Navigating Process Development for Separation Operations

Designing new separation operations involves several key steps, including proper naming of the process steps, identifying suitable technologies and supporting studies needed for unit-operation selection

Chemical engineers must be vigilant when reviewing first-of-its-kind chemical processes to ensure that processing steps are accurately named. In the author’s experience, finding one or two inaccurately named process-step flowsheets containing a dozen processing steps is common. For example, I once worked with a client who spent many hours carefully adjusting liquid ratios in what they believed was an extraction process. However, the process was a solid-liquid dissolution, requiring an entirely different approach. Solids dissolve much faster under vigorous agitation, so by simply reclassifying the unit operation as a “dissolver” and increasing the agitator horsepower by ten times, the vessel sizes were significantly reduced, as was the capital cost of the design. 

Other common engineering misnomers I have encountered include the following:

• Adsorbers mistakenly referred to as extractors
• Solids described as fluidizing when they are pneumatically conveyed
• Flashing described as distillation

Process naming rigor is particularly important for new and first-of-its-kind processes. As an example of the importance of a name, in the 1800s, lymphoma was mistakenly classified as an infection, which caused significant delays in discovering effective treatments and cures. Correctly renaming the condition (cancer, a type of cellular growth disorder) was a major breakthrough that greatly accelerated further medical research.

Chemical engineers should always begin by describing the process in the most fundamental terms, with special emphasis on the verbs: a simple 2-phase mixture of vapor and liquid may be flashed (at the boiling point), separated, rectified, stripped or dispersed (gas in liquid) just to name a few typical processes. Costly delays may result if a process description or name is altered after weeks of engineering. Vendor confusion is likely when receiving questions that refer to their adsorbent as a catalyst or their demister as a coalescer. Correctly naming and classifying each step from the start is a simple but necessary part of process development, ensuring that the appropriate design and optimization activities can follow.

Evaluate unit operation selection

With each step in the chemical process correctly named and classified, it is critical to identify the most suitable equipment for each operation. Often, the mass flowrate plays a significant role in determining the appropriate machine. For example, antibiotics, organic impurities or precious metals dissolved in an aqueous acid are often extracted into an organic phase to concentrate and purify, requiring specific liquid-liquid extractors. At very low flowrates (<1 gal/min), a single-stage or series of centrifugal extractors might be appropriate. An extraction column will be more cost-effective for moderate flowrates of 1–1,000 gal/min. However, at flowrates exceeding 1,000 gal/min, a mixer-settler, constructed similarly to a series of swimming pool-sized tanks, becomes necessary. Scaling a centrifugal extractor to handle such high flowrates isn’t feasible, and using dozens of parallel extraction columns would be impractical.

Given the significance of mass flowrate on equipment selection, it is helpful to have reference tables or charts for various unit operations where one axis of the chart is mass throughput, and the other axis might be pressure ratio (for compressors), particle size (for solid-liquid separators) or stage efficiency (for liquid-liquid extractors). Chemical engineers developing new processes can benefit from updating and maintaining a repository of these references over time. Beyond throughput, a second class of properties, such as temperature, pressure and viscosity, also play crucial roles in selecting the right equipment. Only after thoroughly addressing these criteria should cost be considered, though it is often seen as a primary concern. The process requirements must take precedence; if the cost is too high, the process should be revisited. Choosing cheaper equipment that is not well-suited for the intended service is not recommended.

Consider the following example: A solid material must be filtered from a liquid stream. Based on prior experience, the engineer selects a filtering centrifuge and uses a small 1-ft³ pilot-scale unit to process several gallons of slurry successfully. However, centrifuges exert significant stresses on rotating components, requiring that they be constructed of metal rather than plastic. Should the liquid prove corrosive, the metal components will deteriorate, and the centrifuge must be upgraded or replaced. Instead, a gasket and recessed filter press with impervious polymer plates and filter cloths would have been a superior option. Selecting a plastic filter at pilot testing phase is preferred as it will engender more confidence (if it operates successfully) when scaling up to a larger unit. The filter press also costs less than a tenth of a filtering centrifuge.

Regardless of which equipment performed well in previous projects, it is essential to consider which machine is the least expensive, the most resistant to the processed fluids, and appropriately sized for the flowrates. Process development often requires evaluating multiple viable options simultaneously. Since not all available options are available, good references, such as Couper’s “Chemical Process Equipment Selection and Design,” [1] are essential. Once the several viable options are laid out for evaluation, it will become apparent that a small amount of additional testing is required to make a fully informed selection.

