Canals and powerlines

It’s nice when blogging pays off in the form of published output, but it rarely happens as quickly as this. On Tuesday night Ken Parish alerted me to an amazing infrastructure proposal, similar to Colin’s canal. The plan is for a 3000-km transmission line connecting Darwin to the National Grid. Even a cursory examination was sufficient to show that the economics of this idea were crazy, and some late-night work produced an analysis ready for publication in today’s Fin (article over the fold).

I’m grateful to Ken for pointing this post out to me in the first place, and to commenters Robert Merkel and Derrida Derider (both regulars here) among others, who raised points that help me clarify my argument. Unfortunately, you can’t give this kind of credit in an opinion piece, but I can do so here. This is the kind of thing that shows the power of blogging.

American philosopher George Santayana once observed that those who do not learn from history are condemned to repeat it. I imagine he had a historical span of decades or centuries in mind, but it appears NT Opposition leader Denis Burke’s historical memory does not even go back six months.

In January this year, Colin Barnett looked set to become the next Premier of Western Australia. Despite a reasonable performance in its first term, the Gallop government was lagging in the polls and looked doomed to defeat.

Then Barnett came up with the idea of a 3700km canal to bring water from the Kimberleys to Perth. Running on the slogan ‘decisions not delays’, Barnett disdained such bureaucratic red tape as benefit-cost analysis or environmental impact studies, promising to implement the project regardless.

Working on the limited public information available, I estimated that water from the project could cost as much as $10/kilolitre, a price at which it would be cost-effective to tow icebergs from the Antarctic, Mirages don’t win polls 10/2/05. Any credibility Barnett had left after this proposal was shredded when it was found he’d left a $200 million hole in his costings. Labor was re-elected and Barnett is now an ex-opposition leader.

Amazingly, Northern Territory Country-Liberal Party Leader Denis Burke is following in Barnett’s footsteps. He’s proposed a 3000km transmission line to connect the Northern Territory to the national electricity grid, via Queensland, at a cost of $1.3 billion. Burke suggests that the proposal could reduce NT electricity costs by 30 per cent.

Although there’s much that is murky about this proposal, it’s perfectly clear how this number was obtained. As Burke’s statement observes, the cost of power in the Territory is 14 cents/Kwh, compared to a national average of 10c/Kwh. If Territorians could get their power at the national average price they’d save 30 per cent.

Of course, this could only happen if the transmission line, including both capital and operating costs came free of charge, that is, if it were paid for by somebody else. Getting someone else to pay has been the traditional CLP approach to infrastructure finance, a point emphasised in Burke’s statement, which notes that the line will be ‘the one of the biggest Territory building projects since the CLP built the Alice Springs-Darwin rail link.’

In the present case, the assumption is apparently that the costs of the NT link can be borne by the existing users of the grid. Unfortunately for Burke, the National Electricity Market doesn’t work like that, so the only plausible source of financing is a direct handout from the Commonwealth. As Colin Barnett could tell him, this is unlikely to be forthcoming.

Let’s assume then that the cost of the transmission line is passed on to consumers. At a conservative average cost of capital of 9 per cent, the capital charge for a $1.3 billion dollar line would be about $120 million per year. With depreciation and operating expenditure of $40 million total cost would be $160 million per year.

The Territory’s total consumption at present is about 1600 Gigawatt hours per year. Assuming, very optimistically, that 50 per cent of this demand was switched to interstate suppliers, the transmission charge alone would be 20c per Kwh. Against this, buyers would save the 4c a Kwh difference between the NT and grid prices, implying that the total charge would be about 30c/Kwh. At this price, Burke’s other idea of relying on solar photovoltaics looks considerably more sensible.

Why are the economics of this proposal so awful? First, although long-distance transmission using high-voltage direct current (HDVC) is feasible, this project is unprecendented in its scale. According to the ABB Group, leaders in this field, the world’s longest transmission line is a 1700 km line transmitting power from the Inga falls in the Congo river to the copper mining district of Katanga in the Democratic Republic of Congo (DRC).

Second, the market is tiny. Darwin’s peak electricity demand is around 200 MW, and most of this will continue to be met by existing generators. The Inga-Shaba line has a power rating of 560 MW and the optimal scale for this transmission technology is even larger.

