Doe mee met de fietstelweek!

Door mux op zondag 13 september 2015 11:44 - Reacties (34)
Categorie: -, Views: 3.271

Een heel kort bericht, maar niet minder belangrijk. De komende week - van 14 t/m 21 september - organiseert de Fietsersbond samen met ANWB, gemeentes en het ministerie een grote meetsessie voor fietsbewegingen. De Fietstelweek. Hierin wordt burgers gevraagd om een speciale app te installeren op hun telefoon en deze te gebruiken om hun fietsbewegingen in kaart te brengen.

Ondanks dat we in Nederland de meeste kilometers fietsen per persoon in de hele wereld, hebben we nagenoeg geen informatie over exact hoe mensen zich met de fiets bewegen. Je ziet wel eens telmachines - die kleine metalen kastjes die aan een paal zitten vastgeketend met twee rubberen slangen die schuin over het fietspad liggen. Deze machines worden gebruikt om fietsbewegingen op specifieke punten in de infrastructuur te meten, en hier hebben we een hoop informatie van. Maar hoe mensen zich buiten deze routes bewegen? Geen idee.

Dit is belangrijke informatie, want de aantrekkelijkheid van fietsen en alle voordelen die fietsen met zich meebrengen voor de volksgezondheid en infrastructuur is afhankelijk van hoe we de fiets gebruiken. Als we het fietsen beter willen maken moeten we weten wat het probleem is. De enige manier om hierachter te komen is meten. Meten, meten, meten.

Dus hierbij mijn oproep: Iedereen die mijn blog leest: installeer de app. Het is maar voor een week, en het geeft belangrijke informatie. Zelfs als je niet van plan bent te fietsen de komende week - installeer hem. Dit is ook belangrijke informatie. Het heeft geen zin als alleen fanatieke fietsers meedoen aan de Fietstelweek.

Veel fietsplezier!

City cycling in London is a joke

Door mux op donderdag 09 juli 2015 21:25 - Reacties (67)
Categorie: Fietstechniek, Views: 28.857

Earlier this year I was involved in a small discussion thread on bicycle paths on Reddit. The thread was about new 'cycle superhighways' in London. I commented that to me, these bike paths looked - at best - about on par with mediocre Dutch bicycle paths. Some redditors, presumably from the immediate area, convinced me that actually there is a lot going on in London to improve matters. In fact, in their opinion biking in London is alright. And the cycle superhighways are great!
Boris Johnson, the current Mayor of London, revolutionized city cycling in London (image reproduced from The Standard without permission)

Well, this is interesting. I'm an avid cyclist - one of millions in the Netherlands - and I have a more than passing interest in infrastructure. Also, I happened to be in London with the missus for about 5 days. The stars aligned, taking our folding bikes turned out to be not only possible but also one of the cheapest options, so we went for it. Let's try out London on a bike.

Can you tell by the title of the blog how it turned out?

So, London has terrible, abysmal infrastructure. So much is wrong with it that my very limited amount of time there I was easily able to collect enough for a meaty blog post. I'll start off this blog with general remarks about the infrastructure and finish off with some awesome 'bicycle' infrastructure pictures. I'm in full-on rant mode, so bear with me. There will be pictures to offset the rambling.
So, exactly how does this bike path work?

No explicit right of way
One of the first very obvious problems with London infrastructure is the lack of explicit right of way marking and signage. For instance, I have not seen a single priority sign ([XXXXXXXXXXXXX]) or temporary priority sign ([XXXXXX]) at all in the city. This has the consequence that in a lot of places, car traffic becomes Indian. Cars just merge and split willy-nilly. Side roads with optional 'we tried to give you the feeling that you might want to probably wait for other traffic before you get on the main road'-markings - can apparently be used as priority roads.
Explicit priority signs. Known in the Netherlands as B-series traffic signs (left to right: B03, J08, B01, B02)

In places where right of way is explicitly handled, the inconsistency boggles the mind. For instance, priority pedestrian crossings (often without zebra stripes) have these little yellow light bulbs on stalks to indicate to drivers that there's a pedestrian crossing. Only problem: these yellow bulbs have the world's dimmest light bulbs in them and are invisible during the day and even in lightly cloudy skies. Some crossings have added brighter yellow LEDs around the bulbs, but that's just an afterthought at best.
Typical pedestrian crossings; this one with zebra stripes and definite stopping lines for motor traffic. Note the poles with yellow bulbs; even in this image they are not easy to make out against the sky!

Worse still, cars do not observe right of way for unequal traffic, e.g. pedestrians and cyclists. Cars coming from a side road onto a main road will just stand right on the pavement, so any pedestrians or cyclists have to go around. Pedestrians do not get explicit right of way at crossings either, the crossings are just crossing indicators, not zebra crossings. This essentially means that the deadliness order becomes the de factor pecking order, which is exactly the wrong way around. In a well-designed traffic situation, pedestrians are gods, cyclists/minor motor vehicles have second dibs and large motorized traffic just has to wait for the rest. In London, motor traffic rules and even within motor traffic the biggest cars win out. Which is buses. Who drive like lunatics. Speaking of which...

Traffic consistently drives WAY over the speed limit
So I've been told that the general within-city-limits speed limit is 30MPH (on par with most EU countries - about 50km/h). Because of a combination of extremely lax policing and very loose tolerances on the speed limits, you can apparently get away with consistently driving at least 10MPH over the speed limit. This means that on most city roads, cars drive way too fast. I didn't have any proper measurement equipment for this, but just judging by eye, the majority of drivers on roads like Clapton Ave. are driving in excess of 70km/h. Combine this with bicycle paths generally being either absent (i.e. you have to drive in the gutter) or unsegregated and very tight, and by Dutch standards you will be overwhelmed with a feeling of thorough unsafety when driving on the roads.

London is in a permanent state of disrepair
Can you count the repairs? Can you tell what the primary road surface was?

Words like 'abysmal' and 'horrendous' will frequently occur in this blog, and not without reason. Anywhere you go in the city, one thing is almost perfectly consistent: road surfaces are a patchwork of 15+ year old primary road surfaces littered with repair upon repair upon modification upon modification. I've seen patchworks where the primary road surface has had both transverse and longitudinal repair patches, repair cobbling (this is NEVER supposed to be permanent in an asphalt road!), closed gutter resurfacing and crack repairs. Oh, and the primary road surface was so incredibly far worn-in that the top surface was basically gone and the lower - coarse-grained - asphalt showed. All in a 20-meter stretch of road. Was this the worst part? No, this was not far off the average. Really!

It seems like large stretches of London have had recent(ish) re-done waterworks. So a sizeable minority of the roads have these very long repairs where obviously the road surface was removed, a gutter was dug, stuff was done, gutter re-closed and the road surface repaired. But they didn't properly fuse the asphalt together, so lots of roads are now left with giant longitudinal ridges that are just begging to catch your bicycle tires and cause you to fall over, preferably when a large speeding bus is just behind you. Horrendous. Abysmal.
This manhole cover - with sharp ridges - was at least 3-4 inches (8-10cm) lower than the road surface. Imagine biking in heavy traffic with these kinds of sudden obstacles!

The state of disrepair goes further than just the roads, though. My girlfriend commented that almost on every road, somebody is busy fixing or maintaining something. A huge amount of work seems to be going on just to keep things from falling apart completely, which leaves little resources to modernize. But that's just conjecture on our part.

Haphazard placement of infrastructural features
And once you get through all the big, generalizable problems with London infrastructure and you really start looking at the road in more detail, it becomes so clear why the roads feel so bad. It's not just the road surface, it's not the crazy drivers, it's not the fact that animal instincts prevail over design and order. The biggest problem is that infrastructure is not laid out holistically. Everything is just placed haphazardly, as needed, as if the rest of the road does not exist.

Hey, we need to make pedestrian crossings. Let's improve safety by extending the pavement onto the road a little bit and adding a little guard rail. We are smart! Eh, no, you have now made things a million times worse for cyclists who have to navigate an even narrower road together with motorized traffic. Also, even if cyclists want to get out of the way of danger they can't because there is a fence in the way!
So, where do I bike? By the way: note the 'give way' sign! I complained about that earlier!

Hey, we need to put this big, immovable box somewhere. What is the best possible location for this box? We can choose between 500 yards of empty pavement, or this one spot where we already placed a lamp post and an electrical installation. Right where the pavement is the narrowest anyway. Can you guess where they put the box?

Hey guys! I have an AWESOME idea that will improve traffic safety for pedestrians and cyclists who want to cross one of the busiest roads in London. Let's put a traffic light here, just for them! Let's also put it just about 50 yards over from where people actually want to cross the road and make sure it takes at least 3 minutes before they can go, so that they can just give up and cross the road weaving through traffic anyway. Seriously, I timed it. 180 seconds, to the second.

So now let's talk about bicycles

Cycling in the Netherlands
I come from - depending on who you ask - the number one or number two (after Denmark) biking country in the world. Bicycles in the Netherlands are very popular in pretty much every form. You can go from anywhere to anywhere in the Netherlands on 99+% segregated, safe, well-maintained bike paths. Here, let me show you a random bike path a few hundred meters from my house:
The bike path along the Grindweg/Bergweg Zuid between Bergschenhoek and Rotterdam

This is pretty much totally representative of average bicycle paths. A lot of the Dutch infrastructure budget goes towards this, because almost everybody uses them regularly. Everybody has a bike. We don't use them just for couriering, or just for pleasure, or just for getting a work-out. We use them mostly just to get from A to B because they are convenient, comfortable and fun. And for sub-5km rides - often faster and a lot cheaper than a car.

