With Laurent Segoland from London and Gerard Reed from Berlin.

This is Redefining Energy.

Today on Redefine Energy, part our summer series.

We re air all the episodes, but this one is fascinating because it's an interview by Michael Barnard on our sister show, Redefining Energy Tech with the great Professor Ben Fleiberg, the author of How Big Things Get Done.

And Lauren was read that book last year or there before.

Yeah, amazing, absolutely amazing.

So listen to Michael Bernard interviewing Professor Ben Fleidberg.

Welcome back to Redefining Energy Tech with your host Michael Barnard.

Hi, welcome back to Redefining Energy Tech.

I'm your host, Michael Barnard.

This is a repeat of a discussion I had with Professor Ben Flobier, author of twenty twenty three's top selling business book, How Big Things Get Done.

Listen in for their discussion of nuclear modularity, scaling and mega projects.

Thank you very much, Thank you for having me.

Now the excuse for having you because I gladly spend hours and hours and hours talking to the about a bunch of topics that sadly we don't have time for is the publication of your upcoming book with Dan Gardner, and so you know that book is How Big Things Get Done, and it explores your decades of research and experience in dealing with megaprojects, their failures, what works, and what doesn't.

And today we're going to focus on some clean tech aspects out of that.

But before we get into that, how did you end up being the person that is the go to person for governments globally when megaprojects need to be started effectively, when portfolios and megaprojects need to get moving, or when they just go off the literally in some cases.

So, like you mentioned, I'm an economic geographer, and economic geography is a special branch of the economics that focus on what we call spacial economics.

So you look at how how do how do economics play out in space?

And that could be urban space or regional space, on national space, so you basically can look at what's going on on the surface of the planets in economic terms.

And that's what I was raised on, you know, as a student, undergraduate, graduate PhD post of everything, And at one stage I started noticing that the citis and regions were being built in larger and larger chunks.

So the projects that make up a city, to take that city as an example, just got larger and larger.

And I noticed this early on and I identified it as a mega trend.

You know, I said, okay, this is going to be the future.

It's actually not planning that is so important as such, you know, which you talk about planning as lendings planning.

It's much more what are the specific projects that a city decides to do and how they're delivered.

That's what's going to be defining cities and actually regions and ultimately nations in the future.

This is a long time ago.

I saw this just by observing, you know, what was happening on the ground, including in my home country, which is Denmark, but also other countries.

And I started looking around and found I couldn't find any data on this, even though already at that stage trillions and trillions of dollars were spent on these big projects, and nobody who was collecting data.

You couldn't get data on them.

So as a scholar, you know, this is an ideal of situation.

That's There's that why the area on the map nobody has startied it before, let me get in there and study it.

So that's how it started.

Why was there in irritated?

It strikes me as you know, it's an obvious insight, but why do you think that was?

We've been building big stuff for thousands of years.

Why haven't people been collecting even good anecdotal data.

For For several reasons.

One is that it's very hard to connect this data.

I didn't know this at the beginning, but I got educated quick about how difficult it is to get high quality data.

There's so much crap data out there, and in a big project, there are so many versions of data.

So just to choose the right version of the data, which would be the version that has validity and reliability from a scholarly point of view, is a major job in its own right.

So that's one thing.

And that meant that they were only like, there might be some data on individual projects or a handful of projects, but there was nothing that would allow statistical analyscence, you know, for which you need a large example.

And so it wasn't really on the ratar of scholars either.

And practitioners are so in the here and now, they're very huge oriented.

They're always thinking about the next project.

Nobody thinks once you deliver a project.

Actually, it's very very rare to have anybody stop and think, Okay, what can we learn from what we are what we just did?

You know, why don't we connect some data on the outcomes and then see, you know, how the outcomes compared to the plans.

This is almost never done.

So this this is a combination of reasons, you know why, data that just were not available until we started collecting data.

And it actually took us five years to collect the first data set of just two hundred and fifty eight projects, which was the largest data set in the world, you know, when we polish this in two thousand and two, you know, so this is just over twenty years ago now and today it's ludicrous to think that two hundred and fifty eight projects could be the largest data set the world.

Now, just my group, you know, might see we have a data set that has now grown to over sixteen thousand projects, you know, which still is the largest database in the in the world.

But that's how we started, and two hundred and fifty eight was actually enough to start doing statistical analysis and come up with some valid and reliable answers to questions like how much does it actually cost to deliver a big project, what's the likely cost over on the big project, how long does it take, what's the likely schedule over on?

What are the benefits generated?

And is it the benefits that were plans or do we get more benefits or less benefits than we're playing.

So these were the kind of questions we could start rigorously analyzing after we got those first two hundred and fifty eight projects.

Let's just take an example, like we connected over nuclear energy, and we'll get into that in a bit, but let's just take the history of nuclear generation buildouts.

You know, I've looked at that, You've looked at that.

Take Ontario, where I've lived many times.

That place they had a provincial strategy and the richest largest province in Canada equivalent to Baden.

It's not Rickenstein, but that big the biggest province in Germany.

And they took twenty billion dollars off the books and shoved into general accounting at one point, and you know, fifteen years ago or so, they pushed it back into the utilities books.

But it's now visible and they're trying to pay it down.

But I look at France, and you know, if we take Macron, when he was in charge of the organization that should know this, has admitted that he could never figure out how much they actually spend because as national strategies or major political strategies, money kind of flows in in odd ways in different accounting systems.

So how do you reconcile some of those challenges in a way that was academically sufficiently of quality.

So we just look at that, and you're right, there are myriad ways that you can hide costs, take them off the books, or you can include them in the books obbitrarity, and that's a huge problem.

So basically, we started this in detail for each and every project and see if we can make sense of me.

Sometimes we can't, then we don't include the project.

We have clear criteria that that data have to be valid and reliable, So if there's something dodgy that we can't explain, we will simply leave out that project because we're very, very teenly aware that we don't want to produde our database.

Which is the problem with most data that consultants are using is that it's garbage in, garbage out that they are so they're so hungry for just having data that they will accept data even if the quality of the data might be a dubious quality of dubious status, and they include it anyway, And in that way you get a database that is not you know, really worth it.

And I decided from the very start that we were not going to do that because I felt, you know, we're going to be spending a lot of time on this and that would be completely meaningless if we didn't know the validity and the liability about data, So we would not include anything that had unacceptable validity and reliability.

That doesn't mean that you know, the data are perfect.

There there's no such thing as perfect data.

On big projects like like we're talking about here, there would always be things that you know could be better and that are not entirely clear.

But still you know the day it would have to be above as thrashold where you could say, hey, this this data actually contains a really worthwhile information.

It tells you know, the truth about what the costs were here, and it couldn't if it couldn't claim that, it wouldn't include it all the schedule, all the bit of its.

Yeah, and it's interesting what you get up with because, as you say, sixteen thousand data points and now you can do statistical analysis.

One of the things I thought was most fascinating in the book was, you know, you not only looked at costs and bud and schedule overruns, but you also looked at benefit of cruels and you actually did a fundamental piece of analysis to disprove a theory about unlocking creativity and doing stuff, even though it appears to be a bad idea.

That insight was.

I mean, zero point five percent of projects, if memory serves, actually come in on time, on a budget or close to it, and actually deliver benefits point five percent.

Can you unpack that a bit and unpack that argument that you countered.

Yeah, So the point five percent we call the iron law projects, and the iron law goes like this, over budgets, over time, on the benefits, over and over again.

That's the Iron law.

And about forty eight percent of projects projects come in on budget are better.

About eight percent of projects come in on budgets and on time or better.

