March 2016

Special Report: Corrosion Control

Improved corrosion prevention with acid-aided regeneration technology

Enhancing regeneration in amine treating systems has proven beneficial in many applications by improving operations and relieving design limitations via reduced steamrates and/or improved treating performance.

A commonly applied process for the removal of acid gas contaminants from gas is amine treating. At its heart lie the absorption of acid gases into an amine solution and the regeneration of this solvent to be fed back to the absorption column. The interplay between these two steps is important, as the treating performance can be majorly influenced by the amount of acid gas dissolved in the lean solvent returning from the regenerator.

In some cases, achieving the treated gas specification using standard reboiler duties in the regenerator column is difficult, as the loading of the solvent entering the absorber may not be lean enough. To obtain a leaner solvent, the regeneration can be enhanced by the implementation of proprietary tail gas treating technology that removes sulfur compounds downstream of the Claus unit. In other situations, this technology offers the ability to optimize amine processes by achieving a certain solvent leanness with lower regeneration steam requirements.

The proprietary technology relies on applying a controlled concentration of certain acidic additives that enhance regeneration. Elements of the technology were originally developed more than 40 years ago; the technology has evolved in application over the past decades to incorporate corrosion-avoidance and improved-operability strategies.

The mature technology is particularly useful in existing plants, as it can often help meet tighter specifications without requiring hardware changes, thereby offering valuable benefits in a world in which specifications and emissions limits change over time, and the pressure to reduce capital expenditures (CAPEX) and operating expenditures (OPEX) is ever increasing.

Here, the application of the technology in actual operation is examined, and trends and observations are drawn from operating experience.


In absorption, the acid gas reacts with the amine in an equilibrium reaction to form ions; for example, H2S is removed by reaction with a tertiary amine, as shown in Eq. 1:

H2S(g) «» H2S(aq) + R3N «» HS + R3NH+      (1)

This reaction is then reversed in the regenerator, stripping the acid gas out of the solvent. In the bottom of the column, this reverse reaction is favored by a lower pH, and, therefore, by the addition of acidic regeneration-enhancing agents—for example, a strong mineral acid, as shown in Eq. 2:

HX + R3N «» X + R3NH+     (2)

The R3NH+ concentration increases, forcing the equilibrium of the reaction in Eq. 1 to the left and leading to a greater release of H2S—and, therefore, a leaner solvent leaving the regenerator.


The technology company’s experience with adding regeneration-enhancing additives to amine treating systems is mainly based on its tail gas treating units (TGTUs). These units remove H2S, using a selective amine solution to recycle it back to the sulfur recovery unit, thereby minimizing SO2 emissions from the plant. In the process, deep solvent stripping is applied to remain comfortably within environmental limits.

Good plant data describing plant performance and solvent analysis are required to successfully apply acid-aided regeneration. A comparison of the experience at several different amine units clearly showed the effect of additive presence, and the observations from these units and case studies are presented here. Case studies 1 through 6 address plants removing H2S (with a focus on TGTUs, but also discussing high-pressure applications). Case Study 7 reviews the technology company’s experience of applying regeneration enhancement to CO2-only applications.

Case Study 1: Lean H2S loading

This case study concentrates on improving the controllability of H2S lean loading in multiple TGTUs. The unit’s regenerator performance was optimized at various locations to decrease SO2 emissions.

Regeneration enhancement was applied in the multiple TGTUs, which resulted in improved H2S performance of the absorbers. Structured performance demonstrations were also conducted; the concentration of regeneration-enhancing additive was manipulated in controlled tests. H2S performance in the absorber (and, therefore, SO2 performance of the overall tail gas treating) was improved in each system where the technology was implemented. From this experience, an overall relationship was established between the leanness of the solvent and the H2S performance of the TGTU absorber. However, the operability of the overall TGTU system was also considered in the demonstrations. It was discovered that when the additive concentration increased beyond a certain maximum value, regenerator performance became increasingly sensitive to variation in applied steamrate.

This increased sensitivity to energy input is demonstrated conceptually in Fig. 1, which shows trends observed from the collated plant test data from the TGTUs. Fig. 1 illustrates the overall relationship between the amount of regeneration-enhancing additive, energy input and resulting H2S lean loading. It was observed that, with the same energy input, the solvent becomes leaner when a higher concentration of additive is employed; and, at the same leanness, less steam is required with higher additive concentration.

