Written by: Jackson Udy, PE

Process Engineer at Process Engineering Associates

May 12, 2025

One common decision made during the front-end engineering and design (FEED) of a project is selecting the appropriate means of heat transfer to and from the process. However, selecting a heat transfer fluid can feel difficult due to the many options available. This article outlines the benefits and drawbacks of the most common heat transfer fluids and quantitatively compares various heat transfer fluids for heating, cooling, and refrigeration service.

Introduction

Heat transfer fluids are used for a variety of services in the process industry such as:

• Fluid heating and vaporization
• Fluid cooling, condensing, and crystallization
• Reboiling liquids for distillation
• Combined Heating and Cooling of Batch Systems
• And many other additional services

The variety of process conditions encountered in the process industry means that the optimal heat transfer fluids may be different for each situation. Proper selection of heat transfer fluids results in lower capital and operating costs, and process engineers should evaluate the choice of heat transfer fluid carefully. The sections below describe the benefits and drawbacks of the most common fluids to help provide a guide for selecting a good heat transfer fluid.

Steam

Steam is an inexpensive and widely available heat transfer fluid. Due to its high film coefficient, using steam often results in smaller heat exchanger sizes. Steam is also available at most process plants and, if additional boiler capacity is available, it can be a very economical option.

Steam is quite useful for heating fluids to temperatures as high as 450°F. However, as process temperature requirements increase past 450°F, the applicability of steam becomes problematic. The vapor pressure of steam increases exponentially, and very high pressure is required to meet process temperatures above 450^oF. For example, saturated steam at 600°F has a vapor pressure of ~1,530 psig. A typical hot oil system operating at the same temperature will often operate below 150 psig. Thus, using steam for high temperature applications often requires more expensive equipment and piping compared to other available heat transfer fluid options.

Another item that should be considered  with steam use is when steam is being used in a dual-use jacket or coil, which are commonly applied to batch reactors.  Dual-use jackets use steam for heating and then switch to cooling water service for cooling. Swapping between steam and cooling water, for example, can lead to hammering in equipment jackets and coils due to the rapid condensation of steam. This can cause additional stress on equipment and should be considered if using steam in a system requiring both heating and cooling. Some designers will provide separate heating and cooling coils for batch systems, but this comes at an additional capital cost.

Finally, steam systems typically require chemical treatment to limit corrosion and condensate recovery systems to route condensate back to the boiler. These systems can add significant maintenance costs to the steam system.

Water

Water is the most common fluid used for process cooling in the process industries. Water is cheap and widely available. Additionally, water can be cooled by evaporative cooling via cooling towers, eliminating the need for refrigeration units. Some locations may even allow once through water cooling, where water is pulled directly from nearby bodies of water for process cooling and then returned to the body of water.

Water’s advantages are its low cost and high heat transfer coefficient. Water is orders of magnitude cheaper than other commercial heat transfer fluids. Additionally, water almost always has a higher heat transfer coefficient than heat transfer oils, glycol/water mixtures, or brine systems.

One of the main disadvantages of water is its limited operating range. Cooling water systems operating above 120°F are often subject to increased scaling and fouling. Typical supply temperatures for cooling water systems are about 85°F depending on the season and location, and chilled water systems cannot operate below 32°F. Water can also be highly corrosive and promote biological growth, therefore carefully monitored chemical treatment systems are often required to mitigate these issues. Finaly, cooling water systems often require substantial water blowdown rates that may require discharge permitting and have other environmental considerations.

Suppose that the force balance inequality above is not met? Then the PSV may be expected to reseat during relief, possibly leading to unstable operation. However, an acoustic analysis can determine if the reflection of the pressure wave will reach the PSV inlet before or after the valve closes. To perform an acoustic analysis, a ‘critical’ line length is calculated. A line length near or above the critical line length indicates that instability may occur.  API 520 Part II defines the critical line length as follows:

Heat Transfer Oils

Organic heat transfer fluids systems are common in many industries. The main advantage of organic heat transfer fluids in heating services is the reduced operating pressure compared to steam and subsequent reduced equipment costs. Many organic heat transfer fluids can also be used in cooling service, providing flexibility for batch processes. Heat transfer oils require no blowdown and may operate for many years without being changed out if operated within their recommended temperature range. Some heat transfer oils are designed to operate in chilled service, well below the freezing point of water. Most modern heat transfer oil systems operate in the liquid phase, but vapor phase systems are not uncommon. Due to the low vapor pressure and good stability of many heat transfer oils, many systems are operated open to the atmosphere and use equipment with low pressure ratings. The main drawback of heat transfer oils is their reduced heat transfer coefficient compared to steam and cooling water systems. Another drawback is the cost of heat transfer oils compared to water and glycol-based fluids. Most heat transfer oils are also combustible, which may be a drawback for some facilities.

Brine and Glycol Systems

Calcium chloride (CaCl₂) brine systems are water-based heat transfer fluids that decrease the freezing point of water and allow for use in low temperature applications.

CaCl₂ brine has a higher heat transfer coefficient but can be more corrosive to piping than glycol systems and requires chemical inhibitor programs to ensure mechanical reliability. However, with the right inhibitors and proper monitoring in place, carbon steel can be used in calcium chloride brine systems. Glycols on the other hand have much lower corrosion rates than salt brines, particularly inhibited glycols. Ethylene Glycol water mixtures are generally the preferred glycol due to low viscosity and increased heat transfer rates. Propylene glycols have lower heat transfer coefficients but are generally considered not toxic and are often used in food and beverage industry where incidental food contact is possible. Glycols also have higher maximum operating temperatures than brines, giving them some extra flexibility as they can be used for both cooling and low temperature heating below 300°F.

Selecting a Heat Transfer Fluid

Selecting a heat transfer fluid involves balancing a variety of concerns. Some of the factors to consider when selecting a heat transfer fluid are:

• Service temperature
• Film coefficient
• Fluid Cost
• Fluid Vapor Pressure
• Waste generation, blowdown requirements
• Food Contact Rating Required?

A primary concern when selecting a heat transfer fluid is to ensure a good heat transfer film coefficient at the required service temperature. Tables 1, 2, and 3 below show the heat transfer coefficients for various heat transfer fluids flowing at 6 ft/s inside ¾” tubes at 70°F, 20°F, and 400°F respectively. Various heat transfer fluids are shown in Tables 1, 2, and 3 to give a quantitative example of the tradeoffs between different fluids. Note that viscosity is usually the best predictor of heat transfer coefficient and, when looking for a quick evaluation between fluids, looking at viscosity will usually give a general idea of relative heat transfer coefficients.

Table 1: Heat Transfer (Film) Coefficients for Cooling at 70°F, 6 ft/s in 3/4″ ST HX tubes

Table 2: Heat Transfer (Film) Coefficients for Cooling at 20°F, 6 ft/s in 3/4″ ST HX tubes

Table 3: Heat Transfer (Film) Ceofficients for Heating at 400°F, 6 ft/s in 3/4″ ST HX Tubes

While film coefficients for each fluid can drive the economics of heat transfer fluid selection, some other factors to consider when selecting heat transfer fluids include:

• The required service temperature.
  
  -Does the fluid need to operate in both heating and cooling service?
  
• Fluid cost.
  
  -Generally synthetic heat transfer fluids are the most expensive, with water-based fluids being more economical.
  
• Process side film coefficient.
  
  -If the process side has a very low film coefficient (Such as vacuum condensing service or cooling a viscous liquid), then the process side may be the controlling resistance to heat transfer. In this scenario, selecting the heat transfer fluid with the highest film coefficient may not provide additional heat transfer, and choosing a more economical fluid with a lower film coefficient may make more sense.
  
• Is HT-1 food contact rating required?
  
  -Some processes require heat transfer fluids to be rated for incidental food contact in case of a leak.
  
• Does significant wastewater generation or discharge permits need to be avoided?
  
  – This may preclude steam or cooling water systems.
  

Final Thoughts

The variety of process conditions in industry means that the optimal heat transfer fluids are different for each process and location. Heat transfer fluids selection can have a significant effect on capital and operating costs and should be evaluated carefully. If you need a heat transfer system designed, or are having issues with an existing system, reach out to us at Process Engineering Associates. We have extensive experience with heat transfer fluids in a variety of industries. Give us a call to see how we can potentially help you select the best fluid for your application.

Written by: Jackson Udy, PE

Process Engineer at Process Engineering Associates

September 17, 2024

Over the past decade, an increased emphasis on pressure safety valve (PSV) systems has led to increased scrutiny of existing installations in many plants. Often these evaluations uncover gaps between a plant’s existing PSVs and recognized and generally accepted good engineering practice (RAGAGEP). Remedying these issues can lead to costly modifications. One of the most common issues operating companies run into is the 3% rule. Fortunately, new guidance by the American Petroleum Institute (API) on ‘engineering analysis’ allows for operating companies to potentially exceed the 3% rule and avoid costly facility modifications.

