ELEXON Insight: Sub-zero electricity prices in December
Too much electricity generation in Great Britain pushed prices sub-zero for 13 hours. Occurring on the night of 7 December and morning of 8 December, BSC Parties were paid to consume energy. In this ELEXON Insight our Market Advisor, Emma Tribe, uses ELEXON’s open data from BMRS to explain why the prices were negative.
Published: December 2019
Negative Imbalance Prices on 7 and 8 December
From 23:00 on 7 December until 10:00 on 8 December the Energy Market was being paid to use electricity or reduce electricity generation. There were 26 half hour Settlement Periods with negative Imbalance Prices on these two days. The most negative Imbalance Price over the weekend was -£88/MWh which occurred between 4:00 and 5:00 in Settlement Periods 9 and 10.
The 11 straight hours where the Imbalance Price was negative was the longest continuous period of negative Imbalance Prices. This stretch of negative Imbalance Prices beat the 6.5 hour stretch of negative Imbalance Prices in March this year.
Net Imbalance Volume and Imbalance Prices on 7 and 8 December
The graph above shows the Net Imbalance Volumes and Imbalance Prices between 15:00 on 7 December and 20:00 on 8 December. The Net Imbalance Volume was negative between 22:00 on 7 December and 12:30 on 8 December.
A negative Net Imbalance Volume occurs when the system is long and National Grid were taking more balancing actions to reduce an excess of electricity on the system. The greatest Net Imbalance Volume was -2,283MWh between 4:30 and 5:00 on 8 December (Settlement Period 10). This was also the most negative Net Imbalance Volume in a Settlement Period in a year.
ELEXON uses the Prices that National Grid pays for balancing services to calculate the Imbalance Price. The Imbalance Price, also referred to as System Price or cash-out, is used to settle Imbalance Volumes for BSC Parties.
National Grid pay balancing service providers for Offer volume to increase the level of energy on the system and are paid by balancing services providers for Bid volume to reduce the level of energy on the System. As National Grid are paid for Bid volume, a negative bid price is where National Grid are paying out for that balancing volume rather than being paid.
Energy balancing volumes by price on 7 and 8 December
The graph above shows the Bid volume (coloured volume) and Offer volume (grey volume) that were purchased to balance the system between 15:00 on 7 December and 20:00 on 8 December. The Bid volume has been coloured to show the volume that was positively priced, negatively priced and priced at zero.
Between 21:00 on 7 December and 14:30 on 8 December more than 90% of Bid volume that was bought by National Grid was either negatively priced or priced at zero. The Imbalance Price is calculated to reflect the marginal price of balancing energy in the direction of the Net Imbalance Volume.
Most of the Bid volume (76.8%) provided on 7 December and 8 December was taken to resolve a system constraint. Of the Bid volume that was negatively priced 90% was taken to resolve a system constraint.
Curtailment of Wind Balancing Mechanism Units represented 82% of negatively priced Bid volume on 7 December and 8 December. The graph below compares the volume of curtailed wind to the volume of wind generation for the same period as the two graphs above.
Wind curtailment on 7 and 8 December
Peak wind curtailment in a Settlement Period was 2,342MWh between 4:00 and 4:30 (Settlement Period 9). This curtailment was 33.6% of the potential generation which is the sum of curtailed volume and wind generation.
Trends in negatively priced bid volume since 2014
Balancing services providers choose how to price the balancing volume they submit into the Balancing Mechanism for National Grid to accept. However, balancing services providers can be constrained on how to price their balancing volume if their assets are in part of the electricity system where a network constraint exists.
Percentage of bid volume that is negatively priced
Since 2014 the volume of accepted bids that were priced negatively has increased from 9.9% in 2014 to 25.2% in 2019. This is shown in the graph above. Over the same period the total accepted Bid volume in a year has remained between 10.6GWh and 8.5GWh.
In 2016 the average price of negatively priced bid volume was -£66.14/MWh.
The majority (57%) of negatively priced bid volume has been delivered by Balancing Mechanism Units in Southern Scotland. Followed by bid volume from units in Northern Scotland (30.1%).