There is no uncertainty on the bike: I don't have to plan a route to my destination and specifically seek out bicycle paths. They are just there, everywhere, always, in excellent condition. With plenty of waymarkers.

If you're interested in some very good videos about bicycle paths in the Netherlands - from recordings of bike trips to informative videos about some historical or infrastructure facts - I highly recommend the Youtube channel 'BicycleDutch'.

Both of these videos are a great watch - highly recommended if you like infrastructure porn!

London bicycle infrastructure in general
Switch to London. Car infrastructure is badly designed and maintained, but in most cases bicycle infrastructure is simply absent. It is certainly not a 'London bicycle network' - as they like to proclaim on large signs. Bike lanes start and stop within 100 meters. Or, my personal favourite, little bike drawings are put on the road. I guess to indicate... the existence of bikes? Come on, London - you can't just draw a bike somewhere and expect people to seriously call that a bike lane. It takes more effort than that!

In a lot of places, it's too dangerous to ride on the road, but to go on the pavement would mean:
  • Bothering lots of pedestrians
  • Riding over uneven terrain (plants, uneven pavement)
  • Riding through wildly changing widths of pavement, from road-lane-width to squeezing yourself between obstacles
  • Having to negotiate tall pavement sidewalls
Not having a designated place on the road means that cycling is a free-for-all, and the cyclists consequently will act like they're not bound by any rules. Which upsets drivers. The drivers, by the way, are nice - much nicer than I would have expected. I ride like a Dutch person - I try to avoid being run over, look around, indicate direction, etc. Cars seem to appreciate this and give way to me when I'm overtaking, let me cross, let me turn. I mean, not buses, but most other drivers are just fine.

Low effort
I've mentioned effort in bicycle infrastructure before, and this really is a running theme. Anywhere you go, bike infrastructure is a third, fourth or tenth priority. Actually, I really question the arguments given for the poor design of London roads. A lot of people say it's the age of the city and its road network, but this is obviously bullshit. The same crazy infrastructure extends into areas with plenty of space and besides - we have plenty of equally old cities with orders of magnitude better infrastructure.

So what is low effort? It ranges from simply not giving a shit:
There is no other reason than laziness for the wrong order of painting here

to realizing that they forgot to budget a bike lane into a new road design and just divert bikes over... a construction site?!

Yes, this is the official bike route. Here are the waymarkers:

Giving up after medium effort
Sometimes things go well. For instance, this bridge could be in the Netherlands. Sort of. The bike lanes are tiny:
My folding bike for a size comparison - this two-way bike lane was about 1.70m total width - a tad over the prescription minimum single lane width in the Netherlands

But there is a completely segregated, walled off bicycle lane over this bridge! I must be in Valhalla. This bridge has an awesome view - the O2 and ExCeL on one side, the City and the inner city on the other:

This is awesome! That was a great view. I feel like a tourist now. Let's see if there's more!

Wait... what is that... is that...?

Well, jeebus. You managed to outdo yourself, London. In case it's not obvious yet: what is happening here is:
  • Cyclists descend from the bridge
  • Still on the decline, having gathered quite a lot of speed, you are expected to make a sharp left turn
  • Immediately after the sharp turn at tremendous speed you have a non-priority crossing over a road with very poor visibility and equally poor positioning (oncoming traffic cannot easily see bikes, nor can bikes see the oncoming traffic due to the bushes and fence in the way)
  • After crossing the road onto another segregated bike lane, within 30 meters there are roadworks completely blocking the way for bikes, blocking all sight and forcing you onto the road again with NO indication of this happening when descending from the bridge
They try. They try so hard. For about 10 seconds, and then they just throw in the towel and fuck everything up. London, you're a funny guy.

The London Cycle Superhighways

The London Cycle Shitways
Dear readers of my blog, I present to you, London CycleShitway 3. See? It's not just a bicycle drawn on the road. It's got blue paint around it and a designation in large letters under it. See?

Truly the mark of not just a cycleway, not just a cycle highway, but a SUPER highway. Hey, pay attention, we're bending off to this way now!

No, we didn't have any paint left to give cyclists here right of way or... well, we didn't even have enough blue paint to finish the bike path. But we tried! Honest! Speaking of which, the path didn't really fit well in with the existing road and we really didn't feel like spending the extra 200 pounds to move the parking spots to the other side, so... uh...

Yeah, we decided just to reverse directions. Oh, and we decided to leave one lane out because, you know, paint shortage and all. By the way, SUDDEN INCLINE!

sorry, we couldn't fit something more reasonable. A 12% slope was the best we could do. In the middle of a completely flat section of the city. Besides, cyclists are all 25-year-old bike couriers with a death wish anyway. Right? Anyway, this was necessary to fit this AWESOME bike bridge!

did you blink? I guessed so, because obviously you were supposed to switch lanes on the bridge! Also, if you can just nudge left a bit more.... excellent! Now all the bikes go on the left, pedestrians on the right. Excellent.

sarcasm aside: yes, the lanes switch direction at some indeterminate place between before and after the bridge AND they merge into the one lane. This is a cycle superhighway! This is DESIGNED bicycle infrastructure meant to express London's commitment to be a better cycling city. This is effort!

So that pedestrian/cyclist crossing that was misplaced that I was talking about earlier? It's in CS3. See that cyclist? She's actually going straight over the road as the CS3 path goes - that crossing is completely cock-eyed. Also; no mention of CS3 on the crossing; no blue paint, no kiddie bicycle paintings on the road. Nothing.

Now it just continues over the middle of the road because who gives a fuck at this point anymore.

wait... wait a minute... This is actually half decent!


Ah, I knew you'd fuck it up within 100 meters. Good job, London. You've proven yourself again. In case it's not obvious: there is a right-angle crossing right in the middle of a cycle superhighway. Superhighway. What would you say if a motor vehicle highway suddenly made a right angle turn unannounced?

Well, at least you hired somebody over 7 years old for this drawing. You know there are templates for this? You don't need to draw them by hand!
Excuse the mediocre image quality; I didn't take my dSLR

And in true London fashion, just to tie it all together like a diarrhoea turd, the first of two dedicated bicycle traffic lights has bicycle masks for all of the little bottom lights and two of the top lights, but they forgot to put the sticker on the top red light. Also, the traffic light is broken and NEVER turns green.

By the way, at this moment of writing I'm in the international train from Brussels to Amsterdam. I look out of the window randomly. I see this intersection. This is not even the Netherlands, the road surfaces may be a bit scratchy but literally nothing in London even comes close to being as well-designed as this. Again, I'm stressing; random picture:
A view from the train bridge near Berchem station, Antwerp

Oh, you were still interested in the CS3? It just ends. It. Just. Ends.

By the way, these photos are not cherry-picked locations many miles apart. You can take all these photos in a 5-minute time span (plus 3 minutes of waiting for the traffic light). It's mistake after mistake on this cycleway.

This image is in The Angel, Islington:

This is 50 meters of decent bicycle path. Transmission ends.


London city cycling infrastructure: not impressed in the slightest. Token effort at best, intentionally homicidal by the road 'designers' on average, entirely neglected and absent at worst. Every other country's capital where I've been has better infrastructure, and this includes China.

Now, don't get me wrong. I thoroughly enjoyed my time in London. I even enjoyed cycling, and it can be reasonably safe. I mentioned that most drivers are actually pretty good to us. I just can't accept that the financial capital of Europe, a city that should be an example to the rest of the world, can have such crappy infrastructure. I especially expected much more from a 'cycle superhighway'. I know that CS3 is known to be one of the worst examples, but still: don't call it a superhighway! Don't get people's hopes up, but this will do nothing but invite ridicule.

I'll come back to London one day, and I'll certainly take my bike then. Let's hope things are better by then!
A photo I took near Lijang, China when I visited there in 2005. The cars are decripit, it smells really bad and the area is very poor. But they sure know how to maintain a road!

Why fuel cell cars don't work - part 4

Door mux op maandag 23 maart 2015 10:45 - Reacties (31)
Categorie: -, Views: 5.546

We have arrived at the final station of fuel cell cars. This is the end. We have seen how hydrogen is quite an annoying fuel to use in many respects and how other fuels have their share of drawbacks as well. We've gone over the technical details of a bunch of fuel cell types. I have even talked a bit about the economics of it all. Today I want to talk about what I think will be the future, and what we as a society should strive towards. A bit less technical details, but I hope it will be interesting nonetheless!

This is a continuation of a blog series, here are Part 1, Part 2 and Part 3 if you haven't read those yet!

This is an extremely long, in-depth blog series, so I'll start by giving you a summary. This summary will exist at the top of every part of this series. If you're interested in the technical details, please do read on and make sure to come back for the next parts.

