AC/DC: The unheralded debate in Australia’s future grid

Print Friendly, PDF & Email

The Integrated System Plan being put together by the Australian Energy Market Operator will put a marker down about how Australia will go about organizing its generation future.

It will effectively represent a challenge to all the vested interests in the current system because it will explicitly and implicitly become a formal document that models a very different electricity system to the existing thermal dominated model.

No matter the form of regulation, the NEG, a specified renewables share, carbon tax (my favorite), EIS, CET, it’s the physical design and capacity of the system that will end up driving price.

In this sense we have long seen the ISP as more important than the NEG. Transmission investment is a 5-8 year time frame, getting this right is AEMO’s No 1 task.

In this note we focus on the possibility that  there may be significant advantage in a renewables centric grid to using DC for transmission as opposed to AC.

Right now most believe that consumption growth to 2030 will be flattish.

So new investment – whether in transmission or generation – will likely require retirements of existing capacity.

We note  if consumption is flat and yet we keep investing it will be difficult to manage final consumer prices down. It may be possible but it will be challenging. Prices are less important than total costs.

In California prices are much higher than in Texas but household customer bills are lower because Californians are more energy efficient.

The background challenge, and part of the ISP is coordination of behind the meter with in front of the meter.

Finally, we note that the key bone of contention will be the RIT-T test. This test basically requires that transmission proposals show a net economic benefit essentially looking at the change in consumer costs at both ends of the transmission line.

It is a fact that little new transmission has been built in recent years and that satisfying the requirements of the RIT test have been difficult.

It’s also a fact that the main transmission lines that connect the coal generation in the Hunter valley and the La Trobe valley were never built using the RIT test.

We simply do not see how renewable energy zones [REZ] can be established without significant transmission investment and that a more efficient process than the RIT is required if these zones are to be established in a timely fashion.

The ISP What is it?

The plan had its genesis in the Finkel Report, recommendation 5.1. “to facilitate the efficient development and connection of renewable energy zones across the National Electricity Market”

The concept has grown to include interconnected infrastructure and energy developments including transmission and generation.

The key point is it’s a plan for delivering INFRASTRUCTURE to facilitate an “orderly system transition”.

The first plan is to be released in June 2018.

Why its important

As renewable developers are quickly finding out, transmission issues can be the life and death of wind and PV farms. Recent revisions to marginal loss factors [MLFs] has shown that project economics can be driven by difficult to foresee changes in network utilization.

More importantly its been obvious for years, blindingly obvious but still the AEMC has turned its blind eye against it, that significant development of renewables is going to require transmission links that could never be easily justified under the existing RIT-T.

Transmission takes 5-8 years to get built, even when everyone thinks it’s a good idea. And not everyone does think it’s a good idea.

Vested interests don’t want new competition, the question of how transmission is paid for, the difficulty of forcing new investment into a market seeing flat consumption for another decade are all thorny issues.

Nevertheless, my strong view, based on the success of the ERCOT investment in Texas and the obvious fact, to me, that lots of coal generation has to be replaced over the next 10-15 years makes it imperative that the ISP is a success.

A strong transmission network provides security, flexibility and optionality. Easy access to transmission will incentivize otherwise marginal wind & pv projects to get the go-ahead.

Where is the process up to?

AEMO has published some scenarios and assumptions to be used in the plan, and has run a consultation process. Stakeholder submissions to the process have been received and AEMO has published a summary.

The next step will be publication of  Version 1.0 of the plan in June. After that the arguments will start.

The first argument will likely be about changes to the Regulatory Investment Test [RIT-T] that will surely be required.

We assume that argument away because in our view of the world new transmission is going to be required.

What we see as the key discussion that needs to be held is whether the new transmission should be AC or DC.

Two, in our view, key submissions to the consultation were from Siemens and UNSW. We award the Siemens submission an  ITK gold award with the UNSW submission a worthy contender.

Do yourself a favour and read the excellent Siemens paper. Of course, like every submission it talks its own book. In this case it’s a really good book.

What transmission is proposed?

Broadly speaking at least $3 bn of transmission investment is proposed, but if this was to be done in a DC rather than AC form the initial cost would be higher.

In addition, the Powering North Queensland $1bn of investment, vital to the future of that region and maximizing the diversity of power source in the NEM is generally not included.

Most of the debate is in the NSW-Victoria-South Australia axis, but augmentation of the QLD- NSW interconnection, perhaps by converting the main line to DC is also under consideration.

The following map from the SnowyHydro submission shows the Southern renewable energy zones.

