Operation of power systems - Wind Power Integration: Connection and System Operational Aspects - page 214

its peak). On the positive side, there is often a useful correlation between load
behaviour and weather patterns. So, in northern Europe for example, it is generally
windier in the winter relative to the summer, and windier during the day than at
night (see Figure 5.9). Windy, winter evenings will also introduce a wind chill
factor that increases the space-heating requirement further.
Through balancing the output of an individual wind farm against a local (sto-
rage) load, network impact can be lessened, and it may even be possible to increase
the capacity of the wind farm. For example, at the Mawson base in Antarctica the
large, controllable heating load enables 100 per cent wind penetration for up to
75 per cent of the year, with fuel cells providing longer-term storage (AGO, 2003).
However, the true benefits of such arrangements only accrue when extending and
co-ordinating the scheme across larger network areas. A study for the Northern
Electric regional electricity company (REC) in the United Kingdom indicated that
active control of consumer load could enable an additional 1,600 MW of onshore
wind farms to be accepted on to the network (Milborrow, 2004). Under system
emergency conditions, the ability to switch off load quickly also offers a valuable
source of operating reserve. Within the Long Island (New York) Power Authority
(LIPA) it was proposed that responsive load act as fast/emergency reserve (Kirby,
2003). Residential and small commercial air-conditioning load was mainly exam-
ined. LIPA has over 20,000 residential consumers and 300 small commercial con-
sumers, who have centrally controlled air-conditioning thermostats for demand peak
reduction to alleviate network constraints during the summer. Demand reduction of
25 MW is available, while a conservative estimate of 75 MW of 10 minute emer-
gency reserve should be achievable at low cost during periods of high demand -
reserve provision is highly correlated with the system demand. In the LIPA scheme
demand reduction is achieved by raising the temperature setpoint, while reserve
would be achieved by completely shutting off the load. Hence, the same load can
provide both demand reduction and reserve capability. Similarly, within the PJM
(Pennsylvania, New Jersey and Maryland) control area, curtailment service providers
(CSPs) provide load reduction through control of residential electric water heaters,
water pumps, thermal storage space heaters, industrial freezers, HVAC equipment,
etc., with approaching 8,000 MW of demand response capacity in 2013/14. Com-
munications and dispatch of load control switches is achieved using a telephone/radio
communications system, managing the daily peak demand.
Forecasting of load behaviour can generally be achieved with high accuracy, as
discussed in Section 5.2. Forecasting of the availability of responsive loads should
therefore be equally achievable. It is perhaps even easier, since there is likely to be
less diversity (relative to the total load) in the size and rating of the storage devices,
and they are likely to be driven by the same time of day, time of year and weather
patterns. The difficulty often faced with load management schemes, however, is the
ability to control and co-ordinate switching operations centrally. Traditionally,
communications and control technology made it much easier to monitor the
operation of a few large resources (generators) rather than the multitude of dis-
persed small resources (loads). Radio teleswitching and mains signalling with
ripple control have been applied for many years for water heating, so that, for