Method, assumptions and ranges

A.1 Evidence status

Each figure in the paper carries one of four statuses:

• Sourced: taken directly from a cited document or dataset (Annex C).
• Derived: calculated from sourced or assumed inputs by the method in this annex.
• Assumed: a planning value chosen within a stated range, where no project-specific or published value is available.
• Illustrative: arithmetic that shows a mechanism or an order of magnitude; not a forecast or a commitment.

RC-1 project values (capacities, units, layout reservations, indicative schedule) are planning values from the RC-1 reference configuration [S90] and are labelled planned or indicative.

A.2 The footprint model

For each phase, with finished output P (t/yr):

• Liquid steel L = P / y_r, where y_r is the cast-and-roll yield (0.94).
• Metallic charge M = L / y_m, where y_m is the metallic yield (0.88).
• Purchased DRI/HBI D = M × s, where s is the DRI share of the charge (0.30); scrap = M − D.
• Electricity E (kWh) = L × (e_EAF + e_aux) + P × e_roll + P × e_ind (induction heating in Phase 1 only).
• Gas G (GJ) = P × g_rh for Phases 2 and 3 (gas reheating); zero in Phase 1.
• Scope 1 = L × f_EAF + G × 0.0561 t CO₂/GJ.
• Scope 2 = E × grid factor.
• Upstream iron units = D × f_DRI.
• Intensities divide by P (per tonne of product) or by L (per tonne of crude steel).
• CBAM basis = (Scope 1 + upstream iron units) per tonne, excluding electricity, consistent with Annex II of the CBAM Regulation for iron and steel [S27].

The full build-out combines Phase 1 (300,000 t with induction heating) with 700,000 t of gas-reheated output. The model reproduces the balances in the RC-1 reference configuration to the tonne.

Table A.1 Parameters, central values and ranges

Source: Cameroon Steel reference configuration RC-1 (planning estimate) [S90] and sources as listed.

A.3 Sensitivity analysis

The tornado (Figure 4.4) changes one parameter at a time across its range, holding the others at central values, and reports the change in the Phase 1 intensity including upstream iron units and in Scopes 1 and 2. The DRI-share case adds the furnace electricity penalty for DRI relative to the 30% base. The joint-extreme case in Table 4.5 sets the DRI share, upstream intensity and grid factor simultaneously to their low (20%, 0.5 t/t, 200 g/kWh) and high (50%, 0.9 t/t, 300 g/kWh) values. The heat map (Figure 4.5) varies the DRI share from 10% to 60% and the upstream intensity from 0.1 to 1.0 t/t at a grid factor of 226 g/kWh.

A.4 The 2025 grid cross-check

No 2025 grid intensity was published on the Ember basis at the time of writing. The cross-check in Chapter 3 proceeds in three steps:

1. Total 2024 emissions = 225.9 g/kWh × 7.26 TWh = about 1,640 kt CO₂e [S48][S49].
2. Subtracting renewable generation at IPCC life-cycle medians (hydro 24, solar 48, bioenergy 230 g/kWh [S58]) leaves about 1,504 kt for 1.95 TWh of gas and oil generation, an implied fossil rate of about 771 g/kWh.
3. Applying that rate to ARSEL’s 2025 fossil generation (559.7 GWh of gas, 153.9 GWh of light fuel oil, 66.7 GWh of heavy fuel oil) gives about 602 kt. Divided by 2025 generation of 8,119.9 GWh [S50], this is about 74 g/kWh on a combustion basis; adding hydro and solar at their life-cycle values gives about 96 g/kWh.

The cross-check assumes the 2024 fossil fleet performance applies in 2025 and that the Ember and ARSEL datasets have comparable coverage; neither is certain. It is used only as an indication that the published 2024 value is conservative against the average; it says nothing about the marginal factor (Chapter 3).

A.5 Carbon border arithmetic

The illustrative CBAM charge per tonne in Chapter 5 is:

charge = max(0, E − B × F_t) × price

where E is the embedded emissions used (t CO₂/t) on the CBAM basis for steel, that is direct emissions plus the direct emissions of precursors, without purchased electricity [S27]; B the route benchmark (1.370, 0.481 and 0.072 t/t for blast-furnace, DRI-EAF and scrap-EAF routes, as reported for hot-rolled coil [S31] and used here as a proxy for long products, whose values in Implementing Regulation (EU) 2025/2620 [S104] were not extracted); F_t the CBAM factor for year t (97.5% in 2026 falling to 0% in 2034 [S28]); and the price a constant €82.32/t CO₂ (third quarter of 2026 [S26]). For RC-1, E is the CBAM basis defined above, taking the DRI’s full upstream factor as its embedded direct emissions (a conservative simplification), and the scrap-EAF benchmark applies because more than half of its crude steel derives from scrap [S93]. The all-scrap case is the RC-1 model with no DRI, so that E equals Scope 1. For the ore-based routes, E is a range: the lower end is the IEA’s direct-emission estimate (1.0 t for gas-based DRI-EAF, 1.2 t for BF-BOF [S05]), which omits purchased precursors; the upper end is the worldsteel 2024 route average (1.47 and 2.34 t [S01]), which also counts electricity and other upstream inputs. Each 10 points of DRI in the RC-1 charge add 1 / (0.94 × 0.88) × 0.10 × 0.70 = 0.085 t per tonne of product. The arithmetic ignores any carbon price paid at origin, transport and exchange-rate effects.

