IN this series of four features, Nigel Cornwall is taking us on a journey over the 30-year history of GB Electricity Market.
Nigel Cornwall is a well-known energy industry commentator and founder of Cornwall Insight.
His latest venture is New Anglia Energy, which is a company formed to support local energy markets in Norfolk and Suffolk.
Its purpose is to demonstrate learning by doing, using the knowledge and relationships developed by Nigel in a long and fruitful career in the energy sector to support local stakeholders and initiatives.
This second piece will look at the basic template for the new electricity trading arrangements (or NETA model), which replaced the Pool in 2001, and the establishment of Elexon to act as a manager of the centralised processes.
“The new NETA model was facilitated by passage of the Utilities Act 2000 and implemented on 27 March 2001 in England and Wales after a three-year design process. It was as far removed from that of the centralised Pool as it is possible to get in electricity market terms.
Other electricity market reform programmes around the world looked to refine and enhance pool-based systems, but in England and Wales the regulator headed in a very different direction.
In this piece we look at the market’s first decade of operation, including its extension into Scotland in 2005, focusing on the key features of the new market architecture.
A seismic shift
The design of NETA – the New Electricity Trading Arrangements – was predicated on an assumption that electricity can and should be traded just like any other commodity, with producers and suppliers constantly firming up their contract positions as they move towards delivery. The system operator was a facilitator of transactions but only after market close, and the concept of central dispatch went out the window.
The template was the GB gas market, which had itself been reformed in the late 1990s. Although there was a protracted debate during the market design phase of NETA about whether there should be a centrally established power exchange, the designers decided not to seed one, leaving it to the market to bring this forward.
The approach was “tell us what you want to do and we will deliver it, and in short-run operational timescales we will sort out the bits that are leftover.”
This design was based on a self-commitment model. This meant that, instead of use of a mandatory pool with complex bid structures and generation being dispatched by the system operator based on a merit order, generators self-dispatched against bilateral contracts they had already entered into with simple price/quantity provisions.
To ensure full visibility of the quantum of expected trades to the market, all generators and suppliers above de minimis levels were required to notify their contracts to the newly established market operator Elexon in England and Wales (in reality an extension of the old England and Wales Pool).
The organised trading arrangements themselves was a net pool or balancing market, which in theory at least was simply used as a last resort to shore up physical positions on the day and to deal with any energy imbalance after the traded market closed, which was as close to delivery. The new rules were set out in the Balancing and Settlement Code (BSC), which replaced the Pooling and Settlement Agreement.
Market trading ceased at a predefined time (initially at 3 ½ hours ahead of the delivery, but later with a BSC rule change P11 from 2003 at one hour ahead of delivery) for each half hourly trading period, and any variations against notified contracts had to be dealt with through a Balancing Mechanism (or if they arose from system support reasons through the call off of pre-agreed balancing services contracts).
This deadline is termed Gate Closure. Dependent upon the season and time of day, anything up to 5% of physical volumes – but usually less – were transacted in this way, and the system was typically long.
That said, these arrangements were structured in a way that parties who were out of balance against their contracts or were uncontracted were usually disadvantaged. This was not because of any penal charging – although there is a mechanism to levy surcharges on out of balance parties, the Information Imbalance Charge, but it was set at zero and remains at this level today.
This incentive to contract was simply a reflection of the fact that less frequently dispatched generation that is called on to clear the market during periods of high demand through the Balancing Mechanism is invariably more expensive than baseload because of their higher cost of running. Conversely prices tended to be much lower when demand is low reflecting competition by generators to remain on the system but also suppliers who were over-contracted selling back volumes.
Indeed, the mantra of cost-reflectivity was very important to the rule-makers, and because of this two imbalance prices were calculated for each trading period, one if a party was “short” against its notified contracts and the other if a party was “long”.
In both cases the cost to the trading party was priced against whether the wider system was net short or long. This arrangement meant that, if the system was long and a party was long (or conversely short and short), the trading party would be paid (or pay) the “main” price. In other words, if a trading party was aggravating system length, it could expect to pay more or receive less than market prices.
Similarly, if a party was in imbalance in the opposite direction to the system, it could expect to receive a more benign “reverse” price as it was considered to be helping the system be closer to balance.
