Contributing to the energy security of the UK via STOR – the Short-Term Operating Reserve

Author: Mike Brown
Posted: 19 Jan 2015 | 11:03







Transformer park, Advanced Computing Facility Plant Room C.

The media has reported recently that there is now a potential significant shortfall in UK electricity generating capacity due to the decommissioning of legacy fossil-fuel and nuclear stations and the continued delays in the provision of viable low-carbon alternatives.

The UK electrical demand normally varies (cyclically over a 24-hour period) in the range between around 30GW and 55GW. Although the current maximum theoretical installed capacity is around 80GW (but falling year on year) this includes capacity that is mothballed or due for decommissioning. The safe operating margin between demand and actual available capacity such that the availability of supply can be assured has become significantly squeezed (especially when the considerable distribution losses – which almost equal the amount generated by "renewables" - are factored in). A few years ago the margin was 17% - today it is estimated at less than 5%.

The distributed nature of the UK’s generating capacity, and the resilience and capacity of the National Grid, means that despite the fragmented nature of the generating and distribution industry (a legacy itself of the privatisation policies of the 1980’s) the ability of the existing installed generating capacity to maintain supplies under normal circumstances is generally very high. Problems may occur when circumstances are not normal - perhaps due to serious weather conditions, unplanned shutdowns of complete generating units (such as after the recent fire at Didcot 'B' or the discovery of cracks at three nuclear stations - Hunterston 'B'Hartlepool and Heysham '1') or a combination of events.

The major existing sources of conventional generating capacity such as the coal-fired Longannet (2.4GW) or the nuclear Torness (1.3GW) – operate most efficiently at maximum capacity, thus maintaining idle (but available) additional capacity on standby is a very expensive business. A coal-fired power station has to maintain steam pressure whether the turbines are spinning or not – and the same goes for CCGT (combined cycle gas turbine) based power stations (the other main source of non-nuclear thermal generating capacity). Accordingly there is a high degree of judgement required as to whether (and when) standby capacity should be brought to availability (and risk not being used while still incurring operating costs and releasing carbon), or whether the standby capacity is stood down (with the risk of interruption to supplies if the demand unexpectedly goes above the available capacity).

Pumped-storage systems (like the 440MW installation at Ben Cruachan in Argyll or the 1.7GW station at Dinorwig in Snowdonia) comprise the only effective energy buffering solution available to the National Grid and do provide near instant availability of (very efficient) standby power but the margin between growing peak demand and shrinking maximum capacity continues to narrow, and so other solutions have had to be sought to de-risk the operation of the electrical grid and maximise the security of supply.

One of these is Short Term Operating Reserve (STOR) which is an initiative that makes available to the National Grid a portion of the very significant amount of installed standby generating capacity that usually sits idle at commercial and industrial sites across the UK. Look round the back of your local Tesco – and you’ll probably see a small container housing a standby generating set. National Grid aims to have typically between 1.8GW and 2.3GW of STOR capacity on operational standby 24 × 7 and available for use when required at short notice – equivalent to a medium-large conventional nuclear or coal-fired generating station.

Part of the University’s contribution to the expansion of the Advanced Computing Facility (ACF) in support of the Archer service was the provision of a 2MVA standby diesel generator. This unit is designed to back-up the UPS-protected services and ensure continuity of supply to essential and critical loads in the event of a significant loss of power at the site. Unlike many smaller standby generators – such as the two other standby generating sets at the ACF (which can only operate under load when the mains supply has been isolated) - Generator no 3 has been installed and configured to the electrical supply and distribution industry’s G59 protection standards which allow the unit to operate while synchronised and in concurrent parallel operation with the mains supply from the grid. This means that synchronisation equipment had to be provided as well as approved fault-protection mechanisms to ensure that the operation of the generator does not enable faults to propagate out onto the public high-voltage network and, crucially, that in the event of an external fault on the network that the generator is feeding that the protection relays will disconnect the source of supply from that network. The entire installation has to be checked, trialled and approved by the DNO (Distribution Network Operator) which in our case is SP Energy Networks.

 Generator housing, Advanced Computing Facility Generator no 3.

In the originally installed configuration, the generator would start only on the detection of a mains failure condition and the normal mode of operation ("island mode") meant that the generator would operate under load only while the mains supply was unavailable or isolated (although synchronised parallel generation, through G59, is permitted when undertaking regular on-load tests of the equipment).

