Scholarlink

<p>The performance and economic viability of utility-scale photovoltaic (PV) plants are critically influenced by location-specific climatic factors, yet conventional site selection often overlooks thermodynamic quality and long-term degradation. This study introduces an integrated exergoeconomic and machine learning framework for climate-resilient PV siting, combining transient exergy analysis, financial risk modeling, and Gaussian process regression (GPR). Applied to six climatic zones in Saudi Arabia for a standardized 100&#xa0;MW plant, the analysis reveals that the highest energy-yield location (Riyadh, 224.7 GWh/year) does not achieve the highest exergy efficiency (Abha, 16.1%), highlighting the limitation of energy-only metrics. The real-time cost of exergy destruction exceeds $85/h during peak conditions, underscoring the financial impact of thermodynamic losses. All sites demonstrate strong economic resilience, with a 95% Value at Risk for the levelized cost of energy below 3.1 US¢/kWh. GPR forecasts a 15–20% long-term performance divergence between optimal and sub-optimal sites. Statistical analysis confirms that location and season interact significantly (<i>p</i> &lt; 0.001) to drive performance variation. This framework provides a robust, data-driven decision-support tool for optimizing national PV portfolios under climatic uncertainty.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Exergoeconomic and machine learning approach to optimal PV siting under diverse climatic conditions

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月20日

查看全文
  • Marwa Ben Slimene,
  • Mohamed Arbi Khlifi

摘要

The performance and economic viability of utility-scale photovoltaic (PV) plants are critically influenced by location-specific climatic factors, yet conventional site selection often overlooks thermodynamic quality and long-term degradation. This study introduces an integrated exergoeconomic and machine learning framework for climate-resilient PV siting, combining transient exergy analysis, financial risk modeling, and Gaussian process regression (GPR). Applied to six climatic zones in Saudi Arabia for a standardized 100 MW plant, the analysis reveals that the highest energy-yield location (Riyadh, 224.7 GWh/year) does not achieve the highest exergy efficiency (Abha, 16.1%), highlighting the limitation of energy-only metrics. The real-time cost of exergy destruction exceeds $85/h during peak conditions, underscoring the financial impact of thermodynamic losses. All sites demonstrate strong economic resilience, with a 95% Value at Risk for the levelized cost of energy below 3.1 US¢/kWh. GPR forecasts a 15–20% long-term performance divergence between optimal and sub-optimal sites. Statistical analysis confirms that location and season interact significantly (p < 0.001) to drive performance variation. This framework provides a robust, data-driven decision-support tool for optimizing national PV portfolios under climatic uncertainty.