After the rhetoric comes supply and demand

Print Friendly

The tumult and the shouting dies;

  The Captains and the Kings depart:” Kipling 1897

Australia’s National Electricity Market brings power to millions of people. Shuttershock


New generation needs to be built in the NEM

A 10 year old debate about the future of generation in the NEM continues to this day. A 9 per cent fall in consumption enabled the decisions to be deferred for several years but now consumption is once again rising and plant closures means that the market will ensure solutions. In this note we just look at the near-term supply backdrop. The economics of new generation are deferred to  a forthcoming piece. There is much being said on the topic.

Supply is tight, coal costs rising, not enough new renewables

  1. Rising population, growing GDP, not enough energy efficiency in new apartments and houses, global warming are all increasing electricity demand
  2. But supply capacity is falling, obviously with the closure of Hazelwood but also a bunch of other cutbacks that we have previously discussed.
  3. Thermal fuel costs have risen pushing up the cost of thermal electricity. That’s very obvious in the case of gas, but its also the case for NSW thermal coal.
  4. We think some NSW generators particularly Eraring, Vales Point and possibly even Mt Piper are relatively tight for coal supply and if they expand output in response to the closure of Hazelwood may be forced to the expensive spot market.
  5. Transmission flows are going to change, although the implications are not yet that clear.
  6. About 1.6 GW of new wind and about 0.7 GW of utility PV is under construction or fully committed. Most of this will come on line by June 2018. Over two years there might also be another 1 GW of rooftop PV. However the energy from all of this won’t even replace what is lost from Hazelwood.
  7. Futures prices can be expected to embody all this information and have risen considerably but our conclusion today is that its more likely that prices will rise further, and volatility will remain high.
Figure 1 New renewable output compared to Hazelwood. Source ITK

Figure 1: New renewable output compared to Hazelwood. Source ITK

Where did our network supplied electricity come from last year?

Fig 1 shows the electricity market share by power plant for plants covering 90% of NEM output over the past 12 months. We’ve  aggregated water and wind plants by region.

Figure 2 NEM market share by generator. Source: NEM Review

Figure 2: NEM market share by generator. Source: NEM Review

These 22 “generators” account for 90% of output last year. Hazelwood is the 5th largest with a 5.3% share. The largest 8 are all coal and account for 49%  market share.

Capacity utilization in the top 10

Once Hazelwood goes the question is what will replace it? Not in the long term but over the next 2-3 years. We will get on to new generation in a moment but to start with the obvious place to turn is the existing large (as measured by market share)  power stations. The chart below is colour coded by State. Clearly Victorian generators operate at > cap use than QLD > than NSW. This reflects variable cost. NSW coal is more expensive than QLD which is more expensive than Victoria (without a carbon tax). NSW also benefits from strong transmission connections.

Figure 3 Capacity utilization: Source NEM Review, ITK

Figure 3: Capacity utilization: Source NEM Review, ITK

We calculated theoretical capacity as nominal capacity (not distinguishing between Summer and Winter) and multiplying by hours in year. We used energy sent out in the past 12 months as the numerator. The numbers are only approximate and only meant as an analyst’s guide.

Thermal power stations don’t generally run at 100% of theoretical  capacity utilization for extended periods.  ITK’s guess is that some power stations such as Bayswater and Eraring might manage 85% capacity utilization over a two year framework and others like Vales Point and Liddell might struggle to achieve much over 75%.

NSW thermal coal is export parity priced and relatively tight

All of this assumes adequate coal availability. NSW thermal coal availability has been limited by some older mines petering out, and a failure to develop new mines. Specifically the Cobbora coal scheme was abandoned in 2013. ORG and Vales Point (then Delta Coast) received compensation for the abandonment but the upshot is they have to scramble to find thermal coal. The current thermal coal price is around $125/ t up from $75 t a year ago. At around  .47 t/MWh this can add $10 MWh to the cost of electricity. A coal plant in NSW that paid the export price of coal would have a fuel cost of $47 MWh.

Figure 4 Export coal price NSW. Source: Factset

Figure 4: Export coal price NSW. Source: Factset

We would argue that the NSW generation plants most capable of lifting output, Eraring and perhaps Vales Point face sharply higher marginal fuel costs. That is they will have some contracted coal but depending on size of stock piles will have to pay spot prices to lift output on a sustained basis. This might take 12 months to show up in the case of Eraring but we guess more quickly for Vales Point. Of the other NSW generators Liddell has had long term technical issues. Mt Piper was going to be coal constrained but the closure of Wallerawang has probably freed up some supply. Even AGL will be conscious that lifting Liddell output overly will impact the longer-term coal position of Bayswater.

