxy resin, liquid, at plant [RER] 1.13 (3,3,2,1,1) [—]
Oxygen, resource in air input 3.7003 kg Oxygen 1.22 (4,3,2,1,1) [—]
Nitrogen, liquid, at plant input 8.822 kg Nitrogen, liquid, at plant [RER] 1.13 (3,3,2,1,1) [—]
Ammonium bicarbonate, at plant input 0.01842 kg Ammonium bicarbonate, at plant [RER] 1.13 (3,3,2,1,1) [—]
Ethylene glycol, at plant input 0.0076 kg Ethylene glycol, at plant [RER] 1.13 (3,3,2,1,1) [—]
Water, off-gas vapor (to air) output 1.4239 kg Water 1.57 (4,3,2,1,1) [—]
Nitrogen, off-gas (to air) output 0.73784 kg Nitrogen 1.57 (4,3,2,1,1) [—]
Nitrogen oxides, fossil (to air) output 0.00012 kg Nitrogen oxides 1.57 (4,3,2,1,1) [—]
Hydrogen cyanide (to air) output 0.00191 kg Cyanide 1.57 (4,3,2,1,1) [—]
Carbon monoxide, fossil (to air) output 0.04726 kg Carbon monoxide, fossil 5.06 (4,3,2,1,1) [—]
Ammonia (to air) output 0.001357 kg Ammonia 1.57 (4,3,2,1,1) [—]
Methane (to air) output 0.001444 kg Methane, fossil 1.57 (4,3,2,1,1) [—]
Ethylene (to air) output 0.001366 kg NMVOC, non-methane volatile organic compounds 1.57 (4,3,2,1,1) [—]
Ethane (to air) output 0.001314 kg Ethane 2.05 (4,3,2,1,1) [—]
Carbon dioxide, fossil (to air) output 3.3479 kg Carbon dioxide, fossil 1.22 (4,3,2,1,1) [—]
Heat, natural gas, at heat plant input 68.772 MJ Heat, natural gas, at heat plant [ENTSO-E] 1.13 (3,3,2,1,1) [—]
Nitrogen, liquid, at plant (to air) output 8.822 kg Nitrogen 1.52 (3,3,2,1,1) [—]
Wastewater treatment, class 3 input 2.6e-05 m3 Treatment, sewage, to wastewater treatment, class 3 [CH] 1.13 (3,3,2,1,1) [—]
PAN precursor fiber, wet spun input 2.802 kg tdc dataset cf-v2-precursor-rer 1.22 (4,3,2,1,1) [—]

4.4 Data quality assessment

Energy verification per sub-process (policy B): Fiber stabilization, carbonization, electrolysis and washing: physics, 3 independent reference(s); Exhaust gas treatment 1: literature, 1 independent reference(s); Exhaust gas treatment 2 (regenerative thermal oxidizer): literature, 2 independent reference(s); Fiber drying and sizing: literature, 3 independent reference(s); Graphitisation (UHM grade): literature, 1 independent reference(s). Pedigree scores: 22 of 22 exchanges scored by the developer; the others carry the tdc-doc default (3, 3, 3, 3, 3) (marked *). Proxies (BAFU datasets that do not represent exactly the flow): Heat, natural gas, at heat plant → Heat, natural gas, at heat plant (Heat, natural gas, at heat plant: ‘Heat, natural gas, at heat plant’ used at ENTSO-E (no RER version) — geographical proxy); Wastewater treatment, class 3 → Treatment, sewage, to wastewater treatment, class 3 (Wastewater treatment, class 3: ‘Treatment, sewage, to wastewater treatment, class 3’ used at CH (no RER version) — geographical proxy).

4.5 Data gaps and assumptions register

Item Class Value used Reason / queries tried Effect (Section 6.3)

4.6 Validation of data (ISO 14044 4.3.3.2)

  • Mass and element balances: Section 4.2.1.
  • Cross-check with independent data: BAFU datasets of similar processes (Carbon fiber production, fiber stabilization, carbonization, electrolysis and washing, Carbon fiber production, fiber drying and sizing, Carbon fiber production, exhaust gas treatment 1, Carbon fiber production, exhaust gas treatment 2) and published values (Section 6.6); deviations beyond a factor of 3 are justified with a cited excerpt.
  • Independent second-model review (Gemma-4-26B-A4B-it-GGUF): 3 finding(s), review of 2026-10-28 02:15 (Annex E).
  • Citation test: every quoted excerpt was found in its source text (all found).
  • Independent recalculation by tdc-doc: independent recalculation of GWP100 = pipeline LCIA (lcia.json): PASS.

5 Life cycle impact assessment (EF v3.1)

Characterisation results per 1 kg of UHM PAN-based carbon fiber, computed with bw2calc on the BAFU 2026 v1 database (Mutel 2017; European Commission, Joint Research Centre 2022).

Impact category Indicator Unit Result
acidification accumulated exceedance (AE) mol H+-Eq 0.2019
climate change global warming potential (GWP100) kg CO2-Eq 62.4
climate change: biogenic global warming potential (GWP100) kg CO2-Eq 0.3563
climate change: fossil global warming potential (GWP100) kg CO2-Eq 62.03
climate change: land use and land use change global warming potential (GWP100) kg CO2-Eq 0.01324
ecotoxicity: freshwater comparative toxic unit for ecosystems (CTUe) CTUe 255.9
ecotoxicity: freshwater, inorganics comparative toxic unit for ecosystems (CTUe) CTUe 231.6
ecotoxicity: freshwater, organics comparative toxic unit for ecosystems (CTUe) CTUe 24.32
energy resources: non-renewable abiotic depletion potential (ADP): fossil fuels MJ, net calorific value 1189
eutrophication: freshwater fraction of nutrients reaching freshwater end compartment (P) kg P-Eq 0.02157
eutrophication: marine fraction of nutrients reaching marine end compartment (N) kg N-Eq 0.03575
eutrophication: terrestrial accumulated exceedance (AE) mol N-Eq 0.5043
human toxicity: carcinogenic comparative toxic unit for human (CTUh) CTUh 2.526e-08
human toxicity: carcinogenic, inorganics comparative toxic unit for human (CTUh) CTUh 7.811e-09
human toxicity: carcinogenic, organics comparative toxic unit for human (CTUh) CTUh 1.745e-08
human toxicity: non-carcinogenic comparative toxic unit for human (CTUh) CTUh 2.738e-07
human toxicity: non-carcinogenic, inorganics comparative toxic unit for human (CTUh) CTUh 2.554e-07
human toxicity: non-carcinogenic, organics comparative toxic unit for human (CTUh) CTUh 1.841e-08
ionising radiation: human health human exposure efficiency relative to u235 kBq U235-Eq 29.68
land use soil quality index dimensionless 172.9
material resources: metals/minerals abiotic depletion potential (ADP): elements (ultimate reserves) kg Sb-Eq 7.619e-05
ozone depletion ozone depletion potential (ODP) kg CFC-11-Eq 1.507e-06
particulate matter formation impact on human health disease incidence 1.56e-06
photochemical oxidant formation: human health tropospheric ozone concentration increase kg NMVOC-Eq 0.135
water use user deprivation potential (deprivation-weighted water consumption) m3 world eq. deprived 6.046

Contribution of the inventory items to climate change (GWP100) and the other main categories.

