Selecting an Activated Carbon for Regenerable Steam Systems

In adsorption processes using steam to regenerate activated carbon, the carbon’s physical characteristics can greatly impact efficiency and optimization

Processes using activated carbon as an adsorbent (Figure 1) can be ideal candidates for regeneration using hot fluids like CO₂, CH₄, N₂ or most commonly, steam. Adsorption is a unique process because it is almost completely reversible (although reversal may not always be economic). This means, given enough energy input, the bonds that hold adsorbates to the surface of the carbon can be broken. These physical forces are generally accepted as Van der Waals type and increase as the pore size of the carbon product becomes narrower [1]. For this reason, the term “adsorption potential” is used to describe the pore energy of a given type of carbon — in other words, its potential to attract an adsorbate on its surface, as shown in Equation (1):

ε = RTln(P_i/P_{v, sat}) (1)

where:

ε = adsorption potential energy (cal/cc)

R = gas constant, mol/J·K

T = temperature, K

P_i = vapor partial pressure, mm Hg

P_{v, sat} = saturation vapor pressure at T (K) and P (mm Hg)

Pore energy becomes an important factor in selecting activated carbon for steam regeneration because the product must have properties that align with the adsorbate of interest, as well as be compatible with and economically regenerated by steam. This also allows users to determine the working capacity for the adsorbate of interest, which directly informs the cycle times and steam demand.

Steam is an ideal choice for a regeneration fluid because it carries significant thermal mass (1.996 kJ/kg·K), it can be separated from the desorbed components and is widely available industrially, with established processes. Steam is used as a regeneration fluid in instances where one-time-use systems are not economical, or the user wants to recycle a chemical component back into the manufacturing process. Examples include industries with heavy solvent use, such as adhesives manufacturing, packaging, surface coatings and gravure printing. Additionally, regenerable steam systems are used to purify and recover valuable feedstocks in the polysilicon deposition process. They can also be employed for emissions and pollution control.

Presently, there are two main categories of emissions control for volatile organic compounds (VOCs). These include destruction of the VOC by conversion to CO₂ and water (thermal or catalytic oxidation) and separation and recovery via activated carbon temperature-swing adsorption (TSA) cycles. The second option allows users to recycle valuable chemical feedstocks. Steam regeneration provides economic and sustainability benefits, and helps companies meet various state and federal emissions regulations.

The selection of activated carbon has a significant impact on the cyclic efficiency of regenerable steam systems. The chosen product should align closely with the properties of the VOCs of interest while maximizing working capacity and minimizing specific steam demand (SSD). The physical properties of the activated carbon adsorbent, such as mechanical strength, particle geometry and purity also play important roles in overall efficiency.

 

Steam system components

The major components of a regenerable steam system include fans, heat exchangers, particulate filtration, vertical or horizontal activated-carbon adsorbers, steam boilers, condensers, decanters and, in certain cases, distillation units. Emissions of these systems can be monitored via infrared (IR) or other appropriate vapor-analysis methods to ensure emissions targets are met and signal the end of adsorption cycles. Each recovery system is unique and can present various challenges. These challenges include complex solvent mixtures, various unit-operation designs, safety considerations and ensuring proper alignment of carbon adsorbent with the VOC or steam system of interest.

 

Steam system operation

The operation of carbon-based regenerable steam systems can be complex, but the theory of operation is relatively straightforward. A simplified scheme is shown in Figure 2. VOC-laden air (typically > 1,000 ppmv) is diverted from an industrial process where it passes through pretreatment steps to: (1) remove particulates and (2) heat or cool the air to the desired adsorption temperature (generally 20–40°C). This step ensures the feed to the carbon adsorber is at optimal conditions for adsorption to take place. Process streams can be quite high in volume and may require multiple adsorbers working in parallel to treat.

Once in contact with the activated carbon, the VOC of interest is attracted to the carbon surface and subsequently adsorbed. Adsorption time is a function of the flowrate, VOC concentration, working capacity of the selected activated carbon and total carbon volume.

Once emissions targets are exceeded or after a preset cycle time, the adsorber beds cycle to allow for steam regeneration of the exhausted carbon, while standby adsorbers are brought online for seamless transition without operational delays. This type of sequence allows continuous operation with typical steam regeneration systems consisting of at least two and up to six beds.

In desorption mode, low-pressure steam is contacted countercurrent to the flow of the VOC-laden stream. The temperature of the steam is dependent on the solvent/carbon system configuration, but generally ranges from 120–180°C. Thermal energy carried by the steam through latent heats of condensation and vaporization allow for rapid bed heating and provide a strong driving force for displacement of the adsorbed VOC. Steaming should be carried out at a flowrate that can sufficiently carry desorbed VOC from the carbon bed (at least 0.1 m/s). Steaming generally requires 60–90 minutes, after which the VOC is sufficiently desorbed — or more precisely, economically desorbed.

