Just because something is cheap to buy, does not mean it will be cheap to own, and just because the sticker price of some renewables is coming down does not mean we can keep adding them without costs. A reliable electricity supply is more than just electrons, and most of our reliability is provided by the dirtiest generators that we need to retire to deal with climate change. We can clean up our electricity supply and maintain reliability  provided we use all the technology available to us. Understanding more about ancillary services and the great concentration of greenhouse emissions in a relatively small number of generators in Australia illustrates a likely sensible floor for a nuclear sector in Australia. There is more than enough space for that, as well as a strongly growing renewables sector. We desperately need a sensible centre in our energy discussion that respects the way our system actually operates.

Ancillary adjective : providing necessary support to the primary activities or operation of an organization, system, etc.

Suddenly it seems like everyone is talking about ancillary services.

Ok, that’s an exaggeration. I am sure there are more people talking about the colour of the diving pool in Rio. In my little world of energy though, ancillary services have seemingly gone from nowhere to very interesting.

What the heck are ancillary services? You can read this piece to get a little idea, you can read this from AEMO to get much more of an idea or you can go with this…it’s all the important ingredients in running a reliable electricity system that you don’t know about. It is only one part of the game to provide the right number of megawatt hours in a given hour. The system needs to run at the right frequency, the voltage needs to stay within limits and the system needs to be able to get back up and running again in case of serious failure. That’s all called ancillary services.

So why are we suddenly talking about them?  Like most necessary support services, we acknowledge their necessity when they start to go away. In the Australian National Electricity Market, all of the ancillary services for the control of frequency are provided by 116 connected generating units that bid to provide that service; a mixture of coal, gas and hydro electricity units. No solar or wind generators are bidders to provide this service because, at this time, they can’t: they are non-synchronous, non-dispatchable generators.

South Australia has taken on a very large share of wind generation which has contributed to the exit from the market of several dispatchable, synchronous generators. The availability of ancillary services has declined, and our dependence on the rest of the NEM to provide those services has gone up. The latest 2016 Electricity Statement of Opportunities for the National Electricity Market (Australian Energy Market Operator 2016) has led with this issue:

The 2016 NEM ESOO highlights that as intermittent generation (such as wind and rooftop photovoltaic (PV) generation) continues to increase, and thermal synchronous generation (such as coal and gas-fired generation) withdraws:

  • Total installed generation capacity alone becomes a less reliable indicator of supply adequacy.

  • Availability of plant to supply energy when needed, and capability to provide ancillary services, are both key factors to consider when assessing opportunities related to secure operations and supply adequacy in the NEM. (AEMO 2016)

The future risk of load shedding is projected to be greatest between 2.00 pm and 8.00 pm, if high demand coincides with low wind and rooftop PV generation, unplanned generation outages, and/or low levels of imports from neighbouring regions (AEMO 2016)

AEMO goes on to highlight that there are credible security risks in future and this situation will exacerbate should Australian fulfil its obligations to cut greenhouse gas emissions, which will require the withdrawal of a lot more synchronous generation in the form of coal-fired power stations.

Playford Power Station
The Playford and Northern coal-fired power stations have been retired from the energy market in South Australia. Great for greenhouse gas emissions, not so great for system reliability, liquidity of the electricity market and provision of ancillary services

At the moment we are looking at low-credibility scenarios… but at the moment, we still have a very low penetration of wind and solar in the National Electricity Market overall. Were Victoria, NSW and Queensland to have gone with the penetration of wind that South Australia has achieved (some 40% of total electricity supply)…the situation would be different. This is an early, important warning: just because a generation source is getting cheaper, it doesn’t mean adding more of it is low cost. Just because a generator produces electricity, doesn’t mean the electricity it produces has the same value of what it will displace.

On that note let’s leave South Australia for another high-penetration renewable jurisdiction: Germany. In case you missed it, Germany is telling the world it will run on renewables. Lots of people, (me included) think it will run on fossil and renewables and the actual priority is closing nuclear. But whatever, the point is they have a plan for lots and lots of renewables. So how is the ancillary services situation shaping up?

