As professor Adam Harvey puts it, “The definition of PI is basically, for the same production target, doing something in a piece of technology that is orders of magnitude smaller. You often take a large batch reactor and convert it to a continuous reactor – and I’ve got examples of reactors which are 200 times smaller”. For example, he says, “a 75m3 batch reactor… can be replaced with something which is 0.3m3 – a very different thing.”
The advantages are clear: “In terms of the actual amount of metal, it’s a lot cheaper,” he says, “but if you think about all the other costs around that – the piping, the civil engineering – well, if something’s only 0.3m3, there is no civil engineering. It’s all the on-costs that you reduce.”
Harvey is leader of the Process Intensification group (PIG) at the University of Newcastle, a leading resource in the field: “We’ve got nearly 20 members of academic staff and about 60 to 70 researchers in the area”.
Much of the pioneering work in PI was done at ICI in the 1970s, where apparatus such as the HiGee Rotating Packed Bed (RPB) was developed. The RPB uses a spinning vessel (diagrammed on p23) to create centrifugal force which can be used for distillation or absorption – in this way, a relatively small unit can replicate the function of an extremely tall fractionating column. Other devices such as the Spinning Disc Reactor (SDR) and Annular Centrifugal Extractor (ACE) have since been developed for process intensification.
“It was very much an economic driver at that time,” says Harvey. “Since then, there are other drivers as well – a key one is safety. When you shrink down pieces of technology, you’ve got lower inventories of material in a hazardous state, and tend to have much better control, which makes things safer. With that better control, you’ve also got less waste and higher product quality. We’ve often seen those as drivers for changing processes, and sometimes obviously it’s combinations of all of these.”
There are secondary advantages too: “One of the key ideas around PI is enabling distributed production, and ideas about having a whole chemical plant in a shipping container.
“It can be portable, or it could be a number of smaller distributed units; the accepted wisdom has always been ‘the bigger the better’ – economies of scale – and in many cases that’s still true. But for some products it’s not necessarily true. One idea is about converting oilseed crops to biodiesel directly on the farm.”
USES
For example, the EU is funding the BL2F project, which aims to produce biofuels for aviation and shipping directly from a chemical by-product of the wood pulp industry called black liquor. The intention is to integrate a chemical plant into a pulp mill to undertake a hydrothermal liquefaction process, with a potential “83% reduction in CO2 emissions compared to fossil fuels and a very competitive production cost per litre.”
Similarly, Carbon Clean produces modular systems for carbon capture, utilisation and storage (CCUS) systems which are prefabricated off-site, containerised and delivered to the site ready to install. The firm’s CycloneCC system uses rotating packed beds (RPBs) built on to a skid mount. The firm says this approach “minimises site disruption and execution time, allowing your facility to continue its regular processes. Modular equipment is also much smaller than a traditional open-plant system,” and a unit can be “fully operational in less than eight weeks.”
It’s not all about reducing cost, says Harvey: “It’s often a product quality issue – if you can achieve a product quality that just can’t be achieved otherwise, that’s a great driver.” And some PI systems generate products which can’t be made by traditional means.
A large batch-processing stirred tank reactor (STR) is “basically a big bucket with a stirrer in it, and within that you’ve got a range of different temperatures. You can’t mix it up so that it’s completely uniform above a certain scale. You can do it in the beaker in the lab – that’s easy; you can put a lot of energy in per unit volume.” But with a batch reactor, he says, “you simply can’t. So what you get is not necessarily what you expect: lots of different by-products and fouling.” A development of the STR is a Continuous Stirred Tank Reactor (CSTR), in which reactants are constantly added and product extracted; but this means that some chemicals have a longer residence time in the vessel than others, further compromising the reaction quality. “So, if instead of this big bucket with a mixer in it, you have a continuous tubular reactor, then you’ve got very good control of how long the stuff is in there and how well mixed it is. And that generally helps these reactions go faster.”
ANOTHER EXAMPLE
Edinburgh-based NiTech Solutions makes a Continuous Oscillatory Baffled Reactor (COBR), which pumps reactants through a tube fitted with internal baffles, with a further oscillating pump providing a varying pressure to encourage mixing. The system can deal with solids, liquids and gases, and has developed from a lab-sized unit to a pilot-scale machine and now a production-scale reactor: this uses a nickel-alloy tube with a 43mm internal diameter and up to 40m or more in length. The whole unit is substantially smaller than an equivalent CSTR.
NiTech claims startling possibilities, quoting “reductions of “more than 70% in footprint, 50% in CapEx, 30% in operating costs and 10% improvement in yield”. It also reports crystallisation times down from eight hours to 12 minutes for one client.
“Process intensification is about every step in a process: heat transfer, separation, any step really.” The implications for ancillary plant and equipment can be significant, in removing process steps: “Because of these better reactors, you can actually often get rid of solvents,” says Harvey. “If you can mix something well enough, you can just react A with B without it all being dissolved together in C.”
Solvent dilutes reactants, slowing down reactions and soaking up energy from exothermic reactions. “But with more sophisticated types of reactor you can get the heat out more easily. The heat transfer is the limiting factor.”
This also shows how PI can address the 12 ‘Principles of Green Chemistry’ which have been adopted by many organisations, among which is to minimise or eliminate solvents.
While Harvey describes oil and gas as well-developed in this field, he points out that there’s been very little intensification in pharmaceuticals – he calls them ‘very inefficient’ – perhaps because profits are not process-dependent.