------------------Fundamentals of Oil and Gas Processing Book
Basics of Gas Field Processing Book
Prediction and Inhibition of Gas Hydrates Book
Basics of Corrosion in Oil and Gas Industry Book
https://www.amazon.com/Yasser-Kassem/e/ ... scns_share
6.6 General Considerations for Solvent Process
6.6.1 Solution Filtration
Filtration of the treating solution to remove entrained solids is essential to the successful operation of a gas treating plant.
Filtration rates should be as high as practical and may range from 5 per cent of circulation to full stream. Removing particles down to 5 microns in size is recommended. In order to do this efficiently, two stages of filtration may be required. The first stage, typically a cartridge-type or precoat filter, is designed to remove particles down to the 10 micron or less range.
The second stage of filtration, typically an activated carbon filter, removes hydrocarbons and other contaminants. This is accomplished by adsorption. The carbon filter can also remove smaller particles from the amine stream. The carbon granule size can be selected to remove particles down to the 5 micron range. The activated carbon filter should always be located downstream of the first stage filter because the deposition of solids would plug the carbon filter.
The carryover of carbon fines can be controlled by either locating a second cartridge-type filter immediately downstream of the carbon filter or using a graded carbon bed. In a graded bed, larger granules are placed at the outlet of the filter to trap fines. Large carbon granules produce fewer fines but are less efficient for adsorption.
Basic degradation products are identified by gas chromatography and mass spectrometry. Acidic degradation products are identified by ion chromatography exclusion. These tests are recommended when the amine solution appears to lose its ability to pick up acid gas. Degradation products affect the results of the conventional estimation of amine concentration by titration. This may cause artificially high or low apparent amine concentrations Also, the carbon bed will adsorb very little strong acid degradation products. In this case, purging or reclamation of the solution is recommended.
Carbon filters can be partially regenerated with steam, which removes hydrocarbons and other adsorbed contaminants. Regeneration or bed change out is recommended when foam tests on the inlet and outlet streams show no improvement. This indicates carbon bed saturation.
Filters may be located on the lean or rich solution side. Filtration on the rich side seeks to remove particles that are more insoluble under the rich solution conditions. It also prevents solids accumulation in the hot environment of the stripper.
However, proper design and operating procedures for personnel protection during filter maintenance is mandatory when H2S may be present.
Filtration equipment should be used continuously beginning with the first day of plant operations. When starting up the plant, the full flow filter, even if temporary, may prove its worth by removing the scale and other solid particles and allowing much quicker and easier start-up of the plant.
6.6.2 Flash Tank
Rich solution leaving the contactor may pass through a flash tank. A flash tank is more important when treating high pressure gas. Gases entrained in the rich solution will be separated.
In addition, the amount of absorbed gas will be decreased because of the lower operating pressure of the flash tank. Using a flash tank will:
Reduce erosion in rich/lean exchangers.
Minimize the hydrocarbon content in the acid gas.
Reduce the vapor load on the stripper.
Possibly allow the off-gas from the flash tank to be used as fuel (may require sweetening).
When heavy hydrocarbons are present in the natural gas, the flash tank can also be used to skim off the heavy hydrocarbons that were absorbed by the solution. Residence times for flash tanks in amine service vary from 3-10 minutes depending on separation requirements. Inlet gas streams containing only methane and ethane require shorter residence times. Rich gas streams require longer time for the dissociation of gas from solution and/or the separation of liquid phases.
Corrosion is an operating concern in nearly all sweetening installations. The combination of H2S and CO2 with water practically ensures that corrosive conditions will exist in portions of the plant. In general, gas streams with high H2S to CO2 ratios are less corrosive than those having low H2S to CO2 ratios. H2S concentrations in the ppmv range with CO2 concentrations of 2 percent or more tend to be particularly corrosive.
Because the corrosion in sweetening plants tends to be chemical in nature, it is strongly a function of temperature and liquid velocity. The type of sweetening solution being used and the concentration of that solution has a strong impact on the corrosion rate. Increased corrosion can be expected with stronger solutions and higher gas loadings.
Hydrogen sulfide dissociates in water to form a weak acid. The acid attacks iron and forms insoluble iron sulfide. The iron sulfide will adhere to the base metal and may provide some protection from further corrosion, but it can be eroded away easily, exposing fresh metal for further attack. High liquid velocities can erode the protective iron sulfide film with resulting high corrosion rates.
CO2 in the presence of free water will form carbonic acid. The carbonic acid will attack iron to form a soluble iron bicarbonate which, upon heating, will release CO2 and an insoluble iron carbonate or hydrolize to iron oxide.
In general, design velocities in rich solution piping should be 50% of those that would be used in sweet service. Because of the temperature relationship to corrosion, the reboiler, the rich side of the amine-amine exchanger, tend to experience high corrosion rates. Because of the low pH the stripper overhead condensing loop also tends to experience high corrosion rates.
Acid degradation products also contribute to corrosion. A suggested mechanism for corrosion is that degradation products act as chelating agents for iron when hot. When cooled, the iron chelates become unstable, releasing the iron to form iron sulfide in the presence of H2S. Primary amines are thought to be more corrosive than secondary amines because the degradation products of the primary amines act as stronger chelating agents.
Several forms of stress corrosion cracking are possible in amine sweetening systems. Amine stress corrosion cracking can occur and is worst in hot solutions, but cracking can occur in cooler lines and both rich and lean streams. Wet sulfide cracking and blistering can occur due to hydrogen generated in corrosion reactions. The hydrogen can collect at small inclusions in the steel which delaminate and then link in a step-wise pattern to create blisters. This is called HIC or hydrogen induced cracking. Sometimes stress influences the cracking to cause SOHIC or stress oriented hydrogen induced cracking. HIC resistant steels are available. Seamless pipe is less prone to HIC than plate steels.
Corrosion in alkaline salt processes, such as the hot carbonate process, has been reported to range from none to severe. Corrosion can be expected where CO2 and steam are released through flashing. Severe erosion can take place when carbonate solution strengths exceed 40% because of the tendency to form bicarbonate crystals when the solution cools.
Many corrosion problems may be solved using corrosion inhibitors in combination with operating practices which reduce corrosion.
Following are some guidelines to minimize corrosion.
Maintain the lowest possible reboiler temperature.
If available, use low temperature heat medium rather than a high temperature heat medium or direct firing. When a high temperature heat medium or direct firing for the reboiler is used, caution should be taken to add only enough heat for stripping the solution.
Minimize solids and degradation products in the system through reclaimer operation and effective filtration.
Keep oxygen out of the system by providing a gas blanket on all storage tanks and maintain a positive pressure on the suction of all pumps.
Ensure deionized water or oxygen/chemical-free boiler condensate is used for make up water. If available, steam can be used to replace water loss.
Limit solution strengths to minimum levels required for treating.
Pipe solution exchangers for upflow operation with the rich solution on the tube side.
Monitor corrosion rates with coupons or suitable corrosion probes.
Maintain adequate solution level above reboiler tube bundles and fire tubes; a minimum tube submergence of 12" is recommended.
Corrosion inhibitors used include high molecular weight amines and heavy metal salts. These inhibitors offer potential savings in both capital and operating costs for these special cases.
A sudden increase in differential pressure across a contactor or a sudden liquid level variation at the bottom of the contactor often indicates severe foaming. When foaming occurs, there is poor contact between the gas and the chemical solution. The result is reduced treating capacity and sweetening efficiency, possibly to the point that outlet specification cannot be met.