The German Energy Agency gives us some idea of their thinking with their 2014 Ancillary service study (Deutsche Energie-Agentur GmbH (dena) – German Energy Agency 2014). The study scenario for the year 2033 is, itself, revealing. I have reproduced it below:

Installed generation capacity in Germany in the study year (total: 259 GW)
Lignite Hard coal Natural gas Pumped storage Others Onshore wind Offshore wind Photovoltaics Hydroelectric power Biomas and others
12 GW 20 GW 41 GW 11 GW 3 GW 66 GW 25 GW 65 GW 5 GW 11 GW

The first thing to note is that while nuclear has disappeared there remains 32 GW of coal and 41 GW of gas. That’s what I regard as seriously warped priorities.

The second thing to note is that this scenario retains 28% of installed capacity as synchronous, dispatchable fossil fuels, as part of a grid that is well-connected to surrounding markets. This is hardly a 100% renewable scenario, and Germany is anything but an island in electricity terms. Even in those conditions, the challenge is not lost on them:

Conventional power plants, which today still largely provide the ancillary service products for stable grid operation, will be on-grid for fewer and fewer hours in future. The question as to how the scope and type of these ancillary service products must change to provide the ancillary services frequency control, voltage control, system restoration and system control.

So, what do they find? In summary, they think it is do-able, however it’s amply uncertain in all areas and the planning needs to start now. It will also come at a cost:

The costs for maintaining and providing ancillary service products must be economically bearable both for the system operators and the operators of decentralised energy units and flexible loads. …Further provision of ancillary service products and associated expenses must be made economically viable with a suitable compensation system. Investments and ongoing operating expenditures required on the part of the system operators for secure and stable grid operation in a power supply system with an increasing proportion of renewable energy sources (e.g. grid monitoring and development of system control tools) must be incorporated suitably in the regulatory framework.

Bear in mind, this is just Germany…in a scenario that maintains lots of fossil fuels… connected to several of their neighbours who can provide these services to the grid and currently have no-such plans. This is not 100% renewable Germany and certainly not 100% renewable Europe.

So let’s come back to Australia and the concerns being raised by AEMO. If we act, as we must, to close our coal fired power sector so as to meet our responsibilities to deal with climate change, we will exacerbate, drastically, a problem that is already appearing on the horizon for South Australia as we axe several of the major providers of ancillary services.

AEMO has a point. The chart below illustrates the make-up of Australia’s electricity sector in terms of electricity generation and greenhouse-gas emissions. Note the left y-axis is a log scale (Note also this data is all of Australia, not just the NEM).

Data 1
Chart by Heard & Bradshaw (2016, first appearing here). Data from AEMO designated generation facilities report 2014-2015

This tells us  that a really large percentage of both electricity generation and greenhouse gas emissions in Australia’s electricity sector comes from a remarkably small number of locations. To be precise, 50% of the greenhouse gas emissions from the electricity sector in Australia (corresponding to 37% of the electricity generated) comes from just seven locations: Loy Yang Power station and mine (VIC); Eraring Power Station (NSW); Bayswater Power Station (NSW); Yallourn Power Station (VIC); Stanwell Power Station (QLD); Loy Yang B Power Station (VIC).

Those  locations have 34 generating units totalling 15,000 MWe, producing 84 million megawatt hours, giving an average capacity factor of 64%. They are also all registered bidders for the provision of ancillary services.

So…riddle me this. If you were concerned about climate change and wanted to slash greenhouse gas emissions from Australia’s electricity sector with the greatest certainty and the least messing about… what would you be looking for?

I would be looking for, at the minimum, 10-15 GW of synchronous, dispatchable supply, capable of providing ancillary services, with high capacity factor, that could be spread across perhaps 7-10 key generating nodes, that didn’t emit greenhouse gas. In other words, I would be looking for a medium-sized sector of electricity generation from fission technologies. Nuclear power.