Contribution of the inventory items to climate change (GWP100) and the other main categories.

Climate change 62.4 kg CO2-eq per kg. Main contributors: PAN precursor fiber, wet spun 49 %; Electricity, medium voltage, RER, at grid 33 %; Heat, natural gas, at heat plant 10 %; Carbon dioxide, fossil (to air) 5 %; Nitrogen, liquid, at plant 3 %.

6 Interpretation (ISO 14044, 4.5)

6.1 Significant issues

The inventory items that dominate climate change are PAN precursor fiber, wet spun 49 %; Electricity, medium voltage, RER, at grid 33 %; Heat, natural gas, at heat plant 10 %; Carbon dioxide, fossil (to air) 5 %; Nitrogen, liquid, at plant 3 %. The parameters with the largest effect (Section 6.3) are: Graphitisation (UHM grade): electricity_kWh (per kg step output) (-14 / +0 %); PAN precursor fiber, wet spun (pedigree 95 % interval) (-9 / +11 %); Electricity, medium voltage, RER, at grid (pedigree 95 % interval) (-4 / +4 %).

6.2 Completeness check

board stage ‘inventory’ (closed + automatic check re-run): PASS; inventory completeness (classes, no zero amounts): PASS; every exchange scored in EF v3.1 (linked to BAFU / biosphere): PASS

6.3 Sensitivity check

One-at-a-time variation of (a) every assumption of Section 4.6 over its plausible range and (b) every parameter for which the literature gives a range or several values (Annex D); scenarios for alternative routes, feedstocks and regions.

Parameter Base Low / high Source of the range Δ GWP100 (%) Δ other categories (max %)
Graphitisation (UHM grade): electricity_kWh (per kg step output) 35.0 8.0 / 35.0 [Li2024 (Applied Sciences 2024, 14(6):2528, inducti; YongXian2026 sourced specific power range 15-35 kW] -13.8 / +0.0 30.7
PAN precursor fiber, wet spun (pedigree 95 % interval) 2.80201 /1.22 / x1.22 [pedigree (Annex H)] -8.7 / +10.6 19.9
Electricity, medium voltage, RER, at grid (pedigree 95 % interval) 63.52914 /1.14 / x1.14 [pedigree (Annex H)] -3.9 / +4.4 9.8
Exhaust gas treatment 2 (regenerative thermal oxidizer): heat_MJ (per kg step output) 5.9082 6.93 / 6.951 [Benitez2021 benchmark sub-process ’carbon fiber pr; harmonizing_lca_cf_notes: ’Natural Gas: Exhaust tr] +1.5 / +1.5 2.5
Heat, natural gas, at heat plant (pedigree 95 % interval) 68.77205 /1.13 / x1.13 [pedigree (Annex H)] -1.1 / +1.2 2.0
Carbon dioxide, fossil (to air) (pedigree 95 % interval) 3.34793 /1.22 / x1.22 [pedigree (Annex H)] -1.0 / +1.2 1.2
Fiber stabilization, carbonization, electrolysis and washing: electricity_kWh (per kg step output) 26.4566 28.2264 / 28.2264 [Benitez A, Wulf C, de Palmenaer A, et al. Ecologic] +0.9 / +0.9 2.0
Exhaust gas treatment 2 (regenerative thermal oxidizer): electricity_kWh (per kg step output) 0.0262 0.0262 / 0.1885 [Benitez A, Wulf C, de Palmenaer A, et al. Ecologic; RSC_ESI_2016 Table 1 ‘Abatement system 25 kW’ at 1] +0.0 / +0.9 1.9
Fiber drying and sizing: electricity_kWh (per kg step output) 1.5922 0.6257 / 0.6257 [Benitez A, Wulf C, de Palmenaer A, et al. Ecologic] -0.5 / -0.5 1.1
Nitrogen, liquid, at plant (pedigree 95 % interval) 8.822 /1.13 / x1.13 [pedigree (Annex H)] -0.4 / +0.4 1.2
Fiber stabilization, carbonization, electrolysis and washing: heat_MJ (per kg step output) 6.6946 4.6509 / 4.6509 [Benitez A, Wulf C, de Palmenaer A, et al. Ecologic] -0.3 / -0.3 0.5
Exhaust gas treatment 1: electricity_kWh (per kg step output) 0.023 0.02 / 0.0229 [Benitez A, Wulf C, de Palmenaer A, et al. Ecologic; Benitez2021 benchmark sub-process ‘carbon fiber pr; RSC_ESI_2016 Table 1 ’Abatement system 25 kW’ at 1] -0.1 / -0.0 0.1
Fiber drying and sizing: heat_MJ (per kg step output) 0.0 0.3351 / 0.3351 [Benitez A, Wulf C, de Palmenaer A, et al. Ecologic] +0.0 / +0.0 0.1
Sizing agent, epoxy (pedigree 95 % interval) 0.016 /1.13 / x1.13 [pedigree (Annex H)] -0.0 / +0.0 0.3

6.4 Consistency check

Final control: 22 item(s) passed; no failure.

6.5 Uncertainty analysis

Monte Carlo simulation (1000 runs) on the lognormal distributions of all exchanges:

Impact category Median 2.5 % 97.5 % GSD²
acidification / accumulated exceedance (AE) 0.2028 0.1818 0.2283 1.12
climate change / global warming potential (GWP100) 62.65 56.57 69.55 1.11
climate change: biogenic / global warming potential (GWP100) 0.3563 0.3252 0.3924 1.1
climate change: fossil / global warming potential (GWP100) 62.28 56.21 69.13 1.11
climate change: land use and land use change / global warming potential (GWP100) 0.01327 0.01209 0.01462 1.1
ecotoxicity: freshwater / comparative toxic unit for ecosystems (CTUe) 257.3 225.9 293.3 1.14
ecotoxicity: freshwater, inorganics / comparative toxic unit for ecosystems (CTUe) 232.8 205.3 264.1 1.13
ecotoxicity: freshwater, organics / comparative toxic unit for ecosystems (CTUe) 24.43 20.27 29.08 1.2
energy resources: non-renewable / abiotic depletion potential (ADP): fossil fuels 1193 1073 1336 1.12
eutrophication: freshwater / fraction of nutrients reaching freshwater end compartment (P) 0.02159 0.0197 0.02379 1.1
eutrophication: marine / fraction of nutrients reaching marine end compartment (N) 0.03588 0.03281 0.03935 1.1
eutrophication: terrestrial / accumulated exceedance (AE) 0.5068 0.4524 0.5699 1.12
human toxicity: carcinogenic / comparative toxic unit for human (CTUh) 2.532e-08 2.311e-08 2.803e-08 1.1
human toxicity: carcinogenic, inorganics / comparative toxic unit for human (CTUh) 7.863e-09 7.057e-09 8.858e-09 1.12
human toxicity: carcinogenic, organics / comparative toxic unit for human (CTUh) 1.745e-08 1.595e-08 1.92e-08 1.1
human toxicity: non-carcinogenic / comparative toxic unit for human (CTUh) 2.75e-07 2.493e-07 3.061e-07 1.11
human toxicity: non-carcinogenic, inorganics / comparative toxic unit for human (CTUh) 2.565e-07 2.325e-07 2.853e-07 1.11
human toxicity: non-carcinogenic, organics / comparative toxic unit for human (CTUh) 1.851e-08 1.641e-08 2.1e-08 1.13
ionising radiation: human health / human exposure efficiency relative to u235 29.68 27.06 32.71 1.1
land use / soil quality index 173.2 158 190.7 1.1
material resources: metals/minerals / abiotic depletion potential (ADP): elements (ultimate reserves) 7.684e-05 6.689e-05 9.469e-05 1.19
ozone depletion / ozone depletion potential (ODP) 1.514e-06 1.369e-06 1.681e-06 1.11
particulate matter formation / impact on human health 1.567e-06 1.388e-06 1.776e-06 1.13
photochemical oxidant formation: human health / tropospheric ozone concentration increase 0.1366 0.1233 0.1524 1.11
water use / user deprivation potential (deprivation-weighted water consumption) 6.057 5.505 6.665 1.1

