With high gas prices partly to blame for the electricity blackouts that hit South Australia this month, and gas-fired generators caught short in New South Wales two days later, it is hardly surprising to hear calls for Australia to expand production.
Following bipartisan passage of the Victorian moratorium on onshore gas developments, federal Energy Minister Josh Frydenberg called on all governments to support unconventional gas. He has talked of an “urgent” need to increase gas supplies to improve energy security.
But how much extra gas do we need? And how do we square this with the equally pressing need to reduce our use of fossil fuels?
How much gas can we burn?
On average, the National Electricity Market (NEM) emits about 800 grams of carbon dioxide per kilowatt hour of electricity produced. This is almost double the OECD average of 411g per kWh.
According to the International Energy Agency (IEA), this average needs to fall drastically to 15g per kWh by 2050 to achieve the goal of limiting the global increase in temperatures to 2℃. Indeed, the IPCC Fifth Assessment Report shows that limiting global warming to less than 2℃ will require the electricity sector’s greenhouse emissions to reach zero by 2050.
By any measure that is a huge task, particularly for a country like Australia. Currently, around 11% of the electricity in the NEM comes from gas. Even if every coal power station were closed and replaced with zero-emissions technology, the NEM’s emissions intensity would still be three times this 15g per kWh limit.
This means that at current demand levels we need to burn roughly 70% less gas if we are to stay in this emissions intensity range. That’s a particularly small amount when compared with current total gas demand, as shown in the figure above.
Given this constraint, we need to think about how to maximise the amount of electricity we get from this limited amount of gas, and what new technologies can help us do it.
There are several technologies for converting gas to electricity. Older power stations, such as Torrens Island in South Australia, are similar to coal-fired power stations. Energy from combustion is used to heat water, which in turn powers a steam turbine.
Today, gas is generally converted to electricity using two different technologies. First, there are open-cycle gas turbines (OCGTs). These work in a similar way to jet engines: the gas is mixed with air and burned, producing a stream of hot, high-pressure exhaust gas that drives a turbine.
OCGTs are very flexible and can ramp up and down very quickly. They are sometimes described as “peakers”, because they can respond rapidly to peaks in electricity demand. But because of this, they are typically not used very much – some OCGTs in the NEM run at full load for just a few hours a year.
Their thermal efficiency – the proportion of energy from combustion that is converted to electricity – is relatively low, at around 30%. This also means their emissions intensity is relatively high, at 580-670g per kWh.
A second type of gas power station is combined-cycle gas turbines (CCGTs). These power stations effectively recover extra energy from the exhaust stream of an OCGT turbine. This makes them more efficient than OCGTs, typically recovering 50% of the energy from the gas. As a result, their emissions intensity is lower, at roughly 400g per kWh.
One downside, however, is that CCGT technology is less flexible. It cannot stop and start as easily as an OCGT. Hence it tends to run for more of the time, operating more as a source of “baseload” power than as a response to peaks in demand.
As this chart shows, the efficiency of Australia’s gas power stations depends more on their technology than their age. CCGTs are more efficient. OCGTs are less efficient but more flexible, and typically use less gas overall because they are switched on more sparingly. OCGTs are also better suited to load-following and balancing renewable energy production and providing capacity to the market.
This is not necessarily a question of persisting with one type of gas-fired power station over the other. But it is important to think carefully about how we burn our gas as well as how much of it we burn. We also need to think about how that will help us meet our other energy objectives such as reducing greenhouse emissions and integrating more renewables into the grid.
Using more efficient technologies where possible makes sense. Several old and more inefficient plants are currently being used instead of newer, more efficient ones. Torrens Island and the newer Pelican Point in South Australia are a good example.
Pelican Point is a CCGT that is running considerably below its nominal capacity, while Torrens Island is running at high rates. While the decisions on operation of these plants are commercial and made by private companies, the same gas consumed by Torrens Island could be much more efficiently used by Pelican Point. A unit of gas burned at Pelican Point could theoretically deliver around 50% more energy than the same unit of gas burned at Torrens.
Part of the reason for this situation is that different companies own the plants. Engie has cut capacity at Pelican Point in response to high gas prices, whereas AGL has opted to keep Torrens Island running at full steam.
This highlights the difficulty, in a privatised market, of ensuring that power is drawn preferentially from the most efficient facilities. Solutions such as managed closures or forced divestment are politically unpalatable. The much-discussed “emissions intensity scheme” would theoretically help push the market in the right direction.
What about the competition?
Many of the services and capabilities that gas turbines provide are also available from other technologies. Flexibility, dispatchability and capacity (as well as other services such as inertia and frequency control) can be provided by storage, other renewable technologies and the cheapest – demand-side management.
Some of these technologies include concentrated solar thermal, battery storage and pumped hydro, which Turnbull also mentioned in favourable terms in his Press Club address.
Indeed, some of these technologies may be able to outcompete traditional sources of capacity like OCGTs.
Whatever the case, the role of gas will need to be carefully considered, and its use will necessarily be limited. In the longer term, the need to increase gas supply is far from certain.