In a generally unedifying campaign, we can hope for one positive outcome. If the CLP is defeated, as seems likely, the idea of presenting uneconomic white elephants as election fodder will seem a little less attractive next time around.

John Quiggin is an ARC Federation Fellow in Economics and Political Science at the University of Queensland.

41 thoughts on “Canals and powerlines

  1. Thanks, Robert, I’d missed that one.

    It just gets worse and worse (or, depending on your viewpoint, better and better) doesn’t it?

  2. “the only plausible source of financing is a direct handout from the Commonwealth”

    We paid for their $100 million parliament building ($150 million in today’s money) for what is nothing more than a glorified local council. Don’t rule out us paying for their crackpot transmission lines, on “national development” grounds.

    And especially don’t rule out Fat Kim supporting the idea.

  3. Dave, part of the object of writingit was to do what I could to foreclose this possibility.

  4. If the NT gets gas from the Timor Sea, as is planned anyway, won’t that give them all the cheap electricity they need anyway?

  5. Apparently, there’s some concern that the Timor deal will fall through, or that BHP will produce LNG offshore, making a pipeline to Darwin uneconomic. But they could also connect to the Moomba field via an existing Alice-Darwin pipeline.

    In any case, even going back to burning oil would be cheaper than this idea.

  6. Thanks John, you’ve just inspired my to study economics next year rather than arts. I’ll be doing Physics and the History and Philosophy as Science majors. Haven’t yet made up my mind whether to go Melbourne or Queensland, but leaning towards Melbourne and it’s physics department.

    Would a VTF at a top speed of 300-350 km/hr transporting both freight and passengers be a with government bonds for extra funding. white elephant? I am looking at a scoping study at the moment by the department of transport and regional services.

    Entirely government funded and owned by the federal government,

    Route: Melbourne, Albury-Wodonga, Canberra, Sydney, Newcastle, Toowoomba, Brisbane.

    Extra stations to service freight

    Ticket price somewhere between $50-$80 for a one way adult Melbourne to Brisbane trip, to make it competetive with air travel.

  7. Liked this one. A great AFR start to my morning. You are at your best John when you attach numbers to things.

  8. Nice work John. When I first saw this at Troppo the canal was the first thing I thought of. These Liberal opposition leaders, they don’t learn do they.

    Looks like another resounding victory for the ALP.

  9. So why is it that the libs are dead stately and territorialy but very much alive federally?

    After the last federal election labor has been returned in the ACT, WA and now NT.

  10. John

    Very good work. You’ve quickly come to the correct conclusion (in my view, at least).

    There are a few not insubstantial nit-picks that you may have to deal with in letters to the editor, though.

    a) You’ve used retail electricity prices as the basis of comparison (14c/kWh v 10c/kWh) when you should have compared wholesale prices. According to the National Electricity Market Management Compan (NEMMCO) the average wholsale price in Qld for 04/05 was $29.34/MWh=2.934c/kWh. The proper comparison is to the oil-fired generation in NT which would cost something like $500/MWh=50c/kWh in fuel alone.

    b) The length of the east-coast network is already something like 3,000 km, an extra 1,700 km is no big deal, and anyway it’s less than length of the network across the USA, which is something like 5,600 km

    Like I said, I agree with your conclusion, but there are some difficulties working through the numbers.

  11. the average wholesale price in Qld for 04/05 was $29.34/MWh=2.934c/kWh.

    (This is a running tally for the current year. In the last hour the number has dropped to $29.31/MWh=2.931c/kWh, just in case anyone is scratching their heads over the numbers on that page.)

  12. SJ, I think the retail comparison is OK. Distribution needs to be paid for in both cases, as does use of existing transmission lines.

    Your cost for oil-fired generation looks way too high to me, do you have a source? In any case, the relevant comparison is with gas – the possibility of going back to oil remains a backstop if existing gas supplies run out before one of the new options becomes avaialble.

  13. Back of the envelope calculation follows.


    – Fuel oil price = $1/L
    – Energy content of fuel oil = 40 MJ/L
    – Thermal efficiency of turbine = 0.3

    Cost of electricity = $1/L / 40 MJ/L / 0.3 x 3,600 s/h = $300 / MWh

    So, not quite as high as I’d guessed last night, and you’re right about the gas, too.