First of all, HFC cars are perceived to be a good bridge between fossil fuels and full electric because:
  • You can still fill up like you do with a gasoline or diesel powered car
  • The mileage you can get out of hydrogen is perceived to be more adequate than what you get from batteries
  • Hydrogen fuel cells are thought not to wear out as quickly as batteries (or conversely, batteries are thought to wear out very quickly)
  • Hydrogen as a fuel is perceived to be a relatively small infrastructural change from gasoline and diesel
  • Hydrogen is perceived as a cleaner solution than gasoline, diesel or natural gas
In reality,
  • You cannot fill up like you do with gasoline or diesel. It is actually pretty ridiculous how hard it is to fill up a HFC powered car
  • You won't even go 100 miles on current tech hydrogen tanks that are still safe to carry around in a car
  • Fuel cells wear out crazy fast and are hard to regenerate
  • Hydrogen as a fuel is incredibly hard to make and distribute with acceptably low losses
  • Hydrogen fuel cells have bad theoretical and practical efficiency
  • Hydrogen storage is inefficient, energetically, volumetrically and with respect to weight
  • HFCs require a shit ton of supporting systems, making them much more complicated and prone to failure than combustion or electric engines
  • There is no infrastructure for distributing or even making hydrogen in large quantities. There won't be for at least 20 or 30 years, even if we start building it like crazy today.
  • Hydrogen is actually pretty hard to make. It has a horrible well-to-wheel efficiency as a result.
  • Easy ways to get large quantities of hydrogen are not 'cleaner' than gasoline.
  • Efficient HFCs have very slow response times, meaning you again need additional systems to store energy for accelerating
  • Even though a HFC-powered car is essentially an electric car, you get none of the benefits like filling it up with your own power source, using it as a smart grid buffer, regenerating energy during braking, etc.
  • Battery electric cars will always be better in every way given the speed of technological developments past, present and future

The shape of things to come

I have alluded to this before: I am a very big fan of CGP Grey and his robot future. Even without an impending singularity - the point at which computers have similar cognitive capabilities to humans - it is very clear that self-driving transportation machines - autos - are here, they work and they will only get better, cheaper, safer and more popular. As much as car enthusiasts will try to tell you otherwise, most people use cars to get from A to B and not much more. It is unnecessary to have to drive yourself. It is tiresome, you are very limited in speed because of the unstable human-car control system, it uses roads exceedingly inefficiently and people tend to make a bunch of mistakes on every journey, long or short. Computers are better. Self-driving cars will dominate in the near future. I'm betting money (and I have bet Reddit Gold on this with some random internet stranger already) that a significant proportion of human-transport will be autonomous within five years, and a majority will be driving in autonomous cars in 10 years.

The pace at which autonomous vehicles in general are improving is mind-boggling. Just a few years ago (2011) Caterpillar started a pilot program with self-driving hauling trucks in a single mining operation; today the majority of copper mining haulers are autonomous. In the entire world. Just a year ago the Google self-driving car clocked in its 700 000th mile without incidents. The two incidents it did have? Of course, they occurred when the car was being driven by a human.

Back in April 2014 it couldn't handle rain and certain traffic situations very well - this has mostly been solved by now. In less than a year. It was already better than most human drivers, now it's roughly as good as the most experienced drivers in the world working at the top of their game - but it can sustain this level of competency all the time. And other companies are competing as well; all software/tech companies of course. Because autonomous vehicles are not a car problem, they are a software problem.

I can't place the accent of this narrator. She sounds strange, doesn't she? Is she a robot, too?

And think about it. Cars are stationary almost 95% of the time. They not only cost a bunch of money to buy and operate; they take up the majority of valuable space in cities. Roads and parking spaces take up a giant proportion of urban land area. This doesn't have to be. A single autonomous car can service dozens of people, having to stop only to recharge once in a while. Even if this autonomous car needs to contain a million bucks worth of electronics and batteries - which it doesn't, but just for the sake of argument - it would still be significantly cheaper than everyone having to have their own car. There are very large economic incentives to make this a reality as soon as possible, both on the service side as well as on the user side. And as we know, economics ALWAYS win. In the future, cars will not have to be ubiquitous. The landscape doesn't have to be littered with these scars upon the name of engineering.

This is not to say that cars as a status symbol or cars for fun driving will go away. Of course people will have hobbies. But they will be hobbies, in places where people do hobbies. On tracks, on designated road spaces. Not on the main traffic arteries.

BEVs are the future
Battery electic vehicles, or BEVs, are going to be the dominant type of car in the future. The two biggest reasons for this are:
  1. EVs give practically unlimited design freedom
  2. Electric drivetrains are the most efficient and most versatile drivetrains
Let me expand on this a bit. EVs - whatever actual energy carrier you use - can be made in any shape because unlike internal combustion engine cars, the actual engine is tiny and can be placed very near or - in the near future - inside the wheels. This is then connected by means of wires to the energy source, which can be anywhere and in any shape. This frees up a bunch of space in places that were traditionally reserved for essential drivetrain stuff. All the engine gubbins in front can be transformed into a much more effective (and shorter) crumple zone and storage space. The torsion frame in front of/underneath the car can be greatly reduced, as the full engine torque doesn't need to be transferred through the car frame anymore. You still need to fit in a large amount of batteries or something like a fuel cell, but this can be positioned much more favourably. The Tesla Model S demonstrates this design freedom to a great extent - even though it's only a very early EV design.
OK, the Model S is a giant car, but despite its performance-driven nature it still has more luggage space than most 'practical' family cars

But design freedom goes much further than just the physical. Electric drivetrains have much more ideal and predictable properties. Their torque-speed curves are basically straight lines. Power control is immediate and precise, with greatly reduced drivetrain inertia to slow down the response. This makes EVs much easier to use for self-driving cars than combustion engine cars.

The versatility of electric drivetrains stems from the fact that any type of fuel or even fuel-less energy sources can be made into electricity quite efficiently. You don't have this kind of versatility in gasoline or diesel powered cars. Even though those two liquids are chemically strikingly similar, you can't fill up either car with the other fuel. Let alone use coal or nuclear pellets. This leads to all kinds of perverse economic constructions like the OPEC; who have the freedom to put any price on their scarce resources because of nothing else than geography and culture. With electric cars, there is basically infinite competition from anybody with free view of the sky. Talking about solar.

The solar singularity is here

Solar energy is taking off like nobody's business at the moment. Fueled by a 15-25% year-on-year price drop over 8 years now, a system that would have been economically unviable in 2007 (§4/Wp) is now better than grid parity (§1/Wp) and still dropping double-digit percentage points per year. Actually, module price drops have been accelerating, with installation and electronics costs seeing only minor cost reductions (which is the most important reason for prices not dropping faster). Energy from new static solar installations is approaching §0.05/kWh in the Netherlands, and about §0.035 in southern Europe. This is considerably cheaper than energy from any other power source, and there is no technical reason that stops prices from dropping further considerably in the near future. Of course, the sun only shines during the day, so solar energy is no solution for the general energy problem. But it sure as heck is a great way to charge your electric car for almost-free.

In general, total vehicle ownership costs can be broken down as:
  • 35-40% depreciation
  • 30-35% fuel
  • 25-30% other
In the Netherlands, fuel is actually a significantly larger part of the entire equation, as the Dutch drive quite a lot and fuel is relatively expensive. Fuel costs clock in at a little more than 40% here. Imagine that part being basically free. Of course, there will be costs associated with electricity distribution and other practical concerns, but the energy itself is free. You can even put solar panels on the car (or car manufacturers can integrate them), providing up to 20% of the energy required for driving. The electric drivetrain and batteries (or some other directly-charged-by-electricity) are essential to this kind of tech working. This is the great versatility promise of battery-electric cars. It would be much, much harder to for instance do on-board water splitting in a fuel cell powered car with those same power sources, and because of the inherent inefficiency of such systems you would need about twice the energy to get just as far. Another way of saying this is that solar BEVs are a very short-cycle way to get energy for driving.

How 'free' is free? At the moment, residential installations in the Netherlands and Germany hover between §1 and §1.50/Wp; at the roughly 1500 hours of insolation we get per year this yields 1kWh/Wp per year. The economic lifetime of such an installation is 20 years, with typical maintenance costs hovering between §0.10-§0.30/Wp over the entire installation period. This means that you pay between §1.10 and §1.80 for 20kWh - effectively. About §0.055-0.09/kWh. Residential installations have relatively good pricing as there are no costs associated with land lease or ownership, nor infrastructure costs. Costs of commercial installations are significantly higher because of this.

For the next 10 years, depending on who you believe solar prices will at least halve. Complete installed price for an economically sensible installation (small installations will always have more overhead). This translates to residential prices of §0.03-0.04/kWh over lifetime. This is about a third of the price of utility electricity in the US and about 1/5th to 1/9th of the price of electricity in Europe. For a typical vehicle, fuel costs would go from about §0.10/km (15km/L @ §1.50/L) to about §0.005/km (125Wh/km @ §0.04/kWh). Of course, as demand for oil-based fuels reduces prices will go down significantly, but it is unlikely that ICE car fuel prices will ever be able to match solar electricity prices.

So, about batteries
Lots of Tesla Model S pictures in this post.

The reason why people like fuel cell cars and don't like batteries is the public perception that batteries don't get them far enough and cost a lot. This is true to some extent, certainly at this moment. Vehicle range of BEVs - affordable ones (not looking at Tesla) - is pitiful compared to even the crappiest ICE car. However, range anxiety - as this is called - is not actually warranted in most cases. And because of the charging versatility of cars, it's not likely to be a problem for BEVs either way in the future.

First of all, any range argument can be quite easily counterargued by saying that depending on where you live, between 90 and 99% of all vehicles can be functionally replaced by a 100-mile range BEV without any travel move being impacted by battery range. That is to say: the vast majority of cars never drives more than 100 miles in one go in their lifetime, and most of the long-distance driving is done by a small group of drivers in specific cars. Which can use something else. That's fine. These changes don't happen overnight.

