Análise da viabilidade de geração conjunta eólica/solar em Microgrid conectado à rede

Abstract

The fossil fuels depletion, the environmental concerns, the difficulty to match supply with demand, and the pursuit of more reliable systems are some of the driving forces that question the current electricity business model. In this context, the wind and solar power generation appear as clean and inexhaustible alternative sources of energy with distinct features such as strong generation potential and the possibility of being generated through decentralized sources, as distributed energy resources (DER). Despite these distinct advantages, these two sources of power generation are intermittent, hard to predict, and not dispatchable. As a consequence of such drawbacks, the energy generated through wind and solar sources may not match the real time electricity demand. Therefore, to optimize the integration of these sources in electrical systems, the key uncertainties and the market characteristics have to be taken into account. The current literature focuses mainly on wind and solar forecasting - usually for operational decisions - or on designing hybrid stand-alone systems. To the best of our knowledge, there is almost no paper that investigates these sources power potential combined with the market policies to shed light on the investment decision process. This work aims to evaluate alternatives to promote on-grid hybrid wind/PV systems as microgrids. We propose a three steps methodology as follows. First, wind speed and solar radiation databases are collected, validated and forecasted using linear and quantile regression models. Then, using the forecasted curves, three optimization models for hybrid systems design are suggested and illustrated with a case study in the city of Petrolina. The models capture three scenarios: (i) no incentive policies, (ii) net-metering and (iii) net-metering together with time of use electricity tariff. Finally, the designed systems are assessed through reliability and cost/efficiency metrics. Once the tree step approach is concluded, we investigate the impact of capital subsidies, tariff adjustments and generation uncertainty in the system feasibility and metrics. The results are contrasted with commonly used methods and indicate that quantile regression generates less forecasting errors when fitting the data. In addition, net-metering policy is interesting for a balanced design of the system, but it does not enable grid parity, which therefore fails to make PV panels feasible. We conclude that capital subsidies are necessary. Furthermore, the Brazilians time of use tariff has a negative impact in the system and reduces the technologies attractiveness because they have to generate more power during the off-peak hours to compensate the peak hours. Finally yet importantly, the tariff adjustments perspectives for 2015 and beyond place the alternative generations as option to the prevailing model, which has been increasingly expensive in the past years.