March 2025
Process Optimization
Optimize sulfur recovery: Sub-dewpoint sulfur recovery with interstage membrane
The author’s company’s new sub-dewpoint sulfur recovery with interstage membrane (SSRIM) technology is a breakthrough for the sulfur recovery industry, as it is the first technology capable of achieving an overall recovery efficiency of > 99.9% without requiring a tail gas treating unit. This article discusses the basic SSRIM configuration and its advantages when compared to a traditional Claus/TGT configuration.
The Claus process has been, and continues to be, the “workhorse” of the petroleum industry for the recovery of sulfur from hydrogen sulfide (H2S). Traditionally, the process includes a thermal stage followed by several catalytic stages that allow for cumulative recovery efficiencies of 95%–98% to be obtained. With the advent of very stringent environmental regulatory requirements regarding sulfur dioxide (SO2) emissions, tail gas treatment (TGT) technologies were added to the sulfur recovery unit (SRU) configuration so that overall recovery efficiencies of > 99.9% could be achieved. Today, the most common SRU design incorporates a two-stage Claus plant followed by a reduction absorption amine-based TGT.
The capital expenditure (CAPEX) and operational expenditure (OPEX) of a conventional TGT unit (TGTU) is cost prohibitive, so a more attractive solution would be welcomed by industry. The author’s company’s new sub-dewpoint sulfur recovery with interstage membrane (SSRIM) technology accomplishes this while maintaining simplicity of design and operation. It relies solely on the Claus reaction—i.e., there is no requirement for TGT, which is an industry first. The technology is based on maximizing the Claus reaction via a combination of sub-dewpoint temperatures and interstage process water removal using membranes, resulting in overall recovery efficiencies of > 99.9%. By removing process water and knowing the Claus reaction is favored by colder temperatures, extremely high Claus conversions are achievable in a multi-stage sub-dewpoint converter, multi-membrane configuration. This article discusses the basic SSRIM configuration and its advantages when compared to a traditional Claus/TGT configuration.
Background. In 2021, Saudi Arabia’s Ministry of Environment, Water and Agriculture (MEWA) updated the “point source” regulations for SO2 emissions. The new regulations state that the maximum allowable SO2 emissions from any source should be < 250 ppmv on a dry and oxygen (O2)-free basis. For an SRU, this typically equates to an overall sulfur recovery efficiency of > 99.9%.
The SSRIM process. The basis of the author’s company’s SSRIM process is the removal of process water between catalytic stages. The removal of the process water—which is a product of the Claus reaction, along with elemental sulfur—allows the Claus reaction to move further to the right across the catalytic converters compared to a traditional Claus plant.
Since the early days of sulfur recovery via the Claus process, it has been understood that water removal would be very advantageous with respect to unit conversion across each catalytic stage. However, until now, process water removal in an SRU has been elusive. One idea that has been patented, but not commercialized, is a high-pressure (90 psig) SRU that allows for the condensation of both process water and sulfur vapor to occur simultaneously. While this idea is theoretically feasible, it introduces issues with corrosion and, therefore, has not been adopted by the sulfur industry.
For the SSRIM technology, interstage process water removal is obtained via membranes, which are placed within a sub-dewpoint SRU configuration:
- Traditional sub-dewpoint processes can achieve an overall recovery efficiency of 99%–99.5%, which is notably higher than a traditional three-stage Claus plant that can achieve an overall recovery efficiency of 98%.
- The sub-dewpoint process produces an unusual sulfur dewpoint depression effect—with respect to the sulfur vapor—in the final bed in adsorption, which results in negligible amounts of sulfur vapor in the outlet stream. This eliminates the typical loss in the overall recovery efficiency of approximately 0.2% from the final condenser due to sulfur vapor.
Combining the sub-dewpoint process with interstage membranes for process water removal results in ultra-high Claus conversion/recovery efficiencies—i.e., overall recovery efficiencies of > 99.9% (FIG. 1).
FIG. 1. The SSRIM process.
Simulations. TABLE 1 compares the theoretical cumulative recovery efficiencies that can be achieved for a medium-quality acid gas (55% H2S) and a high-quality acid gas (90% H2S) in a traditional two-stage and threestage Claus plant, a three-stage Claus with direct oxidation, a traditional sub-dewpoint Claus plant, and an SSRIM Claus plant.
Combustion air control. The SSRIM can achieve exceptionally high overall recovery efficiencies if the Claus reaction stoichiometry is precisely controlled. Higher recovery efficiencies require a higher accuracy of combustion air control, as shown in FIG. 2.1 Traditional combustion air control for a Claus plant includes a tail gas analyzer after the final condenser. A direct oxidation configuration requires the tail gas analyzer to be upstream of the direct oxidation reactor, while the sub-dewpoint process normally places the tail gas analyzer after the first catalytic stage (i.e., second condenser).
FIG. 2. The effect of air demand on Claus conversion efficiency.1
The tail gas analyzer “air demand” (the required air flow to maintain the targeted Claus reaction stoichiometry) feedback control loop for traditional SRU Claus configurations can control the combustion air flow to within +/- 0.3%–0.5%. As per FIG. 2, improved control is necessary to maintain overall recovery efficiencies of > 99.9%.
