Network Operators’ Ownership of Storage

DNOs and TSOs CAN own storage. In the 2019 Energy Directive (to which we remain a party at least until the end of the year), any grid operator can request a derogation from their regulator to build and own storage if insufficient is being built. This is a once-off opportunity which your members can seize.

Orkney Project

The costs and benefits of batteries were evaluated in relation to the Orkney Energy Storage Park project by SSE – specifically the 27 June 2016 document entitled Trial of Orkney Energy Storage Park (Phase 2) Closed Down Report. It is on the basis of such reports and analyses that the DNOs come to these views of CAES, which is completely off the mark. Our review is as follows.

Storage Evaluation

It’s interesting in evaluating a stand-alone battery (small scale, short duration) with 4 revenue streams. Ours is CAES (large scale, long duration) with 12-18 revenue streams (most of the 6 possible ones are subjects of Pathfinders at present).

A short-duration battery (< 4 hours) can only balance brief spikes in supply and/or demand. Long-duration storage (> 4 hours) can balance an entire evening peak on a windless winter evening. Therefore they do entirely different jobs, in addition to the long-duration storage being able to provide most of the services of a short-duration battery. The one service long-duration storage does not provide is exceedingly quick response time (which used to be Enhanced Frequency Response, EFR); however it does provide natural inertia that makes EFR unnecessary.
Moreover, batteries have a life of 6-8 years; ours is 40-50 with a mid-life overhaul.

Battery suppliers measure their efficiency (typically 85-92%) terminal-to-terminal on day 1; we measure ours (63% @ 40MW, rising to 70% @ 500MW) grid-to-grid. The differences are that batteries incur numerous ancillary loads that we don’t, of which the biggest is heating/cooling and the second is power conversion. These can reduce round trip efficiency by 15-30% absolute.

  • Northern Power Grid measured the difference between nameplate and measured efficiency, finding in 2017 that measured grid-to-grid efficiencies were 42-60%, though technological advances would have reduced those differences;
  • However, they measured efficiency on “day 1” rather than at mid-life;
  • And their cooling load was measured in Northumbria, not renowned for its sultry weather – hotter locations would require more cooling;
  • Larger installations also require increased cooling;
  • Batteries would benefit from construction at renewable grid connection by eliminating their power conversion (AC-DC-AC) losses, typically 7-10%.

During the life of the battery, cell efficiency (losses) and capacity (revenue potential) deteriorate such that they need three times the cooling (ancillary load) by swap-out in year 6-8. Our efficiency deterioration is of the order of 2% over the life of the plant – just like a power station; our capacity doesn’t deteriorate and nor do our ancillary loads grow other than the mild deterioration in efficiency.

Building Storage In Conjunction with Renewables

If large-scale long-duration storage is built at (or operating in conjunction with) the grid connection of a renewable farm, it has a number of benefits not evaluated here including:

  • Capex offset
    • Halving (for wind; reducing by 2/3) grid connection size (and hence grid reinforcement costs) for the renewables;
    • Eliminating grid connection and reinforcement costs for the storage
  • Opex offset
    • Eliminating UoS charges for export from the farm
    • Eliminating UoS charges for import to the storage
    • Residual UoS charges for export from the storage are half those of the wind farm (1/3 for solar)

Therefore the business case of our technology and proposed setup is very different. and needs modelling and evaluating differently.

DNO Benefits

The benefits to the DNO are very different when comparing our proposed (at/before grid connection) and Orkney setup:

  • Orkney requires full grid capacity for the renewables, and again full capacity for the storage; we reduce grid capacity required for renewables and then piggy-back on that (i.e. no additional capacity needed) with the storage.
  • This has knock-on effects throughout the grid.

The constraint management evaluated by the Orkney project is merely reduction in curtailment, because both size and duration are small. Our size and duration are sufficient to enable grid reinforcement avoidance, and therefore our constraint management benefits are immensely larger.

Our storage also provides grid benefits not offered by batteries, or not evaluated by the Orkney project, including:

  1. Stability services (real inertia, real voltage and frequency control, all 24/7 regardless of whether charging or discharging) – while eliminating or reducing the DC-connected in-feed to the grid.
  2. Ancillary services (real reactive power/load).
  3. Black start, if designed/contracted in at the outset.
  4. All the services currently being tested on PicloFlex.

All this means that:

  1. Storelectric’s CAES is of potentially huge benefit to the DNO;
  2. It offers sufficient benefits to the renewables operator to ensure their cooperation and potentially capex cost-sharing;
  3. It has sufficient revenues on its own not to be a net cost to either DNO or renewables operator – indeed, it has sufficient return on investment to interest an outside investor provided that contracts are of sufficient duration.