Trends in Tragedy – An in-depth Study of Amine System Failures

Authors: PHILIP LE GRANGE, MIKE SHEILAN, BEN SPOONER – Amine Experts International (Netherlands, Canada, USA, Chile, New Zealand)

Equipment malfunction or unplanned shutdown of an amine system can have a devastating effect on a production company’s profitability. The goal of this paper is to determine, and focus, the industry’s attention on the highest probability threats to their facility’s operability and reliability. The threats were identified by analyzing 400 cases of major amine system failure investigated by Amine Experts and others industry leaders over the last 20 years. These include detailed root-cause analysis related to corrosion, foaming, hydraulic restrictions and incorrect specification that have limited or crippled facilities the world over. The most prevalent causes of failure in amine systems are determined and strategies to prevent these are given.

INTRODUCTION

Failure of an amine system to deliver its rated capacity of on-specification product can have an adverse effect on engineers’ bonuses. To avoid this travesty, the authors have analyzed several hundred cases from Amine Experts’ extensive field troubleshooting history, as well as cases from literature, to ascertain the most frequently occurring root causes of these failures. The intention behind this paper is to use these historical lessons from industry to focus the attention of plant personnel on the best strategies to provide the most significant risk reductions to production reliability.

To the authors’ knowledge, this is the first time that such a comprehensive database has been compiled and published. While the root causes of these failures are not new knowledge to the industry, the frequency and likelihood of specific root causes occurring relative to other common root causes has not previously been qualified to this level of detail. The database contains failures from amine systems in many different services (refining, natural gas, LNG, ammonia/urea production, steel mills, biogas and power plants) and where significant differences between the failure mechanisms were observed this is highlighted in the paper.

The paper also differentiates between sweet and sour amine systems. For the purposes of this paper, a sour amine system is defined as containing more than 1000ppmv of H2S in its feed gas. Amine systems in oil refineries, many gas plants, tail gas units and steel mill coke oven sections are thus considered sour. Amine systems used in the production of ammonia/urea and many biogas facilities are considered sweet.

FAILURE TYPES

Of the 400 cases compiled in the study, 297 come from Amine Experts’ field troubleshooting reports with an additional 103 being sourced from literature or other industry experts. Only cases where production was heavily impacted and there was an in-depth on-site analysis done by the Amine Experts consultant or paper’s author were included in the database. Anecdotes or papers in which it is not apparent that the majority of potential causes were considered were not included in the database. In the cases supplied by Amine Experts, the cost in terms of lost production ranged from USD 250 thousand to 250 million per case. It is easily possible that the lost production from the failures in this database represents several billion dollars.

The types of production failures experienced fall broadly into five categories: off specification hydrocarbon product, corrosion, foaming, flooding and excessive amine loss (excluding loss caused by foaming). Note that amine losses are often also a side effect of a foaming system, however, amine replacement costs tend to be eclipsed by the cost of production losses. For this reason, amine losses caused by foaming were not considered in the amine losses category as these cases were covered in the foaming area and double counting of cases was not considered by the authors to be desirable. Any of these problems can result in expensive losses of production. The relative frequency of their occurrence in all amine plants as well as in sweet and sour amine application is shown in Figure 1 (14 cases were not easily classed as sweet or sour and have hence been excluded from Figure 1B). Corrosion accounts for almost half of all failures seen in sweet service, figure 1B, and is responsible for a higher proportion of failures than in sour service.

From Figure 1, it is clear that the most frequent failure areas facing both sweet and sour amine treating systems are foaming, product quality and corrosion. As a result of this relative frequency, these three failure mechanisms will be the primary focus of this paper. A clear difference between the failures on sweet and sour systems is the higher relative frequency of corrosion incidents faced in sweet amine systems.

These cases predominantly occurred within the last 20 years, as a result, this database represents a view of current challenges in industry. Historically, other severe problems may have played a larger role (e.g. hydrogen stress corrosion cracking prior to introduction of post-weld heat treating as a standard in amine plants). Since the early 1980s, there has also been a shift from MEA and DEA toward MDEA-based solvents.   This shift has impacted on the frequency of certain root causes. Analysis of the frequency of failure of the different amine types will not be undertaken in this paper and may form the topic of future work. There are approximately 60 root causes that are a factor in system failures that are generally accepted in industry. The most prevalent of these in sour systems are determined and explained in detail in the following sections of this paper.

 

Fig. 1A: Cases of critical failure by type for all amine systems (sweet and sour)

 

Fig. 1B:       Cases of critical failure by type for sour systems (left) and sweet systems (right)

CORROSION

Corrosion is described in many ways, but primarily, it is ‘the destruction of a metal by the electrical or electrochemical reaction with its environment’. In amine units, the environment includes a number of serious corrosion promoters, such as acid gases (H2S and CO2), heat stable salts and their acid precursors, chelants (that can remove any protective films that may be formed), velocity and high temperature. Steel is also prone to corrosion because it is not the natural state of iron in the environment. Iron is found as iron oxide in nature, so its tendency to oxidize (corrode) is simply the iron returning to its natural state. For iron to corrode is a completely natural process.