What is the 3% rule?

The 3% rule states that the non-recoverable pressure drop in the inlet piping to a PSV should be less than 3% of the valve set pressure. This guideline was developed as a way to ensure PSV stability and prevent valve chatter. Further, PSV relief capacities are based on the stagnation pressure at the nozzle inlet. Inlet losses that exceed 3% cannot be ignored when determining PSV capacity (see 520,II, 7.3.2 ). The 3% rule is cited in various standards, most notably API 520 Pt II and ASME Sec VIII Div I Appendix M. OSHA considers the 3% rule RAGAGEP, and thus it is considered a requirement for installed PSVs. The conventional way to meet the 3% rule for existing installations is through physical modifications such as:

• Installing a remote sensing pilot operated relief device or other relief device (such as a rupture disk) that is expected to operate in a stable range with inlet line losses above 3%.
• Reduce the capacity of the PSV, such that inlet line losses are reduced. This may be accomplished with a smaller orifice size or a restricted lift.
• Installing larger and/or shorter inlet piping to the PSV.

All of the above solutions require physical changes that may require significant capital expense
or may not even be feasible for some systems. However, there is another option to
consider to potentially avoid these costly modifications.

Engineering Analysis of PSV Stability

API 520 Pt II states that the 3% may be exceeded if “an engineering analysis is performed”. Historically many installations throughout the chemical and refining industries have not followed the 3% rule. In the past, it was not uncommon for operating companies to accept inlet pressure drops up to 5%. For example, one common rule of thumb is to keep PSV inlet line losses below the PSV blowdown percentage, plus some safety factor. This was often considered a sufficient ‘engineering analysis’ to exceed the 3% rule. However, API 521 Pt II has provided more definition on what should be included in an engineering analysis in the latest (7th) edition.

An engineering stability analysis usually consists of qualitative and quantitative assessment. The qualitative assessment typically includes a review of PSV relief history and inspection/repair reports to identify any history of chatter. If a qualitative analysis shows no signs of chatter the quantitative analysis is performed. A quantitative analysis consists of both a force balance and acoustic analysis.

What is a PSV Force Balance?

A PSV force balance is used to determine if a PSV is expected to stay open during a relief event by
comparing the sum of forces acting to open the valve against the sum of the forces
acting to close the valve. API 520 Part II gives an equation for a conventional PSV force balance as:

P_open – ΔP_f,wave – ΔP_wave – ΔP_built-up > P_close

Where P_open is the relief pressure, ΔP_f,wave is the frictional pressure drop as the PSV is opening, ΔP_wave is the recoverable wave line losses, ΔP_built-up is the built up backpressure, and P_close is the reseating pressure. A relief device would ‘pass’ the force balance if the relief pressure minus all of the pressure losses was still greater than the reseating pressure. This indicates that the PSV will stay open even as the pressure wave moves through the inlet piping, indicating that the valve will likely be stable during relief.

Performing an Acoustic Analysis

Suppose that the force balance inequality above is not met? Then the PSV may be expected to reseat during relief, possibly leading to unstable operation. However, an acoustic analysis can determine if the reflection of the pressure wave will reach the PSV inlet before or after the valve closes. To perform an acoustic analysis, a ‘critical’ line length is calculated. A line length near or above the critical line length indicates that instability may occur.  API 520 Part II defines the critical line length as follows:

L_crit = Ct_o / 2

Where C is the speed of sound, and t_o is the PSV opening time.

Putting it all together

Based on the new API guidelines, an engineering analysis including both a qualitative and quantitative assessment could allow non-recoverable pressure losses in the inlet piping to exceed the 3% using conventional PSVs. The flowchart below outlines the process of performing such an engineering analysis.

By performing these steps, operating companies can avoid costly physical modification to their facilities, while ensuring that they are meeting RAGAGEP guidelines and preventing PSV instability. If you have PSV systems that do not meet the 3% rule, or you are unsure what the inlet pressure drop for your system is, reach out to us at Process Engineering Associates. We have extensive experience with relief system evaluation in a variety of industries. Give us a call to see how we can potentially help you avoid costly modifications and have peace of mind that your PSVs will operate as expected.

References

1. American Society of Mechanical Engineers, ‘Boiler and Pressure Vessel Code, Section VIII, Division I, Rules for the Construction of Pressure Vessels’ (2019).

2. American Petroleum Institute, ‘API Standard 520, Sizing, Selection, and Installation of Pressure-Relieving Devices, Part II – Installation’, 7th edition (2020).

Written by: Christopher J. Muntean, P.E.

Senior Process Engineer at Process Engineering Associates

March 18, 2024

Background

Various assessments are conducted to promote the safe and reliable operation of process plants in the industry. Some of these assessments include: Hazard and Operability (HAZOP), Process Hazard Analysis (PHA), and Layer of Protection Analysis (LOPA).

During these workshops, a group of cross-functional individuals (often engineers and safety professionals), work together to identify the major hazards within process units. Once these hazards are identified, the team develops effective preventive measures. The identification of such hazards serves as a crucial step in mitigating potential safety and environmental incidents.

Estimating theoretical consequences during a hazards assessment is often done qualitatively. While value can be provided in this manner, it must be noted that other tools exist for a process engineer to support these events in a more quantitative approach. Namely, detailed dispersion modeling. Dispersion modeling can provide numerical results as well as visual graphical representation of the identified consequences.  The results from such modeling can be overlaid on plant plot plans or even Google Earth imagery during PHA/HAZOP/LOPA workshops to provide further clarity in evaluating hazardous scenarios.  Team members can then further enhance their understanding of an incident by visualizing the impact of the release to the three-dimensional surroundings, such as equipment or buildings.

PROCESS has experience using such modeling tools to support our clients during these assessments.

Practical Examples of Dispersion Modeling

A detailed dispersion modeling effort using well designed software generally allows for a wide range of equipment and consequence scenarios to be evaluated to support these hazard assessments. PROCESS has utilized industry accepted dispersion/modeling tools (i.e. Chemcad, ALOHA, DNV PHAST, etc.) to support clients with their hazard assessments. Some specific examples include: (1) pressure relief device (PRD) atmospheric release dispersion modeling, (2) flare thermal radiation and/or flameout modeling, and (3) In-building chemical releases. Such examples are further discussed as following:

1. Pressure Relief Device Modeling

Relief devices that are not connected to a closed relief system (flare header, knock out pot, etc.) should have the tail-pipes directed to a safe relief location. Determining if the atmospheric release is truly to a safe location is a requirement defined by OSHA and ASME. This practice is also recommended in API 520, Part I. Dispersion modeling can provide results to determine relief discharge piping height and location requirements to comply with these codes and standards. PROCESS has conducted various dispersion modeling assessments for clients who wish to understand the potential impacts of a toxic chemical release from a safety valve to the atmosphere.  PROCESS recently conducted dispersion modeling on a large anhydrous ammonia storage tank with 3 PRDs (atmospheric relief). The model indicated that during a fire scenario, the atmospheric relief could potentially result in toxic  conditions to nearby structures or locations in the plant. Our client used the data from the modeling to support their safety plan.

In addition, atmospheric PRD dispersion results can be overlaid on satellite imagery to provide further insight when evaluating hazardous scenarios.

2. Flare Thermal Radiation or Flameout Modeling

Flares provide a certain amount of thermal radiation to nearby pipe racks, buildings, people, and process equipment. This level of radiation changes depending on flaring amount, duration, and certain environmental factors such as wind or time of day. Dispersion modeling can support a project by allowing such radiation effects to be plotted over various radiation levels. This may prove to be beneficial should new equipment, buildings, or pipe racks need to be installed or re-routed.

Additionally, during a flare flameout scenario, a large flammable cloud may persist for quite some distance. Dispersion modeling can support the effects of such a scenario.  PROCESS has recently conducted a similar exercise for a client who was considering a hypothetical large hydrocarbon release occurring while the pilots were distinguished on the flare. Dispersion modeling for that exercise revealed where flammable regions may exist on the site plot plan.

3. Inside Building Releases

Inside building releases can be modeled. For example, an indoor leak in a lab may build up a large concentration prior to being released outside. The outdoor release and effects can be modeled, along with possible explosion effects.

In addition, indoor releases can be modeled considering asphyxiation, by showing oxygen depletion over-time at certain portions in a room or building. This may be valuable if there is a need  to consider a nitrogen or argon line rupture in an occupied room. PROCESS has conducted such an exercise for a client with a lab environment where an argon pipe rupture was considered. The room oxygen concentration was estimated over time as the leak event unfolded. This helped the client determine safety planning if such an incident should unfold.

