10:30am - 10:50am
On the nature of innovations affecting photovoltaic system costs
1Massachusetts Institute of Technology, United States of America; 2National Renewable Energy Laboratory, United States of America
Energy technologies such as photovoltaics (PV) modules and wind turbines have been improving and their costs have been declining rapidly in the last five decades. Literature suggests that innovation activity has been one of the most important drivers of this rapid cost reduction. Previous studies have focused on identifying the effects of innovations on a technology’s unit cost at an aggregate level as represented by R&D investments and patenting. However, they do not offer a framework for connecting specific innovations, such as diamond wire sawing for cutting ingots into wafers, to technology costs. Identifying the links between innovations and cost is critical to characterize past sources of improvement and to inform future innovation efforts. In this work, we ask: Through which channels have specific innovations influenced PV systems costs in the past, and what type of innovations were most influential?
We conduct the first study to characterize in-depth the innovations and other factors affecting PV technology costs. To do this, we develop a framework that maps out the relationships between innovations and technology cost determinants to uncover how specific innovations influence costs. We apply this framework to identify the innovations that affected the cost of photovoltaic (PV) systems in the last five decades. We first build a conceptual model for distinguishing innovations from other factors that affect cost. We then compile a large set of innovations based on a literature review and feedback from experts. We connect innovations and other drivers to the variables determining PV system costs, in order to identify the channels through which PV system costs changed. Finally we develop a typology of innovations to investigate which innovation types have been more prevalent. The typology classifies each innovation according to how it changed the variables in the cost model, including improvement processes such as material quality changes, process development, and automation.
We investigate innovations at both module and balance-of systems (BOS) levels and highlight the differences in the nature of innovations across modules and BOS. Unlike modules, which are mass-produced and standardized goods, BOS components (e.g. mounting systems) are often customized to a specific site.
Our results show that there is a diversity of innovations at both the module and BOS levels. We find that hardware innovations have been prevalent across PV modules and BOS since 1980. Module innovations mainly addressed materials challenges through developing tools and processes. BOS innovations, in contrast, focused on reducing part counts and complexity. Innovations aimed at improving soft variables such as task durations have been on the rise more recently, including online permitting tools, but these innovations haven’t been adopted widely enough to influence costs. Our results point to a need for continued efforts to improve soft variables as well as hard variables in the future. The method developed in this work to identify and connect innovations to cost determinants can be applied to other technologies.
10:50am - 11:10am
Design for Recycling Guidelines for Photovoltaic Modules
National Renewable Energy Laboratory, United States of America
The global growth of photovoltaic (PV) capacity has a parallel growth of PV waste at its end-of-life (EOL) that brings both challenges and opportunities. By 2050, there is an estimated stock of 78 million MT of raw materials that will become available as PV systems reach their EOL (IRENA reference). The challenge lies in that PV modules themselves are often stubbornly resistant to recycling efforts and there is concern that the projected increasing growth of PV installations could become functionally constrained by availability of raw materials, despite ongoing dematerialization efforts. Particularly for silicon PV (Si-PV), the most commonly deployed style of module, several factors make recycling challenging, not the least of which being the design of the module itself. In response, this study endeavors to identify potential design changes that could improve the recyclability of these PV modules at their EOL via a literature review of Design for Recycling (DfR) best practices as used in other industries. As such, this study endeavors to identify if DfR practices that could be adopted today in order to better mitigate tomorrow’s PV resource scarcity.
DfR involves technical considerations in choice of materials used in a product, as well as the capacity for liberation of components, subcomponents, and associated materials during a product’s respective recycling process (e.g., design for disassembly). DfR strategies are a function of both the product and materials in question, as well as the nature of available recycling processes. The challenge is to modify the design of the product to allow for recyclability without compromising the functionality, capacity, and commercial viability of the product. Importantly, DfR should not exist in a vacuum, requiring additional case by case considerations of market conditions, legislative backdrops, and geographic distribution of the EOL products; not every jurisdiction will be subject to the same rules, nor will they all have access to the same recycling technologies.
When considered in the specific context of Si-PV, some module design considerations are likely to better facilitate recycling than others. For example, the predominant use of ethylene vinyl acetate (EVA) laminate dictates that some form of delamination is required during recycling. Although multiple methods exist for achieving this, it is generally considered a challenging process regardless of whether it undergoes either a physical or pyrolysis treatment. From a DfR perspective, any module designs that managed to reduce, eliminate, or substitute the laminate with a more easily treated alternative would be considered favorable provided the resulting design exhibits a sufficiently acceptable operational performance. In another case, the DfR priorities differ considerably as a function of recycling processes. In processes where module polymers undergo pyrolysis, there is an advantage to avoiding fluorinated materials and associated offgassing, where it would not be as critical for a purely physical separation shredding process.
