Baseload and firming prices: Is there really a market signal for storage?

Print Friendly, PDF & Email

We are writing a couple of notes looking at the generation price outlook for both baseload and firming capacity under present policies, never mind endless Federal political machinations.

We think that the new renewables being built, recommissioning of some gas plant, and Queensland Government policy will lower baseload prices in Qld, NSW and particularly Victoria.

In Victoria, open cycle gas is used today for baseload purposes. The new supply, particularly Stockyard Hill, will see gas going back to a peaking role and not much more supply is needed to get rid of it (on an average basis) altogether.

That won’t do much for carbon emissions but it will be massively helpful to price. In South Australia, though, gas will still set the price much of the time.

The second of these notes will hopefully look more closely at the need and potential value of investing in more firming capacity in today’s world.

As only 9% of electricity in the NEM today comes from wind & solar PV, on the face of it it’s going to be hard to justify building too many pumped hydro or solar thermal plants. They are an investment in a future which we are still only walking towards.

Building for the future, not today

With the economics of wind and PV well established and backed up by 6GW, or $12 billion of recent investment, it’s clear there is going to be a lot more renewable energy.

The financial market questions that arise from this are not about energy security or decarbonisation, important though those topics are. No, they are the more prosaic questions: 1. What does it mean for base load prices? And 2. Is there any reward for building firming capacity?

Good answers to these questions will help financiers to see the likely reward for building more new capacity. However, it will also help to answer questions around existing assets. For instance, what is the value of Snowy? What is the value of Loy Yang B? Both of these assets or stakes in them are likely for sale.

By firming capacity we mean “pumped hydro”, “solar thermal” and “lithium storage” or even new gas peakers. Alternatively a new ~$1 bn transmission line could be built to Tasmania to enable the NEM to better access Tasmania’s extensive already built existing hydro resource

In our view, it’s not clear that – with a couple of exceptions – there is a sufficient reward to justify much new investment. In our view there is very extensive firming capacity already existing in the NEM in the form of Snowy, Tas Hydro, existing gas generation and existing coal generation.

Some market evidence for this is in the price of “caps” that is the insurance product that guarantees the buyer that the buyer won’t pay more than $300 MWh for a specified number of MW over a given period.

However, in order to think about the value of new firming investment we start by looking at the outlook for baseload. As such this article will be broken into at least two parts.

New investment in process

Our latest sums show 6GW of new generation investment is underway or commissioned in 2017 onwards. These numbers are well in advance of most of the formal modelling reports, but thanks to Reneweconomy’s research we can be ahead of the pack.

Figure 1 The boom in Australian renewables investment. Source: ITKe

Figure 1: The boom in Australian renewables investment. Source: ITKe

Our under construction numbers still include Horncastle stages 1&2 even though these projects are now commissioned as we are looking at the change relative to 2016.

We anticipate mainland NEM wind and PV penetration to reach in excess of 17% share of total electricity supply by the end of calendar 2018 although wind capacity has a way of not always doing what it’s supposed to.

Figure 2 Estimated wind + pv supply share of demand Cal 19 Source: ITK, NEM Review

Figure 2: Estimated wind + pv supply share of demand Cal 19 Source: ITK, NEM Review

We say in excess because this clearly is not the end of the renewables investment. The Victorian and Queensland State reverse auctions have yet to take place. They should be good for close to 0.8GW of capacity assuming they come from projects yet to reach FID (financial close).

In addition, we think there is a significant number of NSW and Victorian projects that are to some extent held back by lack of easy access to transmission. In fact, we think new transmission investment is the key to seriously getting the renewables scene going in the NEM but that’s for another day.

More than 800 MW of gas will be back in operation

Under current rules we don’t expect any significant curtailment of renewables production and once built the variable cost is close to zero for wind and PV. In other words all that production is going to be dispatched.

In addition  to the new renewables we also have other factors driving down price.

  • Recommissioning of Swanbank E 385 MW of combined cycle gas in Qld;
  • AGL 250 MW fast build/fast start reciprocating engine project on the TIPS A site in South Australia although possibly not in operation until FY20
  • Recommisioning of 100% of Pelican Point 485 MW combined cycle in South Australia (this alone will make South Australia secure in my opinion)
  • Queensland Govt directing Stanwell Power operator of 3.3 GW of Queensland coal generation plus Swanbank E to put downwards pressure on prices.

