Konstantinos N. Genikomsakis

Wind Power Forecasting in a Residential Location as Part of the Energy Box Management Decision Tool

A number of location characteristics, such as buildings, mountains, and trees, are likely to influence the wind flow that reaches a microwind turbine located at a residential area, and as a consequence they may affect the actual wind speed that is potentially utilized by the turbine. In this context, simple regional predictions for the wind energy from the nearest location available can easily lead to unacceptable modeling errors. There is thus a need to develop a framework for predicting the values of the wind speed at the desired location. This work addresses the aforementioned issue by employing a multilayer feed-forward back-propagation neural network for classification that utilizes the global forecast system (GFS) predictions on wind speed and direction to identify patterns of the wind behavior at the location considered in order to obtain a stochastic distribution of the daily wind speed. The proposed approach aims to support the implementation of an enhanced energy box (EB) management decision tool, while its feasibility is demonstrated through a case example for a region in the south of Portugal.

Electromobility and carsharing/carpooling services at the University of Deusto: A preliminary exploratory survey

The widespread deployment of electric vehicles is typically viewed as an approach that could help decarbonize the transportation sector, which is becoming increasingly linked not only to environmental problems, but also to social issues that contribute to the degradation of the quality of life. Alternative modes of transport with the aim to increase the occupancy of private cars are emerging as a means of alleviating the problems from traffic congestion, mainly in urban areas. At the same time, the university members are often considered to be more receptive to alternative transportation services such as carsharing, compared to the general population. Considering that academic communities offer fertile ground for the promotion of innovative actions, this paper presents a preliminary survey conducted at the University of Deusto, Bilbao, Spain, that aims to explore the potential interest and willingness to adopt the use of carsharing/carpooling services and electric vehicles. The results of this exploratory study provide the indication that the population under study maintains a positive attitude towards the transition to electromobility and use of alternative modes of transportation.

Simulation and design of a fast charging battery station in a parking lot of an e-carsharing system

Carsharing has the potential to reduce the total number of cars on the road, with significant benefits to the society and the environment, while at the same time relevant studies show that university communities are often more receptive to alternative transportation services compared to the general population. With the growing interest in electromobility, as a means of decarbonizing the transportation sector, this paper considers the case of combining carsharing with electric vehicles (EVs) to serve the commuting needs of students, employees and faculty of a university in Bilbao, Spain. The aim of the present work is to conceptualize the design of the charging infrastructure of the e-carsharing system under a fast charging scheme and define its components, their attributes and interactions. To this end, a MATLAB/Simulink based simulator is developed incorporating the dynamics of a real-world scenario based on arrival and departure data from the university parking lot.

Solar production forecasting based on irradiance forecasting using artificial neural networks

There is a growing awareness that forecasting of solar irradiance is of special importance for forecasting the power output of photovoltaic (PV) systems and thus for optimizing their operation. This work presents the development of solar irradiance and PV power output forecasting models, based on artificial neural networks (ANNs), operating with a time horizon of 24 h in order to be integrated as part of home energy management systems (HEMS). The key characteristic of the proposed approach consists of employing statistical feature parameters to reduce the size of input data, while the results obtained indicate that it provides a reasonable balance between computational requirements and forecasting accuracy of the PV power output within the considered time frame.

Short-term wind speed forecasting model based on ANN with statistical feature parameters

The intermittent and unstable nature of wind raises significant challenges for the operation of wind power systems, either residential installations or utility-scale implementations, necessitating the development of reliable and accurate wind power forecasting techniques. Given that wind speed forecasting is typically considered the intermediate step for wind power forecasting, the present work proposes a novel short-term wind speed forecasting model based on an artificial neural network (ANN), with the key characteristic that statistical feature parameters of wind speed, wind direction and ambient temperature are employed in order to reduce the input vector and thus the complexity of the model. The results obtained indicate that the proposed model strikes a reasonable balance between accuracy and computational requirements for a forecasting time horizon of 24 hours, providing a light-weight solution that can be integrated as part of energy management systems for small scale applications.