The challenge with corrosion is that there are multiple types of corrosion that can occur and many times it is a combination of several factors, rather than a single factor, that generates the corrosive environment in the plants. The following types of corrosion are found in an amine unit:

  • General ionization corrosion
  • Galvanic corrosion
  • Crevice corrosion
  • Under-deposit corrosion
  • Pitting corrosion
  • Erosion corrosion
  • Oxygen-related corrosion
  • Stress corrosion cracking
  • Hydrogen damage
  • Acid attack
  • Ammonium Bisulfide Accumulation

Except for ionization, corrosion in amine units is always highly localized. The reason why overall corrosion is not seen is because amines are excellent corrosion inhibitors, and if left contaminant-free, provide excellent inhibition of all amine wetted surfaces. The key to troubleshooting corrosion incidences is to try to determine the logical mechanism, and then determine why that particular part of the plant failed to be protected from corrosion. In amine plants, the majority of corrosion is related to acid gas breakout and the subsequent attack of the metal surface, usually at areas of elevated temperature. Since corrosion is a chemical reaction, high temperatures will always accelerate corrosion activity because reactions occur faster and more aggressively at higher temperatures.

Historically, hydrogen embrittlement or hydrogen-induced cracking (HIC) has been a frequent cause of corrosion in the amine industry. In this process the steel becomes brittle and fractures as a result of the introduction and subsequent diffusion of atomic hydrogen into the metal. In amine service, this corrosion is manifested as blisters or cracks in the steel walls of vessels and piping. Atomic hydrogen, which is formed by a number of mechanisms in an amine system, is a small enough atom that it can diffuse through the steel wall to the environment and then re- combine with another atomic hydrogen to form the more stable molecular hydrogen gas. If, however, atomic hydrogen finds an anomaly in the steel (discontinuity or crack caused from an impurity in the manufacturing and forming process) it may stay in the crack until another atomic hydrogen enters the same location causing the unstable atoms to combine to form the stable H2 molecule. This H2 molecule is now too large to diffuse through the steel matrix so it is trapped within the steel walls of the vessels where it develops a gas pressure. Ultimately, if enough hydrogen diffuses into and forms hydrogen gas in this discontinuity, the gas pressure exceeds the tensile strength of the steel, leading to the cracking and blistering seen in Figure 2.

Fig. 2: Hydrogen-Induced Cracking (left)  and blistering (right)

Following numerous failures and a particularly tragic incident in 1984 in which an amine LPG treater at a USA refinery ruptured, causing an explosion and fire that killed 17 people [6,7] a NACE survey [10] of 294 amine units was conducted. The survey focused on refineries (272 out of 294 cases), which were predominantly operating MEA and DEA. The survey found that in 98% of the cases the cracking was linked to the use of non-Post Weld Heat Treated (PWHT) steels.

Most facilities built post-1990 use HIC resistant steels. Typically, the plate is tested to meet the requirements of TM0284, which is the NACE standard for the “evaluation of pipeline and pressure vessel steels for the resistance to hydrogen-induced cracking”. Reputable steel manufacturers have full control of the entire process, from the steel’s metallurgy to rolling, stress relieving and annealing.

There are, frequently, several factors contributing to failure by corrosion and are summarized for all systems in Figure 3A and for sweet systems in Figure 3B. Figure 4 shows the frequency of corrosion failure by location in the plant and shows that almost 50% of corrosion incidents occur in the hottest part of the plant: the reboiler and bottom of the amine regenerator.   In amine plants, most corrosion is related to H2S/CO2 breakout and the subsequent attack of the metal surfaces, usually at areas of elevated temperature and high pressure drop. Since corrosion is a chemical reaction, elevated temperatures will always accelerate corrosion activity because reactions occur faster and more aggressively with increasing temperature. The same tendency for corrosion to be predominantly located in the reboiler was also found in a survey of 80 United States Refineries conducted by a refining industry work group [8].

Fig. 3A: Root causes of the 131 corrosion cases (all amine systems)

Fig. 3B: Root causes of the 47 sweet system corrosion cases

Fig. 4: Locations of the corrosion (131 cases – all systems)

‘Sweet’ systems processing significantly more CO2 than H2S, or only CO2, face more corrosion challenges than plants with a significant amount of H2S (Figure 1B). In sweet plants, 48% of all incidents were caused by corrosion versus 29% in sour plants. The locations where corrosion occurred also varied, with sweet plants experiencing more corrosion on the lean amine coolers, pumps and piping (18% of cases versus 3% for sour systems) and absorber (13% for sweet versus 4% for sour) while there was proportionally less corrosion in the hot lower regenerator/reboiler section (39% for sweet versus 53% for sour). Corrosion of the reflux system in sweet systems is not common as it is typically a result of ammonia or cyanide trapping in the overhead system and these are not present in conventional sweet system feeds.

 H2S absorbed in the amine will react with carbon steel vessels and piping to form various types of iron sulphide (FexSy), depending on the location in the amine plant. FexSy can either help prevent corrosion by creating a stable film that ‘passivates’ the metal surface and prevents further corrosion (pyrrhotite), induce further corrosion (pyrite), or provide limited passivation and contribute to fouling (mackinawite) [4], Table 1. Stoichiometric iron sulphide (FeS; troilite) is rare outside of meteorites; if detected in an amine system, it is a sign of much bigger problems!

