A new paper published in Science raises the prospect of solar power meeting a substantial fraction of the world’s energy needs within decades, forecasting at least 3 terawatts (TW) of cumulative installations of photovoltaic generation (PV) by the year 2030.
The dramatic growth rates in PV in recent decades have blitzed forecasts, but the paper – produced by a team led by Nancy Haegel from Berkeley’s National Renewable Energy Laboratory (NREL) – outlines a collection of looming hurdles that must be overcome for growth to be sustainable. These political, technical, infrastructure and economic challenges might limit solar’s ability to expand.
According to the researchers, the global growth rate of PV outstripped that of total demand and that of all non-hydro renewables, with its capacity growing “by a factor of ~57, from 4 to 227 GW [gigawatts]” between 2000 and 2015.
The expansion of PV is driven by highly regionalised pockets of policy encouraging rapid expansion in manufacturing and deployment. Germany’s ‘Renewable Energy Law’, for instance, is responsible for an order-of-magnitude expansion, from 3 to 30 GW, between 2007 and 2012. The ‘Investment Tax Credit’ in the United States and Japan’s generous feed-in-tariffs (FiTs) have drive similarly dramatic changes.
The steep acceleration in this global industry tends to make something of a mockery of over-cautious forecasts – the International Energy Agency is singled out in the paper as a repeat offender, with a chart showing pessimistic expectations dramatically failing to match the sharp incline of historical deployment.
Rapid change can come with sudden challenges. Japan’s 11 GW of growth saw grid constraints alongside an increase in surcharges, and policy had to be tweaked to reduce the FiT. California now mandates storage installed in parallel with future renewable energy capacity. In addition to the subtle hazards of over-eager policy mechanisms, the paper identifies the need for serious cost reductions, if solar PV stands a chance of becoming a terawatt-scale technology.
The levers of cost reduction
The ‘experience curve’ is a simple idea: the more experience an industry has producing a technology, the lower costs become.
Using the historical ‘experience curve’, assuming the industry follows historical patterns, we can expect that the deployment of 8 terawatts of solar PV will result in a module cost of approximately $0.25 per watt.
Haegel’s team also analyse the changes needed to bring about big reductions in the life-cycle costs of energy from solar PV, also known as ‘levelised cost of electricity’ (LCoE). As opposed to the cost of the machine, measured in dollars per unit capacity ($/W), LCoE is measured in dollars per unit of energy delivered ($/kWh).
Increased efficiency, lower module prices and decreased degradation rates are all plausible pathways for silicon panels to reach an LCoE of around $0.03/kWh unsubsidised cost.
Using different materials to manufacture panels is another potential cost-reduction lever. It’s a crowded field with diverse possibilities.
Silicon now competes with cadmium telluride (CdTe), copper indium gallium diselenide (CIGS) and concentrating PV modules with multi-junction III-V cells. III-V materials such as gallium arsenide (GaAs) are clambering up efficiency curves as research expands into a forest of competing materials. Perovskites show the most promise with regards to scalable manufacturing, but questions around efficiency remain, with more research required.
Technological pathways to reduced manufacturing costs, above and beyond assumptions around ‘experience curves’, are key to breaking terawatt-scale barriers. Even if these cost and efficiency targets are met, a further set of grid integration challenges remain.
The storage question
The task of plugging a large quantity of variable-output renewable energy into the grid currently dominates the coverage of energy in Australia. The research paper looks past the politics and discusses sound technical questions that must be addressed, if solar PV is to reach the terawatt scale globally.
Implementing technologies that shift demand, increased interconnection, flexible conventional generation and improved solar-resource forecasting tools are all key components of integrating solar resource into existing grids. Replacing the service of ‘inertia’, essentially the stored energy of big rotating masses in the electricity system, with a synthetic equivalent service provided by rapid-response power electronics is a looming challenge.
The authors are optimistic about storage – they forecast a stored-electricity price of $0.025/kWh by 2030, with the potential for electric vehicles to service many terawatt-hours of storage alongside industrial and domestic battery systems. A blend of improvements in grid management and a switch to fully dispatchable renewable energy will vastly improve the chances for solar PV.
Hitting the target
Business as usual would see the PV industry hit 3 TW of cumulative PV installations by 2030, according to the authors. The pathways for cost reductions, efficiency improvements and grid integration would see that curve bend sharply upwards to reach somewhere between 5 and 10 TW by 2030. These challenges are a complex, multi-faceted mix of moving pieces but, as the paper highlights, they are surmountable.