Some reasons for foaming are:
Soap-based valve greases
Makeup water impurities
Foaming problems can usually be traced to plant operational problems. Contaminants from upstream operations can be minimized through adequate inlet separation. Condensation of hydrocarbons in the contactor can usually be avoided by maintaining the lean solution temperature at least 10°F above the hydrocarbon dew point temperature of the outlet gas.
Temporary upsets can be controlled by the addition of antifoam chemicals. These antifoams are usually of the silicone or long-chain alcohol type.
The following test for foaming should be run with the various types of inhibitors being considered for a given application. This test should give the operator an indication of which antifoam will be the most effective for the particular case. Place several drops of antifoam in 200 ml of treating solution contained in a 1000 ml cylinder. Bubble oil-free air through the solution at a constant rate. After five minutes have elapsed shut off the air and start a timer. Note the height of foam at the time the air was shut off and the amount of time required for the foam to break. The foam height is the difference between the height of the foam and the initial height of the liquid. The time for the foam to break is an indication of the stability of the foam. A comparison of antifoams will let the operator select which inhibitor will best solve his foaming problems. Between antifoam tests, care should be taken to clean the test cylinder thoroughly because a very small amount of inhibitor may affect the test.
Treating plants normally use carbon steel as the principal material of construction. Vessels and piping should be stress relieved in order to minimize stress corrosion along weld seams. Corrosion allowance for equipment ranges from 1/16" to 1/4", typically 1/8". In some instances, when corrosion is known to be a problem, or high solution loadings are required, stainless steel or clad stainless steel may be used in the following critical areas:
1. Reflux condenser
2. Reboiler tube bundle
3. Rich/lean exchanger tubes
4. Bubbling area of the contactor and/or stripper trays.
5. Rich solution piping from the rich/lean exchanger to the stripper.
6. Bottom 5 trays of the contactor and top 5 trays of stripper, if not all.
Usually 304, 316, or 410 stainless steel will be used in these areas, even through corrosion has been experienced with 410 stainless in DEA service for CO2 removal in the absence of H2S.
L grades are recommended if the alloys are to be welded.
Controlling oxygen content to less than 0.2 ppmw is effective in preventing chloride SCC in waters with up to 1000 ppmw chloride content, at temperatures up to 570°F. There has been an increased use of duplex stainless steels, and they have been successfully used in the water treatment industry to prevent chloride SCC in high chloride waters. This suggests duplex stainless steels could be utilized in amine plant service where high chloride content is expected. As with any specialty steel, proper fabrication techniques and welding procedures are required.
6.7 Solid Bed Processes
6.7.1 General Process Description
A fixed bed of solid particles can be used to remove acid gases either through chemical reactions or through ionic bonding.
This process flows the gas stream through a fixed bed of solid particles, which removes the acid gases and holds them in the bed. When the bed is spent, the vessel must be removed from service and the bed regenerated or replaced. Since the bed must be removed from service to be regenerated, some spare capacity is normally provided.
Commonly used processes under this category are as follows:
Acid gas removal through chemical reactions include:
Iron Sponge Process, SulfaTreat®, Zinc Oxide Process, Chemsweet®, & PuraSpec®
Acid gas removal through Ionic bonding “Adsorption” include:
Molecular Sieve Process, & impregnated activated carbon.
6.7.2 Iron Sponge Process
The iron sponge process is economically applied to gases containing small amounts of H2S (<300 ppm) operating at low to moderate pressures in the range of 25-500 psig. This process does not remove CO2.
The reaction of iron oxide and H2S produces iron sulfide and water as follows:
Fe2O3 +3H2S → Fe2S3 +3H2O Eq. 6-12
FeO+H2S → FeS+H2O Eq. 6-13
The reaction requires the presence of slightly alkaline water (pH 8-10) and a temperature below 110 0F (47 0C). If the gas does not contain sufficient water vapor, water may need to be injected into the inlet gas stream. The pH level can be maintained through the injection of caustic soda, soda ash, lime, or ammonia with the water.
Although the presence of free alkalines enhances H2S removal, it also creates potential safety hazards and promotes the formation of undesirable salts, adding to capital costs.
Ferric oxide is impregnated on wood chips,which produce a solid bed with a large ferric oxide surface area. Several grades of treated wood chips are available, based on iron oxide content. Ferric oxide wood chips are available in 6.5, 9.0,15.0, and 20 lbs iron oxide/bushel.Chips are contained in a vessel, and sour gas flows downward through the bed and reacts with the ferric oxide. Figure 6-11 shows a vertical vessel used in the iron sponge process.
The bed can be regenerated with air; however, only about 60% of the previous bed life can be expected. The bed life of the batch process is dependent upon the quantity of H2S, the amount of iron oxide in the bed, residence time, pH, moisture content, and temperature.
Ferric sulfide can be oxidized with air to produce sulfur and regenerate the ferric oxide. Regeneration must be performed with care because the reaction with oxygen is exothermic (i.e., gives off heat). Air must be introduced slowly so that the heat of reaction can be dissipated.
If air is introduced quickly, the heat of reaction may ignite the bed.
For this reason, spent wood chips should be kept moist when removed from the vessel. Otherwise, the reaction with oxygen in the air may ignite the chips and cause them to smolder.
The reactions for oxygen regeneration are as follows:
2Fe2S3 + 3O2 + 2H2O → 2Fe2O3.(H2O) + 6S+Heat Eq. 6-14
4FeS +3O2 + 2xH2O → 2Fe2O3(H2O)x +4S +Heat Eq. 6-15
S2 +2O2 → 2SO2 Eq. 6-16
Some of the elemental sulfur produced in the regeneration step remains in the bed. After several cycles, this sulfur will cake over the ferric oxide, decreasing the reactivity of the bed and causing excessive gas pressure drop.
Typically after 10 cycles, the bed must be removed from the vessel and replaced with a new bed.
It is possible to operate an iron sponge with continuous regeneration by the introduction of small amounts of air in the sour gas feed. The oxygen in the air regenerates the iron sulfide and produces elemental sulfur. Although continuous regeneration decreases the amount of operating labor, it is not as effective as batch regeneration, and it may create an explosive mixture of air and natural gas. Due to the added costs associated with an air compressor, continuous regeneration generally does not prove to be the economic choice for the typically small quantities of gas involved.
Cooler operating temperatures of the natural gas, for example, during the winter, create the potential for hydrate formation in the iron sponge bed.
Hydrates can cause high-pressure drop, bed compaction, and flow channeling.
When the potential for hydrates exists, methanol can be injected to inhibit their formation. If insufficient water is present to absorb the methanol, it may coat the bed, forming undesirable salts. Hydrocarbon liquids in the gas tend to accumulate on the iron sponge media, thus inhibiting the reactions.
The use of a gas scrubber upstream of the iron sponge at a gas temperature slightly less than that of the sponge media may prevent significant quantities of liquids from condensing and fouling the bed.
There has been a recent revival in the use of iron sponges to sweeten light hydrocarbon liquids. Sour liquids flow through the bed and are contacted with the iron sponge media and the reaction proceeds as above.
Fig. 6-11 Iron oxide acid-gas treating unit.
The Sulfa-Treat process similar to iron sponge process. It is economically applied to gases containing small amounts of H2S. This process utilizes a proprietary iron oxide co-product mixed with inert powder to form a porous bed. Sour gas flows through the bed and reacts with iron as in iron-sponge reaction.