Electricity futures, unlike oil or other storable commodities are indicative

The question in financial markets is always whether to see the futures market as an indication of prices in the future or as just an arbitrage.

Unlike oil or gas though, electricity can’t be stored (I know, I know …) and so the futures prices is actually an indication of where the markets are going to go as opposed to an arbitrage between the cost of storage and the present price.

We exclude South Australia as small and Queensland as a special case driven more by the Queensland Govt than market forces then we can look at the average of Victoria and NSW (62% of NEM demand) as follows:

Figure 3 Futures curves. Source: ASX

Figure 3 Futures curves. Source: ASX

FY18 futures are up $66 MWh from a $48 MWh base a year ago, but FY19 futures are up a smaller $41 MWh from a $49 MWh base and FY20 futures which are almost liquid enough to trade are at $80 MWh and even less at  $34 MWh.

However, a year ago no one was forecasting that FY18 futures were going up $66 MWh so the futures prices are not necessarily a guide. We argue that $20 MWh of the increase was driven by the decision to keep the Portland smelter running when market expectations were for closure. Replacement of Hazelwood with largely gas fired power has contributed to the rest.

State level aggregate demand supply balance adjustements

Across the States we expect some significant changes in the demand supply balance.

Figure 4 Supply demand balances Source: ITKe, NEM Review

Figure 4 Supply demand balances Source: ITKe, NEM Review

We allowed relatively low capacity utilization at Swanbank and Pelican Point. Gas prices are high and the plants will only be run grudgingly.

Also in South Australia we have not adjusted for pre Dec 2016 production from Hornsdale or the Northern production in Cal 16. We assume new renewable supply will replace gas in South Australia and Victoria lead to no net change in export/imports  It also looks as if there will be a significant increase in net exports to NSW from QLD.

Overall we infer from the date there will be a 2% net increase in supply of 4 TWh in 2019 relative to 2016.  For simplicity we assume the increase in distributed PV offsets any increase in demand and just look at utility. However when we come to price we will see that simplification won’t do.

We also need to consider:

  • Impact of higher gas prices;
  • Changes to market structure
  • Impact of intermittency

Its only been 75 days since Hazelwood shut down. A comparison of generation market share by fuel for the combined NSW, QLD, VIC, SA market compared to the same 75 days last year shows that gas and hydro have gained market share at the expense of coal. Wind output is down significantly in both South Australia and Victoria which we assume is due to wind conditions as there is 200 MW additional capacity.

Figure 6 NSW, QLD, Vic, SA, combined generation share. April 1 -June 15. Source: NEM Review

Figure 6: NSW, QLD, Vic, SA, combined generation share. April 1 -June 15. Source: NEM Review

If we focus in on Victorian supply by time of day Fig 7 below shows the remaining coal is totally flat supply and how gas is now going virtually 24 hours a day with hydro supplying the morning and afternoon peaks. These are average half hourly numbers and obviously vary from day to day.

Open cycle gas is now performing base load duties in Victoria.. On average  wind supply in Victoria is relatively constant by time of day.

Figure 7 Victorian generation April 1 - June 15 half hourly average MW. Source: NEM Review

Figure 7: Victorian generation April 1 – June 15 half hourly average MW. Source: NEM Review

Half hourly prices show duck curve but not much demand for extensive time shifting

Figure 8 Average half hourly pool prices April 1 to June 15 2017. Source:NEM Review

Figure 8: Average half hourly pool prices April 1 to June 15 2017. Source: NEM Review

  • The impact of PV on middle of the day prices is now very visible. Much more so than a year or two ago. Its shoulder season so this impact will be more pronounced in Summer.
  • The $160 MWh for South Australia at midnight is due to controlled load hotwater. Surely these days it would be better shifting that load to the middle of the day?
  • The average duration for peak pricing is relatively short, in Qld maybe an hour in the morning and maybe 2 hours in the evening. In Victoria maybe 2 hours in the morning and an hour at night. These are average numbers but if you build a pumped hydro plant you want to run it every day as its purpose is to shift load.
  • We think there is plenty of opportunity for Stanwell to lower prices further in Qld. Recall that QLD demand is actually flat or even down a bit compared to a year ago because of the reduction in Gladstone aluminium smelter output.

