Drivers of photovoltaics cost evolution
Massachusetts Institute of Technology. Engineering Systems Division.
Jessika E. Trancik.
MetadataShow full item record
Photovoltaics (PV) have experienced notable development over the last forty years. PV module costs have declined 20% on average with every doubling of cumulative capacity, while global PV installations have increased at an average rate of 30% per year. However, costs must fall even further if PV is to achieve cost-competitiveness at high penetration levels and in a wide range of locations. Understanding the mechanisms that underlie the past cost evolution of PV can help sustain its pace of improvement in the future. This thesis explores the drivers of and constraints to cost reduction and large-scale deployment of PV. By developing novel conceptual and mathematical models, we address the following questions: (1) What caused PV's cost to fall with time? (2) How may materials constraints influence PV cost and deployment? These questions are addressed in the analyses presented in Chapters 2-4. Chapter 2 assesses the causes of cost reduction observed in PV modules since 1980. We develop a new model that identifies the causes of improvement at the engineering level and links these to higher-level mechanisms such as economies of scale. The methodology advanced can be used to evaluate the causes of improvements in any technology. By developing a model of PV modules, we find that in the early stages of the technology (1980-2001), improvements in the material usage and module conversion efficiency played an important role in reducing module cost. These improvements were mainly driven by research and development (R&D) efforts. As the PV technology matured (2001-2012), economies of scale from larger manufacturing plants resulted in significant gains. Both market-expansion policies and public R&D stimulated cost reduction, with the former contributing the majority of the cost decline from 1980 to 2012. Chapter 3 turns to assessing the materials constraints to PV cost reduction. We ask how fast metals production should be scaled up to match the increasing demand by the PV sector, if installations grow to meet a significant portion of energy demand. Unlike previous studies, which primarily used inherently uncertain factors such as reserves to estimate limits to technology scalability, we use past growth rates of a large set of metals as a benchmark for future growth rates. This analysis shows that thin-film PV technologies such as CIGS and CdTe that employ rare metals would require unprecedented growth rates in metals production even for the most conservative PV growth scenarios. On the other hand, crystalline silicon PV can provide 100% of global electricity without silicon exceeding the historical growth rates observed by all metals in the periodic table. Chapter 4 assesses the risks that material inputs bring to technologies today. This study develops cost-riskiness metrics based on the price behavior of metals along two dimensions: average price and price volatility. We first compare a large set of metals using these cost-riskiness metrics. We observe that metals obtained as byproducts have higher risk than major metals. We then apply these metrics to different PV technologies by treating them as a portfolio of metals. We find that designs such as CIGS and CdTe, which use byproduct metals with high average prices and price volatilities, show signals of cost-riskiness. The approach advanced here can serve as an assessment of the cost-riskiness of technologies introduced by their materials inputs.
Thesis: Ph. D. in Engineering Systems, Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society, 2018.This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.Cataloged from student-submitted PDF version of thesis.Includes bibliographical references (pages 99-109).
DepartmentMassachusetts Institute of Technology. Institute for Data, Systems, and Society.; Massachusetts Institute of Technology. Engineering Systems Division.
Massachusetts Institute of Technology
Institute for Data, Systems, and Society., Engineering Systems Division.