Critical testing

One of the most common questions posed during the design of a larger-scale chemical process is: “What information do I need to scale up?” In the previous example, a filter selection was considered, highlighting the importance of knowing the corrosivity of the fluid on unit operation selection. One may wish to know many more parameters, and it is not uncommon to encounter a multi-step chemical synthesis process where one or two unit operations lack experimental evidence to support their selection. The table below provides examples of  recommended experimental tests for some of the most commonly encountered unit operations:

Unit Operation	Test
Distillation	Vapor-liquid equilibria, tested at a range of relevant compositions
Crystallization	Solubility curve of a product as a function of temperature
Extraction	Liquid-liquid equilibria, tested at a range of solvent: feed ratios
Adsorption	Isothermal adsorption curve as a function of concentration

When designing a distillation tower to separate two components, it is essential to know their relative volatilities at concentrations ranging from the feed to the desired product(s). Without vapor-liquid equilibrium (VLE) data, it is impossible to determine whether distillation alone can achieve the separation. If it becomes apparent that further engineering and design cannot proceed without experimental measurements, the time and cost associated with this testing may be significant. In such cases, consider an alternative separation method that offers simpler or quicker confirmatory tests as the primary option.

Consequential results – or not?

Experimental measurements are often considered essential in designing a new chemical manufacturing plant. While this is usually the case, it’s not always necessary. Engineers should carefully evaluate whether specific tests impact design decisions — or if the outcome remains the same regardless of the results. One way to identify such optional tests is by asking, “If we had the results now, what design decision would change?”

Consider the following example: an engineer is tasked with designing an acid gas absorber column to remove H_₂S and CO_₂ from a nitrogen stream, ensuring emissions meet permit limits. The 100-kg/h gas supply pressure is 20 psi, but the concentration of the acid gases is still unknown, expected to range between 0.05% and 5.0%. Although the gas will be tested, with results expected in three weeks, it’s worth asking: What if the result is 0.05%, 0.5%, or 5.0% — would the design change significantly?

Consulting the Process Selection Chart for Simultaneous Removal of H_₂S and CO_₂ [2] reveals that an amine absorber is suitable for acid gas fractions ranging from 0.05% to 5.0% at 20 psi. The selection might differ if the pressure were higher or the gas composition richer in acid gases. Since the gas flowrate to the absorber remains roughly constant — whether it’s 0.05 or 5 kg/h of acid gas mixed with 95–99.95 kg/h nitrogen — the absorber’s diameter is likely unaffected by variations in feed composition.

This leaves only the absorber height and amine circulation rate as variables determined by the feed analysis. A sensitivity analysis can still be conducted to determine the optimal liquid-to-feed ratio and stage count for both minimum and maximum feed cases. The results may show a negligible difference in stage count (for example, 10 versus 12), making it unlikely to significantly impact capital costs. This is because absorber stage counts are typically more sensitive to outlet specifications than to inlet concentrations.

In this situation, the project could benefit from a three-week schedule reduction by proceeding with the design based on the worst-case feed scenario and revisiting the analysis when more data are available. This accelerated schedule was made possible by addressing the question, “If the results were available now, what design decision would change?”

Final thoughts

The process of developing solutions for new separation challenges involves the following key steps:

• Identify and assign appropriate names to each step in the processing sequence
• List the potential separation technologies that could be applied at each step
• Determine the necessary experimental measurements required to choose the most suitable unit operation
• If certain data is unavailable, consider how its absence would impact the design and make adjustments accordingly

While it’s rare to have the complete information needed for a perfectly streamlined process design, effective selection diagrams and thoughtful inquiry can help overcome apparent roadblocks and compensate for missing data. ■

Edited by Dorothy Lozowski

References

1. Couper, James, and others, “Chemical Process Equipment: Selection and Design,” 2012.
2. Gas Processors Suppliers Association, “GPSA Engineering Data Book,” 2004.

Author

Chris Rentsch is the process development manager at Koch Modular Process Systems (45 Eisenhower Drive, Suite 350, Paramus, N.J.; Phone: (877) 780-8654; Website: kochmodular.com). He has spent the last 20 years in various production and process engineering roles in many industry sectors, including oil and gas, herbicides, lithium-ion batteries and biomass-to-renewable chemicals. Rentsch holds a B.S.Ch.E. from the University of Michigan.