And then you know the famous share point five percent.

Those are the projects that come in on budget, on time and deliver the promised benefits or better.

So that's how little it is.

Half a percent.

There's a there's a diagram in the book.

And the interesting thing is that that half percent, you know, when we had to show that in the diagram, it's invisible because you need to show the one hundred percent, you need to show the forty eight percent, you need to show the eight percent, and then the half percent just becomes invisible.

And that's pretty much a good metaphor, you know, for how things are in real life, that the products that tick all the boxes regarding success, which is on budget, on time, and on benefits, are better.

I'visible.

That's how few they are.

I want to emphasize that it doesn't mean that they don't exist, especially when you started thousands of projects.

Even half a percent will give you like a couple of handfuls.

And that's what's really interesting to me is how on earth to the people who are doing these projects, how do they beat the odds?

You know, because the share point five percent, that's your base rate, that's your objective base rate.

When you look at these products like going to the casino, if you play the if you play the game of building a big project, your odds are shirru point five percent that you will win it.

By winning, you define it as you know, being on budget, on time, and on benefits.

So those are pretty steep odds, you know, and if you are going to go in and claim I can beat those odds, you better have something and I want to know what that's something is.

And we spend a lot of time that in our research, and when we wrote the book, you know, picked the brains and started the data of people who actually beat the odds.

Just as you would if you were in a casino, as you were interested in winning a casino.

There's somebody walking around the casino and they beat the roulette, they beat blackjack, and so on consistently.

You want to know what's going on there.

And believe me, the casinos have detectives that are on this immediately and cameras that are on people.

They'll be on this immediately and study what's going on here in a can see in a casino, it's usually fraud, you know, somebody's gaming the system of counting cards or whatever, and they will be expelled from the casino very quickly.

So we were like those detectives and those cameras, we were watching all the people doing projects and finding who are the ones that actually are winning here, And we didn't kick them out.

We actually sought them out and said, can we please talk, can we please look at your data, can we please look at your documentation, and so on, and then we figured out what is it that they're doing that they are successful at this.

It's an interesting question speak institutional investors.

Occasionally they reach out for, like you know, somewhat often contrarian perspectives on the decarbonization future.

But one of the things that occurs to me is you've got a set of seventeen or twenty five categories of projects, and only five of those are in the sweet spot of usually delivering on time, on budget, and on you know, in delivering benefits.

And you know, that's an easy list.

I mean, transmission is in there, for example, and but you know then we kind of going together stuff Nuclear is way down the bottom of the list.

You know, what are some interesting examples of things that you'll usually go wrong or usually go right in your perspective.

This is for me, this is the most interesting result in the book, you know, And it's the first time we publish these, and we look at twenty five different projects.

We present the data for those twenty five different projects, and we find, like you say, four or five projects type do very well and all the risks are doing rather poor to different degrees to be shure and be show different degrees.

And so some of the successes are as you mentioned, transmission, and the most successful is actually solar power, you know, solar energy, that's the most successful.

After that wind wind energy.

So now we have solar wind transmissions.

Wow, that's interesting because that's exactly what we need in order to solve the climate crisis.

Right, So I just I went through but I saw this, I said, WHOA, what a result?

How lucky?

Can I mean?

It might have been the opposite, you know, and at the opposite and which you also ask for what are the examples are really underperforming projects?

You know, one of the worst is nuclear power, only only you know, surpassed in worseness by nuclear storage, you know, so nuclear is way at the other end.

So basically in nuclear and solar at opposite ends of the scale.

Yeah, but lots of other stuff is at the bad end of the scales.

Who like the Olympic Games is one project type we study that's that's has terrible performance consistently it's not so hard to understand.

I mean you always do it.

In a new country.

We talk about something called the eternal beginner's syndrome.

You know that that it's always beginners who are doing this, so you don't get accumulated experience for the Olympic Games.

Hydroelectric power same things.

So big dams, whether they are hydroelectric or not.

You know, big dams are really difficult to build and they also have bad performance airports, high speed rail, lots of other project types.

So there it's a bad and at the other end we have these good performing, well performing types.

And the secret to the difference is we also uncover that is actually fat jales, so thin tailed projects.

Projects that have thin tail distributions for their performance are doing well.

So they have more or less like a normal distribution of course and coursed overrun and scheduled, scheduled oberon and benefits, and that means that they're managing.

So they are normal in that sings.

They it is statistically right.

The other projects that are going bad have fat chains.

So what Ne's in talent calls next ones, and they are some by some people, they are considered outliers.

We don't consider them outliers because they're actually part of the distribution.

It's not like some some error, statistical error.

It's actually real performance, real projects who performed like that, So they are part of the distribution, and they are extreme like you could have on IT projects, which is the worst project type.

Regarding outliers.

You know, if you look at the fat tail, the average in the fat tail is more than four hundred percent cost overron.

That means that there are products with much higher cost over runs than the four hundred because that's an average, right, So in IT to find cost over runs of four or five six hundred percent is not unusual, and that's a completely different board game should be in, you know, compared to the Living Solar, where you never find that kind of cost of run, at least not so far in the data that we have seen.

Yeah, it's interesting because I prefer the Michelle Wooker's gray Rhino metaphor completely understand why, you know, for a more Western origed book, which I think your book is.

You know, black Swan has become the dominant metaphor for risk since Talent published, you know, around the same time as the subprime mortgage crisis.

My experience with the book is that people have used it as an excuse to say we couldn't have possibly known about these things, as opposed to say, we ignored these things that were pretty obvious if we thought about them, and we didn't build a resilient system that would accommodate fat tail risks knowing that they're present.

And so, you know, Wooker's metaphor, you you redefine, you define the way people should think about black swans in your book, and I recommend people read it obviously, but I would say that you're actually redefine it much more like Walker's gray rhino metaphor, which oddly it's much more dominant in Asia than it is in the way She's She's had massively more penetration there.

It's been discussed at the highest levels of the Chinese pollup Borrow, for example, in terms of guiding their strategies.

You can kind of see some of those long termism behaviors and the way they're structuring out stuff versus you know, some of the short termism that often democracies indulging around electoral cycles.

We're seeing those kind of implications.

So have you actually read a worker's book and have many examples or contrasts there that you you know, to to.

Pull that apart.

So the way I understand it is that the gray rhinos are less extreme than the lex ones, so that they are they are extreme, but they are not as extreme.

Actually, our definition, we don't have to call it lex ones or gray rhinos or anything.

Our definition is actually statistically.

You know, statisticians have a very clear definition of what extreme values are.

So there's something called extreme valued theory and statistics.

And you define an outlier.

So if it's tossed over run, for instance, would you would say it's a The technical definition is that it's one point five interquartile ranges about the third quartile.

And that's you know, that's objective.

That's just the ways that it is.

This and they don't care what you call it, whether you call it gray or black or whatever it is.

That's what's called an extreme value, an extreme value theory.

As I understand, gray rhinos they come in, they're not necessarily that extreme, but they're still extreme enough to create lots of problems, you know, in practical policy and planning.

And I would say the way we do it, we consider the whole range you're familiar with reference stars forecasting.

I've seen in what you're writing, and if you read about the way we do that, you will know that we look at the total distribution, the total probability distribution, or whatever variable it is that we are forecasting.

And I find that gives me really peace of mind because I know there's nothing I don't look at, you know, So whether it's gray or black or what it is, it's being considered.

And we do pay special attention to the most extreme values because we know that these are the most damaging.

If you end up in the tail, you are in trouble, and you are not only in a little trouble, you're in big trouble.

