Ask the experts extraction

Q. When should I consider liquid extraction? Please give the general rules-of-thumb.

The most cost-effective method of separating or purifying a liquid mixture is normally distillation or a related evaporative process, such as steam stripping. Always evaluate distillation first. But don't fail to consider liquid extraction when distillation appears to be cumbersome and/or expensive. Here are some examples of this situation:

Low relative volatility: Capital and operating costs of a distillation process are highly dependent on the relative volatility of the mixture. A low volatility leads to large stage and reflux-ratio requirements. A high number of stages equates to tall, expensive columns. High reflux ratios call for large heat exchangers and high energy costs. In general, consider liquid extraction when relative volatility is less than 1.2.

Removal/recovery of low-volatility components from water: Relative to distillation and stripping, liquid extraction can have real advantages for recovering heavy, essentially non-volatile components from water. Distillation becomes unattractive because so much water has to be evaporated (with its high energy and equipment costs). The heat required to vaporize a pound of water is often 3-4 times greater than that for a typical hydrocarbon. Steam stripping is likely to be unattractive. This application of liquid/liquid extraction may require wastewater cleanup. Roger Cusack ("Solve Wastewater Problems with Liquid/Liquid Extraction," CEP, Apr. 1996, pp. 56-63) provides an excellent discussion of wastewater applications for extraction.

Removal/recovery of low-volatility polar components from organics: Liquid extraction may be used to recover (or remove) salts and acids from an immiscible, organic liquid. This process is often called washing. Water with or without a complexing agent is used to recover the salts and acids. Caustic soda is often used as the solvent.

Recovery of thermally sensitive components:. Distillation can cause problems with heat-sensitive materials. An example is recovering antibiotics from fermentation products. Extraction is a much-gentler means of recovery.

Q. New, advanced models and computer programs for liquid extraction have been developed in recent years. Is it still necessary to pilot a new extraction process?

Yes, I highly recommend that all new extraction processes be pilot-tested before commercialization. These computer models are especially useful when coupled with experimental verification. Use pilot tests to address three critical issues:

1. to demonstrate the full separation process

2. to detect the effects of impurity buildup in the extraction loop

3. to evaluate scaleup of the contacting device (e.g., the packing material).

In general, an extraction process involves an extractor and a solvent-recovery operation. The recovered solvent is recycled back to the extractor, making an extraction loop necessary. The feasibility of the loop must be demonstrated. This is especially important for chemical systems with complex and poorly understood phase equilibria. For example, a system where the slope of the equilibrium line changes significantly with the solute concentration may be prone to pinching.

If the solvent is non-volatile, it can cause accumulation of heavy impurities in an extraction loop that are surfaceactive. Even in trace concentrations, these culprits can have a devastating effect on extractor performance. They can reduce the coalescing rates of drops - and thus reduce column capacity. Since most flooding models are based on pure-component tests, these models tend to be overly optimistic. Relative to a clean system, the presence of impurities can lower column capacity by 20% or more and efficiency by as much as 60%.

Pilot-testing of the extractor is useful for detecting other unforeseen problems. Create an experimental design using available mass-transfer and hydraulic models, and then use an experimental vessel that permits viewing, if at all possible. Run the design solvent-to-- feed ratio at a variety of loadings. In particular, note the mechanisms and location of the flooding condition.

After successful completion of the pilot tests, analyze the data carefully and compare them against those predicted by the models. Address and then resolve all deviations carefully before designing the commercial system. For column-type extractors, don't fail to correct for axial mixing effects. Axial mixing will reduce the concentration difference between the phases and, as a result, reduce the apparent efficiency of the contacting device. Axial mixing can be especially critical in the design of packed columns. The correction for axial mixing in packed extractors is described by Becker and Seibert (Chemie-- Ingenieur-Technik, 75:35, 2000). Vendors of extraction equipment can be especially helpful regarding the scaleup of their devices.

[Author Affiliation]

A. Frank Seibert, PhD, P.E., is the technical manager of the Separations Research Program at the Univ. of Texas at Austin. (Phone: (512) 471-7063; Fax: (512) 471-1720; E-mail: fseibert@mail.utexas.edu). His research is focused on scaleup of liquid extraction and distillation processes. His 22 years of experience covers fundamental research and industrial consulting in several areas of separation process technology.