Other Potential Uses for Dispersion Modeling

While a few practical examples were listed, dispersion modeling may also prove to add value for various other incidents or planning efforts. These include, but are not limited to:

• Emergency Response Planning
• Incident investigation
• Spill and Loss of Containment Events
• Pipeline Leaks
• Facility Citing and Occupied Building Risk Assessments

Conclusion

Whether the end user is in oil and gas, petrochemical, pharmaceutical, the public sector, or a storage facility; accurate dispersion modeling can be an important tool to help ensure safety is not compromised during the design and operation of a facility.

A detailed dispersion study can provide numerical and visual representation of potential consequences identified in hazard assessments (e.g. PRD relief, flare upset condition, inbuilding rupture releases, etc.).  Such results allow safety professionals to develop strategies to prevent future re-occurrences, develop best practices, and make informed decisions to emergency response plans and overall safety measures.

References

1. American Society of Mechanical Engineers, ‘Boiler and Pressure Vessel Code, Section VIII, Division I, Subsection A, Part UG-135(f) (2009).

2. Occupational Safety and Health Administration (OSHA), 1910 Subpart H, Hazardous Materials, Dispensing Devices, 1910.110(h)

3. American Petroleum Institute, API 520, Part I, 10^th Ed. (2020), Section 4 – Pressure Relief Devices

Written by: Jackson Udy, PE

Process Engineer at Process Engineering Associates

October 24, 2020

Due to the current economic climate, many refineries and petrochemical plants are running at reduced capacities. Operating companies are more closely scrutinizing operating expenses in this low margin environment. As a process engineer or operations professional, we have the opportunity to find ways to reduce operating expenses by reducing energy consumption.

One of the areas inside process plants that takes up significant energy is distillation and fractionation. Distillation consumes over 40% of the total energy in the refining and chemical industry(1). In fact, distillation is 6% of the total energy usage in the United States. This is an incredible amount of energy! Thus process engineers should be consistently asking, am I minimizing energy usage in my distillation columns?

Two ways to reduce distillation energy usage are eliminate excess reflux and reduce tower pressure.

Is Your Distillation Column Over-refluxed?

Is your distillation column over-refluxed? To answer this question, determine the specifications required for each stream exiting a column. For example, a depropanizer column in a refinery may require a 95% propane product purity and at least 95% butane concentration in the bottoms stream. Suppose the column is operating at 99% propane purity while meeting bottoms specifications. How much energy is being wasted in this scenario?

The figure below shows the difference in duty for the two cases described above. The 99% purity case consumes 33% more energy than the 95% purity case. For the 10,000 BPD example below, this translates to a savings of 5 MMBTU/HR. If energy in this scenario costs $3/MMBTU, this would come out to $130K per year in energy savings.

Typically operators will target above the required product purity to absorb any swings in tower operation while keeping product streams on spec. Reducing this ‘overshoot’ as much as feasible can result in great energy savings.

Over refluxing towers wastes a significant amount of energy, and sometimes with little or unnecessary improvement in product purity or yields.

Reducing Energy Usage by Decreasing Operating Pressure

Another way to reduce energy usage is by reducing tower pressure. This can be especially effective in the winter months, when cooling water and ambient temperatures are lower, increasing the available process cooling. Reducing tower pressure reduces energy consumption because the relative volatility of hydrocarbons increases at lower temperatures. This makes them easier to separate, i.e. requiring less energy. This is especially true of lighter hydrocarbons (C1-C6).

The Cox Diagram(3) above shows why decreasing column pressure reduces energy consumption. For any given compound, as column pressure decreases, the flash temperature also decreases. The diagram above shows the relationship between flash temperature and vapor pressure. As flash temperature decreases, so does the vapor pressure for each hydrocarbon. But more importantly, the difference in vapor pressure between the compounds increases. For example at 180 deg F, The relative volatility between propane and butane is

Relative Volatility @ 180 F= VP C3/ VP C4

= 400 psig/150 psig

= 2.7

Suppose the overhead pressure in a column is lowered until the overhead temperature reaches 100 deg F.

Relative Volatility @ 100 F= VP C3/ VP C4

= 200 psig/ 50 psig

= 4.0

As shown above, decreasing operating pressure will increase relative volatility. This in turn reduces the required energy to perform a given separation. Let’s return to our depropanizer example. Suppose we drop the operating pressure in this tower from 300 psig to 275 psig. This reduces the required energy in our tower by 7%, saving $25k/year in our example.

One thing to consider when reducing tower pressure is that the reduction in tower pressure will cause an increase in vapor velocity inside the tower. This increases the chance of flooding inside the tower.

Savings energy is not only good for the bottom line, it is also good for the environment, conserving natural resources. If you find this article useful, let me know in the comments below. What ways have you reduced energy usage in distillation columns?

References

1) https://www.emersonautomationexperts.com/2010/industry/downstream-hydrocarbons/reducing_distil/#:~:text=Did%20you%20know%20that%20there,energy%20consumed%20by%20U.S.%20manufacturers.

2) Lieberman, Norman and Elizabeth Lieberman. A Working Guide to Process Equipment. McGraw-Hill Education, 2014.

3) https://petrowiki.org/Vapor_pressure

Recently I saw a safety article that resonated with me. The topic, Trusting Plant Instruments.

Few days go by where myself or a coworker don’t question an instrument reading at the plant I work at. And how could you not? Everyday, process engineers and operations personnel are looking at dozens or hundreds of instrument outputs. And every so often, we come across a reading that doesn’t make sense. But how operation personnel respond to an instrument that “doesn’t make sense” can have serious process safety implications.

I had an experience that made this principle quite clear to me. The process flow diagram below shows the process I was working with.

I was helping with the start up of a new bypass around an exchanger. We were monitoring the feed temperature to the stabilizer tower. If the stabilizer feed temperature got too high, liquid traffic in the stripping section of the tower would be reduced, and stripping efficiency would suffer. Our goal was to keep the feed temperature at the design value of 215 deg F.

As shown above, the unit described above has two feed temperature indications. But one was reading above the design value, and one was below. So then, which one is correct? Both values can not be correct, as there is no heat exchange occurring between the temperature indications.

I looked through the process graphics and trends, self-assured that I would determine which instrument was faulty. But in my search to justify my own assumptions, I realized something unexpected, both instruments were correct.

How can this be? Well dear reader, if you study the process flow diagram above, you will notice a control valve between the two thermocouples. A small portion of the stream flashed as it flowed through the control valve, cooling the stream before it entered the tower. Indeed, both thermocouples were correct!

So why am I telling this story? I believe it teaches a valuable concept. I didn’t understand how both instruments could be correct, so I immediately assumed one reading was false. This is the wrong way to think about instrument readings. When plant personnel ignore instrumentation, it is often because we don’t understand how the reading is possible. We then discard the reading as erroneous to justify our own notions of what is occurring inside our processes. This can lead to serious process safety implications, as the two examples below show.

In the 1960’s, a fire and explosion occurred chemical plant in Tennessee during a unit start-up (1). a thermocouple inside a distillation column was reading 250 degrees when it “should have” read 215 degrees. An instrument technician was sent into the unit to troubleshoot the instrument just before the explosion. A post-incident investigation revealed that the high temperature readings were consistent with increased Nitrobenzene content on that tray, likely due to tray damage. Had plant personnel not ignored this information, the incident may have been avoided.

Before and after of a Tennessee chemical plant fire

Another incident occurred in the 1990’s at a California Oil Refinery (2). A hydrocracker reactor outlet temperature went off-scale high, indicating a run away reaction. The control board operator was hesitant to believe the temperature reading, and did not depressure the reactor to stop the reaction. The ensuing fire and explosion resulted in one fatality and 46 injuries.

Not trusting process instrumentation can lead serious incidents. This is especially true for instruments that indicate an unsafe condition. So what can you do to make sure this doesn’t happen to you?

1. Assume an instrument is working until proven otherwise, especially when the reading could indicate an unsafe process condition. Ask yourself, What are the consequences if this instrument is reading correctly? Then take appropriate action.
2. Use other instruments and samples to confirm suspicious readings.
3. And finally, don’t assume an instrument isn’t working because you don’t understand how the reading could be correct!