6.6 Comparison with other data

Source Scope Value (GWP100 or energy) This dataset Ratio Explanation
[BAFU 2026 dataset] Carbon fiber, weaved, at factory (RER), cradle-to-gate incl. weaving 51.5 kg CO2-eq 62.4 1.21 BAFU analog (Benitez 2021 inventory); includes weaving
[tdc dataset] cf-v2-uhm-ras (same product, other region) 89.7 kg CO2-eq 62.4 0.70 regional variant (electricity, heat, transport)
[tdc dataset] cf-v2-uhm-rna (same product, other region) 65.5 kg CO2-eq 62.4 0.95 regional variant (electricity, heat, transport)
[Benitez et al. 2021] electricity, whole CF chain incl. precursor (per kg CF) 43.2388 kWh 63.5 kWh (this unit process) — different boundary when the precursor is a separate dataset

6.7 Conclusions, limitations and recommendations

GWP100 62.4 kg CO2-eq per kg (95 % interval 56.6–69.5). Dominant contributors: PAN precursor fiber, wet spun 49 %; Electricity, medium voltage, RER, at grid 33 %; Heat, natural gas, at heat plant 10 %; Carbon dioxide, fossil (to air) 5 %; Nitrogen, liquid, at plant 3 %. Most sensitive parameters: Graphitisation (UHM grade): electricity_kWh (per kg step output) (-14 / +0 %); PAN precursor fiber, wet spun (pedigree 95 % interval) (-9 / +11 %); Electricity, medium voltage, RER, at grid (pedigree 95 % interval) (-4 / +4 %). Dataset status: ready for external review.

7 Critical review

Reserved for the external reviewer (ISO 14044 clause 6; ISO/TS 14071 (ISO/TS 14071:2014 Environmental Management — Life Cycle Assessment — Critical Review Processes and Reviewer Competencies// 2014)): reviewer, date, scope of the review, statement, and the response of the dataset developers.//

Reviewer comment Response Change in the dataset

8 References

Analysis of an Existing Method of Producing Carbon Fiber (Stabilization Furnace Energy Efficiency). n.d. https://ijaem.net/issue_dcp/Analysis%20of%20an%20Existing%20Method%20of%20Producing%20Carbon%20Fiber.pdf.

Bell, Caleb. 2016a. Fluids: Fluid Dynamics Component of Chemical Engineering Design Library (ChEDL). Released. https://github.com/CalebBell/fluids.

Bell, Caleb. 2016b. Thermo: Chemical Properties Component of Chemical Engineering Design Library (ChEDL). Released. https://github.com/CalebBell/thermo.

Benitez A., Wulf C., de Palmenaer A., Lengersdorf M., Roding T., and Grube T. 2021. ‘Ecological Assessment of Fuel Cell Electric Vehicles with Special Focus on Type IV Carbon Fiber Hydrogen Tank’. In Journal of Cleaner Production, vol. 278. https://doi.org/10.1016/j.jclepro.2020.123277.

CN102051711B. 2011. Process for producing polyacrylonitrile based carbon fibers. issued 2011. https://patents.google.com/patent/CN102051711B/en.

Commission Recommendation (EU) 2021/2279 of 15 December 2021 on the Use of the Environmental Footprint Methods to Measure and Communicate the Life Cycle Environmental Performance of Products and Organisations, L 471 Official Journal of the European Union (2021).

Energy Consumption of a 100 t/a PAN Carbon Fibre Pilot Line (Per Stage, Measured), ESI of J. Mater. Chem. A. n.d. https://www.rsc.org/suppdata/c6/ta/c6ta06842a/c6ta06842a1.pdf.

European Commission, Joint Research Centre. 2022. Environmental Footprint (EF) Reference Package 3.1. European Commission JRC.

Frischknecht, Rolf. 2023. FOEN LCI Data — Data Quality Requirements (DQR) V2.1. Treeze Ltd. on behalf of the Swiss Federal Office for the Environment (FOEN/BAFU).

ISO 14040:2006 Environmental Management — Life Cycle Assessment — Principles and Framework (Incl. Amd 1:2020). 2006. International Organization for Standardization.

ISO 14044:2006 Environmental Management — Life Cycle Assessment — Requirements and Guidelines (Incl. Amd 1:2017, Amd 2:2020). 2006. International Organization for Standardization.

ISO/TS 14071:2014 Environmental Management — Life Cycle Assessment — Critical Review Processes and Reviewer Competencies. 2014. International Organization for Standardization.

Japan Carbon Fiber Manufacturers Association (JCMA). 2022. LCI of the PAN-Based Carbon Fiber Manufacturing in Japan (2022). https://carbonfiber.gr.jp/assets/pdf/tech/lci2022_EN.pdf.

Li R., Zhang Y., Chu X., and Gan L. 2024. ‘Design and Numerical Study of Induction-Heating Graphitization Furnace Based on Graphene Coils’. In Applied Sciences, 14(6). https://doi.org/10.3390/app14062528.

Mutel, Chris. 2017. ‘Brightway: An Open Source Framework for Life Cycle Assessment’. Journal of Open Source Software 2 (12): 236. https://doi.org/10.21105/joss.00236.

Paul, Edward L., Victor A. Atiemo-Obeng, and Suzanne M. Kresta, eds. 2004. Handbook of Industrial Mixing: Science and Practice. Wiley-Interscience. https://doi.org/10.1002/0471451452.

Prenzel T.M., Hohmann A., Prescher T., et al. 2023. ‘Bringing Light into the Dark - Overview of Environmental Impacts of Carbon Fiber Production and Potential Levers for Reduction’. In Polymers, 16(1). https://doi.org/10.3390/polym16010012.

US5266294A. 1993. Continuous, ultrahigh modulus carbon fiber. issued 1993. https://patents.google.com/patent/US5266294A/en.

Annex A — Energy consumption model

The tdc energy model computes the heat and electricity demand per functional unit from process physics, in two layers, as process simulators do.