And it's very hard to, you know, protect yourself against through contingencies because the contingencies just have to be too high.

You will never get those kinds of contingencies.

It's not even rational within an organization to allocate that much contingency to anyone project because it means you can't use the resources for other projects.

Right, So what we recommend is that when that like eighty percent of the probability distribution, you can protect yourself with contingencies, and we calculate the exact size of those contingencies, and then we say the remaining part of the probability distribution, you really go you got to go in and mitigate so that doesn't happen.

We call that black sworn management.

And then then there's slogan of blacksworan management is cut the tail.

You need to cut the tail.

And how do you cut the tail?

You cut the tail through mitigation.

So let's take a really simple example like Fukushima, you know, and you need to protect the nuclear power plant against a tsunami wave.

Simple you build a wall that's tall enough and then that's it.

You know, that's cutting the tails.

Or move it uphill because that it uphill, you know, that's even easier, even easy.

They build the plant downhill from a tsunami high level marker from the seventeenth century.

So we go through an exercise like this, and of course most of the time it's not a tsunami wall that is the solution.

I'm just mentioning that because it's so easy to understand, or building in a further in country or up you know, at a at a higher level and with different projects and project types.

You know, mitigation means different things, but you can always sit down and think, how do I cut the chain.

It's the most it's the most valuable exercise you can go through when you're delivering a project.

It's very exciting and stimulating too when you get so this is something you do like a whole team of people who are experts in what you are, what you're talking about here, and they just sit down and for whatever time it takes, goes through like how do we get that tail off?

How do we cut that tale?

Are we confident that the tail is now cut?

And come up with all the different issues that could help cut the team.

Well, it's interesting because then I know you worked with Daniel Canneman, and I totally envy the two different academics who have had the opportunity to work with Cannaman.

It just must be such a delightful person.

And that's where I first ran across your work.

In Cannonan's Thinking Fast and Slow and Cantman's Thinking Fast and Slow, he talks about the precursor kind of thinking behind reference class forecasting, but he also articulates the use of pre mortems as a mechanism for doing that.

And I kind of was reading your book and saying, oh, he's not mentioning pre mortems.

The reference class forecasting is excellent because it gives you a statistical view of the actual risk and the variability of the fat tales for your project class.

But the pre mortem allows you to cut the tail.

But you don't use that term.

You call it cutting your tail, which is probably more evocative.

Anyway, it's a process.

Yeah, I think that I'm I seem to remember that we do mention the pre morgim somewhere so but anyway, whether we do or not, I do think pre morgons are very important methodology, very legitimate methodology.

And yes, Canaman has been extremely helpful for my thinking.

He's been very generous, you know, and giving me feedback and so on, and that has been very helpful obviously when you're using his ideas and that really works.

And you probably saw his latest book with co authors called Noise and now, Yeah, he's using a more general term in that book and in previous publications after thinking fast and show cause it.

It's decision hygiene.

So I need a certain hygiene when you're making decisions, and typically humans just skip thatts we have no decision higene, We have this thing called availability bier.

So we just jumped straight to the thing that just pops into our minds, you know, and then we start working on that and we feel that we've got going.

We just we have such a birch.

I don't know what it is is, but we seem to be hardwired just to get going with things and not to sit back and think and reflect over But what is it that we are doing here and think it through before we get going.

So we have this availability bias that tends to get us going, and that the bypasses the decision hygiene, and decision hygiene is the solution to that problem.

And decision hygiene is pre Morton explains, and so pre mortg is an example of decision hygiene, and the reference das forecasting is another example of decision hygene.

And there are all That is.

One of the most funny I've written about this in the past because I read Canam and I dealt with cognitive scientists like John Cook who's the PhD cognitive scientists behind Skeptical Science the Timate Change myth debunking site, and Stephen Lewandowski is PhD advisor.

The availability bias that always amuses me is that so many thought leaders around decarbonization are North Americans, and they tend to be affluent and live in suburbs and have solar panels on their roots, and so their availability bias leads them to assume that rooftop solar can be a more impactful solution than it is.

Now, do you live in Copenhagen itself?

I live in Copenhagen itselfias in Oxford and Copenhagen.

It's a bunch of five story apartment buildings where the population density is much more like the rest of the world than like North American suburbs.

And so you know, I live in much higher density downtown Vancouver.

I every once in a while at count I think it's nineteen eighteen story plus condo buildings are visible from my windows in my home office.

Right, So we have I think thirty six hundred people on my block, for example, And this is much more like an Asian pattern of living.

It's much more like the population density of Paris.

But the availability bias of people who like distribute energy, they see detached homes, so they seem a lot more people in the world living detached homes.

Just one of those interesting things.

And that's the way we are.

I mean, like like Canamans is in his book, like his life has been spent on studying these biases, but he's getting shipped up by them all the time anyway, you know, And I think we just have to realize this.

That's no excuse for letting them run rampant, obviously, quite the opposite.

It means that we really have to have this decision.

I see.

And if we don't do that, we're not doing our jobs.

Yeah.

I spent a lot of time figuring out what the denominators are and then you know, figuring out scale up.

But so let's go back to a specific example.

We talked about the the projects which the project classes like wind and solar, which tend to be very you know, get on time, get on budget, and achieve benefits versus nuclear.

And the reason you know, you and Dan Gardner your cook originally reached out to me was because I published on the natural experiment of those three technologies in China over a dozen years.

I've been looking at that since twenty fourteen.

You included someone like you, You extended it in time and you know, transform some of the stuff because it made more sense.

But what do you talk about what you observe from the experience in China and other places about why wind and solar are so successful and nuclear ism.

Yeah, and first let me say we couldn't believe our luck when we came across that part of your work, because that's that natural experiment that China is exactly what we needed in order to, you know, get a decision on whether nuclear works, basically, because people usually say that China would be the place where it would work if it works, you know, and you just had the data to show that so and obviously your data showed that it doesn't work even in China if by working we mean being scaled up to the level that it needs to be scaled up to if it's going to be a major fix or the climate crisis.

And you compare directly to wind and solar, and you show that they scale up much faster.

Okay, So we explain why in the book.

We say, okay, why why is that the case?

That's that's kind of strange, you know, on first side, why would it be so different?

But it relates back to what I said earlier, that nuclear is fat tailed and the wind and solar are thingsade, so they're much more managedble and they scale much better, much faster, much more easily than nuclear.

Then we went deeper and we looked at what's called learning curves, which basically our courage that show, you know, the more you do a thing, you know, how much easier does it get.

And by easier you often say cheaper, it's gets cheaper and faster.

So and this is our normal experience as human beings.

Just think if you're doing some handiwork around your house.

You start on a Saturday morning, you need to get something done.

Everything feels wrong, you know, the tools in your hand and whatever you're doing, but you get going and you get some experience with what it is that is your your task that day, and even within that kind of a short period of time, you get better at what you're doing.

Now as you expand that and think about doing it day after day after day, you do something that our general experience is that it becomes easier to do something the more times we do it.

It becomes cheaper to do something the more times we do it.

It becomes faster to do something the more times we do it.

Now, that's exactly what happens has happened with solar and wind.

You know that we have become much much better.

It used to take much longer even to put up a single wind turbine that it does now.

You know, you put up turbine, say you know, one a day now and you build.

So this is actually the first time I've ever seen multi billion dollar projects being built in less than a year.

No, those are wind files.

That's unusual.

I would say usually you would have at least at fast makeup project.

It's four or five years.

That's fast.

And here you have in the most difficult waters on the planet.