The material of SulfaTreat is a dry, free-flowing granular substance used for selective removal of H2S and mercaptans from natural gas in the presence of CO2. It is not affected by CO2, and it does not produce sulfur or nitric oxide (NOx). Also SulfaTreat will not ignite or "cement up" in the vessel. Other advantages include longer bed life and lower cost.
SulfaTreat’s particle size varies from 4 to 30 mesh and has a bulk density of 70 lb/ft3. These physical properties give uniform porosity and permeability, which offers small resistance to flow and resistance to bed compression at normal velocities.
Applications for SulfaTreat include: natural gas treating, amine treater off gas, high concentration CO2 streams, and any other H2S-containing system.
The reaction works better with saturated gas and at elevated temperature up to 1300F (54.40C). There is no minimum moisture or pH level required. The amount of bed volume required increases as the velocity increases and as the bed height decreases. Operation of the system below 400F (4.40C) is not recommended. Beds are not regenerated and must be replaced when the bed is spent.
6.7.4 Zinc Oxide Process
The equipment used in the zinc oxide process is similar to the iron sponge process. It uses a solid bed of granular zinc oxide to react with the H2S to form zinc sulfide and water as shown below
ZnO+H2S → ZnS+H2O Eq. 6-17
The rate of reaction is controlled by the diffusion process, as the sulfide ion must first diffuse to the surface of the zinc oxide to react. Temperatures above 2500F (1200C) increase the diffusion rate, which promotes the reaction rate. The strong dependence on diffusion means that other variables, such as pressure and gas velocity, have little effect on the reaction.
The zinc oxide is contained in long thin beds to lessen the chances of channeling. The pressure drop through the beds is low. Bed life is a function of gas H2S content and can vary from 6 months to more than 10 years. Beds are often used in series to increase the level of saturation prior to change out of the catalyst. A spent bed is discharged by gravity flow through the bottom of the vessel.
The zinc oxide process is seldom used due to disposal problems with the spent bed, which is classified as a heavy metal salt.
Chemsweet® is the name for another batch process for the removal of H2S from natural gas. Chemicals used are a mixture of zinc oxide, zinc acetate, water, and a dispersant to keep solid particles in suspension. Natural gas is bubbled through the solution where H2S reacts with zinc oxide. Though several reactions take place in solution, the net result is that zinc oxide reacts with H2S to form zinc sulfide and water.
Chemsweet® can treat gas streams with H2S concentration up to 400 ppmv. and has been operated between pressures of 75 and 1,400 psig. Mercaptan concentrations in excess of 10% of the H2S concentration in the gas stream can cause a problem. Some of the mercaptans will react with the zinc oxide and be removed from the gas. The resulting zinc mercaptides [Zn(OH)RH] will form a sludge and possibly cause foaming problems.
Johnson Matthey Catalysts supplies the PuraSpec range of processes and products for desulfurization of hydrocarbon gases and liquids. The processes use fixed beds of granular, metal oxide-based chemical absorbents which are developments of the ‘high temperature zinc oxide’ used for purification of hydrocarbon feedstocks to steam reformers in ammonia, hydrogen, and methanol plants. PuraSpec absorbents are effective at temperatures down to 32°F, so no added heat is necessary, and are in service at pressures from atmospheric, treating vent gases, to 1800 psi treating dense phase gas feed to a gas processing plant.
PuraSpec units are in service treating natural gas to pipeline or petrochemical specifications. Because the absorbents remove H2S and COS irreversibly, they are best suited to polishing duties.
6.7.8 Molecular Sieve Process
Acid gases, as well as water, can be effectively removed by physical adsorption on synthetic zeolites. Applications of acid gas removal using molecular sieve are limited because water displaces acid gases on the adsorbent bed. (Chapter 5 provides more details on adsorption and its use in dehydration.)
Crystalline structure of the solids provides a very porous material having uniform pore size. Within the pores the crystalline structure creates a large number of localized polar charges called active sites. Polar gas molecules, such as H2S and water vapor, which enter the pores, form weak ionic bonds at the active sites. Nonpolar molecules, such as paraffin hydrocarbons, will not bond to the active sites.
Carbon dioxide molecules are about the same size as H2S molecules, but are nonpolar. CO2 will enter the pores but will not bond to the active sites. Small quantities of CO2 will be removed by becoming trapped in the pores by bonded H2S or H2O molecules blocking the pores.
CO2 will obstruct the access of H2S and H2O to the active sites, thus decreasing the overall effectiveness of the molecular sieve. Beds must be sized to remove all H2O and provide for interference from other molecules in order to remove all H2S.
The adsorption process usually occurs at moderate pressure. Ionic bonds tend to achieve an optimum performance near 450 psig but can operate in a wide range of pressures.
Hydrogen sulfide can be selectively removed to meet 0.25 grain/100 scf (4ppm) specification. However, this reduction requires regeneration of the bed at 600°F (315°C) for extended time (Usually one hour or more depending upon process conditions) with the potential for COS formation if 4A is used (Molecular sieves are classified by their nominal pore diameter in Angstroms. See Chapter 5 for details.).
The problem of COS formation during processing according to the reaction in Equation 6-18, has been extensively studied.
H2S + CO2 ↔ COS + H2O Eq. 6-18
Molecular sieve products have been developed that do not catalyze COS formation. The central zone in the regeneration cycle is most favorable to COS formation.
Regeneration of a molecular sieve bed concentrates the H2S into a small regeneration stream which must be treated or disposed of.
Figure 6-12 shows a typical flow diagram for removal of H2S from natural gas. The configuration is similar to that for dehydration but with the significant difference that the regeneration gas contains high quantities of H2S as well as water as it leaves the adsorbent bed and, thus, must be treated. (Chapter 5 presents a thorough discussion of adsorption.) The flow configuration shows the first bed in the adsorption cycle, the second bed cooling down after regeneration, and the third bed undergoing regeneration with hot gas.
Fig. 6-12. Integrated Natural Gas Desulfurization Plant
Chi and Lee (1973) studied the coadsorption of H2S, CO2, and H2O on a 5A molecular sieve from a natural gas mixture under a variety of conditions.
Figure 6-13, from their paper, shows a typical concentration versus time curve.
In the figure, y is the concentration in the exit stream and yo is the concentration in the inlet to the bed. The gas that entered the bed was saturated with H2O and contained both CO2 (1.14 mol%) and H2S (0.073 mol%.). Because the CO2 content of the gas was 15.6 times that of the H2S, the bed quickly saturated with CO2, and its breakthrough was almost instantaneous. As the H2S was adsorbed and moved down the column, it displaced the CO2 and, consequently, after approximately 30 minutes, the CO2 exit concentration peaked at a value greater than its inlet concentration.
The same phenomenon occurs when the H2S is displaced by the water.
Because the H2S must be removed to extremely low levels, the bed is effectively exhausted from the perspective of H2S purification shortly after H2S breakthrough occurs and, thus, well before the bed is totally saturated.
Fig. 6.13 Effluent H2S and CO2 concentration from adsorption bed as a function of Time.
In general, the sieve bed can be designed to dehydrate and sweeten simultaneously.
The Engineering Data Book (GPSA) notes that a key point in adsorber design is to properly design for treatment of the regeneration gas because the peak H2S concentration may be 30 times the H2S concentration in the feed.