Layer 1 — thermodynamic duties of each step. Sensible heat Q = m cp (Tout − Tin); latent heat Q = m ΔHphase; reaction heat Q = ξ ΔHr with ξ the extent of reaction from the stoichiometry and ΔHr from a cited value or from formation enthalpies (Hess’s law, atom-balanced); vessel and line losses UAΔTt or a cited share. Pure-component properties (cp, density, viscosity, ΔHvap, ΔHf) come from cited sources or from the thermo/chemicals property library (Bell 2016b), whose correlation is reported with each value.

Layer 2 — conversion of duties into purchased energy. Heat supply Qpurchased = (Qdemand − ηrecQexotherm)/ηsupply; electricity of each consumer = power × operating time:

  • agitation: P = Np ρ N3D5/ηdrive with Re = ρND2/μ and Np(Re) from the impeller correlations of the Handbook of Industrial Mixing (Paul et al. 2004); the batch volume sets the vessel and impeller size; the specific power of the mixing duty (blending 0.1, heat transfer 0.5, solid suspension 0.8, liquid–liquid 1, gas–liquid 2 kW m⁻³) is a lower bound;
  • pumps: P = ρgHV̇/(ηpump ηmotor), motor efficiency from premium-efficiency tables (Bell 2016a);
  • vacuum: isothermal compression of the vapour/gas load P = ṅRTln (p2/p1)/η (Bell 2016a);
  • cooling: electricity per MJ of heat rejected; installed equipment at part load; plant utilities per tonne, each only with a source.

All expressions are unit-checked with pint. Result per kg: heat, electricity, primary energy demand (with the primary energy factors of the BAFU energy datasets). Parameters of this dataset:

Parameter Value Unit Source / assumption
[Fiber stabilization, carbonization, electrolysis and washing] Fiber sensible heat, stabilization (25 to 300 C) (sensible) {‘m_kg’: 2.8683, ‘cp_kJ_kgK’: 1.5, ‘T_in_C’: 25, ‘T_out_C’: 300} per kg of step output Stabilization 200-300 C in air, 1-2 h residence (Prenzel2023 sec 3.1.1). Precursor mass 2.8683 kg per kg of core fiber = 2.82244 kg per kg product / 0.984 (sourced C-fraction ratio, spec.json). PAN so
[Fiber stabilization, carbonization, electrolysis and washing] Stabilization oven air heating (recirculating oxidizer air) (gas_heating) {‘m_kg’: 156.94, ‘cp_kJ_kgK’: 1.03, ‘T_in_C’: 25, ‘T_out_C’: 250, ‘recovery’: 0.0} per kg of step output ‘the air is heated in an electric heater to a predetermined temperature and heats the processed fibers due to convective heat exchange’ (IJAEM2021); electric carrier confirmed by CN102051711B (heating
[Fiber stabilization, carbonization, electrolysis and washing] Fiber sensible heat, low-temperature carbonization (300 to 1000 C) (sensible) {‘m_kg’: 2.5815, ‘cp_kJ_kgK’: 1.0, ‘T_in_C’: 300, ‘T_out_C’: 1000} per kg of step output Stabilized fiber 2.5815 kg per kg core (2.5402 kg per kg product / 0.984; grade-common 10% stabilization mass loss, precursor 2.82244; DATA GAP - no sourced UHM stage split, same harmonizing loss frac
[Fiber stabilization, carbonization, electrolysis and washing] Fiber sensible heat, high-temperature carbonization (1000 to 1800 C) (sensible) {‘m_kg’: 1.2908, ‘cp_kJ_kgK’: 1.0, ‘T_in_C’: 1000, ‘T_out_C’: 1800} per kg of step output Mid-route fiber 1.2908 kg per kg core (1.2701 kg per kg product / 0.984; grade-common 50% LT-carbonisation loss; DATA GAP - same harmonizing split as IM/HM-rer). 1500-1800 C in the high-temperature gr
[Fiber stabilization, carbonization, electrolysis and washing] Carbonisation + surface-treatment stage powers: N2 purge heating, wall losses, electrolysis, drives (sourced RSC stage kW residual) (duty) {‘MJ’: 28.97} per kg of step output RSC_ESI_2016 Table 1 sourced stage powers. SHARED grade-common physics of the cf-pan-ws-v2 family (same ~29 MJ as IM/HM-rer).
[Fiber stabilization, carbonization, electrolysis and washing] Stabilization furnace NG burner (on-site gas assist, 0.22 kg NG per kg product) (duty) {‘MJ’: 6.6946} per kg of step output Sourced on-site NG for the stabilization oxidation furnace (carrier fuel). Per user decision 4 the on-site fuel is carried by the energy model as BAFU heat per MJ (fuel_supply), not as a per-kg stoich
[Fiber stabilization, carbonization, electrolysis and washing] Inert gas make-up compression (N2, slight overpressure of the carbonisation atmospheres) (compressor) {‘p1_bar’: 1.0, ‘p2_bar’: 1.2, ‘V1_m3_h’: 55.33, ‘gamma’: 1.4, ‘eta’: 0.7} per kg of step output All nitrogen purchased externally (JapanCFMA_LCI_2022 sec 5-3-6: ‘All nitrogen was purchased externally’); inert N2 for carbonisation (Prenzel2023). Gas flow 55.33 Nm3/h make-up per kg/h of core (N2 c
[Fiber stabilization, carbonization, electrolysis and washing] Fiber line tensioner drives, furnace drive trains, electrolysis and washing drives (motor) {‘P_kW’: 3.04, ‘load’: 1.0, ‘motor_eta’: 0.9} per kg of step output Continuous line under constant tension (Prenzel2023 sec 3.1.1) with drive trains per furnace; 3.04 kW per kg/h of core (3 kW per kg/h product / 0.984, line speed 6.5-8.5 m/min, CN102051711B): declared
[Exhaust gas treatment 1] First off-gas treatment (quench/wash/scrubber) specific energy (specific) {‘quantity_per_batch’: 1.0, ‘kWh_per_unit’: 0.023} per kg of step output First exhaust-gas treatment of the continuous CF line (before the RTO); BAFU dataset ‘Carbon fiber production, exhaust gas treatment 1’ has only an electricity input.
[Exhaust gas treatment 2 (regenerative thermal oxidizer)] Regenerative thermal oxidizer NG fuel (on-site, 2.06 kg NG per kg product) (duty) {‘MJ’: 5.9082} per kg of step output Sourced on-site RTO fuel (carrier fuel). Per user decision 4 carried by the energy model fuel_supply as BAFU heat per MJ. This plus the step-1 stabilization NG (0.22 kg) = the on-site fuel combusted i
[Exhaust gas treatment 2 (regenerative thermal oxidizer)] RTO fans and blower (off-gas and oxidising-air circulation) (specific) {‘quantity_per_batch’: 1.0, ‘kWh_per_unit’: 0.0262} per kg of step output RTO fans and post-treatment blower of the regenerative thermal oxidizer; the sourced RSC_ESI_2016 abatement system (25 kW at 12.5 kg/h) includes these fans - independent check. SHARED equipment (same
[Fiber drying and sizing] Sizing-bath water heating (technical drying, recovered closed loop) (loss) {‘MJ’: 1.2} per kg of step output The sizing-bath water is a recovered closed loop: ‘the water evaporates in technical drying, is recovered and fed back; the coating remains on the fiber’ (Prenzel2023 sec 3.1.3). 1.2 MJ per kg product
[Fiber drying and sizing] Drying air handling + winding drives (sourced RSC stage powers) (motor) {‘P_kW’: 1.08, ‘load’: 1.0, ‘motor_eta’: 0.9} per kg of step output RSC_ESI_2016 Table 1 sourced stage powers (measured, 100 t/a pilot line) - independent of the Benitez2021 benchmark.
[Graphitisation (UHM grade)] Graphitisation furnace electric power, UHM grade (graphite-block/induction furnace at 2500-3000 C in Ar) (installed) {‘kW_installed’: 35.0, ‘load’: 1.0} per kg of step output YongXian2026 sourced specific power range (the defining energy sink that makes UHM need more than HM). No UHM graphitisation furnace dimensions, U-value or Ar purge flow are available in the sources (

