Can Lego save the world? This is one idea that stuck out in reading How big things are done, a new book by Bent Flyvbjerg and Dan Gardner. Flyvbjerg is perhaps the world’s leading authority on mega-project failures – or how big things get done, but woefully late and at woefully cost – and so he makes it unlikely he’d be an optimist.
For decades, Flyvbjerg, a professor of management at the University of Oxford, has compiled a database of large projects from high-speed railways to hosting the Olympic Games. His findings were so bleak that he proposed the “iron law of megaprojects”: these are outpaced by time and budget, time and time again. Even worse, there is a long tail to these disappointments. A significant minority of megaprojects are not only late and costly, but disastrous as well.
Despite this bleak evidence, he and Gardner have proven that we can work miracles if we instead use a principle more familiar from Lego sets. This principle is modularity: a complex Lego model is assembled from a limited set of bricks, each of which is precisely made and interchangeable with other bricks.
Modularization has a number of advantages. The first is that individual components can be manufactured on a large scale, which quickly reduces costs. In the 1930s, an American aeronautical engineer named T.B. Wright made a careful study of aircraft factories. He concluded that the more often a particular model of aircraft is assembled, the faster and cheaper the next one will be.
Workers learned the best ways of working, and special tools would be developed to help with certain tasks. Wright found that the second plane was usually 15 percent cheaper than the first. The fourth plane will be 15 percent cheaper than the second, and the eighth plane will be 15 percent cheaper again. Every time backlog production doubles, unit costs fall by 15 percent. Wright called this phenomenon the “learning curve”.
The researchers later found learning curves in more than 50 products from transistors to beer. Sometimes the learning curve is shallow and sometimes it’s steep, but it always seems to be there. Because modular projects frequently use the same plans and structures, they mock the learning curve for efficiency.
There are other advantages to modular projects. They’re more likely to be able to use factory-made components, and when you build complex things in factories, you’re less inclined to fancy the unexpected when you make them on a construction site – especially if that construction site is deep underground or offshore.
By their very nature, modular construction projects are likely to be able to continue even when there is a problem with one element of the structure. This helps explain why, in Flyvbjerg’s database, benchmark projects are immune to the most dramatic “black swan” cost overruns, which are always a risk to other large projects.
These are the typical joys. Now turning to the problem of climate change, an interesting pattern emerges. Low carbon energy projects include some of the most modular and least modular designs in Flyvbjerg’s database. Solar and wind power are at the normative end, while nuclear and hydropower are at the opposite pole. Perhaps it’s no wonder, then, that prices for solar and wind projects are falling rapidly.
I have no objection in principle to nuclear power, but I do wonder if clean and safe nuclear power can be made affordably, unless nuclear plants are able to switch to a much smaller and more modular design. Nuclear power plants have been supplying power to the grid since the mid-1950s, but they don’t seem to be getting much cheaper, perhaps because we haven’t been able to replicate the same designs often enough to climb up the learning curve. I keep reading news stories about companies having big plans for small reactors, so maybe it’s not impossible.
However, the contrast with solar energy is striking. Silicon photovoltaics began providing practical power around the same time: The US satellite Vanguard 1 was the first to use them, carrying six solar panels into orbit in 1958. (The sun always shines in space, what else would you use it for?) Turn on a satellite Millions of dollars?)
At the time those solar panels produced half a watt which was undoubtedly a painfully high cost. By the mid-1970s, solar panels had dropped to $100 a watt, or $10,000 for enough panels to power a light bulb. By 2021, the cost will be less than 27 cents a watt.
This is the learning curve in action. The learning curve for photovoltaics has been estimated to be about 20 percent steeper than that of aircraft. Chris Goodall, author of a keynotes that the world produced a hundred times more solar cells between 2010 and 2016 than it did in all the decades before 2010. Batteries—an important standard complement to solar photovoltaics—are also racing down a steep learning curve.
There is a similar story to be told about wind power. The wind turbine is made of standard components, and the wind farm is made of standard turbines. The price of wind power has also fallen faster than most proponents could have dreamed of two or three decades ago.
I’m no nuclear power expert, but I’m sure modular reactors should be possible. I hope that. We need big things to happen in our ability to generate clean energy. And the best way to progress is to start with small, repeatable blocks.
Tim Harford’s new book is “How to make the world add“
Continued FTMag On Twitter to catch our latest stories first