Care should be taken to minimize mechanical damage to the solid crystals as this will decrease the bed’s effectiveness. The main cause of mechanical degradation is the sudden pressure and/or temperature changes that may occur when switching from adsorption to regeneration cycles. Proper instrumentation can significantly extend bed life.
The molecular sieve process is limited to small gas streams operating at moderate pressures. It is generally used for polishing applications following one of the other processes.
Oxorbon is an alternative solid bed material which consists of activated carbon impregnated with potassium iodide (KI). Donau Carbon markets such a carbon for the removal of H2S and mercaptans. The adsorbed H2S is converted to elemental sulfur by catalytic reaction under the presence of oxygen. The resulting sulfur is fixed on the pores of the activated carbon. Sulfur loadings as high as 60% of the carbon mass have been reported and sulfide concentrations below 1 ppmv are claimed.
6.8 Direct Conversion Processes (Liquid Redox)
Chemical and physical solvent processes remove acid gas from the natural gas stream but release H2S and CO2 when the solvent is regenerated. The release of H2S to the atmosphere is limited by environmental regulations. Acid gases could be routed to an incinerator/flare, which would convert the H2S to SO2. Environmental regulations restrict the amount of SO2 vented or flared. Direct conversion processes use chemical reactions to oxidize H2S and produce elemental sulfur. These processes are generally based either on the reaction of H2S and O2 or H2S and SO2. Both reactions yield water and elemental sulfur. These processes are licensed and involve specialized catalysts and/or solvents.
The redox agent is then regenerated by reaction with air in an oxidizer vessel.
Liquid oxidation−reduction (redox) processes use iron as an oxidizing agent in solution. The process involves four basic steps:
1. Removal of H2S from the gas by absorption into a caustic solution
2. Oxidation of the HS− ion to elemental sulfur via the oxidizing agent
3. Separation and removal of sulfur from the solution
4. Regeneration (i.e., oxidation) of the oxidizing agent by use of air
The processes share the general chemistry shown below.
• H2S absorption using an alkaline solution:
H2S (g) ↔ H2S (sol’n) Eq 6-19
H2S (sol’n) ↔ H+ + HS− Eq 6-20
Note that this is a non-selective, basic scrubbing solution; thus, additional acidic components (CO2 , HCN) of the sour gas will be soluble in the scrubbing solution. While CO2 is partially absorbed in the alkaline solution, it does not participate in the redox reactions. CO2 may however increase the rate of consumption of caustic or buffering compounds. The fate of other reduced sulfur species (COS, CS2, RSH, RSR) depends on the particular process being considered.
• Conversion to Elemental Sulfur:
HS− + H+ + 1/2 O2 (sol’n) → H2O (l) + S Eq 6-21
Note that in the absence of an auxiliary redox reagent (ARR) the reaction shown as Eq 21-24 is slow and nonspecific. The addition of an (ARR) increases the rate of reaction and directs the oxidation to elemental sulfur.
HS− + ARR(OX) → S + H+ + ARR (RED) Eq 6-22
where (OX) denotes the oxidized form and (RED) denotes the reduced form of the ARR.
• Regeneration of the Spent ARR Using Air:
O2 (g) ↔ O2 (sol’n) Eq 6-23
ARR (RED) + 1/2 O2 (sol’n) + 2 H+ → ARR(OX) + H2O(l) Eq 6-24
The overall simplified chemistry of liquid redox processes is thus:
H2S(g) + 1/2 O2 (g) → H2O(l) + S Eq 6-25
While the oxidizing agent forms the desired reaction, it also reacts with the sulfur species to form metal sulfides that precipitate from solution. To avoid the metal sulfide reaction, the solution contains chelating agents (organic compounds that bind with the metal ion to restrict its reactivity but still permit electron transfer for oxidation−reduction reactions). A common chelating agent for iron is EDTA (ethylenediaminetetraacetic acid). Several processes, including Lo-Cat II®, SulFerox ®, and Sulfint-HP, use iron and chelating agents.
6.8.1 Stretford Process
The Stretford process involves the use of vanadium salts as the ARR. The process has been extensively used in Europe. However environmental concerns around the discharge of vanadium compounds has limited its use.
Figure 6.14 shows a simplified diagram of the Stretford process. The gas stream is washed with an aqueous solution of sodium carbonate, sodium vanadate, and anthraquinone disulfonic. An oxidized solution is delivered from the pumping tank to the top of the absorber tower where it contacts the gas stream in a counter-current flow.
The bottom of the absorber tower consists of a reaction tank from which the reduced solution passes to the solution flash drum, which is situated above the oxidizer. The reduced solution passes from here into the base of the oxidizer vessel. Hydrocarbon gases, which have been dissolved in the solution at the plant pressure, are released from the top of the flash drum.
Air is blown into the oxidizer, and the main body of the solution, now reoxidized, passes into the pumping tank. The sulfur is carried to the top of the oxidizer by froth created by the aeration of the solution and passes into the thickener.
The function of the thickener is to increase the weight percent of sulfur that is pumped to one of the alternate sulfur recovery methods of filtration, filtration and autoclaves, centrifugation or centrifugation with heating.
Chemical reactions involved are:
H2S+Na2CO3 → NaHS+NaHCO3 Eq 6-26
Sodium carbonate provides the alkaline solution for initial adsorption of H2S and the formation of hydrosulfide (HS). The hydrosulfide is reduced in a reaction with sodium meta vanadate to precipitate sulfur
HS- +V+5 → S +V+4 Eq 6-27
Anthraquinone disulfonic acid (ADA) reacts with 4-valent vanadium and converts it back to 5 valent
V+4 +ADA → V+5 +ADA (reduced) Eq 6-28
Oxygen from the air converts the reduced ADA back to the oxidized state as shown below:
Reduced ADA+O2 → ADA+H2O Eq 6-29
The overall reaction is
2H2S+O2 → 2H2O+2S Eq 6-30
6.8.2 Lo-Cat process
This is a liquid-redox process. The LO-CAT® process is a patented, wet scrubbing, liquid redox system that uses a chelated iron solution to convert H2S to innocuous, elemental sulfur. It does not use any toxic chemicals and does not produce any hazardous waste byproducts. In general, the LO-CAT process can directly treat a gas stream, or treat the H2S containing stream from an Acid Gas Removal unit, although the direct treat capability is limited to low pressure streams.
Processes employ high iron concentration reduction-oxidation technology for the selective removal of H2S (not reactive to CO2) to <4 ppm in both low and high pressure gas streams. Figure 6-15 shows a simplified flow schematic of the LO-CAT process.
Acid gas stream is contacted with the solution where H2S reacts with, and reduces the chelated iron and produces elemental sulfur. The iron is then regenerated by bubbling air through the solution. Heat is not required for regeneration. The reactions involved are exothermic (give off heat):
Absorption/reduction: 2Fe3+ +H2S → 2Fe2+ +S+2H+ Eq 6-31
Regeneration/oxidation: 2Fe2+ +½ O2 +2H+ → 2Fe3+ +H2O Eq 6-32
Overall chemistry: H2S + ½ O2 → S + H2O Eq 6-33
Fig 6.14. Simplified flow schematic of the Stretford process.
Fig. 6.15. Simplified flow schematic of the LO-CAT process.
6.8.3 Sulferox process
SulFerox is a Shell proprietary Iron Redox process, whereby a sour gas stream, containing hydrogen sulphide, is contacted with a liquid, containing soluble ferric (Fe3+) ions. In the process the H2S is oxidized to elemental sulphur and the Fe3+ is reduced to ferrous (Fe2+) ions. The system is regenerable, Fe2+ is subsequently reconverted to Fe3+ by oxidation with air. Sulphur is recovered from the aqueous solution as a moist cake.
