CONFIDENTIAL
Best Practice Guide Polysulphide Use BPG MAT-06-I Revision 0 September 2005
Document Revision Log Number Rev 0
Issue Date: September 2005
Comments / Changes International Version of BPG MAT-06
This report has been classified as Confidential and is subject to US Export Control regulations and has been classified as ECCN EAR 99. This document is for use by the operating units constituting Shell Global Solutions and the companies with which they have a licence or a service agreement only, and is provided subject to the terms of the relevant licence or service agreements. It should not be applied in any specific situation without having obtained further clarification and advice from Shell Global Solutions. No member of the Royal Dutch/Shell Group of companies accepts any liability for the application of this document by anyone. Copyright of this document and the accompanying documentation is vested in one or more companies of the Royal Dutch/Shell Group of companies. All rights reserved. © 2005 Shell Global Solutions International B.V. and Shell Oil Company.
1
CONFIDENTIAL
Table of contents PAGE SCOPE ------------------------------------------------------------------------------------- 4 1.0 POLYSULPHIDE DESCRIPTION, SAFETY AND HANDLING CONSIDERATIONS ----- 4 1.1 POLYSULPHIDE DESCRIPTION AND STABILITY -------------------------------- 4 1.2 SAFETY CONSIDERATIONS ----------------------------------------------------- 5 1.3 HANDLING CONSIDERATIONS -------------------------------------------------- 5 2.0 POLYSULPHIDE PROTECTION MECHANISM ----------------------------------------- 6 2.1 WET HYDROGEN SULPHIDE CRACKING ---------------------------------------- 7 2.2 NH4HS CORROSION AND THE ROLE OF CYANIDES --------------------------- 8 2.3 EFFECT OF POLYSULPHIDE ON STABILITY OF FeS FILM ---------------------- 9 2.4 EFFECT OF POLYSULPHIDE IN REMOVING HCN-------------------------------- 10 3.0 GUIDELINES FOR POLYSULPHIDE USE---------------------------------------------- 11 3.1 TEMPERATURE GUIDELINES ---------------------------------------------------- 12 3.2 pH GUIDELINES ----------------------------------------------------------------- 12 3.3 DOSAGE RATE ------------------------------------------------------------------- 13 3.4 WETTING OF STEEL SURFACES------------------------------------------------- 14 3.5 INJECTION METHOD------------------------------------------------------------- 14 3.6 INJECTION POINT METALLURGY ------------------------------------------------ 14 3.7 CHOICE OF INHIBITOR CATION ------------------------------------------------ 15 3.8 CO-INJECTED WATER QUALITY ------------------------------------------------- 15 3.9 INJECTION SYSTEM CONFIGURATION ----------------------------------------- 16 3.10 LIMITS ON THIOCYANATE CONCENTRATION -------------------------------- 16 4.0 POLYSULPHIDE USE IN CATALYTIC CRACKING UNITS ----------------------------- 16 4.1 MAIN FRACTIONATOR OVERHEAD---------------------------------------------- 17 4.2 INTERSTAGE OF THE WGC ------------------------------------------------------ 18 4.3 AFTERSTAGE OF THE WGC------------------------------------------------------ 19 4.4 RECTIFIED ABSORBER (RA) COLUMN ------------------------------------------ 20 5.0 POLYSULPHIDE USE IN COKING ---------------------------------------------------- 21 6.0 POLYSULPHIDE USE IN SOUR WATER STRIPPERS --------------------------------- 22 6.1 INTRODUCTION ----------------------------------------------------------------- 22 2
CONFIDENTIAL
Table of contents (cont’d) PAGE 6.2 GUIDELINES FOR POLYSULPHIDE ADDITION TO SOUR WATER STRIPPERS----------------------------------------------------------------------- 23 6.3 SPECIAL CONSIDERATIONS ON MONITORING IN SOUR WATER STRIPPER SYSTEMS ------------------------------------------------------------- 24 7.0 MONITORING POLYSULPHIDE EFFECTIVENESS ------------------------------------ 24 7.1 SOUR WATER ANALYSES-------------------------------------------------------- 25 7.1.1 BASE-SOLUBLE IRON---------------------------------------------------- 25 7.1.2 COLOUR TEST FOR RESIDUAL POLYSULPHIDE------------------------- 26 7.1.3 IRON CYANIDE “SPOT TEST”-------------------------------------------- 26 7.1.4 TOTAL SOLUBLE IRON (UNFILTERED) ---------------------------------- 26 7.1.5 pH ------------------------------------------------------------------------ 27 7.1.6 CHLORIDE, AMMONIA, SULPHUR --------------------------------------- 27 7.1.7 NON-ROUTINE CYANIDE ANALYSES ------------------------------------ 27 7.2 DIRECT MONITORING OF HYDROGEN ACTIVITY------------------------------- 27 7.2.1 INSERTION PROBES ----------------------------------------------------- 28 7.2.1.1 SHELL PROBE --------------------------------------------------- 28 7.2.1.2 OTHER INSERTION PROBES ----------------------------------- 28 7.2.2 PATCH PROBES ---------------------------------------------------------- 29 7.2.2.1 ELECTROCHEMICAL -------------------------------------------- 29 7.2.2.2 PRESSURE ------------------------------------------------------ 29 7.2.2.3 VACUUM -------------------------------------------------------- 29 7.2.3 PLACEMENT OF PROBES------------------------------------------------- 29 7.2.4 READING OF PROBES --------------------------------------------------- 30 References
--------------------------------------------------------------------------- 31
APPENDIX A
ANALYTICAL PROCEDURE ---------------------------------------------- 41
APPENDIX B
IN-SITU GENERATION OF POLYSULPHIDE ----------------------------- 42
APPENDIX C
POLYSULPHIDE USE AT DPR CAT CRACKING UNIT ------------------- 44
APPENDIX D
POLYSULPHIDE USE AT MARTINEZ REFINERY CCU ------------------- 48
APPENDIX E
POLYSULPHIDE USE AT MARTINEZ REFINERY FLEXICOKER ---------- 50
APPENDIX F
CORROSION OF AIR COOLERS AND POLYSULPHIDE USE AT MARTINEZ REFINERY SWS -------------------------------------------- 52 3
CONFIDENTIAL
SCOPE This Best Practice Guide (BPG) covers various aspects of polysulphide technology applied to refinery sour water systems. Covered are properties of polysulphide important for its intended use, its activity as a film-forming inhibitor against carbon steel corrosion, and its reactivity with cyanides. Also discussed are techniques to be considered in the practice of injection of polysulphide into vapour and water streams, and analytical methods for detecting whether polysulphide activity is adequate. This document compliments the Best Practice Guide MAT-13-I, "Guideline to Ammonium Hydrosulphide Corrosion Control", to be issued.