This is also wind I'm talking about that is being built in the Irish Sea and the North Sea, which are both part of the North Atlantic, which is like part of the world as you you don't want to be out in the winter, like basically you can't do anything from October till April March April.

And nevertheless, you know now in those seven months that they have to do things, they will actually put up a multibillion dollar wind farm in that period of time.

That is totally off the chart.

It's being done.

And this is through this positive learning that I'm talking about getting better by doing it over and over and over again, and sola even more so because it's even more modular than wind.

I usually call it legos.

That's why the last chapter in our book is called what's your Lego?

That's what you need to do when you're doing things.

You need to have a lego in order to be to be effected.

And you know, a wind turbine is four pieces of lego.

There's a foundation, there's a tower, there's a nat shell that's the turbine itself, and then there are the blade or the wings that it's called sometimes also that click onto the cell, and that's that's your win turbine.

So it's like click, click, click, and there it is.

And Okay, those are big pieces of lego, so there are some difficulty in getting them out to see and there might be a lot of wind when they're raising them.

So they have a lot of strong tools and so on, powerful tools, big tools to get these things done.

But basically that's the way it's done.

And it's because it's done like lego, click click, click, over and over that it's so efficient and that we get these economies of scale and economies of learning.

So that's wind and solar at the other end back to nuclear.

It turns out when you start looking at the learning cursion nuclear, it's the exact opposite.

And this is weird.

This is actually something that should not be happening, but it is happening.

The more you do nuclear, the more difficult it gets.

That's called negative learning.

So you do want nuclear plants, and you find out this is actually more then we thought.

For instance, you know, getting the safety requirements right is much more difficult than we had estimated.

So now that we're doing an estimate for the next nuclear plant is actually going to be more expensive than the first one because now we have learned these things about safety, and we build that into the next estimate, and then we try again that's the second one.

So you build the second nuclear plant at higher court takes longer because it's more difficult, and so on.

What happens.

Fukushima happens.

You know, you have or three mile Island or Chernobuild.

You know, you have an accident.

All of a sudden, the rules and regulations change and they are being rammed up.

You get much higher safety standards that you have to live up to.

So now all of a sudden, the standards have gone up and you have to live up to those So now when you then the next nuclear power plant, the third one, it gets even more expensive because you have to live up to these standards.

Right.

That's negative learning.

That's the more you do it, the more expensive it gets.

And that is what has happened with nuclear over the past several decades, except maybe in one place, and that is Korea, but all the countries for which we have data, it looks like negative learning is the case that it gets more and more difficult, it gets more and more expensive.

In fact, some of those nuclear power plants are causing bankruptcies for the companies that are building, like Westlinghouse with bankrupt because of the nuclear reactors that couldn't be built to the budgets and to the time that was in the United States.

In Europe we have to we have to have a few more now.

But the nuclear reactors that have progressed and should be close to being finished now they are not being finished.

You know.

They there's always a new story of why there's another delay and why the body is going up.

So that's negative learning for you.

We just don't have anything on the ground in North America and in Europe that would encourage us to think that we can deliver nuclear cheaper in the future.

So that's why nuclear looks the way it does in our data.

And that's why.

And this is purely for economic reasons.

Many people think this is something ideological being against nuclear.

It's not even about nuclear waste.

That's not even taken into account.

This is surely economics of building nuclear power plants.

On that criterion alone, nuclear is losing out.

Yeah, it's interesting I characterize.

Here's what I characterize is the success requirements for a nuclear rollout.

And this is before I write your work, which you know, confirmation bias suggests we think identical on this, because of course what you wrote confirms what I think.

One of those other biases that canon points out, which I try hard to overcome.

But the articulation I make is that it's possible to do gigawatt scale nuclear reactor buildouts under the following conditions with reasonable experiences.

First, you have to do lots of dozens in a country or a region.

Second, it has to be a national strategy so that you get override over a whole bunch of local regulatory things that's typically related to to nuclear weapons build out or capacity to build out so the federal government can step in and throw its weight around.

Third, it has to be exactly the same design.

You have to avoid the desire to do bespoke engineerings for a specific site where the technology you just have to do dozens of the same thing.

There was a period in the United States where the light pressurized water reactors were built in sufficient quantities that it worked out.

And this gets to another point you make in the book was that you need a master builder and their team.

Building the capacity of resources who are certified and security credentialed enabling them to work on nuclear plants is deeply non trivial.

Like Holland is looking at a nuclear reactor and they have nobody who has any of those skills and where are they going to get them from DF Well, let's think.

Working out really well in Hinckley.

And so that master builder and their team of experienced people and shared stuff has to be done in a short period of time, which is getting introduced in the next subject, small modular reactors.

They have to be built as components at the maximum viable scale for the physics of it.

Right.

So wind turbines, as you mentioned, are very big these days.

Onshore the average is two point six megawatts.

Offshore, it's much higher, and that's a logistical choice.

If they could.

Build wind turbines the size offshore wind turbines and get them along roads and on railways be hinder bridges, they would, But they can't deliver the blades and math and the nay cells to the sites onshore, but they can offshore.

So that's why we're now seeing the in Wind project, which has been going on for decade decade or so.

Now they're edging up towards twenty megawant individual wind turbine, massive, massive piece be You can deliver that offshore and build them in a port area and put them on ships that are special tenders.

And then you have automated solutions heavy lift automation which enables you to construct them.

But wind turbine it's getting bigger.

It's not like we left wind turbine small because we need to maximize the physics of them.

Now nuclear reactors from a thermal generation perspective, coal plants want to be bigger because that boiler and then generating the steam is more efficient.

The bigger they get up to the maximum one to one point two gigawats and nuclear reactors are just coal plants with radiation instead of coal plants with CO two.

Right, it's just thermal generation.

So small modular reactors want to get that modularity and constructibility and deliver manufacturabilibility, deliverability, but they don't they forego the thermal dynamics, so they lose that efficiency there.

And it's an interesting question for me.

Have you seen other examples where modularity has trumped the physics in your data sets or do you have a perspective on that angle.

I think it's an unknown question to me.

I think it's significant, but I haven't done the math sufficiently ever seen another data set where smaller works outside of small onshore wind energy where the limits are logistics.

So when you when you say that smaller works, do you mean that smaller works even when there is you know, so a larger level you know, like the like the one one point two giga what you're talking about.

Yeah, physics tells us we.

Haven't studied that.

It's a very good point.

It's very interesting to me, and and I understand your argument for small modular reactions, And all I can say is the nuclear industry must hate you because this is the only hope in my That's what I say.

You know, if nuclear is going to work, it's going to work.

It would be small, modular reactions.

But now you're saying, like, no, that's that's it's not going to work there.

You even say it's especially not going to work there, because that goes against the basic physics of the whole thing in terms of where is the sweet pot as spots of prenotting size.

Right, Well, if if if you're right, then they I don't hesitate, you believe that you are, and then I would say it's game over for nuclear.

Yeah, it's interesting because I've looked at data sets, you know, like yours, and the initial reactors being built in the fifties and sixties were tiny.

They were much more like the ones on nuclear powered submarines and nuclear powered ships.

Because guess where they nuclear industry started.

They started mostly in the United States, and they repurposed their PWRs for electrical generation.

But they found the electricity was really expensive at that scale, so they made them bigger, and then the electricity dropped, and then negative learning occurred and they tried to innovate instead of just building the same thing over and over again.

They kept saying we could make it better, and then it all blows up again from a fast budget perspective.

And so that's kind of that interesting thing right now.

I see the small modular reactors getting bigger and bigger and bigger because they're doing the math and realizing small doesn't work.

So it's fascinating to watch it balance back and forth.