Post the renewables build

ITK doesn’t have access to the fancy modelling tools of Jacobs or Frontier economics. We don’t have a “Whirlygig” or “Prophet” or “Strategist” program so we have to take a much simpler approach which is nothing more than some rough guesses.

Our view is:

  • Expanded PV production in Qld will result in less gas use in the middle of the day and more middle of day exports to NSW
  • Those extra exports together with higher wind production in NSW will mean that NSW coal generators will cut back their production. Prices in NSW and QLD will ease a bit as a result.
  • In South Australia gas will continue to set the price. Fig 8 below shows our estimate of average output by time of day post the build out of a further 100 MW of Hornsdale, Lincoln Gap, Tailem Bend  & Bungala and a recovery in wind ouput to more normal levels.
Figure 9 Average MW generation in South Australia 2019. Source: NEM Review for FY17 and ITK adjustments

Figure 9: Average MW generation in South Australia 2019. Source: NEM Review for FY17 and ITK adjustments

On average gas is still used every half hour so it continues to set the price. However, midday prices will fall significantly. But if you want to assume that some of the gas is Pelican Point and it runs in an efficient, combined cycle shoulder mode, the average cost of the gas fired generation could reduce. As stated this baseload analysis is based on averages and makes no allowance whatsoever for wind volatility.

  • Finally in Victoria on the current numbers and before the reverse auctions are implemented we see that Stockyard hill and the PV projects already announced will largely eliminate expensive gas. By the time reverse auctions are done gas will be gone and renewables will be eating further into carbon intensive brown coal
Figure 10 Estimated suply Vic post stockyard hill, Mildura, Ganwarra. Source: NEM Review and ITKe

Figure 10: Estimated suply Vic post stockyard hill, Mildura, Ganwarra. Source: NEM Review and ITKe

The next article will deal to the implications for investing in firming plant and what that might mean for assets like Loy Yang B and Snowy.

David Leitch is principal of ITK. He was formerly a Utility Analyst for leading investment banks over the past 30 years. The views expressed are his own. Please note our new section, Energy Markets, which will include analysis from Leitch on the energy markets and broader energy issues. And also note our live generation widget, and the APVI solar contribution. 


  • solarguy

    David, So overall, we won’t see an easing in electricity prices until a critical mass of renewables and perhaps some firming is reached. Is this correct?

    • David leitch

      Pool and futures prices are already easing and will likely ease more over the next 18 months. This may result in lower consumer prices relative to those just announced but only by a few percent.

      • Nemo

        To take the floor lower on prices there needs to be cheaper gas or gas needs to get displaced as the marginal price setter and that is going to take some time. If OCGT fuel costs is around $70-90/MWh will be hard to take pricing materially lower than that in the short term.

        That being said period of acute shortages should be largely over by 2019-2020 in time for Liddell shutdown.

  • alexander austin

    David. Great article. I liked the chart on SA. My only question was whether you have perhaps left the Lyon Solar farm out of the projections for SA. This was in the list from CEC as a committed project

  • George Darroch

    Great explanation of a complex topic.

    28% and 15% sound about right for utility and rooftop solar, but what’s the basis for choosing these numbers?

    • David leitch

      Analysis of existing data and other research. Moree provides single axis tracking output data. Rooftop pv capacity factors are well known and maybe 14% is a better number.

  • George Darroch

    We could start to see large installs on commercial rooftops creating an even larger duck curve over the next several years. Interesting times are ahead.

  • Rod

    “The $160 MWh for South Australia at midnight is due to controlled load
    hotwater. Surely these days it would be better shifting that load to the
    middle of the day?”

    Too many dumb meters here in SA to do much about that. Needs to be manually changed at the meter.

    There is a scheme replacing electric HWS with heat pumps but not sure what the uptake has been.

    SAPN also have a voluntary smart meter replacement program happening.

    • BushAxe

      There was a program a couple of years ago by SAPN to retime about 100MW of the HWS, but that graph shows much more can be done even if they move it to the 3-5am period just before the morning peak.