Chalk and cheese
In terms of market design, the regulator now reformed as Ofgem (housing both the electricity regulator OFFER and the gas regulator OFGAS) tried to address the acknowledged shortcomings of the Pool. The key features of the new arrangements, some already touched on above, are summarised below:
- “self-commitment” by individual market participants rather than central dispatch
- an “anything goes” market where trading parties are free to enter into any contracts they choose over any time-frames up until market close (or Gate Closure as it is termed in the BSC)
- the energy contracts between trading parties were physical contracts specifying volumes and prices, which had to be settled bilaterally, not financial contracts as under the Pool that managed risk of price variation against a single cleared price
- there was no central power exchange but trading platform operators and price reporters were invited to come forward and became central to informing the marketplace of prices and volumes transacted across the forward curve
- notification of these bilateral contracts at Gate Closure to the central trading arrangements as these are the basis against which physical imbalances are identified, costed and paid
- to clear the market a two-sided Balancing Mechanism for incremental and decremental power with the system operator in which sale and purchase notifications notified at Gate Closure are firm (unlike the one-sided Pool where a centrally calculated demand curve, and not suppliers’ accepted bids, was used to set each half-hourly price)
- trading parties whose metered volumes did not match their notified contracts at Gate Closure (plus or minus any accepted transactions in the Balancing Mechanism) or who were uncontracted were exposed to uncapped energy imbalance prices
- for most of the period post 2001 there have been two imbalance or “cash-out” prices
- governance was redesigned in the form of new code, the BSC, overseen by the regulator rather than a multi-party contract with high voting thresholds that deterred change, with public good objectives against which all changes had to be tested by the regulator
- a committee of key industry stakeholders was established to advise on these rule changes against hard-wired processes and deadlines and to make non-binding recommendations to the regulator on rule changes. Any trading party had the ability to propose changes (no thresholds), and panel members were obligated to act independently, and
- Elexon was formed to replace the England and Wales Pool to manage the balancing rules and the settlement processes and to ensure appropriate governance of the BSC.
The two energy imbalance prices were a notable feature of the new arrangements. They have been calculated by Elexon for most of the period until recently (there is now a single price, that is paid by or to out of balance parties) and are known as the System Sell Price (SSP) and System Buy Price (SBP). SSP was paid to market participants who had a net surplus of energy above their notified contracts in any half hour trading period, while market participants who were short against their notified contracts paid SBP.
The price reflecting the net length of the system was termed the Main Price. The second price was called the Reverse Price. There was often a wide spread between the two (SBP which was often high and SSP which was often low), and this tended to incentivise trading parties to over-contract and drive the system long (because it was better to receive less for being long than pay big prices if you were short).
After NETA market start, there were various patches developed through rule changes intended to reduce SBP by taking out small actions and high cost acceptances that were considered to be unrepresentative. But Ofgem has also been keen to sharpen the cost signal at the margin.
This has given rise to various Ofgem-instigated Significant Code Reviews since it was granted these powers in 2010. Notably these led to a reduction in the volume of offers taken into account in the Balancing Mechanism in setting the main imbalance price (the most expensive generator offers when the system was short, and the least expensive supplier bids when the system was long).
But it was only relatively recently that that the reverse price calculation was changed so that it reflected a traded price of electricity and a period after that that the single price was introduced. We will pick up both of these later changes next time.
The transmission system and its use remained subject to a different operating regime and a separate rule book, in the form of the Connection and Use of System Code (CUSC) overseen by its own governing board since 2001, which also makes recommendations on rule changes to Ofgem. Ancillary services were rebadged as balancing services, but they remained procured through tenders and bilateral contracts by National Grid under its own licence.
National Grid in its role as system operator was no longer responsible for dispatching generation; it was concerned only with the electrical balance of the system after Gate Closure. This role was termed by some “the residual balancer”. To carry this out it had three procurement markets to buy the operational and security services it needs to balance the system.
In addition to the Balancing Mechanism that it uses to source offers to increment power and bids to decrement power, these are, first, forward contracts for balancing services purchased predominantly from generators and, second, forward energy trades undertaken anonymously prior to Gate Closure to balance the system, which is unusual in market design terms. But again, as we will see in Part 3, there has been steady evolution in the way these balancing services are defined and procured in a rapidly changing system with a growing range of providers.
How did NETA perform in its first decade?
The main claim for NETA was that it would reduce prices at both wholesale and retail level. Its record here has been decidedly mixed.
Although prices did reduce over the period from 1998 ahead of NETA to 2002, studies have found that the main factor that had a significant impact on reducing prices was the further reduction in concentration of ownership of gas and coal generation.
This was as a result primarily of the large divestments by National Power and PowerGen agreed in 1998 that we noted last time. Also, in the light of those sales, there continued to be significant levels of new entry of new generation over this period as merchant generators developed new generation hoping for a continuation of inflated Pool prices, which of course did not materialise.