A life cycle assessment of a Li-ion urban electric vehicle battery

This study presents an approach on the life cycle assessment and environmental impact of lithium-ion batteries for electric vehicles, specially the iron phosphate technology based battery (LFP), through evaluating the different stages in the whole life of the battery starting with the manufacturing stage and then proceeding with the evaluation of its use in Spain until reaching the end-of-life stage, when the battery cannot continue offering a service in the electromobility sector (the actual capacity is under the 80% of the initial one). In this context, different end of life scenarios are considered in order to examine the feasibility of a second hand use for the already-worn battery that consists of reducing its environmental impact by extending the life of the battery in less stressful conditions to ensure the lower effect of the degradation. To this end, the various utilities in the life cycle of this battery are examined with the help of the SimaPro software simulation tool in order to quantitatively assess the potential benefits from an environmental point of view.

Introduction of plug-in hybrid electric vehicles in an isolated island system

ABSTRACT This paper considers the case of São Miguel in the Azores archipelago as a typical example of an isolated island with high renewable energy potential, but largely dependent on fossil fuels incurring high import costs, in order to assess and analyse the potential impact of the plug-in hybrid electric vehicle (PHEV) technology on the local power supply system. To this end, the present work employs The Integrated MARKAL-EFOM System (TIMES) to examine a number of scenarios with different levels of PHEVs penetration under the grid-to-vehicle (G2V) approach, taking into account the established Government policies, regarding the increase in renewable energy production quotas, for the evolution of demand and supply over time. The results obtained indicate that the PHEVs integration into the local grid system under the G2V energy transferring paradigm can be realized without immediate technical barriers and bears the potential to yield significant benefits to the energy mix, reducing thus the environmental impact.

Regression Based Emission Models for Vehicle Contribution to Climate Change

The reduction of carbon emissions within the transportation sector is one of the most important steps against the threat of global warming. Unless strict emissions-reduction and fuel economy policies are in place, the resulting pollution is expected to increase dramatically along with the amount of vehicles on the roads. An accurate quantification of the emissions produced by each type of vehicle is essential in order to evaluate the social and environmental impacts derived. The literature shows a wide range of pollutant emission models, whether empirical, database centric or regression based. In this paper, we propose and analyze 3 regression based models built on data from pollutant emission databases and knowledge models. The first model is based on an exponential regression that improves the results given in the state of the art. In contrast, the other two models are based on different Artificial Intelligence techniques, namely Artificial Neural Networks and Support Vector Regression, which further improve the results.

Utilizing plug-in electric vehicles for peak shaving and valley filling in non-residential buildings with solar photovoltaic systems

This paper examines the concept of utilizing plug-in electric vehicles (PEVs) and solar photovoltaic (PV) systems in large non-residential buildings for peak shaving and valley filling the power consumption profile, given that the energy cost of commercial electricity customers typically depends on both actual consumption and peak power demand within the billing period. Specifically, it describes a hybrid approach that combines an artificial neural network (ANN) for solar irradiance forecasting with a MATLAB/Simulink model to simulate the power output of solar PV systems, as well as the development of a mathematical model to control the charging/discharging process of the PEVs. The results obtained from simulating the case of the power consumption of a university building, along with experimental parking occupancy data from a university parking lot, demonstrate the applicability and effectiveness of the proposed approach.

A non-myopic approach for a domotic battery management system

The foreseeable dynamic pricing of electricity combined with the emergence of the so-called electricity “prosumers” that not only consume, but also have the ability to produce and store electricity, necessitates the development of energy management systems even at residential level. In this context, the present work considers that in a modern residence, or a “smart house”, electricity is generated from micro-renewable energy sources, while the electricity storage options include a fixed battery as well as the battery from a plug-in electric vehicle. This paper presents the theoretical framework for the implementation of a domotic battery management system with the key characteristic that energy management decisions are made every 10 min, based on the forecasted conditions for the next 24 h. The results obtained from the simulation of the proposed system highlight the contribution not only of the fixed battery, but also of the electric vehicles battery to the maximization of electricity cost savings.