Issue link: https://fhpublishing.uberflip.com/i/819678
NETWORK / 27 / may 2017 Compressed-air energy storage Compressed-air energy storage (CAES) may not hit the headlines – or even fire the imagination – like other technolo- gies, but it is exceptional insomuch as it can bridge applications for fast frequency response and generation, especially when connected to intermittent solar and wind generation. Ireland has a total capacity of 3.7GW from 272 wind farms, and it needs to har- ness that power. That is where Gaelectric's planned 330MW CAES plant comes in. The island has abundant renewable re- sources, and its natural salt caverns on the wind generation-rich east Antrim coast are the ideal site for a CAES system. The system's ability to cope with Ireland's high renewable penetration, its flexibility in terms of services available, long deployment window (6-8 hours) and fast response will make it a good fit with renewable genera- tion, particularly compared with other types of storage such as batteries, says Gaelectric's project manager Mark Byrne. From a standing start, CAES takes about 10 minutes to reach full output but can be on grid within six minutes. Quick compared with coal and gas-fired plants, but a slouch compared with other types of storage. To counter this, Byrne says the CAES plant could be connected to the grid at a mini- mum generation level of 10% of peak out- put, this allows it to respond in less than a second to a frequency event. Byrne says Ireland is the ideal applica- tion for CAES because of the technology's flexibility, not just for peak 'lopping' or shi–ing, but also for halting curtailment, contributing to system stability, offsetting traditional peak demand assets and offering a solution to system inertia. CAES technology is also proven. A 320MW system in Huntorf, Germany has been running since 1978. Air in a cavern is compressed – Gaelectric's project includes two caverns in salt deposits more than 1,400 metres below ground – to about 70 bar. The compressed air can be released through an air turbine that drives a generator. Gaelectric is looking at three revenue streams. Two will involve successfully bidding in the DS3 system service auc- tion, and the third is energy trading –the company plans to store energy purchased at low-cost times or periods of excess high generation, and then sell it back to grid operators at times of low wind generation or high-costs. a substation on site to an offsite substa- tion near Pentir, Wales. The scheme at Glyn Rhonwy will be connected at the DNO level, with the connection being the responsibility of Scottish Power Energy Networks. Despite relatively high deployment costs, payback times are about 15 years. The busi- ness proposition includes peak-shi–ing, black-start, grid balancing, triads and FFR, "but mainly grid balancing," says Holmes. Response time is measured in seconds, and the facility will run for 40 years before its turbines need changing at a cost of about £50 million. The design life of the project is 125 years. ever, lithium-ion remains less cost effective for providing longer responses where large energy volumes (such as MWh) are required, such as peak shi–ing, says Laguna. The project also showed how storage can be used to defer reinforcement costs, about £6.2 million in the case of SNS. As with all storage, the project specifics determine the size of the storage unit, which in turn determines the cost of the system. In the case of the SNS project, a payback time of less than ten years was expected – important because the life of the asset was ten years. It means that payback of an investment, such as SNS, will ultimately depend on the extra revenues obtained from providing services to the whole system, says Laguna. The project showed that grid-scale energy storage could be commercially viable as battery costs fall and revenue streams become accessible. Laguna says storage in general has great benefits, not only for DNOs but also for the whole system. week, not months or years, adding that the technology can be deployed where a weak point in the grid has been identified, effec- tively deferring reinforcement costs. SILVER BULLET? There's no single solution for storage. Here are three benefits of each of the technolgies discussed LITHIUm-Ion l Cost EV adoption is continually driving down the cost of batteries. l Scalable Can be used from the kW to mW scale depending on the application. l Small footprint Lithium-ion's power density makes it a winner when space is a priority. REdox fLow BaTTERy l Scalable from 5kW up to 100mW, so it can match a DNO's requirements. l Costs Limited maintenance needs and potential 40-year life make it cost effective in the long term. l Transportable Can be deployed to rectify a weak point in the network within a week. pUmpEd SToRagE l Speed Can be deployed from 0 to 1,728mW in 12 seconds. l Long life High start-up costs offset by a 125-year design life. l maturity The technology is tried and tested after more than 30 years. CompRESSEd-aIR l Large-scale Can be deployed at the triple-digit megawatt scale. l Long life although start-up costs are relatively high, a long deployment life brings costs down. l flexible Can be used for balancing services as well as capacity storage. LEad-aCId/fLywHEEL (oVERLEaf) l Green credentials The flywheels are under two metres tall and almost silent. l Cost an estimated 20-year lifetime makes the system cost effective. l Speed The system can provide rapid frequency response and voltage control.