Figure 1 REZ and Transmission in South East Australia. Source: Snowy

Figure 2 QLD, Northern NSW REZ. Source: AEMO

The AC v DC debate

Most of Australia’s electricity transmission network is AC. Australia has one of the longest, if not the longest, interconnected transmission grids in the world.

Historically, the debate about AC v DC has simply been about costs. Simply put, DC connection has higher capital costs but lower line losses.

Siemens provided the figure below in their submission and noted that the current economic cutover distance, depending on load, was around 500-800 km.

The figure shows that the DC terminal station costs are more than twice the cost of AC terminals.

Figure 3Typical cost comparison between HVDC and HVAC. Source: Siemens

Wait, there’s more  – DC does “power controlability”

Factors other than cost are becoming important. Siemens notes that DC can do:

  • Precise power flow control
  • Enhance AC stability
  • Reactive power control including AC voltage support
  • Frequency control
  • Black start (hello South Australia? Hello North Queensland?)
  • Oscillation damping

The UNSW submission notes all the above advantages and adds that DC transmission can operate under “low short-circuit ratios”

Your analyst is about as far from being a power engineer as it’s possible to get, but surely there must be some cost saving for PV farms that initially produce DC electricity to connect to a DC transmission line without having to go through an initial AC conversion process.

Most readers will be aware that the Blakers pumped hydro work also talks to, and costs a DC transmission network.

If HVDC converter stations are required for every PV farm, then DC transmission is expensive

Your analyst has assumed that PV farms, in particular, can connect to an HVDC line cost effectively.

However the Powerlink submission noted in connection with the option of a Parallel HVDC network along the Queensland coast, that although it might be cheaper on a capacity equivalent basis, “the ability for generation to connect into a HVDC circuit along its length is limited due to the high cost of HVDC converter stations”.


Siemens points to

  • a 75km, 1000MW DC line being built to enhance Belgium-Germany connectivity. Just 75 km but they still chose DC
  • Conversion of an AC link to DC in Germany (“the Ultranet project”). This conversion increased capacity on that line by 20% plus the increased control and security. We do note that the main DC link in Australia is Basslink and its hardly been a byword for security in the past year or two. Nevertheless…..

UNSW pointed to:

  • Expansion of the France-Spain interconnection. This was only a 64km distance and was successful with flow doubling over the link for only a 70% increase in notional capacity.
  • Various China projects of which the Nan’ao multi terminal project was notable for the low distance, less than 50km and only 350MW of wind power. It was the non cost advantages that won the day, efficiency, flexibility, reactive compensation and independent configuration.

Call now our operators are waiting – the medium voltage DC concept [MVDC]

The Siemens submission went on to mention its MVDC concept. Potentially, Siemens states in 33KV to 132 KV networks MVDC is advantaged because it allows “seamless control of the active power flow”.

We think this means in practice its better suited to bidirectional or omni directional power flow than the traditional one way AC approach. Siemens claims it may offer higher capacity at lower voltage.

The Siemens concept shares my values. Microgrids and a transmission backbone

In the Siemens world an efficient DC backbone is connected to lots of micro grids (I define a micgro grid as a load that can operate independent of the main grid for some time interval).

Siemens notes that the thermal efficiency of centralized coal generation is around 33% but that the overall efficiency of microgrid power can be as high as 80%. Siemens therefore claims that distributed power has a higher value.

As such Siemens states that even strongly connected regional demand centres, eg Newcastle or Ballarat or Gladstone or Murray Bridge might develop their own resources and be managed as a microgrid.

The figure below is very futuristic as today, in my view, hydrogen electrolysis and storage is clearly uneconomic but the concept is clear.

Figure 4 Grid integrated town using MVDC. Source: Siemens

AEMO’s demand scenarios don’t leave much room for new supply

Various submissions to the AEMO process questioned the internal validity of the scenario assumptions.

The point we want to make is that even in the “strong” in front of the meter scenario the cumulative  increase in demand by 2030 is 21TWh which would require around 7 GW of new wind & PV – not a lot for 12 years.

The neutral scenario, which tends to be where the average policy maker or reader ends up, assumes a very small fall in consumption over the next decade.

The weak scenario for in front of the meter demand, sees about an 18% cumulative fall in in front of the meter demand.

Figure 5 AEMO demand scenarios for in front of meter. Source: AEMO

Net demand for renewables then is likely to heavily influenced by thermal retirements.

Forecasting is a mug’s business.