The presence and type of FexSy layer is likely a major factor in the corrosion differences observed between sweet and sour systems. Detrimental pyrite forms in the lower regenerator and reboiler, while beneficial pyrrhotite forms in the absorber of sour systems.

Under-stripping is a situation where a large amount of the H2S/CO2 in the rich amine is stripped in the reboiler rather than in the regenerator column. Excessive amine stripping in the reboiler will often result in pitting-type corrosion (potentially initiated by the bursting of some of the large number of gas bubbles [15]) in the reboiler and lower regenerator. This is avoided by maintaining adequate heat medium flow to the reboiler. Industry guidelines vary between advising that 95 to 99% of the H2S/CO2 present in the rich amine be stripped in the regenerator rather than in the reboiler, or that the vapour return from the reboiler should contain less than 1 mol% CO2.

Under-stripping is the most common root cause of corrosion in amine systems and is frequently self-inflicted through not addressing fouling in reboilers and lean/rich exchangers, overly aggressive energy-saving campaigns that have not understood system limitations, or the mistaken idea that leaving H2S in the lean amine will create a more stable and protective film. The last is especially fatal, as CO2 is more difficult to strip than H2S, so intentionally leaving a small amount of H2S in an amine results in far higher quantities of CO2 not being stripped. It is also counter productive as the type of FexSy that does form in the lean sections is not beneficial.

Fear of ‘over-stripping’ amine and causing corrosion is prevalent in the industry, but unfounded. There is no published data fundamentally supporting the theory of over-stripping corrosion. There have been several instances of plants with low H2S in the lean amine experiencing corrosion and mistakenly assuming that this implies causation. Looking at each of these cases in more detail revealed systems with an elevated HSAS content. These salts increase the stripping of the amine by a phenomenon known as acid-assisted regeneration; they also corrode the hot lean sections of the plant. In this instance, the solution is to limit the HSAS level in the amine, not reduce stripping. Further work on FexSy passivation layers is needed to fully quantify their impacts and stability in real systems.

Figure 5A shows examples of reboilers that have been corroded due to excessive acid gas break out in the reboiler. In both cases, a large proportion of the stripping was occurring in the reboiler rather than the regenerator.

Figure 5B shows corrosion on the lean amine cooler and lean amine pump discharge line on a sweet amine system. This corrosion was the result of CO2 breaking out of the amine solution due to excessive lean amine loadings. High solvent loadings (especially in the high temperature lean amine) allow for small amounts of acid gas to escape from the solution as it undergoes pressure drop moving through the piping and auxiliary equipment. This phenomenon is described in detail by Sheilan and Smith [5]. These acid gas then form an area of localised high concentration that corrodes the metal in a pit like pattern (visually this resembles bites being taken out of the metal).

Fig. 5A: Reboilers Corroded by Excessive Acid Gas Break Out

Fig. 5B: CO2 attack on lean amine aerial cooler (left) and lean amine pump discharge (right) in a sweet system due to excessive lean loadings

Another major factor in preventing corrosion in amine systems is amine system hygiene. Good hygiene effectively requires limitation of other potentially corrosive contaminants. These contaminants include: solids, heat stable amine salts, degradation products and amino acids (including bicine).

Solids in the amine system can contribute to corrosion via erosion of the metal or passivation layers protecting it. Total suspended solids should ideally be kept below 10 ppmw based on use of a 0.45µm absolute filter to test the solution.   Note that actual solids levels in the system will differ from the test as a result of increased solubility of many solid precipitates with agitation and temperature. Adequate mechanical filtration and maintenance of the filters is indispensable in maintaining solution quality [3].

Heat Stable Amine Salts (HSAS) form via reaction of the amine with a contaminant acid that typically enters the system with the feedstock. The resulting acid-amine salt that forms is not readily regenerable at normal amine system stripping conditions and is hence given the name ‘heat stable’. While the acid precursors that enter the unit may be present only at trace levels the HSAS that forms will continue to build up in the system until they are removed.

Above 3 wt% HSAS in the amine the corrosiveness of the amine rapidly increases. Figures 6 and 7 show a section from the bottom part of a corroded regenerator which was replaced as a result of HSAS based corrosion, salts had been allowed to build up in the system to approximately 5 wt%. This is a standard industry guideline that is consistent with the information in the database in sour applications.

Heat stable salts can be removed from an amine system by: Amine replacement, Solid Ion exchange beds, Electrodialysis, Vacuum distillation or Thermal reclamation (traditionally only MEA and DGA™). HSAS may also be neutralized via the addition of strong base species (usually sodium or potassium containing) to form Heat Stable Salts (HSS).   There are, however, limits to the amount of HSS that a system may contain before HSSs precipitate out of the solution and plug parts of the unit. Neutralization also has implications for the HSAS removal methods listed above and is best applied intelligently as part of a heat stable salt management strategy that takes into account the system size, geographic location and HSAS build-up rate. (Note that reclamation of amine to below 0.5wt% can adversely affect the removal of acid species via eliminating the beneficial effects of acid assisted regeneration, most sites consider 0.5wt% HSAS a good end-point for reclamation)

Degradation Products will build up within the amine system over time as a result of thermal exposure and the reaction of amine with CO2 or Oxygen. Some of these products may be corrosive and they should be monitored regularly and kept within vendor recommended limits.   Limiting reboiler tube temperatures to less than 165°C (330°F) will largely eliminate thermal degradation in an amine system. Degradation products are normally removed via installed thermal reclaimers in MEA and DGA® systems. For DEA, DIPA and MDEA, vacuum distillation or inventory replacement are the established options. MDEA is extremely resistant to chemical degradation and degradation rates in DIPA are lower than MEA, DEA or DGA™.

