BY DAVID BOTT
There is an old saying that “no strategy survives first contact with the enemy” (though I prefer Mike Tyson’s more descriptive version “everyone has a plan until they get punched in the mouth”!). When we were planning the Flue2Chem project, we drew out a detailed Gantt chart with deliverables, dependencies and deadlines. Sadly, the elapsed time between the last post in this series and now is a vivid example that Mike Tyson got it right.
To recap, the aim of the Flue2Chem project is to collect the carbon dioxide from flue gases and turn it into a simple non-ionic surfactant for use in detergents and other consumer goods – potentially eliminating the need to extract more fossil fuel to make them. Key intermediates are dodecanol and ethylene oxide, but the first step is to capture the carbon dioxide.
What has gone before
The idea of extracting carbon dioxide from flue gases has been around for a long time, and has actually been “practiced” since the 1950’s. For a long time, it has been driven by the idea that carbon dioxide emissions can be captured and stored underground, thus avoiding adding them to the atmosphere and causing climate change. Nowadays, if you go to a Carbon Capture Utilisation and Storage (CCUS) conference, the few talks on utilisation mostly talk about the direct use of carbon dioxide either in fizzy drinks or for pumping into greenhouses as a feed to horticulture. We are aiming for something different.
The first need is for a source of flue gases. When we were putting together the Flue2Chem consortium we understood that different sources would have different chemical compositions, so we sought out different types of sources to maximise the spread of data for both the techno-economic and life cycle analyses. There are two paper companies in the consortium, Holmen and UPM. Their carbon dioxide emissions are classified as “biogenic”. Both have biomass combined heat and power plants, built in the days when biomass was regarded by the government as a renewable power source, and subsidised under the renewable obligations scheme that formed part of the 2009 Renewable Energy Directive. They each generate about 1000-1500 tonnes of carbon dioxide a day.
We also had the Port Talbot site of Tata Steel as part of the project. You could classify the carbon here as “used fossil carbon”. Coal is used both as a source of heat and as a reducing agent to turn the iron ore into iron. It is a complex process and so there are many sources of carbon dioxide on the Port Talbot site, some mixed with carbon monoxide In total, they generate about 15000-20000 tonnes of carbon dioxide a day.
The basic requirement for a process to capture carbon dioxide is easy to state – you need a system that will reversibly absorb carbon dioxide, and some good engineering!
The liquid amine route for capturing carbon dioxide uses a mixture of amines to react with the carbon dioxide to form carbonates. The absorption is usually carried out in a vertical column where the amine trickles down in a packed column and the carbon dioxide flows up. The resulting carbonate is then moved into another column where it is heated to decompose the carbonate to reform the amine and release the carbon dioxide. The energy efficiency of the process is largely determined by the energy required to decompose the carbonate. Over the years, different companies have optimised their mixture to minimise the energy costs and often keep this as “black art”.
Solid state absorption systems rely on physisorption. They used to be based on zeolites, but many recent ones use metal-organic frameworks. The early ones used a similar temperature driven process to control the absorption and desorption, but there are now systems based on pressure swing, where the absorption is driven by higher pressure and the desorption by much lower pressures. These are suggested to use lower energy than the more conventional temperature driven systems.
So, how is it going?
Early on in the project, one of the two companies providing the capture systems we wanted to include in the project – Carbon Clean – ran into an issue with the Environment Agency’s policy regarding solvent disclosure. They use an amine based solvent and, as mentioned above, they want to protect the confidentiality of their IP from this major commercial risk. However, the Environment Agency requires disclosure of any chemicals that might be emitted in any process, and most amines have a measurable partial pressure at the temperature used in the carbon capture process, so although they might have been able to get an exemption for a research or test use, once they go commercial in the UK with their system, they will have to disclose to the Environment Agency. AND, the Environment Agency is subject to Freedom of Information requests and would have to disclose Carbon Clean’s proprietary information. This is why Carbon Clean chose to withdraw its technology from the project, while continuing to provide techno-economic analysis.