After the carbon bed has been steamed, it is crucial to dry the bed out to ensure adequate performance for the next cycle. Adsorbed water on the carbon surface will reduce efficiency for the next adsorption cycle and can also set off emissions alarms if water vapor carrying miscible components is desorbed upon cycle startup. Dry utility air or nitrogen gas is passed concurrently to the operational flow to remove any adsorbed water.

Finally, a cooling step is used to bring the carbon adsorber back to ambient conditions where it will sit idle until the next adsorption cycle. It has been shown that drying and cooling can be achieved into a single step with just ambient air [2]. The VOC/steam mixture is separated further in a decanter or using distillation to recover the VOC for reuse back into the process. Condensed steam can then be discharged with facility wastewater for further downstream treatment. A properly designed regenerable steam system can recover thousands of pounds of VOCs per hour and recycle them back into a process with greater than 99% efficiency.

 

Activated carbons’ impact

Some of the highest associated costs of a regenerable steam system are attributed to energy consumption from blowers and steam boilers, as well as carbon media changeouts. A regenerable steam system can be operated more efficiently by ensuring the proper design around an engineered activated-carbon adsorbent that aligns with the demands of the application. The main points to consider with activated carbon selection are the following:

1.VOC type

• Molecular weight
• Boiling point

2. Carbon working capacity (lb VOC/lb carbon)

• Pore size
• Pore volume

3. Specific steam demand, SSD (lb steam/lb VOC)

• Function of (1) and (2) above

 

The carbon type selected for the VOC should be optimized to ensure the best performance and safe operation of the system. Activated carbon contains a broad range of pore sizes that are defined by the International Union of Pure and Applied Chemistry (IUPAC; www.iupac.org). Pore sizes can range from microporous (pores < 2 nm in diameter) to mesoporous (pores 2–50 nm in diameter) and macroporous (pores > 50 nm in diameter) [3]. Mean pore diameter (MPD) is an important characteristic that greatly impacts operation. Activated carbons used in regenerable steam systems will typically be made of coconut shell, coal or wood-based media. Each material possesses its own unique pore structure and pore-size distribution. The roles of these different pore sizes in a regenerable steam system are two-fold. First, the pore size dictates the pore energy, as described in Equation (1), and the resultant VOC adsorption capacity. Secondly, the pore size dictates the retentiveness of the carbon, which is its ability to hold onto the adsorbed VOC. These two functions of the carbon pore determine the working capacity, which can be defined as the difference between the saturation capacity under adsorption conditions and the heel (see Figure 3). The heel is the non-economically recoverable portion of VOC left on the carbon surface.

Carbon selection criteria

It is important to select a product that contains the best pore structure for the application in question. Microporous coconut carbons are ideal when the VOC of interest is of low molecular weight and boiling point because the very narrow pores possess enough potential energy to capture these compounds in an appreciable quantity, and they can be easily regenerated. Coal and wood-based carbon do not possess as many of these high-energy pores, thus their working capacity for low-molecular-weight VOCs will be far less.

Examples of commonly encountered VOCs that require coconut-based carbons are acetone, methyl ethyl ketone, methylene chloride, ethyl acetate, carbon disulfide and tetrahydrofuran. Figure 4 shows that the potential adsorption energy remains high on the coconut regardless of whether it is a low- or high-boiling-point VOC. However, in the case of toluene, the retentiveness of microporous coconut-shell carbon can lower the overall working capacity, making a mesoporous product more desirable. This higher proportion of mesoporosity allows for more efficient desorption of the VOCs under steam regeneration conditions, thus making the process more efficient. Examples of common VOCs that are better suited for mesoporous carbon types are hexane, toluene and xylene.

SSD is the most important factor for regenerable steam systems, and process optimization builds on the properties of the VOC and of the carbon product. SSD refers to the pounds of steam required to recover one pound of VOC. The goal in carbon selection should be to minimize the steam demand. Figure 5 shows a comparison of various carbon types (AC 1–4) and their steam demand for a particular VOC. The minimum value given by each curve indicates the ideal working capacity around which to design a regenerable steam system. As steam demand increases past the minimum point, the working capacity of the carbon increases, but it does so to the detriment of steam usage and energy consumption. SSD can range from as low as 1.3 lb steam/lb VOC to as high as 4.0 lb steam/lb VOC, depending on the system in question. For this reason, a given VOC-recovery system should be designed so that the steam demand is as low as possible. This is achievable only with proper alignment of the VOC properties with carbon properties.

