Events & Media

Fighting Climate Change: Which Technologies, and at What Costs?
Bren graduate Joe Bergesen integrates life-cycle assessment thinking and Learning Curve theory to predict the long-term economic and environmental performance of technologies designed to mitigate climate change

Joe Bergesen

The coming decades will undoubtedly witness the continued arrival and evolution of technologies intended to mitigate climate change. As a result, policy-makers and companies will have to determine which technologies offer the best investment opportunity in terms of both cost and effectiveness.

The environmental benefits of technologies we can use to combat climate change are clear, but sometimes unintended tradeoffs can negate the benefits of even the greenest technologies. The most widely known example is perhaps that of electric vehicles (EVs). When running on electricity generated by coal-fired power plants, which emit greenhouse gases (GHG), the reduction in overall emissions from EV cars are offset by the GHGs emitted by the power plant in generating electricity to fuel the cars.

Going forward, therefore, it will be important for decision-makers to have all the relevant facts relating to the economic costs of a technology over the long run, how well it will do its intended job of reducing emissions, and what if any unforeseen “offsets” associated with a given technology might call into question the wisdom of pursuing it on a large scale. Research by Bren School alumnus Joe Bergesen (PhD 2015, MESM 2011) for his doctoral dissertation may be applied to those kinds of decisions.

Bergesen recently had a scholarly article published based on one of three chapters that made up his doctoral dissertation. “A framework for technological learning in the supply chain: A case study on CdTe [cadmium-tellurium] photovoltaics” appeared Feb. 27 in the journal Applied Energy. In the article, Bergesen and his faculty advisor and co-author, Bren associate professor Sangwon Suh, incorporate learning-curve theory in a model intended to estimate the potential changes in both the costs and the environmental performance over time of technologies designed to mitigate climate change. Bergesen tested his model on photovoltaic (PV) cells to generate solar power.

At its most basic, the “technological learning curve” concept refers to the fact that products and processes improve over time as their “cumulative production” increases. Said another way by Bergesen, “The more you make of a technology, the better you learn to make it.”

Technological learning curves, Bergesen notes, have long been used to understand how cumulative production improves a technology and reduces its cost, the result of many small innovations that coincide with production over time. For instance, he explains, a constantly increasing number of Gigawatts has been generated by all the PV cells that have been manufactured and installed since the technology became available. Learning-curve theory holds that as that number increases (as more PV cells are produced), the cost of the technology decreases. “The total-production figure can be used to predict the rate of learning [as reflected in decreasing costs], but it doesn’t show the underlying innovations and changes that are causing those costs to drop.”

One of Bergesen and Suh’s hypotheses is that if learning is occurring for one technology, it probably is also happening for every technology in the economy, since any given technology depends on many others in the “upstream” supply chain. A small piece of the cost of a given technology, therefore, is at least partially the result of cheaper components used to make it. Or, as Bergesen refers colloquially to these “upstream” causes, “It takes a thing to make a thing that’s used to make another thing.”

His point is that the upstream learning that occurs in many products or areas of the economy also affects the price of the final technology we are interested in, such as a PV panel. That economy-wide learning potentially contributes to the price of each of those many products — and any products made from them — to drop over time. “If all processes throughout the economy are learning, then that shared learning may be contributing to what we’re seeing for PV technologies,” Bergesen says.

While the rate at which a technology’s cost falls is important to companies and policy-makers trying to make sound investment decisions, it is only one important metric in Bergesen’s learning-curve approach to technology. For his dissertation, Bergesen developed a method to forecast how the environmental, in addition to the purely economic, performance of technologies designed to mitigate climate change will evolve as they are deployed at an unprecedented scale over the next thirty to forty years. It involves introducing life-cycle assessment (LCA) thinking into the learning curve of a technology.

For the PV example, Bergesen explains that improvements in metals mining and semiconductor materials manufacturing that resulted in reduced environmental impacts would reduce the environmental impacts of PVs, too, which depend on them. That environmental component of the upstream supply chain would not have been considered previously in applying learning-curve theory on a traditional economic basis. By integrating LCA, Bergesen has introduced a more holistic approach to technology-based learning curves.

“If you’re looking at various learning curves in a technology’s supply chain, some of the underlying changes shown by learning will affect the environmental impact of technologies more than others,” Bergesen says in identifying a nuance of his approach. “So for instance, if labor becomes more efficient, it will lower the cost of the product but won’t really affect the environmental impact of a technology. But if you were able to use less energy and fewer materials to make a better PV panel, as we are seeing over time with photovoltaics, then that would affect the life-cycle environmental impacts; you’d probably see reduced impacts as learning occurred.

“What we’re getting at is that environmental impacts ‘learn’ at a rate different than costs, and we’re coming up with a mathematical framework for analyzing that,” he continues. “We did a simplified case study for photovoltaics, but even with our limited data, we could start to see the trend of reduced environmental impacts and reduced costs.”

A number of low-carbon electricity-generating devices, such as some PV solar technologies, are still in the early stages of their development and have a long way to go in terms of technology and costs, Bergesen explains. The model he has proposed can help, first, to better predict costs over time as prices change for materials, such as rare metals, and can also let us see how other technologies in the economy might improve at the same time.

“If we as a society decide to invest in photovoltaics as one approach to mitigating climate change, we might see innovations in semiconductor technology that would also be useful in, say, computing, thus reducing the prices and, potentially, the environmental impacts of other technologies,” Bergesen notes. “That is referred to as ‘spillover,’ and accounting for it and the broader economic and environmental benefits of investments in technologies used to mitigate climate change could strengthen the case for developing renewable energy in the coming decades.”