Real-time monitoring of power production in modular hydropower plant: most significant parameter approach

Abstract

The uncertainty in the water-based renewable energy systems reduces the plant capacity. However, real-time monitoring of hydropower plants ensures optimality and continuous faultless performance from the plant. But the implementation of real-time systems has always increased the overall operation cost of the power plant due to the continuous monitoring, analysis and decision-making (MAD) to assure prolonged and in situ detection and solution of uncertainties. The requirement to observe multiple indicators which represent the plant performance, elevate the cost of managing and impact the economical returns from the power plant. Also the infrastructural adjustments required to enable real-time monitoring of a power plant will also induce increased expenditure. The present study aimed to reduce the cost and infrastructural requirements of a smart system to represent the plant performance for instant mitigation of system failures by replacing the requirement of multi-indicator tracking by single weighted function monitoring. This monitoring upgradation will reduce the process cost of the system, thereby elevating the profitability of the power plant. The functional tracking will also increase the efficiency of the MAD and minimize the memory requirement of the real-time monitoring as single pointer will be required to be analysed and evaluated before taking a decision. In this aspect, an objective multi-criteria decision-making technique was used to find the significance of each indicator in hydropower production such that they can be tracked as per their potential for destabilizing the system. The results show that the new multi-criteria decision-making method which hybridizes with polynomial neural networks can identify uncertainty based on the significance of parameters by a portable and independent platform that can be integrated with supervisory control-based systems to monitor uncertainty in a hydropower system. According to the results, operation and maintenance cost followed by the discharge indicator was found to have the highest significance among the indicators considered in the study. The results depict that the new multi-criteria decision-making method with polynomial neural networks can identify uncertainty based on the significance of parameters with the help of a portable and independent platform that can be integrated in supervisory control systems to monitor uncertainty in a hydropower system at real time.