The relative volumes of negatively priced bids is shown in the colour map and bar chart below. The date slider on the right hand side can be used to adjust the date range shown in the graphs.
Negatively Priced Bid volume
In southern Scotland, 95% of the negatively priced bid volume was from wind Balancing Mechanism Units. In northern Scotland 87.3% of negatively priced bid volume was also from wind Balancing Mechanism Units.
Explore Imbalance Prices Data further
ELEXON publishes the latest Imbalance Prices and the volumes and prices of actions used to balance the GB electricity system on the Balancing Mechanism Reporting Service
The 2010s were a period of significant change for both ELEXON and the wider industry. The shifts seen during the span of ten years were both unprecedented and unpredicted. ELEXON provides a wealth of data to industry via the Balancing Mechanism Reporting Service (BMRS), ELEXON Portal and through data flows. Using some of this data we have created four graphs that reflect on significant changes to generation, demand and balancing between 2010 and 2019.
Changes to how electricity is generated
The first graph shows the rise of low CO2 fuels as a proportion of Great Britain’s fuel mix. In the span of ten years, Great Britain’s fuel mix has gone from 76.3% of generation using fossil fuels to 41.3% from fossil fuels. This has coincided with increases in the proportion of the fuel mix generated from low CO2 fuels, which was 21.6% in 2010 and 49.8% in 2019.
Electricity fuel mix proportions: 2010 – 2019
Generation from coal and gas fuelled power stations form the fossil fuel generation type. The low CO2 fuels include generation from: wind, solar, nuclear, hydro, and biomass. Biomass has been classified as a low CO2 fuel despite its emissions as Drax, the largest generator using biomass in Great Britain, has a carbon neutral process. The CO2 absorbed by the trees planted to produce the wood for the biomass pellets counteract the CO2 produced in combustion.
The ‘other’ fuel type shown as the pink line on the graph, includes:
- Generation in Great Britain where we don’t record the fuel type.
- Generation from outside of Great Britain transferred over interconnectors to meet Great Britain’s electricity demand.
Electricity imports over interconnectors should be considered as part of Great Britain’s fuel mix as it forms the majority of the ‘other’ fuel type. The percentage of Great Britain’s electricity demand met by imports has increased from 2.1% in 2010 to 8.6% in 2019.
Graph number two compares the volume of electricity produced in 2010 to 2019.
Electricity generation: 2010 to 2019
This graph is interactive and you can use the year and fuel type filters to the right of the graph to look at the changes in generation from a fuel type more closely, or compare two years.
The biggest changes in volume of electricity produced relate to coal, which was generating 137 TWh in 2012 at its peak. Coal-fired generation was the second largest producer of electricity in 2010 and 2011, and the largest overall from 2012 to 2014. Since 2015 the decline in coal generation has been rapid and in 2019 it was the sixth greatest generator contributing just 6 TWh.
We are also able to view the significant increase in generation from wind, interconnectors, biomass and solar. These four sources produced 112 TWh of electricity in 2019, compared to 17 TWh in 2010. These fuels have replaced the majority of coal fired generators.
Gas generation still plays a big role as shown in the graph and it is unlikely that our electricity system will ever be run completely without gas power stations. Greater use of carbon capture and storage and alternatives to gas are required in order to achieve Net Zero.
Looking at electricity generation overall, and excluding imports, the total electricity generation per year in Great Britain has decreased from 334TWh in 2010 to 266TWh 2019.
Changes to demand
The third graph shows that annual demand for electricity decreased by 54TWh (16.2%) over the last ten years.
You can use the day and night filters to see how the annual day time and night time demand has decreased. Annual day time demand has decreased by 38TWh (15.6%) and night time demand has decreased by 16TWh (17.3%).
Demand is calculated as the sum of metered volume from Balancing Mechanism Units where there is net demand. A Balancing Mechanism Unit represents a group of customers’ metering systems used by ELEXON in Settlement.
Small scale generation embedded in a Balancing Mechanism Unit that has overall net demand has some of its demand netted out by solar generation. Embedded generation from wind and solar has increased from 6.4TWh in 2010 to 23.5TWh and is responsible for 31.7% of the total decrease in electricity demand.