So the conclusion here is that NSW generators have some capacity to lift output but that it will be far from costless.

QLD generators particularly Stanwell and Tarong can also lift output. However the QLD market is also very strong and prices in NSW have to be higher than QLD to get QLD electricity to NSW. In Summer in the afternoon in Qld that’s going to be difficult.


Our understanding is that historically overnight power flows into NSW from lower variable cost Victorian and Queensland power stations. Queensland has historically supplied a significant portion of power to NSW. However over the past couple of years that flow has started to decline and even reverse during the Sunshine State’s early evening peak. We have several times drawn attention to the emerging  “duck curve” in QLD and our expectation that this would push up late afternoon and early evening prices, particularly in Summer. Fig 3 shows the March quarterly average flow by time of day.

Figure 5 NSW QLD flow, quarterly average by time of day. Source: NEM Review

Figure 5: NSW QLD flow, quarterly average by time of day. Source: NEM Review

By contrast flow from Victoria to NSW runs at a consistent 400 MW average during the day and then doubles in the overnight market.

Most of that flow is going to go away once Hazelwood closes.

Figure 6 Vic NSW March quarter average flow. Source NEM Review

Figure 6: Vic NSW March quarter average flow. Source: NEM Review

Renewable energy supply

Renewable energy is getting very cheap. The Silverton wind farm’s $65MWh (increasing with inflation) is a beacon as well as being a significant boon to NSW. Compare that to ERM which is paying $85MWh plus the cost of black electricity, so probably around double AGL’s cost.

However the total quantity of renewable supply is still not enough. Basically you need about 2.5MW of renewables for every 1MW of brown coal in terms of electricity produced.

So to replace Hazelwood’s nominal 1600 MW needs around 4000MW just to maintain the same status quo.  Never mind the other power stations that have closed.

We don’t ignore rooftop PV, it’s incredibly important, but in our view 700MW of new rooftop PV per year basically doesn’t even offset population growth.

Wind projects committed, not yet operational

We can count about 1.5 Gw of new wind projects that have been announced and we expect a couple more to be announced. Frankly the prices of $73 MWh fixed for 20 years at Hornsdale Stage 3 and the $65 MWh real for Silverton NSW are still pretty attractive prices. The good news for energy security is that we are seeing these wind farms being developed in NSW, QLD and Victoria reducing the “synchronous nature” of over concentration in South Australia.  Still the MW being developed are insufficient. Some of these projects are developed using the Victorian purchase of LGCs, some for the ACT Govt 100% renewable program and some under the Federal scheme.

Figure 7 Wind farms under construction in the NEM. Source ITK, AEMO

Figure 7: Wind farms under construction in the NEM. Source ITK, AEMO

PV projects

Its not easy keeping track of the list of PV projects, however on our numbers we can only get to 678 MW of utility projects definitely proceeding. Even that includes a couple of projects only confirmed in the last couple of days.

Figure 8 PV farms under construction in NEM. Source ITK

Figure 8: PV farms under construction in NEM. Source: ITK

There remain 300 MW of  PV farms that have ARENA funding but nevertheless have not yet gone ahead. The largest of these is ORG’s Darling Downs project.

As time emerges the benefits to AGL of the Powering Australia Fund and in particular  its  strong financial partner in the form of QIC  are increasingly apparent. AGL is getting low cost energy and getting projects built. Talk remains long and action short from  a number of the major retailers.   ORG which is sitting on Stockyard Hill for many years and now Darling Downs PV is not getting very far. For CLP progress is also very slow. And we now know the ERM story.

Conclusion – Not enough renewable energy is being built to even replace Hazelwood.

Confirmed renewable projects won’t produce enough energy to replace even Hazelwood, never mind the reduction in gas fired output and the closure of Northern power.

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.  

RenewEconomy Free Daily Newsletter

Share this:

  • GlennM

    good analysis…and of course when prices spike or there are shortages RE will be to blame not the closing of FF plants.

  • trackdaze

    How do you arrive at the 1gw for rooftop over two years. Last two years saw about 1.4gw?

    & the 15% capacity? Is this a net figure that excludes consumption from the household removed from the nem?

