Australia: where too much wind will never be enough

With the onset of a new Federal Government what might we expect from an Environmental Minister who wanted to limit our use of off-peak electricity for heating water? Perhaps an early warning signal of what to expect comes in a paper by Mark Diesendorf.

The paper is The Base Load Fallacy and explains how there is a real alternative to coal fired power stations. It is of course not nuclear energy but all the renewable energy sources of biomass, wind, solar, geothermal and gas.

Unfortunately, present experience does not support this view.

The starting point is the broad brush statement in the paper that no power supplies are perfectly reliable. This is correct provided you don’t ask about the details. If you did, the devil would point out that there is a difference between a naturally intermittent supply and a supply which trips or goes off line unexpectedly. There is a difference in scale and time. Contemporary distributed electricity systems have devised ways of insuring continuity of supply for the latter events but are struggling to deal with the former. This is not comparing like with like.

Electricity systems have to deal with varying demand over the day. There are two parts to this, the broad movement where demand shifts from an early morning low to the broad working day, to evening and back to the early morning. Power stations are scheduled to pick up the supply which in Victoria would move from 5,000MW in the early hours of the morning to 8,000 or 9,000MW later in the day. There are also variations in demand in the order of 10 to 100MW within five-minute intervals: several generators are given the task of following the demand while keeping the voltage and frequency stable.

Intermittent supply adds an extra stretch for the control of a network. Wind farms illustrate the problem. A standard measure of the performance of a generator is the capacity factor. This is the annual averaged power achieved as a percentage of the installed (or maximum) capacity. For wind farms in the southeast of Australia the capacity factors range over 30 per cent to 40 per cent. So 300MW of installed capacity averaging an output of 100MW gives a 33 per cent capacity factor. The comparable figure for thermal power stations is 90 per cent or more.

But this factor gives no indication of the detailed performance. A measure that helps give an indication of this is a reliability figure. This is the minimum percentage of power that may be relied upon for 90 per cent of the time. For wind farms it is about 5 to 10 per cent: so, for example, for 300MW and a reliability figure of 10 per cent we can rely on 30MW. The comparable figure for thermal power stations is again 90 per cent or more. So planning to match demand with supply may require substantial reserves to back up the unreliable wind farm output.

For instance, if Victoria had over 1,000MW of installed wind power and the wind farms were supplying 1,000MW to the grid, near peak performance, all this could be lost in an hour when the wind stops blowing.

Filling a supply imbalance requires conventional thermal power stations. The loss in the example is beyond the supply reach of gas turbines that are usually clusters of 50MW units and used for short term peaking while loss of wind can be a matter of ten hours, a day or more.

The limit for wind power contributions is thought to be about 15 per cent to 20 per cent of supply. As an example of German experience, the E.ON Netz, group, with 7,000MW of wind farms in its region, states in their Wind Report 2005 that:

In order to guarantee reliable electricity supplies when wind farms produce little or no power, eg. during periods of calm or storm-related shutdowns, traditional power station capacities must be available as a reserve. This means that wind farms can only replace traditional power station capacities to a limited degree

Further, the Renewable Energy Foundation in the United Kingdom states: "Wind energy cannot replace conventional power stations to any significant degree."

So, the final point on reliability of supply is that perhaps a geographical dispersal of wind farms will smooth out supply. In principle this is correct as the wind will be blowing somewhere but remote sites require expensive connections.

But the data recorded by NEMMCO for four wind farms in South Australia from north of Port Lincoln to the Victorian border, a distance of about 500km, shows strong correlations for wind output and no significant smoothing of supply. The reliability factor is less than 10 per cent. Even more interesting the Victorian wind farms also show a response correlated to the South Australian wind power output. It is no better in heatwaves which are times of peak demand.

None of this should actually be surprising as weather systems in south eastern Australia extend over 1,000km.

Wind power has, so far, been the front runner for renewable power. But it can be seen that wind power is inappropriate for base load power. Policy makers should proceed cautiously gathering information. Theory should not be allowed to get in the way of experience.