As solar can only generate during the day we can also conclude that the 11.6TWh increase in embedded solar generation is responsible for 30.5% of the 38TWh decrease in day time demand.
It is difficult to attribute the rest of the decrease in day and night time demand to a particular source. Part of the decrease in day time demand will be due to a decrease in large scale manufacturing taking place in Great Britain with many companies moving factories overseas to avoid increasing costs. Both day and night time demand will have decreased due to an increase in uptake of energy efficient technology and appliances.
Nonetheless, if the decrease in demand for electricity continues in the next ten years, achieving Net Zero may become easier.
Rising costs for managing the system
The final graph shows the increase in the costs to manage the system through the Balancing Mechanism. ELEXON calculates the cashflows for the Balancing Mechanism as part of electricity Settlement. The annual cost to National Grid ESO of managing the system has increased threefold from £215 million per year in 2010 to £672 million in 2019.
National Grid ESO use the Balancing Mechanism to pay for flexible generation and demand side response providers to increase or decrease their generation or demand. This helps the ESO to manage supply and demand on the electricity system and relieve constraints (for example, when there isn’t enough network capacity to transport electricity that is being produced).
Balancing Mechanism Cashflow in the 2010s
This graph can be filtered to show the annual Balancing Mechanism cashflow during the day and overnight. The cost of managing the system during the day has increased from £171 million in 2010 to £386 million in 2019, while the overnight cost has increased from £43 million in 2010 to £286 million in 2019.
In 2010 the ratio of overnight costs to day was 20:80, in 2019 that ratio was 43:57. The increase in the cost of balancing the system overnight has meant that it has become nearly as expensive to balance the system overnight as during the day.
The costs for balancing the electricity system have increased, partly due to the natural unpredictability of when renewable generators will be available. When there is plenty of wind or sun there is an abundance of generation on the system, which is cheaper to produce than electricity from fossil fuels.
If this occurs when demand is low, the ESO may need to pay renewable generators to reduce their output. This can happen in the day time and at night, where a large surplus of wind generation output may be available when demand has tailed off. During December 2019, generation at night was so high at some points that there was a negative Imbalance Price for a record 13 hours, you can read more in our sub-zero electricity prices in December article.
At times of peak demand, renewable sources may not be available to generate and therefore fossil fuel generation has to be increased. As an industry we need to encourage development of electricity storage which can absorb excess renewable generation and export it back to the networks when it is needed.
The 2010s have been filled with rapid change for the electricity system and data such as this provides important insights which both ELEXON and the industry can apply, as we work together to help deliver on the commitment to ‘Net Zero’ carbon emissions
It is difficult to predict exactly how the electricity system will develop. So together with the industry we work to anticipate changes to rules and process that will be needed to support a smarter, greener system and deliver reforms.
This includes our work on the Target Operating Model for moving to Market-wide Half Hourly Settlement which would allow consumers to take up a wider range of ‘time of use’ tariffs. We are also proposing that nationwide electricity ‘flexibility platforms’ are set up so offers of demand-side response, output from electricity storage facilities, and spare network capacity can be traded easily.
Nearly half of all the electricity generated in Great Britain in 2019 was generated by low CO2 generation. In this ELEXON Insight, our Market Advisor Emma Tribe uses electricity generation data to show how 2019 was powered demonstrate how the electricity market is on the road to Net Zero.
The pie chart below shows that 49.83% of electricity generation in 2019 was from low CO2 fuels. The second most dominant fuel was fossil fuels which represented 41.32%.
2019 electricity generation proportions
The generation volumes from the different fuel types have been grouped into emissions type as follows:
Low CO2 fuel:
- Wind (embedded and transmission connected)
- Solar (embedded)
- Imports over interconnectors
- Other fuel types
The fuel sources used in the county of origin for imported energy is unknown so imports from interconnectors have been grouped with other fuel types.
