One aspect of my discussion with Bent Flyvbjerg this week dealt with the odd variance between transmission projects and high-speed rail projects in his data set. Flyvbjerg (Linkedin, Twitter), is a global megaproject expert and consultant and co-author of How Big Things Get Done: The Surprising Factors That Determine the Fate of Every Project, from Home Renovations to Space Exploration and Everything In Between. The book includes an analysis that everyone engaged in big clean technology projects should look at.
For over two decades, Flyvbjerg and his team have been collecting data on megaprojects globally, and now have cleansed information on about 16,000 of them spread across 25 categories. The variance in likelihood of projects in these categories to deliver results on time and on budget is stark. The very good news is that wind, solar, and transmission projects are among the top-performing projects. These are key technologies for climate action. The bad news is that so is fossil fuel generation.
But there are three project categories that are interesting for this discussion: transmission, rail, and road. We had limited time together, so didn’t get deeply into all three, but as the data shows, road and transmission projects do well, and rail projects don’t.
I was perplexed by this for a couple of reasons. Part of my background includes having had Canadian National Railway (CNR) as a client, something I wrote about in a distribution sector magazine. At the time, in 2011, it had the the highest revenue to expenses ratio in the industry globally. It managed this despite the unusual situation where it owned and maintained its own rail tracks instead of having them be a nationalized infrastructure asset, something which boggled the mind of freight rail companies globally. A lot of this had to do with the operational obsessiveness of Hunter Harrison, the then CEO of CNR. People I worked with at the company told me stories about yard operations types getting emails or phone calls at 3 AM from Harrison as he obsessively trawled through the key performance indicators (KPIs) and operational stats in their internal systems and calling on them to get blockages fixed.
One of the key projects I launched for the company was establishing an iterative and incremental program to deliver its next-generation, map-based, operational status, asset, and KPI visualization and exploration system. It was running hundreds or thousands of locomotives and tens of thousands of cars across tens of thousands of kilometers of track in 10 Canadian provinces and 18 US states. I spent a bunch of time learning about freight rail. Oh, and the software project was a success too.
At the same time, Canada’s fading passenger rail service, VIA Rail, was another client. I was working with it to re-architect and modernize its reservation system, which was the typical legacy crud that had accrued components over time and that was limping along on dying hardware, which is fairly standard for IT systems. I spent a bunch of time learning about passenger rail as well.
I’ve spent time assessing the global state of high-speed rail, comparing China’s now 40,000 km of high-speed, grid-tied, electrified rail with the USA’s one, occasionally kind of high speed run, Morocco’s high-speed rail from Tangier on the Mediterranean down to Casablanca, the patchwork of gauges in the 19,000 km of high-speed rail in Europe, and the challenges facing high-speed rail in India.
And, of course, I’m an urban transit wonk. I wrote my first analysis of Toronto’s transit system — an essay in a University of Toronto urban planning course — about 30 years ago, and have delightedly taken urban light and heavy rail in cities across the world. I’ve been on subways underground on more continents than most people have visited.
I’m not a trainspotter, but it’s safe to say I’m a bit of a rail nerd. And, obviously, I’m a transmission nerd as well.
So I was perplexed about the failures of rail, especially high-speed rail, compared to transmission and roads. They are all linear assets. They all run mostly along the surface of the earth. They all feature highly modularized, repeatable construction approaches. There’s lots of ability to improve processes as more kilometers of track are laid.
Reading Flyvbjerg’s book, it’s easy to be left with the impression that high-speed rail projects are all disasters in the making. He includes case studies of two of them, the Hong Kong to Pearl River Delta project he was called to consult on when it was undergoing yet another convulsion as it fell further behind time and schedule, and the California high-speed rail project, in which it’s going to be declared a victory in perhaps a decade when two towns that few outside of California have ever heard of that are each over 100 miles (160 km) from major population centers are connected. Tons of wasted money. Vastly over schedule. In the case of California’s solution, approaching none of the promised benefits when complete.
And as his data shows, that’s par for the course.
And yet. And yet. China has built 40,000 km of high-speed rail in the past 15 years including linking in other countries including Vietnam and Laos, and also extending freight rail through eastern Asia, although not all high-speed. And it’s building dozens of railways in countries around the world under its Belt and Road Initiative, including in multiple African countries, the ‘stans in Asia and Indonesia.
And China was seeing massive usage of its high-speed rail pre-COVID-19, with 2.3 billion passenger trips in 2019, many around Chinese New Year (gong xi fa cai to my Mandarin-speaking friends around the world, and kung hei fat choi to my Cantonese speaking ones). Those enormous passenger volumes are part of the reason my global projection of aviation demand through 2100 is lower than most analysts’, as the fastest growing market in the world is diverting enormous numbers of passenger trips to ground rail (while traditional markets in the west are relatively stagnant, as I discussed with a global aviation forecasting expert).
Benefits are clearly occurring. Could Chinese high-speed rail test Flyvbjerg’s thesis to destruction?
