APPLICATION OF COST FUNCTIONS FOR LARGE SCALE INTEGRATION OF WIND POWER USING A MULTI-SCHEME ENSEMBLE PREDICTION TECHNIQUE - Energy Research Developments: Tidal Energy, Energy Efficiency and Solar Energy - page 340

To conclude, a single wind power plant on its own will not be able to compete with a
pool of many plants, unless it is located in unusual predictable weather conditions and hence
could be considered a semi-scheduled plant. However, there are also means to operate as an
individual on the market by outsourcing the handling of the wind power trading to parties
that are specialised in this and may pool the energy of their aggregated customer's power
to a larger portfolio. Sufficiently large wind generators may in the future also consider to
forecast and trade themselves in order to get the market price in addition to some incentives,
but also the penalties. Nevertheless, this is expected to be the most profitable way forward
for wind energy in the future.

3.3. The Skew Competition in the Trading of Wind

The wind generators are some of the weakest parties on the energy market seen from
a trading perspective. Everybody with a good weather forecast and with a reasonable ap-
proximation and experience on the market can predict the wind energy generation in windy
periods and therefore also the impact on the market price from wind. This is not so much a
problem as long as the fraction of wind energy of the total generation is less than ∼10 %.
Above this level, this becomes a more serious problem, since the wind power traders are
forced to bid on the energy to a relatively low price or the traders risk to be forced to sell
the energy to a pool for a lower price after gate closure.
As previously mentioned, it may be the most beneficial way to trade wind energy with a
pool that contains sufficient storage to create additional uncertainty on the market regarding
the available energy than what the weather forecast provides. In such cases, the market
can no longer predict the energy generation from wind power and speculate as aggressively
against the wind power trader. The market can as an example not know, if the wind energy
pool decides to store energy for some hours instead of delivering the power and thereby
take the peak off the wind power generation. The scheduled generators can in time periods
of strong wind speculate against the wind power trader by setting a price that lies below the
marginal costs, because they know that the wind will be available for a marginal price, if the
wind generation was bid in with a higher price than their own. If the scheduled generators
know that wind generation may not be available for a near zero price, because it may be
withheld from the market for storage, then the risk becomes too high and the scheduled
generator will stop such speculations.
Pooling can hence be regarded as a security measure against unfair speculation. The
pool will in that case be used in daily operation to phase shift wind generation such that the
correlation between the demand and total output of the pool is highest.

3.4. Uncertainty Considerations

Another factor that has impact on the market price and competition level is the uncer-
tainty of the weather forecast. Uncertain weather requires that more energy generation is
active, either as primary energy or as reserve. This again increases the average energy price,
because a larger fraction of the available generation will be required for balancing.
Steep ramping of wind power is an indicator of uncertainty in the weather and often
even a cause of such. Nevertheless, steep ramping requires more scheduled capacity online,
because the efficiency of the ramping generators is lower during the ramp.