Carbon Fibre's Carbon Footprint

I've been meaning to do this one for a while, but I'm afraid that I don't have any real answers for you (or me) yet. I wanted to answer the question: "What is a carbon fibre bicycle's carbon footprint?" under the conditions of me buying a new carbon fibre bicycle and riding it, using similar approximations to my estimates about metal bikes: http://towalestwowheels.blogspot.co.uk/2011/06/low-impact-riding.html

It's not that simple, though.

First of all, I cannot find any data for the lifecycle carbon emissions of carbon fibre composite, even from my old, faithful source, the Bath University Inventory of Carbon & Energy. The best I can do is Glass Reinforced Plastic - GRP - which, at 8.10 kg CO2e/kg of material, is better than virgin aluminium, but carbon fibre cannot come from a recycled source for a large component (frame, fork, wheel, crankset... pretty much anything other than a bottle holder), so I would expect the composite bike to come out worse overall. More importantly, though, working off this number ignores how difficult carbon fibre is to produce compared to glass fibre.

I should probably give a quick explanation about what a fibre composite material is, for those who don't know.

Glasses and ceramics can be pretty strong. Think about the portholes and viewing domes on submersibles - they are under enormous pressures, yet barely compress at all. The downside of that rigidity is that if you do try to bend them, they tend to crack and break, very suddenly.

To overcome this, we can make these brittle glasses and ceramics into fibres. This does several things - primarily, it aligns the crystals of the structure along the fibre (especially with carbon); secondarily, by placing fibres side be side instead of in one homogeneous structure, if there is a defect in one of the fibres, it shall not propagate through the whole structure.

These fibres tend to be very strong in tension, but obviously, when compressed, they buckle. To hold them together, we need to put them in a "matrix" - a bonding material. This is typically something like epoxy resin, and the resulting material is a "composite" of the fibre and matrix. Steel-reinforced concrete is a large-scale fibre composite material.

As mentioned, the fibres tend to be very strong in tension along the axis that they are aligned. To obtain strength in other directions, we bond layers of composite one on top of the other, with the plies of the fibres in different directions.

The balance between the ratio of fibre to matrix, and the diameter of the individual fibres, also affects the mechanical properties of the resulting material, notably affecting the ratio of its tensile, compressive and shear strength, as well as how brittle the overall material is. I'll come back to this later.

Ok, so there's your primer. So how close is glass fibre composite, in manufacturing, to carbon fibre composite?

Glass fibres are made by melting (primarily) silica glass beads (marbles) at approximately 1250 degrees C, which gravity feeds through a sort of strainer and gets pulled into a strand.

So, glass needs to get pretty hot, right?

Not compared to graphite.

There are a few different type of carbon fibres, but one process for structural-grade graphite fibres is the polyacrylonitrile (PAN) process. This starts with a reel of PAN fibres, run it through a controlled oxidisation furnace at 250 degrees C, then into a pyrolysis furnace to drive off any non-carbon atoms at 250 degrees to 1500 degrees, and finally a graphitization furnace to draw out those crystals which warms the strand from 1500 degrees to a whopping 2500 degrees C. Yes, that is considerably more than the melting point of most (if not all?) steels.[1]

So carbon is a totally different animal to glass.

What about the other end of its life, then? What about recycling?

Well, that's actually been pretty well covered in the cycling press. Most large manufacturers offer some form of carbon composite recycling, with Specialized even offering to take in frames from manufacturers who aren't running their own schemes. However, it's not a panacea. Here's a thoughtfully-written piece on the issue:

http://www.redorbit.com/news/science/1112493049/the-dirty-secret-of-carbon-fiber/

I do wish, however, to come up with a slight rejoinder to some of Peter Suciu's closing remarks, where he claims that carbon fibre takes some of the artistry out of design and production.

In my opinion, this is utterly incorrect.

The argument is that 3D CAD design and 3D printing takes humans out of the loop. In reality, we are taking about being out of the loop on things we never really had a say in anyway. How many bike builders cast their own lugs, let alone draw their own tubes? If anything, fibre composites give us more control over the design and feel of our bikes - and it is certainly a lot easier to make a bad one. Invisible things like layout and fibre choice make almost continuous differences, as opposed to the very discrete options of alloy and tube diameter. The laying of "pre-preg" plies or even (as I believe occurs for Look) individual fibre layup is an incredibly taxing job, requiring high dexterity and attention to detail. Manufacturing the moulds alone is a high-precision task, and developing new ways of keeping the internal surfaces of carbon fibre objects "clean" has been one of the major areas of innovation for the bicycle industry over the past few years.

Carbon fibre certainly hasn't taken the human out of the loop. If anything, it is its necessity for hand-layup that has caused so many manufacturers to outsource production to the far east. The design is now more intricate than ever, but based on invisible properties such as durability and stiffness. There will come a day when you can be an artisan carbon-fibre manufacturer - when there is a standard set of lay-ups, moulds and mandrels to iterate on without worrying that your tinkering is going to ruin the ride. Until then, fibre based composites are still an exciting engineering challenge.

I feel I should draw this post back together, as responding to that article has drawn it off-course somewhat.

I cannot guess the carbon impact of a carbon-fibre bicycle, but I have good reason to believe, based off of the GRP values and the relative energy intensity of carbon fibre as well as its inability to be recycled in long strands, that it would be more carbon-intensive than an aluminium bicycle, and therefore than a steel bicycle.

On the other hand, a well-made carbon-fibre bicycle could conceivably outlive an aluminium one, assuming that it is well-sealed. So don't feel too bad.

At least it's not titanium...


[1] Composite Materials for Aircraft Structures (1986); Brian C Hoskin & Alan A. Baker; AIAA Education series,  Washington DC