Fig. 6: Severe Acid Attack in Regenerator Bottom Section, Opposite Reboiler Return Line

Fig. 7: Regenerator Wall Thinned to Breaking Point by Erosion and High HSASs

Amino Acids develop in amine systems because of the ingress of oxidizing components e.g. O2, SO2 and SX into an amine system causing oxidative degradation of the amine followed by subsequent reaction to amino acids. Bicine (one of several amino acids found in amine systems) has been linked to corrosion in several amine systems [11,12,13]. Bicine functions as a chelant, increasing the solubility of iron in amine by a factor of just under 500 [13]. More recently, several other amino acids have also been linked to corrosion in amine systems [13,14]; with hydroxyethyl sarcosine (HES) being thought to be the most frequently occurring [14].

While several industry limits have been proposed for bicine, review of the data does not reveal any consistency in the level at which bicine will corrode a system, Figure 8. This is likely due to the role of other amino acids (e.g. HES) that have historically not been analyzed for in amine systems.

Amino acids are zwitterions, neutral molecules with both positive and negative charge, and can thus be removed or partially removed by ion exchange and electrodialysis techniques. Neutralisation is not considered an effective method for amino acid removal. Amine replacement or vacuum distillation will also remove amino acids.

Fig. 8: Bicine level at which corrosion was reported

PRODUCT OFF-SPECIFICATION

The treated product leaving an amine absorber must meet certain specifications that depend on its destination, whether gas, LPG, or NGL. Failure to meet these specifications often results in valuable product being flared (figure 9). The most common specifications operators are concerned with are: CO2 (or the overall energy value of the gas), H2S and total sulphur (COS, mercaptans, etc.)

Fig. 9: Off-specification gas being flared to atmosphere

There are many interrelated causes for an amine system not to work to specification, as summarized in Table 2. The two most common reasons for gas to be off-specification are high lean amine loading and insufficient heating or cooling, or in other words, improper amine temperature (see Figure 10).

Fig. 10: Number of instances of a root cause putting an amine system off specification

Table 2

Root Causes of Amine Plant Product Not Meeting Specification

Root Cause Explanation
Insufficient contact There is not enough contact area in the columns for the solvent and product / steam to mix sufficiently for adequate treatment / stripping.
Incorrect formulation The solvent (or mix of solvents) chosen has the wrong / suboptimal composition for the application.
Solvent Contamination The presence of elevated levels of high viscosity degradation products or glycols from upstream will make treating more difficult via increasing mass transfer resistance in the amine solvent. Bulk hydrocarbon contamination is also a problem as the hydrocarbons phase is an additional mass transfer barrier to treating and can damage downstream units treating the acid gas.
Phase envelope A condensed hydrocarbon layer floating on top of the amine will add an additional mass transfer resistance to the absorption of H2S / CO2 from a gas
Mechanical damage Mechanical damage to the unit has reduced treating or stripping efficiency. Similar to insufficient contact but a result of damage rather than design.
Insufficient heating/cooling There is a heating or cooling equipment constraint in the system that is adversely affecting treating.
Under-circulating Circulating too little amine: Not enough amine present to react with H2S / CO2
Over-circulating Circulating more amine than needed resulting in a self-inflicted heat transfer limitation that increases lean amine temperatures and loadings, putting the product off-specification.
Understrength The amine strength is not high enough: Not enough capacity to remove H2S / CO2. Sufficient fresh amine needs to be added periodically to prevent this.
High lean loading The regenerated amine still has high levels of H2S / CO2 in it, limiting the level to which H2S / CO2 can be removed.
 

 

The lean loading is the amount of H2S or CO2 remaining in the (lean) amine after regeneration. This is important because the partial pressure of H2S in the product is in equilibrium with that of the H2S in the lean amine when it leaves the top of the absorber. This means the more H2S in the lean amine entering the top of the tower, by default, the more H2S will remain in the treated product. It can also be important in CO2 systems with a tight CO2 specification, e.g. LNG production facilities.

In cases where a high lean loading is the root cause of off-specification product, there is typically insufficient heat medium being sent to the reboiler, insufficient contact with steam in the regenerator column, or contamination of the amine with a strong base. Strong base contamination is often the result of over-vigorous addition of sodium or potassium ions to neutralize acidic components that build up in the system over time. Monitoring and control of the lean loading in an amine system is fundamental to meeting most specifications.

Amine temperature control is also crucial to meeting product specification. Undersized, pinched, or fouled exchangers will not be able to supply sufficient duty to adequately heat or cool the amine. For example, insufficient heat duty will decrease the degree of amine stripping in the regenerator.   This will, in turn, result in high lean loadings and consequently off-specification product. Contrarily, insufficient cooling duty will raise the lean amine temperature and will shift the equilibrium in the absorber to favour CO2 removal over that of H2S. Usually, the H2S specification is much lower and more strictly controlled than that of CO2. Higher pressure absorbers and better regenerated amines are more resistant to temperature effects.

