A knowledge-based system for low-grade waste heat recovery in the process industries

Abstract

The ever-increasing price of energy, combined with increasingly stringent legislation to reduce greenhouse gas emissions, is driving the UK process industries toward increasing energy efficiency. Significant gains can be made in this sector, as up to 11.4TWh per annum (4% of total energy use) of the UK process industries’ energy consumption is lost as recoverable waste heat. Substantial recovery of this waste heat would have economic benefits of the order of £100s of million/year, and environmental benefits of the order of 100s of thousands of tonnes of carbon dioxide equivalent per year. This thesis describes the development of a knowledge-based system for the selection and preliminary design of equipment for low-grade waste heat recovery in the process industries. The system addresses two of the key barriers to low-grade waste heat recovery in the UK. Firstly, it provides a readily accessible and zero cost tool to replace expensive, time-consuming expert consultancy in the initial stages of waste heat recovery projects, and, secondly, it educates users regarding the range and benefits of novel waste heat recovery technologies. The system requires an input of easy-to-access data from the user. Based on this data, it then selects the most appropriate technologies for waste heat recovery for the case study in question from a database including various types of heat exchanger, vapour compression heat pumps, mechanical vapour recompression and organic Rankine cycles. It also generates a preliminary design including equipment size, efficiency/effectiveness, capital cost, cost savings, payback time and potential reductions in carbon emissions. This provides sufficient information to allow the user to make an educated decision regarding whether or not waste heat recovery is suitable for their needs. The knowledge-base of the system was built using a decision tree method that has been proven to be successful in the building of decision making tools for various engineering applications. The software is programmed using the Java language which allows widespread free dissemination to computers running all common operating systems. The system was tested using case studies based on data from both existing publications and collaborating companies. The results were validated against published results, common modelling software results and the views of expert consultants. Broadly, in terms of equipment specification and cost, the knowledge-based system produced the same results as the other methods. Furthermore, the preliminary designs generated were generally within 5% of the final figures from the other sources. In certain cases, the knowledge-based system suggested alternative technologies that were more viable (economically and/or practically) than those considered by the authors of published case studies. In all cases, system operating time (data input, and processing of results) was of the order of minutes, whereas studies by consultants or the use of existing modelling packages would be significantly more time-consuming (of the order of hours or days). Hence, the system can be used as a rapid optioneering tool for investigation of waste heat recovery technologies, requiring substantially less time than current available methods.