The future - Wind Power Integration: Connection and System Operational Aspects - page 260

will be fewer, but larger, turbines. Use of variable-speed technology will increase
energy capture by perhaps 10 per cent, while the greater wind speeds intercepted by
greater turbine height may increase output by a further 10 per cent. The original
electrical infrastructure has a life of 50 years at least, but may need to be uprated.
However, a more cost-effective option may be to employ some storage at the wind
farm to absorb the extra 20 per cent of rating. A battery store, connected through an
inverter at the wind farm substation, could be topped up when wind is plentiful. The
store could then be used for reactive power support and fault ride-through, appli-
cations which don't require much energy storage. The store could then benefit from
high system marginal price at peak demand times by generating. In effect, the
energy store performs many of the functions of a diesel generator, but without the
emissions. The combination of improved wind turbine technology and storage will
improve capacity factor. This, and avoidance of infrastructure investment, should
deliver cheaper energy than the original wind farm.

8.2 Grid codes and beyond

A wind energy penetration over 15 per cent implies that wind power could pre-
dominate over conventional generation at certain times. The main operational
problem then is that frequency regulation falls on a narrower base of thermal plant.
To obviate the problem, grid codes have for some time required that wind farms
should be capable of providing frequency response. While many wind farms now
have the relevant capability, it is not utilised at the time of writing. However, there
is an economic, as distinct from operational, incentive to operate in such a mode.
When wind power displaces a significant tranche of thermal plant, system marginal
price can decrease to the point where the cost of reserve exceeds the cost of energy.
In these circumstances, frequency-responsive wind plant should be backed off so
that its emergency reserve becomes available to the system. In effect, the wind
sector is providing emergency reserve to cover the possible forced outage of ther-
mal plant, while the thermal sector provides the regulating reserve needed to follow
demand and wind power variations. Such a symbiotic arrangement has the further
advantage that it reduces thermal plant ramping (Tang et al. , 2013). Energy markets
should encourage such co-operative behaviour by paying for reserve. Unfortu-
nately, it is more usual to compensate wind generators when curtailed, removing
any incentive on their part to contribute to frequency regulation. This is short-
sighted, and limits long-term wind power development. It may be noted that energy
storage located at re-powered sites is well suited to emergency reserve provision.
Involvement of the wind sector in frequency regulation should allow wind
capacity to further increase towards an energy penetration of 30 per cent. With
wind power now on occasion able to provide over 70 per cent of the demand, the
dynamic and voltage stability of the system become the over-riding concerns
(Flynn et al. , 2013). Grid code requirements for fault ride-through tend to assume a
strong synchronous generation backbone. High penetrations of non-synchronous
renewable generation will require detailed study of the relevant system. There are a