    • Kevfromspace

      The Solar Bonus Scheme is up. Unless we see commercial installations pick up dramatically (to offset rising electricity prices), there’s no reason to think that 2017 will be as good of a year for small scale solar than 2016.

      • Jonathan Prendergast

        Entry to the solar bonus scheme finished a few years ago, so that will not effect the difference between 2016 and 2017.

        I think residential will continue, or maybe even grow a bit. And as you suggest, commercial could scale up even more in reaction to rising elec prices.

        • Kevfromspace

          I didn’t realise that – sorry for misleading and thank you for clarifying.

          On that note, perhaps 2017 will see solar owners with reduced Feed In Tariffs resort to adding new (more efficient) panels and possibly storage to their existing systems in order to keep energy bills down.

          • Jonathan Prendergast

            Agree. Some retrofit additions will be common and add to 2017 capacity.

            Mainstream media were quite confused with ending of solar bonus scheme FITs and reduction in STCs 31 December 2016, and gave out lots of misleading info. Finn from Solarquotes wrote an angry article about it all but I can’t find it now.

      • trackdaze

        Its scaled down roughly in line with cost reductions in hardware costs. No?

    • David leitch

      I have just taken a slightly conservative view. Household saturation will kick in at some point. Perhaps commercial will offset. It makes next to know difference to the overall conclusion. 15% DC capacity factor is realistic for household solar. When IPART looked at it a couple of years ago in NSW it was under 15% I guess it will be better for commercial if that becomes a bigger share.

      • neroden

        I’m not sure it’s wise to take a “conservative” (low) estimate of solar deployment. I guess it depends what you want your projections to be used for.

        A low estimate gives too high an estimate of future grid electricity prices. This is great if you’re a producer trying to lock in a fixed contract prices, terrible if you’re a consumer trying to lock in a fixed contract prices, bad if you’re a producer signing a variable contract, good if you’re a consumer signing a variable contract, and so on….

  • Jonathan Prendergast

    And of course, Hazelwood can be replaced by Energy Efficiency and Conservation, as a reaction to higher prices. This occured when prices went up leading to the 2009 decline in demand (obviously only part of the reason of reduced demand).

  • Jonathan Prendergast
    • David leitch

      Thanks Jonathan. There is always something slipping through the cracks.

  • David Osmond

    Wonderful analysis David. One point though – I suspect your assumption of 35% capacity factor for wind is likely to be on the low side going forward. If you look at wind farms built in the last few years (say post 2014), they are getting capacity factors in the range 34-41% (in 2016), and the wind farms currently under construction are likely to do even better, thanks to the ever increasing blade lengths and hub heights. I’d be surprised if the 1.6GW of new wind didn’t achieve an average CF>40%. If they do that, then we’re starting to get close to new renewables making up for Hazelwood.

  • alexander austin

    Great article David. On the demand side, is there potential for further significant loss of commercial/industrial demand? For example, will Portland continue beyond the current 4 year bail out? Queensland smelters? etc

  • Robert Comerford

    Well a good start is to put up some solar thermal plants where Hazelwood stood. There is obviously water there and transmission lines to the grid so why waste it?

    • Ren Stimpy

      I think the water in solar thermal plants is recycled (unlike coal or nuclear plants). Better to put it in the best geographical spot for success rather than the best political spot.

    • Malcolm M

      Solar thermal needs a particular kind of solar energy that can be reflected on mirrors into a concentrator. In southern Victoria in July and August, only 25% of our solar radiation is direct (and suitable for solar thermal), the remainder diffuse (but suitalbe for solar PV). Even in summer, these percentages aren’t much better. Solar thermal needs cloudless skies, and Port Augusta is marginal, whereas Olympic Dam and Leigh Creek are better.

  • Malcolm M

    And in 5 years time Liddell is scheduled to close, which according to your chart above produced nearly as much energy as Hazelwood. So not only do we need to catch up with new capacity to replace Hazelwood, we also need enough to replace Liddell. All while gas is pricing itself out of anything but peaking.

  • neroden

    At these prices, it’s clear that new PV and wind projects will be highly profitable, no subsidies needed.

    So the question is: what’s obstructing the construction? It can’t be finance. Land acquisition? Utility connection? Construction firms busy? Environmental permitting? What?

  • Malcolm M

    There is a report in the Portland Observer 6 February that wind tower manufacturer Keppell Prince was working on Kiata followed by Gellibrand, “…and then some larger projects such as Dundonnell after that”. They may know something about the stage of these post-Gellibrand projects that hasn’t yet been released into the public arena.