This paper presents a portfolio-based approach to the harvesting of renewable energy (RE) resources. Our examined problem setting considers the possibility of distributing the total available capacity across an array of heterogeneous RE generation technologies (wind and solar power production units) being dispersed over a large geographical area. We formulate the capacity allocation process as a bi-objective optimization problem, in which the decision maker seeks to increase the mean productivity of the entire array while having control on the variability of the aggregate energy supply. Using large-scale optimization techniques, we are able to calculate - to an arbitrary degree of accuracy - the complete set of Pareto-optimal configurations of power plants, which attain the maximum possible energy delivery for a given level of power supply risk. Experimental results from a reference geographical region show that wind and solar resources are largely complementary. We demonstrate how this feature could help energy policy makers to improve the overall reliability of future RE generation in a properly designed risk management framework

The intensity and duration of hot weather and the number of extreme weather events, such as heatwaves, are increasing, leading to a growing need for space cooling energy demand. Together with the building stock’s low energy performance, this phenomenon may also increase households’ energy consumption. On the other hand, the low level of ownership of cooling equipment can cause low energy consumption, leading to a lack of indoor thermal comfort and several health-related problems, yet increasing the risk of energy poverty in summer. Understanding future temperature variations and the associated impacts on building cooling demand will allow mitigating future issues related to a warmer climate. In this respect, this paper analyses the effects of change in temperatures in the residential sector cooling demand in 2050 for a case study of nineteen cities across seven countries: Cyprus, Finland, Greece, Israel, Portugal, Slovakia, and Spain, by estimating cooling degree days and hours (CDD and CDH). CDD and CDH are calculated using both fixed and adaptive thermal comfort temperature thresholds for 2020 and 2050, understanding their strengths and weaknesses to assess the effects of warmer temperatures. Results suggest a noticeable average increase in CDD and CDH values, up to double, by using both thresholds for 2050, with a particular interest in northern countries where structural modifications in the building stock and occupants’ behavior should be anticipated. Furthermore, the use of the adaptive thermal comfort threshold shows that the projected temperature increases for 2050 might affect people’s capability to adapt their comfort band (i.e., indoor habitability) as temperatures would be higher than the maximum admissible values for people’s comfort and health.

An integral part of econometric practice is to test the adequacy of model specifications. If a model is adequately specified, it should not leave interesting features of the data-generating process in the errors. Despite the common tradition, the importance of diagnostic checking as a safeguard against mis-specification has only recently been recognized by neural network (NN) practitioners, possibly because this type of semi-parametric methodology was not originally designed for economic and financial applications. The purpose of this paper is to compare a number of analytical statistical testing procedures suitable to diagnostic checking on a neural network regression model. We present the standard Lagrange multiplier (LM) testing framework designed under the assumption of identically distributed disturbances and also examine two modifications that are robust to heteroskedasticity in errors. One modification also gives the researcher an opportunity to incorporate information concerning the volatility structure of the data-generating process in the testing procedure. By means of a Monte Carlo simulation, we investigate the performance of these tests under GARCH-type heteroskedasticity in errors and various distributional assumptions. The results show that although the primary concern of the researcher may be to design a regression model that accurately captures relations in the mean of the conditional distribution, developing a good approximation of the underlying volatility structure generally increases the efficiency of tests in detecting non-adequacy of a NN model.