Sources

Center for Chemical Process Safety April 2019 Safety Beacon
EPA Chemical Accident Investigation Report Tosco Avon Refinery, Martinez, California

written by Michael Tanzio, Process Engineering Associates, LLC; February 2021

The industrial literature is full of stories about poor separation in systems consisting of drops or particles in a separate continuous phase (gas or liquid).  Poor disengagement of the dispersed phase can lead to major, unwanted, economic consequences (poor product quality, reduced throughput, environmental and safety consequences, plugged piping, damaged equipment, etc.). Many times, this poor performance can be traced to the following:

• Use of poor design methods for sizing the separation vessel
• Lack of complete and valid information on the system (volume fraction of dispersed phase, particle size distribution, particle shape and density, etc.).

But how do you know a design method is poor? Or what data and information is required for a high-quality design? These can be answered, in part, by understanding the assumptions and limitations inherent in the design method being used. If the assumptions and limitations for the system is within those of the design method, then your design is probably working. If not, then you are probably reaching for your wallet!

For this discussion “particle” refers to both solid particles and liquid drop particles.

The majority of disengagement calculation methods are limited by the following assumptions inherent in the calculations:

• The particles are rigid, undistorted spheres.
• The drag force is the major force contributing to momentum exchange between the phases. Particle rotation (Magnus lift force), gradient forces (concentration, velocity (Saffman force), pressure, temperature), virtual mass effects, Basset force effects, collisional forces, electrostatics, all friction, viscous stress and turbulent stress are neglected (references 15, 16).
• The fluid system is dilute with the particles existing at very low volume fractions in the total fluid.
• The continuous phase is a quiescent fluid and relative velocity is negligible.
• The particles are not hindered in their settling by other particles.
• Dispersion bands are not considered.
• The fluids are Newtonian.
• The particles are larger than the mean free path of the continuous gas phase and Brownian motion is not important.
• Wall effects on particle settling are negligible.
• There are no phase changes or mass transfer occurring in the vessel.

A number of other assumptions also limit the use of many design methods. The ones listed above are considered the major factors limiting current design methods.

Although these assumptions should satisfy the majority of the systems analyzed, there may be cases where these assumptions are not valid. A qualitative discussion of some of these effects is presented next with some guidance on when they should be considered. References are also cited for further information.

Particle Distortion

The equations used to estimate terminal velocity are for rigid, spherical particles in a stagnant medium. However, liquid drops in motion may not be rigid or spherical. Up to a certain transition drop size, the terminal velocity can be higher than estimated due to internal circulation within the drop. As the drop size increases, a transition point is reached where the drop starts to oscillate and distort from a spherical shape. At this transition, the terminal velocity is a maximum and starts to decrease as the drop size gets larger.

Complicating the phenomena are the effects that contaminants or surfactants may have on the internal circulation and drop distortion. The demarcation between a spherical particle and a distorted particle is a function of both the Eotvos (Eo) number and particle Reynolds (Re) number. Below a particle Re of about 1.2, particles should remain spherical at all Eo numbers. At Eo numbers greater than about 0.4, particle distortion may be important. Either the Clift correlation (References 1 and 2) or the Hu & Kintner correlation (Reference 3) can be used to estimate the effect of particle distortion.

Dilution and Hindered Settling

The majority of design methods assume the concentration of drops in the dispersed phase is very low. These methods are probably valid only for dispersed phase volumetric hold-ups below 0.03 in the Stoke’s regime and below 0.05 in the Newton regime. This is due to two factors:

1. As a design requirement, many current design methods set the maximum superficial velocity of the continuous phase to be equal to the terminal velocity of a single particle in a stagnant fluid. However, the real velocity of the continuous phase is the proper velocity to use, and, in a moving fluid, the drag force depends on the relative velocity between the drops and the continuous phase. In a multiphase system, when the drag force balances the gravity and buoyancy forces, the resulting velocity is the relative velocity between the phases.

But most methods assume the resulting velocity is the terminal velocity of the dispersed phase in a stagnant fluid. A particular terminal velocity applies only to one particle diameter. By making the continuous phase velocity equal to the terminal velocity, all particles having that terminal velocity will not move or settle out. If particles and drops having only that particle diameter exist in the system, nothing gets separated until solid particles collide or liquid drops coalesce.

• At high dispersed phase hold-ups, dispersed phase settling becomes hindered and the terminal velocity is significantly reduced. Unlike hindered solid settling, as the dispersed phase hold-up increases, drop coalescence becomes more important. At some point coalescence starts affecting the drag force on the drop. Initially, at low coalescence rates, the terminal velocity decreases due to the hindering by other drops.  As coalescence becomes more important, the terminal relative velocity starts to increase and dominates the hindering effect.

For solid particles, unless they stick together, no such coalescence occurs. The terminal velocity is then reduced and the solid particles become entrained in the continuous phase.

In the design of many systems, these effects are usually ignored. Either a conservative, effective, design particle diameter or a disengagement “K” value is used. The K-value method (reference 17) estimates the required continuous phase velocity to be used for design.

The design particle diameter and / or K-value is based on past experience. Although providing a workable design, the use of these methods can be very conservative, or they can provide for an unworkable design.

What if a design value for a particular system is not known? Guidance on design values are available only for a relatively small number of industrially significant systems. If a system deviates from these “norms”, then the dispersed phase hold-up, and the real particle diameter to be disengaged, needs to be known. A rough estimate of the dispersed phase hold-up can be made by assuming a homogeneous flow regime similar to vapor/liquid, two-phase pipe flow system analysis. The hold-up is then just the fraction of the total volume that is the dispersed phase.

However, if significant slip between the phases occurs, the homogeneous hold-up may not be correct. Relatively small changes in the dispersed phase hold-up can result in large changes in the dispersed phase settling velocity with ramifications for vessel sizes. For example, past experience (Reference 4) on a liquid/liquid system showed the ratio of dispersed phase velocity to that of a single particle to be about 0.7 at a dispersed phase hold-up of only 0.05 and about 0.5 at a hold-up of only 0.1.

If the homogeneous hold-up is greater than about 0.03, then experimental measurement of the actual hold-up needs to be considered.

The determination of solid particle diameters is relatively easy compared to that of liquid drops. Liquid drop diameters are difficult to estimate since they depend on the upstream process history of the fluid entering a vessel. It is well known that drop size within a system is inversely proportional to the energy dissipated per unit mass in that system. Therefore, upstream equipment with high energy input (agitators, pumps, etc.) or with high energy losses (high pressure drop control valves, pipe frictional losses, etc.) will tend to produce drops with small diameters. Various correlations are available to estimate drop sizes but they should be used with caution. Further information on drop size estimates can be found in References 5, 6, 7, and 8. Agreement by the client for the drop size estimated and to be used for vessel design is imperative. Note that this must be a true diameter and not an effective design diameter. 

Once particle hold-up and diameter are known, the terminal velocity for a single particle in a stagnant media can be corrected to account for the presence of multiparticles. A number of methods are available:

• Ishii and Zuber (reference 9) developed drag coefficient and relative motion correlations from simple similarity criteria and a mixture viscosity model. They showed that the drag law for dispersed two-phase flows of bubbles, drops and particles could be put on a general and unified basis. Their unified method also accounts for the distortion of bubbles and drops.

• A Richardson / Zaki type of correlation as presented by Maude & Whitmore is described in references 3 and 10.

• The Zenz correlation for particulate fluidization (reference 11) can be used to directly estimate the relative velocity. The use of Zenz’s correlation to adequately determine the relative velocity in a counterflow, spray tower extractor is described in an older edition of Perry’s (reference 12). Note that much of the data Zenz used for his correlation was based on sedimentation data.

• The continuous phase superficial velocity can be estimated for vertical liquid/liquid separators directly from the extractor flooding correlation of Minard and Johnson (reference 13). The design velocity can be chosen at about 40 percent of the flood velocity.

• Rietema discusses other correlations in his review article (reference 14). This article is also an excellent discussion of dispersed two-phase systems.

None of these methods have been evaluated or compared. They are offered as starting references. Applicability to the particular system being analyzed should be determined.

As the dispersed phase hold-up approaches 0.2 to 0.3, drop coalescence effects may become important. Larger drops form due to the “crowding” caused by the high dispersed phase volume fraction. This author does not know of any current method for determining these coalescence effects.  Reader suggestions are welcomed. However, a workable, conservative design can be obtained by neglecting coalescence and using a smaller particle size. For hold-ups of this magnitude, two separators may be required: one to knock-out the bulk of the dispersed phase and a secondary one for removal of the smaller particles.

Dispersion Bands

Liquid / liquid gravity separator performance depends on two phenomena:

1. the movement of drops to the liquid / liquid interface;
2. the coalescence of drops at the interface.