The process offers low capital and operating costs through the use of a high concentration iron chelate solution and effective control of chelate degradation. The process also offers a patented contactor design to improve the overall efficiency of the process which further reduces capital costs. The optimum application for SulFerox is in the one to twenty tons per day range of recovered sulfur.
Typical Sulferox process equipment is illustrated in Figure 6.16. In this process, the sour gas is contacted in a small contactor in co-current flow with a water solution containing about 4% ferric iron ions, held in solution by proprietary chelate ligand. The H2S ionizes in the solution, and the ferric ions exchange electrons with the sulfur ions to form ferrous ions and elemental sulfur. The gas and the solution leave the contactor and are flowed into a separator. The gas out of the separator is sweet and requires further treating for dewpoint control. The solution is flowed into additional vessels for separating the elemental sulfur and for restoring the ferrous iron ions back to the active ferric iron state by contacting the solution with air.
In general, the Sulferox process can directly treat a gas stream, or treat the H2S containing stream from an Acid Gas Removal unit, although the direct treat capability is limited to low pressure streams.
Fig. 6.16. Simplified flow schematic of the sulferox process.
6.8.4 IFP Process
The IFP process was developed by the Institut Francais du Petrole. The process reacts H2S with SO2 to produce water and sulfur.
The overall reaction is H2S+SO2 → H2O+2S
Figure 6.17 shows a simplified diagram of the IFP process. The process involves mixing the H2S and SO2 gases and then contacting them with a liquid catalyst in a packed tower. Elemental sulfur is recovered in the bottom of the tower. A portion of this must be burned to produce the SO2 required to remove the H2S.
The most important variable is the ratio of H2S to SO2 in the feed. The ratio is controlled by analyzer equipment to maintain the system performance.
Fig. 6.17 .Simplified flow schematic of the IFP process.
6.9 Distillation Process
Distillation processes uses cryogenic distillation to remove acid gases from a gas stream. The process is applied to remove CO2 for LPG separation or where it is desired to produce CO2 at high pressure for reservoir injection or other use.
The process consists of two, three, or four fractionating columns. The gas stream is first dehydrated and then cooled with refrigeration and/or pressure reduction.
6.9.1 Three-Column System
The three-column system is used for gas streams containing <50% CO2. The first column operates at 450-650 psig and separates a high-quality methane product in the overhead. Temperatures in the overhead are from 0 to -140 0F (-18 to -95 0C). The second column operates at a slightly lower pressure and produces a CO2 stream overhead, which contains small amounts of H2S and methane. The bottom product contains H2S and the ethane plus components. The third column produces NGL liquids, which are recycled back to the first two columns.
NGL liquids recycled prevent CO2 solid formation in the first column and aid in the breaking of the ethane/CO2 azeotrope in the second column to permit high ethane recoveries.
6.9.2 Four-Column System
The four-column system is used where CO2 feed concentration exceeds 50%. The initial column in this scheme is a de-ethanizer. The overhead product, a CO2/methane binary, is sent to a bulk CO2 removal column and de-methanizer combination. CO2 is produced as a liquid and is pumped to injection or sales pressure.
6.9.3 Two-Column System
The two-column system is used when a methane product is not required and is thus produced with the CO2. Very high propane recoveries may be achieved; however, little ethane recovery is achieved. These processes require feed gas preparation in the form of compression and dehydration, which adds to their cost. Such systems are finding applications in enhanced oil recovery (EOR) projects.
6.10 Sulfur Recovery (The Claus Process)
Sulfur is present in natural gas principally as hydrogen sulfide (H2S) and, in other fossil fuels, as sulfur-containing compounds which are converted to hydrogen sulfide during processing. The H2S, together with some or all of any carbon dioxide (CO2) present, is removed from the natural gas or refinery gas by means of one of the gas treating processes pre-described. The resulting H2S-containing acid gas stream is flared, incinerated, or fed to a sulfur recovery unit.
The Claus process is used to treat gas streams containing high (above 50%) concentrations of H2S.
The Claus process as used today is a modification of a process first used in 1883 in which H2S was reacted over a catalyst with air (oxygen) to form elemental sulfur and water.
H2S+1/2 O2 → S +H2O Eq. 6-34
Control of this highly exothermic reaction was difficult and sulfur recovery efficiencies were low. In order to overcome these process deficiencies, a modification of the Claus process was developed and introduced in 1936 in which the overall reaction was separated into:
1- A highly exothermic thermal or combustion reaction section in which most of the overall heat of reaction (from burning one-third of the H2S and essentially 100% of any hydrocarbons and other combustibles in the feed) is released and removed, and
2- A moderately exothermic catalytic reaction section in which sulfur dioxide (SO2) formed in the combustion section (step 1) reacts with unburned H2S to form elemental sulfur. The principal reactions taking place (neglecting those of the hydrocarbons and other combustibles) can then be written as follows:
Thermal or Combustion Reaction Section
H2S+3/2 O2 → SO2 +H2O Eq. 6-35
Combustion and Catalytic Reaction Sections
SO2 +2H2S → 3S+ 2H2O Eq. 6-36
3H2S+3/2 O2 → 3S+ 3H2O Eq. 6-37
Figure 6-18 shows a simplified flow diagram of a two-stage Claus process plant. The first stage of the process converts H2S to sulfur dioxide and to sulfur by burning the acid gas stream with air in the reaction furnace. This provides SO2 for the next phase of the reaction.
Gases leaving the furnace are cooled to separate out elemental sulfur formed in the thermal stage. Reheating, catalytically reacting, and sulfur condensation removes additional sulfur. Multiple reactors are provided to achieve a more complete conversion of the H2S. Condensers are provided after each reactor to condense the sulfur vapor and separate it from the main stream.
Conversion efficiencies of 94-95% can be attained with two catalytic stages while up to 97% conversion can be attained with three catalytic stages.
The efficiencies are dictated by environmental concerns; the effluent gas (SO2) is either vented, incinerated, or sent to a “tail gas treating unit.”
Fig. 6-18. Simplified process flow schematic for a two-stage Claus process plant.
For the usual Claus plant feed gas composition (water-saturated with 30-80 mol % H2S, 0.5-1.5 mol % hydrocarbons, the remainder CO2), the modified Claus process arrangement results in thermal section (burner) temperatures of about 1800 to 2500°F.
Sulfur recovery would be expected to be lower for a feed gas from a refinery than for a wellhead treater because of higher hydrocarbon content.
Conversion of H2S to elemental sulfur is favored in the reaction furnace by higher operating temperatures of 1800°F and in the catalytic converters by lower operating temperatures of less than 700°F.
To attain an overall sulfur recovery level above about 70%, the thermal, or combustion, section of the plant is followed by one or more catalytic reaction stages. Sulfur is condensed and separated from the process gases after the combustion section and after each catalytic reaction stage in order to improve equilibrium conversion. The process gases must be reheated prior to being fed to the catalytic reaction stage in order to maintain acceptable reaction rates and to ensure that the process gases remain above the sulfur dewpoint as additional sulfur is formed. Figure 6-19 is the flow sheet of an example three-stage Claus sulfur recovery plant; Figure 6-20 shows the mechanical arrangement of an example small, package-type, two-stage Claus plant.
