The significant potential for Australia to utilise solar electricity production has long been known. Inland, arid locations in Australia receive high levels of solar radiation, on par or greater than significant locations in terms of solar thermal development, such as California, USA. Despite the favourable solar resource, solar thermal technology has not yet seen commercial development in Australia. The key barrier to commercial development is the high Levelised Cost of Electricity (LCOE) from solar thermal plants, which would need to compete in a market with some of the lowest electricity prices in the world.
Even as economic feasibility of solar thermal technology approaches, via a combination of reduction in costs and increased electricity prices (through premium pricing for renewables and the internalisation of carbon costs for fossil fuel based generation), a number of issues remain unclear. These issues included; the need to distil the vast array of literature on solar thermal technology; lack of local experience; uncertainty in cost estimates; uncertainty in Government policy mechanisms; availability and accuracy of solar data; and the optimal configuration of plants. This study sought to examine these issues and provide answers.
Through a review of literature six solar thermal technologies were identified and described. These included the low temperature solar tower and solar pond systems, and the high temperature (concentrating) parabolic trough, power tower, parabolic dish and compact linear Fresnel systems. Through a multi-criteria selection process, parabolic trough and power tower systems were selected for further analysis with respect to costs and performance.
A review of Government legislation and policy identified a range of initiatives which have promoted the development of solar thermal power in Australia, including Government support of demonstration plants such as the Cloncurry power tower in Queensland. The two government initiatives with relevance to solar thermal projects were the Mandatory Renewable Energy Target, which creates a premium electricity price for renewable generators, and the Solar Flagships program which is set to provide significant grant funding to utility scale solar thermal plants.
A number of sources of solar data were identified which covered a range of locations around Australia. In order to properly understand the data and its accuracy, the measurement and derivation was discussed and the data sets compared. It was found that reasonable agreement was present between data sets with differences typically less than +/- 20%.
Capital costs were developed for parabolic trough and power tower technologies based on published cost estimates and published costs of actual plant. Difficulty was encountered in developing the cost estimates due to the scarcity of cost data and the questionable accuracy of that which was available. The resulting estimate were:
- 50 MW Parabolic Trough plant, 12 hrs thermal storage – 10,980 AUD/kWe
- 13.7 MW Power Tower plant, 16 hrs thermal storage – 15,619 AUD/kWe
Similar difficulty was encountered in developing operations and maintenance (O&M) costs, which limited data being somewhat restrictive. As a result, O&M costs were based on 1% of direct capital costs plus staff costs, per year.
Three Australian sites were chosen to form the basis of a plant performance analysis and configuration optimisation exercise. Longreach, Mildura, and Port Hedland were chosen for their high annual average direct normal insolation, proximity to transmission infrastructure and also to provide a spread of sites around the Australian continent. Solar data for the three sites was collected from the data sets previously identified. The data was compared and a single source selected for use within the analysis, the EnergyPlus data.
Performance modelling of a standardised 50 MW parabolic trough systems with a range of storage sizes (0 – 12 hrs) based on the developed capital and O&M costs produced a real LCOE in the range of 388 AUD/MWh for Mildura to 268 AUD/MWh for Port Hedland, and 290 AUD/MWh for Longreach. As the plants were standardised for each location, it was concluded that the differentiating factor in LCOE was the solar resource for each location, with an inverse relationship between LCOE and annual average direct normal radiation.
Performance modelling of a standardised 13.7 MW power tower systems with a range of storage sizes (0 – 16 hrs) based on the developed capital and O&M costs produced a real LCOE in the range of 523 AUD/MWh for Mildura to 431 AUD/MWh, and 430 Hedland, and 430 AUD/MWh for Longreach. As the plants were standardised for each location, it was concluded that the differentiating factor in LCOE was the solar resource for each location, with an inverse relationship between LCOE and annual average direct normal radiation.
With respect to the performance modelling results, the significance of location was discussed. It was concluded that while annual average direct normal radiation had the strongest influence on annual production and therefore LCOE, the range of annual variation was also influential. A high annual variation decreased annual production and increased LCOE. The modelling revealed that depending on the proportion of marginal storage cost to fixed plant costs, increasing storage may reduce, increase or have little effect on LCOE. In the case of the trough plants where the cost of additional storage was relatively high, it had little effect, while for the power towers where storage was proportionally cheaper, additional storage reduced the LCOE. In addition to this, it was discussed that the addition of storage was also important for Australian plants in order to provide dispatchable generation which could generate at periods of peak demand and price. It was shown that by generating between 6 am and 9 pm, rather than 6 am to 6 pm, the average electricity price obtained could rise by 10 AUD/MWh, based on 2007 QLD electricity market price data.
Finally it was concluded that significant reductions in the costs presented would be required to deliver solar thermal plant, even with large Government funding under the Solar Flagships program.