1.0 POLYSULPHIDE DESCRIPTION, SAFETY AND HANDLING CONSIDERATIONS 1.1
POLYSULPHIDE DESCRIPTION AND STABILITY
Polysulphides are metastable compounds of sulphur and ammonium or sodium sulphide. The chemical formula is (NH4)2Sx or Na2Sx, where Sx represents a sulphur chain, with x ranging from 2-5. Activity increases with increasing sulphur in the molecule. Due to this activity, the polysulphide sulphur is able to act as a sulphiding agent in stabilization and enhancement of surface films, or to react with cyanide ion to form thiocyanate. The “active” sulphur atoms are those most likely to break off the sulphur chain of the polysulphide molecule. For example, when the S5 ammonium polysulphide molecule decomposes, the following occurs: (NH4+)2 ⋅ [S ⋅ S ⋅ S ⋅ S ⋅ S]
-2
→ (NH4+)2 ⋅ [S ⋅ S ⋅ S ⋅ S]
-2
+ S0
Reactivity decreases significantly with decreasing sulphur on the molecule; S5 and S4 are most likely to break off the chain and are the most reactive, S3 and S2 react only slowly, and the sulphide sulphur [(NH4+)2S-2] is not reactive at all. In the present context “active” is considered the same as “reactive”; one standard test for polysulphide “activity” of a sour water sample is to measure the conversion of a known amount of cyanide added to the sample. Commercial ammonium polysulphide typically contains 40-50% (NH4)2Sx, 25%-35% ammonium hydroxide, NH4OH, and balance water. The ammonium hydroxide is used to keep a high pH (pH ca. 11), necessary to prevent the (NH4)2S5 from breaking down into lower sulphur species, (NH4)2S4 and (NH4)2S3. NH4OH is preferred to other bases, so as not to introduce another cation (e.g., Na, K, etc.).∗
∗ An example of a commercial ammonium polysulfide formulation is 49%w (NH4)2Sx, 33%w NH4OH, and balance (18%w) water, according to the Materials Safety Data Sheet (MSDS) from one manufacturer (Hickson Kerley). Sx is almost all S5, which has been verified by tests. Thus the total sulfur in this formulation is [5 * MW S / MW (NH4)2S5] * 49% = (160/196) * 49% = 40%w. Since all sulfur in commercial polysulfide is S5 and considered active, the active sulfur in this formulation is (128/196) * 49%w = 32%w (all sulfur but the sulfide sulfur), leaving the inactive sulfide sulfur (S-2) as 8%w. 4
CONFIDENTIAL
The stability of polysulphide solutions is affected by temperature and pH. At low pH (at approximately pH < 7), elemental sulphur will precipitate, virtually independent of temperature: (NH4)2Sx + H+ → 2NH4+ + HS- + S0x-1,
(1)
where x implies all polysulphide sulphur is subject to reaction (1). At high temperature, polysulphide will disproportionate into thiosulphate and bisulphide, with greater disproportionation at higher pH: (NH4)2S5 + 3OH- = (NH4)2S2O3 + 3HS-
(2)
The formation of thiosulphate is kinetically very slow and becomes significant only above about 149°C [300°F]. However, for injection of polysulphide at temperatures above about 110°C [230°F], polysulphide can decompose to elemental sulphur, due to water evaporation and/or a drop in pH after injection. The stability of polysulphide as a function of temperature and pH is summarized in Figure 1, modified from Reference 1, with Shell Global Solutions recommended operating range and experience superimposed. This is the basis for the temperature and pH guidelines in this document. It is important to note that Figure 1 is a snapshot based upon a certain ratio of S-2 to S0 (in this case 10:1, which is a typical sour water sulphur ratio). The incipient sulphur precipitation curve will probably be shifted further to the right for typical polysulphide solutions, but no calculations have been done. Ionic equilibrium modelling is needed to track the sulphur precipitation and/or polysulphide stability under varying conditions. The Shell Global Solutions recommended operating region is cut off at about pH 10, because above that pH stability of the sulphide film at the metal wall weakens, and there is the possibility of dropping out carbonate salts, as will be discussed later.
1.2
SAFETY CONSIDERATIONS
Ammonium and sodium polysulphide are both highly alkaline. Some commercial solutions of ammonium polysulphide contain “anti-freeze” additive(s) to enable their use at lower temperatures. Material Safety Data Sheets (MSDS) for the actual polysulphide solution used should be obtained and updated when the supplier changes or the formulation is changed. Samples drawn off sour water systems for analysis of active polysulphide sulphur usually have a significant amount of hydrogen sulphide present. Such samples, as well as any leak in the system, should be treated as potentially hazardous.
1.3
HANDLING CONSIDERATIONS
Ammonium polysulphide, rather than sodium polysulphide, has most often been used in Shell, although sodium polysulphide has been used in particular applications. When ammonium polysulphide is exposed to air, it will oxidize to thiosulphate and sulphate: (NH4)2S3 + 3O2 + 2NH4OH → (NH4)2S2O3 + (NH4)2SO4 + H2O
(3)
where x=3 was chosen for stoichiometric convenience only.