Total volume of electricity generated
The graph below gives the total volume of energy generation in terawatt hours (TWh) over the year for each of the emissions types. There was 145.11TWh generated by low CO2 sources in 2019, the majority of this generation came from wind (58.33TWh) and nuclear (52.71TWh).
Of the 120.33TWh generated by fossil fuels, 114.4TWh was generated by gas and the rest by coal.
2019 electricity generation by emissions type
Use the drill down and hierarchy arrows in the top left and right corners of the graph to view the volumes of generation from the different fuel types that make up each emissions group.
Gas generated more electricity than the combined generation of the next two highest generating fuel types, wind and nuclear. While more electricity was generated overall from low CO2 fuels, Great Britain is still heavily reliant on gas fueled generation as a major part of the fuel mix.
Trends in generation
The monthly volume of electricity generated by each type of fuel is shown below. January saw the highest output from fossil fuels (15.09TWh) and was the only month where more energy was generated from fossil fuels than low CO2 sources.
December had the highest volume of energy generated by low CO2 sources in a month (14.69TWh). It also had the day with the highest generation from low CO2 fuels 0.61TWh on 8 December. In December gas generated 8.74MWh (32.61% of total generation) wind generated 6.91TWh (25.79%) and nuclear 4.88TWh (18.22%). Which was the highest volumes of wind and nuclear generated in a month in 2019.
January and December had the highest total volumes of generation, 28.78TWh and 26.80TWh respectively, as there is higher demand for electricity in these two colder months.
Monthly electricity generation trends
Low CO2 fuels generated more than 50% of electricity in six months: March, April, August, September, October and December.
Generation output at different times of day
The time of day when electricity is generated is also important. The half hour with the greatest total generation over the entire year was 7.10TWh between 17:30 and 18:00, this is the time with the greatest demand for electricity. The graph below shows the volume of energy generated by each type of fuel in each Settlement Period over the year.
Between 17:00 and 21:30 there was more electricity generated by fossil fuels than by low CO2 fuels. This is because there is less sunlight to fuel solar panels, but still high demand for energy. For the rest of the time more electricity was generated by low CO2 fuels.
Generation over a day
The time of year changes the profile of demand for electricity over a day. You can use the filters on the right hand side to filter for month.
Fossil fuel and low carbon generation use during the day
The graph below compares the percentage of the time of the day that fossil fuels and low CO2 fuels are the dominant form of generation.
For most of 2019, there was more low CO2 generation on the system than fossil fuel generation. Particularly overnight between 23:30 and 6:00 when low CO2 generation exceeded fossil fuel generation for more than 70% of the year. Overnight there is also less demand for energy so less generation is required.
Percentage of time as dominant fuel over a day
The times when the split between fossil fuels or low CO2 fuels being the greatest generator was closer to 50% is during the morning and evening peaks. This is also the time of day when flexible generation is required to ramp up and down to meet the demand peaks.
How 2019 compares to other years
The graph below shows the proportion of the fuel mix by emission type for the past 10 years.
2010’s electricity generation proportions
Looking at the fuel mix percentage over the last decade, there is a clear trend of decreasing generation from fossil fuels, and an increase in generation from low CO2 fuels. In both 2018 and 2019, there was a higher proportion of generation from low CO2 fuels than from fossil fuels.
The contribution of fossil fuels to the GB fuel mix has reduced from 76.27% in 2010 to 41.32% in 2019. This decrease is due to a reduction in generation from 259.92TWh in 2010 to 120.33TWh, which is a 53.7% decrease. Over the same period generation from low CO2 has increased from 73.83TWh in 2010 to 145.11TWh in 2019.
Until November 2017 Biomass was considered an ‘other’ type fuel. The growth of generation from this low CO2 fuel has meant that ELEXON changed how we reported the electricity generation fuel mix to industry. This combined with increases in wind generation resulted in low CO2 generation overtaking output from fossil fuels, in 2018 as shown in the graph.
Road to net zero
The commitments made by the government and the trends in fuel mix show that the electricity sector is on the road to net zero.
The times of the day and year when low CO2 fuel is available and generating is as important to achieving net zero as the total volume generated. Low CO2 flexible generation, or flexible demand will be needed for Great Britain to achieve a net zero electricity system. ELEXON is supporting industry in achieving this goal.