The remaining 63% of the electricity supply… who knows. Maybe the rational use of nuclear in Australia is higher than 10-15 GW. After all, 85% of our greenhouse gas emissions from electricity come from just 22 locations that provide 70% of our electricity. That’s a remarkably elegant match for the rule of thumb that 30% of supply that can be cost effective with variable renewables, supported by papers such as Ueckerdt, Hirth et al. (2013) and Nikolakakis and Fthenakis (2011). On the other hand we are going to get better at using variable renewables and some of the novel ancillary service solutions may come to fruition. That rule of thumb may have to change.

But to maintain that Australia, or anyone, is going to get the fossil fuels out of the electricity supply without fission technologies is, quite evidently, highly uncertain at best and at worst? It’s hopelessly ideologically rigid or perhaps just incredibly rent-seeking and greedy on the part of the renewables industry. We need much smarter policies to reflect what we actually need in the market place.

No one gave a damn about ancillary services until they started to go away. Now that their absence and future availability is being noticed, hopefully we will better-appreciate just what we  need to climb the mountain of decarbonisation. It is clear to me is we are not getting there on wind and solar alone.

Like what you see here? Please subscribe to the blog, Like Decarbonise SA on Facebook and follow @BenThinkClimate on Twitter.


Australian Energy Market Operator (2016). 2016 Electricity Statement of Opportunities for the National Electricity Market. Melbourne, Victoria.

Deutsche Energie-Agentur GmbH (dena) – German Energy Agency (2014). Summary of key results of the study “Security and reliability of a power supply with a high percentage of renewable energy”. Berlin, Germany.

Nikolakakis, T. and V. Fthenakis (2011). “The optimum mix of electricity from wind- and solar-sources in conventional power systems: Evaluating the case for New York State.” Energy Policy 39(11): 6972-6980.

Ueckerdt, F., L. Hirth, G. Luderer and O. Edenhofer (2013). “System LCOE: What are the costs of variable renewables?” Energy 63: 61-75.


  1. The main problem with nuclear as it stands in the western world today from what I see..


    Hinkley Point C $24 billion USD

    To get the equivalent of 3200MW for 1 year from wind turbines. Assuming 2 million/ 1MW and 30% average usage only costs $21 billion.

    At $500/Kwh storage you can get about 6GWh of batteries with the other 3 billion.

    Now this is assuming 100% nuclear running time and not ramping up and down to match reneweables and load.

    For 90% (mostly base load with a bit of load following and 3 days of storage you need turbines at 1.5 Million per MW and storage at $100/KWh(of capacity and not delivered) . Those figures provide you with 3 days worth of time to use the power unlike the nuclear station where it is use it or lose it.


    It has taken a long time for Hinkley Point C to be built (November 2012 from wikipedia was the licence granting time)


    So as it stands yes Nuclear is cheaper if you use it as base load. Use it as load following and the economy is not as good.

    Nuclear also takes more time to build while with renewables you can build small and later buildings get better and cheaper tech.

    1. Didn’t read the article, Catprog?

      Your comment contains absolutely no mention of ancillary services. This is a good example of the style of analysis-free opinion that got South Australia and Germany into their current nonsensical impasse.

      Allow me to summarise it for you:
      Given that ancillary services such as frequency control are essential for synchronous grids, and that these are only available in Australia from lignite, black coal or the limited hydro power stations, future ancillary services must come from either
      1. Existing hydro plus coal; or
      2. Existing hydro plus nuclear.

      Hydro is and will remain less than 5% of the total.

      Batteries are not synchronous – they cannot and will not provide the ancillary services.

      The choice, for 65% of Australian electricity, is thus either coal or nuclear. Only one is carbon-free.

      1. @singletonengineer Appreciate the response you have made there, saved me the trouble.

        Batteries are slated to assist with voltage control at the distribution level, provided that can be done with some smarts and presuming penetration forces this and makes it necessary.

      2. I’ve noticed a trend this year in citing Hinkley Point C as the “nuclear disaster” story over the Fukushima Daiichi triple meltdown.

