The construction of efficient energy storage systems is key to the efficient and flexible operation of power grids in various countries. Currently, existing power grid systems worldwide cannot handle large-scale intermittent energy integration. When the penetration rate of intermittent renewable energy exceeds 20%, it can greatly disrupt the stability of the power grid, leading to severe damage. The current consensus is that large-scale electrical energy storage systems can effectively alleviate many inherent inefficiencies and defects in power grids, improve grid reliability, promote the comprehensive integration of intermittent renewable energy, and efficiently manage power generation [1]. The cost-effectiveness of liquid flow battery energy storage systems for power grid applications has been widely studied. Avista Turner EES is a typical example of a successful vanadium redox flow battery project in the United States. This paper will use this demonstration project as a case study to introduce and analyze the actual operation situation of liquid flow battery projects.
"Avista Turner EES System Overview"
The Avista Turner EES system is a vanadium redox flow battery system with a capacity of 1 MW / 3.2 MWh, located at the Turner substation in Pullman, Washington. It supports the intelligent campus operations of Washington State University and was purchased from United Energy Technologies (UET) [2]. The system consists of two sub-systems with an actual output of 0.5 MW / 1.6 MWh each, made up of four battery modules (the first four modules in Figure 2) and a power conversion system with a battery management system (BMS) (the fifth independent module in Figure 2). Each subsystem operates independently without interference and has external storage tanks for electrolytes. The rated discharge power of a single system is 500 kW, the rated charge power is 400 kW, and the peak power is 600 kW. Each battery module is composed of three 50 kW stacks connected in series, containing 50 cells connected in series within the module. The open-circuit voltage limit for a single cell is 1.25 V/cell (0% SOC) - 1.49 V/cell (100% SOC). According to data from Washington State University, the cost of this VRB-EES system is $700 per kilowatt-hour [3].
（Figure 1）
（Figure 2）
"Avista Turner EES System Technical Specifications [4]"
The main technical parameters for the Avista Turner EES system are as shown in Table 2.1. For the UET full system, its rated power and peak power are 1 MW and 1.2 MW respectively, with a maximum output energy of 3.2 MWh, a system efficiency of 65-70%, a design lifespan of 20 years, a DC voltage range of 465V-1000V, AC output range of 4160V—34.5kV, and suitable operating temperature range of -40℃—50℃.
For the next generation of EES systems used in MESA 2, its main technical parameters are as follows: Table 2.2 shows that its AC charging and discharging efficiency is 70%, the output AC voltage is 12.47 kV, the response time is less than 100 ms, the footprint is 78 m2, the dimensions are 12.5 m x 6.1 m x 2.9 m, the total volume is 221.125 m3, the weight is 170,000 kg, the volumetric energy density is 9.9 Wh/L, the volumetric power density is 2.7 W/L, the mass energy density is 12.9 Wh/kg, the mass power density is 3.5 W/kg, the design service life exceeds 20 years, the operating temperature range is -40°C — 50°C, and the self-discharge rate is less than 2% of stored energy.
The discharge capacity of the Avista Turner EES system and RTE
Regarding discharge capacity, under different discharge rates (fixed 600kW charging power), the discharge energy varies nonlinearly with SOC, and the discharge energy capacity range is between 2020-3600 kWh. Discharging at a rate of 1000kW provides approximately two-thirds of the rated energy of 3200 kWh, while discharging at 520kW provides 94% of the rated value, and at 400kW, the discharged energy exceeds the rated energy by 4%. On the PCS DC side, excluding auxiliary loads, the vanadium redox battery ESS system delivered its full rated energy at a discharge power of 520kW, and at 400kW, it provided energy that was 11% higher than the rated energy.
For RTE (round-trip efficiency), in practical operation, Subsystem 1 had a cumulative energy output of 125 MWh after 78 complete discharges, and Subsystem 2 had a cumulative energy output of 165 MWh after 103 complete discharges (as shown in Figure 2.3). The accumulated RTEs (charge-discharge efficiencies) without considering auxiliary loads for both subsystems showed some differences, being 52% and 60% respectively, indicating good system stability (as depicted in Figure 2.4).
Figure 2.5 shows the cumulative charging and discharging capacity of subsystem 2 as well as the coulombic efficiency. Initially, the coulombic efficiency is close to 100%, gradually decreasing over time and reaching 95% at the end of the test, indicating electrolyte cross-penetration in the system. Electrolyte cross-penetration is an inevitable issue in vanadium redox flow batteries due to the presence of proton exchange membranes and concentration differences on both sides, causing a certain degree of crossover between positive and negative electrolytes. In the scientific community, efforts are made to minimize ion penetration by optimizing the composition of ion exchange membranes, making highly fluorinated sulfonic acid proton exchange membranes essential for high-performance flow batteries and significantly increasing the overall cost of all-vanadium redox flow batteries. Although some non-fluorine ion exchange membranes have been proposed, their overall performance is not satisfactory and requires further exploration and research. From the results of actual operation of Avista Turner's EES system, the coulombic efficiency maintains a relatively high retention rate, suggesting that the current impact of ion crossover on the overall operation of the system is minimal, and the control effect of ion exchange membranes on ion crossover is good. This phenomenon can also be further improved through regular maintenance.
