ALP battery subsidy could save $100m a year, and reverse auctions will also deliver cheaper energy

What was announced?

The ALP has announced it will part subsidise 100,000 household batteries and use the CEFC (Clean Energy Finance Corporation) to run reverse auctions to help procure renewable and low carbon energy to meet a 50% renewable target.

These are excellent policies. I might have gone bigger on the batteries but perhaps keeping it small to start with, as Simon Corbell did in the ACT, and making sure the policy works as designed is a good idea. The other criticism of the battery policy is the means test. The idea is to get the industry volume up and early adopters are those that can afford to be early adopters.

It’s not a matter of supporting the ALP in this, it’s a matter of supporting clever policy for a clever country. South Australia has a Liberal Govt and a similar battery policy and I support that equally.

The batteries get the attention and will kickstart an industry that is perhaps for the first time, ready to be kickstarted, but it’s the CEFC reverse auctions that will have by far the bigger impact for the next few years.

What are the problems the policy solves?

The policy solves several problems:

– Reverse auctions reduce risk and increase planning oversight during a critical economy impacting transition, and at the same time minimises political risk;

– Household batteries are thought to be a useful tool, but have not really taken off. The federal Labor scheme adds to the Victorian, South Australian, Queensland and ACT schemes and should give the industry a real start. This will be developed below.

Here it is necessary only to say that the products and their integration into solar and electricians’ training have come a long way in the past few years and in my view, they are ready to be scaled up.

Lowering the price of renewables by accessing the low-cost of capital of the federal government and providing clarity around a steady flow of new investment

We published as far back as June 2016 an article suggesting using the CEFC to run reverse auctions. We think the CEFC is an excellent vehicle for this, if it must be done at a federal level. It has the board and management structures and is used to running a big balance sheet. It also has a low cost of capital.

We demonstrate some advantages of this below;

You can argue about the parameters but the point is there is roughly a $12/MWh saving to consumers from lowering WACC (the weighted average cost of capital) by 2 per cent on a solar project and nearly as much on wind.

Although the detail is yet to be revealed, announcing a reverse auction rolling three-year program has the following advantages:

a) It will allow everyone to see the new investment coming.

b)The transmission investment can be structured to keep up.

c) Boom and bust conditions are avoided.

d) Existing thermal players can make their exit strategies with more confidence.

e) Because of the fast build of wind and solar it minimises one of the main problems of govt infrastructure, that is poor ability to adjust to changes in market conditions either of cost, price, supply or demand;

f) New dispatchable generation needs like competitive bids for pumped hydro can be accommodated and planned as required by AEMO

g) Competition remains but at the bid stage rather than in the post market. So in fact oligopolies can perhaps be avoided.

So to me, not much more needs to be said on this point. Reverse auctions have been shown to be successful around the world and Simon Corbell showed the way in the ACT.

New low-cost variable supply will make prices go down, but as coal generators leave the system prices will spike.

There will be a big price for dispatchable power, so keeping up competition in this sector will be important.

No matter how much wind and solar go into South Australia, prices stay high. Why? Because gas sets the price. Everyone knows this in the electricity industry and the Coalition know it as well as the ALP.

So there will be no excuse for not having budgeted and managed it properly. No excuse.

So in South Australia what is needed is more competition in supply at the times when wind isn’t blowing. That could be pumped hydro or it could be batteries or it could be new transmission. Wait… there’s a policy for that!

Batteries can’t do the job, can they?

It’s on batteries that the skepticism of heavy industry falls most heavily. The likes of Bluescope, Orora or perhaps NextDC might, reluctantly, conclude that wind, or ideologically preferable solar, have a role to play, but batteries?… great for phones but they won’t run a steel plant.

Of course this kind of skepticism is absolutely typical of the establishment reaction to disruptive change in general.

And yet we now see that in California, utility-scale batteries are in the same economic ball park as older gas plants. And that’s utility-scale battery plants.

It’s long been a subject of some debate about the scale advantages of batteries. What we can say with confidence is that

(i) there is no scale advantage in cells or inverters other than the buying power of large quantities.

(ii) There may well be significant scale advantage in balance of system costs.

However, and this remains to be proved, once household battery manufacturing really gets going the balance of system cost will come down dramatically. It will come down in the factory, not in the field.