5
CONFIDENTIAL
Thus, when stored in tanks or vessels, the product should be blanketed with nitrogen or covered with a layer of diesel oil to minimize odour emissions, air uptake, and breakdown of the polysulphide. Nitrogen blanketing is mostly used at Shell because of concerns that the diesel oil could contaminate the injected polysulphide. With nitrogen blanketing, care should be taken to avoid pressuring out polysulphide when pumps are down. Below about 3°C [38°F], ammonium polysulphide, but not sodium polysulphide, tends to salt out of the concentrated solution. Equipment handling and delivering the material should, therefore, be insulated and heat traced. Storage and delivery temperatures of polysulphide should be kept below 54°C [130°F] to avoid polysulphide decomposition. Above about 54°C [130°F], the S5 in the polysulphide molecule starts to break down to the less reactive S4. Therefore, electrical tracing is recommended, rather than steam tracing, if lines need to be kept warm. A proprietary method of dealing with the salting-out problem during cold weather conditions is the addition of morpholine to stored ammonium polysulphide solutions. If ammonium polysulphide with morpholine is used, the handling temperature can be lower. Consult manufacturer for specifics. Some locations prepare ammonium polysulphide in-situ, by passing high pH sour water over sulphur sticks or flowers of sulphur; more details of this preparation method are given in Appendix B. In-situ generation is not recommended, however, because of uncertainty over the active sulphur content, difficulties in system control, and need to maintain the equipment. In all cases, sufficient training should be given to those personnel involved in handling and storage of polysulphide. Important items to stress are maintenance of storage temperature, (10 to 54°C [50 to 130°F] is recommended), and maintenance of any tracing. A key consideration in the use of polysulphide is to add it sparingly. Only as much polysulphide as needed should be added. Sulphur precipitation problems, plugging and/or corrosion in downstream equipment can occur if polysulphide is overdosed or if polysulphide decomposes because pH is too low or temperature is too high. Caution: Spills of ammonium polysulphide will liberate NH3 and H2S. Any spilled polysulphide solution sent untreated to an effluent treatment plant will contain sulphides, a reduced sulphur species that increases oxygen demand at the effluent treatment plant. Consult MSDS sheets for guidelines as to treatment of spills.
2.0
POLYSULPHIDE PROTECTION MECHANISM
Injection of aqueous polysulphide solutions into refinery process streams is a method of inhibiting most forms of wet hydrogen sulphide cracking in carbon steel equipment. Protection of carbon steel equipment occurs by two basic mechanisms: (1) stabilizing iron sulphide films, and (2) reaction with cyanides. Polysulphide also finds use in reducing concentrations of cyanides in effluent waters. These two mechanisms are discussed below. For the role of polysulphide in stabilizing iron sulphide films, first some background is given on wet H2S cracking in the absence of polysulphide, NH4HS corrosion and role of cyanides in accelerating corrosion.
6
CONFIDENTIAL
2.1
WET HYDROGEN SULPHIDE CRACKING
“Wet hydrogen sulphide cracking” is a summary term for a variety of stress corrosion cracking failure modes of carbon steel exposed to aqueous sour refinery environments. These are sulphide stress cracking (SSC), hydrogen blistering, hydrogen-induced cracking (HIC), stress-oriented hydrogen-induced cracking (SOHIC), and intergranular stress corrosion cracking. Sulphide stress cracking is generally addressed by controlling material hardness, but polysulphide has been shown to be of benefit in mitigating hydrogen blistering, HIC, SOHIC, and intergranular stress corrosion cracking. Figure 2 summarizes the critical pH ranges in wet hydrogen sulphide cracking for relatively low concentrations of ammonium hydrosulphide. Note from Figure 2 that cyanides only have an effect on wet hydrogen sulphide cracking in a relatively narrow pH range, approximately pH 7.5-8. 5. However, this pH range is typical for many refinery sour water systems. Hydrogen Blistering, HIC, and SOHIC These cracking mechanisms, like sulphide stress cracking, are all associated with corrosion reactions which cause hydrogen entry into steel. They result from reaction of iron at the steel surface with ammonium hydrosulphide (bisulphide), liberating atomic hydrogen as a reaction product in addition to iron sulphide. Fe + 2HS- → FeS↓ + S-2 + 2 Ho
(4)
In the presence of hydrogen sulphide, the rate of combination of hydrogen atoms to form hydrogen molecules at the metal surface is greatly reduced. Therefore, a significant portion of the atomic hydrogen diffuses into the steel where it recombines to form molecular hydrogen at imperfections or inclusions in the steel. Since molecular hydrogen is too large to diffuse through the steel, it builds up internal pressure which eventually may lead to the formation of blisters. In hydrogen-induced cracking, blisters link up internally and may ultimately break through to the surface. The effect of cyanide on these corrosion mechanisms is indirect and complex. It is discussed in more detail in Section 2.2. The reaction product of iron and bisulphide, ferrous sulphide (FeS), lends corrosion protection, but the degree of protection provided is a function of pH (see Figure 2). At higher pHs the resiliency of the corrosion inhibiting sulphidic film increases. Figure 2 also shows that in the pH range of 7.5 to 8.5, where the surface film is already relatively unstable, cyanides may further contribute to its removal.∗ That the iron sulphide film becomes more stable at higher pH has been proven experimentally using solutions containing 50 ppm to 3.5% ammonium hydrosulphide. At higher concentrations, i.e., at solution concentrations >3.5% ammonium hydrosulphide in polysulphide-free solutions, the sulphidic surface films become less protective, particularly for carbon steel. General corrosion rates in stronger ammonium hydrosulphide solutions are higher, and the protective quality of the surface films becomes dependent on process flow conditions. A more detailed account of ammonium
∗ More exactly, the iron sulfide corrosion product which lends most protection is thought to be the FeS2 film. The corrosion products which are likely to occur with carbon steel in wet H2S environments are a combination of FeS/FeS2. In the intermediate pH range of approximately 6-9.5, HS- is the predominant sulfur species in solution (as compared to H2S and S-2), and the more protective FeS2 films appear to be suppressed (References 2, 3). 7
CONFIDENTIAL
hydrosulphide corrosion is given in the Best Practice Guide MAT-13-I, "Guideline to Ammonium Hydrosulphide Corrosion Control", to be issued. In summary, corrosion by ammonium hydrosulphide depends on its concentration in aqueous solution and the solution pH. At constant pH, corrosion becomes more severe as ammonium hydrosulphide concentration increases. At a constant ammonium hydrosulphide concentration, corrosion rates decrease as pH increases. These two effects are linked in a refinery processing high nitrogen feeds. There, raising the pH of an already alkaline aqueous solution will likely increase corrosion rates. Refinery systems are always hydrogen sulphide dominated and, therefore, a higher ammonia production leads to an increase in the ammonium hydrosulphide concentration, which overcompensates the mitigating effect of the higher pH. Further, at high ammonia partial pressures, FeS surface films become fluffy and are charged with ammonia. They are less protective, and lead to increased corrosion and hydrogen permeation. Intergranular Stress Corrosion Cracking This form of cracking is not related to hydrogen permeation in the steel, but instead is dependent on moderate to high carbonate concentrations in alkaline sour water. As summarized in Figure 2, critical pH levels for this type of cracking are pH 9-11, and carbonate levels above about 800 ppm (for total sulphur <0.5%w), or 1200 ppm (for total sulphur above about 0.5%w). A carbonate film on carbon steel surfaces is moderately protective against general corrosion, but may not be resistant to (intergranular) stress corrosion cracking within a certain surface potential range. The tendency to stress crack is enhanced by the incorporation into the carbonate film of sulphidic impurities, which act as crack initiators.