This includes work to:
- Allow customers to be served by multiple suppliers through the same meter
- Proposing flexibility exchanges to trade demand side response and spare network capacity
- Design metering arrangements which fully account for services provided by smaller generators, batteries behind the meter
Explore the data further
- Half hourly generation by fuel type data can be found on ELEXON’s Balancing Mechanism Reporting Service
- Historical generation data going back to November 2008 is published on the ELEXON Portal.
The data published by ELEXON doesn’t include generation from embedded solar and wind generation, this data can be found on National Grid ESO’s Data Explorer.
The System Price reached £2,242/MWh and £1,708/MWh in the early evening of Wednesday 4 March 2020. This was the first instance of the System Price being over £2,000/MWh since 2001. In this ELEXON Insight our Market Advisor, Emma Tribe explains how the System Price was calculated and how it compares to historic System Prices.
System Prices on 4 March 2020
ELEXON calculates System Prices every 30 minutes. They are used by ELEXON to settle Energy Imbalance Volumes for electricity market participants. System Prices reflect the marginal price of energy used by National Grid ESO to balance the net energy Imbalance of the Electricity System.
System Prices give a financial signal to electricity market participants to reduce their Energy Imbalance Volume, or to have Energy Imbalance Volumes that are helpful to the system.
A high System Price will lead to high energy imbalance cashflow for market participants. This cashflow could either be a large charge if the participant had a deficit of energy, or a large pay out if the participant had an excess of energy.
Looking at the graph below, which shows System Prices for 4 March, there is a clear spike at Settlement Period 37 and 38 (18:00 to 19:00). The System Price reached £2,242/MWh in Settlement Period 37 and then £1,708/MWh in Settlement Period 38. For comparison the average System Price in February 2020 was £33/MWh and the maximum System Price was £120/MWh.
Looking at the System Prices earlier in the day, System Prices had remained between £0/MWh and £72/MWh in Settlement Periods 1 to 32. The System Price then increased between £119/MWh and £144/MWh in Settlement Period 33 to 36 before going up to £2,242/MWh.
The difference in Settlement Periods 37 and 38 compared to the rest of the day was the Reserve Scarcity Pricing Mechanism. The Reserve Scarcity Price was able to set the marginal price of energy used in balancing the net energy Imbalance of the Electricity System. Hence, the Reserve Scarcity Price helped set the System Price.
There was also a Buy Price Adjuster of £19.14/MWh that affected the System Price in these two Settlement Periods. The Buy Price Price Adjuster is added to the System Price as a final step in the calculation.
Reserve Scarcity Pricing
Energy scarcity will generally cause the cost of balancing energy to increase as balancing services providers adjust their prices to match the scarcity. However, there are certain balancing services providers that cannot adjust their prices because they have a fixed price agreement with National Grid ESO.
Short Term Operating Reserve contracts fix the price of balancing energy in advance during pre-agreed Short Term Operating Reserve Availability Windows. The Short Term Operating Reserve providers are then paid an availability fee to be available with their contracted prices during these windows.
These fixed Short Term Operating Reserve prices for balancing energy could then have a knock on impact on System Prices. This is because System Prices may not reflect the energy scarcity, and therefore not give the right market signal. The Reserve Scarcity Price was introduced in 2015 to stop this happening.
Getting the Reserve Scarcity Pricing
To get the Reserve Scarcity Price, National Grid ESO calculate forecasts of the De-Rated Margin and Loss of Load Probability for a Settlement Period. ELEXON then publish these forecasts on bmreports.com as an information tool for industry. Using the Loss of Load Probability calculated one hour ahead of the Settlement Period, ELEXON calculate a Reserve Scarcity Price to be fed into the System Price calculation.
The Reserve Scarcity Price is equal to the Loss of Load Probability multiplied by the Value of Lost Load, currently £6,000/MWh. On the 4 March 2020, during Settlement Period 37, the Loss of Load Probability was 0.3705 (37.05%).This meant that the Reserve Scarcity Price was calculated to be £2,223/MWh.