Annex B — Stoichiometry specification and balances

{
 "product": {
  "name": "Carbon fiber, PAN-based, wet spinning, ultra high modulus (UHM), sized",
  "formula": "C0.082424H0.001984N0.000214O0.000313"
 },
 "inputs": [
  {
   "name": "PAN precursor fiber, wet spun",
   "kg": 2.80201,
   "formula": "C0.056535H0.056548N0.018848",
   "io_type": "raw_material",
   "in_product": true
  },
  {
   "name": "Sizing agent, epoxy",
   "kg": 0.016,
   "formula": "C0.057069H0.085605O0.014267",
   "io_type": "raw_material",
   "in_product": true
  },
  {
   "name": "Oxygen, resource in air",
   "kg": 3.70034,
   "formula": "O0.062504",
   "io_type": "resource",
   "in_product": true
  },
  {
   "name": "Nitrogen, liquid, at plant",
   "kg": 8.822,
   "formula": "N0.071393",
   "io_type": "resource",
   "in_product": false,
   "fate": "to_air"
  },
  {
   "name": "Ammonium bicarbonate, at plant",
   "kg": 0.01842,
   "formula": "N0.012649H0.063246C0.012649O0.037948",
   "io_type": "resource",
   "in_product": false
  },
  {
   "name": "Ethylene glycol, at plant",
   "kg": 0.0076,
   "formula": "C0.032223H0.096668O0.032223",
   "io_type": "resource",
   "in_product": false
  }
 ],
 "outputs": [
  {
   "name": "Carbon dioxide, fossil (to air)",
   "kg": 3.34793,
   "formula": "C0.022722O0.045447",
   "io_type": "emission"
  },
  {
   "name": "Water, off-gas vapor (to air)",
   "kg": 1.42387,
   "formula": "H0.111019O0.055513",
   "io_type": "emission"
  },
  {
   "name": "Nitrogen, off-gas (to air)",
   "kg": 0.737845,
   "formula": "N0.071393",
   "io_type": "emission"
  },
  {
   "name": "Nitrogen oxides, fossil (to air)",
   "kg": 0.00012,
   "formula": "N0.033327O0.033327",
   "io_type": "emission"
  },
  {
   "name": "Hydrogen cyanide (to air)",
   "kg": 0.00191,
   "formula": "C0.036999H0.036999N0.036999",
   "io_type": "emission"
  },
  {
   "name": "Carbon monoxide, fossil (to air)",
   "kg": 0.04726,
   "formula": "C0.035702O0.035702",
   "io_type": "emission"
  },
  {
   "name": "Ammonia (to air)",
   "kg": 0.001357,
   "formula": "N0.058716H0.176149",
   "io_type": "emission"
  },
  {
   "name": "Methane (to air)",
   "kg": 0.001444,
   "formula": "C0.052431H0.209724",
   "io_type": "emission"
  },
  {
   "name": "Ethylene (to air)",
   "kg": 0.001366,
   "formula": "C0.056908H0.113815",
   "io_type": "emission"
  },
  {
   "name": "Ethane (to air)",
   "kg": 0.001314,
   "formula": "C0.054638H0.109275",
   "io_type": "emission"
  }
 ]
}

Annex C — Transport calculation (raw-material provenance)

Provenance of each raw material in the supply of the region: supply_transport (Eurostat Comext / UN Comtrade provenance shares, PEF legs). Legs: default scenario of the PEF method (Commission Recommendation (EU) 2021/2279 of 15 December 2021 on the Use of the Environmental Footprint Methods to Measure and Communicate the Life Cycle Environmental Performance of Products and Organisations 2021) (European suppliers 130 km truck, 240 km train, 270 km barge; overseas suppliers 1000 km truck plus the sea route from the origin’s main port, computed on the marine network).

Raw material HS code Regional share Overseas share Main partners Mean sea distance (km) Source

Annex D — Source excerpts

Every value taken from a source, with the verbatim passage it relies on (also attached as notes to the references in the Zotero library).