Furthermore, the temperature profile within the absorber itself can also determine the rate at which the amine will remove H2S / CO2. Excessive temperature “bulges” in the column render large portions of the tower useless for removing H2S / CO2 from the process feed since the amine loses its capacity for absorption. Such a situation is shown in Figure 11, where two new LNG trains were unable to process more than 65% of their design capacity due to excessive temperature bulges in the absorbers.

Fig. 11: Simulated (left) and measured surface temperature (right) absorber temperatures

FOAMING

Foaming occurs when gas is incorporated mechanically into the amine liquid phase, resulting in a froth in which gas bubbles are surrounded by a liquid film. The formation and stability of the foam tends to be more strongly dependent on surface characteristics of the liquid phase than physical conditions like temperature and pressure [1].

A foaming amine plant is characterized by high and erratic differential pressure measurements in the absorber and/or regenerator columns [2]. It may also display fluctuating levels and erratic flow valve positions and generally results in reduced treating efficiency, amine carryover to downstream systems and loss of production capacity.

The common causes of amine plant foaming have been divided into 12 categories (see Table 3 for an explanation of each). Of the 108 recorded cases of amine system foaming in amine treating plants in the database, 56 of these were found to have foam promoting contaminants entering the plant with the feed, see Figure 12. From this finding it seems clear that proper conditioning of the feed gas is a key focus area for lowering the foaming risk in amine plants, especially in amine systems in gas treating service.

There are numerous potential inlet contaminants in the feed gas (Table 4) that can promote or stabilize foaming. In a gas plant these include: compressor lubrication oils, brine water from downhole, drilling and pipeline chemical additives, soap sticks, iron sulphides / oxides, silica and sand. In refineries surfactants include: sponge oils, catalyst fines, thermal cracking demulsifiers, upstream corrosion products and polar hydrocarbons. It is critical to remove these before they enter the amine system where they can accumulate and induce foaming incidents. To do this, a combination of 3-phase separators, knock out drums, particle filters and coalescing filters is typically used. For certain feed streams with unconventional contaminants, water washes and silica gel beds may also be necessary. For a plant with a high feed ingress rate of foam causing contaminants an inlet 3-phase separator followed by a particle filter followed by a coalescing filter (see Figure 13) represents an optimal approach to resolving this issue.

Interesting to note is that particulates circulating in the amine solution itself are rarely a root cause of foaming in and of themselves, rather they tend to exacerbate foaming once it occurs by making the foam that does form more stable and harder to break. Small suspended particulates will become incorporated into a foam bubbles structure, increase its mechanical strength, and making it more difficult for the foam to break.

Fig. 12A: Number of instances of a root cause causing foaming in all amine systems

Fig. 12B: Number of instances of a root cause causing foaming in oil refinery amine systems

Fig. 12C: Number of instances of a root cause causing foaming in gas plant amine systems

Table 3

Root Causes of Amine Plant Product Not Meeting Specification

Explanation  
Many contaminants will be stripped from the solvent in the regenerator only to condense in the reflux cooler and be trapped in the system. This leads to a build-up of surfactants in the regenerator reflux section. Purging this section will prevent build-up of contaminants and limit regenerator foaming.  
Design deficiencies that allow or contribute to foam formation such as: flow loops with no purge (e.g. purging absorber product knockout vessel liquids back to the feed drum), absent or incorrectly designed filters and separators in the system and at the inlet.  
Heat Stable Amine Salts (HSAS) are formed by the reaction of amine with acids stronger than H2S or CO2. These enter or form in the system in a variety of ways, and their bond with the amine molecule will not break under amine regenerator conditions. Some organic acid-based HSAS have surfactant properties that may induce foaming. These acids will build up over time and need to be routinely monitored. An appropriate strategy to deal with them when their level rises should be employed.  
Gas streams should not be sent to an absorber at their hydrocarbon dewpoint temperature. Gas at its dew-point should ideally be preheated. Failing to do this could cause hydrocarbons to condense in the absorber, and eventually promote foaming.  
This flow regime occurs in columns when there are high gas and low liquid flow rates and results in an agitated, dispersed liquid phase that can easily foam. This is not a healthy region to operate a column, as liquid entrainment significantly increases. There is also risk of flooding and vibrational damage to the column.  
Particles that enter the system or are generated via corrosion can stabilise a foam that has formed. Solid particles can become incorporated in the bubbles surface and lend mechanical strength, causing the foam not to break. The amine should be adequately filtered to prevent this.  
Use of acid-activated carbon, the incorrect antifoam (or one that is past its shelf life    or    being    heavily    overdosed),    filter    cartridges    manufactured    with

inappropriate binder or containing trace lubricants from manufacture, are some common examples of incompatibilities that induce foaming

 
Surfactants and hydrocarbons should be skimmed off the top of the amine in the flash drum, and trace levels will be adsorbed in carbon beds.   If these are not working, surfactants will inevitably build up in the system  
If the amine is cooler than the hydrocarbon dewpoint of the gas being treated, condensation of some of the gas is possible and may promote foaming. Solvent temperature should be 5 to 10°C higher than the gas feed (exceptions exist)  
Contamination of the amine with surfactants during manufacture or transportation, contaminated chemical truck transfer hoses, not flushing detergents from a system prior to start up, or using poor quality make-up water can all lead to foaming incidents in the columns.  
Benzene, Toluene, Ethylbenzene & Xylene (BTEX) are present in the gas phase in some process feeds. These are soluble in the amine solvent and once in solution affect surface tension, thereby inducing foaming  
Solid particles, slugs of liquid and aerosols entering with a process feed could induce foaming.  
Many contaminants will be stripped from the solvent in the regenerator only to condense in the reflux cooler and be trapped in the system. This leads to a build-up of surfactants in the regenerator reflux section. Purging this section will prevent build-up of contaminants and limit regenerator foaming.  