      • Rod

        Ideally they would all be under the control of local network controllers
        to either soak up excess RE or smooth out the demand curve.
        Until we
        have enough smart meters in place to do this, the very least they could
        be doing is staggering the timers on new meter installations.
        This time of year even those with solar HWS would need a little boost.

    • Rod


  • Mark Roest

    You could have at least mentioned that battery storage is the best, and is or will soon be the cheapest form of firming (headed for $100/kWh capacity, and actually less than $100/kW power). And while utilities probably don’t want it because they profit from volatility of net demand, the wind and solar farm owners, and rooftop solar owners, will truly want it when the economics are obvious to them in that coming price regime. They can use both batteries and energy management systems to keep the utilities from being able to economically justify use of gas, let alone coal. That is based on continued rapid, compounding growth of renewable energy, battery storage, and energy efficiency, in addition to demand response (or should we call it demand anticipation?). If the energy management system providers seize the opportunity to look at the entire system and deliberately pit their clients against all fossil fuels, they can both drive fossils out of the market, and take profits from the utilities and give them to their clients, creating a durable symbiotic relationship about which songs will surely be written.

  • bedlambay

    Blatant price gouging by private operators. Qld still has government operators. Abbott and his fellow deniers has made it much worse with their renewables investment uncertainty.

  • Peter F

    Great article: Two comments.
    1. Batteries on the grid with a primary purpose of firming renewables are not yet necessary, we have more than enough gas, hydro/pumped hydro until we double existing solar and wind capacity and if the renewables are widely dispersed and have high capacity factors (i.e. Low wind turbines and tracking solar) you may be able to triple the current renewable capacity on the NEM. Southern Power pool in the US is regularly reaching 50% power supply from wind without storage, The CEO of 50Hz (the north German equivalent of the NEM says he can get to 70% share without storage

    However there are many other uses for batteries and other storage in avoiding distribution upgrades, providing end of grid power stability, behind the meter solar storage, peak price and demand charge mitigation and most of all “stuff the power company”. Even at gas generators, the ability to provide zero fuel cost spinning reserves is close to economical.

    Then of course there is demand response which has the same value and by some estimates that alone will account for more than 3GW

    By my guess there will be more than 2-3GW/4-6GWhr of new non hydro firming capacity in the next 3-4 years. That, with existing hydro and gas will take us to the point where we could supply a typical windy spring Sunday without coal. It won’t happen because we will run coal instead of gas but it will be possible.

    2. I beg to differ on the question of major transmission upgrades. The balance between transmission and generation investment has swung sharply in favour of excess renewables and local storage with a thin grid moving mainly off-peak power around. Local storage (power to heat, batteries etc.) is not only cheaper than transmission investment but it reduces peak loads and therefore losses in the existing network.
    As an example, new low wind turbines on 110m towers in the Latrobe valley can generate as much power per year as the 4 year old turbines with shorter blades on shorter towers do at McArthur wind farm, in spite of lower wind resource. Therefore now you don’t have to chase the best possible wind speed. If the Latrobe Valley had the same density of wind turbines and solar panels as Germany has now, the valley would generate the same amount of energy as it does today and use the existing transmission network without any upgrades.

  • Michel Syna Rahme

    Just a thought, and excuse me for I have no time to check the figures to see if what I’m saying is correct, but in terms of the Snowy Hydro 2.0 and expansion of hydro storage and pumped hydro in Tasmania – which is a great addition and an essential part of the solution – when designing the inevitable future storage network essential for future renewables integration, in some respects the more spread the better.

    i.e future design of the storage network must take into consideration the probabilities of worst case scenarios, and we all know weather patterns today will be different and vary into the future and extremes potentially amplified. Therefore, if not already factored in, the future storage network, and in particular the expansion of Snowy Hydro, must have contingencies for a repeat of those rare periods/events, such as the Tasmanian drought surrounding the years of 2007, where all of that added capacity may become limited for a significant period of time.

  • dyemanoz

    Very informative analysis. Thanks!

    Nitpick: There are a couple of references to Horncastle in the article – I assume you mean Hornsdale?

    • David leitch

      I do and thanks for the correction.