STOR allows approved generating equipment to be operated in production concurrently with the normal incoming mains supply. Under a STOR agreement, generators like that at the ACF (and many hundreds of similar machines across the UK) can be brought on-line remotely to provide a near-instant uplift in capacity across the grid (eight hundred of these machines would generate more than all of the output of Torness) and even if they are only required to run for a short period (such as around the advertising breaks during a major TV spectacular - although typically they will operate for up to two hours) then it is far more efficient than bringing extra boilers on-line at a coal-fired station, or firing up an open-cycle gas turbine station from cold to operate in standby. Older coal-fired power stations may require up to 12 hours’ notice before they can generate, and unlike maintaining a conventional power station on operational standby, the STOR generating capacity releases no carbon until it is actually engaged and brought on-line.

Capacity against demand on the National Grid is regulated through the supply frequency (nominally 50Hz). A rising demand will cause the frequency to fall (indicating capacity starvation) while a falling demand will induce a frequency rise (indicating excess capacity that needs to be shed). STOR capacity is typically invoked by National Grid when the frequency is falling and other measures (pumped storage, spinning reserve (from partly loaded conventional power stations), and Frequency Service - whereby large industrial consumers can be paid to elect to drop off-line automatically when the frequency falls below a defined threshold) have failed to achieve frequency stability.

Bringing the ACF generator on-line under STOR is triggered through receipt of a remote enable signal to initiate the start of the diesel. Only once the generator had synchronised with the grid will the supply breaker close. At this point the unit will be generating in parallel with the grid supply and it will immediately run up to its full 2MVA capacity. It typically takes less than 30 seconds between receipt of the enable signal (a "call" in National Grid parlance) and the unit being on-line and generating at full capacity. Since 2MVA is typically less than the site baseload, the actual effect is that the amount of power being imported into the site from the grid is simply reduced by that amount – we would only export power should the site’s own demands be less that the output capacity of the generator. Either way – the load on the grid has been reduced by 2MVA.

Why should the University be interested in providing part of the UK’s standby generating capacity? There are two good reasons why it is in our interest to be part of this scheme. Firstly, the University gets paid a standing fee for making its capacity available (and charges per-unit costs when actually on-line) and so an income stream is generated that helps to offset maintenance and associated running costs. This turns what had been a cost item and a maintenance liability into a revenue earner. Secondly, regular use of the generator will ensure that it remains fully in the correct operating and serviceable condition (starting and operation of the machine, exercising of supply breakers, turning over of fuel stocks etc) – plant should never be left idle and the “standby generator” that is never spun until the time it is needed in earnest has a not insignificant likelihood of failure at the very time it is needed.

The ACF generator went live as an available STOR resource at the beginning of January and the first live production use of the generator under STOR was on 12th January with the "call" signal being received at 1650 and the "cease" at 1810. It is expected that the generator will be brought into operational use for around 50 hour’s total runtime in any year.

The generator is in live readiness within defined service windows (one in the morning and one in the late afternoon/evening) – with the windows being longest during the winter when demand is at its peak, with a maximum continuous run-time of two hours.

ACF Generator no 3 is one of three similar systems that are planned to be made available by the University as STOR resources. Our agreement is with a STOR aggregator (who manage the resources from multiple sites so as to provide National Grid with a "virtual power station") and who handle the calls from National Grid and selects appropriate generating units to meet the requested capacity demands.

Although normally made available to STOR, the standby requirements from the site always take precedence and we retain complete operational control of the equipment despite its allocation to STOR. Should we need to take the system off-line for any reason (such as maintenance) at a time when it might have been called for under STOR, the aggregator will be aware of our temporary unavailability and simply select another provider, although generally maintenance and on-load run testing activities will be scheduled to take place outside of the defined service windows so as to maximise our availability.

From a personal point of view, making available Generator no 3 to STOR is a rather satisfying outcome. I originally went to University with a view to embarking on a career in power engineering. After a direction change leading to a lengthy (but at times amusing) diversion through the provision of high-performance computing services, it is rather fitting that I end up after all, in effect, managing operations at a UK power station directly connected to the National Grid!

The prime mover is a 61litre V16 turbocharged Perkins diesel of 2,300bhp coupled directly to a 3-phase 50Hz generator. Power is generated at 415V but stepped up to 11,000V and exported through the local HV network and potentially out onto the public HV network should the site demand be less than the generator’s output when operating. The nominal output is 2.0 MVA (1.6MW at 80% PF). The combined unit is installed in a purposely-designed container that also contains starting batteries, isolation and protection equipment and control gear as well as the primary fuel tank. A much larger secondary fuel tank is installed adjacent to the generator housing.


Mike Brown, Operations & Services Manager, EPCC