Jeff Masters Weather Blog

Technological learning and policy together can advance clean energy » Yale Climate Connections


Year after year, the costs for climate technologies fall dramatically, continually undercutting forecast prices and accelerating the clean energy transition. Now, with the Inflation Reduction Act, costs are likely to drop even faster – good news for a climate-safe future. Nevertheless, technology forecasts consistently overestimate the future cost of clean energy, potentially leading policymakers astray.

Specifically, decision-makers may fail to support strong climate policies based on their mistaken fear they will lead to higher energy costs for their constituencies. However, by better predicting the rapid cost declines seen in the real world, policymakers can better understand the huge economic opportunities a fast clean transition offers.

So how can we more accurately determine future clean energy costs to chart a safer path forward?

The answer may lie with the economic principle known as “technology learning curves.” Learning curves represent the rate at which the costs of a new technology fall in relation to how much of the technology is produced over time. Put simply, the “learning-by-doing” and economies of scale that result from more production tend to generate rapid cost declines that are hard to predict through other modeling methods.

From ‘Moore’s Law’ to ‘Wright’s Law’

Innovative technologies have followed different learning curves in recent history, with perhaps the most famous example relating to computer chips. This curve, dubbed “Moore’s Law,” accurately predicted that the amount of information a computer chip could contain would double every two years as producers gained more experience with the technology.

Researchers have found that a similar principle holds for technologies that produce and use clean electricity. Producers are learning and reducing costs so quickly that global society could save trillions of dollars by accelerating the transition to a fully clean economy.

For clean energy, researchers have found a learning curve known as “Wright’s Law” may hold the key to understanding clean tech’s cost progression. Under Wright’s Law, named for the engineer Thomas Wright who first discovered it, a technology will advance at a pace that follows an S-shaped curve. In other words, it will first experience slow development, followed by exponential growth, and eventually taper as the market saturates.

A faster clean energy transition could save trillions

Based on past cost declines and deployment rates, just how quickly can we clean up our energy resources? To examine how Wright’s Law could predict cost declines of four vital climate technologies – solar cells, wind turbines, batteries, and clean fuels – an Oxford team extended its observed rates of progress via S-curve forecasts.

By accounting for the past improvement rates for more than 50 clean technologies, the authors were able to determine likely cost declines for the four key technologies. They found that the expected cost declines of these technologies are so significant that they could enable a cost-effective 100% clean energy system within just two decades.

As a result, the key factor to reducing emissions is how quickly we can break down barriers to faster deployment. In doing so, we can speed these improvements and cost declines. The researchers found that the faster the transition, the cheaper it is – to the tune of $26 trillion in savings – as a result of the cost declines that come with technology learning.

Although the authors project an optimistic future, they note that while a fast transition “is aligned with market forces, policies that discourage the use of fossil fuels will likely still be needed to fully decarbonize energy.” In other words, current institutions that stick by traditional energy technologies, like coal and natural gas, could dampen continued clean technology learning and innovation, thus limiting the speed of cost reductions and slowing their deployment.

Cheap, clean electricity enables faster electrification

The principle of learning curves applies not only to energy-producing technologies but also to energy end uses such as cars and appliances. Transitioning all vehicles, buildings, and industrial processes away from burning fossil fuels to running on electricity powered by renewable energy, known as “electrification,” is crucial for cutting greenhouse gas emissions across the entire economy. But how fast might that electrification take? 

In the past, electrification projections have reflected traditional economic assumptions that electricity is expensive relative to fossil fuels. Hewing to those conventional assumptions slows down electricity’s ability to out-compete fossil-fuel-powered technologies. As a result, models have been pessimistic about electrification growth rates.

Not only does that approach negatively affect predictions about the electricity transition, but it also goes hand-in-hand with an overly optimistic view of the need for bioenergy and carbon sequestration, each of which raises major environmental and equity concerns.

To better represent the electrification’s likely future, a research team in Potsdam used learning curves to modify the traditional electricity versus fossil fuel cost assumptions. They also studied the effect that supportive policies could have on electrification.

First, the researchers incorporated more likely projections of the cost of solar, wind, and battery storage by carrying observed learning rates into the future. That approach brought down the cost of electricity. Second, they incorporated potential global climate policies: a hard cap on remaining carbon budgets to achieve climate targets, a constraint on bioenergy and carbon sequestration, and strong and consistent policy incentives to scale electric vehicles.

The results show substantially faster growth in global, economywide electrification is possible. The fastest electrification growth rate occurs with a carbon budget that aligns with a safer, 1.5°C increase in global warming. Here, the policies built into the model drove up the cost of fossil fuels when compared to electricity, with electricity becoming the cheapest way to transport and use energy by 2040.

The researchers’ novel study shows that accelerated climate ambition is possible when it comes to clean energy deployment and electrification. However, it would be a mistake to simply rely on technology costs to fall without adopting appropriate policy incentives. While we’ve seen many clean technologies follow Wright’s Law-like learning curves, such a rate of learning is not guaranteed. Real-world constraints can lead to stalemates, such as barriers designed to support incumbent fossil fuel interests. Technology can improve quickly, but policies will help ensure the promised results of the most promising technologies.

Climate success is possible with accurate cost predictions and strong policy

Clean energy technology has improved faster than ever imagined, but policies often fail to account for its rapid improvement. To set ambitious carbon-free policy, society and its decision-makers must be able to accurately predict future cost declines – a role that observed learning rates can fill. In turn, smart policies can help clean energy stay on its forward trajectory. With learning curves and policy informing one another, society can transition away from polluting fossil fuels fast enough to preserve a livable climate, all the while cutting household energy costs.


James Arnott is Executive Director of the Aspen Global Change Institute, and Michelle Solomon is Policy Analyst with Energy Innovation Policy and Technology LLC®. Both organizations are Yale Climate Connections content sharing partners.



Source link