However, I am not saying batteries aren't actually limited. As much as battery technology has advanced in the last 10 or so years because of the sharp rise in lithium ion battery production, by far most improvements have been process tech and cost reductions. Lithium ion batteries are going to be, barring any very fundamental breakthrough, limited to about <300Wh/kg. Why? This is actually a very fun calculation. Lithium ion batteries are, like fuel cells, reduction-oxidation or redox cells. The two technologies aren't that dissimilar. As such, batteries store charge by ionizing lithium and some other oxidizer at the electrodes in the battery.

Lithium can 'store' one electron per atom, so you need 6.24 x 1018 lithium atoms to store one coulomb of charge. There are 6.022 x 1023 atoms in one mole of lithium, which stores 96508 coulombs. One mole of lithium weighs 6.94 grams and has a half-reaction redox potential of -3.05V. This means that 6.94 grams of Li can store E = Q x V = 96508 x 3.05 = 294kJ or 81.8Wh, which gives us the incredible energy density of 11781Wh/kg for lithium as a chemical energy carrier.

So... why... what!? This is awesome! This is about on par with other chemical energy sources like fossil fuels. Well, the devil here is in the phrase 'half-reaction'. This is only half of the story. For a redox reaction you need both the reduction reaction (which is the ionization of lithium) as well as an oxidation reaction to happen. And that's where things go wrong pretty quickly. But, just to give quick closure to this chapter: theoretically, a battery with only a lithium anode can exist. It would use oxygen from the air as the oxidation agent, and as such this is called a 'lithium air'-battery. As of now this is a fairytale; there are numerous practical problems with actually making this a reality and there is absolutely zero outlook on an actual working lithium air battery within the foreseeable future.

In actual practical lithium ion batteries, we cannot use lithium metal directly. The anode is usually made from a lithium salt, for this example we'll be looking at LiCoO2. The second side of the equation, the one missing above, is generally performed by carbon in the form of graphite. This is generally called the cathode (although more accurately we should be referring to the electrically positive and negative electrodes, as the two sides switch roles whether they charge or discharge). For each electron 'stored' in the reaction, we need to lug around one carbon atom, one cobalt atom and two oxygen atoms. These weigh 12.011 + 58.933 + 2 x 15.999 (+6.94 for the Li) g. This accounts for a 15.83x increase in reagent mass for the same amount of charge, to get to a maximum theoretical energy density of 744Wh/kg. Unfortunately, even that is way too optimistic for any kind of future battery technology, as a couple of quite severe technical problems (e.g. short circuiting through dendrite formation) don't allow the electrodes to be so close to each other that they can quickly exchange ions. So we need to introduce a lithium ion-conducting electrolyte in between the electrodes, which necessarily increases the mass again. A few different types exist, from polymer membranes (hey, remember PEM fuel cells? These are surprisingly similar!) in lithium-polymer cells to lithium halogen impregnated paper-like electrolytes in the familiar round lithium ion cells found in e.g. the Tesla Model S. This is a surprisingly large contribution to the mass of a lithium ion cell, and limits the theoretical energy density to around 300-350Wh/kg.

As long as we make fully contained complete redox reaction pair batteries, i.e. recheargable lithium ion batteries, this is pretty much an unavoidable brick wall. With current generation battery packs peaking at about 175-200Wh/kg, the best possible improvement we will ever be able to make is about a twofold increase in capacity and that's it. In other words: battery packs in electric vehicles will necessarily always weigh a couple hundred pounds, whatever you do.

So why do I think batteries are not a dead end? Well, contrary to fuel cells, batteries have a pretty bright future as far as cost reduction goes. As battery production has ramped up, vehicle-grade battery packs have fallen from $450/kWh (2007, A123) to $140/kWh (2014, Tesla). With the raw materials being plentiful, relatively widespread and very cheap, the majority of cost goes into process tech and packaging. This is something that is very optimizable as production volume goes up. So even though weight can't necessarily be reduced that much, cost can easily halve in the next 5-7 years with some speculating that Tesla will announce a sub-$100/kWh price point this year already for its residential battery pack (battery only).
As you can see, this is a very big case of 'depending on who you ask'. Predictions vary quite wildly

There are still some concerns; some more important than others. Environmental concerns around battery production and the associated pollution of lithium mining are mostly unimportant; the amount of pollution generated by the considerably higher amount of fossil fuels required for ICE powered cars easily offsets this. Recycling is an increasingly hard problem as optimal battery technologies make it hard to recover materials from lithium ion batteries. Lithium in general has fairly poor recycling characteristics. But again; the environmental and user benefits have been shown to, even now that the technology is still in its infancy, still outweigh the environmental downsides of traditional vehicles. And there is no fundamental reason why EVs wouldn't become better in the future whereas fossil fuel use is a guaranteed dead end with unescapable environmental concerns on both short and long term.

But our infrastructure isn't up to snuff!
Another often heard problem with electric cars is that our infrastructure will not hold up to the high peak demands from charging cars. This is slightly true, but not likely to cause big problems in the long run. This kind of runs into a ocmmon misconception in that cars/mobility are a huge drain on resources/large cause of greenhouse gas emissions. It's kind of sad that I have to touch on this so late in this blog series, but: cars ain't that bad. Yes, certainly, cars are incredibly inefficient and guzzle seemingly enormous amounts of energy from unsustainable sources. But if we look at the total CO2 output of all of humanity, all transportation put together accounts for only about 11-14%. Of that, only about 35% is embodied in personal transportation by passenger car. The rest is trucking, commercial use passenger cars, aviation, shipping and light motor vehicle use. That is: only about 4% of all CO2 emissions can be attributed to cars. If we look at actual pollution, cars amount to almost nothing. Actually, tire wear and emissions from the production and disposal of cars is a larger weight on the environment than the actual use of cars.
One of the clearer illustrations I could find of the relative impact of EV charging on UK infrastructure

This parallels electricity use by electric cars. If all of our cars suddenly become BEVs, electricity use won't increase tenfold. It wouldn't even increase twofold. Of course, even a twofold increase in capacity does require some extensive retooling, especially in third-world countries like the USA where the electrical grid is woefully undermaintained. But the investment in infrastructure to make this happen is absolute peanuts as compared to the infrastructure changes we'd need for, for instance, hydrogen or methanol fuel cell cars.

The real challenge here is not that we need to build twice the infrastructure we have; it is that we should decide right now how we intelligently charge our cars. If everybody hooks up their car to a charger when they get home, the peak demand will increase dramatically. If instead smarter charging strategies are used - spreading the load over for instance an entire night - the infrastructural problems will be negligible.

Mythbusting: Fuel cells are a conspiracy by Big Oil

Right at the end of this blog series I'd like to tie up some loose ends in the general discussion of fuel cell vehicles. One of the most important observations about fuel cells are that at least for the first few decades, the majority of hydrogen production will have to be done by reforming natural gas. With Big Oil - the OPEC, Russia, Nigeria, Norway, Brazil and the US - making so much money off of oil production, they don't want to see us going to free energy. So they invent something that looks and smells 'green' but actually isn't: fuel cells. That way, they can keep selling us oil, in the form of reformed natural gas. Sounds like a credible conspiracy? I'm not buying it.

For one, the costs and technical challenges that hydrogen production, storage and sale encompass are astronomical. Big oil has had decades of hundreds-to-thousands of percents of profit margin on their oil products to subsidize the oil and gas infrastructure we have today. Hydrogen is fundamentally incompatible with most of this infrastructure, but the catch is: there is no guaranteed market, no large-scale dependence and no revenue stream to bootstrap such a big infrastructure project.

Another big red flag is the fact that traditional oil companies have historically shown very little interest in this market. All of the sponsors for our hydrogen-powered race karts Forze I and Forze II were technology and tool/hardware companies. Only one or two companies can be tangentially associated with the oil industry; DSM being the only large one (who supplied resin for the carbon fiber body parts). Other international teams as well as the Formula Zero organization saw barely any interest. The hydrogen supplier was Linde, a company who mostly supplies fertilizer companies and other industrial purposes. And this goes for most of the hydrogen fuel cell market; the main players are struggling medium-sized companies like Hydrogenics and Nuvera who, if anything, have only seen a lot of competition from Big Oil.

If hydrogen fuel cells are going to become a big thing in the future, I don't expect oil companies to have much of anything to do with it. A hydrogen economy requires radically different thinking from traditional oil and natural gas-based industry.

Toyota's recent decision to go all-in on FCVs

Another interesting thing that has happened very recently, is Toyota's announcement of the Mirai FCV:
Today, we are at a turning point in automotive history.
A turning point where people will embrace a new, environmentally-friendly car that is a pleasure to drive.
A turning point where a four-door sedan can travel 300 miles on a single tank of hydrogen, can be refueled in under five minutes and emit only water vapor.


Our fuel cell vehicle runs on hydrogen that can be made from virtually anything, even garbage!
It has a fuel cell that creates enough electricity to power a house for about a week.