  Fig. 1. Summary of TGTU test data results.

Case Study 2: Absorber performance

Enhancing regeneration in an amine unit can benefit treating performance only if the absorber operates at a close approach to lean solvent loading. In practice, this means that the absorber must have enough stages to allow for deep treating of H2S. An additional requirement is that improvement in regenerator performance must be greater than the loss in the absorber top.

If these requirements are not met, then performance in the absorber can actually worsen when acidic additives are present. This is because the presence of the additives also results in a higher H2S vapor pressure, which disfavors absorption. The strength of this impact depends on several interacting factors, such as the relative amounts of H2S, acidic additive and CO2 in solution, along with the design and operating conditions.

In TGTU absorbers, the condition of the top trays is most important for control of absorber performance, and, in this region, coabsorption of CO2 plays a large role. CO2 interferes in H2S removal, thereby decreasing the effect of enhanced regeneration. If substantial amounts of CO2 are coabsorbed, then the approach to H2S equilibrium over the lean solvent worsens, minimizing the deep H2S treating benefit of the technology. This effect is illustrated in Fig. 2 from simulation, which is compared against known data;1, 2 partial pressure of H2S increases with increasing CO2 coabsorption.

  Fig. 2. Effect of CO2 coabsorption on H2S absorption.

Case Study 3: TGTU with near-flooding conditions

After a major capital change in a TGTU, the amine regenerator began to experience near-flooding conditions, which posed difficulty in achieving deep H2S stripping. In this particular situation, the operators found that changing the steamrate had only a minor effect on the depth of H2S stripping. When the ambient temperature was high, this unit experienced challenges meeting environmental targets.

A project to address the limitations concluded that the best remedy was to formulate the methyldiethanolamine (MDEA) solvent with the appropriate dosing of proprietary additive. This action succeeded in reducing the regeneration steam by 40% to achieve the desired depth of stripping while the specification of the treated gas remained unchanged, which also lowered the regenerator differential pressure and reduced the risk of flooding (Fig. 3).

  Fig. 3. Effect of formulation on differential
  pressure, steamrate and H2S performance.

No change was observed in the decomposition rate of the solvent, and no noticeable accumulation of soluble materials was found. Also, following formulation, the operators now observe a clear relationship between the steamrate and the resulting H2S concentration in the treated gas, improving operability of the unit.3

Case Study 4: Energy savings

Steam savings possible for an example case are illustrated in Fig. 4. Based on the previously outlined conclusion to target additive content for controllability, the bulk of the possible steam savings can still be realized, as the steam savings curve flattens at higher acid concentrations. In this example case, 40% of the steam can be saved while maintaining improved operability and preserving a minimum concentration of H2S in the lean solvent.

  Fig. 4. Steam ratio savings.

Case Study 5: Application in high-pressure selective treating

Although most of the technology company’s experience is with TGTUs, it has also applied enhanced regeneration in other applications, including selective H2S removal in high-pressure natural gas. As with the tail gas treating technology, that application requires deeply regenerated lean solvent to meet the H2S specification in the treated gas.

This effect is illustrated in Case 5, in which H2S performance of the system was optimized by manipulating the extent of regeneration enhancement. Fig. 5 shows the result of changing the amount of the additive: low H2S lean loadings that occur at high additive concentrations on an example high-pressure integrated system. This plant was dependent on regeneration enhancement to reach targeted H2S concentration in the lean solvent. Test runs demonstrated that higher levels of enhancement resulted in too-deep regeneration—deeper than the company’s technical governance for the application.

  Fig. 5. Relationship between H2S lean
  loading and additive concentration in a
  high-pressure application.

Case Study 6: Inferring corrosion effects

It is well known that amine systems processing H2S tend to form a layer of iron sulfide on exposed carbon steel (CS) surfaces. This layer is thought to help protect CS surfaces against certain types of corrosion. Past publications have introduced the concept that maintaining a minimum level of H2S in the lean solvent facilitates preserving this iron sulfide layer.4, 5

This operating philosophy was applied to the plant in Case 5. The depth of regeneration was maintained on target by controlling the amount of regeneration enhancement. Monitoring of corrosion coupons, filter changes and solvent-quality tests demonstrated a substantial decrease in corrosion in that location.