        Of course, nobody is suggesting building either EPRs or vintage BWRs in Australia. Australia’s electricity market isn’t intertwined with that of France. We can also be confident that ARPANSA’s future expanded regulatory scope won’t be modelled on pre-2012 NISA in Japan. Australian stakeholders have the opportunity to ensure this technology – with the fastest proven roll-out rate of all http://www.columbia.edu/~jeh1/mailings/2016/CaoJ.China-U.S._cooperation_to_advance_nuclear_power.Science.2016.pdf – is effectively regulated and integrated into decisive national climate action with maximum benefit to emissions reduction AND stability/quality of electricity supply.

  2. Ben,

    You make an excellent point that is rarely mentioned even, and that is that Germany’s intermittent penetration not only depends on the vast Euroepan grid, but also on their vast dispatchable electricity dependence which makes you wonder why they would press their neighbors so hard to switch to intermittent sources. If anything, they should be pressing the rest of Europe to have more nuclear.

    The other point that is often ignored, is that if you have to spend much more than you would otherwise on ancillary services to keep coal power plants alive, on top of massive expansion of transmission, massive subsidies and tax exemptions for coal https://www.odi.org/sites/odi.org.uk/files/odi-assets/publications-opinion-files/9954.pdf), then it seems that just accounting for the massive renewable subsidies is not enough to measure the cost of renewables.

    1. [Germany] should be pressing the rest of Europe to have more nuclear.

      That’s obvious to any power systems engineer. But it’s a political challenge – how can a German politician justify encouraging say Czech Republic to build nuclear?

  3. My understanding is that paid ancillary services for voltage and frequency support (FCAS) is on a short time scale like minutes perhaps hours. When there is a wind lull for days or a cloudy week that needs costly gas backup that’s lack of planning. My guess is we didn’t need ancillary services so much in the late 20th century when most power came from coal and hydro. We’ve created the problem by forcing a large part of our electricity to come from unreliable sources.

    Before building new interstate powerlines and subsidising batteries I think we should drop the renewable energy target. Make emissions the main driver of policy. That way wind and commercial solar no longer get preferential treatment but will have to compete on basic cost and contribution to low carbon. If all technologies including nuclear are allowed we might find the least cost generation mix in which paid FCAS is a minimum.

    It’s not just the need for ancillary services but the LGC subsidy as well. Why should we be paying an additional $85 per Mwh to wind power when it could go missing in action when most needed in heatwaves? Why should we subsidise solar when peak electrical demand is in the early evening? Make them earn their place by paying their associated costs.

    1. There is a “causer pays” methodology for ancillary services in the NEM.

      For frequency control you have it split in the raise/lower for call the following: regulation; 6 second; 60 sec; 5 minute. Those are the timeframes.

  4. Some readers will have noticed the lack of discussion about black start capacity. It is nowhere near as much a constraint as are other ancillary services. Small generators can start large ones, hence 23MW from a single Hunter Valley gas turbine can start a 500MW Liddell unit. Two can start a 660MW Bayswater unit, and from there the whole system can be re-started from black, with loads and generating capacity progressively ramped up. Black start capacity exists at multiple points within the NEM – it is not in short supply and is not seriously endangered by increases in wind or solar generation.

    Also, black starts are nowhere as often at less than 1/year, system-wide than are the occasions when frequency or voltage correction are required, ie virtually continuously.

    1. Mmmm… except that it was probably the main challenge identified in the DENA report and one they do not yet have an answer to beyond keeping the existing fossil generators in the system.

    2. That’s in the AEMO ESOO report called System Restart Ancillary Services. I shudder to think of the loss of those assets that provide this service. I didn’t realise that you needed generators online to restart other generators in blackout conditions.

      Page 36 of the ESOO report on SRAS is cautious reading for SA’s energy market transition.

  5. Have you seen any resources that say batteries can’t provide FCAS? My understanding is that there are batteries providing FCAS in the market right now

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: