Quantifying the cost savings of global solar photovoltaic supply chains

John Paul Helveston, Gang He*, & Michael Davidson

November 10, 2022

Speed and scale

10 \(\times\) today’s level by 2050

Source: IEA (2021)

Cost is key for renewables to compete

Source: Lazard LCOE 15.0

PV is getting cheap

Between 2010 - 2020, global levelized cost of energy (LCOE) of utility-scale solar PV fell by 85%

Partly thanks to the global supply chain dominated by China

China is manufacturing leader in almost every sector

Technology Scale
Solar Panels From 1% to 78% (2001 - 2021)
Wind Turbines 1/3 of global supply (2020)
Electric Vehicles 51% of global sales (2021)
Lithium-ion Batteries 70% of global production (76% by 2025)
Nuclear Reactors From 45 to 88 plants by 2030

China’s “gift to the world”

China comprises ~78% of global PV production in 2021


Response:

  • US and EU tariffs on imported Chinese PV panels
  • June 2022: Biden Administration invokes the Defense Production Act to accelerate US PV manufacturing

Diversifying supply chain, domestic manufacuring

Rationale:

  • Localizing benefits in terms of growth, employment, and trade surpluses
  • Supply chain resilience to disruptions

Approach:

  • Restricting the free flow of products, capital, talent, and innovation
  • Tariffs, trade barriers, industrial regulations/policy

Research Question

  • What are the costs of national versus global supply chains?
  • What does it mean for achieving climate goals?

Learning curve model

Source: Yelle (1979)

Learning curve model

In context of solar PV:

  • X: Cumulative installed cap.
  • Y: Price per kW


Log transformation:

\[\ln Y = \ln a + b \ln X\] Learning rate:

\[lr = 1 - 2^{b}\]

Source: Grübler, Nakićenović, and Victor (1999)

Two-factor learning curve model:


\[\ln p_{it} = \ln \alpha_i + \beta_i \ln q_{t} + \gamma_i \ln s_{t} + \varepsilon_{it}\]

price ($ / kW) = intercept + installed capacity + silicon price

for country i and year t

Learning rate:

\[L_i = 1 - 2^{\beta_i}\]

Data Sources

Country Data Source
Global Installed PV capacity and prices International Renewable Energy Agency (IRENA)
U.S. Installed capacity Solar Energy Industries Association (SEIA)
U.S. Module prices Lawrence Berkeley National Laboratory (LBNL) & National Renewable Energy Laboratory (NREL)
China Installed capacity & module prices Energy Research Institute (ERI) & China Photovoltaic Industry Association
Germany Installed capacity IRENA
Germany Module prices Fraunhofer ISE50

Note: All prices are in $2020 USD; inflation adjustments from IMF, exchange rates from Federal Reserve Bank.

“National Markets” Counterfactual Scenario

Assumption: learning-related price decreases in country i in year t are derived from incrementally more nationally-installed PV capacity

\[q_t - q_{t-1} = \lambda_t(q_{i,t} - q_{i,t-1}) + (1 - \lambda_t) (q_{i,t} - q_{i,t-1}) + (1 - \lambda_t) (q_{j,t} - q_{j,t - 1})\]

\((q_{i,t} - q_{i,t-1})\): Amount installed in country i

\((q_{j,t} - q_{j,t-1})\): Amount installed in all other countries

\[q_t - q_{t-1} = (q_{i,t} - q_{i,t-1}) + (1 - \lambda_t) (q_{j,t} - q_{j,t - 1})\]

Source: Helveston, He, and Davidson (2022)

Global Markets vs. National Markets

Global markets

\(\lambda_t = 0\)

Capacity from all countries

\[(q_{i,t} - q_{i,t-1}) + (q_{j,t} - q_{j,t - 1})\]

National markets

\(\lambda_t = 1\)

Capacity only from country i

\[(q_{i,t} - q_{i,t-1})\]

\(\lambda_t\) -> 1 over 10-year period

Source: Helveston, He, and Davidson (2022)

Model results


United States China Germany
Est. (Std. Err.) Est. (Std. Err.) Est. (Std. Err.)
(Intercept) 15 (1.04)*** 18 (1.58)*** 12 (0.96)***
log(cum_capacity_kw) -0.44 (0.045)*** -0.57 (0.070)*** -0.33 (0.042)***
log(price_si) 0.15 (0.058)* 0.23 (0.079) 0.21 (0.054)

Asterisks indicate the level of significance: *5%; **1%; ***0.1%.

A National-market Strategy Would Increase the Costs

Higher prices in 2020:

  • 54% higher in China ($387 versus $250 per kW)
  • 83% in higher Germany ($652 versus $357 per kW)
  • 107% higher in the U.S. ($877 versus $424 per kW)

Source: Helveston, He, and Davidson (2022)

Total Savings by Global Supply Chain: $67 billion ($50 - $84 billion)

Source: Helveston, He, and Davidson (2022)

Future projections

Two future projection scenarios out to 2030

Country 2030 Target (GW) Implied CAGR
U.S. 295 12%
China 750 12%
Germany 103 7%
World 2,115 11%

Sustainable Development (SD)

Country 2030 Target (GW) Implied CAGR
U.S. 411 16%
China 1,106 17%
Germany 147 11%
World 3,125 16%

(Sustainable Development Scenario in the 2020 IEA World Energy Outlook)

Source: Helveston, He, and Davidson (2022)

Higher prices in 2030

~20% higher in each country

  • China: $162 versus $135 per kW
  • Germany: $298 versus $251 per kW
  • U.S.: $320 versus $262 per kW

Sustainable Development (SD)

~25-30% higher in each country

  • China: $136 versus $108 per kW
  • Germany: $276 versus $221 per kW
  • U.S.: $287 versus $221 per kW


For comparison, NREL’s 2021 Annual Technology Baseline report predicts $170, $190, and $320 / kW by 2030 in advanced, moderate, and conservative improvement scenarios.

Source: Helveston, He, and Davidson (2022)

Future Savings

Source: Helveston, He, and Davidson (2022)

Take home messages

  • Globalized supply chain has saved solar installers in the U.S.($24B), Germany($7B), and China ($36B), total $67B from 2008 to 2020
    • China’s mass production accelerates the cost decline
    • China also benefits from the global supply chains
  • Solar prices will be 20-30% higher in 2030 if countries move to produce domestically
    • Confronting climate change relies on international collaboration

Open science

https://www.nature.com/articles/s41586-022-05316-6

Data and code

https://github.com/jhelvy/solar-learning-2021

Sensitivity analysis app

https://jhelvy.shinyapps.io/solar-learning-2021/

More about the paper

https://drganghe.github.io/posts/2022-10-nature-cost-savings-of-global-solar-pv-value-chains/

Contacts

John Helveston
George Washington University

https://www.jhelvy.com

Gang He
Stony Brook University

https://www.ganghe.net

Michael Davidson
University of California, San Diego

https://mdavidson.org

Back up slides

No transition time: historical prices

Source: Helveston, He, and Davidson (2022)

No transition time: historical savings

Source: Helveston, He, and Davidson (2022)

No transition time: future prices

Source: Helveston, He, and Davidson (2022)

No transition time: future savings

Source: Helveston, He, and Davidson (2022)

Drivers of learning

Source: Kavlak, McNerney, and Trancik (2018)

Source of learning

Learning by doing

  • Economy of scale
  • Economy of scope

Learning by researching

Source: Zheng and Kammen (2014), Meng et al. (2021), Zhang et al. (2022)

Historical silicon prices

Source: Helveston, He, and Davidson (2022)

Relation between \(\lambda\) and p

\(\lambda\): share of additional capacity from domestic production
p: share of total capacity from domestic production

Source: Helveston, He, and Davidson (2022)

Alt model 1: cummulative national module production capacity

Source: Helveston, He, and Davidson (2022)

Alt model 2: cumulative national installed capacity

Source: Helveston, He, and Davidson (2022)

Alt model 3: global weighted average plant size

Source: Helveston, He, and Davidson (2022)

References

Grübler, Arnulf, Nebojša Nakićenović, and David G Victor. 1999. “Dynamics of Energy Technologies and Global Change.” Energy Policy 27 (5): 247–80. https://doi.org/10.1016/S0301-4215(98)00067-6.
Helveston, John, Gang He, and Michael Davidson. 2022. “Quantifying the Cost Savings of Global Solar Photovoltaic Supply Chains.” Nature. https://doi.org/10.1038/s41586-022-05316-6.
IEA. 2021. “Net Zero by 2050.” International Energy Agency.
Kavlak, Goksin, James McNerney, and Jessika E. Trancik. 2018. “Evaluating the Causes of Cost Reduction in Photovoltaic Modules.” Energy Policy 123 (December): 700–710. https://doi.org/10.1016/j.enpol.2018.08.015.
Meng, Jing, Rupert Way, Elena Verdolini, and Laura Diaz Anadon. 2021. “Comparing Expert Elicitation and Model-Based Probabilistic Technology Cost Forecasts for the Energy Transition.” Proceedings of the National Academy of Sciences 118 (27).
Yelle, Louis E. 1979. “The Learning Curve: Historical Review and Comprehensive Survey.” Decision Sciences 10 (2): 302–28. https://doi.org/10.1111/j.1540-5915.1979.tb00026.x.
Zhang, Chao, Liqin Xie, Yueming Qiu, and Shuangtong Wang. 2022. “Learning-by-Manufacturing and Learning-by-Operating Mechanisms Drive Energy Conservation and Emission Reduction in China’s Coal Power Industry.” Resources, Conservation and Recycling 186 (November): 106532. https://doi.org/10.1016/j.resconrec.2022.106532.
Zheng, Cheng, and Daniel M. Kammen. 2014. “An Innovation-Focused Roadmap for a Sustainable Global Photovoltaic Industry.” Energy Policy 67: 159–69. https://doi.org/10.1016/j.enpol.2013.12.006.

https://drganghe.github.io