The DC-DC RTE of Subsystem 2 as a function of SOC is shown in Figure 2.9, overall, the bending part of the DC RTE curve appears at the low SOC part and changes less in the high SOC area. The RTE decreases with the decrease of SOC and increases with the decrease of power, and is the highest at 520 kW. For example, at 90% SOC, the RTE at 520 kW is 84%, while at 32% SOC, the RTE is 76%. At 1000 kW discharge, the RTE at 90% state of charge is 73%, while at 50% state of charge, the RTE is 61%.
Avista Turner Energy Storage System (EES) Response Time and Internal Resistance
The response time of the Avista Turner EES system ranged from 3 to 10 seconds during the testing period. For the charging process, as the State of Charge (SOC) increased from 20% to 60%, the response time increased from 4 seconds to 10 seconds, reaching a maximum power output of 800 kW, which then linearly decreased to 400 kW until the SOC reached 100%. The response time range was between 340 kW/s at 90% SOC and 315 kW/s at 30% SOC, with a maximum power output of approximately 1000 kW within the range of 90~40% SOC, while at 30% SOC, the response time slightly decreased to 950 kW.
For the internal resistance of the system, data showed that as the SOC changed from 90% to 40%, the internal resistance increased from 0.100Ω to 0.110Ω. At 30% SOC, an internal resistance peak of 0.125Ω was observed. After standard conversion, the resistance range was from 0.05Ω to 0.055Ω, consistent with research results for Snohomish's MESA 2 systems. Within the SOC range, the internal resistance during charging periods was slightly lower than during discharging, decreasing from 0.11Ω at 20% SOC to 0.095Ω at 100% SOC.
Avista Turner EES System Availability
During the testing period, the overall availability of the flow battery ESS was 56%. The total test duration was 365 days, with 162 days or 44% lost due to various reasons, distributed across different categories as shown in Figure ES3. Among these losses, issues related to the stack, including mismatched stack SOC and stack leakage, accounted for 16% of test time loss (58 days); problems associated with the Power Conversion System (PCS) led to a loss of 11% of test time (40 days), including long-term use of the PCS during high SOC charging and exposure to leaking electrolyte causing corrosion of electronic components; issues related to PCS software resulted in a loss of 9 days; pump-related problems and cylinder housing leakage caused losses of 9 days and 8 days, respectively; failure of thermal management led to a loss of 7 days, while inability to remotely reset AC breakers caused another 7 days of loss; human error and weather contributed to losses of 6 days and 7 days, respectively; maintenance, communication failures, and miscellaneous causes accounted for a loss of 11 days.
Main Operating Status Conclusions
Through the review of publicly available materials, we find that the operating characteristics and performance of vanadium redox flow battery energy storage systems (EES) can meet the requirements. The main operating status conclusions are as follows:
(1) The Avista Turner vanadium redox flow battery ESS system has different energies at different power levels, with a maximum energy of 3395 kWh at 420 kW power, and a discharge depth of 81%.
(2) The available energy for discharge is highly dependent on the power (kilowatts) level during the test cycle. For example, the available energy at full rated 1 MW power is about two-thirds of the energy that can be released at 50% rated power.
(3) Response time depends on the power level, mode, and state of charge (SOC).
(4) The availability factor of the vanadium redox flow battery ESS system is lower than expected.
We have reason to believe that with the continuous development of vanadium redox flow battery technology, practical operation of vanadium redox flow batteries can be achieved and widely promoted, playing a significant role in the coordinated operation of renewable energy and the grid.
References:
[1] Gür, T. M. (2018). Review of electrical energy storage technologies, materials and systems: challenges and prospects for large-scale grid storage. Energy & Environmental Science. doi:10.1039/c8ee01419a[2] Vincent S. Office of Electricity Grid-Scale Energy Storage[EB/OL]. (2018-10-18)[2024-10-14].[3] Chris Q. Case Study: Avista Turner Flow Battery Energy Storage System[EB/OL]. (2023-10-12)[2024-10-14].[4] A Crawford, P Balducci, V Viswanathan et al. Avista Turner Energy Storage System: Assessment of Battery Technical Performance[EB/OL].(2019-7)[2024-10-14]
