Abstract: In this paper, we investigate potential pathways for achieving deep reductions in CO2 emissions by 2050 in the Chilean electric power system. We simulate the evolution of the power system using a long-term planning model for policy analysis that identifies investments and operation strategies to meet demand and CO2 emissions reductions at the lowest possible cost. The model considers a simplified representation of the main transmission network and representative days to simulate operations considering the variability of demand and renewable resources at different geographical locations. We perform a scenario analysis assuming different ambitious renewable energy and emission reduction targets by 2050. As observed in other studies, we show that the incremental cost of reducing CO2 emissions without carbon capture or offset alternatives increases significantly as the system approaches zero emissions. Indeed, the carbon tax is multiplied by a factor of 4 to eliminate the last Mt of CO2 emissions, i.e., from 2000 to almost 8500 USD/tCO(2) in 2050. This result highlights the importance of implementing technology-neutral mechanisms that help investors identify the most cost-efficient actions to reduce CO2 emissions. Our analysis shows that Carbon Capture and Storage could permit to divide by more than two the total system cost of a 100% renewable scenario. Furthermore, it also illustrates the importance of implementing economy-wide carbon emissions policies that ensure that the incremental costs to reduce CO2 emissions are roughly similar across different sectors of the economy.

Abstract: Chile is currently engaged in an energy transition process to meet ambitious greenhouse gas reductions and improved air quality indices. In this paper, we apply a long-term energy planning model, with the objective of finding the set of technologies that meet strong reductions of CO2 emissions and of local PM2.5 concentrations. For this purpose, we use the existing ETEM-Chile (Energy-Technology-Environment-Model) model which considers a simplified version of the Chilean electricity sector that we extend to the residential and commercial sectors and to local concentration considerations. We propose an original approach to integrate in the same framework local and global emission constraints. Results show that to meet the goal of zero emissions by 2050, electrification of end-use demands increases up to 49.2% with a strong growth of the CO2 marginal cost. It should be noted that this electrification rate is much lower than government projections and those usually found in the literature, in certain geographic areas in southern Chile with a wide availability of firewood for residential heating. Regarding local PM2.5 concentrations, our analysis shows that even without a specific emission reduction target, acceptable PM2.5 concentrations are achieved by 2045, due to first the emergence of more efficient, cleaner and cost-effective end-use technologies, in particular, residential firewood heaters, and second the use of drier and therefore less contaminating firewood. Achieving acceptable air quality as early as 2030 is also possible but comes with a high marginal cost of PM2.5 concentration. Our results illustrate the need for implementing effective public policies to (i) regulate the firewood heating market to increase its production and improve its environmental quality and (ii) incentivize the installation of efficient firewood heaters in the residential sector.

Abstract: This paper investigates how the choice of stochastic approaches and distribution assumptions impacts strategic investment decisions in energy planning problems. We formulate a two-stage stochastic programming model assuming different distributions for the input parameters and show that there is significant discrepancy among the associated stochastic solutions and other robust solutions published in the literature. To remedy this sensitivity issue, we propose a combined machine learning and distributionally robust optimization (DRO) approach which produces more robust and stable strategic investment decisions with respect to uncertainty assumptions. DRO is applied to deal with ambiguous probability distributions and Machine Learning is used to restrict the DRO model to a subset of important uncertain parameters ensuring computational tractability. Finally, we perform an out-of-sample simulation process to evaluate solutions performances. The Swiss energy system is used as a case study all along the paper to validate the approach.