A.6 Hydrogen and green-iron arithmetic

Electricity for hydrogen-based reduction = DRI output × 54 kg H2 per tonne of DRI [S17] × 45 to 52 MWh per tonne of H2 [S05]. For RC-1’s full build-out need of about 362,700 t of DRI a year this gives about 0.88 to 1.02 TWh; for a 1 Mt/yr module, about 2.43 to 2.81 TWh. The electricity used by the direct-reduction plant itself is excluded. Gas for a 1 Mt/yr gas-based module = 9.2 to 10 GJ per tonne [S16][S15], compared with 2025 national gas production of 75.5 billion cubic feet [S83], converted at about 1.09 GJ per thousand cubic feet.

The hydrogen-based iron intensity of about 0.2 t CO₂ per tonne of DRI used in Chapters 4, 6, 8 and 10 is derived from Energiron’s stated reductions relative to the blast-furnace route (53% with natural gas, 86% with hydrogen [S15]): the hydrogen route emits about 0.14 / 0.47 ≈ 30% of the gas route, and 30% of 0.7 t is about 0.2 t. The derivation is approximate because the stated reductions refer to the whole DRI-EAF route.

A.7 Pathway arithmetic

The pathway in Chapter 10 applies the following parameter changes cumulatively to the full build-out, recomputing the model at each step: furnace electricity 450 to 420 kWh/t, auxiliaries 70 to 60 kWh/t and rolling 110 to 95 kWh/t; furnace direct factor 0.10 to 0.08 t/t; gas reheating 1.3 to 1.15 GJ/t; grid factor 226 to 100 g/kWh; DRI share 30% to 25% (with the furnace electricity penalty); upstream DRI intensity 0.7 to 0.5 t/t; and finally 0.5 to 0.2 t/t for hydrogen-based iron units. All values except the last lie within the ranges of the RC-1 reference configuration or the 2025 grid cross-check.

A.8 Other derivations

• Residues: slag 140 kg, dust 15 kg per tonne of liquid steel and mill scale 20 kg per tonne of product [S90]. The slag value sits within the IFC benchmark for solid residue, 110 to 180 kg per tonne of product, mostly slag [S20]; the dust value sits within a planning range of 10 to 20 kg (the European reference document reports 10 to 30 kg) [S19]. Zinc in dust at 20% to 30% Zn.
• Standards thresholds at RC-1’s scrap share: IEA near-zero = 0.40 − 0.35 × scrap share [S06]; ResponsibleSteel Progress Level 1 interpolated linearly between 2.34 t at 20% scrap and 0.96 t at 80% scrap [S37].
• Minor sources: liquid steel × 0.01 to 0.02 t CO₂/t for diesel and LPG, plus an SF6 inventory of 0.5 to 2 t × 0.5% a year × 23,500 [S105]; about 3,250 to 6,620 t CO₂e a year in Phase 1 (central 4,934 t).
• Lender threshold crossing: the grid factor at which Phase 1 Scopes 1 and 2 reach 100,000 t = (100,000 − Scope 1) / Phase 1 electricity = about 236 g/kWh on the RC-1 model, or (100,000 − 31,915 − 4,934) / 288,957 MWh = about 219 g/kWh with the central minor sources.
• Furnace electricity at the upper planning value: +40 kWh/t × 319,149 t of liquid steel = about 12,800 MWh a year in Phase 1, or about 2,900 t CO₂ at 226 g/kWh, +0.010 t per tonne of product.
• Billet basis: per tonne of cast steel, Scope 1 = 0.10 t; melt-shop Scope 2 = 520 kWh × 226 g/kWh = 0.118 t; upstream DRI = 0.30 / 0.88 × 0.70 = 0.239 t.
• Rainfall uplift: station return values × 1.07 per additional degree of global warming [S59].
• Currency: where US dollar equivalents are given, a planning rate of 600 FCFA per US dollar is used [S90]; the FCFA is pegged to the euro at 655.957.

Table A.2 Key derived values used in the text

Source: WP06 model on the Cameroon Steel reference configuration RC-1 (planning estimate) [S90].

A.9 Limitations

The model is a steady-state balance at nameplate output. It does not represent ramp-up, part-load operation, product-mix effects on rolling energy, seasonal grid variation, the emissions of purchased inputs other than iron units, transport, or construction. Benchmarks are drawn from international guidelines and planning assumptions, not from supplier guarantees. All results should be read with the ranges shown and will be replaced by engineering and then measured data (Chapter 11).