The name we’ve given to our new car is Mirai, which in Japanese means “future.”
We believe that behind the wheel of the Mirai, we can go places we have never been, to a world that is better, in a car that is better.
For us, this isn’t just another car. This is an opportunity – an opportunity to really make a difference. And making a difference is what Toyota is all about.
The future has arrived. And it’s called Mirai.
You have to admit, that is some serious tech porn

This announcement was followed in January of this year with an opening and royalty-free licensing of a whole lot of fuel cell patents. This seems to be a large swing in Toyota's R&D, which of course produced battery-ICE hybrids like the Prius. A lot of people go so far as to say Toyota is going all-in on fuel cells and abandoning BEVs completely.

However, reading into it a little more deeply, things start making a lot more sense. Of course, at the current state of technology Toyota would not be able to make a production FCV. For all intents and purposes, the Toyota Mirai is a specialty car that serves more as a public technology demonstration than something you can properly buy. Production volume is announced to be 700 in 2015, going up to 3000 in 2017. For comparison, Tesla is now producing 50 000 Model S EVs annually, and they are an absolutely microscopic car company. Typical production volume for cars nowadays is in the hundreds of thousands.

Toyota aren't bluffing though. They have serious, innovative technology under the hood and I do believe they hope FCVs will be a big thing in the future. As far as I'm concerned, the Mirai is only a very minor step up from the concept that the FCX Clarity was a couple of years ago. They sure aren't going all-in. The Mirai is testing the waters and seeing if this fuel cell thing catches on or if BEVs will prevail. By opening their patents they hope for more competition in the fuel cell camp to fight off BEVs.

I don't think they will succeed. Judging from the information released, they haven't found any solutions to the fundamental problems with FCVs. They haven't made the Mirai magically less complex and they haven't sufficiently reduced platinum loading in the stack to allow for sufficiently large production volume. Maybe they have another trick up their sleeves, but I doubt it. They even doubt it because they're not actually putting much money at risk with their comparatively tiny production volume and R&D budget.

In any case, start stocking up on platinum. Prices are sure to go up as fuel cells become a hot topic once again.

We're done, we are at the end of a journey through the tech inside fuel cell cars - and other future cars. I don't want to leave you with a feeling of negativity. Yes, I am saying that fuel cell cars don't work, in any shape. I'm saying that batteries are better, in every way.

This part of the blog was futurology, i.e. talking about things in the future with a little bit of scientific backing. It's not complete hand-waving. I've discussed essentially two possible futures:
  1. Either cars as we know them are going away completely, being replaced by about 1/10th the amount of completely self-driving, non-owned transportation service autos
  2. or car ownership will remain, BEVs will dominate because of their significant economic, complexity and comfort advantages over all other alternatives
Either way, fuel cell vehicles make little sense. For future number 2, it would just require too much infrastructure for very little benefit to the end user. Cars would have to get more expensive and it will take a very long time before future 2 can be a reality.

Future 1 can be a reality in 5 years. This year already, multiple auto makers have announced production (i.e. you can buy them!) 90% self-driving cars. Tesla and Volvo are at the forefront here, the rest will certainly follow shortly. Uber has announced they want to move in the direction of a completely self-driving car fleet in 5 years. This is possible. The question is: will these self-driving cars be a minority or will it be disruptive?


I am just a dude, I am not an expert in basically any of the fields I have spoken about. I know enough about them to make some general statements and do some general back-of-the-envelope calculations, but a lot of the nuances are at best slightly vague and at worst completely unknown to me. I've been corrected multiple times on my application of diffusivity and catalysts. Not in ways that undermine my point, but just to show: this is not gospel.

I hope you enjoyed my extensive treatise of fuel cell cars and my short overview of battery electric cars. Again, I don't make a single dime on these blogs, I do these because I adore the subject matter. I realize that even with 120kB of text I still haven't even scratched the surface, let along the dozens of handwavy statements and predictions I made without proper scientific evidence to them. Leave a comment if you found a problem, error, false claims or if you just want to engage in a discussion about any of the points I raised. Don't agree at all? Do you have good reasons? Write your own blog post! Be sure to leave a link here.

Because this is actually important stuff.

Why fuel cell cars don't work - part 3

Door mux op maandag 16 maart 2015 12:02 - Reacties (17)
Categorie: -, Views: 6.982

We are at part three of the fuel cell blog series, and it's high time we look at other things than hydrogen going into a PEM fuel cell. Other things like ethanol, methanol, split cycle vehicles, solid oxide fuel cells, using hydrogen in combustion engines and direct solar photolysis. And you'd be surprised in how many ways all these technologies have major to insurmountable problems. Probably the reason why we don't see them used at all.

This is a continuation of a blog series, here are Part 1 and Part 2 if you haven't read those yet! This series concludes in Part 4.

This is an extremely long, in-depth blog series, so I'll start by giving you a summary. This summary will exist at the top of every part of this series. If you're interested in the technical details, please do read on and make sure to come back for the next parts.

First of all, HFC cars are perceived to be a good bridge between fossil fuels and full electric because:
  • You can still fill up like you do with a gasoline or diesel powered car
  • The mileage you can get out of hydrogen is perceived to be more adequate than what you get from batteries
  • Hydrogen fuel cells are thought not to wear out as quickly as batteries (or conversely, batteries are thought to wear out very quickly)
  • Hydrogen as a fuel is perceived to be a relatively small infrastructural change from gasoline and diesel
  • Hydrogen is perceived as a cleaner solution than gasoline, diesel or natural gas
In reality,
  • You cannot fill up like you do with gasoline or diesel. It is actually pretty ridiculous how hard it is to fill up a HFC powered car
  • You won't even go 100 miles on current tech hydrogen tanks that are still safe to carry around in a car
  • Fuel cells wear out crazy fast and are hard to regenerate
  • Hydrogen as a fuel is incredibly hard to make and distribute with acceptably low losses
  • Hydrogen fuel cells have bad theoretical and practical efficiency
  • Hydrogen storage is inefficient, energetically, volumetrically and with respect to weight
  • HFCs require a shit ton of supporting systems, making them much more complicated and prone to failure than combustion or electric engines
  • There is no infrastructure for distributing or even making hydrogen in large quantities. There won't be for at least 20 or 30 years, even if we start building it like crazy today.
  • Hydrogen is actually pretty hard to make. It has a horrible well-to-wheel efficiency as a result.
  • Easy ways to get large quantities of hydrogen are not 'cleaner' than gasoline.
  • Efficient HFCs have very slow response times, meaning you again need additional systems to store energy for accelerating
  • Even though a HFC-powered car is essentially an electric car, you get none of the benefits like filling it up with your own power source, using it as a smart grid buffer, regenerating energy during braking, etc.
  • Battery electric cars will always be better in every way given the speed of technological developments past, present and future

Fuel cell types

In all of the previous posts, all I talked about was PEM or Proton Exchange Membrane fuel cells. But fuel cells, like everything in engineering, exist in many different variations. A fuel cell is defined very simply: an apparatus that converts chemical energy into electricity directly. So using a combustion engine to drive an electric generator doesn't qualify: the conversion must fundamentally be direct. No cheating!

For an extended overview of fuel cells, I highly recommend (again) Wikipedia's great article on fuel cells. I will only go over them very briefly.

Alkaline fuel cell
The AFC used in the NASA Apollo missions. Image from

This is the oldest and most developed type of fuel cell; it has been used in multiple NASA missions (including the Apollo missions and ISS). It is also, by quite a margin, the most electrically efficient type of fuel cell (excluding external systems). In an AFC, pure hydrogen and oxygen are led into two chambers separated by a porous material that is drenched in an alkaline solution (most often potassium hydroxide, KOH). At the electrodes, the following half-reactions occur:

2 H2 + 4 OH- -> 4 H2O + 4 e-
O2 + 2 H2O + 4 e- -> 4 OH-
AFC cell schematic (very simplified). Image originally from Wikipedia

This is different from any other hydrogen fuel cell in that the hydroxide ions (which cause alkalinity) are necessary for the reaction. This reaction produces water at the anode and consumes (half the) water at the cathode, so the most straightforward construction involves an electrolyte that can flow from anode to cathode as well as outwards (to dispose of the excess water). In older designs (and this is a great example of the age of these kinds of cells), asbestos was often used as an electrolyte matrix.

AFCs can be made (and work pretty well) without catalysts and with exceedingly cheap materials, while still attaining pretty good efficiency. This is the fundamental reason why AFCs are still being used (on quite a large scale) for certain applications. This is, however, where the good news ends.

The problem is: hydrogen in water (aqueous solution) does not often spontaneously dissociate to form loose protons and electrons. You wouldn't get large amounts of power from a fuel cell unless you make the fuel cell house-sized. This is where catalysts come in, and AFCs become a lot less interesting. I'll go into how catalysts work in the subchapter 'the platinum problem'. Suffice it to say, the only catalyst that works well in hydrogen and all other fuel cells at low temperature is platinum. You need quite a bit of platinum for a decently sized fuel cell; about 3g/kW for PEM and about double that for AFCs*. Now, this is a problem that all fuel cells have, so it's not unique.

What is unique to AFCs, is the problem they have with using normal air as an oxygen source. Air is composed of about 21% oxygen, 78% nitrogen, a little bit of water and trace amounts of other compounds. Unfortunately, because the electrolyte in AFCs is some alkali hydroxide (potassium, calcium, etc. hydroxide) it will react very favourably with CO2 in the air, producing potassium carbonate or calcium carbonate - chalk-like substances. These are not very soluble in water and will, over time, clog up the electrolyte. This reduces performance and puts a relatively short lifetime on AFCs.

