Multi-objective optimization of ORC Using NSGA-II Algorithm with Python-ASPEN Hybrid Platform

This repository presents an approach to optimize Organic Rankine Cycle (ORC) systems using the NSGA-II algorithm within a hybrid platform bridging Python and Aspen Plus. In this framework, Python and Aspen Plus communicate seamlessly through a COM (Component Object Model) connection, enabling efficient data exchange and integration of optimization algorithms with process simulation.

General formulation

This problem aims to minimize the environmental impact of the ORC while maximizing its efficiency.

$$ \begin{align} \begin{split} \min_{x_i, P_P} ;; & env = \sum_i{x_i} \cdot GWP_i \cdot MW_i \cdot F \\ \max_{x_i, P_P} ;; & \eta = \frac{W_{turb} - W_{pump}}{Q_{in}} \\ \text{s.t.} ;; & \sum_i{x_i} = 1 \\ & 0 \leq x_i \leq 1 \\ & 0 \leq P_P \leq 10 \\ & x_i, P_P \in \mathbb{R} \end{split} \end{align} $$

Where:

• $x_i$ represents the molar fraction of the components.
• $GWP_i$ represents the Global Warming potential in $\frac{kg_{CO_2}}{kg_i}$
• $P_P$ stands for the Pump pressure in $bar$
• $F$ stands for the total molar flowrate in $\frac{kmol}{s}$

Case study 1

For this case study, butane, pentane, isobutane, and isopentane were chosen as working fluids. To replicate the results, run main.ipynb in the Case study 1 folder.

Case study 2

For this case, in addition to the working fluids butane, pentane, isobutane, and isopentane, the refrigerants R113, R114, R123, and R124 were also used. To replicate the results, run main.ipynb in the Case study 2 folder.

Conclusion

This work demonstrates the significant capability of connecting Python with ASPEN, enabling optimization of the intriguing ORC technology using a model-free method. It is worth noting that this work could be further expanded and enhanced, considering factors such as net profit or operational risks (e.g., flammability, toxicity).