Gases leaving the final sulfur condensation and separation stage may require further processing. These requirements are established by local, or national regulatory agencies.
These requirements can be affected by the size of the sulfur recovery plant, the H2S content of the plant feed gas, and the geographical location of the plant.
Fig. 6-19. Three-Stage Sulfur Plant. (Straight-Through Operating with Acid Gas-Fueled Inline Burners for Reheating)
Fig. 6-20. Example Package-Type Sulfur Plant
6.10.1 Claus Process Considerations
The Claus sulfur recovery process includes the following process operations:
• Combustion — burn hydrocarbons and other combustibles and 1/3 of the H2S in the feed.
• Waste Heat Recovery — cool combustion products. Because most Claus plants produce 150-500 psig steam (365-470°F), the temperature of the cooled process gas stream is usually about 600-700°F.
• Sulfur Condensing — cool outlet streams from waste heat recovery unit and from catalytic converters. Low temperature of the cooled gas stream is usually about 350°F or 260-300°F for the last condenser.
• Reheating — Reheat process stream, after sulfur condensation and separation, to a temperature high enough to remain sufficiently above the sulfur dewpoint, and generally, for the first converter, high enough to promote hydrolysis of COS and CS2 to H2S and CO2.
COS + H2O → CO2 + H2S Eq 6-38
CS2 + 2H2O → CO2 +2H2S Eq 6-39
• Catalytic Conversion — Promote reaction of H2S and SO2 to form elemental sulfur.
6.10.2 Process Variations
Several variations of the basic Claus process have been developed to handle a wide range of feed gas compositions. Some of these are shown in Figure 6-21. Straight-through operation results in the highest overall sulfur recovery efficiency and is chosen whenever feasible.
Table. 6-12 can be used as a guide in Claus process selection.
Fig. 6-21. Claus Process Variations
Table. 6-12. Claus Plant Configurations
6.10.3 Combustion Operation
Most Claus plants operate in the "straight-through" mode.
The combustion is carried out in a reducing atmosphere with only enough air (1) to oxidize one-third of the H2S to SO2, (2) to burn hydrocarbons and mercaptans, and (3) for many refinery
Claus units, to oxidize ammonia and cyanides. Air is supplied by a blower and the combustion is carried out at 3-14 psig, depending on the number of converters and whether a tail gas unit is installed downstream of the Claus plant.
Numerous side reactions can also take place during the combustion operation, resulting in such products as hydrogen (H2), carbon monoxide (CO), carbonyl sulfide (COS), and carbon disulfide
(CS2). Thermal decomposition of H2S appears to be the most likely source of hydrogen since the concentration of H2 in the product gas is roughly proportional to the concentration of H2S in the feed gas. Formation of CO, COS, and CS2 is related to the amounts of CO2 and/or hydrocarbons present in the feed gas.
Heavy hydrocarbons, ammonia, and cyanides are difficult to burn completely in a reducing atmosphere. Heavy hydrocarbons may burn partially and form carbon which can cause deactivation of the Claus catalyst and the production of off color sulfur. Ammonia and cyanides can burn to form nitric oxide (NO) which catalyzes the oxidation of sulfur dioxide (SO2) to sulfur trioxide (SO3); SO3 causes sulfation of the catalyst and can also cause severe corrosion in cooler parts of the unit. Unburned ammonia may form ammonium salts which can plug the catalytic converters, sulfur condensers, liquid sulfur drain legs, etc. Feed streams containing ammonia and cyanides are sometimes handled in a special two-combustion stage burner or in a separate burner to ensure satisfactory combustion.
Flame stability can be a problem with low H2S content feeds (a flame temperature of about 1800°F appears to be the minimum for stable operation).
The split flow, sulfur recycle, or direct oxidation process variations often are utilized to handle these H2S-lean feeds; but in these process schemes, any hydrocarbons, ammonia, cyanides, etc. in all or part of the feed gas are fed unburned to the first catalytic converter. This can result in the cracking of heavy hydrocarbons to form carbon or carbonaceous deposits and the formation of ammonium salts, resulting in deactivation of the catalyst and/or plugging of equipment.
A method of avoiding these problems while still improving flame stability is to preheat the combustion air and/or acid gas, and to operate "straight-through". An example of such an arrangement is shown in Figure 6-22. Steam-, hot oil-, or hot gas-heated exchangers and direct fired heaters have been used. The air and acid gas are usually heated to about 450-500°F. Sometimes split flow is combined with acid-gas preheat. Other methods of improving flame stability are to use a high intensity burner, to add fuel gas to the feed gas, or to use oxygen or oxygen-enriched air for combustion.
6.10.4 Claus Unit Tail Gas Handling
The tail gas from a Claus unit contains N2, CO2, H2O, CO, H2, unreacted H2S and SO2, COS, CS2, sulfur vapor, and entrained liquid sulfur. Because of equilibrium limitations and other sulfur losses, overall sulfur recovery efficiency in a Claus unit usually does not exceed 96-97%. Venting of this tail gas stream without further processing is seldom permitted; the minimum requirement is normally incineration, the principal purposes of which are to reduce H2S concentrations to a low level (which value will depend on the local regulations) and to provide the thermal lift for dispersion of SO2 upon release to atmosphere through a stack. Depending upon the size of the
Claus unit, the H2S content of the feed gas, and the geographical location, a tail gas cleanup process may be required in order to reduce emissions to the atmosphere.
Fig. 6-22. Sulfur Recovery Process with Acid Gas and Air Preheat
Incineration of the H2S (as well as the other forms of sulfur) in the Claus plant tail gas to SO2 can be done thermally or catalytically. Thermal oxidation normally is carried out at temperatures between 900°F and 1500°F in the presence of excess oxygen. Most thermal incinerators are natural draft operating at sub-atmospheric pressure with air flow controlled with dampeners; the excess oxygen level varies between 20% and 100%. A typical concentration of oxygen in the stack effluent is 2.0%. Although the Claus unit tail gas contains some combustibles — for example, H2S, COS, CO, CS2, H2, and elemental sulfur (in the case of "split-flow" plants, some hydrocarbons) — these combustibles are at too low a concentration to burn since they generally amount to less than 3% of the total tail gas stream. The entire tail gas stream must therefore be incinerated at a high enough temperature for oxidation of sulfur and sulfur compounds to SO2.
Incinerator fuel consumption can be reduced significantly by utilizing catalytic incineration. This involves heating the tail gas stream to about 600-800°F with fuel gas and then passing the heated gas along with a controlled amount of air through a catalyst bed. Catalytic incinerators are normally forced draft, operating at a positive pressure in order to maintain closer control of excess air. Catalytic incineration is a proprietary process which should be considered where fuel costs for conventional (thermal) incineration are high.
Another method of improving overall fuel economy involves recovering heat from the incinerator outlet gases. Saturated steam at pressures ranging between 50 psig and 450 psig has been produced, and saturated steam has been superheated, using waste heat from the incinerator outlet gases.
Incinerators with waste heat recovery are normally forced draft operating at a positive pressure.
Fuel required for thermal incineration is determined by the amount of heat needed to heat the Claus tail gas, air, and fuel to the required temperature. Normally the incinerator is sized for at least 0.5 second residence time, and sometimes for as much as 1.5 seconds residence time. Generally, the longer the residence time, the lower the incinerator temperature needed to meet the environmental requirements. This is illustrated by Figure 6-23 which shows the relationship between residence time and temperature for a typical installation to meet a maximum H2S requirement of 10 ppmv.
