Progettazione di apparecchiature e sistemi industriali in condizioni di incertezza e di funzionamento variabile

Variability and uncertainty often strongly affect the industrial environment. When this happens, it is impossible to neglect uncertainty, ensuring that this does not lead to incorrect decisions and wrong estimation of the inherent technical and economic risk. Industrial equipment, components, and systems often operate under conditions of deep uncertainty and variable operating conditions. However, they are usually designed based on nominal conditions, neglecting sources of uncertainty. Traditional design methods commonly assume constant process conditions and known values of the various design factors involved. However, this approach may fail to meet specifications when operational conditions change and can lead to inaccurate decisions and underestimation of inherent risk. Using safety factors or worst-case design can result in a conservative design, with components and equipment being oversized. Starting from a literature analysis of the existing methods and frameworks, this research aims to propose a general framework that extends the existing ones, including, incorporating, and systematising already available methods and approaches, dealing with two different macro-problems: incorporating uncertainty during the design phases and performing accurate risk assessments under uncertainty. This thesis also introduces a novel classification of uncertainty based on the variables’ behaviour. This new clustering technique aims to streamline selecting the proper methods to represent the uncertainty sources. The framework comprises different blocks, and each block can be activated or switched off to achieve different objectives. Although the objectives can be divided into several groups, two main clusters of goals are system design optimisation and system performance evaluation. Two case studies have been conducted to demonstrate the capabilities of the proposed general framework. The first study involves the design optimisation of a shell and tube heat exchanger, whereas the second one involves the economic assessment of a wind power system. Following the general framework approach in the application examples shows how this approach may lead to a more manageable selection of the proper methods to model, propagate, and assess the uncertainty effects while implementing optimisation and counteraction for risk mitigation. The results of the general framework’s application to different case studies have shown the subsequent fact. The uncertainty should be considered to obtain a more effective design and assess the economic and technical risk adequately. Including several sources of uncertainty dramatically increases the dispersion of the industrial system’s output, showing the impact of uncertainty sources on the system's performance. A technically and economically viable design obtained under deterministic conditions may be ineffective when uncertainty is included. In other words, this thesis proposes a general framework for designing and evaluating industrial systems under uncertainty using a full probabilistic evaluation method. Future work will apply this methodology to manufacturing plants and include real-time data to adapt mitigation instruments to different cases by selecting the most appropriate ones.