2.2
NH4HS CORROSION AND THE ROLE OF CYANIDES
In situations where the protective sulphide surface film on steel is not completely stable, either due to the sour water pH which is too low or due to the presence of carbonates, cyanides may react stoichiometrically with iron sulphide. The resulting removal of the protective film (which is an iron sulphide corrosion product) exposes fresh steel surface to renewed direct contact with and corrosion by the ammonium hydrosulphide solution. The maximum effect of cyanides occurs in the pH range 7.5-8.5, where the protective iron sulphide film can be destabilized by reaction with cyanide to form ammonium ferrous-cyanide, in the absence of oxygen, such as FeS + 6 NH4+ + 6 CN- → (NH4)4 Fe (CN)6 + (NH4)2S
(5)
This reaction occurs in ionic form in aqueous solution. Below pH = 7.5, HCN remains as a stable molecule and very little ionizes to CN-. Thus there is very little attack on the FeS film. When there are leaks to the atmosphere, or when units are opened for inspection, the ferrous ion (Fe+2) is oxidized to ferric ion (Fe+3), and the ammonium ferrous/ferric cyanide complex NH4 Fe+2[Fe+3(CN)6] is formed: 4 (NH4)4 Fe (CN)6 + 1/2 O2 + H2O → 2 NH4Fe+2[Fe+3(CN)6] + 12 NH4CN + 2 NH4OH
(6)
8
CONFIDENTIAL
Ammonium ferrous/ferric cyanide has a characteristic intense “Prussian” blue colour, and its presence is easily recognized. In practice, the ammonium ferrous/ferric cyanide is usually found combined with a ligand (L), such as chlorides, phenolics, thiocyanates, formates, which replace some of the cyanide in the complex, as NH4Fe[Fe(CN4)L2]. Note that ammonium hydroxide is produced in reaction (6), and that this results in a characteristic ammonia smell when these cyanide complexes are present. When the Prussian Blue colour is seen upon equipment inspection, it is misleading to call it evidence of “cyanide corrosion”. Instead, cyanide has reacted with the iron that was already in the aqueous phase, typically from corrosive attack due to ammonium hydrosulphide or perhaps hydrogen chloride. Cyanide can be thought of as having an “accelerating” effect on corrosion that is already occurring. When no corrosion of this type occurs (in solutions containing no hydrogen sulphide), the corrosion rate of cyanide with steel is <0.05mm/yr [2 mpy] for all cyanide concentrations and temperatures up to 100°C [212°F] (Reference 4). No data are available for higher temperatures. When a well developed blue colour is observed on equipment surfaces, more detailed inspection is needed to assess the extent of damage to underlying steel. Hydrogen cyanide present exclusively in the vapour phase will not contribute to destabilizing iron sulphide films. HCN must be transferred from the gas phase into the aqueous phase, and ionize, in order to contact and subsequently react with the iron sulphide surface film. The transfer of HCN into the aqueous phase is difficult and therefore slow. Also, some cyanide will partition into the hydrocarbon phase. The transfer to the aqueous phase is influenced by the hydrogen cyanide partial pressure, temperature and pH of the water phase (Reference 5). These effects are discussed more fully in Section 2.4.
2.3
EFFECT OF POLYSULPHIDE ON STABILITY OF FeS FILM
Polysulphide reinforces the stability of the protective sulphide corrosion product film, but only if the pH exceeds 8. This holds whether or not cyanides are present. The iron sulphide species FeS2 provides the most protection against corrosion. Polysulphide active sulphur may be thought of as “adding” sulphur to the (less protective) FeS film to form FeS2, making the iron sulphide film more tenacious. The amount of polysulphide needed to reinforce the protective FeS film has been determined indirectly, namely by a study of carbonate type cracking of carbon steel exposed to refinery sour waters. The amount of polysulphide needed to lend corrosion protection was demonstrated in a series of slow strain rate tests. Slow strain rate data show that the needed polysulphide addition rate is temperature dependent: 12 ppm active sulphur at 65°C [150°F] versus 250 ppm at 110°C [230°F]. If a sufficient amount of a strong sulphiding agent such as polysulphide is present in the environment at a sufficiently high pH (above about 8), the entire carbonate surface film will be replaced by a tenacious sulphide film, which lends overall protection. It is assumed that the amount of polysulphide which inhibits intergranular stress corrosion cracking will also adequately mitigate hydrogen blistering, HIC, and SOHIC. Injection rates to obtain stable FeS films are usually defined at moderate temperatures (around 65°C [150°F]). Hence required dosages of Sx-2 are small, and these small dosages are practically ineffective at high injection temperatures. In other words at the hot polysulphide injection point, the usual dosages of polysulphide are insufficient to provide stabilization of the iron sulphide film.
9
CONFIDENTIAL
The amount of polysulphide needed to convert cyanides, which have been transferred into the aqueous phase, needs to be added to these quantities plus a reasonable excess over the stoichiometric requirement. Industry sources recommend 50% excess, which accounts for the different kinetic behaviour of the individual atoms in the polysulphide sulphur chain. When polysulphide is used in particulate-containing streams, some of the polysulphide may adsorb on the solids and not be available to protect vessel walls or react with cyanide.