Reference Page / section Excerpt Used for
(Prenzel T.M. et al. 2023) Sec 4.2 “Various authors have provided quantification for this mass loss between 40 and 55 wt.% (c.f. Appendix A).” Precursor-to-fiber mass loss; UHM uses the high end (55% loss -> yield 0.45)
(Prenzel T.M. et al. 2023) Sec 4.3 “a plausible average value for the total energy demand for the production of 12 k and 24 k HT-fibers was determined to be 170 MJ of thermal and electric per kg carbon fiber” Total thermal+electric energy per kg CF for HT grade; UHM is higher due to graphitisation
(Prenzel T.M. et al. 2023) Sec 4.4 “waste air purification in the manufacturing process is indispensable. A common practice is treatment by proprietary regenerative thermal oxidizers, offering more than 95% efficiency [55]” Off-gas treatment: regenerative thermal oxidiser destruction efficiency >95%
(Prenzel T.M. et al. 2023) Sec 3.1.1 “the precursor is cautiously exposed to temperatures around 200 to 300 °C under air in a continuous oxidizing furnace for one to two hours while being under constant tension” Stabilization/oxidation: 200-300 C, air, 1-2 h
(Prenzel T.M. et al. 2023) Sec 3.1.2 “The first stage is performed at about 1000 °C in a low-temperature furnace… Subsequently, the fibers are heated up until they reach 1500 to 1800 °C in an ensuing high-temperature graphite furnace” Carbonization: two steps 1000 C (tar removal) then 1500-1800 C
(Prenzel T.M. et al. 2023) Sec 3.1.3 “For sizing, the fiber is drawn through a sizing bath in which the sizing agent and water is absorbed. The water then evaporates during technical drying, is recovered, and fed back into the cycle, while the coating remains on the fiber” Process water: the sizing-bath water is a recovered closed loop (no net water input for the UHM heat-treatment stage); supports the water DATA GAP
(Japan Carbon Fiber Manufacturers Association (JCMA) 2022) Sec 5-3-6 “All nitrogen was purchased externally. Unit data of the electricity and steam consumption for nitrogen was determined using the unit data based on the production performance of industrial gas manufacturers (background data).” Transport: the BAFU CF analog datasets carry large lorry/rail transport for their purchased nitrogen and natural gas; our dataset’s only sourced purchased material is the sizing agent (precursor fiber is on-site; N2 and RTO fuel are documented DATA GAPS)
(US5266294A 1993) Claims/Specification (graphitization) “The heat treatment will be conducted in a substantially non-reactive atmosphere to ensure that the fiber is not consumed. The non-reactive atmosphere may be nitrogen, argon or helium, however for temperatures above about 2000° C., argon and helium are preferred.” Water: the UHM heat-treatment stages (carbonization, graphitisation) run in a dry inert (non-reactive) atmosphere with no aqueous reagent; no net process-water input
(Prenzel T.M. et al. 2023) Sec 3.1.5 / 4 “a plausible average value for the total energy demand for the production of 12 k and 24 k HT-fibers was determined to be 170 MJ of thermal and electric per kg carbon fiber” Literature energy reference: 170 MJ/kg (thermal+electric, HT grade) for the verify-stage energy cross-check
(Energy Consumption of a 100 t/a PAN Carbon Fibre Pilot Line (Per Stage, Measured), ESI of J. Mater. Chem. A, n.d.) ESI Table 1 “The energy consumption of our 100 tons per year pilot line displayed on Figure 1 was measured by the Carbon Nexus production facility” Stage powers per kg product (12.5 kg/h, 30x24k spools, 120 m/h): 134 kW stabilisation, 15 kW low-T carbonisation, 45 kW high-T, 25 kW abatement, 5 kW material handling, 43 kW surface/handling/drying, 10 kW winding; total 277 kW = 79.8 MJ/kg
(Analysis of an Existing Method of Producing Carbon Fiber (Stabilization Furnace Energy Efficiency), n.d.) Section I “Over 93% of the total energy consumed spent on heating the air and only about 3% on heating the processed material” Stabilization oven energy split: back-derive recirculating air flow from the sourced 134 kW stage power (0.93 share) -> 35.9 MJ/kg air duty
(Japan Carbon Fiber Manufacturers Association (JCMA) 2022) sec 5-5 “The total primary energy consumption of all raw materials and fuels (electricity, steam, and directly consumed fuels) input for the production of 1 kg of CF, calculated back to the mining and extraction of resources, is 318.2 MJ. In addition, there is 32.1 MJ of feedstock energy, which together makes a total of 350.2 MJ.” Full cradle-to-gate primary energy incl. the PAN precursor stage -> system-boundary reference for the energy-divergence justification
[YongXian2026] Energy Cost in Carbon Fiber Production “High modulus carbon fibre (380-450 GPa) requires an additional graphitization step at 2000-2800 C, which adds 15-35 kWh per kilogram to the energy budget” Graphitisation furnace power (UHM-specific): 35 kWh/kg high end (UHM >2500 C in Ar, US5266294A); resolves the DATA GAP that kept UHM energy below HM
(Benitez A. et al. 2021) workbook lci-carbon-fiber.xlsx, sub-process table “Per-step LCI inventory (workbook lci-carbon-fiber.xlsx) of Toray T700G PAN-based carbon fibre production, based on Ellringmann 2016, Gill 2016, Muetzel 2012. The carbon-fibre heat-treatment chain is split into per-sub-process inventories (per kg of each sub-process’s own output).” Per-sub-process benchmark: stabilization/carbonization/electrolysis/washing 28.23 kWh + 4.65 MJ steam per kg output; exhaust gas treatment 1 0.0229 kWh per kg off-gas; exhaust gas treatment 2 0.0262 kWh + 0.1908 kg natural gas per kg off-gas; drying and sizing 0.6257 kWh + 0.3351 MJ steam per kg product
(Li R. et al. 2024) sec 1 “Graphitization requires heating temperatures typically up to 2800 C or higher, which is an energy-intensive process requiring 3-8 MWh per ton of graphite product produced” Independent graphitisation energy: 3-8 kWh per kg of graphitised product (induction-furnace family, same equipment as UHM fiber graphitisation, US5266294A); supports the 35 kWh/kg (per kg of step output) upper bound chosen from YongXian2026
(CN102051711B 2011) “Said pre-oxidation process adopts electric furnace heating wire and far infrared assistant heating” Carrier of the main heat-treatment step: the stabilization oven (electric heating wire + far infrared) and the carbonisation furnaces (low-T heat-resistant electric furnace heating wire below 1000 C; high-T electric resistance furnace) are all-electric in this UHM line; basis of the cited family-step difference vs the NG gas-assist of the lower grades
[benitez_2021_carbon_fiber_lci] “carbon fiber production, fiber winding and unwinding / carbon fiber, unwound / 2.73691” Precursor anchor of the common stabilisation + carbonisation path: 2.73691 kg unwound PAN precursor per kg unwound CF (Benitez 2021 T700G, LCI workbook) - basis of the UHM precursor ratio 2.73691 x (0.99/0.96) = 2.82244 kg per kg sized UHM fiber
[benitez_2021_carbon_fiber_lci] “carbon fiber production, fiber stabilization, carbonization, electrolysis and washing / carbon fiber, washed / 1.092 / 28.226376 / 4.650902 / 0.198342 / 8.07885” Inert gas quantity (user decision 3): N2 8.07885 kg per kg of this step’s output x 1.092 kg per kg product = 8.822 kg N2 per kg product, bound to BAFU ‘Nitrogen, liquid, at plant’ - same sourced quantity as the verified IM/HM-rer
[groetsch_2021_gas_emission_carbon_fiber] “The carbon content continued to increase to 96% in the HT carbonization phase, where all other elements except for 3% N were removed.” UHM precursor ratio: C mass fraction of the carbonised (pre-graphitisation) PAN fibre = 0.96, denominator of the graphitisation mass-loss factor 0.99/0.96
[toray_m40j_data_sheet] “Chemical Composition: Carbon >99%, Na+K <50 ppm.” UHM precursor ratio: C mass fraction of the graphitised high-modulus fibre >= 0.99 (Toray M40J data sheet, 377 GPa), numerator of the graphitisation mass-loss factor 0.99/0.96
[harmonizing_lca_cf_notes] “Natural Gas: Stabilization 0.22 kg/kg CF (0.34 m3). Exhaust treatment 2 2.06 kg/kg CF (0.29 m3)” On-site furnace/RTO fuel (user decision 4): 0.22 kg NG per kg CF stabilization gas assist + 2.06 kg NG per kg CF RTO fuel, grade-common shared equipment; carried as BAFU heat per MJ (fuel_supply ‘Heat, natural gas, at heat plant’), never a per-kg stoich input
(US5266294A 1993) “The non-reactive atmosphere may be nitrogen, argon or helium, however for temperatures above about 2000° C., argon and helium are preferred.” Inert gas (user decision 3): graphitisation at >2000 C runs in argon (sourced atmosphere; UHM graphitises >2500 C). No sourced Ar make-up quantity in the sources after queries ‘ultra high modulus carbon fiber graphitization argon flow rate Nm3/kg’ and ‘carbon fiber graphitization argon make-up flow L/min furnace purge continuous’ -> Ar quantity = DATA GAP; the sourced N2 quantity (8.822 kg/kg, Benitez) carries the stoich inert-gas input