Liquid drops crossing the interface can form a deep band of dispersed drops (reference 7). Many design methods do not allocate vessel volume for a liquid dispersion band. They usually assume, for a liquid/liquid system, a dilute system and a conservative, effective drop diameter for design. This combination may provide enough volume for any dispersion band that may actually develop. But I have seen no data to confirm that. It is prudent to base the design on the actual, measured dispersed phase hold-up and the real, measured drop diameter. Adequate volume needs to be provided for a dispersion band. But if the band height is a significant fraction of the separator depth, a small increase in feed rate could flood the separator with the dispersion band resulting in poor, or no, liquid / liquid separation.

Davies (reference 5) presents a correlation for estimating the time required for coalescence of drops in a multi-drop system. Reference 7 provides some design criteria for determining the dispersion band volume.

Other Limitations

Other major assumptions are that the fluids are Newtonian and the drop sizes are above 2 mm. Perry’s (Reference 1) provides some guidance on non-Newtonian fluids. For small drops in a gas, the size can become comparable to the mean free path of the gas. Below about 2 mm in diameter, the terminal velocity of a drop in a gas needs to be corrected by applying the Stokes-Cunningham correction factor (see Zenz, Reference 11).

For particle sizes below 0.5 mm, Brownian motion becomes significant and can dominate gravity (see Reference 11). At these very small sizes, normal available disengagement calculation methods are not valid.

References

1. Perry’s Chemical Engineers’ Handbook, 7^th ed., P. 6-53.
2. Handbook of Multiphase Systems, G. Hetsroni (ed), Section 1.3.9, “Hydrodynamics of Drops and Bubbles”, J.R.Grace, M.E.Weber, P.1-204, McGraw-Hill, NY, 1982.
3. Handbook of Multiphase Sytems, G. Hetsroni (ed), Section 9.4, “Liquid-Liquid Separation”, R.A. Jaisinghani, McGraw-Hill, NY, 1982.
4. Private Communication, Fluor Corporation, 3/2002.
5. Handbook of Multiphase Sytems, G. Hetsroni (ed), Chapter 4, “Liquid-Liquid Systems”, J.C. Godfrey and C. Hanson McGraw-Hill, NY, 1982.
6. Perry’s Chemical Engineers’ Handbook, 7^th ed., P. 18-21.
7. Handbook of Separation Process Technology, R.W. Rousseau (ed.), Chapter 3, “Phase Segregation”, L.J. Jacobs, Jr., W.R. Penney, P.148-156, Wiley & Sons, NY, 1987.
8. Mixing in the Process Industries, 2^nd ed., N. Harnby, M.F. Edwards, A.W. Nienow (eds.), Chapter 14, “Dynamics of Emulsification”, D.C. Peters, Butterworth-Heinemann, 1992.
9. M. Ishii, N. Zuber, “Drag Coefficient and Relative Velocity in Bubbly, Droplet or Particulate Flows,” AIChE Jul, Vol 25, No. 2, p 843, 1979
10. Perry’s Chemical Engineers’ Handbook, 7^th ed., P. 6-52.
11. Fluidization and Fluid Particle Systems, F.A. Zenz, D.F. Othmer, P. 236, Reinhold Publishing, NY, 1960.
12. Perry’s Chemical Engineers’ Handbook, 4^th ed., P. 21-24.
13. Perry’s Chemical Engineers’ Handbook, 7^th ed., P. 15-32.
14. K. Rietema, “Science and Technology of Dispersed Two-Phase Systems-I and II”, Chemical Engineering Science, Vol. 37, No. 8, P.1125, 1982.
15. The Handbook of Fluid Dynamics, R.W. Johnson (ed.), Chapter 18, “Multiphase Flow: Gas / Solid”, C. Zhu, L.S. Fan, CRC Press, NY, 1998.
16. Flowing Gas-Solid Suspensions, R.G. Boothroyd, Chapter 2, Chapman & Hall Ltd, London, 1971.
17. GPSA Engineering Databook, 11^th ed., Vol. I, Chapter 7, 1998.

Designing critical OSBL facilities for the Sadara Project presented the UPI team with some unique challenges, which required innovative solutions.

Written by Mr. Joe Matherne, Chief Process Engineer, Process Engineering Associates, LLC

December 7, 2020

The Sadara Chemical Company is a joint venture between The Dow Chemical Company and Saudi Aramco.  Sadara constructed at Al Jubail, KSA the world’s largest petrochemical complex ever executed in a single phase.  Total CAPEX for the Sadara project was approximately $20 billion.  Its 6-km² site features 26 integrated, world-scale manufacturing plants that produce a total of over 3 million metric tons of high value-added plastics and chemical products each year. (1)

The Sadara site includes a world-scale Mixed Feed Cracker (2), which went live in 2016 (7), and produces approximately 1.5 million tons per year of ethylene and approximately 400,000 tons per year of propylene (5, 8).

Figure 1- Sadara Configuration (2)

The design teams for the 26 process plants were located in contractor offices all over the world during the FEED and detailed design phases of the project (5, 8).  The scope of the Sadara Utilities, Power, and Infrastructure (UPI) Project (3), which was developed in the US and UK, included:

• Steam Generation and Utilities
• Power Distribution
• Interconnecting Facilities
• Waste Treatment
• Interface Management

The Sadara Steam Generation plant featured six 80MW boilers, each fitted with advanced NOx and SOx emission control systems (6).  The UPI Interconnecting Facilities scope included distribution of 6,000m3/hour of desalinated water produced by a third party for makeup to cooling towers located throughout the site (4).  The Waste Treatment scope included two world-scale Thermal Treatment Units to process hazardous, toxic, and halogenated hydrocarbon wastes generated across the complex (7).

Designing critical OSBL facilities such as these supporting a megaproject like Sadara presented the UPI team with some unique challenges, which required innovative solutions.

Practically all of the UPI project facilities had to be mechanically complete and commissioned before any of the process plants could begin to start up.  Yet the schedule for the Sadara project required the UPI team to finalize the design of their facilities before the design for many of the process plants had advanced far enough to allow their utility and waste treatment requirements to be firmly set.  To overcome this schedule disconnect, as the UPI team collected utility and waste treatment demand data from the teams designing the individual process plants, they were asked to not only define the expected normal and peak stream flowrates, they were also asked to honestly assess how firm that data was.  Based on this information, the UPI team assigned an “uncertainty factor” to each bit of design data, which reflected the extent to which the utility and waste treatment demands reported by the process plant design teams could grow as their designs progressed.  Essentially a design margin, an uncertainty factor of 1.0 was assigned to data UPI received regarding any process plant whose design was finalized and thus considered firm.  A higher uncertainty factor, up to 1.5, was assigned to design data UPI received regarding any process plant whose design was preliminary, and hence potentially subject to significant growth.  Although subjective and imprecise, this approach nonetheless allowed the UPI team to assemble the design data for their facilities in a manner consistent with the overall Sadara project schedule.

Figure 2 – Sadara Steam Boilers Under Construction (3)

The UPI team next had to aggregate the individual bits of uncertainty-adjusted utility and waste treatment demand data regarding the 26 individual process plants into design requirements for the overall Sadara site utility and waste treatment systems.  UPI was charged with designing cost-efficient facilities that would support up to the peak flowrate for each utility stream going to or waste stream coming from each of the process plants, whether during startup, shutdown, or any other non-steady-state operating condition.  And yet it would have been uneconomical for UPI to design their facilities to accommodate all the peak flowrates for any given utility supply or waste treatment system from all relevant plants occurring simultaneously.  Thus, for the first pass, the UPI team adopted the philosophy of designing each utility and waste treatment system for the “sum of the normals plus the greatest single peak.”  In other words, UPI started with the assumption that all peak utility demand or waste generation scenarios for the various process plants would occur independent of each other.  Then UPI went back and, in collaboration with the design teams for the process plants, identified scenarios in which one or more of the utility or waste generation peak flows could occur simultaneously, and only designed for such scenarios by exception.  Thus, the Sadara UPI team met its obligation to design cost-efficient facilities that would support the full range of expected operations for the process plants.

Figure 3 – Sadara Power and Piperack Interface (3)

Due to its massive scale, UPI’s interface management role presented the team another unique challenge.  There were several hundred interface points between UPI’s facilities and those of the 26 individual process plants to be tracked and managed as the Sadara project progressed.  These interface points included not just utility and waste streams, but also all raw material feedstocks and any process streams exchanged among the process plants that passed through UPI’s Interconnecting Facilities.  At first, the only information that was known regarding each of these interface points was what might be found in a typical process simulator – normal and peak flowrate, composition, temperature, pressure.  Then, as the design progressed, the interface data was expanded to include pipe size and flange rating, which had to match between the UPI and ISBL sides of each interface point.  During detailed design, the interface data was further expanded to include the physical location in space for each interface point, which again had to match exactly – to the millimeter.  During construction, the interface data was finalized by including the date when the section of pipe making up each half of each interface point was expected to be installed.  None of the Sadara interface management work was azeotropic distillation.  But just the sheer volume of data that had to be exchanged, checked, and repeatedly rechecked during each phase of the project with teams scattered all across the globe for the hundreds of individual points in UPI’s interface management database made this critical aspect of UPI’s scope a major undertaking.

