Scholarlink

<p>The global transition toward low-global warming potential (GWP) working fluids in Organic Rankine Cycles (ORC) presents complex thermo-economic and environmental dynamics. This study performs a comparative thermo-economic and environmental assessment of a recuperative ORC system utilizing hydrofluoroolefin (HFO) fluids R1233zd(E) and R1234ze(Z) as sustainable alternatives to the conventional R245fa for low-grade geothermal heat recovery (150 ℃). To overcome the computational cost of rigorous thermodynamic simulations near the critical point, a novel data-driven optimization framework is proposed. A comprehensive dataset of 3000 design points was generated using a robust process model incorporating flow-regime-dependent heat transfer correlations. A Bayesian-optimized support vector regression (SVR) surrogate model was developed to replace the physics-based solver, achieving a near-perfect prediction accuracy of R<sup>2</sup> = 0.9999 while accelerating the simulation speed by over 10,000 times (from 1.45&#xa0;s to 0.12&#xa0;ms). Subsequently, a weighted gray wolf optimizer (GWO) coupled with the SVR model was employed to the thermo-economic performance trajectory. The multi-objective optimization reveals a highly synergistic relationship between efficiency and cost for this specific application; R1233zd(E) is the superior candidate, achieving a maximum exergy efficiency of 40.02%, marginally outperforming the baseline R245fa (39.97%) while operating at a significantly lower evaporator pressure (839&#xa0;kPa vs. 1381&#xa0;kPa). Although this necessitates a slightly larger heat transfer area, increasing the specific investment cost (SIC) by 1.6% to 5442 USD/kW, the environmental benefits are decisive. Life cycle assessment (LCA) results indicate a 99.8% reduction in the total equivalent warming impact (TEWI) for R1233zd(E) compared to R245fa (dropping from 2926 to 4.6 tons CO<sub>2</sub>-eq). These findings demonstrate that AI-driven optimization can effectively leverage the synergistic relationship between thermodynamic perfection and economic viability, validating R1233zd(E) as a technically superior and environmentally imperative successor for next-generation green ORC plants.</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

Thermo-economic and environmental optimization of low-GWP HFO-based ORC systems using a hybrid Bayesian SVR-GWO framework

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月8日

查看全文
  • Bayram Kılıç,
  • Emre Arabacı

摘要

The global transition toward low-global warming potential (GWP) working fluids in Organic Rankine Cycles (ORC) presents complex thermo-economic and environmental dynamics. This study performs a comparative thermo-economic and environmental assessment of a recuperative ORC system utilizing hydrofluoroolefin (HFO) fluids R1233zd(E) and R1234ze(Z) as sustainable alternatives to the conventional R245fa for low-grade geothermal heat recovery (150 ℃). To overcome the computational cost of rigorous thermodynamic simulations near the critical point, a novel data-driven optimization framework is proposed. A comprehensive dataset of 3000 design points was generated using a robust process model incorporating flow-regime-dependent heat transfer correlations. A Bayesian-optimized support vector regression (SVR) surrogate model was developed to replace the physics-based solver, achieving a near-perfect prediction accuracy of R2 = 0.9999 while accelerating the simulation speed by over 10,000 times (from 1.45 s to 0.12 ms). Subsequently, a weighted gray wolf optimizer (GWO) coupled with the SVR model was employed to the thermo-economic performance trajectory. The multi-objective optimization reveals a highly synergistic relationship between efficiency and cost for this specific application; R1233zd(E) is the superior candidate, achieving a maximum exergy efficiency of 40.02%, marginally outperforming the baseline R245fa (39.97%) while operating at a significantly lower evaporator pressure (839 kPa vs. 1381 kPa). Although this necessitates a slightly larger heat transfer area, increasing the specific investment cost (SIC) by 1.6% to 5442 USD/kW, the environmental benefits are decisive. Life cycle assessment (LCA) results indicate a 99.8% reduction in the total equivalent warming impact (TEWI) for R1233zd(E) compared to R245fa (dropping from 2926 to 4.6 tons CO2-eq). These findings demonstrate that AI-driven optimization can effectively leverage the synergistic relationship between thermodynamic perfection and economic viability, validating R1233zd(E) as a technically superior and environmentally imperative successor for next-generation green ORC plants.