  14. SJ, I think your thermal efficiency is a bit low – a turbine can get close to 40% and if you use a combined cycle turbine (you use the water cooling to run a secondary steam turbine) you can get close to 60%.

    You raise an interesting question, though. If the production cost of electricity in Darwin is greater than its retail cost, who’s making up the difference? The taxpayers of NSW and Victoria, as usual?

  15. An extension of 1700km is a big deal for such a small market. The US is heavily populated right across from east to west so a grid there is worth it.

  16. Energy sources in the NT include uranium and abundant sunshine, so into the analysis along with the cheap gas should hopefully go distributed solar and nuclear. In a rational world, that is. Particularly solar as its peak availability is probably an excellent match to the electricity demand peak from air conditioners, but why not consider nuclear for the baseload?

  17. On thermal efficiency, there were experiments in the ’20s and ’30s using secondary systems in which the primary used boiling mercury (it was difficult to keep ultra-hor steam from being too corrosive). Recently there have been experiments with thermionic generation and thermophotoelectric systems (which “fake” operating at higher temperatures than the actual equipment). But of course efficiency is not the same as power, and the economics don’t work out. They are valuable experiments, of course, and show what might one day be practical.

  18. I’m not opposed to nuclear in all circumstances, but it is a non-starter in Australia where our renewable resources are so good, and a complete non-starter in Darwin. I’m not sure of the smallest size of a commercially viable nuclear plant, but I would imagine it would be a lot larger than the entire Darwin market. What is more, given that so much of Darwin’s demand would match sunlight, a sensible system would use solar (not necessarily photovoltaics) for the bulk of usage, so the remaining baseload would be really tiny. Aside from the unpopularity, nuclear would be using an elephant gun to kill a gnat.

  19. Maybe you could put the nuclear plant halfway between Darwin and Mount Isa and connect them both up to the grid.

  20. Have opposition parties always been inventing wildly uneconomic megaprojects or is it something to do with the modern electorate? There have been big projects proposed in the past but I thought they were on a subject to feasiblity basis. It’s a worry, it seems like some charismatic who couldn’t run a tuck shop successfully will get one of these things through the electoral process before long.

  21. Like Pauline Hanson’s plan for a 2% flat tax and cheap mortgages for everyone. “Where will the money come from?”. Hanson: “Print more money”.

  22. The smallest modern nuclear designs are very much of the size of the Darwin market and are most likely quite cost competitive with fossil fuels, particularly in places like Darwin. Their Australian proponents at UIC have plenty of information about one technology, there are others. All that sunlight in the NT though – at least let’s rule out coal or oil fired plants as sensible proposals.

  23. Maybe if governments banned the use and sale of eggnitioners in favour of dehumidifiers and fans. If you live in the tropics then cop it. Even if you live in Melbourne during summer you should also cop it. Cheaper and use far less energy.

    A fan and a dehumidifier saves much energy and are just as effective. dehumidifiers are also perfect for Brisi, no installation costs and cheap to buy.

  24. We had a discussion here previously regarding Geodynamics LTD and their geothermal power project in Central Australia (which incidentally includes use of the Kalian cycle heat energy for scavenging enrgy from low temperature heat – much in the vein of the stuff PM Lawrence was mentioning earlier).

    At the time I suggested a potential use for the power from the project woudl to electrify the Adelaide to Darwin railroad. A power transmission line running from the Cooper Basin along the A-D rail ine could also provide power to Adelaide (which has supply problems currently), Alice Springs (which also depends on oil) and the Olympic Dam mine which is a major power user in its own right.

    This might still not be viable but I suspect it’d be a better bet than the Darwin-Mt. Isa powerline.

    Enviromission is claiming that improvements in the design and enrgy efficiency of their solar tower proposal mean its now feasible for sizes from 20 megawatts up and should deliver power at rpices comparable to coal (much less to deisel).

    You’d hope testing those claims would take priority over the power-line idea.

  26. Robert Merkel:

    SJ, if the cost of oil-fired power was really that high, you’d use solar and/or ship coal up there on purely economic grounds.