Figure 4 – Sadara 2km Interconnecting Facilities Piperack (3)

Hydraulic hammer, or surge, is typically only considered in process plants for long liquid lines.  But since UPI’s Interconnecting Facilities spanned across Sadara’s entire 3km x 2km site, every liquid line was a long line.  UPI conducted over a hundred hydraulic hammer calculations and ten dynamic simulations, and then managed the resolution of all problems that these evaluations uncovered.  Vapor waste streams presented the UPI team a different challenge.  UPI conducted heat-loss calculations on each of the vapor waste streams between the sending process plants and the UPI waste treatment systems, and identified the insulation and/or heat tracing needed to keep each stream safely above its dew point throughout.

Figure 5 – Sadara 35m Wastewater Tank Construction (3)

In the end, the Sadara Project was a resounding success due in no small part to the contributions of the Utilities, Power, and Infrastructure Team that designed and installed critical OSBL facilities supporting the entire $20 billion venture.  The Sadara UPI team learned along the way how to overcome the unique process risks associated with such a megaproject, stitching together Sadara’s 26 process plants so they could work together as parts of a unified whole.

References

• Sadara Corporate Brochure_EN.pdf
• Dow_Sadar_Update20Mar2014.pdf
• Sadara Integrated Chemicals Complex – Fluor.html
• Pentair – Case study about Sadara Codeline.html
• Dow and Saudi Aramco continue to award Sadara packages – 2B1stconsulting 2B1stconsulting.html
• Semi dry scrubber systems Hadek Sadara Steam Station.html
• Sustainability Report 2018 – Landscape.pdf
• Sadara Chemicals Complex, Al Sharqiya, Saudi Arabia – Chemical Technology.html

Adsorption on Activated Carbon – Part 2

Part 2 of this blog topic continues the discussion from last month of some important aspects of adsorption on activated carbon in the process industries and both parts are intended to serve as an introduction to the topic. The references cited at the end of this article are good starting points for further information.

Last month, in Part 1, we discussed some characteristics of activated carbon; the importance of understanding bed capacity; tradeoffs to be considered when choosing an adsorbent particle size; and fixed bed and carbon cannister units.  This month, in Part 2, we will discuss the following:

• regeneration methods
• contaminant disposal
• some tips to keep in mind when designing a carbon adsorption system

Regeneration

Activated carbon which has reached its working capacity can be either regenerated for re-use or disposed and replaced with fresh carbon. Regeneration is the reversal of adsorption and is sometimes described as “desorption“.

Various processing methods are available and can be classified into the following general categories (References 1,3):

Temperature Swing Regeneration

For this method, desorption takes place at a temperature much higher than adsorption. The elevated temperature shifts the adsorption equilibrium resulting in contaminant desorption and regeneration of the adsorbent bed. To remove the thermally desorbed contaminant from the bed, a purge or sweep gas is utilized. A cooling step then returns the bed to its adsorption temperature.

Temperature swing regeneration is characterized by high capacities at low concentrations. This results in relatively long cycle times (hours to days).  Most applications are for systems with low contaminant concentrations (purification).

Pressure Swing Regeneration

In this process, desorption occurs at a pressure much lower than adsorption. This pressure reduction shifts the adsorption equilibrium resulting in contaminant desorption and regeneration of the adsorbent bed. A purge step can also be utilized to increase contaminant recovery during this method of desorption. No heating or cooling steps are required for this method.

Most pressure swing regeneration cycles are characterized by low capacities at high concentrations. This requires that cycle times be short (seconds to minutes). Major uses for this process include purification and feed systems where contaminants are present at high concentration (bulk separations).

Inert Purge

Inert purging desorbs the adsorbate solely by partial pressure reduction. Many times this method is used in conjunction with the temperature swing and pressure swing methods. No heating or cooling steps are required for this method.

Like pressure swing regeneration, this method is characterized by low capacities at high concentrations with a resulting short cycle time. Bulk separations of contaminants not easily separable at high concentration and of weakly adsorbed components are especially suited to inert purging.

Displacement Purge

Regeneration can also be accomplished by using a purge fluid that is adsorbed at the cycle conditions. The purge fluid displaces the previously adsorbed contaminant. The contaminant is then swept away in the remaining purge fluid. A displacement purge can also be used in conjunction with the temperature swing and pressure swing methods

After displacement of the contaminant, some purge fluid remains in the adsorbent material. To remove the purged fluid during the next adsorption cycle, the contaminant in the feed must be more selectively adsorbed than the displacement purge fluid. But not all of the purge fluid may be removed. Therefore, a displacement purge can contaminate both the product stream and the recovered adsorbate. The displacement purge fluid must be carefully selected. A separation step (e.g., distillation) to purify the product during adsorption and another one to recover the purge fluid may be required.

In some displacement processes, steam is used to remove the displacement purge fluid. Sometimes a water wash can be used.

Disposal and Replacement

Disposing of spent catalyst and replacing it is one way of “regenerating” it.  Disposal of spent activated carbon may be difficult in many cases since landfilling is no longer allowed in many countries in the world. Instead of regenerating in place, the carbon can be sent to a central regeneration facility and replaced with either fresh carbon or regenerated carbon

Particular processes for regenerating spent adsorbents include:

• Hot steam
• Hot gas (air, nitrogen, for example)
• Internal thermal heating of the bed
• Chemical and extraction methods
• Biological treatment
• Vacuum systems
• Electric and electrochemical methods
• Use of supercritical fluids
• Use of microwaves
• Ultrasound system
• Use of gamma-ray irradiation
• Photochemical treatment

After desorbing the adsorbed contaminant, the remainder of the regeneration cycle can include drying, cooling, or otherwise preparing the adsorbent to again adsorb.  The regeneration step is the major consumer of energy in an adsorption process and does not depend on whether thermal energy or compression energy is used for regeneration (Reference 4).  Methods to conserve energy are further discussed in Perry’s (Reference 4).

Contaminant Disposal

What one does with the desorbed material from the desorption step depends on the toxicity, flammability and permitting requirements for the compounds being disposed.  It also depends whether or not the desorbed material can be easily and economically recovered. Costs for proper disposal of the desorbed gas or liquid can be significant and need to be considered when choosing a regeneration system. Also, spent carbon transport may require hazardous waste handling, adding additional costs.

There are three basic treatment options for the desorbed material:

Discharge to the atmosphere

For desorption using a hot gas, the desorbed gas, depending on its vapor concentration and permitting requirements, may be disposed of to the atmosphere.

Recovery

For large amounts of adsorbed solvent, or if the solvent is expensive, thermal desorption can be used for solvent recovery. Thermal desorption can also be used to recover any adsorbed residual product. Desorbed gases containing a dispersed solvent are immediately condensed or recuperated via membrane filtration. If steam is used for desorption, the recovered liquid will contain the solvent and water. The pure solvent can be recovered by decanting the water and then by distillation. The recovered material is then recycled to a storage tank or to the process.  However, the water can be contaminated, and you may be left with a water emissions issue.  

If the desorbed composition is complex, or the solvent or residual product quantities are too low, or the solvent cost is low, the solvents or residual product can be disposed by incineration. Burning the solvents directly in the gas phase, via after-burning, is also an option.

Incineration

For this option, steam may not be used to desorb but rather a hot gas may be used. When desorbing, the component to be removed is concentrated above the feed composition first entering the carbon bed adsorber such that the use of a combustion system for final destruction is facilitated. A common practice is to concentrate the recovered component to within 25% of its LEL. This method can be used to significantly reduce the amount of auxiliary fuel required in the combustion system. And it can facilitate economic design of recovery systems if the desorbed gas is captured by a membrane or recovered with a condenser. Commonly used combustion systems include flare, thermal oxidizer, incinerator, and furnace heater systems.

Safety, environmental and economic considerations all need to be considered when choosing a regeneration / disposal system. Important factors include:

• type of adsorbate – toxicity, flammability, corrosive nature, radioactivity

• types of adsorption – physical adsorption or chemisorption

• the costs of regeneration and disposal (contaminant and spent carbon)

Design Tips

1. Keller and co-workers (Reference 1), have developed guidelines, in the form of a process selection matrix, that design engineers can use for choosing the most effective and economical regeneration process. Both gas phase and liquid phase systems are addressed. Perry’s (Reference 4) also discusses the use of these guidelines.

2. Steam for regeneration should be superheated at a temperature higher than the boiling point of the adsorbate at regeneration pressure.

