[bobbyxng]

Changes proposed in this Pull Request

The nuclear/uranium fuel price from PR #259 was about 3x too high. The TYNDP 2024 nuclear fuel price is based on EIA fuel expenses per MWh of electricity generated, but it is used here as a price per MWh of fuel, so the efficiency gets applied twice (effectively about 3x too high). It's a unit mix-up in the TYNDP 2024 price data (ENTSO-E/ENTSOG): an electricity-based EIA cost is listed as a fuel price.
@fneum @euronion

Old value (TYNDP 2024 Prices.zip, sheet IEA - References, cell E24):

AVERAGE(EIA nuclear fuel 2011–2021) × 0.84 EUR/USD ÷ 3.6 GJ/MWh
= 7.20 USD/MWh_el × 0.84 ÷ 3.6
= 1.68 EUR/GJ

EIA Table 8.4 is in mills per kWh of electricity generated, so this is not a fuel price per GJ of fuel.

New value (Lazard LCOE+ 2026, pp. 7 and 28; same value as the 2023 edition, p. 49):

0.85 USD/MMBtu × 3.412 MMBtu/MWh ÷ 1.16 USD/EUR
= 2.50 EUR2026/MWh_th

1.16 USD/EUR is the ECB average for Jan–Sep 2026.

This is also in line with the earlier TYNDP cycles (2020/2022): the same workbook still lists their nuclear fuel price of 0.47 EUR/GJ (≈ 1.7 EUR/MWh_th) in sheet Final Fuel Prices 2022.

Effect on nuclear's marginal cost (VOM 4.46 EUR/MWh_el, efficiency 0.326):

	fuel price	marginal cost
before	7.45 EUR/MWh_th	4.46 + 7.45 ÷ 0.326 ≈ 27.3 EUR/MWh_el
after	2.44 EUR/MWh_th	4.46 + 2.44 ÷ 0.326 ≈ 12.0 EUR/MWh_el

Only the four nuclear/uranium fuel rows change; the outputs are regenerated.

Checklist

• Code changes are sufficiently documented; i.e. new functions contain docstrings and further explanations may be given in doc.
• Data source for new technologies is clearly stated.
• Newly introduced dependencies are added to environment.yaml (if applicable).
• A note for the release notes doc/release_notes.rst of the upcoming release is included.
• I consent to the release of this PR's code under the GPLv3 license.

[fneum]

LGTM.

Could you remove "Replaces the TYNDP 2024 value of 1.6808 EUR/GJ, which the TYNDP workbook derives in sheet IEA - References, cell E24, as AVERAGE(EIA Table 8.4 nuclear fuel expenses 2011-2021) * 0.84 EUR/USD / 3.6 GJ/MWh; EIA Table 8.4 is in mills per kWh weighted by net generation, i.e. USD per MWh of electricity, so using the result as a thermal fuel price applied the plant efficiency twice (about 3x too high)."? It's not necessary to document changes in the data.

[Irieo]

Right, I followed same calculation here. Good it is fixed upstream!

[euronion]

Nice!

• What @fneum said, no need to describe the data changes in the description column or the release notes - shorter is better.
• The efficiency value 3.412142 MMBtu/MWh_th is not from Lazard, right? Add where you are taking it from
• Currency year: Debatable whether 2023 or 2026 is to be used. I'd be leaning towards 2023 (earlier appearance of the number)

Just for my understanding - did ENTSO-X get it wrong in the TYNDP, or did we interpret the number incorrectly?

[bobbyxng]

Thanks @fneum @euronion @Irieo

Regarding the questions/comments:

• I shortened the release notes
• 3.412142 MMBtu/MWh_th isn't from Lazard, it's the unit conversion 3.6 GJ/MWh ÷ 1.055056 GJ/MMBtu, using the International Table Btu (1 Btu_IT = 1055.056 J), see NIST SP 811, Appendix B.9 (https://www.nist.gov/pml/special-publication-811/nist-guide-si-appendix-b-conversion-factors/nist-guide-si-appendix-b9).
• switched to currency year 2023. The rows now cite Lazard v16.0 (2023), p. 49, like the other nuclear rows, converted at 1.09 USD/EUR --> 0.85 × 3.412142 / 1.09 = 2.66 EUR/MWh_th.
• The error is in the TYNDP 2024 inputs, so on ENTSO-X side, The workbook (sheet IEA - References, cell E24) takes EIA Table 8.4 nuclear fuel expenses, which are per kWh of electricity generated, and divides by 3.6 GJ/MWh as if they were per unit of fuel. We used the value as given, so the plant efficiency was applied twice. The earlier TYNDP cycles (0.47 EUR/GJ, still in the same workbook) are in line with Lazard.

[euronion]

Great! 2023 is also better, because the inflation data is already in the past and no longer developing.