Extending lifecycle of flywheel energy storage via average consensus algorithm

Researchers at the Inner Mongolia University of Technology, in China, have developed a new lifecycle parameter that can reportedly help increase coordinated control and service life in flywheel energy storage array systems (FESAS).
Compared to other mechanical energy storage technology, such as pumped hydro and compressed air systems, FESAS have higher values for specific power and energy, power and energy density, lifecycle, efficiency, self-discharge rate, and energy capital costs. By contrast, FESAS have lower values for lifespan, scale, maintenance, and power capital costs. Their impact on the environment is extremely low, while pumped hydro and compressed air energy storage systems have a medium-to-low and a high-to-medium impact, respectively. It is claimed FESAS effectively achieve power smoothing, frequency regulation, and improved power quality.
“A flywheel is an energy storage technology that transforms mechanical energy to electrical energy, and vice versa, through high-speed rotor rotation,” the scientists explained. “During charging, a motor converts electrical energy into the rotational kinetic energy of the flywheel, increasing its angular velocity. During discharging, a generator converts the rotational kinetic energy into electrical energy, decreasing the flywheel’s angular velocity.”
The proposed approach is based on a new type of “doubly stochastic Perron matrix algorithm,” which is in turn relying on an error-balancing iterative algorithm. The latter is an algorithm that facilitates the convergence of auxiliary variables to an average consensus and effectively suppresses unbalanced power, according to the researchers.
The novel algorithm is said to operate with low computational complexity and fast convergence while allowing each flywheel to operate independently, eliminating single-point failures and enhancing system reliability.
“By distributing errors uniformly, our technique reduces errors and ensures that the sum of the row and column elements of the matrix are closer to 1, minimizing error accumulation,” the research team said. “Additionally, constructing doubly stochastic matrices facilitates more uniform information propagation, enhancing the efficacy and stability of the consensus algorithm and making it more conducive to realizing global average consensus, compared to column-stochastic matrices.”
The lifecycle-based average consensus algorithm was tested through a series of simulations based on the operation of a wind farm connected to six 250 kW FESAS, each with an energy storage capacity of 50 kWh, a maximum rotational speed of 7,200 rpm, and a minimum rotational speed of 3,000 rpm.
“Our findings reveal that within a certain range, a larger convergence factor results in faster system convergence,” the academics said. “For both undirected and directed graph topologies, FESAS can effectively suppress unbalanced power only when each auxiliary variable achieves average consensus. In an undirected graph structure, the auxiliary variables converge to [an] average consensus.”
The academics added, the new algorithm can be used for battery and supercapacitor energy storage, and in distributed energy systems.
The findings can be read in the study “Research on the strategy for average consensus control of flywheel energy storage array system based on lifecycle,” published in the Journal of Energy Storage.