3. Steam may be the most commonly used regenerant but there are situations where it should not be used. These situations should be identified and carefully evaluated. For example, steam used in a degreasing operation that emits halogenated Volatile Organic Carbons (VOCs) could cause the VOCs to decompose creating other issues.

4. At the carbon surface, VOCs with a lower vapor pressure will tend to displace those with higher vapor pressure that have been previously adsorbed. Therefore, the bed’s capacity for the higher vapor pressure constituent decreases as the adsorption cycle proceeds. This needs to be considered when sizing the adsorber. A conservative design approach would base the adsorption cycle design on the least adsorbable component in a mixture and the desorption cycle on the most adsorbable component.

5. In most processes, the contaminant feed composition and feed flowrate to the adsorption bed can vary significantly. But the system must always function to prevent contaminant breakthrough. Therefore, the design basis loading for the bed should be the maximum contaminant loading expected over the adsorption cycle — not the average loading over the cycle.

6. Adsorber vessels are erected either in a vertical or a horizontal position. For horizontal vessels the carbon volume usually occupies no more than 1/3 of the vessel volume (Reference 9).

7. Actual vendor quotes should be used for estimating capital costs. It is not likely that carbon cannister costs will inflate with the construction cost index or general inflation.

8. A typical carbon bed has a useful lifetime of about five years. However, a lifetime as low as one to two years is likely if the adsorbed material is very difficult to desorb, polymerizes, or reacts with other constituents.

9. If cannisters are likely to contain benzene they could be subject to the requirements in the Benzene Waste Operations NESHAP regulations. If so, disposal at landfills is not permitted and the carbon contaminated with benzene must be incinerated.

10. Factors limiting the effectiveness of carbon beds include:

• Relative humidity greater than 50% can reduce carbon capacity.
• Elevated temperatures (greater than 38° C or 100° F) inhibit adsorption capacity.
• Biological growth on carbon or high particulate loadings can restrict flow through the bed.
• Some compounds can cause carbon bed fires because of their high heat release upon adsorption.

11. Most adsorber systems are available as modular unit skids. When evaluating the pressure drop provided by a vendor for their unit, be sure the value provided is for the entire skid (inlet skid feed flange to outlet skid product flange) and not just for the carbon bed. Skid piping losses can be significant. Values for both the adsorption and desorption cycles should be obtained – they can be different. And for each cycle, start-of-cycle and end-of-cycle values should be provided.

12. Adsorption testing of different activated carbons from several manufacturers may be necessary to determine the most cost-effective system. However, these tests should be performed with conditioned, regenerated carbon. That will provide a more realistic estimate of the adsorptive capacity during operation.

13. An optimum adsorption cycle time exists. As the adsorption time cycle increases, the amount of adsorbent required increases. Therefore, the size, number, and cost of the adsorber vessels increase. Although shorter adsorption times require smaller adsorber systems with a lower capital cost, higher annual operating costs are required because the beds must be regenerated more frequently. In general, larger systems have better overall efficiency since less energy is used per pound of material adsorbed.

14. Selection of materials of construction needs to consider potential partial decomposition of components during regeneration. For example, chlorinated hydrocarbons can decompose forming hydrochloric acid.

15. During regeneration or adsorption, the temperatures should not be higher than the self-ignition point of the contaminant.

References

1. R.W, Rousseau (ed.), “Handbook of Separation Process Technology”, Chapter 12 by G.E. Keller II, R.A. Anderson, C.M. Yon, “Adsorption”, Wiley & Sons, 1987.

2. “EPA Air Pollution Control Cost Manual”, W. M. Vatavuk, W. L. Klotz, R. L. Stallings, EPA/452/B-02-001, Section 2/ Chapter 1 “Carbon Adsorbers”, 09/1999.

3. CATC TECHNICAL BULLETIN, “CHOOSING AN ADSORPTION SYSTEM FOR VOC: CARBON, ZEOLITE, OR POLYMERS?”, EPA-456/F-99-004, May 1999.

4. Perry’s Chemical Engineer’s Handbook, Section 16, Adsorption and Ion Exchange, 7^th Edition, McGraw Hill, 1997.

5. Wauquier, Jean-Pierre. (2000). Petroleum Refining, Volume 2 – Separation Processes. Editions Technip. A. Deschamps, S. Jullian Chapter 10: “Adsorption”; Chapter 11: “Adsorption in the Oil and Gas Industry”.

6. Cheremisinoff, Nicholas P.. (2000). Handbook of Chemical Processing Equipment. Elsevier. Chapter 5.3: “Adsorption”.

7. Albright, Lyle F.. (2009). Albright’s Chemical Engineering Handbook – Taylor & Francis. K.S. Knaebel, Chapter 14: “Adsorption”.

8. DG 1110-1-2, DEPARTMENT OF THE ARMY, U.S. Army Corps of Engineers,  Engineering and Design, “ADSORPTION DESIGN GUIDE” , 03/01/2001.

9. EPA Air Pollution Control Cost Manual (7th Edition), J.L. Sorrels, A. Baynham, D. D. Randall, K. S. Schaffner, Section 3.1 / Chapter 1, “Carbon Adsorbers”, 10/2018.

Adsorption on Activated Carbon – Part 1, M. Tanzio 09/10/20

Background

As a unit operation, many do not consider Adsorption a glamorous technology – “just put some charcoal in a drum and away you go!”.  But the physics involved and the engineering required for an effective system are quite challenging and sophisticated.

For a long time the unit operation of Adsorption has been a cost effective method for separating molecular components (Reference 1).  Since the fifteenth century, it was known that the color of solution liquids could be improved by adsorption with various materials.  In the late eighteenth century, sugar solutions were decolorized using bone char, and air for hospital respirators was purified in the nineteenth century utilizing wood charcoal.

Today, Adsorption is widely used for many molecular separations in the petroleum, natural gas, petrochemical, chemical and food industries. The most commonly used adsorbent material classes include molecular sieve zeolites, activated alumina, silica gel and activated carbon.

This blog will focus on just some of the important aspects of adsorption on activated carbon in the process industries and serves as an introduction to the topic. The references cited at the end of this article are good starting points for further information.

We will cover this topic in two parts. Part 1, to be covered this month, will discuss:

Activated Carbon

Activated carbon is available as powders, granules and pellets. and is commonly used to:

• reduce Volatile Organic Carbons (VOCs) and other air pollutants from process off-gases;
• reduce water pollution from process waste waters;
• purify off-gas for product recovery and downstream process use;
• remove contaminants from liquid products;
• protect downstream equipment (reactor beds, ion exchange equipment, etc.).

Activated carbon is classified as an amorphous, hydrophobic adsorbent (Reference 4). The extent of hydrophobicity of an activated carbon depends on its ash content and its level of surface oxidation. This type of adsorbent is a microcrystalline, non-graphitic form of carbon with internal surface area from about 400 m2/gm to 1800 m2/gm. It can be made from different materials depending on whether the activated carbon is used in a gas system or a liquid one. The properties of an activated carbon (hardness, density, pore size, particle size, surface area, extractables, ash, pH)  depend on the starting raw material from which it is made. These properties are varied and optimized for a particular service.  For decolorizing liquids, wood, peat, lignite and lignin are the usual starting materials. For use in gas phase systems, nutshells, coal, peat and petroleum residues have been used. Other materials used to produce activated carbon include: animal blood and bones, rice hulls, coconut husks, coal tars, pitches and carbon black.

Carbon activation is usually accomplished by pyrolyzing (carbonization or calcination) the starting material at a temperature of 400 °C to 500 °C to remove all volatile material leaving only the carbon. Depending on the application, the carbon is then partially oxidized to provide the required porosity and surface area. Inorganic chemicals are sometimes used to dehydrate the organic molecules during the pyrolysis step. Oxidation can be achieved by using:

• Air at low temperature;
• High temperature steam, flue gas or carbon dioxide
• Oxidizing agents (alkali metal hydroxides, carbonates, sulfates, phosphates).

Typical physical properties, taken from Reference 4, for carbon adsorbents:

For specific applications, and to improve removal efficiency, activated carbon can be chemically treated or impregnated. Impregnated activated carbon adsorbs and retains the specific components long enough for the chemical impregnant to react with the contaminant component  (chemisorption). Impregnated activated carbon has been specially designed to collect chemical components that are difficult to adsorb with standard activated carbon.

Systems used for carbon adsorption come in three basic types:

• filled cartridges (carbon canisters)
• packed, fixed beds
• powdered carbon injection.

