By Calvin Stephen
The world challenges of global warming and climate change have led to the need for cleaner ways of generating energy. Hydroelectric technology has been a beacon of light in minimizing carbon footprint in both power generation and water supply - the water-energy nexus they call it. In particular, pumped-storage hydro makes use of Pumps-as-Turbines (PATs) to meet power supply needs during peak hours and pumps water to a higher altitude for later use during off-peak hours. PATs are critical in this type of a system, and any failures can lead to undesirable consequences.
With the advent of the industrial internet of things (IoT), information on PATs failure status can be obtained well in advance and can facilitate important decisions on when to carry out maintenance. Machines have produced data for ages. With the decreased cost of sensors, now is the opportune time to make use of these data to ensure machines operate efficiently and effectively. This data availability also applies to PATs. The use of predictive maintenance helps to avoid issues and surprise shutdowns that can have detrimental effects on both the machine and the end-user. Predictive maintenance involves the use of smart sensors and actuators to enhance maintenance. Data from these sensors can help operators to make decisions on the optimal time to carry out maintenance activities before the machine failure damages the machine itself. As a result, the end-user does not have to experience power outages during peak hours as the machine is available to cover the excess demand. The other advantage associated with predictive maintenance is that operators have all of the data in one place, to help deal with repetitive maintenance problems. So, there is an opportunity to design out the problem and to improve machine uptime. PATs, like any other hydro-turbo machines, are prone to mechanical, electrical and hydraulic failures. Mechanical failures include unbalance of rotating components, misalignment at couplings, defective bearings, looseness, mechanical shock, soft foot, impact or fretting. Electrical failures include failures in the generator, such as air gap eccentricity, broken rotor bars, unequal distribution of air-gap flux, inter-turn faults, shorted or open stator and rotor windings, unequal phase currents, magnetostriction and oscillations of torque. Finally, hydraulic failures are associated with fluid flow conditions that lead to cavitation, aeration, over pressurisation and excessive heat. Numerous methods can be used in predicting the machine’s condition in order to decide on whether or not to carry out maintenance activities. Vibration monitoring is a widely used technique to detect most machinery failure conditions. Vibration analysis is well suited as the data being collected from the vibration signature of the machine can provide both maintenance information and insights into operating conditions. Other techniques include acoustic emissions, lubricant analysis, ultrasonic analysis, motor current analysis and thermography. PATs require the utmost care to prevent its failures in operation. PATs are critical components of micro-hydropower systems and their failure in operation can yield unwanted results: for example, no power supply during peak demand power phase. Failure also means that micro hydropower systems cannot deliver on its intended purpose. Hence, there is a critical need to ensure there is in a healthy state of operation always. PATs have been lauded in literature for being associated with fast return on investments, but failures can hinder the achievement of this. As such, the application of predictive maintenance is intended to ensure a fast return of investment and optimize its operation. The application of digitization to hydropower will be a game-changer, and will lead to a technological leap even on PAT-based micro hydropower systems.
Figure 1 information flow & exchange in a digital avatar of hydropower turbines
Figure 2 Research plan
Kougias et al [1] make use of the phrase “digital avatar” (shown in figure 1) to refer to the digital model of the unit made from gathering the corresponding information into a comprehensive set of data and using it to support the flexible unit operation. The digitisation of hydropower will allow gathering of actual realtime data from PATs operation to enhance decision making on its design, development, operation and maintenance. The use of advanced tools in data analysis, advanced modelling, lifetime prediction, predictive maintenance and condition monitoring will dictate the achievement of better overall equipment effectiveness, safety and reliability hence increasing energy production.
The initial stage of my research involves developing digital models of the PATs operation. These models will aid in simulating operation failures to create an in-depth understanding of how PATs respond under different failure conditions, as shown in figure 2. I will also carry out experiments to validate the digital models and to provide confidence in the outputs. This validation is an initial step in creating a PAT digital avatar that will be integrated into both predictive maintenance and condition monitoring. At the final stage, the research aims to create the same digital avatars for our demonstration sites incorporating IoT principles.
References
[1] I. Kougias, G. Aggidis, F. Avellan, S. Deniz, U. Lundin, A. Moro, S. Muntean, D. Novara, J. I. Perez-Diaz, E. Quaranta, P. Schild and N. Theodossiou, "Analysis of Emerging Technologies in the Hydropower Sector," Renewable and Sustainable Energy Reviews, vol. 113, p. 109257, 2019.