The tendency of AFCs to irreversibly deteriorate over time unless you use either large/expensive scrubbers or a separate tank of pure oxygen makes AFCs pretty much uninteresting compared to more modern PEM fuel cells.

* Relatively recently, platinum-less catalysts for AFCs have been developed. Currently they are still more than an order of magnitude worse than platinum, but this is already much better than any other non-platinum catalyst developed so far. The stack cost and size is still very much unsuitable for vehicles, but it may be a solution for stationary AFCs.

Hydrogen PEM fuel cell
Proton Exchange Membrane (PEM) fuel cells use, like most types of fuel cells, the following half-reactions to get from hydrogen+oxygen to water+electric current:

H2 -> 2 H+ + 2 e-
1/2O2 + 2 H+ + 2 e- -> H2O
PEM (and most other) fuel cell overview. Image courtesy of Wikipedia

The electrons are, again, siphoned off via the catalyst surfaces and substrate to power whatever is attached and the hydrogen ions (=protons) are moved through a very, very special material called a proton exchange membrane. This is a fairly revolutionary material: a polymer that effectively blocks everything except protons. It doesn't even conduct electrons. Well, it kind of does a little bit (as well as hydrogen and oxygen), but it works remarkably well nonetheless.

PEM fuel cells can be made extremely compact. Because the membrane material can be made almost arbitrarily thin and the catalyzing surfaces are very thin as well, the only limit on miniaturization of PEM fuel cells is water, reagent and heat management. This makes it theoretically possible to make PEM fuel cells that power cell phones.
Just as an aside: how cool is this. We worked on the fuel cell race kart in the same hall where the Solar Boat, Nuna, Ecorunner, Formula Student and Wasub were being worked on. As well as a bunch of other smaller student projects like DARE, a record-setting rocket project I also contributed to

PEM fuel cells are low-temperature cells, meaning they operate typically below the boiling point of water at near-atmospheric pressure. Because there is effectively no electrolyte solution (the 'electrolyte' is the PEM itself), these fuel cells are fairly insensitive to poisoning by 'normal' trace gases in the atmosphere, so not much scrubbing is needed to purify the incoming oxygen from the air. Because they operate at almost-ambient conditions, start-up times are very reasonable (in fuel cell terms), to the point that it is practically feasible to put them into cars. They're not quite as instantaneous as batteries or gasoline powered cars, but they can be and they are really close already.

There are problems, though. Aside from the platinum problem, two issues that have not seen any improvement in the last 5 years or so are:
  • Gas and electron leakage over the membrane. Hydrogen is a tiny molecule that will eventually find its way through the membrane. But also oxygen leaks through. Because of the relatively large surface area of the membrane, although it is a fairly good insulator the membrane still leaks electricity significantly (up to about 10 percent of the losses in PEMFCs are due to this phenomenon).
  • Activation losses on the oxygen side. Even though platinum is a great catalyst for the hydrogen-side half-reaction, it is not super duper on the oxygen side. It is still the best we have, though. A large part of the losses in PEMFCs are due to the platinum catalyst on the oxygen side. Again; science would love to find a new, cheaper, better catalyst - but has failed to so far. The main issue here is the strong oxidative strength of oxygen, which rules out almost all non-precious metals.
Besides these issues, water and heat management are both significant issues that cause current-generation fuel cell stacks to be quite big. Most of the heat in a PEM fuel cell is produced at the catalyst on the oxygen-side, however in order to maintain good reaction slew rate (i.e. throttling performance) as well as good efficiency, the fuel cell may not heat up much and the temperature should be roughly equal over both reaction sites. This means that high power densities are most likely impossible, so you need a relatively large fuel cell. Water management has similar issues - with especially the membrane needing just the right amount of water in it to function well. So why didn't I put these issues in the nice list above? Well, even though these two issues are, at the moment, by far the biggest practical roadblocks for PEM fuel cell application, I don't view them as important to the discussion because the issues aren't fundamental to the issues with PEM fuel cells. These are, in my view, just engineering challenges that will be solved given enough time and money. The diffusivity and catalyst issues are much less likely to ever be solved.

If you're interested about how this translates into a practical fuel cell, Here is the technical product specification of a fairly widely used fuel cell stack, which at the moment retails for about §40 000-§50 000 (with subsidies).

Phosphoric acid fuel cells
This bus runs on a PAFC. Image courtesy of Institut fŁr KFZ Aachen.

Phosphoric acid fuel cells (PAFCs) are extremely similar to PEM fuel cells, but the proton exchange membrane is replaced by molten phosphoric acid (H3PO4), and the fuel cell works at slightly elevated temperatures of about 200C. PAFCs are advantageous as compared to PEM because they don't need to hydrate the electrolyte (meaning much longer electrolyte life) and the produced water is in the form of steam, which can be used in a small steam turbine or other cogeneration plant to recoup the thermal energy from the exhaust. Maybe even more interesting is that it can run natively on some simple hydrocarbon fuels (most notably methane and methanol), as the electrodes nor the electrolyte are poisoned by CO2. PAFCs, because they run at higher temperatures, also have an easier time getting rid of their heat so the power density (output power per kg) can be better than PEM fuel cells.

All this makes PAFCs more efficient and easier to implement than PEM fuel cells. So why aren't we using them as a prime candidate for cars? Well, the big problem is that they need to run at higher temperatures (=long startup time) and even though power density of the stack itself is very good, the supporting systems are so big that complete system power density (and total efficiency) is actually quite low. It's still compact enough to be used in things like buses and trucks, but most of the applications are in static electricity generation. After AFCs and PEMFCs, PAFCs are the most widely implemented fuel cells.

High-temperature (auto-reforming) fuel cells
Then there is a whole sub-class of very high temperature (typically 500-1000C) fuel cells, with the most well-known type being the Solid Oxide Fuel Cell (SOFC). To start off with the biggest reason why these will probably never be used in cars: the start-up times are in the order of hours, not minutes. A PAFC can start up in about 5-15 minutes, a decently sized SOFC reaches equilibrium after about an hour and will consume a very significant amount of energy during start-up.
A schematic representation of a solid oxide fuel cell. Image originally from Wikipedia

If we just look at the technology: SOFCs and other high-temperature fuel cells are again similar to PEMFCs and PAFCs in that they have two electrodes with an electrolyte in the middle, but in this case both the electrodes and electrolytes are generally ceramic materials. Because they operate at much higher temperatures, just because of thermal energy there is enough activation energy in the fuel inlet to transform hydrocarbon fuels, through gas reformation, into hydrogen (given the appropriate catalysts are present). Combined with the even better immunity against electrode and electrolyte poisoning, this means that basically any kind of volatile hydrocarbon fuel can be fed in and will generate electricity remarkably efficiently.

SOFCs and other high-temperature fuel cells are increasingly being considered for applications like back-up power supplies because of their improved energy efficiency and higher reliability for intermittent operation as compared to diesel generators.
One of a couple Siemens Westinghouse co-generating SOFCs. This particular one outputs 250kW, using the waste heat from the stack to power a turbine, which yields very impressive efficiency (about 50% total system efficiency, including reforming losses which is almost 10% more than a traditional coal-fired system). Note the immense size of the installation, even though its power output is only a couple times more that of a typical car engine

The platinum problem
So, I talked about this already in the AFC section, but this is a really big problem. People are worried that lithium will run out for battery electric vehicles - well, we know for a fact that platinum will run out at leastan order of magnitude sooner.

Catalysts are materials that have a certain shape which allows reagents - materials that want to react together to form some other compound - to do their thing very well. If you have reagents just floating around in a liquid, the chance is fairly small that they will bump into each other and perform the desired reaction. More importantly, in the case of fuel cells we don't just want them to perform a reaction: we want to siphon off those electrons that are involved in the reaction. If the electrons just start floating around in a liquid they will most likely find an undesired alternative way to react and reduce reaction efficiency.

Platinum has the unique property that it is an ideal 'place' for hydrogen molecules to sit down, split up and become H+ ions (which are bare protons). In the case of fuel cells, the other side of the reaction (the oxygen reaction) is also catalyzed particularly well by platinum, although other types of catalysts exist for this half-reaction. Catalysts are not active parts in the reaction; they are merely energetically favourable sites of reaction. An important property of catalysts is thus that they have very large surface areas.

Without platinum catalysts, the reaction rate in fuel cells would be many orders of magnitude slower. Depending on temperature and pressure, the reaction rate is at least 105 higher with a Pt catalyst.

So, how much catalyst do we need? Can we improve on this? Currently, platinum consumption is about 2-3g/kWe (i.e. output electrical kilowatts). This means that a typical electric car (35kWe) will need about 70-105g of platinum. Platinum prices have skyrocketed in recent years because of its popularity as an investment metal, and sits at around 37000 USD/kg. Just the raw platinum metal in a fuel cell (that is rather small, but sufficient for most purposes) costs $3000. But that's not all.

The most important aspect of a catalyst is its reaction surface area. To improve on this (and subsequently reduce required platinum consumption), most fuel cells today use tiny (2-5nm diameter) platinum nanoparticles seated on a carbon (graphite) substrate as catalyst. This is what allows for the relatively small amount of platinum catalyst in fuel cells. Not much possibility for improvement on this front, unfortunately.

