Automated retrofit of heat exchanger networks

Abstract

In a large industrial processing plant, a significant amount of process heat is generated using fossil fuels, contributing to New Zealand’s overall greenhouse gas emissions. While New Zealand’s renewable energy sources present potential alternatives to fossil fuels, reducing the process heat demand of industrial processing plants is another critical step for emissions reduction, with multiple long-term economic and environmental benefits. Reducing process heat demand often centres on retrofitting the heat exchanger network to improve heat recovery and lower the hot utility consumption. This thesis presents a comprehensive automated retrofit design method for heat exchanger networks, fulfilling several gaps in current knowledge. The novel contributions are presented and discussed in four main chapters: (1) two retrofit tools are developed that enable visualisation of the problem – the Modified Energy Transfer Diagram and the Heat Surplus-Deficit Table, (2) an algorithm called Automated Retrofit Targeting that searches for all possible retrofit modifications of the heat exchanger network that unlocks energy savings, (3) a comprehensive energy retrofit planning tool that uses a multi-stage retrofit analysis to provide strategic long-term and cost-effective retrofit plans, and (4) an heat exchanger network simulation method, incorporating using Monte Carlo Simulation, to quantify the effect of variable process flows and temperatures on flexibility and steady state performance. The developed methods are illustrated using a simple four-stream network and then applied to two industrial case studies that are representative of some of the large industrial energy-users in New Zealand: a paper mill at a Kraft pulp and paper mill cluster and a petrochemical complex. In both case studies, numerous potential retrofit designs have been identified. These options are reduced to those that ranked as the most cost-effective, using thermodynamic and economic constraints with Pareto front analysis. By successive application of the retrofit method, long-term retrofit plans have been proposed, enabling strategic sequencing of modifications and minimising process heat demand. For the petrochemical complex, the hot utility consumption could decrease by 9.58 MW (63% of the total possible savings) using four separate stages of retrofits. The total retrofit profit would reach 1.5 million Euros per year with a simple payback of 1.7 years. The paper mill could achieve its process heat demand reduction target using four retrofit stages, achieving a total retrofit profit of 0.78 million NZD per year. These modifications resulted in an inflexibility of 51.3% (probability of the heat exchanger network failing to meet all target temperatures), but this could be reduced to 4.1% using bypass control. This thesis provides a novel retrofit method, which uses graphical and numerical tools, in combination with automated analysis, generating and evaluating cost-effective retrofit designs. The retrofit design method has been implemented in Microsoft ExcelTM as a critical and functional output that can be rapidly applied to complex industrial heat exchanger networks.