The Reserve Scarcity Price is then used as part of the System Price calculation. It represents what the price of Short Term Operating Reserve actions would have been if they could have changed their pricing to match the market scarcity.
Reserve Scarcity Price forecast
Below, I have calculated what the Reserve Scarcity Price would have been during the second Short Term Operating Reserve Availability Window of 4 March (dotted lines). This was calculated using the eight hour ahead, four hour ahead and two hour ahead Loss of Load Probability forecasts. The calculated Reserve Scarcity Prices are then compared to the Reserve Scarcity Price that was used in the System Price Calculation, which uses the one hour ahead forecast.
All of the forecasts are less than the final Reserve Scarcity Prices during Settlement Periods 37 and 38. Although the two hour ahead forecast is very close to the final Reserve Scarcity Prices.
The four and eight hour ahead Reserve Scarcity Prices are around a quarter of the magnitude of the final Reserve Scarcity Price. However, they are large enough to indicate to the market well in advance that there might be energy scarcity.
Reserve Scarcity Price and forecasts: Settlement Periods 34 and 41, 4 March
The Loss of Load Probability and De-Rated Margin forecasts are a signal to the market to make more energy available, and for market participants to have efficient Energy Imbalance Volumes. In this instance, the forecast showed greater energy scarcity the closer in time to the Settlement Period. It is difficult to investigate the impact the forecasts had on participants Imbalance Volumes given there are lots of factors involved in influencing Imbalance Volumes.
Comparison to historic System Prices
There have only been nine other System Prices higher than £1,000/MWh between 2002 and 4 March 2020. All of these high Prices were in November 2016 and May 2017. The November 2016 high prices were caused by French Nuclear outages affecting energy prices in Europe. The May 2017 high prices were due to low wind and solar generation.
I have compared 4 March System Prices to very high System Prices after 2001 in the graph below.
System Prices during 2001 have not been considered in this comparison as this was the first year of the New Electricity Trading Arrangements that introduced the System Price in England and Wales. There were some incorrect market signals created during 2001 with the System Price that were corrected for by modifications to the System Price calculation.
System prices over £1,000/MWh since 2002
None of the prices over £1,000/MWh in November 2016 or May 2017 were set by the Reserve Scarcity Pricing mechanism. Instead, they were a result of National Grid ESO needing to take high priced balancing actions.
Comparison to historic periods of scarcity
Since its introduction, this mechanism has only occasionally affected prices, mostly because there hasn’t been enough periods of energy scarcity in the market to cause market signals. The graph below shows all of the instances where the Reserve Scarcity Price has set the System Price.
There were only three Settlement Periods where this had happened before 4 March, these were in October 2016 and May 2017. Both of these periods coincided with the market conditions that led to the prices over £1,000/MWh described above.
The previous highest System Price set by Reserve Scarcity Price was £843/MWh on 9 October 2016.
Reserve Scarcity Price setting the System Price
In November 2018, the Reserve Scarcity Price calculation changed. The Value of Lost Load was set to £6,000/MWh where it had previously been £3,000/MWh. The method for calculating the Loss of Load Probability was also changed from a static function to a dynamic function.
Hence, 4 March was also the first time a Reserve Scarcity Price calculated with the new calculation set the System Price.
Background information to this article
Find out more background information and context regarding some of the terms mentioned in this article.
Reserve Scarcity Pricing Mechanism
The Reserve Scarcity Pricing Mechanism was introduced as part of Electricity Balancing Significant Code Review Modification P305 implemented in 2015. This Mod also details post implementation reviews conducted after it went live.
System Price Analysis Report
ELEXON monitor the System Price calculation and produce a monthly System Price Analysis Report full of information on the elements of the System Price calculation over a given month.
Balancing Mechanism Reporting Service
ELEXON publish data on System Prices and the Balancing Mechanism, as well as forecasts and REMIT information on the Balancing Mechanism Reporting Service.
Historic System Prices
Historic System Prices dating back to 2001 are published on the ELEXON Portal in Best View Prices.