MATLAB Framework for Thermodynamic Modelling, Optimisation, Sizing, and Economic Assessment of Maritime Waste Heat Recovery Systems

Overview: This repository contains the complete MATLAB codebase developed for the design and comparison of shipboard waste heat recovery cycles. The software models idealised transcritical CO2 cycles: simple, recuperated, reheated recuperated, recompressed recuperated and cascade recuperated (CR); as well as six shipboard cycle architectures: simple, CR, split recuperated, split reheated recuperated, dual pressure (DP) and DP CR; across six different working fluids: CO2, steam, toluene, cyclohexane, cyclopentane and pentane. Cycles recover heat from marine engine exhaust gas and jacket water streams. It performs steady-state thermodynamic analysis, gradient free optimisation, component sizing (heat exchangers and turbomachinery), and costing to evaluate net power output, thermal efficiency, heat recovery efficiency, footprint, capital cost, fuel savings and payback periods across representative marine operating zones. Modelling Approach: The code was created to be modular, with separate functions for each cycle component (condenser, pump/compressor, recuperator, heater, mixer and turbine) that allow simple construction and modification of cycle architectures. These component functions were also made to generate the T-s and P-h diagram representations. Heat exchanger modelling follows a minimum pinch point approach, and sizing is done with an LMTD / enthalpy based effectiveness-NTU approach, phase-change segmentation and Gnielinski convective coefficients. Turbomachinery sizing is achieved with specific speed and specific diameter correlations. Power-law equipment costing correlations were used from Turton, Weiland and Shamoushaki for shell-and-tube heat exchangers, recuperators, supercritical radial turbines, axial turbines, pumps, generators and gearboxes. Costs are escalated to 2026 using CEPCI. Optimisation uses a genetic algorithm to find the optimal basin, followed by refinement with a general pattern search, over each cycle's decision variables. Each constructed cycle has associated optimisation algorithms and enforced feasibility constraints (turbine superheat enforcement, wet expansion correction factors, acid dew point and minimum jacket water return temperatures). A series of post-processing tools span zone-weighted annual performance to figure plotting. Working-fluid properties are calculated using CoolProp via a Python interface in MATLAB. Structure: componentFunctions/ - modular component models cycleFunctions/ - cycle models sizingFunctions/ - heat exchanger and turbomachinery sizing costingFunctions/ - equipment cost correlations optimisationScripts/ - GA and GPS postProcessing/ - plotting and calculations Dependencies: MATLAB R2021b or later MATLAB global optimization toolbox MATLAB parallel computing toolbox Python v3.9 with CoolProp installed Contact: [e-mail] Related Publication: Comparative Assessment of Transcritical CO2, Steam, and Organic Rankine Cycles for Maritime Waste Heat Recovery