    The cost of oil-fired power really is that high, but I hadn’t realised last night that the NT had built a gas pipeline from the Cooper Basin to Darwin, and that they run their major power station at Channel Island on gas.

    It’s worth noting, though, that NT PowerWater “operates 68 remote community power stations, largely diesel powered.” So oil does really get used as fuel for electricity generation in the NT.

    You raise an interesting question, though. If the production cost of electricity in Darwin is greater than its retail cost, who’s making up the difference? The taxpayers of NSW and Victoria, as usual?

    It could just be an internal NT cross-subsidy. I don’t know.


    On thermal efficiency, there were experiments in the ‘20s and ‘30s using secondary systems in which the primary used boiling mercury (it was difficult to keep ultra-hor steam from being too corrosive).

    The binary-cycle stations were more than just experimental:

    And, in 1923, it installed an experimental General Electric mercury-steam generating unit. This was followed in 1928 by one of the first commercial mercury cycle generating units in the world, at HELCO’s South Meadow Station in Hartford. (Although the mercury-vapor cycle offered improved thermodynamic efficiency, its use was discontinued for several reasons. Its health hazards became known and major improvements in conventional steam turbines eclipsed the technology.)

    It wasn’t the only one in the world, either. I can remember reading about the problems cleaning up the site after the decommissioning of one of them in Asia (Phillipines?) back in the eighties.

  27. SJ, I know the mercury things were actually built; but as the quotation itself shows, they didn’t go beyond experimental installations. I didn’t mean lab-only experiments but experimental installations to see just how far things could be taken. And, as we see, they didn’t go so far as regular power production in ordinary use.

  28. but as the quotation itself shows, they didn’t go beyond experimental installations

    Maybe if you read beyond the first sentence, you’d see the second one: “This was followed in 1928 by one of the first commercial mercury cycle generating units in the world…” (emphasis added).

    I don’t think they were using the word “commercial” as a synonym for “experimental”.

    There were dozens of the things, all over the world, operating for decades.

    A few random quotes:

    Such growth called for more generating capacity, so, in 1948, the company built the Schiller Plant, a 40,000-kW mercury binary cycle facility, in Portsmouth.

    The Kearny station of Public Service Electric and Gas Corporation — N. J.
    — operated a mercury binary cycle power plant. OK — this in the early 1930’s now!! The Kearny plant delivers: A 20,000 kw mercury boiler/turbine unit then delivers steam (from second
    stage) to a 33,000 kw turbine for a total of 53,000 kw for the combines
    mercury-steam unit.

    Your insistence on the term “experimental” is rather odd.

  29. Just coming back on the numbers, SJ, $1/L for fuel oil is too high. That’s the price of petrol, tax and all. I think 40c/l would be closer to the mark. Then if you can get 60 per cent efficiency you’re looking at about 10c/kwh for generation costs, which seems reasonable to me.

  30. John:

    I’ll get back to you on the oil price.

    There’s a fundamental problem in using the 60% efficiency figure. You wouldn’t expect the new powerline to displace the most efficient producer in the NT, i.e., the combined cycle Channel Island plant. You’d expect it to displace the least efficient. Leaving the remote diesel plants aside for the moment, the least efficient plant connected to the NT’s grid are (at a guess) the open cycle turbines at Katherine.

    There’s a map of the NT grid here.

  31. (Note, again, John, that I don’t disagree with your conclusion that the powerline isn’t justified).

  32. SJ, I did read the lot. I called that experimental despite the power being sold commercially, simply because it was clearly being operated on a trial basis. I think the disagreement is over what we call experimental, but it doesn’t do to suppose someone isn’t paying attention. I don’t see any inconsistency in things being used commercially and yet being experimental; I would say the Concorde was like that (it wasn’t economic, except on a sunk costs basis).

  33. John:

    I think we’d need some input from someone who knows more about oil pricing than I do.

    The current wholesale price of distillate in the NT seems to be $1.18/L, and the rebate is what, $0.38/L?, giving a final price of about $0.80/L.

    That’d put the fuel cost of electricity generation at 24 c/kWh.

    That’s SRMC, neglecting any incremental maintenance costs.

    It’d be OK to use a figure like that if the NT had plenty of electricity generating capacity, and was not contemplating building more. Otherwise, we’d have to work on capital costs and develop a LRMC.