Annex E — Independent second-model review

# Severity Exchange Finding Answer / change
1 medium Electricity, medium voltage, RER, at grid The ratio of model electricity (63.53 kWh) to the RSC_ESI_2016 pilot line benchmark (22.16 kWh) is 2.87, which is outside the 0.5-2 range. While the energy model justifies this via the addition of the UHM graphitisation stage (35 kWh/kg), the discrepancy is significant. Ratio 2.87 = UHM total 63.53 kWh/kg vs the RSC_ESI_2016 STANDARD-modulus 100 t/a pilot line (22.16 kWh/kg, all-electric, 2016, no graphitisation). Subtracting the grade-specific UHM graphitisation stage (35 kWh/kg, the CF-specific sourced range 15-35 kWh/kg of YongXian2026 at its >2500 C UHM high en
2 medium Nitrogen, liquid, at plant There is a documented DATA GAP regarding the Argon (Ar) make-up required for the graphitisation stage (>2500 C), as per US5266294A. The model currently uses the N2 quantity as a proxy. The 8.822 kg/kg ‘Nitrogen, liquid, at plant’ is the sourced inert-gas make-up of the whole heat-treatment chain, not a graphitisation-only gas: Benitez 2021 LCI gives 8.07885 kg N2 per kg of step output x 1.092 kg step output/kg product (BAFU 2026 carbon-fiber analog basis, identical to the verified
3 low Water, off-gas vapor (to air) Completeness: The model lacks a direct water input (e.g., deionised water for electrolysis/washing), though it justifies this as a ‘DATA GAP’ because the sizing bath is a closed loop. However, electrolysis and surface treatment typically require aqueous reagents.

Annex F — Cross-check with independent data

Independent analogs (BAFU): [‘Carbon fiber production, fiber stabilization, carbonization, electrolysis and washing’, ‘Carbon fiber production, fiber drying and sizing’, ‘Carbon fiber production, exhaust gas treatment 1’, ‘Carbon fiber production, exhaust gas treatment 2’]; tolerance: factor 3.0

exchange / class this dataset (per kg) independent value(s) source status
electricity 63.53 0.02291 - 28.23 Carbon fiber production, fiber stabilization, carbonization, electrolysis and washing, Carbon fiber production, fiber drying and sizing, Carbon fiber production, exhaust gas treatment 1, Carbon fiber production, exhaust gas treatment 2 ok
transport 0.00208 0.007105 - 1.26 Carbon fiber production, fiber stabilization, carbonization, electrolysis and washing, Carbon fiber production, fiber drying and sizing, Carbon fiber production, exhaust gas treatment 2 FLAG
water 0 8.114e-05 - 0.0002505 Carbon fiber production, fiber stabilization, carbonization, electrolysis and washing, Carbon fiber production, fiber drying and sizing FLAG
Electricity, medium voltage, RER, at grid 63.52914 22.16 kWh RSC_ESI_2016 Table 1: 100 t/a pilot line total power 277 kW ok
Wastewater treatment, class 3 2.6e-05 2.6e-05 m3 Analyst-added waste-disposal input (not a stoich/energy flow ok
electricity vs energy.json 63.53 63.53 kWh tdc energy model ok