 

 
Table 4

Potential Inlet Contaminants in Feed Gas

Solid Liquid
Scale Compressor lube oils
Desiccant fines Hydrocarbon condensates
Corrosion products Amines
Dirt (silica, sand etc.) Glycols
Iron sulphides Pipeline chemicals
Iron oxides Completion fluids
Adamantane/Diamantane Condensed water
Catalyst fines Well brines
  Organic acids
  Thermal cracking demulsifiers
  Sponge Oil

Fig. 13: Inlet separation configuration for an amine plant with extreme inlet contamination

 It is critical that inlet separation equipment is attentively monitored and well maintained. A good review of these systems and how to operate them correctly is given in the paper “The Seven Deadly Sins of Filtration & Separation Systems in Gas Processing Operations” by Sheilan and Engel [3]. Specific areas to be cognizant of:

  1. Differential pressure instruments should be in good working order and operating limits not exceeded to prevent particle filter and coalescing cartridge blow
  2. Lubricants containing surfactants should not be used in element
  3. Cartridges that are present in particle filters and coalescers should be installed correctly and securely to prevent bypassing; filter cartridge elements are only as good as the sealing surface between the element and the vessel, so care must be taken during the cleaning and installation stages to ensure very close seals (double O-ring gaskets are preferred in gas phase applications as they provide the best seal).
  4. Level control instruments and valves should be maintained in good working order for equipment to function
  5. It is important to select a filter cartridge where all of the components in the cartridge are compatible with amine
  6. Differential pressure over a filter should rise slowly with time, zero differential pressure for long periods of time is not indicative of healthy filter operation and should be investigated.

In gas plant service the majority of foaming incidents occurred in the absorber or throughout the whole system. There were very few cases of ‘regenerator only’ foaming recorded in the natural gas industry. This is a result of the large role that inlet contamination plays in foaming incidents as the first area to come into contact with surfactants in an inlet contamination incident is the amine absorber.

In refineries regenerators have often suffered foaming incidents with no associated absorber foaming. In these instances, low level refinery contaminants enter in the absorber in quantities too low to cause foaming but become trapped in the overhead of the regenerator where they are stripped and condensed in a continual cycle until the contaminant level builds up to the point where foaming occurs. It has also been proposed that certain cracked LPG feedstocks contain contaminants that can react at stripper conditions to form surfactants and thus induce foaming, while this may be occurring there is, as yet, no published evidence to support this.

Once the amine has been contaminated, foaming will continue in either the absorber or regenerator (or both) until the contaminant has been removed or diluted. In these cases, the flash tank, reflux purges and proper operation / maintenance of the particle filter and carbon bed are critical.

CONCLUSION

Off specification product, foaming and corrosion are the three biggest challenges in reliably operating amine units and have cost the industry billions of dollars over the last 30 years.

For H2S removal the chemical equilibrium limitation is the central factor in meeting specification.   Product not meeting its H2S specification is most commonly caused by high lean amine loadings, predominantly due to insufficient heat in the system by design or operation.

Foaming problems in gas plants have been found to be due predominantly to inlet contamination of the feed with impurities that have surface tension affecting properties. Adequate inlet separation equipment and proper maintenance of existing equipment is critical.

Corrosion was most commonly found to occur in the reboiler and lower part of the regenerator connected to it. The dominant causes of corrosion in sour plants were found to be excessive stripping in the reboiler due to insufficient regeneration in the regenerator column and poor amine hygiene.

Review of the catastrophic failures of 400 amine plants showed that approximately half of the failures could have been eliminated if the following had been in place/performed:

  • Proper regeneration of the solvent such that lean amine loading is low and the bulk of the regeneration occurs in the regenerator
  • Adequate and well-maintained inlet separation equipment
  • Regular solvent quality monitoring and having a strategy to maintain HSAS, degradation products, amino acids and solids within acceptable limits.

Based on these findings it is advised that plant operators recite the following mantra daily:

“HEAT IN, FILTERS ON, SALTS OUT”

ACKNOWLEDGEMENTS

The authors would like to acknowledge the Amine Experts consultants (past and present), whose field work and experience provided the bulk of the cases for this database. We would also like to thank the authors of the external cases who have published their experiences with sufficient care and detail that others might learn from them.