Building on that success, a systematic review of tail gas treating corrosion experience was conducted in the company’s US downstream applications to document CS corrosion in key locations within the amine units. This review was coupled with solvent quality monitoring in the locations.

The study demonstrated that plants maintaining H2S in the solvent were less likely to detect iron in solvent quality analysis. Testing for iron in the solvent samples can give an indication of possible corrosion. Although corrosion may occur without the solubilization of iron, if iron is found in the solvent samples in H2S removal plants, this can be seen as a warning sign of conditions that may lead to corrosion.

Plant data shows that iron content correlates with high acid concentration, as well as low concentration of suppressive H2S in the sample. Fig. 6 shows the relationship between iron content, depth of stripping and additive concentration in the population of TGTUs. A simple relationship is observed: plants that do not strip too deeply are less likely to find iron in solvent samples.

  Fig. 6. Iron prevalence in MDEA-based TGTUs.

Case Study 7: Regeneration enhancement in CO2-removal systems

The technology company has also reviewed the effect of regeneration-enhancing additives with the objective of reducing the regeneration heat requirements, both in addition and in comparison to other options to reduce the energy footprint of the unit at a specific plant. Although the CO2 specifications are somewhat more relaxed, the actual performance examines a similar deep removal compared to the usual H2S specification partial pressures.

Plant test data in a secondary amine solvent system was taken and is presented in Fig. 7, which predicts that, in conditions similar to TGTUs, steam savings exist for CO2, but are much lower than for H2S and follow a much flatter curve. With the lower steam savings, adding acids looks less interesting for CO2 systems—particularly since CO2 is generally easier to strip out than H2S, and since CO2 specifications are often less severe and do not require deeper stripping of CO2.

  Fig. 7. H2S and CO2 lean loading in DIPA
  before and after adding additives.

Note that a secondary amine forms carbamates in the presence of CO2, a relatively more stable component, which is more difficult to regenerate and makes it difficult to strip to very low CO2. Fig. 7 illustrates the effect of the additives. The lines indicate trends derived from plant data, while the performance with additives is shown in data points from the tests. The data indicate that systems using a secondary amine such as DEA or DIPA can also benefit from the addition of acids, but less data are available than for the tertiary MDEA solvents.

Observations have also been made for an MDEA-based solvent and are presented in Fig. 8. In this example, a gas treating unit processing high-pressure gas for deep CO2 removal was analyzed, based on plant performance vs. plant leanness analysis, in a structured way. The data suggests that the presence of acids in the solvent significantly impacts CO2 lean loading to some extent. This data presents another example of the effect of acid content in high-pressure absorption, although it is not as pronounced as in the case of H2S, for the reasons outlined above.

  Fig. 8. Acid impact on CO2 solvent leanness
  and steamrate on an example unit.


Enhancing regeneration in amine treating systems has proven beneficial in different applications by improving operations and relieving design limitations through reduced steamrates and/or improved treating performance.

However, care must also be taken, since an improper dosing of acid can lead to corrosion risk, worse treating performance and reduced controllability of the unit. From experience in operating with acidic additives and controlled plant tests, several observations were reviewed to understand how to avoid corrosion and improve treating results while applying regeneration enhancement. HP


The tail gas treating technology referenced in this article is Shell Claus Offgas Treating (SCOT) technology, and the proprietary units referenced are Shell SCOT units.


1 Huang, S. H. and H. J. Ng, “Solubility of H2S and CO2 in alkanolamines,” GPA Research Report RR-155, September 1998.
2 Bullin, J. A., R. R. Davison and W. J. Rogers, “The collection of VLE data for acid gas—alkanolamine systems using Fourier transform infrared spectroscopy,” GPA Research Report RR-165, March 1997.
3 Bonner, S. and J. Critchfield, “Relieving stripper flooding at Martinez SCOT 3,” presented at Brimstone Sulfur Symposium 2009.
4 Van Roij, J., J. Klinkenbijl, P. Nellen and K. Sourisseau, “Materials threats in aging amine units,” Paper 2207 presented at NACE Corrosion 2013.
5 API Recommended Practice 945, “Avoiding environmental cracking in amine units,” April 2008.

The Authors

Related Articles

From the Archive



{{ error }}
{{ comment.comment.Name }} • {{ comment.timeAgo }}
{{ comment.comment.Text }}