An increasing amount of our work at the CCC is focused on emerging economies, since this is where most of the interesting growth in coal use is taking place. And so sometimes it is refreshing to take a short commute to a local meeting looking at far more regional issues. The relatively new Centre for Energy Policy (CEP) at Strathclyde University in Glasgow, Scotland, is still somewhat finding its feet and defining its purpose, but is increasingly producing reports, blogs and media appearances to provide expert but unbiased information on national energy issues. The CCC sits on the advisory board of the CEP and I must admit to thoroughly enjoying being able to attend and provide input to their meetings.
Simon Gill, of Strathclyde Uni, gave an excellent and thought provoking summary of the balance of the grid in the UK, highlighting the many ways the government was working to ensure that supply meets demand and that output is reliable, despite the growth in intermittent sources such as wind. The talk was particularly timely considering the frequency response issues encountered in the UK last month when the interconnector with mainland Europe dropped 1 GW of power from the grid with little or no warning. A similar significant power drop situation occurred earlier this week in Australia where 1.7 million residents were left without power for several hours, leaving the grid operators with many questions to answer
There is a chain of reactions which occur when electricity demand and supply do not match. The UK grid is balanced in less than 1 second intervals, although a leeway of a few seconds is achieved due to latency in the system (kinetic energy in rotating systems and inertial response lags). Once supply drops, due to a drop in wind, an outage at a baseload plant, or even a sudden unexpected increase in demand, a chain of response options kicks into play. This includes a series of back-up capacity options ranked in order of speed of response. The National Grid works with sources in advance to ensure that sources and costs have been identified and agreed in advance and are guaranteed to provide back-up power, should it be required.
First there is the high frequency response, systems which can react within 10 seconds or less and which are expected to produce up to an additional 10 MW for over 20 seconds. The UK has around 800-1,600 MW of this available at any time, agreed in advance by the grid operators. These tend to be stand-by steam systems such as CCGT (combined cycle gas turbines), OCGT (open cycle gas turbines) or coal plants – plants which are already running and have excess capacity on standby. Over and above this, around 1,200-1,600 MW of fast reserve systems are available, providing up to 10 MW within 30 seconds for at least 30 minutes. Again, this stand-by power tends to come from large generators which are already online. Hydro power has a stand-by time of around 3-4 seconds. As discussed my the previous report on intermittency, coal-fired plants which are already running can ramp power output up and down very rapidly, but, if allowed to turn off and cool, start-up times increase from hours into days and become very costly.
If these frequency response operations are either already in use or not available, then STOR (short term operating reserve) kicks in. Primary STOR systems are expected to provide >50MW of power within 2 mins and maintain this for at least 15 mins. The STOR section of the supply chain also includes systems which can respond within 4 hours, maintaining this >3 MW output per system for up to 2 hours. The STOR market caused controversy earlier this year when 2-3 GW of capacity was procured in the form of small scale back-up generators, including relatively highly polluting diesel gen-sets. These systems can be established purely for STOR purposes but, more often, they are generators which are already acting as back-up in locations such as hospitals and industrial sites. These locations have the option to sell the potential to use their back-up systems to the grid for use in times of need. The payment for power from these systems can be high – up to £250/MWh. However, these systems are not used very often. There are occasions when using the STOR capacity is actually more cost-effective than the balancing market, which can hit £1,000/MWh at times of peak or emergency demand. However, STOR is only used around 1-2% of the time and provided around 0.1% of the power demand of the UK in 2014. Studies have shown that using this small capacity for small periods can actually produce fewer CO2 emissions than the “headroom” emissions from baseload fossil fuel plants running at high capacity to provide the same stand-by availability. Clearly there is a complex balance between keeping large but efficient plants running when they are not actually needed versus running those plants only to the extent required but having to call upon smaller and arguable less efficient sources of power when demand increases.
Finally there is a contingency balance reserve – power systems that can react within 2 hours to produce 2500-3500 MW for at least an hour. These tend to be large sources such as moth-balled generators. Up to 3.5 GW of this capacity is available in the UK at the moment. The “balance” part of the contingency balance reserve includes the option for load shedding, where large consumers or aggregates or consumers have agreed (for a price) to reduce or halt demand during periods when supply is unusually low.
The main capacity market is also complex. Plants in this sector are only required to change their output by up to 2 MW within 4 hours periods for the duration of whatever stress event is occurring. Should a plant in the main capacity market have an outage during an emergency period they lose some, but not all (1/24), of their emergency payment. Payments during emergency situations can be up to £6,000/MWh (1,000 times the average cost of electricity). Nuclear power plants are paid to be baseload in the capacity market, always contributing in what is almost a “don’t switch off” guarantee and the low cost of this power helps to keep less economic plants in operation.
This year, the UK also put out tenders for a new “enhanced frequency response service” – 200 MW of back-up power for delivery in 2018. These are largely energy storage systems such as batteries. Eight projects were awarded which guarantee to produce response power at £7/MWh, guaranteeing to respond within 1 second and produce power for at least 10 seconds. Surprisingly the call for tenders was significantly over-subscribed – suggesting that several companies in the UK have already invested in battery systems or plan to do so within the next year or so.
Simon’s talk was extremely interesting and, judging by the follow-up discussion between UK regulators and power companies, something that everyone is prepare to admit is extremely complex. Keeping the lights on in the UK is a very tight balance of supply and demand. As much as possible is negotiated and agreed days, if not weeks, in advance. However, as wind capacity grows, so does the challenge. Predicting wind output is currently still only around 50% accurate which complicates forward planning for electricity supply immensely. The controversial STOR market is currently a necessary bridge to maintain supply until this unpredictability and intermittency is resolved. The emerging new market for battery and other energy storage systems is vital – if we can store energy from the sun, wind and waves and use it when we need it, rather than having to ramp up large and small scale fossil plants, then the move towards decarbonisation will move more quickly and more smoothly.