    I propose that we wait a week for the election.

  34. BTW, John:

    “That’d put the fuel cost of electricity generation at 24 c/kWh.”

    I did some work for Fred Hilmer back in 1995, and at the time we put that number at 25 c/kWh. Given the increase in oil price since then (offset somewhat by exchange rate changes) I’m surprised that the answer turns out to be about the same.

  35. In a similar vein:

    PETER CAVE: To the Northern Territory now where there are just a few days campaigning left in the elections and amidst the usual issues of law and order, health and education, a handful of more unusual visions have emerged, including plans for a recreational lake in the desert.

    The Country Liberal Party wants to create a large lake near Alice Springs, one of the driest parts of the country.

    Rachel Carbonell reports from the Northern Territory.

    RACHEL CARBONELL: Alice Springs is almost always hot and dry and has low erratic rainfall.

    The local Country Liberal Party member, Dr Richard Lim, believes he may have the solution to the tough living conditions.

    RICHARD LIM: Alice Springs deserves to have the lake. Every time I go door knocking and I speak to people, people tell me that the one thing that they desperately miss is a great body of water that they can play, recreate upon.

    RACHEL CARBONELL: Dr Lim has promised a feasibility study, but ruled out damming rivers or tapping into the massive underground water supply.

    He’s not the first to dream of a desert oasis.

    Former Alice Springs town councillor, Des Rogers, says he too campaigned on a platform of water for fun.

    DES RODGERS: Richard Lim is right in what he’s saying, that if you canvass most people in Alice Springs, they’d all say that yeah it’d be wonderful to have a lake there. But I think when you start looking at the technical side of it, it’s simply not practical for the expense. There’s three metres of evaporation rate. I wonder how you’re going to keep the lake full and maintain that level as well.

    RACHEL CARBONELL: The idea dates goes back as early as 1970, when proposals first emerged to build a flood mitigation dam to protect Alice Springs from the massive “once in a generation” floods that have hit in the past.

    Various feasibility studies have been done but the project has never made it past first base.

    John Brisbin, from the Arid Lands Environment Centre in Alice Springs, says there’s probably a reason for that. It’s not a very good idea.

    JOHN BRISBIN: I mean, do we want to have a malarial mosquito breeding swamp? You know, is that what we’re looking for? This is a big change environmentally. If you’re talking about a lake, that’s a substantial body of water, so you can imagine wetlands on the side of it, you could imagine all kinds of exotic grasses coming in and then, well there’d be birds. Birds are eating what – flies, insects, mosquitos again. You’ve got all kinds of issues there that would arise.

    If we absolutely had to have a lake as a matter of survival, well then, we’d have to look at a lake. But to have a lake out there as kind of a lifestyle option, to just put it into the middle of the environment like that, is asking for problems that we don’t really need.


  36. BENNO (June 10th, 4:41 PM): As a lifelong resident of Darwin I cannot let your statements about the alleged indulgence of “eggnitioners” (air-conditioners I presume) pass without comment.

    1. You say in a previous post on this thread (June 9th, 9:18 PM) that you are going to study physics. Well, you obviously have an awful lot to learn about the basics of this important subject if you think dehumidifiers will solve the problem (“are just as effective”). If they did work, AND were cost effective (as you also assert), we would have started using them decades ago—NOBODY likes paying increased electricity bills. Can you quote a reliable source for your claim about dehumidifiers (in combination with fans) being effective and cost efficient in hot, humid tropical environments? Or provide some solid calculations of your own? I know of only three ways to dehumidify air.

    a) Heating. However, this kind of defeats the purpose for tropical environments, as you then have to somehow cool the air afterwards. And it takes a fair bit of energy to dehumidify air by heating in the first place.

    b) Chemical desiccation. Requires large amounts of desiccant, works too slowly, needs constant replacement of desiccant, and doesn’t reduce the temperature.

    c) Air-conditioners. These both dehumidify and cool at the same time, which is why they are used, often in combination with overhead fans which increase the wind-chill factor, thus reducing the amount that the air needs to be cooled in the first place.