Conveniently, the U.S. Environmental Agency (EPA; Washington, D.C.; www.epa.gov) has published various cost-analysis equations related to regenerable steam systems [4]. Users can calculate steam usage cost using Equation (2) shown below:

 

C_s = S_D m_voc θ p_s (2)

where:

C_s = Steam cost ($/yr)

S_D = Steam demand (lb steam/lb recovered VOC)

θ = operating hours (h/yr)

m_voc = VOC inlet loading (lb/h)

p_s = steam price ($/thousand lb)

Although not covered in detail here, the steam usage costs, as well as those associated with cooling water requirements, electricity requirements, carbon replacement and recovery credits can be found on the EPA website [4].

Additionally, operators should consider other physical properties of the activated carbon, because these will inform other costs associated with changeout frequency and energy consumption.

The hardness of the selected carbon should be high, because mechanically weak carbon will produce fines, causing increased pressure loss and energy consumption, and reduce cyclic efficiency. Activated carbon produced from coconut shells typically shows hardness values > 98% via ASTM D3802 [5]. Carbons produced from anthracite and bituminous coal generally trend >95%, and carbons produced from lignite and peat can range from 85–95%, depending on the quality and supplier. Mechanically compromised carbon will break down more quickly, generate more fines and cause uneven flow distribution through the adsorber. This directly impacts all cost areas associated with a regenerable steam system.

The geometry of carbon plays a role as well. Pelletized carbon affords better pressure-loss characteristics compared to granular activated carbon. This is because of the uniform packing of pelletized products and larger mean particle diameter. Energy consumption can be reduced by selection of proper particle geometry. Where possible, the product with the largest MPD should be chosen where performance metrics such as working capacity and SSD are still optimized.

Finally, there are compatibility considerations when choosing a carbon product for regenerable steam systems. For systems applied in traditional solvent recovery or in polysilicon production, it is crucial to have high-purity carbon with low levels of contaminants. Solvents such as acetone, methyl ethyl ketone and methyl isobutyl ketone (MIBK) are known to react and polymerize on activated carbon surfaces [6]. Use of these industrial solvents requires specially treated, low-ash-content carbon made from coconut shell. This is due to the heat released when in contact with carbon surfaces that contain high mineral content like coal-based carbon. Specially treated coconut carbons can achieve <1% ash content via ASTM D2584 [7], greatly reducing the likelihood of heat excursion or bed fire. For the purification of solar-grade hydrogen, carbon-based regenerable steam systems are used to recover costly silanes used in the deposition of polysilicon. Again, treated, high-purity coconut-shell carbon is required to minimize contamination of the hydrogen gas used in the process as impurities will lower the quality of the polysilicon produced.

Edited by Mary Page Bailey

 

Acknowledgement

All figures provided by author

References

1. Kim, M. J., Choi, S. W. and others, Simple synthesis of spent coffee ground-based microporous carbons using K₂CO₃ as an activation agent and their application to CO₂ capture, Chemical Engineering Journal, Vol. 397, October 2020.

2. Gu, J., Faqir, N. M. and Bart, J., Drying of an Activated Carbon Column after Steam Regeneration, Chemical & Engineering Technology, October 1999, pp. 859–864.

3. Thommes, M. and others, Physisorption of gases, with special reference to the evaluation of surface area and pore size distribution (IUPAC Technical Report), Pure and Applied Chemistry, Vol. 87, January 2015.

4. Sorrels, J. L. and others, Chapter 1 – Carbon Adsorbers, EPA Air Pollution Control Cost Manual, 7 ^th Edition, October 2018.

5. ASTM International D3802-16, Standard Test Method for Ball-Pan Hardness of Activated Carbon, May 2016.

6. Soehlberrg, N., Enneking, J. and Kovach, L., Avoiding Carbon Bed Hot Spots in Thermal Process Off-Gas Systems, 30^th International Conference on Thermal Treatment Technologies and Hazardous Waste Combustors, May 2011.

7. ASTM International D2866-11, Standard Test Method for Total Ash Content of Activated Carbon, June 2018.

 

Author

Lane Flora is an applications engineer at Jacobi Carbons, Inc. focused on air, gas, oil-and-gas and other specialty treatment markets (Email: [email protected]; Website: www.jacobi.net). He is responsible for initial application reviews, product recommendations, product qualifications, implementation, troubleshooting and various field-performance testing. He works closely with end users to provide effective treatment solutions using activated carbon. Lane graduated from the Ohio State University with a B.S.Ch.E. and later returned to complete his M.S.Ch.E.