• Energy Research

Abstract Electric motors are industrial systems’ components widely diffused enabling all productive processes and safety equipment. They are affected by aging effect with a contribution based on the environmental condition on which they work. In order to design efficient maintenance plans, the behaviour of their main components, such as bearings and winding, has to be predicted. Therefore, a model-based methodology is applied aiming at codifying the failure rate of an electric engine, taking into account the thermal aging and relevant environment boundary conditions in which bearings and winding operate. The winding failure mode is coded by means of the Military standard technique while the bearings one is simulated comparing the Military Standard and the Svenska Kullagerfabriken (SKF) techniques. While the former predicts more conservative behaviours, the latter, taking into account lubrication conditions, dynamic loads and a better knowledge of materials quality, enables to capture the evolution of the operative conditions. The proposed reliability model can capture both the deterministic and stochastic behaviour of the electric motor: it belongs to the field of hybrid automaton application; the model is coded by means of the emerging software framework called SHYFTOO. The proposed model and the Monte Carlo simulation process that performs its evolution can support the development of a new class of electric motors: a cyber-physical oriented electric motor.

This paper presents a novel approach to estimating short-term production of wind farms, which are made up of numerous turbine generators. It harnesses the power of big data through a blend of data-driven and model-based methods. Specifically, it combines an Artificial Neural Network (ANN) for immediate future predictions of wind turbine power output with a stochastic model for dependability, using Hybrid Reliability Block Diagrams. A thorough state-of-the-art review has been conducted in order to demonstrate the applicability of an ANN for non-linear stochastic problems of energy or power forecast estimation. The study leverages an innovative cluster analysis to group wind turbines and reduce the computational effort of the ANN, with a dependability model that improves the accuracy of the data-driven output estimation. Therefore, the main novelty is the employment of a hybrid model that combines an ANN with a dependability stochastic model that accounts for the realistic operational scenarios of wind turbines, including their susceptibility to random shutdowns This approach marks a significant advancement in the field, introducing a methodology which can aid the design and the power production forecast. The research has been applied to a case study of a 24 MW wind farm located in the south of Italy, characterized by 28 turbines. The findings demonstrate that the integrated model significantly enhances short-term wind-energy production estimation, achieving a 480% improvement in accuracy over the solo-clustering approach.

Grid-connected low voltage photovoltaic power plants cover most of the power capacity installed in Italy. They offer an important contribution to the power demand of the utilities connected but, due to the nature of the solar resource, the night-time consumption can be satisfied only withdrawing the energy by the national grid, at the price of the energy distributor. Thanks to the improvement of storage technologies, the installation of a system of battery looks like a promising solution by giving the possibility to increase auto-consumption dramatically. In this paper, a model-based approach to analyze and discuss the performance and the economic feasibility of grid-connected domestic photovoltaic power plants with a storage system is presented. Using as input to the model the historical series (2008–2017) of the main ambient variables, the proposed model, based on Stochastic Hybrid Fault Tree Automaton, allowed us to simulate and compare two alternative technical solutions characterized by different environmental conditions, in the north and in the south of Italy. The performances of these systems were compared and an economic analysis, addressing the convenience of the storage systems was carried out, considering the characteristic useful-life time, 20 years, of a photovoltaic power plant. To this end the Net Present Value and the payback time were evaluated, considering the main characteristics of the Italian market scenario.

Abstract Analysis of complex failure scenarios and mitigation procedures of an industrial plant is one of the most important activity for the safety of the factory, the personnel and the surrounding areas. The dependability assessment of such systems is fulfilled by risk experts who, adopting well-known Reliability, Availability, Maintenance and Safety (RAMS) techniques, design and solve the stochastic failure model of the system. Traditional techniques like Fault Tree Analysis (FTA) or Reliability Block Diagrams (RBD) are of easy implementation but unrealistic, due to their simplified hypotheses that assume the components malfunction to be independent from each other and from the system working conditions. Dynamic Probabilistic Risk Assessment (DPRA) is the umbrella framework encompassing new mathematical and simulation formalisms aiming to relax the constraints of traditional techniques and increase the accuracy of dependability assessment. At the state of the art, DPRA cannot boast a well-defined methodology because the nature of a dynamic reliability problem can be so complex to require an ad-hoc modelling and resolution. Moreover, one of the main issues encountered by risk-practitioners is that there is a small support in terms of available tools or expert systems, specifically designed for DPRA problems. To tackle this lack, this paper presents the conception of general framework for the modelling and the simulation of a Stochastic Hybrid Fault Tree Automaton (SHyFTA), one of the most promising DPRA methodologies, able to combine Dynamic Fault Tree (DFT) with the deterministic model of the system process. The logic of the repairable DFT gates and the concepts for the implementation of a simulation engine combining Discrete Event Simulation (DES) and Time Driven Simulation (TDS) are illustrated and, a Matlab® toolbox library (SHyFTOO) has been coded and tested with a thorough validation campaign. Finally, a common case study in industrial engineering has been modelled and analysed under different stand-by configurations in order to demonstrate the modelling flexibility of the toolbox.

Despite the emerging contribution of machine automation, artificial intelligence and information systems, humans remain yet the most fragile ring of any organization. Decision support systems are widespread, supporting us to decide among uncertainties, such as weather conditions, suppliers’ performances and financial opportunities, but how humans take into account this information and, most of all, how they trust their own management knowledge is a controversial issue. This paper assesses, by means of a controlled experiment and ex post interviews, how individuals consider and use decision support systems in the context of the Newsvendor Problem. In accordance with prior research, the results show that individuals’ order quantities are pull-to-center biased. Moreover, ex post direct interviews suggest that (i) the individuals’ trust in decision support systems is not blind; (ii) individuals do not play the business game as a real task, (iii) they are biased by the type of incentive promised and (iv) they seem not skilled or trained enough. Ex post interviews shed a new light on controlled human experiments: they should be better analyzed and re-engineered.

Abstract The contribution of wind power systems to the reduction of the impact of fossil fuels sources has increased more and more during the last decades leading to a greater attention to the estimation of the performances of renewable power plants. However, forecast methods of productivity of onshore/offshore wind farms still suffer, in terms of accuracy, the innate variability of the energy resources and the effect of components failures. This paper proposes a novel “hybrid” approach for the estimation of the energy conversion of onshore wind farms. The model combines the Jensen wake mathematical theory with a stochastic dependability model, a Fault Tree, to better forecast the energy production. A new key index was conceived to optimize the preventive maintenance of wind turbines. This model was tested on a real case study, a wind farm (25.5 MWp) located in the south of Italy. Results were promising because the model achieved a twofold objective to improve the accuracy of the energy conversion forecast and to provide a support decision system for the activities of maintenance planning.