I think there's an optimal size in there, but I don't know what it is.

I think it's over three hundred megawats per reactor.

China, for example, or in India for example, has a bunch of reactors, can do Canadian design reactors that are three hundred megawaks or so maybe there's a sweet spot in there.

I don't know.

It's an interesting question, and I haven't seen a counter example to that.

But I do have one counter example which I think is interesting, and that's the next topic, which is pumped hydra storage versus batteries, right because there's a really interesting one.

You know, pumped hydrat storage.

They're billion dollar plus projects.

You have to you have to tunnel a ten meter diameter pen stock up to eight kilometers through solid rock and you have to build a reservoir four hundred meters or higher below the lower reservoir, and you have to put in a you know, four to twelve pumps or counter that are regenerative pumps.

And building that cunnel thirty ten meters in diameter through kilometers of rock is non trivial.

That's the megaproject part of pump hydro.

Everything else is components and trivial stuff, and how to build reservoirs, you know, how to We can buy pumps off the shelf, We can buy the electronic control systems off the shelf.

So that's thing one.

Thing two, though, is batteries.

Batteries are absurdly modular, like Tesla and Barzilla and all these other companies are now delivery containers full of cell based batteries, you know, thousands or hundreds of thousands of small, tiny modular components plugged into a repeatable tattern in a deliverable container framework.

And so that's going off the shelf.

So this is an interesting aspect.

But I always like to tell people this is kind of that introductory piece, and I have a big question for you at the end of this.

Pumped hydro is a I would say it is modular in the sense that it uses multiple turbines and knows how to do that, and it's highly commoditized hardware non you know, there's not a lot of innovation and pump hidri.

Second, it's by far the largest This is stuff people don't know.

It's by far the largest energy and power grid storage in operation globally today, like orders of magnitude higher than anything else.

And that's because we built a lot to give coalon nuclear plants something to do at night, to justify those investments in gigawatt scale generation.

The last point is it's also by far this is another weird unknown to me in the power and energy world.

It's also by far the largest power and energy form of grid storage under construction today.

Like we hear about all these batteries going in, but the batteries are going in twenty to forty to one hundred megawatts.

And I look at Mark Wilson's, you know, across the water from you in Scotland.

Mark Wilson of Intelligent Lands Investments has three pumped hydro facilities that he's developed and he's currently in the process of selling to people who will construct them.

He's done all the transmission interlocks and got all that stuff done.

But Turkey's nest on top of a hill next to lock nests and Lockness is a lower reservoir, and those three have two point five gigawants of power capacity and sixty gigawant hours of storage, which dwarfs all the battery projects in Europe.

Right, And that's one small developer in one small country.

So it's an interesting question.

Where does the modularity and that repeatability and that fat tail stuff have And do you have pumped hydro projects in your data set yet.

Not as a separate category.

We have dams, as I mentioned, as a category, but we haven't singed it out pumped hydro yet, but we'd be very interested.

Very often the way new projects get into the data set is that somebody from the outside contact we would like to work on this, you know, will you help us, like explain to us how you build a good data set, and then we help them and we build a good data set for whatever it is, what asset they're interested in, in this case, pumped hydro, and I'd be very interested in working on that, but we don't have the data yet.

My advice though, would be to people doing pumped hydro, and they probably already know this and don't need my advice, but it would be.

And you know a lot of organizations are contacting us now trying to help them with the question, how do we mode modularize what we're doing further?

Even if what they're doing is very very modularized already, how do we modelize further?

Yeah, and that's the key question, And that's what I would say if you're doing pumped hydro, that's what you need to focus on.

The thing that immediately gets my attention is like digging is involved involved, and we know anything that involves digging is really risky.

You know, when you start digging, you don't know what you find, and a lot of the fat tails actually in construction come from digging and not knowing what you're going to hit on the ground.

So that's something that needs to be sought through carefully for pumped hydro.

How do we avoid the risks the unique risks that are that I involved in digging.

Yeah.

No, I've spent a lot of time randomly looking at tunnel boring machines and the multiple failures over the years as they run into much harder chunks of rock, you know, igneous intrusions from blow and fault lines and water and stuff.

You document few of those, and that's kind of an interesting thing, because pump tidra is an interesting question, you know, I encourage you to think about it simply because it is so big and it gets a little pressed.

It's one of those hiding under the surface of the battery hype and batteries are amazing and I love them.

I recommend you know.

I've published projections on global storage through twenty sixty and you know they're strongly present.

But I think pump tidra is going to win, except there's this nagging question for me about the modulary perspective.

So well, it doesn't matter if it's working, you know it's going to win, especially given the fact that it can scale to the level that you were talking about there.

I actually I consider pumped hyd group and all the battery.

You know, it is a battery in essence.

Yeah, it gets rid of a lot of the fat tail risks.

The Australian National University study from six projects years ago by Matt Stocks very interesting.

What he did is I was looking at machine learning and clean tech solutions globally.

That one popped up on my radar and I talked to Matt Stocks and what he'd done was he'd actually not used machine learning at all, which was interesting.

He'd just taken a GIS data set and queried a bunch of questions.

And the questions were two sites within this much horizontal distance kilometers, you know, a couple of kilometers horizontal distance, at least four hundred meters of vertical distance, because it's mgh.

It's just a it's just you know, you want higher for more void, more maths to get better storage.

And then he said it has to be too near a transmission, it has to be off protected lands, so getting rid of two or three long long tail risks, and you know, then projected there's Huberd times the resource the worst case scenario for storage global.

Wow.

So it's an intro.

I know, it's just a huge resource.

And what that means is if you have a five hundred megawat meter head height difference and a gig a liter of water, that's a gigawant hour of storage.

Yeah, it's just a huge volume, and it's just water is a cheap commodity.

I'm beginning to question for myself because about this cognitive dissonance between the modularity stuff which I deeply internalize and my preference.

I'm tidro and the question is can I justify it by saying water is the ultimate modular resource?

H That's pretty tough, yeah, anyway, But what I what I think you know is that this is good news.

And this is one of the things that struck me, you know when we wrote the book is that, like I mentioned earlier, we saw, Wow, it actually turns out that solar and wind are thin tailed and therefore easy to deliver it.

How lucky is that, you know, giving the problems we have with the climate crisis, and now you're saying the same for pumped hydro.

So those are really good news.

You know that there actually are things there that can really be scaled and it's exactly what we need.

Right So I'm actually quite optimistic about being able to solve the problems with the climate crisis given the technologies we have.

If we can get our stuff together to skate it fast enough.

That's really what it's about.

That we need to be able to do this at a scale and at a speed that is unprecedented.

And I'd like to lean into this now because you talked about transmission, and you also talk about two examples of high speed rail.

And you know, I look at high speed rail and I say that should be like transmission.

It's a linear project.

It has repeatable pieces along the way, have to prepare a foundation, but it's just a repetitive process of putting in place.

So why does high speed rail fail in your examples and in your data set?

And why doesn't it?

And one of the questions I have to ask you is does your data set include the forty thousand kilometers of high speed rail that China's built?

No, it doesn't.

It does include high speed rail, but not the forty thousand kilometers China has built.

Not because we didn't want it.

We'd actually asked for it.

But you know, it's difficult to get data in China.

That's why we were so happy about your data that you were actually able to get them.

We did, We did get through the back door and and and gods, you know, data for Chinese transport infrastructure projects through the World Bank.

You know, the World Bank is working in China, and they have certain requirements regarding getting data on the products that they fund, which meant that in DC they had they had data on some projects like just under one hundred projects in the in China and but not not from the high speed rain network.

This is conventional rail and rail and road projects that we've got data for.

So we don't have the data for China high speed rail.

We do have data for quite a lot of high speed rail around the world, and it's not performing well.

It's not as bad as nuclear or the Olympics.

It's sort of in between and sort of midway in between.

It's also not as bad as ig projects or defense projects.

Defense protects are also terrible aerospace projects.

It's somewhere in between.

And I would say it's because again of the digging.

You know, there's a lot of digging involved in building a high speed rail line, much more than in transmission is easy.

You know, transmission is basically following the surface of the earth, or if it's if it's being if they're digging in laws, that would be because of landscape reasons.

So in the city you put it underground.

But it's not like it's not like a board tunnel, you know, it's just serious digging.

So it's much more limited digging and if there is any digging at all with transmission, whereas with the high speed rail it's very very serious digging.

It's it'sunnel boring and also bridges.

So there's a lot of bridges and a lot of tunnels on any high speed ray line, and they are the two things that will create problems.

We have enough stories and you know, you know the story about the high speed ray line in China where the trains derailed and got a substantial number of people were killed and top civil servants were jumping out windows and got one person got a death sentence and so on.

So it's not like high speed rating.

It's been non problematic in the China with just the evidence that we have, and based on the evidence that I have from conventional rail in China, I would be surprised if high speed rail in China was a walk in the park.

I don't think that's the case, but we'd have to get the data in order to give a final verdict.

Done that, Yeah, I think that they did well China, being China being very engineering and STEM center can very economically focused.

I would say they cut off some of the fat.

Tale simply besays they unlike Europe, where high speed rail runs on different gauges in every country practically, and in the UK, where I believe high speed rail runs on acting different gages and regular rail et cetera, et cetera, et cetera.

China standardized a bunch of that stuff.

And if you look at some of their big high speed rail stations, like in Beijing, every train is the same as every other train, you know, and you talk about that, but repent repetition, you know.

And if we look at you have an example in the book which another subject unfascinated by, which is subways in urban centers.

Same problem right North America and the Western world.

The subways typically are really expensive.

And elam Berteau actually does.

I feel like Elan Berteau must be an honorary urban geographer.

Do you know Alan Berteu in his work?

I know the name, but I might have come across the work.

There's so many people and so much, so many great people exactly.

So Berteau's point was he was the chief urban planner for the World Bank for like fifty years, and so he went into China when it was going through the economic transformation to assist them with urban planning that made more sense than was there.

He went into the Soviet Union or with Russia after the collapse of the Berlin Wall, to assist them to try and figure out what to do with the Stalinist apartment blocks and the city development.

And he went into African countries with no cities and laid out roadge grids.

And his perspective on subways was interesting.

He said, the subways in New York are just as the price per square meter for land is the same as the land directly above them.

So it was a kind of an interesting urban geography point.

Yeah, you also counter with a specific example of a subway, an urban subway problem project which came in cheap.

You know, I have seventy six stations.

Can you tell us about that one?

Yes, that's Madridge.

But before we go to Madrid, let's just returned to China to round all discussion about the high speed rade in China.

I would say that I would be very surprised given the volume of high speed rail that has been built rather than links, rather of a high speed rail that has been built in China, if there wasn't some positive learning, I would be very surprised.

If we went in and started this and found there's no positive learning.

That would be almost impossible building that much in such a short period of time.

So I would expect that to be some of that, and that would support your argument.

But the bottom line is, Mike, we don't know because we don't have the data from China, and this is too often the problem when we work with China.

China is so important.

As expert on megaprojects, I can tell you there's no other country that is more important, because there's no other country that is building more megaprojects than China is doing and have been doing for decades.

So I would just love to get my hands on those data.

I've been in China.

I've talked to my Chinese tottage about that, and they say, forget it.

You know that it's just up the way China works.

If the central leadership says that this is what the data show, that's what the data show, and they don't want any researchers looking at what the data actually show, you know, if there's an official story about what the data show.

So that's the problem.

You know, work with the projects in China and many other things in China.

This is not just for megaproducts.

Economists who are studying the national economy have the same problem.

Okay, let's China.

Let's let's go to Spain now and to Madrid, because that's one of the places where we found a team that we're totally able to beat the odds in the casino.

So the odds for urban rail are not that good.

You know, you will have large cost overroadens on average in constant prices, you would have like forty cost overrun on average, you would have delays, you would have lots of passengers in the forecasts that never show up in reality.

So that's that's your standard urban rail project on average.

So urban rail actually follow the iron law over budgets over time, on the benefits, over and over again, except with fire.

This example in Madrid, you know an outlier that were built twice as fast as urban rail is normally done to schedule and half the cost to budgets and basically got the passengers that they projected.

Yeah, now, how on earth did they do that?

You know?

And they did it by doing what nobody thought could be done, modularizing on the ground rail like modularizing subway.

So this team in Madrid.

They figured out there's got to be an ideal link for a tunnel boring machine, you know, boring a tunnel for a subway.

So they started measuring that, what's the ideal links for one tunnel boring machine with one team you know running that tunnel boring machine.

They figured it out and then they you know, they were in the process of doing the largest expansion of a subway system in the world ever at the time.

Now China has done more, but but this was this was outside China.

Madrid's expansion was like much larger than usual, and they figured, we got to get this done.

This this is the policy of Madrid.

We need to get it done.

And then they just hired as many tunnel boring machines as they needed for whatever length it was that they were they were building, you know, and they would actually get eight tunnel boring machines and tenes in on theer Madrid, you know, to work at one time, you know, when they had the most teams going, and instead of taking you know, eighten years a building an extension, they would take four years to build the extend.

They would work around the clock, which is actually not common, you know, usually for a different reasons, like not to disturb, you know, not to have construction going on at night and on weekends there's a lot of downtime.

And they negotiated with the local community groups that hey, we can take ten years to do this, so we can do it in four years.

You know, if we work around the clock twenty four to seven, we can do it in four years.

If we abide by the usual rule of not working at night and not working on weekends, it'll take more than twice as long.

What do you prefer?

And they didn't hide that.

They preferred to do it, you know, twenty four to seven, and they got that through the community groups actually accepted that.

So that's one thing, you know, very good collaboration with local community groups.

And then this thing about modularizing the different parts of the metro and also stations.

They just you know, in many metros around the world, you know, in London, in Moscow and so on, you will find that the stations are almost like pieces of arts and each station is different, and you get fancy architects to design the stations.

In Madrid that decided, no way, we're not going to invite signature architecture.

So signature architecture is one of the other areas that typically has very large cost overruns and delays and so on.

And they said, in Madrid, why we would be so stupid that we would invite that kind of economic risk in by having specially designed stations by famous architects.

We'll do the exact opposite.

We'll make a very very nice, big area station.

We're not going to drill it.

We're going to do cut and cover.

So we just take a big take a big hole, and we put in the station and recover it and that's it, and we'll do We'll do the same station pretty much, you know, around Madrid, so that we get positive learning curves every time we do a station, we do it better next time, and better after that and so on, as opposed to if you do the spoke stations that are each designed by what a famous architect, each one will be different and you won't be able to get these positive learning curves.

So they really maxed out on all these things.

They also decided no lawsuits.

And you know in construction lawsuits are so common.

You know, this is the actually you write the wrong contracts up front, contracts that actually encourage conflicts you know that people start thinking about how can we sue each other when things go wrong?

From day one, you know, this is the first thing, but even before day one, this is what they think about when they design the contracts.

And if you design your contracts like that, when things go wrong, and they always do, there's always something that goes wrong on products of the size that we're talking about here, then people start suing each other.

In Madrid, they decided we're not going to write our contracts like that.

We're actually going we the client, are going to take on a lot of the construction risk.

We're not going to try to allocate this to the contractors because we've tried that and it usually doesn't work.

Even if we thought we had signed it over to the contract that it always somehow mysteriously inpact with us, you know, So why don't we just face that back and then accept that's the way it is, and then we get a partnership with our contractors where we collaborate on getting as usos risk to materialized as possible and we pay the contractors to avoid it instead of suing each other, you know, when things have gone wrong.

So there was like a handful of basic things like that they didn't Madrid that worked out when you put them all together, it worked out beautifully, delivering at half the price, twice as far and very functional subway.

If you've been there and you try that, you know that this is acting a system that really works.

It's very large for a city of the size of Madrid.

They have a fantastic you know, mitro system.

Oh, this gets into I want to dig into thinking slow and acting fast because when you talk about acting when you talk about the duration of megaprojects.

My supposition, I don't think it was crisply laid out in the book because I think you assume that it's just so internal to you.

But I think when you talk about acting fast, you're talking about the delivery phase where construction, you know, after the shovel hits the ground until completion.

So the duration, the duration you're using is for that portion.

So for Madrid, it's when the first construction site was had, the first shovel in the ground is to start.

But that thinking slow process is intentionally and rightly excluded from it.

Now, can you characterize thinking slow versus acting fast because it's such a fundamental to a premise in your book.

Yeah, yeah, that's that's a key premise one of the chapters.

It's called that you know.

And if we take the Madrid example again, then the thinking slow is the leadership in Madrid actually thinking up all these rules of some that I just mentioned.

You know, we have to have good stakeholder management with the community groups.

We're not going to sue anybody.

We're going to modularize tunnels intersegments of optimal lengths in relation to what it's very relevant to actually what you talk about, Mike, it's about the basic what are the basic physics of this and and and getting back to the basics.

So what's the optimal links that one tunnel machine will do, and then we'll just hire as many tunnel boring machines that we need to do the total links that we need to do right and the same with the station.

So that's the thinking slow, thinking all this through before you do anything, instead of what usually happens is that people only figure out these things, you know, while they're delivering, you know, while construction is going on in Madrid that did it before, and that's what we find.

This is what intelligent intelligence master buildings are doing and that's that's what we call the people who do it this way, we call it master that they're really mastering what they're doing, and they are mastering it by what masters always do is like they really think things out, you know.

But then once they get going, they know the clock is ticking.

And this is the reason that it's so important to egg fast.

Once you've got the shoveling in the ground, as you say, and for an it project, of course it won't be a shovel, it'll be something different.

But once you start delivering, you need to go fast because that's how you reduce your risk.

We call it the window of the doom in the book.

So there's a window and that's the time window from you start delivering until you finish delivery.

That's the window.

And that's a window of doom in the sense that that's where you can really get hurted.

You and your project can really become expensive.

It is when you so if your tunnel boring machine is flooded, like what happened on on the high speed rail project in Hong Kong that you talked about, and on a tunnel in Denmark that we also talk about.

It's actually surprisingly in common you know that you have tonnel boring machines that get flotted, and these machines are expensive, very difficult repair because they're in a hole on theerground, right, So they create shoes delay if you get problems like this.

But that's the kind of thing, and that's why we call it the window or doom.

All these things can happen in that window.

Like obviously, you want that window to be as small as possible.

You particularly want it to be so small that no fat black swan can fly through it, you know, and mess up your project.

And the smaller you make it, the less risks you have of these things of any type of risk, including black swan risk.

So that's the reason why projects that are able to move fast in delivery have much smaller risks than projects that take all the time.

A lot of people don't think of it, you know, like there we have lots of time.

It's not a troplet.

We take ten years to deliver a project.

Well, let me tell you it is a problem.

If you take ten to fifteen years to deliver a megaproject, which is not uncommon, you can count on that something really bad is going to happen during that period, just because that's the nature of things, you know, that's history.

You know, you would have a major financial crisis, you'd even have a pandemic as we've seen now everybody nobody has been thinking about pandemics saw, you know, eighty ninety years, because we haven't had one for about one hundred years.

Right, really bad one that our public health surreil and system.

And as a you may have noticed, I did actually help build the world's most sophisticated, outbreaking communical disease managements STEAM in the world after SARS.

And since SARS we've had h one n one, we've had a bowl up.

Yeah, and now this what I articulate is we have these amazing resilience built in because we've mostly learned our lessons.

We keep forgetting because climate change and pandemics are gray swans.

Yeah, they're gray rhinos.

They're not black swans.

They are expected.

But I will say, let's just take the duration.

The median duration in decades is it takes ten years to construct a nuclear power plant, and so much stuff is happening so quickly.

A decade ago, it was possible to look at the data and say, we don't know if wind and solar will be viable.

We don't have good data on grid integration, We don't have good data on how they'll integrate with markets, We don't have good data on grid reliability with significant portions of that, and they're still fairly expensive.

But any nuclear reactor started a decade ago in construction, it went into that coming into market today is facing a radically different market competitive situation because wind and solar have proven themselves grid reliable, cheap, stable, and are now starting to take over incillery services on grids as well, which is really interesting but really nerdy and output that aside.

Kind of the point is that's that window of doom.

The more you can shrink that, the more likely that your business case assumptions for some thing are still going to be valid when it goes into production, and things like the Ukraine War for example, don't impact what's going on.

And you can see that, you know.

Yeah, you know, you know.

Fild tedlock Ye whose co author with Dan, my co author on the new book on the book called super Forecasting.

So Dan wrote on filter Ted called super Forecasting before he wrote this book with me called How Big Things Get Done.

And he has a law that I call titlock y'all, and that is, you know, you have a certain reliability of your forecast the first two or three years of the forecast, and after three to five years you can pretty much forget any certainty at all of your forecast.

So that tells you everything.

That means that you actually need to have a substantial parts, maybe the major part of your project needs to be done within two to three years.

You should have as little as possible beyond three five years because that's the completely uncertain part and that's where the window of doom will bite you.

So that's that's why, Yeah, that window needs to be kept real small.

And we can see it in the data.

It's very clear this is not something this is not speculation.

This is something that we can see it in the data.

That the face that you are, the lower the risk you get.

And that's you actually win on two fronts.

You you you reduce risks generally, and you even more dramatically, you reduce black lawn risk.

Well, there's another thing that I'd like call out that you you articulate clearly against the Tesla gigafactory and against wind and solar farms, which is that modularity enables you to start accruing benefits before you've completed everything, you know, so that one's the point where you want to win.

When a wind farm is in, what ten percent of is in it can be generated in electricity all the other ninety percent of the wind farm is completed.

The lord for solar farms similar intesslic Do you want to speak more to that, because I feel that is under under estimated as a shrinking of the window of doom.

I think that the whole, the whole discussion of benefits is hugely under discussed and underestimated.

They're much more important than we think.

You know that to get to the benefits is really important.

And the reason that became so clear with Tesla was that at that time Musk was not the rich guy that he is now.

He was actually in shoot Steads.

And you know, so when he was building his first gigafactory, which was called Gega Factory one at the time and it's still called Giga Nevada because it's the Giga factory in Nevada, and he heard that it would take five years, you know, for the normal construction industry to build a factory like that, he said, no way, I mean, if I have to wait for my revenue stream for five years, I'm dead.

You know, test that is not going to exist if I have to wait for five years to get to my revenue stream.

And he said no, and he didn't talk about, as far as I know, talk about Titlock's law, but he acted as if he understood this law.

