Energy and utility flow modelling in manufacturing systems

Abstract

Over the previous decades, industry has been one of the major drivers of global economic growth and the issue of energy efficiency has become a matter of concern in different industries. Improving energy efficiency in industry, particularly in manufacturing plays a key role in reducing production costs and the environmental impact. Manufacturers also are driven to practise more energy efficient measures because of governmental regulations and policies. In this regard, a holistic view of the entire system is needed in which the dynamic behaviour of the production processes, supporting services, other energy consumers and their interrelationships must be considered. Predominantly, the underlying system is too complex so it is vital to utilise simulation to model the system. Furthermore, a simulation model can be used to design a system with an optimal performance where the performance relies on the value of some input parameters and the point is how to determine the values of these parameters (possibly subjected to some constraints) which contribute to an optimal performance. To address both above-mentioned issues, an integrated simulation-based optimisation framework was proposed in which a simulation model represented the production system in a hierarchical structure and simulated the energy consumed in the system using a bottom to up approach. Six basic modules were embedded, including four modules for main energy consuming equipment (one for machine tools and process chains, and three for TBSs including steam generation unit, compressed air system and HVAC systems). Careful observations of a wide range of diverse equipment were carried out for the first four modules, and then the basic components of a generic state-based energy consumption model were identified for each. Regarding the optimisation part of the framework, a population-based optimisation algorithm called the Cross Entropy method was utilised which treated the simulation model as a black box model to evaluate the system under different settings. A weighted sum method was used to combine different objectives in case of multi objectives. The proposed methodology was applied in two different manufacturing environments; a mass production system where a small number of products with large quantity were produced and a discrete-part production environment where a medium number of products with medium quantity were manufacture.