2.4
EFFECT OF POLYSULPHIDE IN REMOVING HCN
Polysulphide is also used to scavenge cyanide, either to mitigate destabilization of the FeS film, or to meet effluent treating limits on cyanide. Polysulphide is very soluble in the aqueous phase, has only very limited solubility in the hydrocarbon phase, and does not exist in the vapour phase. For polysulphide to be an effective “scavenger” of cyanide, the steps indicated in Figure 3, must take place. In summary, molecular HCN must transfer via absorption from the vapour to aqueous phase, ionize to H+ and CN-, and then the CN- must react with polysulphide to form thiocyanate (SCN-). First, the HCN vapour must be absorbed by the aqueous polysulphide solution; this mass transfer is enhanced by contacting trays or sprays with small, high velocity drops. The absorbed molecular HCN then ionizes to H+ and CN- in the aqueous phase, which is unfavourable at low to moderate pH (7-8). For instance, a concentration of 200 ppmv HCN in the gas phase at 1.7 bar [25 psia] total pressure equilibrates with 20.5 ppm undissociated HCN in the aqueous phase at 65°C [150°F]. Ionic equilibrium calculations show that HCN is 31% dissociated at pH 8, 82% dissociated at pH 9, and 98% dissociated at pH 10. [Henry’s law constants and dissociation constants in Reference 5 were used.] Experience at Shell Global Solutions has indicated that once the cyanide ion is in solution, the reaction with the S4 and S5 species on the polysulphide molecule is relatively fast, complete in a matter of minutes. However, the reaction rate is still not fast enough to overcome typical mass transfer resistance of HCN from the gas to water phase. The reactivity of polysulphide is enhanced by having more of the higher order sulphur species on the molecule, e.g. (NH4)2S5 or (NH4)2S4. This is a strong justification for using purchased polysulphide solutions rather than using in-situ polysulphide generation, since commercial solutions are typically greater than 95% S5. The reaction rate is also enhanced by having a high excess of active polysulphide sulphur with respect to cyanide ion. But high excess polysulphide sulphur can cause problems of sulphur precipitation and corrosion in downstream equipment. The reaction between cyanide ion and polysulphide active sulphur to form thiocyanate (Step 3 in Figure 3) is an irreversible reaction (Reference 6, 7). This implies that SCNwill not revert to cyanide, and the polysulphide wash water can be reused for a second step of scrubbing, if CN- is sufficiently reacted and the SCN- concentration is within limits - see guidelines later.
10
CONFIDENTIAL
3.0
GUIDELINES FOR POLYSULPHIDE USE
The guidelines for use of polysulphide, for either reinforcing the stability of the sulphide film, or reacting with cyanide, are discussed below. Also, the temperature and pH guidelines are summarized in Figure 1. Polysulphide is always added as an aqueous solution; therefore the water wash guidelines apply for injection of polysulphide into a vapour stream. The most important principles are: 1)
Maintenance of <2%w ammonium hydrosulphide (NH4HS) in the aqueous phase in carbon steel equipment.∗
2)
Annular flow in heat exchanger tubes.
3)
>50% of polysulphide aqueous solution remaining as liquid after equilibrium flash at injection point.
In addition, there are criteria relating to temperature and pH: 4)
The temperature of the receiving stream at the injection point should be a maximum of 121°C [250°F] before polysulphide addition, and a maximum of 110°C [230°F] after equilibrium flash of the injected polysulphide.
5)
The pH of the polysulphide solution, after injection, should not be allowed to fall below about pH 8.0. This is a concern especially for injections into hot vapour streams, above about 93°C [200°F].
Reasons for the temperature and pH criteria are discussed in Sections 3.1 and 3.2. Ionic modelling should be done to predict the equilibrium pH of the injected polysulphide solution into hot environments. Contact Shell Global Solutions. Laboratory experiments at Shell Global Solutions US have quantified the dosage of polysulphide active sulphur necessary to stabilize iron sulphide films against NH4HS corrosion (see Section 3.3), so in principle operation at higher levels of NH4HS than 2% should be possible in carbon steel, when the system is polysulphide dosed. But the amount of polysulphide active sulphur needed increases by more than an order of magnitude as temperature goes from 65°C to 110°C [150°F to 230°F], and the effectiveness of polysulphide in corrosion protection depends on maintaining this level of active sulphur at pH >8. Typical vapour phase applications are condensing systems that cool from about 110°C [230°F] and may pass through regions of pH < 8. Also, polysulphide treatment upstream of a separator has no effect on ammonium hydrosulphide precursors (NH3, H2S) leaving that separator. For these reasons, no “corrosion protection credit” should be taken for the presence of polysulphide in condensing systems. Normally, polysulphide should be used where the presence of cyanide might lead to destabilization of iron sulphide protective films and NH4HS corrosion in carbon steel or alloys, as a preventive measure for suspected intergranular stress corrosion cracking, or in liquid streams high in NH4HS, even in the absence of cyanides.
∗ If polysulfide is present at very high levels of excess active sulfur, for instance insitu generation systems, experience has shown that levels of ammonium hydrosulfide greater than 2% can be tolerated in carbon steel. See Appendices B and D. 11
CONFIDENTIAL
3.1
TEMPERATURE GUIDELINES
The temperature at and downstream of the polysulphide injection point should not exceed 110°C [230°F]. If the polysulphide solution is flashed above about 110°C [230°F], polysulphide will decompose to H2S and elemental sulphur, Sno. Also, if given enough time, polysulphide will slowly decompose to thiosulphate. It is recommended to inject polysulphide into vapour phases with temperature no higher than 121°C [250°F], as long as the equilibrated temperature after injection is 110°C [230°F] or below. Experience has shown that polysulphide injection above about 121°C [250°F] may cause sulphur to precipitate from the polysulphide solution. There is the possibility of depositing molten sulphur; the melting point of elemental sulphur is 113-121°C [235-250°F], depending on the sulphur form. It is always best to inject water to drop the stream temperature below 110°C [230°F], and then inject polysulphide solution. If elemental sulphur precipitates from solution, it can be very corrosive to carbon steel, up to several mm/yr [hundreds of mpy]. If elemental sulphur is dry, it is not corrosive to carbon steel, but in the presence of even slight amounts of water, sulphur can be very corrosive. Corrosive attack by sulphur is mitigated by “significant” levels of NH3 in solution (as NH4OH is added, sulphur dissolves in solution at high pH to form ammonium polysulphide). Excess sulphur suspended in the hydrocarbon phase may also cause problems in downstream treaters. The low temperature limit on polysulphide is governed by preventing freezing and salting out of the polysulphide solution. Temperature should be kept above 10°C [50°F] for ammonium polysulphide without freeze protective additives (ammonium polysulphide starts to salt out at 3°C [38°F] and could plug lines). If ammonium polysulphide with morpholine or other additive for freeze protection is used, consult with the manufacturer. Sodium polysulphide freezes at -34°C [-29°F] and therefore needs no special freeze protection.
3.2
pH GUIDELINES
The pH should be between pH 8 and pH 10, measured in the cool (<65°C [150°F]) sour water in the downstream separator boot after polysulphide addition, to achieve optimum polysulphide effectiveness. pH should be measured “closed cup” at sampling locations where the pressure is greater than about 1 barg [15 psig], to avoid false (too high) pH readings due to flashing of H2S and perhaps CO2 when the sample is taken. If tests have shown that there are no differences between closed cup and open cup pH at the particular sample point, then open cup pH measurements are acceptable. Below pH of about 8, polysulphide starts to decompose to form elemental sulphur and the corrosion protective sulphide film on carbon steel surfaces becomes unstable. Also, between pH of about 7.5 to 8.5, cyanides can react stoichiometrically with the iron sulphide film, as per equation (5). Fortunately, the partitioning of cyanides from the gas phase into the aqueous phase is low at low pHs.