These findings represent only a portion of an ongoing evolving investigation. As such, the presentation will include a more comprehensive list of DfR guidelines and strategies from both an overall and PV-specific perspective.
11:10am - 11:30am
Using Big Data to Understand the Variability of Carbon and Energy Footprints of Pulp and Paper Products
North Carolina State University, United States of America
The pulp and paper industry produces a variety of products essential to everyday life. Many life cycle assessment (LCA) studies have been performed for different pulp and paper products to understand their environmental impacts. However, the life cycle inventory data in previous studies are either industry-average or from a few specific mills. The variations across different pulp and paper products, technologies, process configurations, and mills are usually not considered due to the lack of process data.
In this presentation, we will discuss two data-driven approaches that have been developed to address the knowledge gap in understanding the distributions of energy use and Greenhouse Gas (GHG) emissions across diverse pulp and paper products. The first approach is database integration. Facility-level GHG emission data collected from publically available databases were integrated with mill-level production data from private sources. A case study including 165 mills in the United States was conducted. The statistical variances of GHG emissions from five major types of pulp and paper productions were analyzed, including packaging, market pulp, printing and writing, tissue and towel, and specialty paper products. The results can be used as data references for LCA, footprint accounting, and paper-related analysis that commonly needs ranges and distributions of data for sensitivity analysis and Monte Carlo simulation.
The second approach is a bottom-up analysis using mill-by-mill process-based data. Mill data such as energy consumption, fuel sources, chemical usage, and wood sources were collected for most U.S. mills from private data sources. GHG emission factors were collected for fuel combustion (i.e., fossil fuel and bio-based fuels such as black liquor in mills) and upstream production of wood and chemicals. The total energy consumption and GHG emissions in Scope 1, 2, and 3, were estimated and analyzed for major pulp and paper products. Biogenic and anthropogenic carbon were separately tracked for each category of paper products and the differences across all products are shown. Comparative scenarios combined with contribution analysis between different products such as coated versus uncoated paper, recycled paperboard versus virgin paper products were developed to understand the differences of environmental footprints in products that may have similar functions but are produced from different processes/technologies.
Both case studies used large datasets to understand the variations of energy and GHG footprints of different pulp and paper industry in the United States. The differences between the two are the methods and data sources. The results of two case studies will be compared and discussed to highlight the insights provided by the analyses. Although this study focuses on the pulp and paper industry, the methods can be applied to many other manufacturing industries; especially those that have large variations in manufacturing processes and products.
1. Alec Nabinger, Kristen Tomberlin, Richard Venditti, Yuan Yao*, Using a Data-Driven Approach to Unveil Greenhouse Gas Emission Intensities of Different Pulp and Paper Products, 26th CIRP Life Cycle Engineering (LCE) Conference Paper, Procedia CIRP (accepted).
11:30am - 11:50am
Cutting CO2 emissions from U.S. steel consumption 70% by 2050
1Department of Mechanical Engineering, University of Michigan, USA; 2School for Environment and Sustainability, University of Michigan, USA
Climate change mitigation strategies must include the steel industry, as steel production alone contributes 25% of industry greenhouse gas emissions. The United States is the world’s second largest steel consumer, and with plateauing per capita stocks is a potential template for future global steel consumption; other researchers predict plateauing global per capita steel stocks by mid-to-late-century. Therefore, emissions reductions strategies for the U.S. will likely apply globally in the decades to come. This research determines feasible technological, trade, and societal pathways that lead to a 70% cut in CO2 emissions attributable to U.S. steel consumption by 2050 relative to 2010 levels. This reduction is within the range of the International Panel on Climate Change’s reduction requirements to ‘likely’ stay below 2 degrees Celsius. Pathways analyzed include strategies to reduce U.S. steel consumption and strategies to reduce emissions from liquid steel production, which are where the bulk of steel sector CO2 emissions are released. The technological variables include alternative low carbon primary and secondary production as well as improvements to manufacturing material efficiency. Trade and foreign steel production emissions intensities are important variables affecting consumption based emissions and are also included; the U.S. currently imports over 40% of its steel consumption and exports over 30% of its scrap. The societal variables included are population growth, per capita steel stocks, product lifespans, and end-of-life recycling rates.
We use a dynamic material flow analysis of the U.S. steel sector from 1900 to 2050 to understand how our societal changes affect the demand for steel and production of usable scrap. In combination with the technology and trade scenarios, these values, determine how the steel demand is supplied (i.e., whether the steel is primary or secondary and what technologies are used in production) enabling the calculation of total sector CO2 emissions. The results of this analysis provide CO2 reduction pathways to inform sector mitigation polices and illustrate the trade-offs between changing consumption patterns, technology changes and shifts in trade policy.