The power market design column – World-wide experiences

Last August 3600 delegates from 98 countries met in Paris at the 49th CIGRE Session. CIGRE is an association, or in nowadays terminology, a global community for power system professionals. Not only does it cover many countries around the globe, it also gathers experts with different backgrounds from utilities, manufacturers, universities and authorities. Traditionally CIGRE focusses on the technical aspects of large power systems, but since 2002 it is also explicitly addressing markets and regulations in its Study Committee C5. Due to its world-wide presence, CIGRE establishes an interesting platform to discuss the different market designs in order to learn from successes and failures. For example, Study Committee C5 is now working on the application of price caps and mechanisms for price formation in different markets. Such work is important to gain real understanding of the different market designs. Especially as specific design elements that work well in one design cannot automatically implemented in another market design.

Many different market designs have been developed and implemented around the globe. Market rules are continuously evolving and in some cases more fundamental market redesigns took place. A lot of experience has been collected in the South American markets, were Chili was in the fore-front with a cost-based pool design. The England & Wales pool was based on that model but changed cost-based bidding into a price-based bidding. The EU followed suit and played a big role especially with the rapid introduction of 100% market opening and full retail competition. Colombia and several US states introduced capacity markets that are becoming increasingly popular in the EU. The Australian market has recently been under stress because of larger black-outs. And emerging markets like the Indian power market show their unique peculiarities, like the imbalance settlement based on the system frequency deviation.

The US and EU market designs

It is difficult to classify all these different market designs in a simple structure. However one could try to compare two basic models, a typical US and a typical EU market design. The typical US market would be a mandatory pool with nodal pricing. The typical EU market would be a bilateral trading energy-only market model within a larger bidding zone based on bilateral trading. (For the sake of simplicity, the presence of a capacity market is ignored. In the US, capacity markets are established in NYISO, ISONE and PJM, but not in MISO, SPP and CAISO. In the EU France, UK, Italy, Ireland, Poland and Belgium have a market-wide capacity market or are preparing such market.)

Both models are very different but some features make them look alike. In the pool model, market participants can place offers to make sure that they will be dispatched, so the pool becomes less mandatory as it looks. In the EU market the day-ahead market organised by the power exchanges cover large volumes and to some extent become de-facto mandatory as the power exchanges have been granted the monopoly to use cross-border capacity for day-ahead as well as intraday trading.

Nodal pricing

One key element of the typical US design is nodal pricing. This allows for a coordinated dispatch of generation and transmission. The costs for losses and congestions are reflected in nodal prices and directly result in the optimal dispatch. In the EU market design, each bidding zone is treated as if it is a copper plate. Market participants can freely trade within the zone. In case of congestions within the bidding the zone, the TSO has to apply corrective measures, e.g. through redispatch. The TSOs will recover the associated redispatch costs through the grid tariffs. These redispatch costs are no welfare losses, because the costs are caused by an actual grid congestion. Such redispatch however can result in a welfare loss if the ultimate dispatch is less efficient.

The debate on the best market design often focuses on these merits of nodal pricing which should result in a better economic dispatch. However, a “perfect market” requires more than the best economic dispatch. A perfect market also requires that market participants are able to buy or sell for any time period at any time. Absence of liquid forward markets introduces inefficiencies and welfare losses.

In the US market design this challenge is tackled with the introduction of virtual zones and trading hubs that allow for liquid trading of forward products. However they do not provide a perfect hedge as the price risk towards the relevant node remains in place. Products to hedge that price risk are obviously traded with much lower liquidity. The more products are needed for a perfect hedge, the lower the liquidity will be.

Demand-side response

It is difficult to allow for direct participation of demand-side response in a mandatory pool model. The US market design introduced the possibility for aggregators of demand-side response according to certain rules. These rules allow the aggregator to trade on the basis of a virtual, zonal price. This rule entails preferential treatment of demand-side response as the exposure to the nodal price risk is absent. More problematic however is that similar rules were also advocated for the EU market design. Demand side response in the EU market design is however much less problematic. With retail competition, and as each residential consumer is free to change supplier, demand-side response is primarily based on the tariff of the retail supplier. The concept of self-dispatch also applies to the residential consumer that decides on its consumption pattern in response to the retail tariff. The supplier can offer retail tariffs in any shape and form, ranging from flat rate and traditional peak/off-peak tariffs to hourly prices linked to the day-ahead market prices. The roll-out of smart meters allows for such tariff structures. Competition will force suppliers to offer those tariffs that fit the flexibility capabilities of the consumer and thus value will be created. This mechanism is also known as implicit demand-side response.

When developing the new legislative proposals of the Clean Energy Package, the EU Commission seemed to have forgotten about this form of demand-side response and incorrectly assumed that consumers are passive. As a result it introduced the concept of demand-side aggregation, most likely influenced by success stories from the US, but probably unaware of the important differences between the US and EU market designs. Regulation of a specific business model in the EU market is however, at its best, unnecessary.

Price formation

Another difference between the two market models is the way in which prices are formed. The US model with a mandatory pool is characterised by the well-known merit order of generator offer prices. Total demand is matched with the available generation, and a price is formed. The same concept is also believed to apply in the EU model. However, in practice price formation in the EU model takes a very different route. It is a bilateral market where buyers and sellers agree on a price. While moving closer to delivery, market participants (including generators) optimise their portfolio in which also the day-ahead markets - as organised by the Power Exchanges - play an important role. However, these markets are not mandatory and generators may well act as buyers. So, in practice there is no actual merit order for the total market. Still, the easy to explain merit-order curves dominate the thinking. This becomes problematic, for example if EU regulators believe that generation from wind and solar has to be offered at zero Euro/MWh in a day-ahead market. It ignores the fact that a day-ahead market is a forward market and a sale in the day-ahead market is a commitment that causes the generator to be exposed to an uncertain imbalance price.

Conclusion

Comparing different power market designs around the globe is a useful exercise. The aim should not so much be to find the theoretical optimal design as in practice imperfections, for example introduced by external policies, are unavoidable. The aim should be to gain understanding but also to avoid that specific features that are successful in one design are blindly introduced in another design.

This is my 16th column on power market design issues. The earlier columns covered the following topics: cross-border capacity calculation, flexibility, cross-border capacity, electric time and unintended exchanges, EU Network Codes, price formation and zero marginal cost generation, simplicity in the Clean Energy Package, smart grids, storage, auto-generation, balancing, VoLL, demand side response, interconnectors and the Economist on market design.

Disclaimer: The views as expressed in this column do not necessarily reflect the views of Statkraft

Paul Giesbertz

paul.giesbertz@statkraft