References

 General:

  1. Sheilan H., Spooner B.H, van Hoorn E.; “Amine Treating and Sour Water Stripping, 10th Ed.”, Amine Experts 10th Ed., 2015
  2. Kohl, L., Riesenfeld, F.C.; “Gas Purification”, 3rd. Edition, Gulf Publishing Company, 1979.
  3. Sheilan H., Engel D.; “The Seven Deadly Sins of Filtration & Separation Systems in Gas Processing Operations”, Gas Processors Association Annual European Chapter Conference, 2015.
  4. Sheilan, , van Hoorn E., Spooner B.; “Iron Sulphide: Friend or Foe”; 57th Annual Laurance Reid Gas Conditioning Conference; 2007
  5. Sheilan, and Smith, R.F.; “Hydraulic-flow effect on amine plant corrosion”, Alberta Oil & Gas Journal 82(47), 1985.
  6. McHenry, H. I., Shives, T. R., Read, D. T., McColskey, J. D., Brady, C. H., and Purtscher, P. T.; “Examination of a Pressure Vessel that Ruptured at the Chicago Refinery of the Union Oil Company on July 23, 1984”; National Bureau of Standards Report NBSIR 86-3049, March
  7. McHenry, H. I., Read, D. T., and Shives, T. R.; “Failure analysis of an amine-absorber pressure vessel”; Materials Performance, 18-24; August 1987.
  8. Critchfield, E., and Jenkins, J.L., “Evidence of MDEA degradation in tail gas treating plants”, Petroleum Technology Quarterly, Spring 1999, p87-95.
  9. Bosen, F, and Bedell, S.A., “The Relevance of Bicine in the Corrosion of Amine Gas Treating Plants”, NACE paper No. 04481, NACE Corrosion 2004.
  10. Cummings L., Waite S.W., Nelsen D.K.; “Corrosion and Corrosion Enhancers in Amine Systems”; Brimstone Conference, Alberta, April 2005
  11. Closmann F., Thompsen J., Schuette G.; “Advances in the Identification of Amino Acids and other products in degraded amine”; 63rd Annual Laurance Reid Gas Conditioning Conference; 2013
  12. Jepson P.; “The effect of flow characteristics on sweet corrosion in high-pressure, three-phase, oil/water/gas horizontal pipelines”, Prevention of Pipeline Corrosion Conference, Texas, October 1994

Prior industry surveys/reviews:

  1. Kennedy B., Scott B., Tunnel D., Wagner E., Zacher M.; “The cost of poor amine operations”; Gas Technology Magazine,
  2. Kittel , Bonis M., Perdu G.; “Mitigating corrosion in sweet gas units: a comparison between lab data and field survey”; 64th Annual Laurance Reid Gas Conditioning Conference; 2014
  3. Richert, J. P., Bagdasarian, A. J., and Shargay, C. A.; “Extent of stress corrosion cracking in amine plants revealed by survey,” Oil and Gas , June 5, 1989, pg. 45-52.

External case sources:

Weiland R.H., Hatcher N.A.; “Advanced Gas Treating: The Engineering Science, 2nd Ed.”, Optimized Gas Treating Inc., 2012

Kister Z.H.; “Distillation Troubleshooting”; John Wiley & Sons Inc.; 2005

Addington F., Hendrix D.E; “Aggressive Corrosion of 316 Stainless Steel in an Amine Unit: Causes and Cures”; Corrosion 2000 paper 00698 – NACE; 2000

Jones C.E., Hatcher N.A., Weiland R.H.; ”Design Pitfalls”; LNG Industry; 2014

Howard M., Sargent A.; “Operating Experiences at Duke Energy Field Services Wilcox Plant with Oxygen Contamination and Amine Degradation”; 51st Laurance Reid Gas Conditioning Conference; 2001

Epps R., Wimberly M.O.; “UCARSOL Solvents –State of the Art Amine Technology”; 40th Annual Laurance Reid Gas Conditioning Conference; 1990

DuPart M.S, Bacon T.R., Edwards D.J.; “Understanding and Preventing Corrosion in Alkanolamine Gas Treating Plants”; 41st Annual Laurance Reid Gas Conditioning Conference; 1991

Masson W.B.; ”Westcoast McMahon Plant Gas Treating Experiences”; 44th Annual Laurance Reid Gas Conditioning Conference; 1994

Lyengar J.N., Siba P.W., Clarke D.S., ”Operations and Recovery Improvement Via Heavy Hydrocarbon Extraction”, 48th Annual Laurance Reid Gas Conditioning Conference; 1998

Howard I.; “Hannibal’s Experiences”, 48th Annual Laurance Reid Gas Conditioning Conference; 1998

Mykitta R.S.; “Operating Experiences of Shell’s Yellowhammer Gas Plant”; 49th Annual Laurance Reid Gas Conditioning Conference; 1999

Barnes D.R.; “Reduction of Heat Stable Salt Formation in a Monoethanolamine (MEA) CO2 Removal System”; 49th Annual Laurance Reid Gas Conditioning Conference; 1999

Huffmaster M.A.; “Sulfinol-M Conversion Experience at Hythe Brainard Plant”; 50th Annual Laurance Reid Gas Conditioning Conference; 2000

Miller T.E, Roesler K, Holub P.E., McCaffrey C., Covington K.; “Unique Acid Gas Enrichment Application”; 51st Annual Laurance Reid Gas Conditioning Conference; 2001

Chakraborty A., Bagde A.; “Case Studies on Gas Sweetening Process”; 52nd Annual Laurance Reid Gas Conditioning Conference; 2002

