• wind power production is not dispatchable except that the production can be curtailed
• as the fuel costs of wind power are zero, the short-term marginal costs of wind power production are low, i.e., US$.01 per kilwatthour or lower.

• A financial market trading different financial contracts used for risk management and speculation, e.g., forward and future contracts.
• A wholesale day-ahead power market (Nord Pool spot) trading hourly contracts for physical delivery of power during the next day. The day-ahead market is settled once a day by aggregating supply and demand bids from market participants into a supply curve and a demand curve and finding the intersection between the supply and demand curves for each hour of the next day. The day-ahead market is divided into a number of price areas. The available trading capacity of the transmission lines between the different price areas is determined by the TSOs from security of supply considerations before the market clearing process begins. Usage of these transmission capacities is determined by the spot market clearing process, such that transmission bottlenecks cause different day-ahead power market prices between the price areas.
• An intraday power market for continuous trading of one-hour power contracts up to one hour before delivery (named the Elbas market) for each hour of the day. This market enables market participants to trade changes from their planned supply or consumption schedule determined in the day-ahead market, thereby reducing the deviations between planned and realized production and consumption during the actual operation hour.
• A market for regulating power (real-time market) where flexible producers and consumers submit bids for upward regulation or downward regulation of production or supply with an activation time of at most 15 minutes. Bids can be submitted up to one hour before delivery. The Nordic TSOs monitor the frequency and activate upward or downward regulation within the operation hour in question by choosing the cheapest regulating bids available.
• Markets for ancillary services, e.g., spinning (primary reserves), are operated separately by the TSOs in each country.

The owners of wind turbines use wind power forecasts when bidding their wind power production into the day-ahead market. Since the short-term marginal production costs of wind power are very low, economic theory implies that in a day-ahead market with well-functioning competition, the wind power production should be bid into the market with a low selling price to ensure that as much wind power production is sold as possible. The wind power production therefore reduces power prices in the day-ahead market by replacing more expensive types of power production. As more and more wind power production is introduced, power prices decrease more and more, and the production of conventional power plants is reduced more and more. In Western Denmark, characterized by a high share of wind power production, there have been several production hours during the night with a day-ahead power market price of zero. This reflects the fact that the relatively low load existing in these hours plus the scheduled interchange could be entirely supplied by wind power generation, and production from conventional plants which cannot be reduced due to system stability constraints and the need for heat production from combined heat and power plants.

The reduction in day-ahead power market prices is beneficial for power consumers but creates problems for the conventional producers that face both decreasing prices and a decreasing number of operation hours. They can react by taking nonprofitable conventional units out of production. This will create higher power prices in hours with high load and low wind power production and might cause problems with the capacity balance in some hours. Alternatively, they can cover their investment costs by making supply bids with a relatively high price in hours with a low expected wind power production and high load. In the Nordic system, such behavior might cause the attention of the competition authorities monitoring the power market because conventional power producers are expected to sell power at a price close to their short-term marginal costs. Both reactions would cause volatile day-ahead power prices characterized by alternating patterns of high and low prices. Finally, the conventional power producers could cover their investment costs by increasing their revenues in markets other than the day-ahead market, i.e., by shifting from selling energy (active power in the day-ahead market) to selling flexibility (e.g., regulating power). This strategy might be feasible to some extent as the errors connected to wind power forecasting increase the demand for regulating power compared to a situation without wind power.

In summary, the effect of increasing the share of wind power production in the Nordic power system will lead, on average, to decreased prices and increased price volatility in the day-ahead power market, in combination with a higher demand for flexible power production, i.e., higher prices in intraday and regulating power markets.

So how do these changes influence long-term investment behavior? The increased volatility in day-ahead power market prices gives rise to higher volatility in the revenues of the investment. According to investment theory, this will lead to a higher risk premium required by investors, i.e., higher profitability required of new investments. Adding to the investment uncertainty is the fact that the expansion of wind power production is governed by support schemes determined in a political process that can be hard to forecast, i.e., there can be a lot of uncertainty connected to future wind power development. Increasing risk premiums might, to some extent, reverse the trend of decreasing day-ahead power prices because investors will postpone investments until price expectations cover investment cost.

Higher investment uncertainty and more focus on selling flexibility relative to selling energy should shift investments more in the direction of power plants with relatively smaller investment costs and faster start-up times, e.g., open cycle gas turbines, at the expense of base load plants with higher investment costs and higher production efficiencies. Also, with very high shares of wind power production, storage technologies like pumped hydro storage or compressed air energy storage might come into play. Furthermore, utilizing the flexibility of power consumers to shift the load into periods with high wind power production will be very useful. Plug-in hybrid vehicles might offer very interesting possibilities in this area in the longer term.

It will be interesting to observe how market participants will react as wind power production increases, thereby increasing the number of low price hours in the power markets and reducing the number of operation hours of conventional plants. It is an open question as to whether or not the uncertainty concerning power plant investments in the Nordic power market due to wind power development, CO emission permits price development, fuel price development, and building of interconnections to continental Europe will postpone necessary investments enough to threaten security of supply in the longer term. In my view, we need to address these issues sooner rather than later.