Pre-packed filter cartridges are usually used in low-load applications and can be periodically replaced after their working capacity is reached. When a large quantity of activated carbon is required, such as for heavily-loaded systems, a fixed bed of granular activated carbon can be used.  For high flow rate streams with a low concentration of the component to be removed (dioxin removal, for example), powdered activated carbon can be injected into the stream. The spent powder, containing the adsorbed material, is then filtered away.

Vendors can offer these systems as pre-packaged, modular units. Fixed beds and carbon cannisters are discussed further later in this article.

Activated carbon which has reached its working capacity can be either regenerated for re-use or disposed and replaced with fresh carbon. Regeneration can be done on-site at the plant, or can be shipped to a central regeneration facility. Various regeneration methods are discussed next month in Part 2.

Some Important Process Parameters

Many process parameters and phenomena are important for designing a carbon adsorption system:

• Pressure, temperature, flowrate, and contaminant concentration
• Flow direction
• Carbon particle physical properties (hardness, density, shape, pore size, particle size, surface area, extractables, ash)
• Batch vs continuous operation
• Equipment capacity
• Desired recovery
• Cycle times
• Sorption Equilibrium
• Contaminant solubility in a liquid
• Contaminant molecular weight
• Moisture content and humidity of the feed
• Solution pH
• Adsorbent deactivation
• External and internal heat and mass transfer
• Hydrodynamics including axial and radial dispersion and pressure drop
• Reaction kinetics.

The references cited present additional information on adsorption theory, phenomena modeling and the effects of other process parameters on carbon adsorber design. Equipment vendors should also be consulted.

For this short discussion, we consider only two critical  parameters:

• Capacity
• Particle size.

Carbon bed capacity is an  important process parameter which can cause  some confusion. There are several types of capacities important for adsorber design:

• Saturation Capacity

This is the maximum capacity the adsorbent can hold. It should NOT be used alone to size the bed or to compare different carbon adsorbent materials.

• Breakthrough Capacity

This capacity is defined as the amount of material that is actually adsorbed before a concentration of the material exits, or breaks through, the bed that reaches either a limiting pollutant, toxic, or flammable concentration.

The required breakthrough capacity could vary from locality to locality depending on regulations and safety concerns.

• Heel Capacity

This is the amount of adsorbed material remaining in the bed after regeneration.

• Working Capacity

This capacity is the difference between the breakthrough capacity and the heel capacity after the bed has been “conditioned”. It is the amount of material actually adsorbed in each working cycle.

Before a stable amount of material can be adsorbed and regenerated, the bed usually must go through a number of operating conditioning cycles. About five cycles are required but should be confirmed with the equipment vendor. After conditioning, the carbon may only adsorb about 50% of the amount that was adsorbed by the starting virgin material.  This is thought to be the result of continued recapture of molecules (i.e., depressed vapor pressure) in the carbon micro-pores. Molecules, especially VOCs, have a difficult time desorbing from the carbon micro-pores.

This conditioning effect on working capacity must be considered when sizing the adsorbent bed. The conditioning effect can require up to twice as much adsorbent in the bed and will have a significant effect on installation, operating, and maintenance costs. Pilot plant or lab testing by the equipment vendor may be required.

The time required for breakthrough depends on the bed size. So does the time required for regeneration. Both conditioning and regeneration requirements need to be carefully considered when choosing the number of beds and the bed dimensions. The facility design would also need to consider the longer breakthrough times during the initial conditioning cycles of the virgin carbon material.

The working capacity is the capacity to be used for sizing the bed. A typical working capacity is 10-20 pounds of contaminant per 100 pounds of carbon but can vary depending on the application.

Another important parameter is particle size. Two considerations affect the choice of adsorbent particle size:

• system pressure drop
• mass transfer characteristics.

Pressure drop through a fixed bed can be reduced by choosing a large particle size. But the smaller the particle size, the greater the mass transfer rate and the less adsorbent required. This trade-off needs to be considered.

The trade-off is especially important for gas phase systems where compression costs can be significant. Compression costs need to be considered when determining the optimum vessel diameter. Typically, when only considering the mechanical design of the vessel, an optimum vessel diameter is obtained for a fixed quantity of adsorbent based on material stress and L/D ratio. The optimum diameter increases with decreasing pressure for this case.

But for a fixed quantity of adsorbent, the pressure drop also decreases with an increase in bed diameter. By considering the compression costs, the optimum vessel diameter could be larger than that dictated by mechanical stress conditions alone.

Types of Processes

Processes exist for treating both gases and liquids in carbon systems. Carbon used to treat a gas phase fluid usually contains a larger number of small pores than that used to treat a liquid phase.  Carbon systems are also used for treating liquid streams, however, it is very difficult to predict how effective carbon will be in treating a given liquid (Reference 6).  Laboratory tests are necessary and consists of two parts:

1. Preliminary isotherm tests to demonstrate the feasibility of using a carbon system
2. Laboratory column tests to obtain data for designing the commercial plant.

Tests for water-related applications may take several months. But most testing for chemical applications take less than a month (Reference 6).

The following types of adsorption equipment are used:

• Fixed regenerable beds;
• Disposable/rechargeable cannisters;
• Traveling bed adsorbers;
• Fluid bed adsorbers;
• Chromatographic baghouses.

For air pollution control, the most commonly used equipment are fixed-bed systems and cannister types. Since these are the most used equipment types, only they will be discussed here. Information on the other types of equipment can be found in the References.

Fixed Bed Units

Fixed-bed adsorbers can operate intermittently or continuously depending on the operation of the feed source.

In intermittent operation, the adsorber removes contaminant only during the time which the feed source is emitting contaminant. After the the feed source is shut down (e.g., overnight), the unit begins the desorption cycle. When desorption is complete, the bed sits idle until the feed source starts up again.

For continuous operation, the feed source operates continuously and a regenerated carbon bed must always be available for adsorption. At least two carbon beds are provided.  One is adsorbing while the other one is desorbing or idled. Since each bed must be capable of adsorbing the entire contaminant loading, twice as much carbon must be provided than an intermittent operation. If the adsorption and desorption cycle times are significantly different, three, four, or even more beds may be required.

For some units, the process conditions during regeneration are much more severe than those during adsorption. A single regenerator with materials of construction capable of handling the more severe conditions may be more cost effective than to make all adsorbers capable of regeneration by using the more expensive material of construction. Most water and wastewater treatment processes using thermally reactivated carbon utilize a separate vessel dedicated only to the regeneration step.

Carbon Cannisters

Cannister-type adsorbers are relatively small, returnable containers and include carbon beds made from 55-gallon drums. However, systems as large as 18,000 cfm are available in the industry (Reference 2). Cannisters are normally used for lower-volume, intermittent feed sources (for example, storage tank vents), where off-site regeneration is available and economical.

Carbon cannisters are not intended for on-site desorption. The spent carbon is usually sent to a central facility to be regenerated or disposed. Either the carbon or the entire cannister is replaced.

Most systems use two or more cannisters in series. Two or more parallel trains may also be specified.

Next month, in Part 2, we will discuss:

• Regeneration and disposal methods;
• Some tips to keep in mind when designing a carbon adsorption system.

References

1. R.W, Rousseau (ed.), “Handbook of Separation Process Technology”, Chapter 12 by G.E. Keller II, R.A. Anderson, C.M. Yon, “Adsorption”, Wiley & Sons, 1987.
2. “EPA Air Pollution Control Cost Manual”, W. M. Vatavuk, W. L. Klotz, R. L. Stallings, EPA/452/B-02-001, Section 2/ Chapter 1 “Carbon Adsorbers”, 09/1999.
3. CATC TECHNICAL BULLETIN, “CHOOSING AN ADSORPTION SYSTEM FOR VOC: CARBON, ZEOLITE, OR POLYMERS?”, EPA-456/F-99-004, May 1999.
4. Perry’s Chemical Engineer’s Handbook, Section 16, Adsorption and Ion Exchange, 7^th Edition, McGraw Hill, 1997.

5. Wauquier, Jean-Pierre. (2000). Petroleum Refining, Volume 2 – Separation Processes. Editions Technip. A. Deschamps, S. Jullian Chapter 10: “Adsorption”;
6. Cheremisinoff, Nicholas P.. (2000). Handbook of Chemical Processing Equipment. Elsevier. Chapter 5.3: “Adsorption”.
7. Albright, Lyle F.. (2009). Albright’s Chemical Engineering Handbook – Taylor & Francis. K.S. Knaebel, Chapter 14: “Adsorption”.
8. DG 1110-1-2, DEPARTMENT OF THE ARMY, U.S. Army Corps of Engineers,  Engineering and Design, “ADSORPTION DESIGN GUIDE” , 03/01/2001.
9. D.D. Randall, K.S. Schaffner, Section 3.1 / Chapter 1, “Carbon Adsorbers”, 10/2018.