Indeed, the National Audit Office found in May 2003 that, although prices to industrial and commercial customers had fallen significantly, for domestic customers “there have been modest reductions (e.g. 3%) in the bills of consumers who switched away from the incumbent supplier before NETA was implemented … For consumers who remain with the incumbent supplier, our analysis shows that there has been little reduction in their bills since April 2001”.
Prices rose steady through to 2008, but they then fell dramatically with the financial crisis of that year. But, while the linkage of wholesale prices to retail prices has muddied since with the steady addition of policy costs to the retail bill and the steady rise in network costs, wholesale electricity prices do seem to be broadly aligned with commodity prices. So, the bad old days of systemic generator gaming and price setting that characterised the Pool do seem to have long gone.
Again at a high level, the NETA designers also set significant store on the impact of properly integrating consumers and the demand side in the market. However, there was no evidence of any increase in demand-side participation with NETA.
In fact the expected ability for suppliers to bid in demand reduction failed to appear, largely as a result of the glacially slow progression of implementation of mass market smart metering (which Elexon is not responsible for) compounded by the failure of the market operator to address the over-hang of profiling introduced in 1998 to support retail competition.
Even today householders and small businesses customers cannot affordability be settled on a half hourly basis. Furthermore, no large industrials have actively engaged in the market probably due to the cost and complexity of participation (though the market has been actively intermediated by brokers who deal directly with suppliers on behalf of business customers). As we will see next time, the ability to buy flexibility and to turn demand up and down has been actioned through bilateral contracts placed by the system operator.
One of the main attributes of NETA was its ability to reform itself through open governance. The early years were largely focused on addressing teething problems and loose ends in the market rules.
The main developmental issue was the extension of the system to Scotland, which occurred in April 2005. Originally it had been hoped to align the trading system north of the border with that in England and Wales, but under a separate operational structure, in line with the establishment of NETA, and the new Ofgem had consulted on this back in October 1999.
However, this raised many complexities – not least the need to reform the transmission arrangements in Scotland which were very different to those in England and Wales. In the event this initial reform process was timed out.
But by May 2002 after a further Ofgem consultation, the Department of Trade and Industry with Ofgem sought views on specific proposals to establish a single GB energy market but also extending the England and Wales transmission arrangements into Scotland.
This entailed not only aligning the arrangements for network access and use, but also extending the jurisdiction of National Grid as system operator into Scotland. The enabling legislation was passed in 2004 and the single GB market was switched on in April 2005.
Ofgem focused early on governance, advocating the need for “more effective and flexible governance arrangements”, and it introduced significant changes that gave it more direct control over the rules of the market.
It achieved this through the vehicle of the Transmission Licence, which required National Grid to have in place the BSC and the CUSC and through back-to-back license conditions in generation and supply licenses that require holders to comply and be signatories to the market codes. Additionally, all changes to the market rules have to be approved by Ofgem.
Roll in another nine industry codes, energy industry governance is complex and cumbersome, and rule changes can cut across the different agreements governed by different bodies with differing governance conventions. So the theory and the practice have diverged.
Over the period to 2010 there were some 250 modifications to the BSC alone. Although the new governance arrangements have eliminated barriers to change under the Pool on major issues, the new system has its critics. For the first decade of NETA Ofgem was engaged in all rule changes irrespective of their significance.
While it could not propose modifications, stakeholders thought it leaned on parties (especially National Grid on cash-out reform) to bring forward change. Decision-making can take a long time, and the regulator’s legalistic attitude has caused delays to some important modifications. All issues received Rolls Royce treatment whether or not they were significant, and modification reports have tended to become formulaic.
Ofgem’s decisions have not always been consistent. And because of the complexity of the market processes, there is a high embedded cost of change, which has meant that it can be difficult to demonstrate net benefits, which can in turn impose a high threshold of proof for change.
Furthermore, the piecemeal approach to dealing with significant issues through serial proposed modifications which (among other things) meant that critical issues like the form of cash-out prices could not be considered in a structured manner.
There was – and is still – a very poor interaction with the CUSC and other codes, and National Grid has been allowed to administer changes to balancing services without any formal code or checks that come with it.
To some decree the changes that were implemented as a result of the Ofgem initiated Code Governance Review in 2010 (and two further phases of this review) have addressed some of these problems, but governance reform was limited in the first decade and even today remains a work in progress.
“But different – some would say bigger issues – have moved up the agenda, most notably decarbonisation of the electricity system and most recently net zero. European interactions have become more important as systems have become more interconnected. We will return to this issue next time.”