12
CONFIDENTIAL
The calculated pH of the flashed injected solution should be checked. Normally, pH at the polysulphide injection point will be lower than that measured in the sour water in the downstream separator. At the injection point, pH decreases as a result of heating alone. For instance, when an ambient temperature polysulphide solution is suddenly heated to 110°C [230°F], the pH will decrease about 1 pH unit (exact pH change will depend on receiving stream vapour composition), and the resulting pH may be low enough that elemental sulphur will drop out. Injection at receiving stream temperatures significantly lower than 110°C [230°F] reduces this effect. Also, if a large amount of water is evaporated in the injection zone, it is very likely that the pH of the remaining aqueous solution will drop below the stable range for polysulphide and elemental sulphur will be formed. The lower pH at the injection point is due to interaction of the hot vapour stream with the cool injected polysulphide solution, and flashing of various components, including ammonia and H2S, from the polysulphide solution. If there are acids in the vapour receiving stream which readily absorb in water (such as HCl), the pH at the injection point could decrease further. The upper pH limit is based on several factors. The increased cyanide transfer into the aqueous phase at higher pHs is actually desirable, since higher pHs enhance the stability of the protective sulphide film and make it resistant to cyanide attack. However, the enhanced scrubbing ability of the higher pH sour water also transfers more carbon dioxide into the aqueous phase. The scrubbed out carbon dioxide is converted to carbonate which may induce intergranular stress corrosion cracking. The intergranular, carbonate driven failure mode is not a problem as long as residual polysulphide is present in the sour water dropout phases at all times. However, higher pHs can lead to plugging due to ammonium carbonate deposition. Recent data indicate that there may be a slightly lower pH where polysulphide is stable in highly buffered systems that contain substantial quantities of bisulphide ion (HS-) and carbonate ion (CO3-2). This may explain why polysulphide is used successfully in scrubbing of Wet Gas Compressor (WGC) afterstage in cat cracking units, where pH in the afterstage separator is often 7.5-8.0.
3.3
DOSAGE RATE
The minimum polysulphide dosage to achieve corrosion control is primarily temperature dependent: 12 ppm of active polysulphide sulphur at 65°C [150°F] versus 250 ppm at 110°C [230°F]. This is measured in the water boot of the downstream separator. Measurement in the water boot takes account of any partial reaction with cyanides that takes place, namely the conversion of cyanide in the aqueous phase to thiocyanate. The upper limit for residual polysulphide in sour water is usually based on sour water stripper limitations. The total active sulphur in the combined feed to any sour water stripper should not be greater than 25 ppm (NH4)2Sx . Decomposition of polysulphide in neutral pH sour water stripper bottoms may cause corrosion and plugging of reboiler tubes by elemental sulphur. If the goal is HCN scrubbing, to have a “reasonable” reaction rate with cyanide (most reaction occurring in a few minutes), the active polysulphide sulphur (S4, S5) should be at least 50% excess with respect to the cyanide ion. Only the active polysulphide sulphur should be used when calculating polysulphide needs for reacting with cyanides. However, the actual reaction rate is usually limited by mass transfer of HCN to the water phase and the ionization to CN-, which is dependent on pH and temperature. See discussion in Section 2.2.
13
CONFIDENTIAL
The use of purchased polysulphide is recommended, because the active sulphur typically is over 95% S5 . As discussed in Section 1, a typical commercial polysulphide formulation is 40%w total sulphur, of which 32%w is active sulphur and 8%w sulphide sulphur. The manufacturer of the purchased polysulphide should be consulted for the percent S4 and S5 on sulphur chains, and the delivered polysulphide should be checked routinely for active sulphur. In-situ produced polysulphide contains mostly S2 and S3 sulphur chains, and therefore is less reactive and higher excess polysulphide is needed.
3.4
WETTING OF STEEL SURFACES
For polysulphide to have any effect on reducing corrosion and wet H2S cracking, it must be in the aqueous phase and in contact with the metal surface. Polysulphide decomposes in the vapour phase and is only very slightly soluble in the hydrocarbon phase. For it to be effective, the polysulphide solution which remains after the system reaches equilibrium must wet the metal surfaces. This is certain in the bulk water phase, very difficult in the hydrocarbon phase (only by dispersion), and impossible in the gas phase. That is one reason why in coolers with the process stream on the shell side, the upper quadrant of the shell is difficult to protect.
3.5
INJECTION METHOD
Purchased ammonium polysulphide should not be injected directly into refinery process streams, but rather blended first with (sour) wash water. This blended sour water stream is fed, for instance, to CCU wet gas compressor discharges. Industry practice is that the maximum concentration of (NH4)2Sx in the resulting injection water should not exceed 5% to avoid sulphur precipitation (Reference 8). Typical commercial concentrations of ammonium polysulphide are about 50% (NH4)2Sx, and these solutions should be diluted at least 10 times with the co-injected water. Increased water rates should improve the mass transfer of HCN to the water phase. As outlined at the beginning of this section, the injected polysulphide solution is a “water wash” application, and the injection should follow the “water wash” guidelines. In certain cases of well designed spray systems integrated with heat exchangers (e.g., vertical downflow heat exchangers with polysulphide spray above the tube sheet), the criteria of 50% remaining as liquid has been relaxed to 25%. If this criterion is relaxed, it is very important to check pH of the resulting water phase to avoid sulphur precipitation, since pH decreases more as more water is evaporated. Also, inhibition with polysulphide has to be continuous. Attempts at controlling corrosion in a 35% ammonium hydrosulphide system at pH 9 by batch inhibition failed in laboratory experiments.
