An optimal generation dispatch for a reliable and environment friendly microgrid using multi-objective optimization

Abstract

The generation and transmission system play a very important role for a low voltage microgrid, specially for one which is equipped with distributed generation systems. The generation schemes and energy sources define whether a microgrid is environment friendly or not. The renewable energy sources supported distributed generation is environment friendly. But the coordinated generation with different type of renewable power generation sources create an urgency of a cost effective dispatch for the generation systems. Otherwise, the generation system will not be enough to meet the power demand with lower cost and higher reliability. For this reason, an online dispatching technique should be designed for the energy management to supply energy to the customer with least possible cost. The presence of different types of renewable energy sources make it more difficult as the output from these sources are not constant. They vary with the sunlight, wind speed as well as other weather conditions. Moreover, they are not dispatchable sources. Therefore, a special type of algorithm need to be designed for the distributed generation system so that they can meet the power demand and also make the best use of the excessive power when the demand is lower than the power generation from the renewable energy sources. For this reason, a single objective optimal dispatching method has been discussed by proposing two different optimization algorithms for three distinguished scenarios in order to find the best possible dispatching scheme for the renewable energy sources at minimum cost. Three wind power plants, two solar plants and one Combined Heat and Power (CHP) have been considered for this thesis. The CHPs are not technologies, they are approach for applying technologies. In this study, a CHP is considered with solar and wind sources because of its cogeneration capability as well as thermal and overall efficiency. A conventional system of producing power and usable heat has 45 percent combined efficiency, while CHP method can operate at 80% combined efficiency. Therefore, combining CHP with other renewable energy sources is more cost effective and electrically efficient because of the lower efficiency associated with renewable energy sources. Moreover, it is also environment friendly for its cogeneration capability. Three different cases have been considered in terms of supplying power from the CHP and the availability of the renewable energy sources. For the first case, energy storage systems are assumed to be associated with the renewable energy sources in order to make all the sources ava xx cost function has been minimized to find the best economic dispatch for all the available power generation sources at lowest possible cost. For the second case, the same cost function has been minimized to achieve the same objective but no energy storage source is considered. All the renewable energy sources are considered as variable sources and assumed to be able to supply the highest power available at any given time. On the other hand, a third case has been considered for the renewable energy sources where no wind or solar plants are dispatchable. The same quadratic cost function has been considered as objective function to find the cost effective optimal dispatch for the distributed power generation sources. The economic dispatches obtained from these three scenarios are presented and compared in this thesis in order to find the necessity of the energy storage systems and the best distributed generation system at the lowest possible cost. Reliability or risk analysis is also very important for designing an unintervened power supply system for a microgrid. Often the rural microgrids use renewable energy sources where power interruption could occur for many reasons. Therefore, to ensure the continuous power supply to the consumers, reliability analysis should be done for the islanding mode of operation. Moreover, for supplying the power to the customers at highest reliability, risk analysis plays a very important role. Reliability analysis can help the energy management by telling the energy management system what precaution need to be taken in terms of an outage or islanding scenario. For this reasons, the amount of expected energy not served has been forecasted to the energy management system to take precaution and make necessary arrangement to reserve the amount of energy which could not be available in case of failure of any distributed power generation sources and transmission lines. A probabilistic calculation, based on Monte Carlo Simulation has been proposed in this thesis for the corresponding reliability analysis and the results are presented for the above mentioned three different cases. On the other hand, a high voltage microgrid equipped with conventional power generation sources has greater negative impact on the environment because these generators normally use fossil fuel such as coal, oil or gas. All of these fossil fuels have a huge negative impact on the environment. These fuels causes environment pollution to a great extent by emitting pollutant gas such as sulphur oxides and nitrogen oxides. Normally, these gases are emitted from the thermal units of the fossil fueled conventional generation systems. So, in this case, dispatching of the generators are not done depending on the cost of generation or fuel cost alone, it also depends on the amount of pollutant gas emitted in the environment. The primary objective of economic dispatch is the scheduling of generation units according to the load demand and forecast at lowest possible cost. But due to the recent passage of some environmental amendments, utilities are bounded to operate the thermal power generation units with least possible environmental pollution and toxic gas emission. Therefore, researchers have suggested a number of techniques and strategies to reduce the pollutant gas emission. Some of them have proposed to install additional equipment for cleaning pollution. While many researchers have suggested the utility companies to replace the highly pollutant fuels with the low emission fossil fuels. On the other hand, many of the power system researchers have suggested to use cleaner and newer fuel burners for the generation units replacing the old one. All of these suggestions require either installation of additional equipment or complete xxi replacement of the existing equipment causing additional investment cost which finally increase the generation cost and electricity price. Researchers have found a new way to avoid the additional cost in order to balance the tradeoffs between generation cost and pollutant gas emission. They have suggested that economic dispatch considering the environment pollution can solve these issue. To do this, multi-objective optimization techniques need to be implemented to find a cost effective optimal dispatch which will cause the lowest environment pollution. That is why a multi-objective optimal dispatching method has been proposed to find the best trade-off and best compromised solutions between fuel cost and pollutant gas emission from the distributed fossil fueled generation units. Reliability of the generating units have also been considered with the emission as an objective to be optimized. To address all the above mentioned issues, three different illustrations have been considered in this thesis in terms of multi-objective optimization and constraint functions. Firstly, a quadratic fuel cost and emission function have been minimized separately considering restricted generation of the power generation units in order to find the extreme solution points and diversive nature of the pareto optimal solutions resulted from the proposed multi-objective optimization method. This is done to validate the proposed algorithm and demonstrate the effectiveness of the algorithm. The fuel cost function and emission function have also minimized together as multiobjective optimization function in this case too. Secondly, the proposed multi-objective algorithm has been used to optimize three different scenarios considering the power balance constraint. In this illustration, the same fuel cost and emission function have been minimized at first. Then, the fuel cost has been minimized along with the system adequacy parameter, expected energy not served (EENS), as reliability function. System's reliability plays a great role because a huge amount of load receive energy from high voltage microgrids. Therefore, corresponding risk analysis, outage forecast as well as probable amount of unserved energy need to be reserved in the generation system. Therefore, reliability is considered here to make the proposed algorithm more effective. Furthermore, emission function and reliability have been optimized to compare the result with the previous two cases results. Finally, the fuel cost function and emission function have been minimized as multiobjective optimization function considering the power balance and reliability function as constraint functions. The results obtained in this scenario has illustrated a very good similarity with the previous results demonstrating the effectiveness of the proposed multi-objective optimization algorithm. For the system adequacy parameter evaluation, Monte Carlo Simulation based probabilistic approach has been implemented. While the reliability acts as a function, it uses this Monte Carlo Simulation to calculate the EENS for all the possible solution spaces. Then these solutions have been minimized along with the fuel cost function and emission function respectively. On the other hand, when the proposed multiobjective optimization algorithm considered the reliability as a constraint, it has used the same technique to calculate the EENS and try to keep it in a prescribed limit according to the necessity of the algorithm.