Assumptions about how often coal generators will run should also be scrutinised carefully. If it is assumed that a plant will run more often, the cost of building the power station is spread over more hours of operation, so the cost per megawatt hour is cheaper. The AI Group points out that utilisation level has a fundamental impact on levelised cost, for example if an ultra-supercritical coal plant generates at an average of 60% of its capacity instead of 80%, the cost rises from $80/MWh to $100/MWh*.
Some of the modelling we have seen on coal generation costs also assumes highly optimistic levels of utilisation. Assumed levels of plant utilisation for a coal-fired power station of around 87%-90% are unrealistic because it does not meet the requirements of the market. The market has a capacity factor of 60% and this will drop further as households install more solar. To meet the market’s needs, i.e. the energy customers consume, with coal, you either have to build more coal generation than you need but run it at only 60% or build just enough coal generation so it can run 90% of the time and then supplement coal with other generation sources.
Either way, running a coal station at 90% does not reflect the cost of meeting the market’s requirements. If the assumption shifts to a more likely capacity utilisation factor, a new coal power station becomes significantly more expensive on a cost per MW hour basis.
In addition to applying optimistic operating assumptions, the modelling also overlooks some important business risks. One critical risk is carbon pricing. Even if a carbon price is not the policy of the current government, there is a reasonable probability that one could be introduced in the decades following the construction of a new generator, so this needs to be factored in to any commercial evaluation.
A further risk arises from building ‘baseload’ generators in a market that increasingly values of flexibility. Because cheap renewables are only available when the wind is blowing, or the sun is shining, flexible solutions like gas peaking plants and battery storage have become more important than ever. Combined, these sources can deliver reliable, low emissions energy at the lowest possible cost. Conversely, coal generators perform best under steady-state operating conditions, and incur much higher operation and maintenance costs when they need to operate flexibly. The extremes in Australia’s climate require greater flexibility of energy, not steady-state.
Ultimately, the real test of credibility for any model or set of assumptions is whether someone believes them enough to put money behind the project. That’s what AGL has done with our plan to replace Liddell.