3 replies to this topic

[Art Montemayor]

tlp:

I suspect that theoretically all batch processes could be converted (or replaced) by continuous ones - or at least by semi-continuous ones. But that is generalizing based on personal experience. A more proper engineering response would be based on an actual run-down of each and every type of batch operation.

However, bear in mind that installing a batch process instead of a continuous one doesn't necessarily imply that a continuous version isn't available - or possible. The type of process selected for eventual design and installation on an industrial scale is subject to a lot of probing questions and analysis:

1) Is the process selected the most "optimum": will it yield the highest rate of return on the financial investment?

2) Is the process selected the SAFEST? Are all the important safety aspects and risks readily addressed and resolved?

3) Does the process selected exhibit the best operability - i.e., can it be operated with ease as well as safety and does it exhibit acceptable turn-down ratios?

There are other points to consider, but the above are some of the principal ones. I once participated in a change-over of an established batch process to a continuous one in which many factors had to be considered. The process was the classical batch acid hydrolysis of Pentosans to Furfural. The feed was sugar cane bagasse and sulfuric acid - formidable and troublesome substances to handle, meter, and seal off. The break-through was the introduction of a "French" Press that enabled the feeding of the troublesome solids as a virtual, high-pressure plug into a series of reactors operating at over 150 psig. It worked, but a lot of lessons were learned in the initial operations. Problems of this type can be expected in going from batch to continuous and often the effort is not worth it and the process remains batch. However, with the increasing cost of manual labor on a world-wide basis (something that is basically good for social & moral reasons) the incentive to go to continuous processing is always increasing.

I hope this response answers your query efficiently.

[JMW]

Another reason why some batch processes dominate is instrumentation: for example, heavy fuel oils are now most often produced by by blending a high viscosity residual oil with a low value low viscosity cutter stock.
Fuel oil specifications include quite a number of different parameters such as vanadium content, carbon residue, ash, aluminium and silica, sulphur, density and viscosity.
Most of these properties can be calculated for the blended fuel from the known properties of the residual oil and the cutter stock as they are "addative". However the viscosity is not.
Most use a fuel blend calculator to determine the appropriate volume or mass ratio of the feeds to produce a blend of the required quality. The problem is that even with the ratio accurately controlled the viscosity of the blend can be off significantly.
In a significant number of installations the approach taken is to batch the calculated volumes into a mixing tank, mix and test and then correct as necessary, though when this is performed on barges (sparge mixing) there is often no opportunity to correct.
Even where inline blending is used to improve the mixing the blend is still dumped into a storage tank for testing and correcting.
However, many inline blending systems now use online viscosity measurement for feedback control.
The simple reason for such processes being batch rather than continuous has been the dependence on taking samples for laboratory evaluation. New viscometers do make continuous blending possible.
A similar situation has existed in caustic blending where the heat of the reaction has caused problems for process instrument measurements, in this case, density.
So one reason why batch processes may persist is due to a lack of suitable process instrumentation.
Of course, when new instruments become available and are intorduced it is surprising how many are used less than optimally because they are used simply to replace the lab measurement and not to properly automate the process as this could involve a significant capital outlay. What the excuse is for new systems i could not say.

[Ming Hooi]

Yet another reason why batch processes are sometimes chosen over continous processes is because batch processes permit an extensive variety of the product and frequent changes of the product. For example, in the production of resin, the same reactor run in batch can be used to produce resins from types A to Z, depending on the market demand. Such advantage is not exhibited in a continuous process whereby perhaps 26 different reactors are needed to produce all the A to Z types of products. On the other hand, products from a continuous process are generally more consistent in terms of quality, and the throughput is often larger.