One of the limiting control parameters for optimal combustion air control includes the dead time for the traditional tail gas analyzer feedback signal for the combustion air control block, which can be > 60 sec. This makes it impossible to achieve optimal combustion air control during even mild process upset conditions. One solution for this is to use feed-forward control, which uses an acid gas analyzer to measure H2S and hydrocarbons. Unfortunately, these systems have proven to be unreliable due to sample system, configuration complexity and calibration issues. A more practical solution includes a tail gas analyzer downstream of the first condenser, which reduces the dead time of the feedback signal from minutes to several seconds.
While this setup can respond to mild to severe process upsets quickly, it does not account for the additional air demand required for the significant additional H2S produced across the first converter via hydrolysis of carbonyl sulfide (COS) and carbonyl disulfide (CS2). Therefore, an additional tail gas analyzer must be placed at a point downstream of the first catalytic stage. Field tests show that the optimal location for the second analyzer is after the final sub-dewpoint reactor to compensate for changing rates of adsorption and desorption of H2S and SO2 at different points in the sub-dewpoint cycle. In addition, minor amounts of COS and CS2 hydrolysis can occur across the sub-dewpoint reactors, which must also be compensated for in the total air demand. Note: The author’s company has successfully pilot tested tail gas analyzer operation after the first condenser with a close-coupled tail gas analyzer.
The second part of an optimized air control system for the SSRIM process will include the use of upgraded flow control valves (FCVs) for the acid gas, main air and trim air. For example, round ported or V-ball valves would allow for both fast-acting response and a tightened flow control compared to butterfly valves that are traditionally used in SRUs. Other than extreme upset conditions (e.g., liquid hydrocarbon carryover), this upgraded combustion air control system can control the air demand to +/- 0.1%.
Bench scale tests. Bench scale tests were conducted to determine if a suitable membrane could be found for removing water from a Claus process stream. Important requirements for the membrane include a non-reactive material, high water removal efficiency and essentially no crossover of sulfur compounds, including sulfur vapor, into the membrane purge gas stream. Additionally, since the SRU is a low-pressure unit, the driving force for the water removal must be based on concentration gradient across the membrane and not on pressure differential. A schematic of the experimental apparatus is shown in FIG. 3.
FIG. 3. Schematic of the experimental apparatus.
Test results are detailed in TABLE 2.
Field demonstration. To commercialize this technology, a field demonstration is required. This will be conducted in an existing Claus or sub-dewpoint SRU that will be retrofitted with the following equipment:
- Membrane unit installation at condenser outlets
- Plant air or N2 for purge medium
- Upgraded acid gas, main and trim air FCVs (round ported or v-ball)
- An additional tail gas analyzer at the first condenser outlet (and moving the existing tail gas analyzer to the outlet header, if applicable)
- New combustion air control block (fast-acting, prioritization of a tail gas analyzer, etc.).
Successful criteria will include the following:
- Water removal efficiency of > 50%
- Negligible crossover of any sulfur compounds into the purge gas
- Integrity of the membrane after long-term exposure to the Claus process stream
- Adequate control of combustion air under normal operating conditions, as well as during process upsets
- Adequate hydrolysis of COS and CS2 over the Claus converter and cycling sub-dewpoint reactors to achieve an overall recovery efficiency of > 99.9%.
Economics. The SSRIM process can achieve ultra-high Claus conversion efficiencies (recovery efficiency of > 99.9%) without the requirement of a TGTU, which makes this technology very attractive from an economic and operational simplicity standpoint. An economic evaluation2,3,4 indicated that if the traditional two-stage Claus/TGTU CAPEX is set at 100 units, the equivalent CAPEX of the SSRIM process is 70 units. Extending the same basis for OPEX, the traditional two-stage Claus/TGTU will have a lifetime OPEX of 100 units, while the SSRIM process will have a lifetime OPEX of 40 units. The result is that the SSRIM process offers a significant improvement in the net present value (net present cost) of 45% when compared to the traditional two-stage Claus/TGTU.
Takeaways. The SSRIM process is a breakthrough for the sulfur recovery industry, as it is the first technology capable of achieving an overall recovery efficiency of > 99.9% without requiring a TGTU. The economics also indicate that the SSRIM is a very attractive option compared to the traditional Claus/TGT lineup. Finally, the operation of an SSRIM process is greatly simplified when compared to a Claus/TGT process since it is based solely on the Claus reaction.
ACKNOWLEDGEMENTS
The author would like to thank Alberta Sulphur Research for conducting the SSRIM bench scale tests. Their bench scale testing capabilities, together with their knowledge in the field of sulfur recovery, remain unprecedented, and Alberta Sulphur Research continues to be a vital source of information and support to the sulfur recovery industry.
LITERATURE CITED
1 Paskall, H. G., “Capability of the modified-Claus process: A final report to the Department of Energy and Natural Resources of the Province of Alberta,” Western Research & Development, January 1979.
2 Slavens, A. F., D. K. Stevens and K. R. Didriksen, “A novel pairing of technologies to achieve today’s sulfur emissions stringency with minimum facility investment,” 2007.
3 O’Connel, J. P. and S. Verma, “Saudi Aramco simplified sulfur recovery technology,” Saudi Aramco Engineering Report SAER-9151, 2019.
4 Hydrocarbon Publishing Company, Worldwide Refinery Processing Review: Sulfur plant and aromatics production,” 2016.
Comments