Even in the field, though, installation will become a lot more plug-and-play and competition will become more widespread. You would only trust a few installers to put your household battery in today, but in three years it will be different.

Adding up the battery impact

I assume you can’t get a state and federal subsidy, although I have no doubt many will try. Assuming the schemes are additive, my quick running total is about $350 million.

Obviously, NSW is nowhere. Minister Don Harwin was in his element talking about the Art Gallery extension, but asking him to spend some real money on household batteries is a bit much.

Queensland doesn’t have any money, and it shows. Also it is the only mainland state to still mostly own its generators and I will never trust the decision making of an umpire that owns some of the teams.

The grants are generally capped at households with an income of $180 k or less. These people are definitionally less able to afford top of the market solutions.

There are lots of families for whom even $5000 is money that can be spent on things other than batteries. But assuming a full take up… Lets assume 8 KWh per household, then we get over 1 GWh of daily energy produced and a bit more consumed, but more importantly we get close to 500MW of peak demand reduction.

Figure 3 System wide impact of battery subsidy. Source ITK

How much does that matter? It will come down to a state-by-state view, but at first blush we have looked annual average NEM wide.

NEM-wide views need to be treated cautiously. The first point is that Queensland time is the reference. This is in sync with NSW and Victoria when there is no daylight savings, but runs an hour behind when there is.

It is perhaps going too far to suggest that when NSW and Victoria turn off daylight saving, Queensland should move its clocks back to stay an hour behind, but growing up close to the border I’ve heard it said.

Also the annual average makes no allowance for seasonality and as solar penetration grows the differences between Winter and Summer will become more pronounced.

Finally in this set of numbers, rooftop PV is excluded from demand, what’s shown is demand net of behind the meter supply.

Still the numbers as they stood for FY18 are shown below:

Figure 4 NEM, Daily average price and demand. Source: NEM Review

Subsidy could save over $100 m a year, conservative 4 year payback

500MW of battery can reduce peak demand by 2 per cent but, because the batteries are charged from the rooftop systems, this will not impact minimum demand much.

Fig 3 shows that prices rose from $80/MWh at say 2pm QLD time, up to about $125/MWh at 5pm – and were still over $110/MWh at 7pm.

We suspect that 500MW of demand reduction at that time could have quite an impact on average price dropping it say $10 MWh. That benefit would flow to all consumers, after it flowed through hedges. It would flow to industrial users although to a lesser extent as well as households.

If we said 2 hours x 25 GW x $10 x 365 days a year that’s $180 million of savings. Even if we ignored industrial users and said it was only retail and commercial that benefitted, it’s still $90 million a year, which gives a payback on the system subsidy of four years. That is, $365 of subsidy $90 million of system-wide benefit a year.

And that ignores all the other benefits in terms of system strength, less need for distribution augmentation, etc.

Another way to look at things – and one that has an eye to the impact of household batteries on system security – is a price duration curve.

We’ve done this for both demand and price using centiles.

Firstly demand, the overall chart shows about 89 per cent of the time demand was less than 25GW, but we need enough generation, and transmission and distribution capacity for up to 31GW of demand.

Figure 5 NEM demand. Source: NEM Review

Looking at the top 100 half hours of demand across the entire NEM in the past 12 months

Figure 6 Top 100 half hours of demand last 12 months. Source: NEM review

500 MW of battery would only avoid 13 of those intervals. I should add that for whatever reason NEM wide high demand didn’t always mean high prices.

When we turn to price, its more sensitive than demand at the extremes. Looking first at the centiles we see that 93 per cent of the time price in past 12 months NEM wide was < $150/MWh but thereafter rises steeply.

You have to think that 500MW of battery capacity could make a difference to that 7 per cent of prices. Space doesn’t permit here, but the higher prices mostly result from a supply breakdown rather than high demand, or at least there is some problem with supply typically when demand is already high.

So having that 500MW of extra battery supply which would be as reliable as any generator in the NEM would be a big help at those times.

Figure 7 Percentage of time NEM price < than x Source: NEM Review

Looking at the 100 highest price intervals (about the top 0.5%) for 12 months we see 15 half hours of price exceeding $1000/MWh and 43 where price was over $500/MWh. We suspect 500 MW of battery would have made a difference to each and every one of those prices.

Figure 8 NEM top 100 prices last 12 months. Source: NEM Review