This led to another decision – this time by Tata Steel. They had already worked with Carbon Clean in India and were looking to scale up the technology to the 10 tonnes/day envisaged within the project. With that off the table, they wanted to rethink their plans. As they were doing so, the bigger announcement that they would close the blast furnaces and move to use electric arc furnaces at Port Talbot amplified their concerns. Depending on the exact implementation route they choose, there might be minimal emission of carbon dioxide, so they withdrew from the work package to collect carbon dioxide.
Fortunately, in addition to Carbon Clean, we had also included a solid state capture technology, albeit at a much lower state of technology development, in the project. FluRefin had been developed at the University of Sheffield and was being commercialised by Carbon Capture and Utilisation International (CCUI). This had been operated at the small scale but as part of the project, it was being scaled up to 1 tonne/day capture. This required wholly new equipment, some of which had to be imported from India, some from Germany, but was assembled in the UK. It was planned to be installed at the first collection site (Holmen) at the end of November 2023, was actually delivered to site in mid-January, but commissioning issues delayed the first real carbon capture until late April. We have learnt a lot about fast-tracking process development – and the challenges it causes, partners working off different versions of the Process and Instrumentation Diagrams, the design experts being in Sheffield and the equipment being in Workington and so on but, as anyone who has done this before will tell you, this is all quite normal and we were very optimistic in our initial plans! Once on site in Workington, we had the support of some excellent engineers and the various problems were overcome.
One aspect of using a pressure swing process is the need to compress the input gas. This required the use of a number of compressors, but when they arrived we discovered that they had been designed for compressing air to be used as “compressed air” and were a bit “leaky” on the input side. We knew this because the output carbon dioxide concentration was lower that the input and not as we needed and thought we would get. This required more engineering to adapt the compressors for our use.
Another requirement of using the pressure based system is that the input flue gases need to be cooled (from about 150oC to around 30oC) – this recovers a fair amount of heat. More engineering was required!
Over the next few months, we started to capture enough carbon dioxide to supply the chemical conversion work packages. But this led to another “challenge”. We are aiming to capture about 1 tonne per day. This is below the level where we could engage one of the major gas product companies to provide bottling technology, and we were initially not planning to liquefy the gas (so did not have the required equipment). This means we were using a fairly basic “put gas in a pressurised gas bottle” process. Luckily, the University of Sheffield team were also involved in another UK Research & Innovation (UKRI) project – called SUSTAIN Steel. They had a small carbon dioxide liquefaction kit. We have “borrowed” it and are using it to liquefy the captured carbon dioxide, albeit at a very slow rate. We have other ideas for how to do this at a larger scale, but this works, and we are now capturing the required amount of carbon dioxide to send to the two centres where the next stage of our supply chain – the conversion of the carbon dioxide to ethylene oxide and dodecanol – will be carried out, but that’s another post!
So, what have we learnt?
Firstly, that project plans written in a hurry to fit within the proscribed timescale and budget will almost certainly be too optimistic and liable to require drastic adjustment. This was no real surprise to those who had been involved in scaling up processes before, but we rediscovered the saying that “if it can go wrong, it will” is irritatingly true. However, the capabilities of the individual organisations in the project and the creativity of the combined “leadership team” means that we have always found a way out of every “challenge”. We have also applied to Innovate UK to extend the project by 4 months, and have been successful, so have bought a little more time. The next work packages, which would have been squeezed by the 6 month (or so) delays will be “less” squeezed.
Perhaps the biggest learning is the need for flexible resources to enable scaling up the sort of processes we are using. This does not necessarily mean expensive new buildings or plants. A small carbon dioxide liquefaction plant that could handle 1-10 tonnes a day would have saved us about 2 months. More engineering expertise in the consortium might have saved us another couple of months building and commissioning the FluRefin plant. We had some allowance for creep in the original plan but when every month the deadline slipped by a month, I felt sorry for the project manager!
What has really been driven home to us is that the change we are attempting to prove – that it is feasible to move the chemistry supply chain away from virgin fossil carbon as a feedstock – may be scientifically credible, but reducing anything from theory to reality is harder than we think and required even more planning and effort than we imagined.
And we need to use realists, or even pessimists, as planners!!
Written by David Bott, Director of Innovation at SCI and originally published on Linkedin
