Dynamic data analysis™ of large scale data to monitor fouling in heat exchanger networks

Abstract

Data collected from heat exchangers networks in the field are typically used to monitor the thermal performance of the individual units. However, the information extracted from the data is limited in quantity and quality by the simplified models typically used in practice. Key decisions such as cleaning of heat exchangers rely on the calculation of derived quantities such as the fouling resistance which lumps together a number of factors contributing to fouling. This approach has been severely criticised in the past by various authors but it is still widely used in the industrial practice [1]. In this paper it is shown that a significantly larger amount of information and insights can be extracted from the same measurements by using rigorous models and a flexible framework. It is shown how this approach leads to a much deeper analysis of the status of the network which, in turn, helps with diagnosis and troubleshooting. An industrial case study is presented to illustrate the benefits.