Data-driven long-term wind speed forecasting and techno-economics of offshore wind-to-hydrogen production

Prihandono Aditama; Abdul Wasy Zia

doi:10.1049/enc2.70029

欢迎来到能源电力期刊网！

您当前的位置：

首页 >

文章列表页 >

Data-driven long-term wind speed forecasting and techno-economics of offshore wind-to-hydrogen production

更新时间：2026-03-24

- Data-driven long-term wind speed forecasting and techno-economics of offshore wind-to-hydrogen production
- Data-driven long-term wind speed forecasting and techno-economics of offshore wind-to-hydrogen production
- Energy Conversion and Economics 2025年6卷第6期
- 作者机构：
  
  1. Institute of Mechanical, Process and Energy Engineering (IMPEE), School of Engineering and Physical Sciences, Heriot-Watt University,Edinburgh,UK
  2. Abdul Wasy Zia, Institute of Mechanical, Process and Energy Engineering (IMPEE), School of Engineering and Physical Sciences, Heriot-Watt University,Edinburgh,UK,EH14 4AS
- 作者简介：
- 基金信息：
- DOI：10.1049/enc2.70029
  中图分类号：
- 纸质出版：2025
- 稿件说明：
移动端阅览
Prihandono Aditama, Abdul Wasy Zia. Data-driven long-term wind speed forecasting and techno-economics of offshore wind-to-hydrogen production[J]. Energy Conversion and Economics, 2025,6(6).

Prihandono Aditama, Abdul Wasy Zia. Data-driven long-term wind speed forecasting and techno-economics of offshore wind-to-hydrogen production[J]. Energy Conversion and Economics, 2025, 6(6).
Prihandono Aditama, Abdul Wasy Zia. Data-driven long-term wind speed forecasting and techno-economics of offshore wind-to-hydrogen production[J]. Energy Conversion and Economics, 2025,6(6). DOI： 10.1049/enc2.70029.

Prihandono Aditama, Abdul Wasy Zia. Data-driven long-term wind speed forecasting and techno-economics of offshore wind-to-hydrogen production[J]. Energy Conversion and Economics, 2025, 6(6). DOI： 10.1049/enc2.70029.

摘要

The economic viability of offshore wind-to-hydrogen systems depends heavily on the accuracy of wind-speed forecasting. We evaluated three probabilistic scenario generators (historical bootstrapping

parametric Weibull fitting

and calibrated long short-term memory (LSTM) sequence model) using a decision-coupled framework. From 61 years of ERA5

we created 1000 synthetic 23-year hourly scenarios per method

propagated them through a techno-economic model

and scored the continuous ranked probability score (CRPS) on economic distributions (levelized cost of hydrogen (LCOH)

net present value (NPV)

and interest rate of return (IRR)). We report the decision bandwidth and its elasticity to skill

and run a financing price sweep to meet probability targets. Relative to bootstrapping and Weibull

the LSTM lowers CRPS by 30% for LCOH and NPV (25% for IRR)

narrowing risk bands at similar medians. The hydrogen price required to reach more than 90% probability is €7.76–7.78/kg for Pr(NPV > 0) and €9.16-9.18/kg for Pr(NPV > 0 and IRR > 10%). A cross-method spread of less than €0.02/kg indicates a threshold-saturated regime

in which better skills mainly contract risk. Robustness tests shift medians

but preserve method ranking and thresholds. This framework translates probabilistic forecasting skills into finance ability metrics for wind-to-hydrogen screening and policy design.

Abstract

The economic viability of offshore wind-to-hydrogen systems depends heavily on the accuracy of wind-speed forecasting. We evaluated three probabilistic scenario generators (historical bootstrapping

parametric Weibull fitting

and calibrated long short-term memory (LSTM) sequence model) using a decision-coupled framework. From 61 years of ERA5

we created 1000 synthetic 23-year hourly scenarios per method

propagated them through a techno-economic model

and scored the continuous ranked probability score (CRPS) on economic distributions (levelized cost of hydrogen (LCOH)

net present value (NPV)

and interest rate of return (IRR)). We report the decision bandwidth and its elasticity to skill

and run a financing price sweep to meet probability targets. Relative to bootstrapping and Weibull

the LSTM lowers CRPS by 30% for LCOH and NPV (25% for IRR)

in which better skills mainly contract risk. Robustness tests shift medians

but preserve method ranking and thresholds. This framework translates probabilistic forecasting skills into finance ability metrics for wind-to-hydrogen screening and policy design.

关键词

Keywords

references

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

Risk analysis of distribution network outages under a typhoon–rainstorm–flood disaster chain

Deep learning for cybersecurity in smart grids: Review and perspectives