Human judgement tends to use recent experience as the best guide to the future, I guess it’s the  availability heuristic  also  anchoring. In any case it may or may not be correct.

There is money to be made and lost by making a different call but it’s a big bet.

One great advantage of wind and PV is that they are an excellent investment in times of high uncertainty because investment is modular and short time frame.

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.  

  • Tripp Tucker

    The AC v DC debate needs to reconsidered with 3DFS technology that provides users of AC, line losses more in line with DC, and a variety of other benefits.

  • BushAxe

    Unfortunately at this stage HVDC is only cost effective for point to point where cost justifies the low losses and the need for precise control of power flow, frequency control etc that power electronics (thyristors) can provide. DC links are also used to connect systems with different operating frequencies. However they can’t provide in other areas like fault current, inertia that AC can so choice of link is a balancing act of what is more important. I think HVAC power control will improve in coming years as well just to complicate things.

  • wmh

    Renewables do not change the balance of advantages of AC and DC. Transmission losses are important and DC lines don’t have to drive line capacitance and inductance. Every renewables generator will need either a DC-AC or a DC-DC inverter to transform the generator voltage to the required line voltage. DC is hard to switch off so needs more expensive switch gear. Perhaps the need to locate wind generators at sites with good wind resources, rather than proximity to loads, may result in more DC links.

  • Peter F

    If we install wind and solar at the same physical density i.e. wind turbines/ solar panels/square km that Germany has today we can generate all the energy we need within 180 km of the population centres so then the only need for long distance transmission is as an alternative to local storage.

    Behind the meter generation and storage has many economic benefits to customers so it will continue to grow and actions like IPART reducing feed-in tariffs will just accelerate the trend, thus reducing peak transmission demand. In SA and Tasmania it is fairly clear that in state storage even in front of the meter is already more cost effective than increased transmission so it is fairly obvious that power flows into those states will fall quite significantly, thus undermining any investment in new transmission to them.

    Thus the justification for new large scale transmission investments relies on moving generation from the Hunter and Latrobe valleys to REZs which don’t have the same existing transmission capacity. However, if transmission and storage is localised, the scale of REZs will be significantly reduced and the current transmission capacity will probably be largely adequate. At the current rate of progress peak demand on the NEM will be below 30GW well before any new large scale transmission projects could be completed, while wind and solar near Broken Hill, Portland, new England etc will reduce long distance power flows even within NSW and Victoria.

    In summary, while there are sure to be some projects where the grid needs to be strengthened, overall HV grid utilisation will fall due to distributed generation and storage together with energy efficiency. Therefore large scale investments in transmission are likely to increase the cost of power rather than reduce it.

    Another issue in Australia is that DC links seem to take much longer to repair than AC links, not just here but in the US as well. With a long grid like Australia with limited alternative transmission paths, DC links can make the system more vulnerable to disruption.

    • Mike Westerman

      Peter bear in mind that altho’ end use demand may be falling, transformation from FF to RE with storage means that demand for storage will mean total demand during daylight periods will be several times higher than current to provide the surplus of RE needed to cover the evening peak and overnight demand. Hopefully use will be made of both the Hunter and Latrobe infrastructure, with extensions out to REZ ie solar in the areas north west of the Hunter and east of Latrobe (including offshore wind off Gippsland).

      I would hope to see a coastal HVDC link from Williamsdale down Cann River and then on to the Basslink terminal station to open up the Monaro for more wind and pumped hydro on the escarpment.

      • Peter F

        This is an interesting question, my feeling is that peak power flows will probably fall, particularly if storage is located close to the load and can be recharged away from peak demand. However that really depends on how fast local generation and storage becomes the dominant source of supply

        SA VPPs and Highbury pumped storage are examples which will disrupt peak demand on the transmission grids because they can be charged from excess local solar as well as far away wind farms.

        • Mike Westerman

          Certainly in SA and north Qld, PHES will demolish the need for transmission augmentation – the transmission links become merely a backup or route for exports of surplus. Similar to Tas where I can’t see Basslink 2 or PHES being of any value until Vic has retired most of its lignite. But the Qld/NSW/Vic part of the network will most likely need significant links to provide geographical diversity and reduce the need for storage.

          • RobertO

            Hi Mike Westerman, I like the idea of Basslink 2 as a community / security project when put via King Island (if we are going to build it anyway and perhaps as a GW, 800 kV HVDC) I also think that it would help to retire Vic coal earlier if we built it sooner rather than later. The object of new large scale transmission is not to move power from Tas to Qld or SA or to move power from Qld to Tas or SA but to provide security of supply if needed to the idea of power is best if used locally. The Siemens doc stated Germany uses about 80 GW but they are building 400 GW. We in the NEM use about 30 GW and we will need about 100 – 130 GW (we have better RE resources).