8.0.1 Justified deviations (with evidence)

  • off-gas: The toxic off-gas from stabilization/carbonization (HCN/CO/NH3/C2Hx) is NOT released: it is fully combusted in a regenerative thermal oxidiser (>95% destruction, Prenzel 2023 sec 4.4). What leaves to air are the RTO combustion products (CO2 fossil, H2O, N2, residual NOx), all modeled in the stoichiometry. The RTO support fuel (natural gas) is a documented DATA GAP. [evidence: Prenzel2023]
  • transport: The BAFU CF analogs carry lorry/rail transport for their large purchased inputs (inert nitrogen 8 kg/kg, RTO natural gas 0.2 kg/kg, deionised water, epoxy). In this dataset the only purchased raw material is the epoxy sizing agent (0.016 kg/kg -> PEF 130 km lorry leg = 0.00208 tkm). The PAN precursor fiber is delivered on-site (no transport, per the shared-precursor rule) and the inert N2 / RTO fuel are documented DATA GAPS. The lower transport is a genuine, sourced difference (fewer purchased bulk materials), not a missing leg. [evidence: JapanCFMA_LCI_2022]
  • water: The BAFU CF analogs include a small deionised-water input (electrolysis/surface-treatment and washing). In the UHM heat-treatment scope, the sizing-bath water is a recovered closed loop (Prenzel 2023 sec 3.1.3: water evaporates in technical drying, is recovered and fed back; coating remains) and the inert-atmosphere heat-treatment (carbonization/graphitisation at 3000-3500 C, US5266294A) has no aqueous reagent. No net purchased water input -> DATA GAP, not an omitted flow. [evidence: US5266294A]
  • energy-divergence Fiber drying and sizing: The benchmark’s 0.335 MJ steam for drying/sizing is its thermal load; in this line the sizing-bath water heating and drying-air handling are electrically heated (sizing bath, CN102051711B) so the model’s heat_MJ = 0 and the load sits in electricity. The whole-step comparisons agree: total_MJ 5.73 vs 2.59 (x2.21) and electricity 1.59 vs 0.63 kWh (x2.55), both within factor 3. Only the heat-vs-steam form differs because of the carrier choice. [evidence: RSC_ESI_2016]
  • energy-divergence Graphitisation (UHM grade): Li2024’s 3-8 MWh/tonne (3-8 kWh/kg) is for BULK graphite product (powder/felt/foam) graphitised in an induction furnace that itself runs at ~15.8% efficiency (coil cooling water = 70% of total loss), a shorter-residence denser charge. The UHM continuous fiber graphitises at 2500-3000 C with a 1-2 h hold at the TOP of the temperature range (US5266294A), the energy-intensive case the CF-specific YongXian2026 range (15-35 kWh/kg) covers; the model uses its high end (35 kWh/kg of step output). The Li2024 value is a low-end graphite-product equipment-family floor, not a CF-fiber figure, so the 4.4x ratio is a product-basis and temperature-range difference, not a missing/over modelled duty. [evidence: YongXian2026]
  • energy-divergence: Model total 297.5 MJ/kg (electricity 63.53 kWh = 228.7 MJ + heat 68.77 MJ; upstream steps x0.992761 for the 0.967 carbon-fraction decision, 2026-10-28). Vs the 3 independent literature refs: Prenzel2023 170 MJ/kg HT-grade average no-graphitisation (x1.75, within factor 3); RSC_ESI_2016 79.8 MJ/kg measured STANDARD-modulus pilot-line total (x3.73, above factor 3 - explained below); JapanCFMA_LCI_2022 350.2 MJ/kg full cradle-to-gate incl. the PAN precursor stage (x0.85, within factor 3). The RSC x3.73 divergence is grade-specific, not an over-modelled duty: the UHM dataset adds (a) the >2500 C argon graphitisation stage 35 kWh/kg = 126 MJ/kg (absent from the standard-modulus pilot line, YongXian2026) and (b) the on-site NG fuel for the RTO + stabilisation assist = 68.8 MJ/kg (a combustion load the all-electric pilot-line subtotal omits). Subtracting those two grade-specific terms gives ~103 MJ/kg, within x1.3 of the RSC 79.8 MJ/kg grade-common path. [evidence: YongXian2026]
  • energy-divergence Exhaust gas treatment 2 (regenerative thermal oxidizer): Electricity x0.139: the model RTO item is the RTO-only fan/blower (0.0262 kWh per kg treated off-gas, the Benitez2021 published value) whereas the independent RSC_ESI_2016 abatement measurement (0.1885 kWh per kg) covers the WHOLE abatement train (first treatment + RTO + post-treatment blower), so the RTO-only value is lower by design. Heat 5.9082 MJ per kg off-gas = BAFU-published RTO support NG (grade-common shared equipment 2.06 kg NG per kg CF x LHV 35.8 x 0.85), the same published basis as the verified SM/IM/HM grades. No physics is missing. [evidence: RSC_ESI_2016]
  • energy-divergence Fiber stabilization, carbonization, electrolysis and washing: The benchmark’s 4.65 MJ steam heat for this combined step is its thermal load; in this UHM line the same load is supplied electrically (pre-oxidation heating wire + far-infrared, high-T graphite heater, CN102051711B) plus a 0.22 kg/kg on-site NG gas-assist duty carried as BAFU heat per user decision 4. Whole-step comparisons agree: total_MJ 101.9 vs 106.3 (x0.959) and electricity 26.46 vs 28.23 kWh (x0.937), both within factor 3. Only the heat-vs-steam form differs because of the carrier choice, not the total. [evidence: Benitez2021]
  • family-step Fiber stabilization, carbonization, electrolysis and washing: After the 2026-10-28 0.967 carbon-fraction decision, the UHM main heat-treatment step = 99.58 MJ per kg product (upstream steps x0.992761, precursor 2.80201 kg/kg, now equal to HM). Family per-kg-product step totals after the 0.967 round: HM 102.43 / IM 100.63 / UHM 99.58 MJ/kg - spread x1.03, well within tolerance. The per-step-output intensity is identical across the family (101.9 MJ and 26.46 kWh per kg of step output; all-electric furnaces per CN102051711B plus 0.22 kg/kg on-site NG gas assist carried as BAFU heat per user decision 4), so the per-kg-product ordering simply mirrors the kg_per_kg_product scaling (HM 1.004854, IM 0.984, UHM 0.976877) prescribed by each grade’s precursor ratio. UHM being the HIGHEST at this step was an artefact of the rounded 0.96 carbon fraction, which overstated UHM’s precursor ratio (2.82244); with the sourced 0.967 (Groetsch 2021) UHM = HM = 2.80201. UHM’s WHOLE-PROCESS energy remains the highest of the family (297.5 vs HM 248.4 / IM 173.0 / SM 60.5 MJ/kg) because of the grade-specific 35 kWh/kg graphitisation. Step vs Benitez 2021 benchmark: 101.9 vs 106.3 MJ per kg step output (x0.959), electricity 26.46 vs 28.23 kWh (x0.937). [evidence: CN102051711B]
  • energy-literature-based: Dominant thermal terms are sourced specific/stage powers, not UAdT: graphitisation furnace 35 kWh/kg (YongXian2026 ‘adds 15-35 kWh per kilogram’, UHM >2500 C high end), the N2-purge + wall-loss + electrolysis + drives residual (28.97 MJ, sourced RSC_ESI_2016 stage kW less the computed fiber-sensible share), and the on-site NG fuel duties (stabilisation 0.22 kg + RTO 2.06 kg, harmonizing_lca_cf_notes, grade-common shared equipment, carried as BAFU heat per user decision 4). First-principles physics for the UHM furnace/Ar purge is UNAVAILABLE: no UHM graphitisation furnace dimensions/U-value/Ar purge flow, no N2 purge flow per kg (US20150210925A1 gives only a 1-10 Nm3/h minimum per tow), no furnace wall q_W_m2. The fully physics-computed terms (stabilisation oven air back-derived from the sourced 134 kW + IJAEM2021 93%/3% split; fiber sensible heats) are ~40% of the step-1 energy. Queries tried: ‘UHM carbon fiber graphitization furnace dimensions m2’, ‘ultra high modulus carbon fiber graphitization argon flow rate Nm3/kg’, ‘carbon fiber carbonization furnace nitrogen gas flow rate Nm3 per kg fiber’. [evidence: YongXian2026]

Open flags: none

Annex G — Files, tools and audit trail

Item File / tool Version / date
spec.json /home/achile/tdc_datasets/cf-v2-uhm-rer/spec.json 2026-10-28
energy.json /home/achile/tdc_datasets/cf-v2-uhm-rer/energy.json 2026-10-28
energy_result.json /home/achile/tdc_datasets/cf-v2-uhm-rer/energy_result.json 2026-10-28
snippets.json /home/achile/tdc_datasets/cf-v2-uhm-rer/snippets.json 2026-10-27
board.json /home/achile/tdc_datasets/cf-v2-uhm-rer/board.json 2026-10-28
review.json /home/achile/tdc_datasets/cf-v2-uhm-rer/review.json 2026-10-28
lcia.json /home/achile/tdc_datasets/cf-v2-uhm-rer/lcia.json 2026-10-28
cf-v2-uhm-rer.ecospold /home/achile/tdc_datasets/cf-v2-uhm-rer/cf-v2-uhm-rer.ecospold 2026-10-28
tdc-doc /home/achile/tdc/scripts/tdc_doc.py 2026-10-28

Annex H — BAFU DQR v2.1 documentation fields (EcoSpold v1 metadata)

DQR field Content
name Carbon fiber, PAN-based, wet spinning, UHM (RER)
location RER
unit kg
time period 2024-2026
geography Europe
technology Conversion of wet-spun PAN precursor fiber to ultra-high modulus (UHM) PAN-based carbon fiber: oxidation/stabilization (~250-300 C), low-temperature carbonization (~350-400 C), high-temperature carbonization, graphitisation (~2500-3000 C), surface treatment and sizing. From precursor fiber of the re
included processes Fiber stabilization, carbonization, electrolysis and washing; Exhaust gas treatment 1; Exhaust gas treatment 2 (regenerative thermal oxidizer); Fiber drying and sizing; Graphitisation (UHM grade)
representativeness Section 3
sources 16 references (Annex D, References)
data generator tdc pipeline
validation tdc-doc final control: ready for external review

To comment on this report, select a passage and use the annotation panel on the right (registered users; comments are published once the BAFUquery team has reviewed them). To suggest a dataset that does not exist yet, use Propose a dataset.

datasets/tdc/5c838461-ad82-5ce1-ae17-4b14915023ca.1794866394.txt.gz · Last modified: by 127.0.0.1

Donate Powered by PHP Valid HTML5 Valid CSS Driven by DokuWiki