Could Australia end up with synchronous failure across states?
When wind power is maxing out it’s bad for grid stability — it pushes out the reliable spinning inertia — the massive rolling turbines that relentlessly pull the grid back to 50Hz. Here’s a graph of SA and Victoria wind farms last month, and you can see that for all the thousand kilometers that might separate them, they are controlled by much larger common weather patterns.
Tom Quirk looks out our national grid in light of the SA blackout debacle. The message from South Australia is that wind power does not make for nice stable and synchronous grids. As I mentioned before the whole idea of alternating current (or AC electricity) is about the exact push pull of electrons at a set frequency. The grid lives and dies by its frequency. We can’t add a 53Hz current to a 47Hz one and get a 50 Hz average. When different frequencies meet we get interference patterns – a mess of spikes and dips. Say hello to Lumpy Electricity. Say goodbye to your computer.
Indeed when the frequency hit 47Hz the Victorian interconnector said goodbye to the whole state of South Australia. (See graph 1 at that link).
Tom Quirk expands on this and talks about how heavy spinning turbines (like coal, but not wind) are able to share the load of frequency changes in the grid and restore the frequency back to the sacred 50Hz. He estimates that once wind power supplies more than 20% of the total grid, things can get hairy, and with South Australia at 40% and Victoria planning to jump to 20 – 30% now is the time to figure out those limits and the costs. (Ten years ago would’ve been better). Don’t cry now, but Queensland’s target seems to be 50%. Let’s not mention Bill Shorten, potential PM, who floated some fantasy that the whole nation could get to 50%. The least mad state might be WA. Time to secede? Our state is not on the “national” grid — it’s islanded already and every day. That might be what saved us from copying South Australia.
Please can one government somewhere do a cost-benefit estimate on the value of cooling Australia with bat-killing giant fans?
As Tom points out, wind power needs a lot of transmission lines (and ideally the type that don’t blow over too easily). New transmission lines cost $1 to $3 million per kilometre. So add that to the cost of “wind”. Each kilometer of wire buys an awful lot of cheap coal powered electricity.
Victoria currently has a bit less wind power than SA, but it’s a much bigger market and for some baffling reason, wants to increase its wind power to twice or more the size of SA. Remembering that SA is pretty much utterly dependent on the Victorian grid right now. (Don’t miss figure three below, where Victorian fossils make ten times the electricity that South Australian fossils do). Without enough spinning reserve, the dreaded instability infection can spread. Imagine how much fun it will be if the Victorian part of the national network becomes unstable like SA? Consider that weather systems in Australia blow from South Australia across to NSW and Victoria, so wind production is often all on, or all off at the same time. Every MW of wind capacity added needs to be backed up. (Which raises the obvious question of why we don’t just use the damn synchronous back up in the first place and skip the asynchronous, unreliable stuff?). Are we trying to change the weather?
The owners of Hazelwood coal fired plant in Victoria are thinking of closing it. Like Port Augusta, which closed in May, Hazelwood is struggling to return great profits in an artificially distorted market that favours intermittent “climate changing” energy over reliable cheap predictable electrons.
If Hazelwood closes and is replaced with wind power, then South Australia’s unstable network will be paired with another network that will frequently also have “lots” of wind or “not much” and need spinning reserve at the same time. On those days the interconnector might not save one state from the other. If NSW copies the meme, it gets even worse.
Problems and Limits for Wind Power
Guest post by Tom Quirk
ABOUT: A founding director of the Victorian Power Exchange and then Deputy Chairman of VENCorp ( Victorian energy networks for electricity and gas)
The blackout in South Australia on the 28th September has received worldwide attention. The reasons for the blackout are not completely understood but it is a combination of the collapse of transmission lines, the extreme variations in the power output of wind farms and the stability of an inter-connect to the state of Victoria.
There is no doubt that the necessary geographical spread of wind farms requires long transmission lines. This can be seen in Figure 1, where wind farms to the east and west of Spencer’s Gulf need transmission lines of up to 600 km to Adelaide and wind farms in the south east of the state have some 300 km to bring power from near the border with Victoria. There is no “ring main” that gives security of supply for some of the distant wind farms and although the wind farms produce on average only 30% of their maximum output, the transmission lines must be able to handle the maximum possible output for all wind farms that feed into the transmission network. New transmission lines cost $1 to $3 million per kilometre so the economic case for increasing the network may not have justified its cost and the difficulty of assessing the security value of new connections.
The total installed wind farm capacity for South Australia is 1576 MW. The State of Victoria has an ambitious plan to move from the present 1242 MW of installed wind farms to add 3000 to 4000 MW to provide 20% to 30% of the electricity demand from wind power.
For Victoria, which supplies the balancing power to South Australia, an increase of supply from wind will increase the risk of supply failure. This is due to the weather patterns of south-east Australia that give rise to correlated wind in South Australia, New South Wales, Victoria and Tasmania. This has been shown in published analysis from Paul Miskelly in a peer reviewed journal Energy & Environment in 2012.
This correlation can be seen in the wind farm performance in September 2016 shown in Figure 2 where the wind farm variations are larger in South Australia reaching 1400 MW while Victoria lags with 1000 MW.
The performance of fossil fuel generators for South Australia and Victoria is shown in Figure 3. For South Australia the gas fired thermal power station at Torrens Island and gas turbines partially even out the wind power variations. The balance of supply comes from Victoria through the Heywood inter-connect. Victoria has a significant wind farm contribution to meet local Victorian demand. However the bulk of supply comes from the brown coal burning power stations of the Latrobe Valley with Hazelwood and Loy Yang as load-following generators. The gas turbine generators and particularly the Murray hydro power generators match much of the Victorian daily demand variations.
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.
The need for synchronous spinning
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.
Distorting the market
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.
All for what? CO2 “saved” is inconsequential
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?