Fig. 6-23. Typical Relationship Between Incinerator Residence Time and Required Temperature.
The incinerator and stack can sometimes be combined into a single vessel. The incinerator is the
Tail Gas Clean-up Processes (TGCU)
All of the Claus tail gas cleanup (TGCU) processes fit roughly into four categories:
Processes based primarily on the continuation of the Claus reaction to produce additional sulfur under more favorable equilibrium conditions than normally found in the Claus units, either through operation at temperatures below the sulfur dewpoint or in the liquid phase at a temperature above the melting point of sulfur.
Processes based on converting all the sulfur components in the tail gas to SO2 and recovering the SO2 for further processing.
Processes based on converting all the sulfur in the Claus unit tail gas to H2S, then recovering sulfur from this H2S.
Processes that directly oxidize the tail-gas H2S to sulphur.
Overall Claus plant conversion efficiency is maximized by maintaining the stoichiometric H2S:SO2 ratio of 2:1 in the process gas to the catalytic converters. The most suitable point for this determination is at the outlet of the last sulfur condenser because a slight change in the air:acid gas ratio at the front of the plant will result in a significant change in the H2S:SO2 ratio in the tail gas and in the theoretical overall sulfur recovery. An H2S:SO2 ratio in the tail gas of between 1:1 and 3:1 can be considered normal although the desired goal should be a 2:1 ratio.
Because of the effect of temperature upon the Claus reaction equilibrium, control of temperatures at various points in the process sequence is important. Unexpected changes in operating temperatures usually denote changes in conversion efficiency. For example, a decrease in the temperature rise across a catalytic converter bed is an indication of declining catalyst activity which may be caused by adsorption of elemental sulfur on the active surface area of the catalyst. Operating the catalyst bed at a temperature 50-100°F higher than normal for 24-48 hours will remove this sulfur from the catalyst and can restore its activity.
6.11 Gas Permeation Process (Membranes)
Membranes are thin semipermeable barriers that selectively separate some compounds from others.
Membranes are used in natural gas processing for dehydration, fuel-gas conditioning, and bulk CO2 removal, but presently CO2 removal is by far the most important application. In some applications, membranes are used to recover CO2 from EOR floods for recycle injection into oil and gas reservoir.
6.11.1 Membrane Fundamentals
Membranes do not act as filters where small molecules are separated from larger ones through a medium of pores. They operate on the principle of solution-diffusion through a nonporous membrane. Highly solubilized components dissolve and diffuse through the membrane.
Relative permeation rates
• Most soluble (fastest gases)
H2O, H2, H2S, CO2, O2
• Least soluble (slowest gases)
N2, CH4, C2+
CO2 first dissolves into the membrane and then diffuses through it. Membranes allow selective removal of fast gases from slow gases.
Membranes do not separate on the basis of molecular size. Separation is based on how well different compounds dissolve into the membrane and then diffuse it.
Fick’s law (known as Basic Flux Equation) is used to approximate the solution-diffusion process. It is expressed as
Ji (Si Di pi)/L Eq. 6-40
J is the flux of component i, that is, the molar flow of component i through the membrane per unit area of membrane,
Si is the solubility term,
Di is the diffusion coefficient,
pi is the partial pressure difference across the membrane, and
L is the thickness of the membrane.
Customarily, Si, and Di are combined into a single term, the permeability, Pi, and thus divides Fick’s law into two parts:
Pi /L, which is membrane dependent and
pi, which is process dependent.
(Note that Pi/L is not only dependent on the membrane but also dependent on operating conditions, because Si and Di depend on both temperature and pressure. Pi also depends weakly upon the composition of the gases present).
All the mixture components have a finite permeability, and the separation is based upon differences in them. Customarily, selectivity, 1-2, is used, which is the ratio of two permeabilities, P1/P2, a term important in process design and evaluation. An of 20 for CO2/CH4 means that CO2 moves through the membrane 20 times faster than does methane.
6.11.2 Membrane Selection Parameters
High permeability results in less membrane area required for a given separation and a lower system cost.
High selectivity results in lower losses of hydrocarbons as CO2 is removed and a higher volume of salable product.
Unfortunately, high CO2 permeability does not correspond to high selectivity. A choice must be made between a highly selective, or permeable, membrane and somewhere between on both parameters. The usual choice is to use a highly selective material and then make it as thin as possible to increase the permeability. Reduced thickness makes the membrane extremely fragile and therefore unusable.
In the past, membrane systems were not a viable process because the membrane thickness required to provide the mechanical strength was so high that the permeability was minimal.
6.11.3 Membrane Structure Types
Asymmetric Membrane Structure
An asymmetric membrane structure features a single polymer consisting of an extremely thin nonporous layer mounted on a much thicker and highly porous layer of the same material, as opposed to a homogenous structure, where membrane porosity is more-or-less uniform throughout. Figure 6-24 is an example of an asymmetric membrane.
Meets the requirements of the ideal membrane, that is, highly selective, and thin.
Provides mechanical support and allows the free flow of compounds that permeate through the nonporous layer.
Fig. 6-24. Asymmetric membrane structure, and a Composite membrane structure.
Composite Membrane Structure
The disadvantages of the asymmetric membrane structure are they are composed of a single polymer; they are expensive to make out of exotic, highly customized polymers; and they are produced in small quantities.
These drawbacks are overcome by producing a composite membrane.
The composite membrane consists of a thin selective layer made of one polymer mounted on an asymmetric membrane, which is made of another polymer.
The composite structure allows manufacturers to use readily available materials for the asymmetric portion of the membrane and specially developed polymers, which are highly optimized for the required separation and the selective layer.
Composite structures are being used in most newer advanced CO2 removal membranes because the proprieties of the selective layer can be adjusted readily without significantly increasing membrane cost.
6.11.4 Carbon Dioxide Removal from Natural Gas
Many different types of membranes have been developed or are under development for industrial separations, but for CO2 removal, the industry standard is presently cellulose acetate. In these membranes are of the solution-diffusion type, in which a thin layer (0.1 to 0.5 μm) of cellulose acetate is on top of a thicker layer of a porous support material. Permeable compounds dissolve into the membrane, diffuse across it, and then travel through the inactive support material. The membranes are thin to maximize mass transfer and, thus, minimize surface area and cost, so the support layer is necessary to provide the needed mechanical strength.
6.11.5 Membrane Elements
Commercial membrane configurations are either hollow fiber elements or flat sheets wrapped into spirally wound elements. Presently, about 80% of gas separation membranes are formed into hollow fiber modules, like those shown in Figures 6-26 & 6.27.
Flat Sheet (Spiral Wound)
In the spiral wound element shown in Figure 6-25, two membrane sheets are separated by a permeate spacer and glued shut at three ends to form an envelope or leaf. Many of these leaves, separated by feed spacers, are wrapped around the permeate tube, with the open end of the leaves facing the tube. Feed gas travels along the feed spacers, the permeating species diffuse through the membranes and down the permeate spacers into the permeate tube, and the residue gas exits at the end. The gas flow is cross flow in this configuration.
The spiral configuration is inherently more resistant than the hollow fiber membranes to trace components that would alter the polymer permeability. It also allows a wider range of membrane materials to be used. However, the hollow fiber membranes are cheaper to fabricate, and thus dominate the field.
Optimization involves the number of envelopes and element diameter.
Number of envelopes
The permeate gas must travel the length of each envelope. Having many shorter envelopes makes more sense than having a few longer ones because pressure drop is greatly reduced in the former case.