3.6
INJECTION POINT METALLURGY
Process piping at the injection point should normally be alloyed (solid or clad on carbon steel) with Alloy 20, Alloy 825, or possibly type 316L stainless steel in low chloride environments, for a distance of at least five pipe diameters upstream and ten pipe diameters downstream from the injection point (Figure 4), or 1.5m [five feet] upstream and 3m [ten feet] downstream for large diameter pipe (whichever leads to a shorter alloyed section). Alloy 625 solid, weld overlay or clad, or Alloy C-276, are also acceptable. The alloyed section should be flanged in for ease of inspection and maintenance. There may be occasions where guidelines on ammonium hydrosulphide corrosion would allow the use of carbon steel for piping around the injection point. These situations should be evaluated on a case by case basis. 14
CONFIDENTIAL
The exception to the above injection point metallurgy recommendations is polysulphide addition to Sour Water Stripper feed streams, where the polysulphide is injected into a pipe that is liquid-full with sour water at moderate temperature (see Sour Water Stripper section).
3.7
CHOICE OF INHIBITOR CATION
Ammonium and sodium polysulphides are commercially available. Ammonium polysulphide is preferred over sodium polysulphide in most instances. It has an intense yellow colour even at high dilution which makes its presence easily recognizable, whereas sodium polysulphide is comparatively colourless. The sodium cation may have a negative impact on downstream sour water stripper operations because it is not strippable and may require acidification of the feed to the unit in order to neutralize the higher pH water. Ammonium polysulphide is typically used in CCUs. Sodium in the sour water has a negative impact on all cat cracking catalysts, via sour water recycled to CCU risers. However, sodium polysulphide has been used successfully at a Shell (site 1) Refinery Sour Water Stripper and a Shell (site 2) Refinery Coal Gasification unit. For equal active sulphur, no differences in cyanide removal or reactivity were seen. These are special cases where sodium is not a problem in the effluent (site 1 Refinery), or special desire to minimize ammonia to the SWS (Coal Gasification). See further remarks in the SWS section of this document. Sodium polysulphide is also used at the Shell site 2 Refinery Flexicoker in the Venturi Scrubber; this is also due to a desire to minimize solid ammonium salt deposits (see Appendix E).
3.8
CO-INJECTED WATER QUALITY
An alkaline sour water of pH 8.5-9.0 (e.g., cat cracker main fractionator overhead separator water) should normally be used to mix commercial polysulphide solution. Main fractionator overhead water is usually also the water source most advantaged economically. Temperature of the water at the mix point should be between 10 and 54°C [50 and 130°F] (see Section 1.3). Concentrated ammonium polysulphide is soluble in alkaline sour water solutions. Water containing magnesium or calcium is not advised because at pH > 8, Mg and Ca salts will precipitate. Ammonium polysulphide is unstable in water if the pH is lowered to 7, at which point it starts to decompose into ammonia, hydrogen sulphide and elemental sulphur. Use of condensate and highly stripped sour water is not acceptable, even if the pH is within guidelines (8 to 10). Some NH4+ and HS- is necessary to stabilize the polysulphide and avoid precipitating sulphur. In limited scouting experiments at SMRC in 1997 for polysulphide use at the Delayed Coker, ammonium polysulphide was mixed with SWS#7 bottoms (typical pH is 8.2-8.4, with low ppm levels of ammonium and hydrosulphide) , or condensate Elemental sulphur precipitated on mixing the polysulphide with either condensate or SWS#7 bottoms. To estimate the concentration of NH4HS required to hold polysulphide in solution, various dilutions of Delayed Coker MF boot waters were mixed with APS. Results indicated that the lower stability limit of commercial polysulphide in that sour water system was about 500 ppmw (0.05%w) NH4HS. For new or revised systems the dilution water is recommended to be sour water with at least 0.1%w NH4HS, since the individual concentration requirements of NH4+ and HS-, and effect of bicarbonate, etc., are not known at this time.
15
CONFIDENTIAL
3.9
INJECTION SYSTEM CONFIGURATION
In vapour phase injection applications, polysulphide solutions ideally should be injected countercurrent into the centre of horizontal process lines through a full cone spray nozzle to provide good contact with the process stream. Injection into vertical downflow piping should be avoided. Injection location should be at least 5 pipe diameters downstream, and at least 10 pipe diameters upstream from a change in direction at tees or ells (see Figure 4). Acceptable materials for spray nozzles are Alloy 20, Alloy 825, Alloy C-276, or type 316L stainless steel in low chloride environments. A spray nozzle design having a large open area is preferred to reduce the chance of plugging. It is desirable to make spray nozzles retractable on-line as a precaution should plugging be a problem. Strainers which are cleanable on-line without shutting down the polysulphide solution flow are recommended if spray nozzles are not retractable. Strainer holes should be about 1/3 the minimum opening in the associated spray nozzles. Strainers are recommended but optional for retractable nozzles. The method of contacting (co-current or counter-current) and the pressure drop taken over the nozzle depend on the orientation of the equipment to be protected and the particular alloy and isometrics of the receiving line. For instance, co-current injection of polysulphide solution and a moderate pressure drop (about 2.7 bar [40 psi]) is acceptable if the purpose is entirely to protect the immediate downstream heat exchanger, and if the spray nozzle can be directed such that it sprays directly onto the tube sheet. For counter-current injection, a significant pressure drop (2.7 to 4.1 bar [40 to 60 psi]) and perhaps multiple nozzles are recommended to maximize mass transfer of hydrogen cyanide from the gas to the water phase, or if small drops are needed to enhance equal distribution. Polysulphide injections before an air cooler with multiple banks, or parallel heat exchangers, present problems of water distribution similar to the situation in water washing of hydroprocessing air coolers and multiple exchangers. Air cooler inlet piping should be symmetrical and horizontal lines level if the injected polysulphide solution is to be equally divided amongst several banks. In critical applications, it may be necessary to provide individual polysulphide injections to each air cooler bank or to each heat exchanger. Additional instrumentation will normally be needed to ensure proper distribution and control.
3.10
LIMITS ON THIOCYANATE CONCENTRATION
Cyanides which are dissolved in sour water are converted to thiocyanate (SCN-) by addition of polysulphide. Thiocyanate is corrosive to carbon steel above about 1.5% (15,000 ppmw). It is usually present below 500 ppm in the low pressure section of the CCU gas plant and therefore is not a problem there. Fouling and corrosion by thiocyanate is most likely in the high pressure section of the gas plant, where it can concentrate in the debutanizer bottoms, especially in the reboiler. In the event of excessive amounts of thiocyanate, the aqueous solution should be purged to keep its concentration below 1.5%.
