In 1844, Samuel Morse sent the first message over the telegraph, marking the biggest leap in communication technology since Gutenberg added movable type to the printing press. Telegraphs were electrical signals transmitted over wires, which meant they could only be sent between stations where wires had been laid—and then only if the stations used the same system. The telegraph was transformative, eagerly embraced, and very slow to spread. Stringing the wires took an enormous investment of time and money, but once messages could be sent instantly across large distances, telegraphs became essential. By 1861, there was a transcontinental telegraph system in the United States. By 1866, people were sending telegraphs across the Atlantic.

The path the telegraph and its supporting infrastructure took is a pattern that appears in almost all developing infrastructure says Paul Edwards, a professor in the Department of History and the School of Information. Electricity, canals, and the radio are also examples. The pattern interests Edwards, particularly when it’s combined with climate data, computer models, and global warming.

“One of the more interesting things about calling something an infrastructure,” Edwards says, “is that it has become this kind of invisible background to everyday life. Electric power is a great example. What would you do if it just disappeared tomorrow? Our whole society would collapse.”

Eighty percent of the power plants that were retired last year in the United States burned coal. The majority of these were retired either because of environmental regulations or because they had become unprofitable.

A world without electricity is hard to conceive, but having it comes with a cost: carbon. In the United States, we generate almost 70 percent of our electricity by burning fossil fuels, and we generate carbon dioxide alongside it. Everyone agrees we need to transition away from fossil fuels, but there’s very little agreement about how. Edwards sees it as an infrastructure problem, and he uses two different types of models to address it.

Model Behavior

The first model Edwards uses is historical, and it lays out the steps through which infrastructure develops. Infrastructure starts with innovation, followed by a long, slow period when a new technology begins to be adopted by a few people and starts to spread. The technology is still expensive and awkward at this stage, and it doesn’t connect to existing systems or similar technologies very well. When it moves into the take-off phase, the technology becomes widely adopted and the infrastructure required is quickly built. It peaks when nearly everyone who can afford the new technology has it, at which point infrastructure historians consider the infrastructure built out. In general, physical infrastructures take between 30 and 100 years to complete this process. Communication infrastructure has sometimes been faster, taking between 30 and 50 years.

The second model Edwards uses is a climate model in which computers run complex equations that simulate interactions between mathematical representations of, say, ice or oceans. These models demonstrate how current climate systems work, and they illustrate how the future climate might respond to changes, such as a massive reduction in carbon dioxide or a sudden increase in methane.

Edwards says these climate models are among the most complicated ever made. “Some are more than a million lines of computer code.” The trick, of course, is getting the models to resemble the real climate. How do you know if you have the right answer, or that you got the right answer for the right reasons?

There are many different groups of people in the world doing this type of modeling, and about 25 important ones, Edwards says. Each time the Intergovernmental Panel on Climate Change releases a report with new data, these modeling groups run a set of standard experiments and compare their results.

With these climate models, scientists can examine theoretical scenarios and the futures they would produce. One is particularly salient: If we let the existing fossil fuel infrastructure run until its power plants wear out and replace each dilapidated coal or natural-gas power plant with one powered by renewables, how would that affect global warming?

A 2015 report released by the U.S. Department of Energy said that wind energy prices had hit new lows, making them economically viable in several locations across the United States.

The answer got people’s attention because results suggested that this alone might keep the planet within the two-degree Celsius range scientists say is the allowable boundary. And here Edwards returns to history: “This is where the historical model becomes useful.”

A Bundle of [Renewable] Energy

If you map the historical model to renewable energy’s growth, you can see the same developing-infrastructure pattern. “Solar is in the middle of the take-off curve,” says Edwards. “You can see it clearly. It’s the same with wind energy. Those two in particular are really well along and beginning to grow faster than any other energy source right now, even with the steep drop in oil prices and coal’s collapse.” But despite the renewable energy surge and the modeled benefits of phasing out older power plants, Edwards says we’re still replacing fossil-fuel power plants with newer fossil-fuel power plants. “This is one of the things about infrastructure: Once it’s in place it’s hard to veer away from it,” he says. “Developing countries that haven’t inherited this infrastructure can leapfrog over that obstacle entirely.

“However,” he continues, “we already have an electric power grid renewables can use, which might help them to accelerate. It’s easier to add them than to build a new coal-powered plant, and the 1,000 mile-long wires needed to connect it to everything.”

Infrastructures are not discrete systems. They have different histories and champions, as well as financial, political, and cultural influences. But Edwards says we might be at a tipping point with regard to energy. “History shows there are ways to accelerate changes to an infrastructure, and this might be a moment when a lot can happen very fast.”

This article is part of a larger environmental series in honor of Earth Day. U-M kicked off the first nationwide Earth Day in 1970 by hosting a teach-in that drew more than 50,000 people. Also in this series: