How Can A Canadian Transit Think Tank Be So Deeply Incompetent?

Sign up for daily news updates from CleanTechnica on email. Or follow us on Google News!

Recently, I had the opportunity to sit down with Michael Raynor, formerly a managing director of sustainability and thought leadership for Deloitte and author of multiple books on innovation and strategy. We talked about not only an innovative approach to addressing Scope 3 emissions that Raynor has developed, but also an odd case of a bad study of bus transit that is creating problems in Canada, one which recommended a lot of hydrogen buses along with battery-electric ones. Below is the link to the second half of the discussion and a lightly edited transcript. The first half is available here.

Michael Barnard (MB): Welcome back to Redefining Energy – Tech, sponsored by TFIE Strategy. I’m your host Michael Barnard. My guest today is someone whose work has shaped my thinking and approach to innovation for two decades, Michael Raynor. He’s the co author with the late Clayton Christensen of The Innovator’s Solution. Until recently he was a managing director working on sustainability and thought leadership for Deloitte. Now he has a couple of firms he’s invested in and founded, including S3 Markets. Returning for the second half of our conversation, which is focusing on the Canadian Urban Transit Research and Innovation Consortium’s deeply flawed scenario modeling for Brampton’s bus decarbonization that led Brampton to seek to have hydrogen buses as part of their fleet. 

They had a $10 million difference between the two scenarios, and I’ve just identified, just glancing at the first thing, the price of hydrogen they use a potential swing of $200 million multi times for one factor. And if that was all, then I’d say, well, maybe we can rerun it and see what that is. But then I looked at how long drivetrains and batteries were lasting in real world circumstances with full warranties for eight years in Europe and how short a time fuel cells were lasting. Well, they had the same replacement time of six to seven years for both in their study. So that’s two and a half to three times difference. So a fuel cell bus, you’re going to replace the fuel cell over 15 years five times, not twice. 

And battery electric buses, if you’ve got a 15 year life cycle, well guess what, you’re going to replace the battery. 

Michael Raynor (MR): Once, not twice. 

MB: And so not replacing the batteries. CATL just announced they’re warrantying batteries for EVs and heavy transportation for a million miles. Now I did the math. Basic transit bus length of journeys annually in North America in 70,000km while CUTRIC and Brampton agreed on 58,000 km, a slight variance. Don’t know the reasons for that. But assuming 70,000 kilometers, that’s about half of a million miles over the lifespan of the bus. So CATL is delivering batteries this year, which will last twice as long as a bus. And so you kind of sit there and go okay, why are fuel cells, from data, empirical data documented by the EU in their status reports lasting three years being replaced at the same rate as those other things. So it’s $100 million swing in favor of battery electric just by more realistic duration for battery packs. 

There’s a second one though. They also don’t seem to think that batteries are getting cheaper. 

MR: So hydrogen is. But batteries aren’t. 

MB: Yeah. Fuel cells are going to get better and batteries aren’t. The past 15 years has proven that fuel cells are remaining the same and batteries are getting a lot cheaper. You know, there’s the entire Wright’s Law thing. You’ve undoubtedly spent time looking at manufacturing and Wright’s law and manufacturing the experience curve. Why don’t you lean into this one and give your explanation of it? Because I end up saying this too much. 

MR: The technologies that go into making hydrogen are well understood. They’ve been manufactured at enormous volume for long periods of time. And one of the things that I’ve observed working in the climate space for that matter, is that people talk about economies of scale as if it was some magic wand, as if it was created by Harry Potter. That simply making more of anything is going to result in the costs getting dramatically lower for reasons that don’t need to, that are utterly inscrutable and need not be explained. And that is unfortunate. Right. Because I think first of all, as you point out, as the data would show, cost of fuel cells and the cost of hydrogen don’t go down despite the fact that they’ve been manufactured at scale for long periods of time. 

Batteries are getting cheaper on a Wright’s law curve. So every doubling of production cost falls by a predictable percentage for reasons that we understand. The underlying technologies are such that we are much the same. The one that everybody is everybody, that many people are intuitively familiar with is how microprocessors got smaller and cheaper, et cetera. In fact, they got cheaper because they got smaller. And the same sorts of things are happening with the underlying chemistry and design of batteries as well. Right. So they are getting cheaper empirically for reasons we understand. And it’s because you’ve got those two things to lean on that you’re justified in saying that a reasonable person can expect this to continue for some meaningful period of time. If you’ve only got one or the other, that’s a wobbly stool. Right. 

Because you’re like, well, it’s been doing it for this long. Well then you’re in mutual fund territory. Past returns are not indicative of future performance. And if you’ve only got an explanation, well, then it’s a theoretical prediction. But when you’ve got both, it’s done this for a reason we understand. Well, now you’ve got a theory of the case that you can rely on. And I would assert that we’ve got both dimensions for both technologies. It’s just they lead you to completely different conclusions. Right. Hydrogen is empirically not getting better for reasons we understand, and batteries are empirically getting much better for reasons we understand. And so we can extrapolate those curves. Problem is one of them is a horizontal line and the other one’s an exponential cost reduction. And that seems to have been ignored. So does that scratch the itch? 

MB: Yeah, it absolutely does. We’re building billions of batteries just like we’re building billions of solar panels. And every time we double manufacturing volumes, the price goes down by 20% to 27% per unit. Then we assemble them into packs, and the packs are going down in the surround, but all the content, the big contents, the big price driver, the batteries, is going down. I actually did a professional engagement this year with a European investment fund that deals with green infrastructure to test this. And we’re only this year kind of at the tipping point for a diminishing return on battery cost as a percentage of the total battery pack cost. 

We’re now getting to the point where the battery pack cost is the slight majority of the cost of the total battery pack, not the batteries, but the surround, thermal management, the shell, the power management, those types of things which are not subject to the same degree of transformation. We’re going to see a diminishment in benefits. But back to, you know, CUTRIC, their assumption was, you know, hundreds of thousands of dollars to replace a battery pack. The first battery plaque replacements would be in the 2030s, and there’d be two of them, and they were both priced at probably 2022 rates for battery pack replacement, as if there was no reduction in those costs. 

MR: Right. So they’ve done it twice as often at 50% above the cost, which means that basically four times the cost has been loaded onto batteries. 

MB: And to be clear, still they had to game the system to get hydrogen in there at all, because the most expensive scenario was hydrogen only, and battery only was just $10 million more expensive. So this swing, that’s another hundred million dollar swing in favor of batteries. Ten times that, $10 million. And I wasn’t done. Then there were the hydrogen fuel cell replacements. Well, replacing them not every six to seven years, but every three years. Well then that adds another 25 million bucks to your replacement costs. Right. It’s just like you’re doing it more. So that’s another thing which also exceeded the $10 million variance they clearly said made it the cheapest alternative. 

Now I said they didn’t actually price hydrogen. Well, that included not costing Canada’s carbon price on carbon emissions. Gray hydrogen is made from natural gas. Every kilogram of hydrogen is roughly equivalent in energy to a gallon of gasoline. It comes with 10-12 kg of carbon dioxide or equivalents. And it turns out it’s even more. 

Now, you know, they didn’t even price this one. If we just assert the 408 hydrogen buses and their consumption, it turns into another $25 million in costs on top of this once again blowing this $10 million. 

MR: Well, yeah, I mean I’ll permit myself to jump in there too. Now, you’ll suffer the wages of my undergraduate degree. Which is that, which is that again, if the point of this were to minimize the cost of the transformation of the fleet, don’t transform the fleet, just run frickin diesel buses and pay the carbon tax and you’ll minimize your cost. Job done. That’s super easy. That’ll take 12 seconds and we just did it. 

Right. If you’re inflicting all of this pain and suffering on it, but that won’t serve the hydrogen lobby because if you keep burning diesel, you’re not burning hydrogen. Right. So if you’re going to say, look, let’s take an economic lens to this and purely an economic lens, well then just stick with diesel and shut up about it. If you’re going to take a carbon lens to this, well then as we just discussed a minute ago, 100% battery is the only thing that makes any sense. 

If you, if you look at the minimization of the carbon that you’re left with at the end of the transformation and then henceforth in perpetuity. So tell me again how hydrogen snuck in there somehow because it doesn’t make sense economically and it doesn’t make sense ecologically. So now I’m. Now I’m confused. 

MB: Let me just close off on that one extra thing I found. These numbers are benchmarked with global costs for these things. I’ve seen the reports and they’re low balling hydrogen refueling facilities at the stations. They’re asserting liquid hydrogen facilities for storage of hydrogen. For those who haven’t listened to me rant about liquid hydrogen, I’ll allow myself 10 seconds. It’s 20 degrees above absolute zero, probably it’s 290 degrees Celsius below comfortable temperature for you. 

MR: I am from Thompson [a northern city in Manitoba]. But even so. 

MB: Even so. Okay, okay. So 270 degrees. I’m from North Bay and Moosonee [northern towns in Ontario]. So the liquid hydrogen is there to try and achieve density of hydrogen storage. It’s a very expensive way to do it. It takes a third of the energy in the hydrogen to turn it into a liquid to get some benefits of lower volume. And they’re going to have liquefication facilities and capture facilities and boil off facilities on these plants. Now, for the number of buses they’re looking at, they’ve only priced for the refueling facilities, not for liquefication facilities. As far as I can tell, I’m not seeing any evidence. 

Here’s a data point for you. The government of Quebec, another province neighboring where Michael is sitting. I was there for weeks this summer, great place. 

They actually did a hydrogen fleet trial which just ended last year. In 2019, they leased 46 cars, 45 Toyota Mirais and a Hyundai for governmental employees. And they dispersed them across ministries in Quebec City. They paid $5.2 million Canadian for a single pump hydrogen refueling station that had a small hydrogen compressed tank beside it. Not liquefication, but it also had an electrolysis facility there. So they’re actually making hydrogen at the facility with green electricity from the grid. And it is Quebec so this is actually green electricity. It was low carbon, one of the few fleet things where it’s virtuous. But that $5.2 million is for a single car to be able to refill 5 kg at a time and service six cars an hour if it’s lucky. It doesn’t have any liquefication facilities. 

408 buses between two facilities, 300 in one and 108 in the other, so need something a bit more robust than that. And guess what? Each one of those facilities is going to be much more robust than the $5 million facility in Quebec, and they had costs of $3 million for each facility. The numbers just don’t add up from global benchmarks on hydrogen refueling and liquefication facilities. And so I just said that’s probably another $10 million by itself. Yet again, blowing that $10 million variance out. And so that’s kind of my thing. Mine came up to about $360 million based upon global benchmarking against against realistic real world prices, not whatever CUTRIC has in their data. I figured that was enough. 

And I shared that with you, Michael, because, you know, I said, and by the way, there’s that materiality thing. And then your brain, which is shaped differently than mine, saw something very interesting. So now is the time to talk about your 2007 book. 

MR: Oh, yes, so that’s called The Strategy Paradox, which was, as they say in the business, the trade version of my doctoral thesis. So it was an exploration of the implications of uncertainty on strategy. The bumper sticker in the standard strategy space is that strategies about commitments in the interests of being different from your competition, but unfortunately, commitment necessarily implies certain assumptions about the future. The future tends to be somewhat unpredictable. So how do you square that? How do you make bold commitments in the phase of uncertainty without just turning strategy into a trip to Vegas? There’s a bunch of companies, they all make different bets and we see who wins. And that’s not a function of who was the best, who had the best strategery. That just becomes a function of who picked 17 red. 

I don’t play much roulette, so maybe 17 is not red, but anyway, you know what I mean? And what that led me to was a particular take on scenario based planning. And that’s where, when I looked at those three scenarios, to your point, these are all point estimates of what the next 20 years are going to look like, which is necessarily subject to some meaningful uncertainty. But when you call it uncertain, that doesn’t mean you have to throw your hands up and say, I have no idea what the future is going to be. Back to our earlier conversation about under what circumstances you can make credible claims about the range of future outcomes. If you were to embed that into the cost projections that they made, you could come up with a range of outcomes for each of their three scenarios. 

So you, and really, you don’t need to do them independently. You’ve got batteries, you’ve got fuel cells, and you make projections about the costs associated with each and how those are expected to change over time. And all. And each of them will have their error bars. And so you pointed out reasons to suggest that, hey, wait a minute, current costs are much higher than your projected costs. Past performance and what we know about what drives past performance suggests that the outcomes you’re betting on are actually pretty unlikely. Right? So your, the point estimate that you’ve used in your projection is actually kind of in the 95th percentile of all outcomes. And that’s pretty aggressive. And worse, the projection you’ve used for batteries is in the fifth percentile of all outcomes for batteries. 

And so as a result, you are arguably underestimating the cost of hydrogen and overestimating the cost of batteries. And that’s a problem. So it’s a bad idea to make these decisions based on your point estimates. It’s an even worse idea to make these decisions based on highly biased point estimates.

MB: And so that was kind of observation #1 for both of us. It’s like, hey, $10 million bucks on $9 billion is a nothing burger. And the degree of variance would be higher. That was my observation and your entire thesis and the book that resulted from it says the variance on the hydrogen is a lot higher. 

MR: And skewed differently, right? The point estimates that they use are drawn from a different point in the distribution. If you’re going to use the modal outcome of that distribution, fair enough. But even into that, you can say something about the distribution of those outcomes. These are well developed tools. It’s not like it’s some big secret. You can just do a straightforward Monte Carlo analysis and say, okay, what percentage of the time does my mixed fleet come out with a particular cost profile? That’s easy to do, right? So you build the distribution for batteries, you build the distribution for fuel cells, and then you know, put them together in whatever mixture you think is going to make sense.

How you can even run a series of simulations with an optimization function and say, fine, here’s the best possible outcome. But that, of course, ignores another key dimension of the problem. And I’ll guess that this is where we go on next, which is the implicit cost of the complexity. 

MB: Yeah. Why don’t I talk to the first half of the complexity and then you talk to the big problem. I’m going to talk about bus complexity. 

MR: Sure. 

MB: Just briefly and just introduce it. Diesel buses are well known. They’re well understood. They have an engine, they have a transmission, they have a drivetrain, they have brakes, they have a shell, they have seats, they have alarms and horns and stuff. And a battery electric bus gets rid of a lot of complexity. It has a solid state battery. The battery just sits there. Nothing moves. The battery, it’s got some power management stuff and it’s solid state. Nothing moves there. No moving parts in the power management. Then it’s got the electric motor and there’s one moving part in the electric motor. And then there’s all the rest of the stuff. There’s brakes and all the rest of the stuff that a diesel has. We’ve gotten rid of a lot of complexity. 

That’s really nice, which is why everybody’s asserting that battery electric vehicles are lower maintenance or should theoretically be lower maintenance than internal combustion engines. And we’re finding they are California stuff again, like 60 to 70% of the maintenance costs of diesel buses because they’re so simple and the brakes last longer and stuff. 

So then a hydrogen fuel cell bus. One of the little known facts about a hydrogen fuel cell is that it’s not a black box. You’re putting air and hydrogen in, you’re getting water out. And the inside of a fuel cell is like the inside of a battery. It needs pure stuff at certain climatic norms. Otherwise stuff goes wrong. 

We need to take any air that comes in out of city streets, driving behind cement trucks with their diesel engines spewing fumes and we need to purify it to hospital grade air quality and we need to humidify it to exactly the right humidity that fuel cells like and the temperature range that fuel cells like. So that’s sensors and pumps and filters and humidifiers and dehumidifiers and thermal control. All that stuff doesn’t exist in a battery electric bus. 

And then on the hydrogen, we’ve got the tanks which have to be certified for the equivalent to three to seven kilometers underwater. That’s how much pressure is in those things. Then when the gas comes out, it changes temperature radically. So we got to have thermal management for that. 

It’s got pressure management because the tanks reduce in pressure over time. We actually have to be able to get pumps to pump the gas into the fuel cell at the pressure it likes, which is about 3 atmospheres, like being 30 meters under the water because that’s what the pressure it likes. And it’s got to be the right temperature and it’s got to be pure. 

Then we’ve got the water coming out, which, because when you run a fuel cell with hydrogen, it turns water comes out, which is the advantage. No emissions, just water. Well, water freezes in the wintertime. Whistler’s buses, for example, kept breaking because the water kept freezing inside them because water that came out of the fuel cells wasn’t managed thermally and pumped out before it froze. So we have to have thermal management for that. 

And all of those things have to be assembled around the fuel cell. The fuel cell has not one input and output of power and some sensors. It has gas and other sensors and power and a whole bunch of other sensors that have to be assembled for. Need water out, you need air in, need hydrogen in all those types of things. So the complexity of assemblage is higher. That’s all on top of what’s inside of a battery electric bus, because a hydrogen bus still has a battery power management and an electric motor. You’re adding all this complexity, and it’s finicky complexity. You’re driving this across potholes and you’ve got 700 atmospheres. You have to preserve the stability of. 

You’ve got all this HVAC stuff you have to preserve the quality of. It’s much greater climate control than the passengers get.. Imagine just doing this for all the passengers inside. I just feel so good after I get out of there because my lungs clear up because I’m not breathing city air. You know, that’s what the fuel cell gets, not the passengers. 

So this is all complexity. We see that in the reliability factors for fuel cell vehicles versus battery electric vehicles. That’s already been covered. But then you saw the systemic thing, tons of complexity. 

MR: I think that’s a good setup because of the complexity of the technology you’re describing. I’ll assert that it’s somewhat charitable to say that the cost of all of that complexity is captured in the embedded costs of the devices themselves. When you say hydrogen buses cost more, they have slightly higher maintenance costs, we’re going to assert, you know, we will grant the assumption that all of the complexity you’ve described has been accounted for in the cost associated with that bus. 

But the complexity that occurred to me comes from the fact that you’re running multiple technologies simultaneously. And so now you’ve got a whole new ballgame around the organizational complexity required to cope with that technological complexity. And you’re running three different kinds of buses. That means you’ve got three and completely different technologies in each one. 

So now you’ve got completely different maintenance personnel who have to have utterly separate and distinct skills and expertise and experience and all that go, all intuitions and all that goes with it, right? You can’t just pick up the instruction manual and you know, this is not your, your, your flyboard bookcase from Ikea, right? Where you pick up the pictogram and put the thing together. These things are tough to figure out and it takes real expertise and skill to maintain them. And so now what’s that going to look like? So think about the specifics. You’re going to have a maintenance depot. Are you going to have multiple depots for each one? Are you going to put all three technologies in the same depot and the need for the different parts. 

And now you’ve got these really finicky or quality clean rooms for your fuel cells. I don’t think you’re going to put them next to your diesel buses. Anybody who’s ever been in a good old fashioned diesel bus garage, no, that doesn’t sound like a good idea. Are you going to have the right people there at the right time to fix the bus that’s broken down in a particular way? On and on we go, right? And so that kind of complexity, it’s a complexity and expense born of diversification. So think about any company that’s got three or four or five different lines of business that compete and operate in entirely different industries in different markets. There’s a very well known diversification discount that goes along with that kind of complexity and overhead. 

And that’s what they’re inflicting on themselves by running this very gradual rollout of two different technologies and a very gradual wind down of the diesel buses. I didn’t see anything in there that acknowledged a cost arising from that. And what I would say with a pretty high degree of confidence is that those costs are far from trivial. 

MB: There are two points there. The first one was their transformation scenario. Their timeline didn’t respect the discontinuity. This is your observation. The discontinuity that when you get down to a couple of hundred diesel buses, the cost per bus for maintenance shoots up so radically that you want to get them off the streets as fast as possible. 

MR: I think that’s right. That’s good old fashioned economies of scale that most people understand pretty intuitively. If you have to maintain an entire diesel maintenance shop for three buses, those things don’t scale easily. You build it to manage the 600 diesel buses that you’ve got, and when you dial that back down, you’ve only got 50 left. Okay, fine. The labor cost may be somewhat variable, but all of that fixed infrastructure is not nothing. At some point you’re like, this is ridiculous. Time to get out of the diesel business. 

MB: And CUTRIC, I’ll remind everybody, is in the business of fleet transformation. This is a simple observation that should have been caught sometime in the past nine years. They’re thinking about how fleets would transition. This is not rocket science. 

MR: I say different things with different degrees of ferocity based on my level of certainty. Which is that it seems to me that the, the complex, the complexity born of diversification is something to take very seriously. Everything I’ve learned would suggest that we need to look at it. How costly is it? I don’t know, but we might be able to come up with some reasonable estimates. And what I would suggest is that CUTRIC’s been thinking about fleet transformations for nine years, but nobody’s ever done one. So basically it’s data free. What does it cost to transfer? I mean, in Canada, like, what does it cost to do this? They could, they could throw their hands up and say, we don’t know, time to tell. 

But back to where you started the conversation. Right. What about the reference? You know, the best reference case examples. Other people have done it. And so there’s something to learn from them, one would hope. 

MB: Well, I will go back to 2019. One of my weird experiences is that I was in the initial phases of consulting before COVID hit with Translink in British Columbia for their battery electric truck and battery electric bus transition. You know, I’d run into the person who was responsible for it at a conference and I started talking. And so we were having discussions around this and I asked him a simple question. This is the person responsible for replacing this. He’d had this role for a year or so he’s well into planning. I said, so when are you going to China to get the lessons learned from the leading practices of large fleet transformation from, for example, Shenzhen, where they now have 16,000 electric buses on their roads? He said it hadn’t occurred to him. And this is not a slight on this person. 

Then I suggested the next thing. New Flyer, the primary vendor for transit buses in North America, had a peak number of delivered buses in any year of 6,500. You have 2,100 buses in a small geography, there are 72,000 in the United States. New Flyer clearly can’t scale to an unfamiliar technology in the timeframe you’ve got. So obviously the strategy there is to joint venture with Yutong or BYD, two primary bus manufacturers in China, or just get them to establish a factory. Oh well, he said, Canadian customers wouldn’t accept Chinese quality of buses, indicating that he’d never been to China’s current cities and so he had very different ideas about Chinese consumers would or wouldn’t want. 

Anybody who’s been to Beijing or Shanghai or one of the other modern Chinese cities right now, including Shenzhen, where they have 16,000 electric buses. Chinese consumers, for anybody who’s listening, are incredibly picky. They have an amazing choice of cheap stuff that’s insanely high quality and diverse. That’s why Ford’s CEO recently came back from China and said this is an existential threat. They’re doing stuff we can’t for lower prices. This is not, once again, a slight on the individual. This is a systemic challenge. I said at the beginning, transit organizations are geographically bound and they’re very focused on operational excellence. Good on that. But ask them to be transformational people and look up and outside of their patch is challenging. They need help to do that. 

MR: Well, I mean, yeah, not that anybody needs defending from me, but that’s true of all of us. It’s a shared pathology. To your point, it’s not reasonable to ask people running the current system to figure out how to transform it. 

MB: But I would say the next thing. Do you think that should apply to CUTRIC, this well funded think tank whose job it is to go and find those lessons learned globally? Have they been to China? Have they been to Shenzhen? Have they documented and brought back the lessons learned? Have they looked at any of this? I see no evidence of it. 

MR: No, no, absolutely. I mean, they are nominally, at the risk of putting words in their mouth, they are nominally the mechanism. And your question is rhetorical. They’re nominally the mechanism to do precisely that. That’s why it exists, so that the people who are running Brampton and Mississauga Transit can keep worrying about how to run Brampton and Mississauga Transit while it gets transformed. Right. You got to rebuild the ship while at sea. That’s no trivial undertaking. 

MB: And it’s been done to your point. It has been done globally. It’s been done in China. The place to go to learn that is China. This entire question has been thoroughly explored and documented. If you go to Chinese cities, it’s transit. It’s not like their power system. They’re very collegial. They’ll welcome people and tell you stuff. It’s a very different thing. It’s not like their transmission lines which aren’t on Google Maps because they’re secure critical infrastructure. Want to know how they’ve transformed their transit? Just go ask them. You know this is a different thing. So this is yet again a place where I would say the data was out there, CUTRIC didn’t get it. This is a recurring theme. This isn’t all though because we’ve already talked about your observations that materiality is non-existent. 

My observation is that the variance really favors battery electric and it’s completely missing from the report. It’s not that there are no error bars. There’s no mention of confidence in their estimation over 20 years and you know, 1100 buses of three different types. No mention of confidence. It’s a failing grade. I assert this, you used to be an MBA professor. Would you have given this scenario modeling a passing grade? 

MR: Well, there is such a thing as grade inflation. So maybe they might have passed. They wouldn’t have been top of the class, that’s for sure. I would feel that I had failed a student that came up with that as their best effort. 

MB: One of your observations was that they’ve labeled a graph wrong. 

MR: That’s one of those things where it’s a small thing, that’s a big thing. I was thinking about this earlier. Many people have been exposed to that story about Van Halen and the, and the Brown M&Ms. Oh yes, you’ve heard that. I learned of it in Atul Gawande’s book The Checklist Manifesto. 

So for folks who haven’t heard it, this was back in the mid-70s when it was the real Van Halen. When it was in the whole stadium rock concert had just kind of taken off and it was incredibly complex. There was not a well established infrastructure for setting up concert venues with all the pyrotechnics and the lights and the explosions and all that. And so there was kind of a phone book of instructions that went to how to set all this stuff up. And if you did it wrong, people can get hurt or even killed. And so the story is that Van Halen on, you know, page 274, paragraph 838, sub, paragraph 7, sub paragraph I said a bowl of brown M and Ms in the green room. 

And so when they got to the venue, the first thing they did was go look for the brown M and Ms. And if they were there, they’re like, looks like they read the phone book, right? So it’s a way to make sure that people have done what they needed to do. 

CUTRIC has a chart in the report that is labeled net present value. That’s my bowl of red of Brown M&Ms, which is that the vertical axis on their discounted costs associated with each of the scenarios is labeled net present value. It’s a very small thing, but it does potentially betray a misunderstanding of what they’re actually doing because that’s not a net present value, it’s a present value of costs. 

A net present value is where you are looking for the net, hence the name, between a series of negative cash flows and a series of positive cash flows that occur at different times, right? In the stereotypical example, you’ll have a series of investments, negative cash flows for the first few years, and then when you’ll start generating revenue and investment tails off. Then you’ll have a series of positive cash flows in the future. You discount them all back today to get a net, the net of the positive and negative present as of right now, value of the overall investment. Fair enough. 

What they just did is they took all of the negative cash flows, that is to say all of the expenses associated with each of these three scenarios, and discounted them back today. So they weren’t netting anything, they were just calculating the present value of the future expenses. Perfectly legitimate thing to do, by the way. In fact, exactly the right thing to do, just mislabeled. And the funny part is, to me, at least, because I know exactly why that happened, I’ve spent enough time as an Excel jockey. 

The present value function in Excel, the PV function, doesn’t admit of variable revenue streams or cost streams. To do that, you have to use the NPV function. And so they, I can see it because I’ve seen people that have done this stuff for me back in my consulting days, they just, they wrote it out, they say, oh, I use the NPV function, so it must be an NPV, but it’s not. And if everything else in the report were gold plated, then that’s easy to overlook. But when it’s part of a pattern of what seems to be inadvertent oversights, then one can only begin to wonder, right? Once your Bayesian prior is that maybe these folks aren’t quite as expert as one might like. Lots of things I can’t do well either, by the way. 

So nobody’s criticizing anybody personally, but it does suggest maybe these weren’t the right tradespeople for this particular analysis. 

MB: It is part of the deeply substandard work that they’ve done, by my professional opinion, as someone who spends enormous amounts of time looking at hydrogen transportation scenarios and reports, by yours as a scenario modeler and a person who’s written a book literally on it. But it came to another couple of insights. So one of the insights you made was why are we taking so long? If we’re trying to get benefits, why not implement the battery electric buses faster? But then you said, huh, when are they actually buying and implementing fuel cell buses? So what was your observation there? 

MR: Well, they’re all pushed out to, I think it doesn’t start until 2034, 10 years. I mean, call me a liar for a year or two one way or the other. It starts very late in the planning and that has a very interesting consequence for a couple of reasons. First, it may well be a fig leaf behind which one can hide to say that’s why the hydrogen costs are so low. Right. We’re modeling them lower than current costs, but trust us, it’s 10 years in the future, so we’ll get there. But the quantitative implication is that hydrogen buses even then are higher cost. 

And so they’ve taken the higher cost and put it out 10 years, which means those higher costs get discounted back, whereas all of the buses that, the battery electric buses that you’re deploying right now don’t get discounted back. And so the hydrogen looks artificially cheaper if they were the same cost, if hydrogen buses and battery buses were the same cost, that discounting wouldn’t matter because it would just be one for one. Right? The present value of the cost of the total project would be the same regardless if they cost the same. You’re shuffling around one unit here versus there. Doesn’t matter. But because hydrogen buses are more expensive, that’s not what happens. 

If a hydrogen bus costs 1.3 times as much as a battery bus, all in capital, opex, the whole bit, which is not crazy. In fact, that’s generous given what you’ve described. That has a material impact, right? Because now what’s happening is that a $1 battery bus isn’t discounted, but a $1.30 hydrogen bus is. And all of a sudden it has a real material impact. If you were. If you take those differential costs and just say, fine, I’m going to give you the future costs of the hydrogen buses. I’m going to give it to you right now. You can have hydrogen buses for what you think they’re going to cost in 2034, but in exchange, I’m going to make you deploy them now. And when I do that and rerun the discount, the whole scenario blows up, right? 

So the seemingly comparable costs turn out to be an artifact of the timing and the deployment and nothing else. When the viability of your entire scenario turns on that and only that, like, let’s throw everything else out the door now, all of a sudden, everything determines. Everything turns on the timing of the deployment. And if anybody can tell me to a certainty that what they think they’re going to do 10 years from now is exactly what they do 10 years from now, I’m back to my error bar conversation. 

MB: You had an estimate, and I’m running with the estimate whether you’re comfortable with it or not. But Basically it’s a 3 and a half percent discounting rate they use. People can argue up or down. The EU’s guidance is 4%. The United States guidance is 2 and a half percent. 

MR: Big deal. 

MB: 3 and a half percent. 

MR: 3 and a half. 3 and a half of like, not a crazy number. 

MB: Nope, it’s very reasonable. But that’s every year and it’s accumulating and turns into about 40% in 10 years, which means that stuff from 10 years on is discounted by at least 40%. Now, I did math against the full hydrogen fuel cell bus scenario and I took the 408 buses in the blended scenario and I took a chunk of that and I multiplied by 0.4 to find out what percentage would be, you know, what grossed it up. And it’s $1.1 billion. 

MR: It’s now 10% of the entire project that turns on your timing assumption. 

MB: I just look at that. And my stuff adds to that, by the way, because they’re low volume costs and then they’re pushing them out. If we add my $360 million to their hydrogen fuel cell costs and then bring them, don’t discount them by 40%, then we end up with an even higher number. And the total variance that they based the blended scenario on was $10 million. 

MR: Yeah. And when you add it all together, we got something close to like a billion five. 

MB: I look at that and I’d say I personally, if I was a transit agency and looked at our findings, I would be seriously questioning CUTRIC and whether they were biased towards hydrogen. I would be getting third party others to redo any numbers that CUTRIC had provided me because they’re not credible. And if I was CUTRIC, I’d be looking at my governance very seriously and wondering how we’d ended up in this terrible place and getting third party people to assist them to improve their modeling. Because if CUTRIC doesn’t improve this, they don’t have a reason to exist. 

MR: Yep. The last thing I’ll throw in, which is the opposite of the criticism we just leveled, right? Which is that they should be discounting, they’re just doing it inappropriately. When you look at the climate benefits of the transition, which they don’t consider at all there, you actually can’t discount the future impact of the changes. So when you roll this out slowly for whatever you know, there are reasons. I think it was Feynman who said this, or maybe not. Time is nature’s way of preventing everything from happening at once. Right? So you can’t, you can’t just wave your wand and say fine, battery buses, poof. It takes a while. So yes, there has to be a transition, as we’ve discussed. 

But there is real benefit in accelerating the deployment because it reduces carbon sooner. Future harms are just as bad as present harms, unlike future costs, which costs less than current costs due to discounting. Future harms don’t discount because they don’t harm just you. Climate change is all about the externalities. And classic standard finance theory says you can only discount benefits or, you know, benefits or costs if you’re the only one benefiting or being harmed. As soon as you violate that assumption, and the reason for that is that if somebody else is harmed, how do you know they use your discount rate. You don’t know what their discount rate is. Maybe they have a negative discount rate, right? 

Ten years from now, people who are harmed by those emissions are, they’re getting harmed then, not now. So you may think that harm 10 years from now is worth some fraction of what it would be worth 10 years from now because it’s you. But for someone who gets harmed, then it’s worth 100% of the harm to them then, right? And you don’t get to decide for them what that’s actually worth. There are both good financial reasons and good moral reasons to look at those future emissions and say they are just as damaging then as they would be if they were happening now. And that gives you a very good reason to take that into account when deciding how quickly to affect the transition. 

So there’s good economic reasons to do it fast. There’s good moral and ecological reasons to do it fast. There are good economic reasons to make it simple. There are good organizational reasons to make it simple and just. And on every one of those dimensions, it’s just, you get a big red X all the way down the list, which is really disheartening. 

MB: To that point of discounting, I’ll keep you for another two minutes just to express this thought. The European Union with their emissions trading system and their carbon pricing and the core carbon border adjustment mechanism, get this future discounting thing. They say, we have a carbon price, it’s going to go up. And we have, for budgetary guidance, for projects, for lifecycle carbon assessments and for business cases, because the carbon has a price to your business case, here’s your schedule, year by year through 2050 for the carbon price to pay, right? What they did though, is they said, okay, in 2030 it’s going to be around €200. In 2040 it’s going to be around €300 and 2050 it’s going to be a little more. 

But then they said at our 4% discounting rate, we’re going to gross those up according to the discounting rate into the future. And you use the gross up number in your business cases so that when you bring back 2040, €300 per ton to 2023 or 2024, it’s still €300 a ton.

MR: I was not aware of that. That is impressively sophisticated and coherent. That’s amazing. Good for them. 

MB: And they are putting a carbon price in the world with the carbon border adjustment mechanism and they’re looking at getting scope 3 emissions into the CBAM. 

Let’s net this out. CUTRIC did a really bad scenario report for Brampton and Brampton’s transit general manager has accepted their recommendations and made that assertion to the city. So now they’re trying to find $9 billion to spend on this, which is actually $10.5 billion to go down a bad path that emits a lot more carbon and provides worse transit service for the citizens of Brampton. You know, it’s impossible to understand how an organization like CUTRIC has arrived at this position. 

To be clear, I don’t expect transit managers to have written a book on scenario modeling and variance, but they should have people who are doing due diligence on the proposals they receive. So I suspect there’s a quality control issue that the city of Brampton might want to look at a little bit.

Let’s leave this on a different note. At the end of these things I always leave an open ended opportunity for you to say whatever you’d like. So anything we’ve talked about, anything we’ve missed, any thoughts for other people who are, you know, late to the party, you know, I forget the term you use, but arriving at sustainability later in their careers, Something like that. 

MR: Yes, I did. Longtime listener, first time caller. I watched too many episodes of Frasier going through the doctoral program. I won’t generalize from my experience, I’ll just share it, and if that’s helpful to anybody, then that’s a happy coincidence. I think I mentioned in passing, I’ve come to the view that we’re all kind of on our own journey when it comes to thinking about the climate crisis and what we want to try and do to involve ourselves. As they used to say in Chaplin high school, if you’re not part of the solution, you’re part of the problem. A good bumper sticker and we should all be aware of that one. 

And I guess I would say that it’s, it can certainly seem overwhelming. I am frequently overwhelmed by it. It can be depressing. I’m frequently depressed by it. But it seems to me there’s kind of two choices, try or give up. And for now, at least, I choose to try. And my hope is that for anybody who’s on their own journey and thinking about the degree to which it’s worth doing something, I guess I would I’ve come to the conclusion that I will regret not having tried far more deeply than I will regret having failed. So that’s why I’ve got my lance tilted at these particular windmills. 

MB: And on that note, Michael, I’m Michael Barnard and this has been Redefining Energy – Tech. My guest today has been Michael Raynor, former managing director working on sustainability and thought leadership with Deloitte, now co founder and investor in S3 Markets and Canary Medical, author of many books you should read, at least one of which has been guiding my observations of the world of innovation and technology for 20 years. And the windmill he’s tilting at today are turning scope 3 emissions for really affluent white collar firms into avoidance of carbon through S3 markets and the city of Mississauga where he lives, where they’re doing something really stupid with hydrogen buses. Michael, it’s been an absolute pleasure. Thank you so much. 

MR: My pleasure. Take care.