The marketing and trading of electricity on short-term markets such as the intraday market are rapidly growing in importance. This development is driven by legal regulations and the increasing proportion of volatile, renewable production capacity, but also by attractive earnings potential in the new markets.
Traders and energy producers entering this new playing field are confronted with new challenges that also lead directly to additional requirements for IT systems in terms of trade and optimisation.
The rapidly growing trade volumes in the short-term markets have different, mutually reinforcing causes. On the one hand, the increasing share of direct marketing of EEG generation plants leads to higher liquidity and activity in short-term markets. This is already enforced in part by changes in the legal framework, which EEG 2014 has driven forward even further. In particular, the reduction of the limit value for compulsory direct marketing to 100 kW means that only small plants can remain in the traditional fixed remuneration model.
In the long term, the plants that are currently still marketed according to this remuneration model will no longer qualify for subsidisation and will inevitably revert to direct marketing. The resulting focus on short-term markets is a matter of course as the volatile production of EEG plants only allows for short-term forecasts.
It is also to be expected that, in the sense of the required "system convenience" of renewable energies, more opportunities for renewables such as wind power to participate in balancing energy markets will arise in conjunction with ever more powerful storage facilities.
On the other hand, the current market environment, with its very low prices for long-term electricity supplies, means that many operators of conventional production systems will be unable to achieve adequate margins, or that operating the plant will even lead to negative profit margins. This then also forces these operators to develop new forms of marketing which may lie in short-term trading with the associated price volatility. Furthermore, this will enable production flexibilities available at short notice to be marketed and the cost of balancing energy resulting from volatile infeed and limited forecasting possibilities to be minimised.
In general, the ancillary services markets for minute reserves (MRP), secondary control power (SCP) and primary control power (PCP) are suitable for short-term marketing of flexibility, as are the markets for short-term energy supplies, i.e. the established spot market for day-ahead trading, and, more recently, intraday trading as a new form of marketing. In addition, the market premium and the flexibility premium for biogas plants are also available. The increasing diversity of marketing forms creates new revenue potential, but also increasing complexity.
The potential revenues and therefore the appeal of the various markets depend on the respective, significantly changing price levels. In principle, we can assume that marketing on a bid and contract-based market is all the more competitive the more participants there are submitting bids and who are therefore competing in price against their own offers.
The number of pre-qualified suppliers for balancing energy has increased steadily in the last few years, e.g. for SCP, from five in 2007 to 12 in 2010 up to currently 35 suppliers (source: regelleistung.net, 18/03/2016). The situation is similar for PCP and MRP. Accordingly, a general fall in the cost per kilowatt can be observed for SCP and MRP, while the price level for PRL remains stable for the time being due to restrictive technical requirements. In general, the question is whether balancing energy as the sole business model for flexible systems generates sufficient revenue, or whether marketing should also include the spot market and especially the intraday market which currently offers an attractive alternative.
At the moment, both the spot market and the intraday market are significantly more attractive to traders than long-term and balancing energy markets with regard to volatility and in view of a possible concentration on high-priced hours. The transaction-related pricing of the intraday market in particular offers traders interesting opportunities to profitably exploit a price spread for one and the same quarter hour in the course of trading without having to physically deliver, as extreme price spreads can occur in individual quarter hours (see Figure 3).
It is foreseeable that the further increase in volatile infeed from EEG plants will further strengthen this development so that short-term marketing opportunities will become more important. At least the flexibilities left over after the marketing of balancing energy should therefore be placed on the spot and intraday markets.
Different marketing forms should therefore always be considered depending on the technical properties and parameters of a plant. As a consequence, when a plant or a pool of plants is suitable for several or all of the marketing forms, the immediate result is a complex optimisation task: how is plant flexibility, i.e. the flexibly available power and energy, used over the various markets and time horizons to achieve the maximum revenue in a volatile market whilst taking the plant's basic technical and economic conditions into account? This question is all the more essential as plant flexibilities can only be marketed in parallel to a very limited extent unless the cost per kilowatt hour and basic price are intentionally placed in strategic opposition and in acceptance of the associated risk when creating offers for SCP and MRP.