Sargent P.E, Seagraves J.; “LPG Contactor Design and Practical Troubleshooting Techniques”; 53rd Annual Laurance Reid Gas Conditioning Conference; 2003

Crochet B., Bills R., Eguren R.; “A useful Amine Management Tool to Track Important Operating Parameters”; 54th Annual Laurance Reid Gas Conditioning Conference; 2004

Pearson H., Shao J., Norton D., Dandekar S.; “Case Study of Effects of Bicine in CO2 Only Amine Treater Service”; 55th Annual Laurance Reid Gas Conditioning Conference; 2005

Jordan T.J., Nozal P.J., Azodi A; “Handling Trace Oxygen at the Saunders Gas Processing Facility”; 56th Annual Laurance Reid Gas Conditioning Conference; 2006

Hakim N.J., Benmoulay A., Oehlschlaeger F.; “QATARGAS DIPA Losses Minimization Approach”; 57th Annual Laurance Reid Gas Conditioning Conference; 2007

Thomas J.C., Bradley A.; “Operational Experience with Accelerated MDEA Solvents in CO2 Removal Applications”; 58th Annual Laurance Reid Gas Conditioning Conference; 2008

Holub P.E., Richards R., Ralph W., Sweatt G., White L.; “New Design for High Pressure Reclamation of DGA® Agent Solutions”; 60th Annual Laurance Reid Gas Conditioning Conference; 2010

Friloux B., Gall S., Luzardo., Eguren R.; “How to Operate and Clean a Dirty Amine System at the BP Crane Plant”; 60th Annual Laurance Reid Gas Conditioning Conference; 2010

Ness C., Frey C., Holub, P.E., White L.R., Griffie S., Halfast C.; “Plant case Study with THEED Degradation using blended DEA solvents”; 62th Annual Laurance Reid Gas Conditioning Conference; 2012

Bickham C.; “Tips for Troubleshooting Trayed Absorbers”; 63th Annual Laurance Reid Gas Conditioning Conference; 2013

Pack B., Shackelford A.; “More than Meeting Spec: Examination of Recent Process Saftey Incidents in Gas Conditioning Systems should be Cause for Pause”; 63th Annual Laurance Reid Gas Conditioning Conference; 2013

Patio J.A., Sanabria J.C., Rozo J.M., Cruz U.I.; “The Challenge of Troubleshooting the Amine Plant in Cusiana Field (Colombia)”; 63th Annual Laurance Reid Gas Conditioning Conference; 2013

Bickham C., Teletzke E., Smith P.; “Rein in your Unruly Regenerator”; 64th Annual Laurance Reid Gas Conditioning Conference; 2014

Daughtry J., Teletzke E.; “Controlling Corrosion: Case Studies from Amine Plant Operation”; 66th Annual Laurance Reid Gas Conditioning Conference; 2016

Worley Parsons “Sulphur Plant Operations & Lessons Learnt Workshop”; Sour Oil & Gas Advanced Technology (SOGAT) Conference 2014

The Dow Chemical Company, “Gas Conditioning Fact Book”, 1962.

Williams A.F.; Lom W.L.; “Liquefied Petroleum Gas”; Ellis Horwood Limited, Sussex England, 1974

Keller A.; “The Aftermath of SO2 Breakthrough and Ways to Prevent and Mitigate It”; Brimstone Sulfur Symposium, Colorado, 2012

Safruddin S., Safruddin R.; “Twenty years’ experience in controlling corrosion in amine unit, Badnak LNG Plant”; NACE Corrosion 2000 – Paper 00497.

Bulger J., Girgis M., Polvi T.; “Corrosion due to process instability in hot lean amine system”; NACE Corrosion 2005 – Paper 05386

Moore M.A., Qarni M.M., Lobley G.R.; “Corrosion Problems in Gas Treat Systems”; NACE Corrosion 2008 – Paper 08419

Optimised Gas Treating Inc.; “The Contactor – The case of the corroding regenerator”; Vol 10. Issue 6, 2016.

Raut, N., Chaudhari, R.M., Naik, V.S., “Failure of an amine regenerator column of amine treatment unit”, NACE paper No. 09334, NACE Corrosion 2009.

Litschewski, M.J., “More experiences with corrosion and fouling in a refinery amine system”, NACE paper No. 96391, NACE Corrosion 1996.

Russel, M.T., Wortham, G.M., Lawson, D.M., “Experiences with combined corrosion effects on stainless steel due to chlorides and H2S”, NACE paper No. 97340, NACE Corrosion 1997.

de Hart, T.R., Hansen, D.A., Mariz, C.L., McCollough, J.G., “Solving Corrosion Problems at the NEA Bellingham

Massachusetts Carbon Dioxide Recovery Plant”, NACE paper No. 99264, NACE Corrosion 1999.

Wilson, P., “Corrosion of Carb on Steel in CO2 capture plant using MDEA and Amino Acid based solvents”, NACE paper No. C2012 – 0001554, NACE Corrosion 2012.

Saithala., J.R., IIlson, T., Hachicha, M., Abdollah, S., “Corrosion related tube failures in amine reboiler units”, NACE paper No. C2013 – 0002368, NACE Corrosion 2013.

Ammonia Know How