This is the stuff of dreams.
During a dry year, hydropower generation is reduced by about 4000 GWh. Since the 1950s, power stations at Meremere and then Huntly, have maintained a coal stockpile sufficient to substitute for the hydropower shortfall. The government believes that the Onslow scheme will eliminate the need for the stockpile.
The scheme envisages building a high dam at Lake Onslow and a tunnel from the dam to the 1200 MW underground power station near Lake Roxburgh. The generating plant would comprise a number of pump/turbines that rotate in one direction to generate electricity and in the reverse direction to pump the water back up into the lake.
The scheme is based on operating Lakes Tekapo and Pukaki at a much lower level so that they can store floodwater instead of spilling it and then generate hard to bring the level down and provide surplus power to pump water into Lake Onslow. Pumping could also take place when there was surplus wind or solar power.
In a dry year the water would flow back down the tunnel and generate electricity before being discharged back into Lake Roxburgh.
Most pumped storage schemes operate in a much different mode. They have sufficient storage for only 6 to 10 hours of operation and pump when power is cheap and generate when prices are high. Pumped storage stations also operate in a system support mode responding rapidly if the system is suddenly short of power or suddenly loses a significant amount of demand. They can also stabilise system voltage during disturbances. Many of these stations change operating mode dozens of times a day.
There are two major problems with Lake Onslow’s mode of operation. The first is that, unlike a conventional pumped storage scheme, it will spend a lot of time waiting for a flood so that it can pump and waiting for a dry year so that it can generate. It might generate a substantial amount of power once every five years or so. This means that the utilisation factor of the plant will be very low and, when it does generate in a dry year, the return from generation needs to be very high.
The second problem is that because of the way our existing electricity market works the other generators would jack up the power price as soon as pumping started so depriving the pumped storage scheme of cheap power. This problem could only be fixed by a major overhaul of the electricity market and this might delay completion of the scheme by several years.
Evaporation from the lake depletes storage. Evaporation losses in Central Otago are about 1 m per year so that the lake would lose about ~700 GWh (15% of storage) to evaporation every year. When this loss is added to the pumping/generating efficiency of 75%, the overall efficiency drops to around 60%. For every 5000 GWh of pumping power, only 3000 GWh would be generated.
Filling the lake is likely to take several years depending on rainfall and the amount of renewable energy available to fill it. If we get two dry years in quick succession the lights may well go out in the second year.
It cannot be in service by 2030. Feasibility studies including extensive drilling along the line of the tunnel will take at least one year, environmental approval will take two years or more, capital raising another year, final design and tendering another two years, six years for construction and then three years to fill the lake. It might be ready in 2037.
Economically, it is a disaster. It is estimated to cost $4 billion but, after allowances for interest during construction, escalation and contingencies, it is likely to cost $5 billion or more.
Annual capital charges at 6% would be $300 million, operation and maintenance would be about $20m and transmission charges might be $5 million. Making up for evaporation loss at 5 cents/kWh could be as high as $35 million per year. Refilling the lake every five years at 5 cents/kWh would amount to $40 million per year. So the station has to earn an average of $400 million per year ($2 billion every five years) just to break even.
Assuming that it needs to generate 3000 GWh one year in five it needs a price of $0.70/kWh to bring in $2 billion.
Because of the way our electricity market works all the generators would be paid this $0.70 so each dry year would cost the consumer and the economy more than $20 billion and the domestic power price will treble. This would fall heaviest on poor people and decimate our economy.
Onslow would save an average of 300,000 tonnes of CO2 per year at a cost of at least $1300/tonne. The current CO2 price is $32. Keeping the aluminium smelter in operation could save 2,500,000 tons of worldwide emissions each year at a cost of about $100 million or $40/tonne and has the potential to significantly minimise the dry year problem.
It is easy to see why the Interim Committee on Climate Change recommended burning gas during dry years. Why was this advice ignored?
Bryan Leyland is a Consulting Engineer with wide interests in modern technology. His columns expose the truth behind popular misconceptions.