A first possible optimisation strategy can be derived from the market's defined deadlines for bid submission (e.g. Tuesday and Wednesday of the previous week for PCP and SCP, or 45 minutes prior to delivery for intraday). In turn, a "cascaded" procedure can then be derived which markets availabilities in the sequence of deadlines. An attempt is made to place an optimum quantity of power or energy on the market for each deadline; the corresponding optimisation is achieved through the calculation of the margin prices and the respective quantities. A further possible strategy for marketing quantities can be implemented by prioritising the market or markets most lucrative at a specific time. Reliable price forecasts are essential as a basis for this strategy.
Ultimately, the most comprehensive approach lies in optimising marketing at all times in parallel across all markets and time periods, which requires complex modelling and more complex calculations. However, each of these possible approaches demands very sophisticated, high-performance mathematical methods as the relevant market parameters (e.g. prices, demand, load) of the different markets change dynamically and therefore require stochastic treatment.
Short-term marketing places much higher demands on IT systems, firstly with regard to system performance and the possible process automation, and, secondly, with regard to technical requirements and in particular to optimisation. This is illustrated by the following examples using various PSI products.
The product PSIvpp supports the marketing and technical control of virtual power plants in the above mentioned markets for balancing energy and energy supplies. The available power and energy has to be distributed flexibly and optimally across the various market levels of the balancing energy, spot and intraday markets. This requires intelligent and powerful bidding strategies which ascertain the energy quantity and prices of all bids for each market level under volatile market conditions.
PSI handles this task with the stochastic optimisation software TS Energy which markets flexibility in a cascaded optimisation process on the basis of different bid deadlines. The bids for ancillary energy are created using the opportunity costs for different power reserves. Different procedures are available for allocating these costs to kilowatt hour rates and basic prices; some of the procedures make use of historical data. If a bid is not accepted, the flexibility is made available for marketing on the next market level until the final remaining flexibility is ultimately traded on the spot and intraday markets. The existing price structures and volatilities require market participants to make decisions of a particularly high quality — this is guaranteed by the optimisation module TS Energy.
Using a stochastic approach allows uncertain variables such as weather conditions, commodities and electricity prices to be taken into account overall and the optimum margin prices for day-ahead and intraday trading to be generated to support trading activities. The margin prices which, under consideration of stochastic volatility, will maximise revenue in the long term are calculated on the basis of input data such as price forecasts, stored capacity, inflow forecasts and potential operating restrictions. The margin prices produce a bidding curve and can be used for automated trading (cf. Fig. 3). On the intraday market, the margin prices resolved every hour can serve as the minimum or maximum bids for deviations from the schedule. Analyses of PSI customers have shown that this stochastic-based approach achieves significantly better results in volatile markets than other methods.
The following bid submission requires automatic interfaces to the appropriate platforms (e.g. Trayport etc.) for submitting bids and receiving contracts and also for processing the delivery on call. This applies in particular to the intraday market, where continuous trading results in many hundreds or even thousands of bids and concluded contracts in the course of the day. These requirements can only be covered through trade automation and high-performance processing, including the subsequent middle and back office processes such as position management and nomination.
One of the essential instruments for ensuring system performance in the PSImarket trading system is central change monitoring; it logs and analyses changes and ensures the necessary updates to specialised processes occur in good time. The following example demonstrates the resulting quantity structures and performance requirements: tens of thousands of data changes per minute are possible in a trading system with approximately 50 users and around 80 active jobs. A central change management system ensures that these changes are processed and managed efficiently in the system.
These examples represent only a small portion of the high demands that trading on short-term markets places on the IT systems used for this purpose. Energy traders who master these requirements, however, will nonetheless find very attractive opportunities in the new markets.
Photos (top to bottom): Michel Martinez, PSI Energy Markets