Problems and limits for wind power

Managing the supply system requires a constant frequency of 50 hertz within limits of +/- 0.15 hertz while at the same time meeting the electricity demand load.

The regulation of the system is given to load-following generators but their task has become increasingly difficult as increasing variations in wind supply has been added to demand variations. Coal burning generators have borne the brunt of this regulation but gas turbine generators also play a part as do hydro power plants, the latter two forms of generation providing the most immediate response to load changes.

What is also not readily appreciated is the importance of the very real property of "system inertia" that is inherent in a group of generators that are operated at synchronous speed. All conventional generators connected to the grid in Australia spin, very precisely, each at a speed which corresponds to 50 Hz. This figure of 50 Hz can be thought of as 3000 cycles per minute (50 times 60). For a generator with two magnetic poles, its spin, or synchronous, speed is 1500 rpm. For a 4-pole generator, the spin speed is 750 rpm. The generators are electrically locked into this same, hence synchronous, speed. Indeed, any generator which might begin to stray a small amount for some reason from synchronous speed is automatically pulled back into lock. This synchronicity of operation is an inherent property of the operation of these synchronous machines: There is no requirement for any sort of active control system to bring about this speed regulation, that is, it is inherently fail-safe.

This property of synchronicity leads to another property, a property that becomes critically important in dealing with transient faults, overloads, short circuits and open circuits. As the spinning rotors of the generators are all locked together at synchronous speed, the mechanical inertia of the rotating system is the sum of the rotational inertias of all of those locked-in-synchronous spinning rotors. If there is a sudden load increase, then that load is shared by all the generators, completely automatically. In this instance, the speed of ALL of them will drop together, to the same extent. This speed change IS the means by which the change in the load is sensed and a throttling correction applied. But, what is important for transient changes in load, such as a flashover due to a lightning strike, a short or open circuit due to a transmission cable disconnection or a generator dropping out, is that, on a network of synchronous generators, any sudden load change can be absorbed by the collected sum of the operating generators.

This inherent safety in dealing with transient faults seems to have been ignored in the lead-up to the severe weather event that affected South Australia on 28 September 2016. It was important that as much synchronous generation as possible ought to have been powered up, spinning in synchronism with those other, few, synchronous generators that were actually operating on the South Australian grid on the day.

Where wind farms, and other non-synchronous forms of generation are used, it needs to be very clearly understood that these forms of generation do not share this protective safety feature. A significant part of any line or load transient occurring near a given non-synchronous generator must be borne by that generator as if it is acting alone. The result is that such generation may well be far more prone to shutdown in the event of nearby transients than is a synchronous generator. Thus, reliance on such as wind generation to provide a large share of the total generation during any severe weather event, or similar situation where the transmission system is subject to disturbance, or potential disturbance, is not a wise strategy

So if Victoria has a target of 4000 to 5000 MW of installed wind farm supply then the variations of supply will approach the situation in South Australia for load following. This would require the Victorian generators to cope with correlated variations in South Australia and Victoria with variations of as much as 3000 MW. Although the installed capacity of wind farms in New South Wales is only some 500 MW, these will also have a degree of correlation with the southern states so the system will need to be able to handle 4000 MW variations. This is the key question as load-following generators were developed to handle demand changes of 10's of MW per minute but, with the projected increase in wind farm installed capacity, the short term supply changes may increase to a requirement of 100's of MW per minute. The creation of more interstate transmission lines may not help when simultaneous variations occur in all the States.

The conclusion for the proposed Victorian increase in wind supply is that the possibility of blackouts will be increased, even though the Victorian transmission network with a "ring main" around Melbourne and the array of transmission lines from the LaTrobe Valley is comforting, provided that coal burning power stations are not closed down.

The real distortion to the system is the treatment of wind generated power. It is described as non-dispatchable as it must be used when generated. Wind farms do not bid a price into the wholesale market but rather take what is on offer and collect a legislated $40 per MWh from distributors who pass this cost on to the users. The consequence of this is a distortion of the market that drives out high priced generators whose actual costs are less than that of the subsidized wind farms. This occurred on 28 September 2016, when the Pelican Point and much of the Torrens Island gas-powered generators in Adelaide were off-line during most of the day.

The physical and financial integration of wind power into our power networks has not been thought through in any careful or precise way. All that has been put in place has been a series of, seemingly ever-increasing, ad hoc, "renewable energy targets". It is not clear what the physical limit on renewable energy might be but the experience of South Australia suggests that the danger zone starts when more than 20% of supply comes from renewables. Perhaps it is useful to think of the criterion of "spinning reserves", where the concept is that the largest generator supplying power to the system is always to be shadowed by a generator of similar capacity, or a collection of generators whose summed capacity is of similar capacity to this largest generator to guard against a sudden loss of power. For wind the shadow capacity would have to be the total installed wind generation capacity matched by conventional generators and with no interstate support. The behaviour of South Australia, from recent and bitter experience, shows insufficient reserves were available when the network became isolated from interstate support. This test would suggest that if the inter-connect to Victoria supplies some 400 MW then South Australia has 400 MW too much installed wind power and this absence of reserves is a reflection of the distorted wholesale market.

The financial subsidy in the wholesale market is beyond redemption and the subsidy should be eliminated from all proposed future wind farms

There is a cascading series of orders of magnitude that are largely absent from the political approach to the climate change issue. As a world total we generate some 27 gigatonnes of carbon dioxide annually from the use of fossil fuels. Forest and peat fires in the tropics generate 13 gigatonnes of carbon dioxide annually. China current annual production is 9 gigatonnes of carbon dioxide and it plans to have an annual increase that is equal to the total annual carbon dioxide emissions from Australia of 0.33 gigatonnes of carbon dioxide. The contribution from South Australia is 6% of Australia’s emissions and it is of no consequence but what about the cost?