We need to be in business and generate revenues within the first year, the first two or three years, the first year.

So he designed the factory.

He said, like, let's let's not even talk to these guys at conventional instructions.

And I know this for a fact because people from I know, people from Convention Construction who tried to call Mosque and get a dialogue with him about building the first Geka factory, and he would not talk to them.

He said, we're going to reinvent this ourselves.

Not a lot of people know this, but this is actually the secret source of a lot of what Mosk is doing.

Is that he rethinks things to the basics and their modularity is actually or standardization is a key to that, both for Testnet but also for spasics and allus.

You can you can see it if you start looking at what he's doing and including the Geka factory.

So he actually decided on a design where the Geka factory was consist of twenty one modules, where each module could function as a factory in its own rights.

And then you know, so you just build one of the twenty one modules.

And they did that within the first year there a bit sever within the first year, and they were immediately in business.

They were producing batteries and the famous that what is now called the power wall was coming out of there within one year, and they had a revenue stream that cloud back into Tesla, and Andy financed that growth.

And then they would build another module and that would be, you know, combined with the first, and now they would have a larger piece of factory and so on and so forth.

And that's how they that's how it scaled up the factory.

And they also had positive learning to the degree that they realized, we actually don't need as many modules as we thought we did, you know, because we're getting more and more efficient the more we do this, So we can now produce more volume of batteries in thrower modules of factory.

So they got these kinds of efficiencies through the positive learning curves.

So that's that's a that's a clear story about how modularity can work positively for you.

And yeah, we included that in the book.

Well and and for you know, on my primary concern, which is at electrical generation, though I dot all with transportation as well for the question of wind, solar and nuclear.

As soon as you've got the transmission link in and you're putting in your first wind pat wind far wind turbines or solar panels, you can actually start feeding electricity to the market.

But with nuclear you have to be all the way to the end and say a whole bunch of regulatory approule and you turn.

It on on one day.

The gigawat gets.

Turned on in one day at ten years out and all the debt and revenue has foregone, and it's just that problem until then.

Whereas you know, with the modularity solution, that's part of the reason it's so you know, advantageous.

I think the takeaway for institutional investors, policy makers, energy strategists is they should look at that chart in your book.

I'm not going to say what page it is for the simple reason that I read it on kindle and page numbers are wonky on kindle, But it's which chapter is that amazing chart of variants?

And I think's the last chick.

So that's actually I don't have the final book yet, can you believe it.

It just came off the press, So I don't even know if the page numbers are going to be the same, but I think they are, so letd me just find that job for you.

So that's the job with the variants.

It's on page one seventy three and it's the final chapter card What's your Lego?

That's chapter nine.

So chapter nine called What's your Lego?

Page one hundred and seventy three, there's a diagram with the variance on different projects.

Yeah, I think this is such an important part of this and as you say, this is the first time you published it.

I think every policymaker, strategists, institutional investor could buy this book for that charge in that chapter, and then they should look at that and say, what is my risk profile in my portfolio of major infrastructure projects based on nets and what can I do about it besides call up Ben Flusberg and his firm to help me figure out how to modularize this and avoid stuff.

Yeah, it's because.

You're a small you're you're you're as a buddy of mine says, you've created frameworks and you help sell people ladders to help them with their problem.

But there's only so many ladders you can help people with.

Yeah, we talked, we we we we we do talk to institutional investors from time to time, including pension funds.

So pension funds in Benma, my home country, are big, you know, investing in in infrastructure including you know, energy infrastructure, wind farms and so on.

And they're beginning to get it.

But it's actually taken a while, you know that, Like that, like the thinking and the financial six is unfortunately so conventional, and they have all had the same statistics.

One O one course is that they don't understand extreme value theory.

This is what Mesimtelip has been pointing out all the time, you know, like he's been he's really been pounding this message that that this is the problem.

And and I can say that my experience confirms that it is a problem.

But I do think that we be beginning to get at rode through and we are trying to explain that the risk that you're looking at, I'm not the relevant risks you know, is completely different risks, and those are the risks that we try to highlight with this diagram.

Yeah, so I have to say, amazing book.

It resonated so strong with me because I've worked on billion dollar it projects and I've fixed certain numbers of them or tried to, and I killed a few, you know, as a troubled project fix it guy, and I launched a bunch.

But I think that what I was expecting more from the book was more the Code of the Heuristics, because that seems like if so much of your publication is about those types of heuristics, and so I recommend for people who finished chapter eight to keep reading.

The Code of Heuristics is a very useful set of stuff to paste on a wall to remind yourself as you plan and think about projects and delivered projects.

Don't screw up.

They're very useful.

Now I'm going to be respect I have to be respectful of your time.

I know you know you and you're a firmer, are very busy in demand, and your cycle of interviews should be increasing radically as this book comes out.

So I always like to leave an open ended opportunity.

We you know, clean tech talks.

We've got about a fifty percent US audience and about a fifty percent global audience.

We've been talking about the transformation, we've been talking about climate Chaine and talking about risks.

But what if you had like just an open ended opportunity to give guidance to people based upon your perspective.

What would it be?

If I could say only one thing, it would be understand your base rates.

And base rates are like your basic risks.

Like what we talked about people going to the conceine, the cansino, the base rates and again a casino are the odds you know, in the casino for the individual game.

So there's a base rate for playing the lids, there's a base rate for playing blackjack, and so on.

I find that most people, both doing projects and investing in projects don't understand what the base rates are.

And that fits complete with behavioral economics.

So this is there's something called the base rate fallacy.

That's our fallacy.

We are hardwired not to get the base rate base rates rights.

And that's actually the most simple thing we can do is to get the base rates right.

And we know how to do this now, like we have the data for this.

We know how to do it with reference cards, forecasting and so on.

So that would be my first thing, but there's many things.

And even though you mentioned that the heuristics are in the in the code, there are eleven heuristics, to be specific, and I agree, I really encourage people to get there, but we have also spread them out through the book.

You know, they pop up in different places in the book, also in context with specific examples, you know, and with specific people actually using them and being successful using them.

So that's another thing I would I would say, in addition to getting your base rates right, I would say get your heuristics right.

Start working on your heuristics.

This is an individual thing.

Each master builder has his or her own setup heuristics and and I haven't met a master ability that does not have heuristics, you know, So that's another thing.

If you don't, if you haven't worked on your heuristics, start thinking about this, and it might be a good place to start.

In the coda there, it's very just a few pages, and as we say there, we put those heuristics in there to inspire your heuristics, so you can see which one do you o resonate with and which one would you change, and you probably have additional heuristics that you would add to that list.

That's another thing to do.

Thank you very much.

So I'm Michael Bernard and missus Clean Tech Talks, and my guest today is being Bent Filiberg, the first BT Professor of the Science School of Economic Socksford University.

He also has a professorship at the IT University of Copenhagen, but those things are an aid of him being the world's leading megaproject expert.

He consults globally, he assists people.

His intellectual capital on how to manage risk and programs is used globally and it's very applicable and very good news for clean technology and the transformation we have to do.

Ben, thank you so much for your time today.

Really appreciate speaking with you, and I wish I had six more hours.

Thank you likewise, Mike, and thank you for giving me this opportunity to talk about these things.

Thank you so much.

This is Mike Bernard, host of Redefining Energy Tech.

My guest today has been Professor Benflobier.

If you don't have.

His book How Big Things Get Done on your desk, get it now.

Thank you for listening to Redefining Energy, don't forget to rate the show and subscribe on Apple Podcast, Spotify, or the platform of your choice.