Explaining Process Intensification to Process Chemists

kilomentor | 03 October, 2009 17:14

In the 1980s, Colin Ramshaw at ICI coined the term “process Intensification” to describe his engineer rethink about gas/liquid mass transfer. That resulted in aiming for much smaller chemical plants that would be markedly cheaper and safer than existing ones.

Ramshaw’s zero-based thinking moved away from existing equipment. He started thinking afresh. Distillation for example he saw as fundamentally a gas liquid mass transfer for which the key cost drivers for a given system were well established:

· Well mixed liquid and gas phases
· Lots of interfacial surface area
· Thin liquid film
· Counter-current operation

In general gases mix well in all conditions as do low viscosity liquids in thin films. Simple geometry teaches us that smaller, finer, packing gives us more surface area so that would be the obvious way to go - a column with very fine packing with counter-current gas flow.

However, a liquid film running through a bed of fine material floods when the film thickness becomes approximately equal to the clearance between the bits of packing. The limiting factor is the thickness of the liquid film and most of the factors determining film thickness are physical properties of the fluid and are not open to modification. Only gravity was independent. The higher the applied gravity the thinner the film and the smaller the packing could be. If gravity could be varied that would give a lot of mass transfer surface area for volume i.e. an intensified plant. To increase virtual gravity the centripetal effect of rotating the packing in a “high-g” machine was demonstrated to deliver an order of magnitude reduction in size. The idea was a major announcement at the time. An article appeared in Chemistry & Engineering News, Novel Separation Technology May Supplant Distillation Towers March 7, 1983, but the high-g machine never became popular.

Even so, this zero-based engineering that starts afresh from first principles exemplied the essential process of science and had appeal as a creative process. Understanding a process (a reaction, a crystallisation etc.) with sufficient depth so that the key rate controlling steps are understood and then matching that process to the right processor was seen as potentially breakthrough innovation.

Heat exchangers are another example. Obviously one of the keys to performance is heat transfer area so it is surprising that many heat exchangers are based on pipes that have a minimum surface area! It has been proposed that this reflects mechanical engineering considerations rather than process ones. Clearly the plate heat exchanger is a much more effective way of providing area, albeit with some mechanical downsides.

This is 180 degrees opposed to the normal approach in the chemical and pharmaceutical process industry, which creates a process to match standard equipment. Although there are good economic reasons for this in a batch process industry, there was a feeling not to lose sight at the design stage of the possibility that intransigent difficulties operating in the standard way may become trivial with different equipment. For example, the ubiquitous batch reactor might be used to carry out a polymerisation in the laboratory but the recipe used on plant scale will be adjusted to match the relatively poor heat transfer performance of a larger reactor. Here, the process has been tuned to match a characteristic of the processor. Perchance in some particular instance, the rationale for this matching process may even be lost in corporate history. Perhaps a batch takes a certain length of time to complete because many years ago it was matched to a particular reactor or type of reactor.

Just as important is the corrollary that the process that has been matched to a particular processor cannot be simply transferred to a different processor without adjustments. For example, for exothermic reactions rate is proportional to temperature. A reaction temperature is selected so that the heat can be removed and the reaction condition kept under control. One can make an order of magnitude change in the rate and still disipate heat , by going to a plate reactor. Thus a higher operating temperature can be held in control and a much shorter reaction time becomes practicable. The reaction time may become so short that continuous processing becomes possible.

In fact, the new reactor will not work unless the process conditions are changed to harmonize with its new character. In the above example of an exothermic reaction, the matching of process temperature is key. Other characteristics that might need adjustment are mass transfer, mixing, diffusion etc. Often the controlling step is obvious, sometimes it is completely unknown and sometimes there are different rate controlling steps during the course of a reaction. What the critical variables are, constitutes fundamental understanding.

Batch reactors or in their continuous form continuous stirred reactors (CSTR) will match a process that inherently needs long times (perhaps diffusion controlled with real maximum temperature limitations). Oscillating columns offer moderate residence times with better than batch heat transfer. Plate heat exchanger type reactors (HEX reactors) are a good match for clean high heat transfer duties. Spinning disc reactors offer good heat and mass transfer as well as good mixing. It is erroneous to claim that one is inherently better than another, anymore than to claim a Posidrive screwdriver is better than a crosshead. What is required is to match the process and the processor! All the benefits of process matching, precision processing or process intensification are not always obvious. Clearly capital cost saving is the classic rationale with smaller reactors, less civil costs, less safety systems but improvements in yield, higher conversions, less or no solvent use are also important along with energy reduction. Improvements to product properties and even novel products that competitors find difficult to match are other potential major benefits.

As discussed above, the matching of the process to the processor is key to precision processing. It is also important to recognise that the way a business is run often reflects the processor. For example multi products on a batch reactor with the need to clean between batches usually means some form of campaign operation of the reactor and a warehouse is needed to meet customer delivery demands.

The way the business works is matched to the characteristic of the processor. Change the processor to a low inventory continuous reactor and it might be possible to move to just-in-time (JIT) manufacturing with all those benefits. The business operation has been properly matched to the new processor characteristics.

Comments

HiGee

gyges | 04/10/2009, 11:11

Interesting post.

I heard that Prof Ramshaw was working for NASA; he was investigating whether or not it was possible to do any chemistry in space, ie zero g conditions. A gravity (force) field of some sort was required, hence the spinning. He quickly realised that the force could be varied and hence (as you point out wrt flooding) throughput could be varied.

It is surprising that HiGee distillation isn't more popular; from what I can gather from reading PIN reports, this technology is used in China.

I think that the technology would be useful in biodiesel production, in order to separate methanol from methyl ester.

«	May 2013					»
Mo	Tu	We	Th	Fr	Sa	Su
		1	2	3	4	5
6	7	8	9	10	11	12
13	14	15	16	17	18	19
20	21	22	23	24	25	26
27	28	29	30	31

kilomentor