Estudio de la variación de la demanda energética en las Islas Canarias con el cambio climático

Abstract

In the study presented below, an analysis of the variability of the increase in energy demand in the Canary Islands due to meteorological causes has been carried out. The seasonal and daily conditions of energy demand in the period 2013-2019 have been studied. Then, some future estimations were made. The variability of the energy demand has been studied through three standard indexes, monthly seasonal variation index (MSVI), daily seasonal variation index (DSVI) and hourly seasonal variation index (HSVI). The results confirm the seasonal increase in energy demand in the long summers of all the Canary Islands. This variability is greater in the non-capital islands, probably due to a greater flow of residents during the summer period. The hourly and daily variability of the energy demand is also very remarkable in all the islands, which has been reflected by the imposition of different quotas in several time slots by the Spanish Government and the Electricity Companies. The energy consumed during weekdays remains fairly constant throughout the week, decreasing on Saturday and, even more so, on Sunday. Even in some islands, such as Fuerteventura or La Gomera, the drop is only really noticeable on Sundays. The direct relationship between the increase in demand and the increase in temperatures, through the cooling degree-days, CDD, has been proven. This gives rise to quite disparate values, with particularities being observed on each island. For this purpose, the comfort temperatures of the different islands have been calculated, using a degree 3 polynomial fit of the energy demand data versus the average temperature of each island, once demand time series have been detrended. Establishing in this way a base level of energy consumption of cooling systems from which temperature changes involve an increase in energy demand. Both by the use of refrigeration systems giving rise to the CDD, and heating systems that give rise to the so-called Heating degree-days, HDD. The latter having a relatively lower weight in the Canary Islands due to climatic conditions. The time series for the mean temperature of each island were calculated from the ERA5 reanalysis data, averaging all those grid nodes that correspond to land and that are below 1000 masl. In this way, the possible bias produced by the low temperatures at higher elevations, which are not significant for the relationship with electricity consumption, since it does not occur in those areas, is reduced. After the analysis of the historical 2013-2019 period, the study has been extended to the analysis of possible future scenarios of the evolution of electricity demand and its variability due to climate change. Regionalized climate projection data were provided by Grupo de Observación y la Atmósfera (GOTA), that you belong to the Universidad de La Laguna (ULL). These projections were made using the WRF mesoscale model and different boundary conditions provided by three global climate models (GFDL-ESM2M, IPSL-CM5A-MR and MIROC-ESM) were used to simulate two future periods under study: 2030-59 and 2070-99. In addition, two possible socio-economic scenarios of greenhouse gas emissions, the RCP4.5 (Representative Concentration Pathway) scenario, a more hopeful scenario, and the RCP8.5, a more catastrophic one, were taken into account. They correspond to additional 4.5 and 8.5 W/m2 radiative forcings by 2100, respectively. Data from the WRF simulations, which have a much higher resolution, were aggregated to create a grid equivalent to that of the ERA5 and the same process was applied to calculate the mean temperature time series for each island. Furthermore, a bias correction method was applied to these time series, using the scaled distribution mapping (SDM) technique, which outperforms previous methods based on quantile mapping and preserves raw climate model projected changes to meteorological variables such as temperature and precipitation. An increase in CDD was predicted in both cases, in the RCP4.5 scenario in a more moderate and assumable way with a stabilization of the values. On the other hand, in the RCP8.5 scenario, the increase in CDD is exponential and without stabilization, reaching values of almost 5 more CDD in the summers of the 2070-2099 period. In addition, in the first three months of the year, when there are currently very few days with non-zero CDD values, the energy demand for cooling in the future could be significant. At the end of the century and in the least favorable scenario, the corresponding CDD could take values between 2 and 5 during those first months. For future work, it would be interesting to have more disaggregated data on energy demand; it would be interesting to study the relationship between energy demand and temperature in different socioeconomic environments: rural, residential areas, industrial or commercial areas, etc. In addition, the use of projections of technological and socioeconomic evolution, that allow us to estimate a future trend in the use of cooling systems and their efficiency, would make it possible to translate project results, currently based on CDD, into estimates of future energy demand. This approximation would be much more appropriate than simply assuming that energy uses and technologies remain unchanged.

En el estudio que se presenta, se ha hecho un análisis de variabilidad de la demanda energética en las Islas Canarias por causas meteorológicas. Se han estudiado las condiciones estacionales y diarias de la demanda energética en el periodo 2013-2019. Se ha estudiado dicha variabilidad de la demanda a través de tres índices estándar, monthly seasonal variation index (MSVI), daily seasonal variation index (DSVI) y hourly seasonal variation index (HSVI). Obteniendo la confirmación del aumento de la demanda energética de manera estacional en los veranos alargados de todas las Islas Canarias. Asumiendo una relación directa entre el aumento de la demanda y el aumento de las temperaturas, a través de los cooling degree-days , CDD, se han calculado las temperaturas de confort de las diferentes islas. Estableciendo en esta un nivel base de consumo energético de los sistemas de refrigeración a partir del cual los cambios de temperatura implican el aumento en la demanda energética. Realizado un análisis en el histórico 2013-2019 se ha extendido el estudio al análisis de los posibles escenarios futuros de la evolución de la demanda eléctrica y la variabilidad de estas por el cambio climático. Se han supuesto diferentes condiciones de contorno para la simulación de los datos de la temperatura media diaria de los dos periodos futuros a estudio 2030-2059 y 2070-2099. Condiciones provenientes de tres modelos climáticos globales de nivel internacional, los modelos GFDL-ESM2M, IPSL-CM5A-MR y MIROC-ESM. Además se han usado datos basados en dos posibles escenarios socioeconómicos de emisiones de gases de efecto invernadero, los escenarios RCP 4.5, un escenario más esperanzador, y el escenario RCP 8.5, un escenario más catastrofista. Se estima un aumento de los CDD en ambos casos, en el escenario RCP 4.5 de una manera más moderada y asumible con una estabilización de los valores. En cambio en el escenario RCP 8.5 el aumento de los CDD se produce de manera más abrupta y sin estabilización, alcanzando valores de casi 5 CDD más en los veranos del periodo 2070/99.