    2. Heating buildings is also a major user of energy. Would those who live through long cold seasons (most of Australia’s population) also be required by government fiat to go without heating? To paraphrase your own line: ‘If you live in a temperate zone then you should just cop being cold.’ At least in temperate zones people have the option of putting on more clothing.

    3. Unlike heating, peak cooling requirements generally coincide with maximum sunlight, which means energy requirements for cooling could be substantially offset by using renewable solar power (photo-voltaics, solar collectors, or solar towers—see Ian Gould’s post on this thread, June 10th, 5:26 PM).

    4. The efficiency of aircons has improved considerably in the last few years, and will probably continue to do so. (By the way, reverse-cycle aircons are extremely efficient heaters. Has the efficiency of conventional heaters, especially radiant heaters, similarly improved? I doubt it.)

    5. Building insulation materials have also improved considerably in the last few years. If you want to reduce energy use for both cooling and heating buildings then mandating much higher insulation standards would be a VERY good place to start. Say a minimum R rating of 2.5, with rebate incentives for higher ratings. Insulation can also be retrofitted to most buildings fairly easily (at least in the roof and often the floor, if not the walls). I am strongly in favour of energy efficiency and conservation, and practice what I preach—the total R rating on my house is around 2.5-3.

    6. By the way, the total heat load on an organism (in this case humans) is not only dependent on air temperature and humidity, there is a third factor, radiant energy, which peaks in the tropics and when it is hottest. A 32 degree, 60% humidity day, can be comfortable if the radiant load is low, when the sun is low in the sky (May-August). But if the radiant load is high, when the sun is high in the sky (October-March), then a 32 degree, 60% humidity day can be very uncomfortable. Simply getting into shade during the summer doesn’t solve the problem, although it helps. (Have you ever experienced a full tropical summer, or 40, especially the build-up from about late September until the rains break? It is not pleasant and seriously reduces people’s capacity to function efficiently, effectively and enjoyably. It is also when violent behaviour and suicides peak.)

    7. On simple political grounds alone no government of any political persuasion would have a hope in Hades of banning air-conditioners anywhere in Australia, let alone in the tropics. And the same goes for heaters.


  37. (By the way, reverse-cycle aircons are extremely efficient heaters. Has the efficiency of conventional heaters, especially radiant heaters, similarly improved? I doubt it.)

    Just an aside, here. The efficiency with which pretty much any kind of electrical heater (radiant, fan, oil filled, etc.) converts electricity into heat is 100%. So there never was any room for improvement in the first place.

    Reverse-cycle air conditioners are the exception. They don’t try to convert electrical energy into heat. What they do is pump heat from one place to another, i.e., from the air outside your house to the air inside your house. They can do this even though the air outside is colder than the air inside.

    A particular model might use 1 kW of electricity to drive a compressor and fan to transfer 2 kW of heat from the outside air to the inside air. The 1 kW of electricity ends up as heat inside the room, too. In effect, you use 1 kW of electricity and you end up with 3 kW of heating in the room.

    So there’s no question that reverse-cycle air conditioners are more efficient than any other kind of electrical heater (in most circumstances: I’ll come back to this point).

    But where did the energy come from that made the electricity? The bulk of the electricity used in Australia comes from some kind of fossil fuel. Only about 30% to 60% of the energy contained in the fuel ends up as usable electrical energy. If all you’re going to do with that electrical energy is run an ordinary electric heater, it would’ve been more efficient for you to burn the fuel yourself, at home. There are environmental reasons why we wouldn’t want to encourage a move back to coal as a direct home heating fuel, but gas does the job pretty well in the areas where it’s available.

    The point I promised to come back to: for an area that isn’t continually occupied, like the bathroom in your house, it doesn’t make sense to keep the area continually heated. In cases like this, radiant heaters like the combined fan/light/heat globe things are the way to go. The heat globes are just radiant heaters that start up instantly, and operate at very high temperature. The heat globes don’t really attempt to heat up the air in the room (although they will, eventually). They act directly on the skin, making you feel instantly warmer, even though the room is still just as cold as it was before you turned it on. It’s the same effect that you feel when moving into sunshine from shade on a cold day.

    (On a re-read through your comment, Eskay, I can see that you probably know all of that already. I’ll submit my comment anyway, because someone else may find it informative.)

Comments are closed.