A larger bundle diameter allow better packing densities but increases the element tube size and decreases cost. A larger diameter also increases the element weight, which makes the elements more difficult to handle during installation and replacement.
As shown in Figures 6-27 and 6-27, very fine hollow fibers are wrapped around a central tube in a highly dense pattern. Feed gas flows over and between the fibers and some components permeate into them.
Permeate gas travels within the fibers until it reaches the permeate pot, where it mixes with the permeate from other fibers. The total permeate exits the element through a permeate pipe. Gas that does not permeate eventually reaches the element’s center tube, which is perforated. In this case, the central tube is for residual collection, not permeate collection.
The low-pressure, bore-feed configuration is a countercurrent flow configuration similar to a shell-tube heat exchanger with the gas entering on the tube side. It has the advantage of being more resistant to fouling because the inlet gas flows through the inside of the hollow fibers.
However, the mechanical strength of the membrane limits the pressure drop across the membrane. The configuration is only used in low-pressure applications, such as air separation and air dehydration.
To handle high pressures, the permeate flows into the hollow fiber from the shell side. This feature makes the membrane much more susceptible to plugging, and gas pretreatment is usually required. The gas flow is cross current and provides good feed distribution in the module. This configuration is widely used to remove CO2 from natural gas.
Fig. 6-25 Spiral wound membrane element. (UOP - LLC)
Spiral Wound Versus Hollow Fiber
Spiral Wound Hollow fiber
• Installed in horizontal vessels
• Operate at higher allowable operating pressures 1085 psig (75 barg) and thus have higher driving force available for permeation
• More resistant to fouling
• Have a long history of service in natural gas sweetening
• Perform best with colder inlet stream gas temperatures
• Do not handle varying inlet feed quality as well as hollow fiber units installed in vertical vessels
• Require extensive pretreatment equipment with high inlet stream liquid hydrocarbon loading • Characteristics of hollow fiber membranes
• Installed in vertical vessels
• Offer a higher packing density
• Operate at lower inlet stream pressures 580 psig (40 barg)
• Handle higher inlet stream hydrocarbon loading better than spiral wound units
• Require inlet feed gas chilling
• Hollow fiber based plants are typically smaller than spiral wound-based plants
• Handle varying inlet feed quality better than spiral wound units installed in horizontal vessels.
Finer fibers give higher packing density, but larger fibers have lower permeate pressure drops and so they use the pressure driving force more efficiently.
Table. 6-13. Spiral Wound Versus Hollow Fiber.
Once the membranes have been manufactured into elements, they are joined together and inserted into a tube (Figure 6-28). Multiple tubes are mounted on skids in either a horizontal or vertical orientation, depending on the membrane company.
Fig. 6-26. Hollow-fiber membrane element.
Fig. 6.27 Cutaway view of the two module configurations used with hollow fiber membranes.
Fig. 6-28. Cutaway view of spiral wound membrane module. (UOP - LLC.)
6.11.6 Membrane Design Considerations
Process Variables Affecting Design are:
184.108.40.206 Flow Rate
A maximum acceptable feed gas rate per unit area applies to the membrane, and required membrane area is directly proportional to the flow rate. Membrane units perform well at reduced feed rates, but their performance drops when design flow rates are exceeded. Additional modules are added in parallel to accept higher flow rates.
The percentage of hydrocarbon losses (hydrocarbon losses/feed hydrocarbons) remains the same at different flow rates.
220.127.116.11 Operating Temperature
Increased operating temperature increases permeability but decreases selectivity.
Membrane area requirement is decreased, but the hydrocarbon losses and recycle compressor power for multistage systems are increased (Figure 6-29).
Because membranes are organic polymers, they have a maximum operating temperature that depends upon the polymer used. Exceeding this temperature will degrade membrane material and shorten the useful life of the unit.
18.104.22.168 Feed Pressure
An increase in feed pressure decreases both membrane permeability and selectivity, but at the same time creates a greater driving force across the membrane that results in a net increase in permeation through the membrane and a decrease in the membrane area requirements (Figure 6-30). Increasing the maximum operating pressure results in a less expensive and smaller system. Limiting factors are the maximum pressure limit for the membrane elements and the cost and weight of equipment at the higher pressure rating.
Fig. 6-29. Effect of operating temperature.
Fig. 6-30. Effect of feed pressure.
22.214.171.124 Permeate Pressure
Exhibits the opposite effects of feed pressure Lowers the permeate pressure Increases the driving force, and lowers the membrane area requirements. Unlike feed pressure, permeate pressure has a strong effect on hydrocarbon losses (Figure 6-31).
Pressure difference across the membrane is not the only consideration. Pressure ratio across the membrane is strongly affected by the permeate pressure.
For example, a feed pressure of 1305 psig (90 bar) and a permeate pressure of 43.5 psig (3 bar) produce a pressure ratio of 30. Decreasing the permeate pressure to 14.5 psig (1 bar) increases the pressure ratio to 90 and has a dramatic effect on system performance.
Fig. 6-31. Effect of permeate pressure.
Desirable to achieve the lowest possible permeate pressure Important consideration when deciding how to further process the permeate stream.
For example, if permeate stream must be flared, then the flare design must be optimized for low pressure drop. If permeate stream must be compressed to feed the second membrane stage or injected into a well, the increased compressor horsepower and size at lower permeate pressure must be balanced against the reduced membrane area requirements.
126.96.36.199 CO2 Removal
For a constant sales gas CO2 specification, an increase in feed CO2 increases membrane area requirement and increases hydrocarbon losses (more CO2 must permeate, and so more hydrocarbons permeate). This is shown in Figure 6-32.
Fig. 6-32. Effect of CO2 removal.
Membrane area requirement is determined by the percentage of CO2 removal rather than the feed or sales gas CO2 specifications themselves.
For example, a system for reducing a feed CO2 content from 10% to 5% is similar in size to one reducing the feed from 50% to 25%, or one reducing a feed from 1% to 0.5%, if all have a CO2 removal requirement of about 50%.
Traditional solvent or absorbent-based CO2 technologies have the opposite limitation.
Their size is driven by the absolute amount of CO2 that must be removed. For example, a system for CO2 removal from 50% to 25% is substantially larger than one reducing CO2 from 1% to 0.5%. For this reason, using membranes for bulk CO2 removal and using traditional technologies for meeting low CO2 specifications makes a lot of sense. Depending on the application, either one or both of the technologies could be used.
An increase in CO2 content in feed gas of an existing membrane plant will results in sales gas with higher CO2 content. An additional membrane area can be installed to meet the sales gas CO2 content, although with increased hydrocarbon losses. For example, if heater capacity is available, the membranes can be operated at a higher temperature to also increase capacity.
If an existing non-membrane system must be de-bottlenecked, installing a bulk CO2 removal system upstream of it makes good sense.
188.8.131.52 Environmental Regulations
Environmental regulations dictate what can be done with the permeate gas, specifically whether it can be vented (cold or hot vent) to the atmosphere or flared either directly or catalytically. Ninety five to ninety nine percent CO2 yields low Btu/scf content (flare requires a minimum of 250 Btu/scf to burn).
Location often dictates a number of other issues, such as space and weight restrictions, level of automation, level of spares that should be available, and single versus multistage operation.
Fuel requirements can be obtained upstream of the membrane system, downstream of the pretreatment system, downstream of the membrane, and from the recycle loop in multistage systems.