Picking winners

Australia's largest CCGT, air-cooled

A few commentators to my Picking Losers blog earlier in the week quite rightly pointed out the potential for gas-fired CCGT plant to be a sensible baseload replacement for brown coal-fired power. I have looked at the numbers, and agree.

Taking the cost of new gas CCGT at around $1million per MW, replacing 6GW of coal-fired plant would cost around $6billion. You also have to add in the higher cost of fuel, which at $4/GJ and a heat rate of 7GJ/MWh adds $23/MWh or just over $1b per year of fuel cost to this baseload fleet. With the shutdown costs of the coal-fired fleet thrown in, and all capital costs amortised to annual, it amounts to about $3b per year. The savings are less than the 60MT if replaced by zero emission plant, and are around 45MT (with gas having about a quarter of the greenhouse emissions of the existing fleet of Latrobe Valley generators in aggregate). So the effective abatement cost is $69/t CO2e.

A few interesting observations here. Firstly, doing the numbers in aggregate disguises the fact that the abatement cost of replacing the oldest, less efficiency and therefore higher emission intensive plant is much lower than that of the newer more efficient plant.  (eg $55/t for the 60 year old Hazelwood plant, over $80/t for the more recently built Loy Yang). It therefore would not make sense to try to replace the entire fleet, but start where abatement costs were lower.

Secondly, it may not be that shut down is the most economically efficient strategy for the brown coal-fired fleet. As Garnaut pointed out in his paper on energy, released last week, there might be more efficient abatement from merely having the units operate more as intermediate rather than baseload units. This would be more closely aligned to the role some of these units were in prior to the start of the national electricity market. It’s mainly the interconnectedness of our energy regions that the NEM brought in that has allowed these units to truly baseload by exporting power to other states (eg SA, NSW, TAS) when the local demand is not sufficient to utilise the entire output of the fleet. By doing this, perhaps during times when gas prices are seasonally higher (winter), and when capacity is required (eg summer peak), it allows new gas-fired CCGT to see enhanced operation rather than being relegated to traditional role of intermediate.

As an aside, it’s actually the switching of fuel at the margin (ie between least efficient coal being displaced by most efficient gas) that is one of the main drivers of the EU ETS permit price (the EUA). Utilities (who are the main traders of EUAs, not “banksters” as others would have you believe) are continually optimizing their power generation portfolio on a daily and even hourly basis depending on the carbon price and the price of gas and coal.

Thirdly, if we assume that instead of paying generators to shut down we just instead allow them to reach the end of their economic life (eg the SECV had planned to run Hazelwood until 2005), then the costs are much lower. Removing this requirement to pay generators takes the abatement cost from almost $70/t down to $40/t.  Another assumption underpinning the idea of paying generators to shut down is that their shareholders will necessarily be the same investors that will invest in the new plant. However, even if these owners of coal-fired plant went on a capital strike, there would be other investors (that may or may not already own coal-fired assets) that would see a positive investment case on the new capacity.

All of this again raises the question of the appropriate mechanism for allowing the provision of electricity to transition to a carbon-constrained world. It seems to this coyote that having government attempting to direct which capacity additions are to be funded and which generators to be shut down, or whether or not to run them differently rather than shut them down, and also balancing up the relative long term prices of gas and coal, would be extremely inefficient. In a sense, this is the main reason the National Electricity Market was created; to have a market-based and orderly augmentation of capacity as required by the market, and at least cost. Much gas-fired capacity has been added in the last decade, by the main generator-retailers in the market as well as independent power producers.

There is a clear analogy when it comes to applying the carbon constraint. If the electricity price, gas price, coal price and carbon price are all observable, then the market (ie operators of existing plant, and proponents of new capacity) can make the appropriate and least cost decisions about how to generate the next unit of power in the short term and the next large augmentation of capacity for the longer term.

Latest posts by _EcoRon_ (see all)


  1. Alex Heyworth

    Thoroughly agree that, if we are determined to reduce CO2 emissions, a market based price mechanism must be the way to go. However, to achieve Australia’s 2020 emission target is going to be extremely difficult (perhaps impossible) Most commentators underestimate both the pain involved and the hidden obstacles along the way.

    For example, consider this scenario. A coal fired power plant operator has an old 600MW power station that becomes uneconomic due to the rising cost of emission permits, writes down its value to zero and offers it for sale to the highest bidder. New operator (possibly a subsidiary of the original owner) buys it for a token amount, say $20-30m, starts it up again and runs it profitably because the low capital outlay makes it economic to do so. Result – higher power costs (because we have to pay, indirectly, for the permits) but no reduction in CO2. I will be surprised if we don’t see this one play out somewhere along the path.

    Roger Pielke Jr has done a good analysis of the scale of change needed (see his book The Climate Fix, or a later version online at http://citeseerx.ist.psu.edu/viewdoc/download?doi= The task is daunting. Pielke thinks a carbon price alone cannot achieve Australia’s 2020 target in the time frame, and I agree. I expect the government is anticipating buying credits from other countries to offset our emissions.

  2. Picking winners?

    GE is going to build a 400 megawatt thin film solar module plant in the U.S. that will bring its investment in thin film solar to $600 million ( link )

    The NYT reports that:

    The plant, whose location has not been determined, will employ 400 workers and create 600 related jobs, according to G.E. The factory would annually produce solar panels that would generate 400 megawatts of energy, the company said, and would begin manufacturing thin-film photovoltaic panels made of a material called cadmium telluride in 2013. While less efficient than conventional solar panels, thin-film photovoltaics can be produced at a lower cost and have proven attractive to developers and utilities building large-scale power plants….
    “We see our way to much higher efficiencies than that,” he said. “We probably can cut costs 50 percent over the next several years.”

    • See that big orange thing in the middle of Oz.. I think it is known as a “sun drenched desert”.

      Again I ask if the Chinese are aiming for 50GW from solar and now I see the Yanks can manage a baseload capacity by 2013 why can’t we ??