The cost of solar photovoltaics (PV) electricity in Australia in 2030 is on track to be about A$30 per megawatt-hour (MWh). This conclusion arises from current trends in PV module efficiency and cost.
Importantly, $30/MWh is below the operational cost of most coal and gas fired power stations. Continued installs of rooftop and utility scale solar will likely lead to a wave of retirements of existing coal fired power stations during the 2020s. This price is also competitive with industrial gas heating.
The price of PV modules continues to fall as cumulative shipments expand (Figure 1). Each doubling of cumulative production causes a 23% decrease in module prices. The doubling time is currently about 5 years but is likely to accelerate.
The silicon PV industry has an industry roadmap [International Technology Roadmap for PV] that helps to guide industry development and component standardization.
Like the roadmaps in many other industries, it is partly a self-fulfilling prophecy.
The 11th edition was recently released. Sustained improvements in manufacturing technology are being implemented throughout the 2020s including the growth of larger silicon ingots, diamond wire sawing of ingots into wafers, thinner wafers (using less silicon), larger wafers, higher throughput of cell manufacturing equipment, less silver in the cell conductors, better surface passivation, fewer wafer defects, higher cell yields, lower cell degradation rates and better module materials.
Sophisticated computer control of manufacturing is being widely implemented. Most PV module factories now have annual production capacity above 2 Gigawatts, and most will be above 5 Gigawatts by 2030. Even larger fabs may be built soon.
A major change is the rapid transition from BSF cell technology to PERC cell technology which now has 70% of the global solar market (about 70 Gigawatts in 2019). Another large change is the rapid transition from multi crystalline silicon wafers to single crystalline silicon wafers (which have higher quality).
Together, these changes allow an increase in typical silicon module efficiency from 17% in 2018 to 23% in 2030 (Figure 2). Many modules are bifacial, allowing the harvesting of a few percent more light that enters the module through the rear surface.
Increased module efficiency shows up as a cost reduction across most of the value chain of a solar farm. Apart from the modules, the cost of a solar farm includes transport, land, mounting structures, module mounting and wiring, inverter, grid connection and project costs (approvals, finance, contingencies).
Most of these costs depend on the area of module that needs to be deployed. When module efficiency increases, most of these costs decline in proportion.
A 1% increase in module power (Watts per square metre) has a similar effect to a 2% decrease in module cost ($ per square metre) on PV farm energy cost ($/MWh). The mooted increase in panel efficiency to 23% by 2030 is equivalent to a PV energy cost decrease to three quarters of the current value.
On top of increasing module efficiency, the cost of module manufacturing per square metre is slowly declining due to improving manufacturing technology, process control and scale. These two factors combine to provide confidence that the cost of PV energy in Australia in 2030 will be about two thirds of the cost today, about A$30/MWh.
About 95% of solar modules are based on silicon. Alternatives to silicon PV are much discussed. However, the road to commercialization of perovskites, organic PV, CZTS, PV concentrators, solar thermal and others is tough. Silicon PV has left behind CdS, CdTe, CIGS, a-Si, III-V, dye cells and others over the years.
Silicon PV is a moving target with respect to both efficiency and energy cost. Silicon has high and stable cell efficiencies, low toxicity, infinite raw material supply, incumbency and the bankability accorded by 600 GW of installed capacity to date. An alternative cell material still incurs encapsulation costs and balance-of-systems costs, which together comprise more than 80% of the PV farm value chain.
This means that a new solar cell that is literally free must still have a mass production efficiency above 20% in 2030 (allowing for modularization losses). This efficiency would need to be stable for 30 years with only a slow degradation rate.
Niche applications such as lightweight and flexible modules for buildings are suggested for non-silicon modules. However, silicon modules are also lightweight and flexible if the glass superstrate is deleted. However, the absence of glass greatly reduces module lifetime for any PV device.
Tandem cells offer cell efficiency above 30% when a high bandgap cell is paired with a silicon base cell. The problem is that the upper cell must be very efficient, low cost and stable. Perovskite and III-V cells are interesting possibilities, but further research and development is required.
Access to cheap renewable energy unlocks emission reductions beyond electricity. Taking advantage of low-priced solar PV and wind, the Australian electricity grid may double in size as electrification of energy services proceeds.
Electrification of land transport (via electric vehicles), urban heating (via electric heat pumps) and industrial heating (displacing gas) adds about 40%, 10% and 50% respectively to current demand, leading to a reduction in Australian Greenhouse emissions of 70%.
The resulting generation growth means that the Australian grid may not reach 100% renewables before the mid 2030s because the increasing supply of PV and wind is chasing the rapidly expanding demand for electricity.
The cost of balancing an 80-90% renewable electricity grid is kept low (~$10/MWh) by using strong interstate interconnection (to smooth-out local weather), storage (batteries, pumped hydro, thermal storage), demand management and legacy coal & gas (to cover the occasional wet and windless week in winter). When the legacy coal & gas is finally phased out the balancing cost rises to about $25/MWh.
The steady fall in the cost of PV electricity has major economic implications. $30/MWh is below the operational cost of most coal power stations and many will retire earlier than expected in response.
The use of gas for industrial heat is under threat because $30/MWh is equivalent to $7 per Gigajoule assuming a gas burning efficiency of 85%. Access to cheap energy changes economic drivers. Industrial companies can purchase compact thermal storage such as hot water, hot rocks, molten salt or molten silicon, sufficient to last for 24 hours, and heat them with low-cost daytime solar electricity.
Prices of $30/MWh for PV electricity will be available in many countries in the sunbelt (lower than 35° of latitude). This threatens Australia’s exports of thermal coal and gas. Australia could put its head in the sand and have its Kodak moment around 2030.
Alternatively, Australia can recognize that it is deploying PV and wind at a rate that is 10 times faster than the global average, and leverage this experience to develop different economic models. Our renewable energy advantage can unlock new opportunities in low carbon exports and associated jobs.
Professor Andrew Blakers and Matthew Stocks work at the research School of Electrical, Energy and Materials Engineering at Australian National University