Another, as of now, insurmountable problem is the actual production capacity of platinum in the world. Production at the moment is at an all-time high, but cannot keep up with demand. We're producing about 150 000 - 200 000 kg of platinum per year, and this is not expected to be able to go up. Half of this is already being used by, mostly, ammonia production and diesel exhaust scrubbers. The rest is split between investment and jewellery.
World production of platinum 1990-2015. Image courtesy of Wikipedia
Platinum prices 1992-2012. Image courtesy of Wikipedia. 1 troy ounce == 30g

We're producing about 50 million cars per year in the world. If all these cars were to run on fuel cells, we would need 5 million kg of platinum per year. This is physically impossible. With the current state of technology, even producing 1 million fuel cell cars per year would be impossible. But wait, we can recycle the platinum from fuel cells, right?

Yeah, absolutely. With about 50-75% efficiency. The big problem here is that in order to use amounts of platinum that a mortal human can pay for in a fuel cell, we need to use either platinum nanoparticles or other microstructured forms of platinum to generate the surface area needed for the reaction to take place. As a fuel cell is used, some of these tiny particles erode from the electrodes and wash away, either through the exhaust or into the cooling system. The platinum particles that stay put are still extremely tiny and during the regeneration process, they cannot all be recovered. Of course, this can be improved upon and platinum recovery from fuel cells is hardly a big industry right now. But even in applications where platinum exists as a gauze or sponge - all in one place, easy to recover - the recovery rate is still usually less than 95% (higher for very large industrial gauzes). This means that we will still never be able to produce anywhere near the amount of cars we produce right now using fuel cells.

Even if platinum consumption can be reduced in the future - even if it could be reduced tenfold - the extremely limited supply and complicated substrate structures will still mean that electrodes in fuel cells will be a majority cost of your car. As it stands, the electrodes in production cars account for about $10 000 of the final bill of materials. If fuel cells are ever going to become a thing, the single biggest hurdle that needs to be overcome is not a reduction, but a complete elimination of precious metals as catalysts.

Using other fuels

Hydrocarbon fuels in PEMFCs using on-board reforming
In the above chapters we've fairly generally gone over different types of fuel cells and although no amount of blog posts can truly convey the technical details of this all, the general gist is that only PEM fuel cells are small, light and fast enough to be able to be used in smaller vehicles. We've also established in the previous parts of this blog series that hydrogen is a bitch. The logical step is to look for other types of fuels to use in fuel cells.

This is a great idea because - and for some reason I haven't said this already - there are more hydrogen atoms in a liter of uncompressed gasoline than there are in a liter of hydrogen gas compressed at 700 bars. Gasoline is an awesome storage medium for hydrogen. As are most other hydrocarbon fuels; simply the fact that the molecules are larger and packed together more tightly (because they can exists stably as a liquid at low pressures) counts for a lot of convenience.

Well, there's a little problem. PEM fuel cells only accept hydrogen and no hydrocarbon fuels (e.g. natural gas, gasoline, diesel, methanol). Fundamentally, this is because the fuel side and oxygen side are physically separated by the PEM which only conducts hydrogen cores (protons), whereas for a full hydrocarbon redox reaction you want the carbon from the fuel side to go over to the oxygen side and bond with the oxygen to form CO2. So in order to make PEM fuel cells work with other fuels, there is no other option than to bring your own hydrogen reforming plant into the car.
Renault-Nuvera's on-board multi-fuel reformer concept from 2002-2006. This project was abandoned when Renault-Lexus went all-in on battery hybrids. Funny enough, just recently Toyota has partnered up with this exact same company.

Great. Like fuel cell cars weren't complex and inefficient enough already. This is obviously not an elegant solution, nor a cheap one. Hydrogen reforming is again something that depends quite a bit on (expensive) catalysts and needs a bunch of extra supporting systems. However, you do save on having to store hydrogen, which takes away a lot of problems. But... let's do some quick calculations. Let's take the LHV of a bunch of hydrocarbons, and the LHV of the recoverable hydrogen that can be extracted from them using gas reforming (all units per mole of original fuel):

FuelLHV [kJ]LHV of recovered H2 [kJ]Efficiency fuel->H2
Methane 802.34 488 60.8%
Methanol 638.55 488 76.4%
Ethanol 1329.8 732 55.0%
Gasoline (octane) 5074.9 2196 43.4%

Not that great. As it turns out, people have written papers on this. A lot of hydrocarbons contain very large amounts of hydrogen, but the actual stored chemical energy in the form of hydrogen you can extract from them is very limited. This is because a great deal of the chemical energy is in the form of C-H (and C-C) bonds, which is energy you lose when converting to pure hydrogen (and cannot recover using a fuel cell!). Especially if we look at the efficiency of using gasoline directly in a fuel cell car, we find something incredibly interesting: it's actually worse to use gasoline in a reformer and then generate electricity in a PEM fuel cell than it is to directly burn it in an internal combustion engine - technology that has been made very cheap and abundant already. ICE efficiency (especially more modern engines that use the Atkinson cycle) can get thermodynamic efficiencies in excess of 35%. If reforming is already just 43.4% efficient and requires a ~60% efficient hydrogen-to-electricity step after that, the fuel cell alternative is worse in every respect.

The table is quite limited, but there are two important trends to observe: 1) larger hydrocarbons are less useful for gas reforming, and 2) alcohols are better at hydrogen energy storage than other hydrocarbons. This is why methanol in particular has gotten quite some attention over the years for use in fuel cells. It has none of the storage problems that pure hydrogen has, it is fairly easy to produce from other chemical energy sources (like biomass) and it is pretty good at 'storing hydrogen' for hydrogen fuel cells. Mind you though, you still lose about a quarter of the energy stored in the methanol outright, and you need to get rid of this substantial amount of extra heat with a larger, heavier and more expensive cooling system. The fuel cell car only gets more complex. So why not do...

Direct hydrocarbon use in high-temperature fuel cells
Autoreforming fuel cell types, i.e. fuel cells that can reform hydrocarbon fuels in the stack itself instead of needing a separate reformer, relieve the need for additional complex subsystems. Very roughly, there are two types: proton exchange and cation exchange. Proton exchange autoreforming fuel cells only conduct hydrogen ions through the electrolyte, meaning that we end up with the same problem as before: a big proportion of energy in hydrocarbon fuels is 'wasted' because we cannot harvest the energy from the creation of CO2. However, specifically solid oxide fuel cells (SOFCs) transport the oxygen ions through the electrolyte, which means these fuel cells work very efficiently with hydrocarbon fuels. This is why SOFCs are such a popular research topic for static and e.g. shipbound applications; they can run directly on e.g. Diesel fuel with much greater efficiency than even very large, nearly perfect turbine and piston combustion engines. Unfortunately the exceedingly long start-up times, very high weight of the stack and high required temperatures fundamentally rule out SOFCs for vehicle use.
The Kraftwerk portable SOFC. Courtesy of Kraftwerk

There are portable direct hydrocarbon fuel cells, e.g. direct methanol fuel cells. These have even found some moderate success in commercial applications, a recent example being the Kraftwerk mobile direct hydrocarbon fuel cell power supply. These suffer from quite low efficiency and power density, though. This product especially is marvellous... to demonstrate the poor efficiency and near-obsolescence of direct hydrocarbon fuel cells in portable applications. Let's do some calculations again!

First of all, Kraftwerk demonstrates how much supporting structure you need to even make a low-power USB charger. The entire charger is 160g (200g filled up with LPG) and only produces 2W continuous (with a quoted 10W peak, which is most likely a 1-second rating as the electrolyte depletes, although this is conjecture as they give almost no specs). The total energy that can be generated from the 40g of fuel is about 56Wh. LPG has an energy density (LHV) of 46.28MJ/kg, or about 12.9kWh/kg. 40g of LPG then contains 514Wh of energy. Only 56Wh can be regenerated from that, giving an overall system efficiency of 11%. A lot of this energy loss is in reforming; LPG is usually a mix of propane and butane which have a thermodynamic reforming efficiency of 47.7% and 45.9% respectively. This leaves an actual fuel cell efficiency of about 23.5%. This is much, much worse than any combustion engine, but indicative of typical lower-temperature SOFCs.

The really interesting thing about Kraftwerk in particular is that they claim the exhaust is only water vapor and CO2. Typical SOFCs do not have 100% complete combustion and require CO scrubbing, as well as always leaving a small proportion of the fuel unburnt (which is exhausted as is). The description isn't quite clear on this, but the fact it is advertised as being usable indoors leaves me to think they have done something very innovative to optimize for clean exhaust gases. Either that, or the eventual product will actually say it cannot be used for prolonged periods of time in an enclosed space. The unit claims protection against overheating, which is consistent with the entire energy cycle being internal to the unit (i.e. it will produce 18W of heat when it's generating 2W of electricity), with the majority of the heat not being exhausted as gas. Also interesting to note is that the unit only has a 2-year specified working life 'during normal use'. This is also consistent with low-temperature autoreforming fuel cells, which tend to have a useful life of 1000-2000 hours.
This is actually better at a lower price. Image courtesy of STX

But for all this innovation, a 200-g lithium ion battery unit with solar panel stores almost the same amount of energy (40Wh), can literally be used everywhere without even needing to fill up with gas and has a much higher power density, being able to fast-charge phones and even tablets in much less time. And these units are commodity and cost about half of the Kickstarter price. Cutting-edge (truly innovative) fuel cell technology cannot even start to compete with commodity battery tech.

Using hydrogen in existing ICEs

A couple of people have asked why we don't use hydrogen directly in ICEs. I don't need to devote much text to this; the simple answer is that it's inefficient. But we can get into a little bit more detail.