Probably not. I had the opportunity to speak with Flyvbjerg because he and his co-author stumbled across my assessment of a dozen years of the natural experiment of wind, solar, and nuclear deployments in China. They extended the data a bit and included it in chapter 9, “What’s your Lego?” They were delighted to find it for two reasons, per my discussion with Flyvbjerg. The first is that it so clearly shows that when a bunch of western excuses for slowness of nuclear build-out are controlled for, the highly modularized, parallelizable, repeatable wind and solar programs deliver a lot more actual generation of electricity.
The second reason was that it was quality data from China. And there’s the rub. Flyvbjerg has visited China numerous times and attempted to get data on Chinese megaprojects multiple times. It is, after all, building more megaprojects than the rest of the world combined, so it’s incredibly central to his academic and professional focus.
But while China runs on quality data internally, and undoubtedly knows to the renminbi and hour how much and at what cost the rail projects were completed, it doesn’t share that data with the rest of the world. A common refrain from western analysts, including me, is that it’s trivial to get data about everything in the always-broadcasting USA, pretty easy for Europe, and very hard for China, and not just because of the language barrier. Some of this is simple strategy. As Sun Tzu advises, “Appear weak when you are strong, and strong when you are weak.” Telling the world details about its internals is a degree of transparency that the world would love, but there’s little reason for China to divulge the details.
So we don’t know how much China’s high-speed rail lines cost, whether they were over budget or if they completed on time. Flyvbjerg pointed out the major high-speed rail crashes, but in retrospect those are operational issues, not construction issues. The Wenzhou collision in 2011 dampened Chinese citizen confidence in high-speed rail safety and hence ridership for a bit, but the 2019 results show that that was temporary. And given the 1.4 billion Chinese citizens and billions of passenger miles, the accumulated 41 deaths and 170 injured passengers I can see from public data is tragic at a personal level, but vastly safer than driving per passenger mile and at minimum approaching commercial aviation’s stellar and hard-won statistics, while not contributing nearly as much to climate change.
World Bank data from 2014 showed that China’s highest-speed rail, greater than 300 kph, cost $17-21 million per kilometer even with a high-ratio of tunnels and viaducts, while France’s equivalent is at $25-39 million and California’s project was at $52 million per kilometer. But this was aggregated data, not project data, and so not useful for inclusion in Flyvbjerg’s data set and analysis.
Given that purchasing power parity — a comparison of purchasing the same basket of goods and services in local currencies between countries — in China was the same as in the United States in 2013, as Graham Allison points out in the one good chapter in his book, Destined for War, and now is higher, Chinese renminbi go a lot further than US dollars domestically by comparison. (If someone gives you Destined for War, try hard to ignore the title, read the first chapter and throw away the book.) That means Chinese can build a lot more things for the same money, and there’s little reason to believe that this isn’t true for high-speed rail.
The gray rhino of COVID-19 also significantly dampened benefits from the high-speed rail, with most internal travel prohibited to contain the spread of the disease. It’s unclear what the implications of that are beyond an obvious revenue shortfall, but as the rail system is nationalized in China and the debt is in their internal currency, it’s a very manageable problem for them fiscally. China will not become bankrupt due to it (or anything else) despite Sinophobes’ constant assertions to the contrary.
Flyvbjerg did differentiate rail from transmission and roads in a specific way that has strong merit. Rail requires a lot more tunnels, and as the data shows, any time you are digging tunnels, fat-tailed risks abound and projects operate more slowly. You can wind roads around hills and have steeper grades on them than for rail. That’s even more true for transmission, which can ignore a bunch of cliffs entirely, putting towers at the top and bottom and having the wires go up and up and up. As a result, when you come to hills or mountains while building rail, there’s a much greater tendency to go through them rather than around them. And so, costs and durations typically explode.
But we also talked about subways, another place where western costs tend to explode. I believe it was in Alain Bertaud’s (former chief urban planner for the World Bank) book Order Without Design that he calculated that the cost of Manhattan’s subway platforms and rail lines was roughly the same per square meter as the ground costs above it, and so he considered it economically reasonable. However, most western cities quail at the estimates that come back for subway extensions, even before their budgets explode in construction (the unsurprising reality for Toronto’s latest, otherwise pretty well thought out, extension).
And there Madrid in Spain shines. Its subway is a rare beast of a megaproject in that it delivered on time and on budget, in both 4-year incremental build-out programs, and is seeing the promised ridership (unlike Toronto’s subway extension I assessed three decades ago which is still not seeing promised rider volumes after 20 years in operation). The heuristics that Madrid applied to deliver dozens of big stations and almost 300 km of track are strongly aligned with Flybjerg’s guidance, and it’s showcased in his book. Well worth reading, as is the very positive case study of Terminal 5 at Heathrow. Both address the team and motivational issues that Irish firm Vision Consulting, an infrastructure project optimization consulting organization focused on communication and commitment-based management, pointed out in a 2021 article.
It is possible to do big things, even in the 21st Century. At least, it is in some places, and when the builders think slow and act fast.
Don't want to miss a cleantech story? Sign up for daily news updates from CleanTechnica on email. Or follow us on Google News!
Have a tip for CleanTechnica, want to advertise, or want to suggest a guest for our CleanTech Talk podcast? Contact us here.