The challenge in Australia, as in every other carbon heavy economy (i.e. China) is to provide energy security and reduce carbon emissions. Hopefully this can be done in a cost effective away. Hence with energy trilemma.
Despite 5GW of new supply under construction, the industry is presently operating without any indication of what long term policy will be. In one sense this is the private sector making its own judgement and the market is responding appropriately to a price signal.
On the other hand many barriers stand in the way of the transformation of the electricity system.
The underlying emissions target is zero but the journey is long
We focus on the 28% reduction in total Australia emissions from 2005 levels by 2030 which is the agreed Australia target. Most people understand that this 28% reduction is just a stepping stone and significantly more is required. So we also consider 50%, even though that too is not enough.
We go with the general assumption that electricity/stationary energy has to bear a proportionate share. Nevertheless, the electricity industry would be right to point out that it gets about 95% of the attention but is responsibly for less than 40% of emissions.
For electricity this translates to reductions of about 50 million tonnes for a 28% reduction and 90 million tonnes for a 50% reduction – 2015 electricity emissions being little changed from 2005 emissions.
Figure 1 Electricity emission targets. Source: Emissions.gov.au, ITK
Closure of Hazelwood removes about 14mt of emissions. In the short term this is largely replaced by open cycle gas and black coal from NSW, but we also know that that the 4.5 GW of utility scale renewables under construction, or just about to start, should produce around 12 TWh of carbon free electricity. So this means the closure of Hazelwood is a net carbon saving.
On that basis we calculate emissions intensity and targets as follows:
For a given absolute level of carbon emissions the emissions intensity moves inversely to electricity demand. Ie if demand is higher intensity is lower. So, if demand was flat in 2030 compared to 2017 the required emissions intensity for 128mt of emissions is 0.67 rather than 0.63.
Low Emissions target
The following section draws heavily from Jacobs modelling for CCA
The low emissions target (LET) is similar to the RET and would operate from 2020 to 2040, with the difference being that low emission fossil technologies are also eligible to earn certificates as a fraction of their emissions intensities below an emission intensity threshold of, say, 0.6t to 0.7 t CO2-e/MWh.
This includes CCGTs, CCS plant, both coal fired and gas fired, and also nuclear generation technology. In practice though we expect the vast majority would continue to be wind and solar PV. And as this will increase the volatility of pool prices it will provide an indirect reward for existing gas generation.
Eligible generators receive certificates adjusted by how much the generation outperforms the threshold intensity, such that a zero emissions generator would receive one certificate for every MWh of output, while a generator with half the threshold intensity would receive half a certificate.
To avoid windfall gains for pre-existing generation, zero- and low-emissions generators existing at 2020 would be eligible to produce LET certificates only for any generation above pre-specified historic baselines (similar to pre-existing renewable generation under the current LRET).
We note that Jacobs states that energy security/reliability would not be an issue in their modelling under any of the potential policies.
How would this interact with the State Schemes?
The short answer is we don’t know and we doubt if anyone does. However ,an easy assumption is that renewable energy built pursuant to a State scheme will reduce emissions intensity. The LET target could be reduced by energy procured under a State scheme but whether that can be made to work practically is another matter.
How much new supply would be required to lose 40MT CO2?
Our calculations are entirely back of envelope. We assume that the technologies available until 2030 are combined cycle gas, wind and solar PV.
Because of increasing renewables penetration the underlying value of dispatachable gas generation is increasing and so whereas in much of the world gas is displaced by renewables before coal because the variable cost of gas is higher than coal, in this case we see coal being displaced.
That said, history shows that making forecasts based on what you think you know about technology and fuel costs is a fool’s business. One day gas is cheap and the next it is expensive. One day futures prices are low the next high. One day new Utility PV is $120 MWh and the next it is $90 MWh.
We assume black coal is largely displaced, which in practice will mostly mean NSW as it is higher cost than Queensland open cut being displaced before brown coal. However, Yallourn will also come under pressures.
NSW black coal has an emissions intensity of about 1. This suggests that an LET target of about 40 TWh of new LET or 40 M certificates. This would require around 12-14 GW of wind a bit more utility PV and even more gas because combined cycle gas only gets 0.4 certificates for every MWh of production.
The November 2016 Jacobs modelling for the CCA referenced above shows 20 GW of new supply being built by 2025, the vast majority of which is wind and maybe 60 GW by 2040 with a lot of wind and PV but also some gas and biomass.
However, we are going to mostly resist the temptation to just copy and paste from the Jacobs modelling because, although it was very thorough modelling and independently reviewed and also allowed for distributed energy, it still seems better to look at things our own way. The Jacobs modelling has more new generation than my guess possibly because it assumes more demand growth.
Carrot and stick
The disadvantage of the LET as with the RET is that it does not directly put extra costs on high emissions generation. By contrast an Emissions Intensity Target [EIT] has high emissions generation making payments to low emissions generation.
So as with the RET the LET works by forcing down the price of electricity in the pool and on the futures market. Eventually this means that coal generators don’t have enough revenue to cover their fixed operating costs and ongoing capital expenditure ,and so they go out of business.
This will usually happen as with Hazelwood or Wallerawang at the time some major item of maintenance capex is required or when bank refinance is required.
The new LET generation will need a price that covers its LCOE. There are plenty of arguments about that but notwithstanding the incredible sub $55/MWh price that Goldwind is taking from Origin for the Stockyard Hill wind farm, we still think PPA prices of around $60-$80 MWh real are a good benchmark.
Once built ,new wind and PV will be dispatched unless curtailed by the operator and so will displace the highest variable cost generation. That will be gas in first instance but the main point is it will drive down the average pool price and this will feed through to lower futures prices.
Existing black coal generation, particularly Liddell (scheduled to close in 2022) and Vales Point have annual costs in the $20/MWh to $30/MWh range (very short run costs are a bit less).
So as for where this leaves the balance between LET certificate prices and the pool price is hard to say.
An LET could well be supported by the ALP as it ticks various boxes. It’s good for investment. It’s good for regional employment and it helps to ward off the greens. It may also wedge the Coalition to an extent.
The Coalition is stuck between having to do something to meet the COP21 goals and its right wing rump. Public opinion, manufacturing industry and the utilities industry all want some clear carbon policy.
It’s not a big vote winner but doing nothing may well be a vote loser. These considerations may outweigh the risk of alienating the rump and the coal lobby.
Learning from the RET mistakes
The RET in our view suffers from the fact that higher certificate prices can be passed on to consumers and that retailers can choose to pay a penalty price and avoid compliance.
We also think that the uncertainty around future certificate prices contributes to booms and busts in the sector and makes the cost of capital higher than it needs to be.
The increasing vertical and horizontal integration of the electricity supply industry also gives a degree of bargaining power to the PPA providers, that is the existing retail/generators.
If an LET is introduced we hope the scheme can be improved over the existing scheme.
David Leitch is a regular contributor to Renew Economy. He is principal at ITK, specialising in analysis of electricity, gas and decarbonisation drawn from 33 years experience in stockbroking research & analysis for UBS, JPMorgan and predecessor firms.