  • Mark Roest

    Consider battery storage at USD100/kWh within 3-4 years, vs cost of transmission.
    Also, consider every building (and PV canopies over parking where necessary) providing sufficient solar electricity to power both the building and all the vehicles associated with it, with stationary battery storage where needed to fill demand when solar input is low. That should pay for all parts of the system within 3 to 5 years, if not less, through savings. Then energy is effectively free for decades!

  • TW

    You are correct – you are a long way from a power engineer. Maybe a technical check before you publish may be worth it, just so that the good parts of your article are not spoilt. There could well be a future for DC.
    There are only a few reasons to use DC.
    – Control of power flows, voltage and frequency – eg the DC links in the eastern grid. The Murray Link (S A to Vic) and the North Cost Link ( NSW to Qld).
    – Separate different grids – eg Britain to Europe
    – Long distance – as used in South Africa and China
    – Undersea cables – the capacitance of a long ac cable would be so large that power to flow along the cable is not possible. Eg Bass Link, Japanese Islands, Scandinavia to Europe.
    If (say) the Qld to NSW main interconnection was converted to DC then it would be very expensive for projects such as the Saphire Wind Farm to connect to Dc at the high voltages used for transmission.
    The short circuit ration of DC can be an advantage in the grid ( eg if connecting tow grid) or a disadvantage within a grid because a flash over or other fault on the system will result in a much large voltage dip with a small short circuit ratio.
    Low voltage and medium voltage DC systems might be efficient for small renewable projects connecting into the local system.

    • David leitch

      Hi TW

      Thank you for your comments. If you have a moment to list the bad/erroneous parts of the article i’d be grateful. I was hoping someone would. It seems that connecting to dc lines is more expensive. Some of the examples I pointed to were short distance eg the island in China.

    • Malcolm M

      1. Is the reason for the low reliability of HVDC in Australia that our expertise is very thin ? We need to wait for experts or specially made parts to come form Germany, whereas if we had our own HVDC industry we could keep the links going more reliably, as Germany and China seem to do?
      2. Could some existing transmission lines be reconfigured as DC and double capacity? For example from the Murray power station to South Morang there are 2 x 330 kV circuits and 6 conductor wires. The rating is ~1500 MW, but 8% of energy is lost in transmission. If instead there were 3 conductor wires energised to +330 kV and 3 conductor wires to -330 kV, would this effectively be 660 kV ? If so, the capacity could be doubled to 3000 MW, and we wouldn’t need a new transmission line for Snowy 2.

  • TW

    Cables underwater are a challenge. They are expensive to manufacture and even more expensive to repair.

    An ac cable needs to be three phase and hence three conductors and three sets of insulation. Read Expensive. In addition, an ac cable is a perfect capacitor – that is a conductor surrounded by insulating material and an earthed shield. Current has to be supplied to charge the electric field in the capacitance every half cycle. This reduces the amount of current that can be transmitted as power to zero over a (relatively) short distance.

    DC cables can be single conductor with an earth return ( read cheaper) and the cable capacitance is charged once when the line is energised.

    Other advantages are that the grids can be kept separate from a frequency control and voltage control perspective.

    As a result, So, short distances between islands can be “economic” for DC.

    As far as the Qld – NSW interconnection is concerned, DC would have some interesting consequences. At the moment, the interconnector is tied in to quite a few substations along the way eg Armidale, Tamworth, Muswellbrook. Building that many AC – DC stations would be expensive.

    • David leitch


      The repair point is interesting for proponents of the North West shelf to Indonesia project.

    • RobertO

      Hi TW, Having the Earth Return in the long run is not the best option given that if you have two cables (one in, one out) you can also have Earth return. If one cable fails then you can continue with just the one while you repair the second cable, so you still have some income. NZ Cook Strait Cable is 4 cables plus earth return and corrosion is an large issue. They use 2 power and one return and one spare and earth return.

  • trigger 2014

    We are seeing the result of selling off an essential service. In the good old days when we owned the network and had an independent public service all these things would be worked out, without the complexity of many profit driven counter competing companies. Now we are left with in-effeciencies and politicly driven objectives we didn’t even think of. Sure they may well have got it wrong, however they would have done some thing, and it would have been done for us not some overseas corporation.