If we just quickly gloss over all the practical problems with storing and injecting hydrogen into the cylinders of your existing car, let's just look at the raw theory. The idea behind burning hydrogen in your car is that hydrogen has a much higher combustion temperature than gasoline or diesel; whereas gasoline burns at about 550K, hydrogen burns at roughly 800K. Because an ICE is a heat engine, we can express the theoretical maximum efficiency of the engine as:

efficiency = 1 - TC/TH
Carnot cycle efficiency. Image courtesy of hyperphysics

Where TC is the ambient temperature and TH is the combustion temperature (this is a gross simplification for practical engines, but bear with me). If we calculate this for gasoline, diesel and hydrogen, we get the respective efficiencies:

FuelCombustion temperatureCarnot efficiency
Gasoline550 K46.7%
Diesel673 K56.4%
Hydrogen800 K63.4%

Theoretically, burning hydrogen in an ICE is quite a bit more efficient than using gasoline or diesel. By the way, here's one of the biggest reasons why Diesel fuel is 'better' (in terms of efficiency) than regular gasoline. Actually, things get even better for hydrogen. One of the simplifications in this approach is that the actual combustion inside a cylinder doesn't happen simultaneously in the entire mixture. Instead, there is a flame front going through the mixture over a not-insignificant period of time. During this period of time, the piston actually moves which means that the combustion process doesn't happen under the most ideal circumstances everywhere in the cylinder. Diesel has a leg up on gasoline because it can achieve autoignition at multiple points in the cylinder instead of just having the single point of ignition (the spark plug). Hydrogen, though, has even more of a leg up in that its flame front speed is much, much higher than either hydrocarbon fuel. This improves considerably on the practical efficiency.

However, this is where the party stops. Because even though it's theoretically over 60% efficient, practical efficiency will most likely never surpass about 47% because of multiple technical issues (e.g. the actual practically usable heat engine cycles). Even the much more mature gasoline technology has not yet surpassed 38% efficiency in practical engines. And all of this while PEM fuel cells already, even in its relative infancy, achieve about 50-60% practical efficiency. Fuel cells are simply better than ICEs and will always be better because ICEs have this unbreakable efficiency limit. Not to mention the significant inefficiency of generating and storing hydrogen, which makes the comparison between gasoline/diesel and hydrogen in ICEs extremely unfavourable for hydrogen.

A commonly heard argument is that we already have ICE cars, and that we don't need to change much to regular cars to let them drive on hydrogen. Should I even type a rebuttal after 3 blogs full of the technical details here? No, it actually is quite complicated to adapt existing cars. This is not a sane argument.


I think it's best to leave this side of the technical discussion on fuel cells here. I was planning on talking about direct solar photocatalytic conversion, Sabatier carbon capture and a few other things but the post is already long enough and these subjects - in my opinion - don't add too much of value to the discussion. But if you disagree, I encourage you to type up a comment. In general, comment away! I don't make any money off of these blogs, I do these just for the love of the subject. I love to hear your thoughts, comments, corrections!

Today we've looked at alternatives to hydrogen and alternatives to PEM fuel cells. Basic conclusion: they're generally not that great, to be honest. Actually, I'm wrong here. They are great. SOFCs and AFCs are excellent fuel cell types that, in my opinion, will certainly come to dominate the fuel cell market. But the fuel cell market won't dominate vehicles. This will be for stationary applications. Fuel cells are great when they're sitting still.

In the next and last post in this series, I will tell you all about the way I think the future of vehicles will be. I will tell you why battery electric vehicles are better than both combustion engine-driven cars as well as fuel cell vehicles. I will also give you some sound investment advice. Seeya!

Mmmm, lekker eten!

Door mux op zondag 01 maart 2015 15:31 - Reacties (6)
Categorie: -, Views: 3.793

Ik heb nu al tweemaal geblogd over hoe je kunt besparen bij de supermarkt. Een vraag die Šltijd opkomt is wat wij dan eten. Pannenkoeken? Iedere dag macaroni? Nee, natuurlijk niet. We zijn luie luxebeesten en zowel vriendinlief als ik vinden koken leuk. In dit eerste deel van (misschien) een blogserie over voedsel:


Sinds een paar maanden bakken we regelmatig brood(jes). Brood bakken is eigenlijk kinderlijk eenvoudig en het is me niet heel duidelijk waarom we nu pas - na bijna 10 jaar uit huis te zijn - het broodbakken hebben uitgevonden. Anyway, brood en broodproducten zelf bakken geeft eindeloze gelegenheid om creatief te zijn. Een paar van onze favorieten:

1. Simit
Voor alle afbeeldingen geldt: klik voor groot!

Simit is een gezoet witbrood met sesamzaadjes in een gevlochten ringvorm. Al op de middelbare school vond ik dit ťťn van de lekkerste dingen uit de Turkse bakkerij vlakbij school, maar tegenwoordig wonen we in zo'n blank dorp dat we niet eens een echte Turkse bakker hebben. Dus bakken we ze maar zelf. Recept:

Standaard witbrood, voor 2 pers. (4-5 simit)
  • 250g bloem (Aldi/Lidl, §0,35/1kg)
  • 5g droge gist (Sahan, §1,25/125g)
  • 5-10g suiker
  • 3g zout
  • 150ml lauwwarm water
  • eetlepel olie (Aldi, arachideolie §2/1l)
Voor simit komt hierbij:
  • ca. 20g sesamzaad (Sahan, §2/200g)
  • eetlepel Pekmez (Sahan, §3,99/500ml)
  • evt. theelepel honing
  • twee eetlepels water
(Totaalprijs voor deze portiegrootte is ca. §0,28 excl. elektriciteit voor de oven)


Doe het water in een flinke kom en zorg dat het lauwwarm is. Strooi de gist en suiker in het water en meng het tot er geen klontjes meer zijn en al het suiker opgelost is. Doe de bloem erbij en daarna het zout. Kneed nu minimaal 15 minuten het deeg. Het deeg moet aan het einde soepel zijn, niet meer plakken en je moet er dunne vliesjes van kunnen trekken. Breekt het nog te gemakkelijk als je het uit elkaar trekt? Kneed dan nog een paar minuten door; je voelt het deeg in je handen veranderen. Langer kneden = luchtiger kruim in je eindproduct.

Olie nu een bak in en leg het deeg erin. Dek de bak af met een vochtige theedoek en zet het op een warm plekje om te rijzen voor minimaal 45 minuten. Aan het einde van de eerste rijs moet het deeg ongeveer 2x zo groot geworden zijn.

Neem het deeg uit de kom en kneed alle bellen eruit. Eventueel kun je het nog een minuut of twee doorkneden. Deel het deeg in het aantal simits dat je wilt maken; maximaal 5 (dat worden hele kleine ringetjes), minimaal 3. Deel vervolgens ieder stuk deeg nog eens door tweeŽn en rol ieder stukje uit tot een deegsliert van ca. 25cm. Neem de twee deegslierten en draai ze om elkaar. Druk de uiteinden aan elkaar. Het hoeft er niet perfect uit te zien. Leg de voltooide simitringen op een bakpapier.

Pak een klein kommetje, doe de pekmez, eventueel een beetje honing en twee eetlepels water erin en meng het goed door. Pekmez en honing zijn van zichzelf net even te dik om goed over de broodjes te kunnen smeren, vandaar deze stap. Smeer nu met een kwastje de pekmez over de broodjes en strooi het sesamzaad over de ingesmeerde delen. Draai de broodjes om en herhaal hetzelfde voor de onderzijde.

Laat nu de broodjes voor een tweede keer rijzen, wederom met een theedoek eroverheen, voor ongeveer 30-45 min. Bak de broodjes daarna gedurende 12-15 minuten af in een voorverwarmde heteluchtoven op 175 graden. Je resultaat moet er ongeveer zo uitzien:

2. Pittenbrood

Zoals alle broden in deze blogpost is dit niets anders dan een variatie in vorm en decoratie. In dit geval is het bovenstaande recept genomen, maar is het deeg in tweeŽn gedeeld en van ieder stuk deeg een kort stokbrood gevormd. Vervolgens is het deeg lengtegewijs diep ingesneden zodat het de ruimte krijgt om te rijzen en een mooi patroon krijgt op de bovenkant. Brood zal altijd naar boven of naar een insnijding toe rijzen, dus op deze manier is de vorm na de tweede rijs goed te controleren. Als decoratie is een gemengde pittenmix (zonnebloempitten, pompoenpitten, pijnboompitten) over de broden gestrooid.

Erg lekker bij een kop snert!

3. Maanzaadbolletjes/pistolets

Er is niks lekkerder dan 's ochtends wakker worden, de oven aan te zetten en 20 minuten later knapperende, vers gebakken zelfgemaakte harde bolletjes bij het ontbijt te hebben. Af en toe maken we het deeg en vormen we de broodjes in de avond zodat we 's ochtends alleen nog het brood hoeven af te bakken. Sterk aanbevolen! Het recept is niks bijzonders; op de foto staan standaard witte bollen met pekmez en maanzaad als decoratie. Tip: maanzaad is tegenwoordig heel goedkoop te krijgen omdat het blijkbaar verheven is tot superfood, en dus in grootverpakkingen te krijgen is (bijv. Xenos: 200g voor §2). Maak daar gebruik van zolang de superfoodrage duurt!

4. Hamburgerbroodjes