4.0
POLYSULPHIDE USE IN CATALYTIC CRACKING UNITS
Polysulphide is normally used for equipment protection and cyanide scrubbing in cat cracking gas plants. The cat cracking process typically generates 0.9 – 4.5 kg [2 - 10 pounds] of cyanide per thousand barrels of feed. Cyanide production increases with reactor temperature, conversion level and the concentration of nitrogen compounds (cyanide precursors) in the feed. 16
CONFIDENTIAL
Polysulphide addition to cat cracking units follows the general principles described earlier. Special attention needs to be placed on the best polysulphide addition location as regards temperature and pH, realizing that the pH can be affected by absorbing vapour components (e.g. NH3, H2S, CO2) at different temperatures and pressures. Figure 5 is a typical polysulphide injection system in a CCU gas plant. Every cat cracker is different, and polysulphide systems should be evaluated for each particular CCU according to the principles discussed earlier and further illustrated in this section. Examples for Shell site 1 Refinery CCU are in Appendix C, and for the Shell site 2 Refinery CCU in Appendix D. This section is divided into 4 parts to address polysulphide use in the following areas of a CCU: •
Main Fractionator Overhead
•
Interstage of the Wet Gas Compressor (WGC)
•
Afterstage of the WGC
•
Rectified Absorber (RA) Column
4.1
MAIN FRACTIONATOR OVERHEAD
Polysulphide addition to the main fractionator overhead should be done with care, keeping in mind the principles of temperature and pH discussed earlier. Even when the overhead temperature is about 110°C [230°F] before injection, the pH can decrease when even a moderate percentage (greater than approximately 20%) of the wash water vaporizes, causing elemental sulphur to drop out. Further details and an example from Shell site 1 Refinery are discussed in Appendix C. Because of typical main fractionator vapour temperatures, it is recommended that a water wash, with no polysulphide, be done immediately upstream of the exchanger where the dewpoint is expected. Usually the corrosive species of concern are ammonium chloride, NH4Cl, and hydrogen chloride, HCl, at dew point conditions. A small portion of the HCN in the main fractionator overhead vapour will be absorbed in this alkaline wash water at low pressure. At higher pH of wash water, more HCN will be absorbed. CCU main fractionator overhead sour water is a good source of wash water. It is low in sodium and calcium. This water should normally be used to wash the main fractionator overhead condensers, the WGC interstage condensers, and the WGC high pressure condensers. Also, main fractionator overhead sour water is often injected into the CCU riser reactor, where it is used as a tool to heat balance the unit. Sodium poisons the cat cracking catalyst, so sodium polysulphide should not be used as the source of polysulphide at the main fractionator wash step, when main fractionator water is used to heat balance the reactor. The recycle of sour water to the riser reactor is common, even though it potentially causes an increase in corrosive species (e.g., the NH4+ and Cl- in sour water can add to NH4Cl deposition in the main fractionator overhead). This potential increase in corrosive species should be kept in mind when sour water recycle is utilized. Note that if the main fractionator wash does not contain polysulphide, then sour water recycle to the riser will allow HCN to be recycled, raising HCN concentration in the wet gas.
17
CONFIDENTIAL
4.2
INTERSTAGE OF THE WGC
The interstage of the WGC is a good starting point for cyanide scrubbing by polysulphide because the pH is normally 8.5 - 9.0 at this point in the process. The total pressure, and therefore the partial pressure of HCN, is high enough for a significant portion of the HCN to be scrubbed into the water phase. pH is in the range 8.5 to 9.0 due to the preferential absorption of NH3 , as compared to CO2 and H2S, at this moderate pressure. The ideal total wash water to this stage is about 11 to 22 litres [3 to 6 US gallons] per barrel [1 barrel = 159 litres] of CCU feed. Another reason why interstage polysulphide addition is preferred is that some HCN is soluble in the interstage condensate hydrocarbon phase; non-dissociated HCN condenses along with the C3’s.∗ This is one route HCN gets to the RA column, and then to the debutanizer and depropaniser. Since polysulphide is only active in the water phase, there needs to be good contacting of the dilute polysulphide and hydrocarbon liquid, as well as good contacting of polysulphide with the hydrocarbon vapour. Scrubbing at the interstage allows much of the HCN to be recovered into the polysulphide water phase and reacted to SCN-. If HCN scrubbing is not started at the interstage, the interstage hydrocarbon phase will contain more HCN, which will enter the RA column. Two steps are recommended for maximum cyanide scrubbing: 1.
Polysulphide Wash of Interstage Condensers (Intercoolers)
Assuming the compressor outlet temperature is low enough to avoid sulphur precipitation at the injection point, the condenser should be washed with a combination of main fractionator (MF) overhead water and purchased polysulphide. Otherwise, a sour water wash similar to that for the MF overhead should be used. Sour water with polysulphide should be added to the interstage at a rate that satisfies the usual water wash guidelines, such as at least 50% of the injected water remaining as liquid (see Section 3). The water rate is typically 1.1 to 3.8 litres [0.3 - 1.0 US gallons] per barrel [1 barrel = 159 litres] of CCU feed. Polysulphide should be added to the makeup water to hold an excess polysulphide concentration of 15 to 25 ppm active polysulphide in the water sump of the interstage separator. The ratio of net makeup/purge water should be adjusted to keep the ammonium hydrosulphide concentration below 2% so that carbon steel equipment can be utilized. Also, the purge should be such that it prevents excessive ionic build up of SCN- (> 1.5%w) and ammonium hydrosulphide (ammonium bisulphide, NH4HS). The cyanide in the water phase will react with polysulphide to form SCN-. This water can be reused to scrub additional cyanide from the vapour phase, provided the cyanide ion has been sufficiently converted to SCN-. The water can be reused for additional cyanide scrubbing because the SCN- will not revert to HCN, as discussed earlier. This leads to the second recommended step.
∗ Since HCN tends to fractionate with the C2’s and C3’s, poor C3 recovery in a rectified absorber will result in less cyanide going with the liquid hydrocarbons and more remaining in the dry gas. 18
CONFIDENTIAL
2.
Separator Water as an Additional Vapour Scrub
In normal situations with low excess polysulphide, a large volume of recirculated separator water should be injected into the vapour stream before the separator, to scrub out additional cyanides. The water rate is the most important variable, because the absorption of HCN into the liquid polysulphide solution (Step 1 in Figure 3) is not influenced by the concentration of polysulphide when low excess polysulphide is present. This large volume of water is not added before the condensers for two reasons. First, the high volume of water would cool the hot inlet temperature of the exchangers, which hurts performance of the exchangers. Second, the HCN scrubbing is favoured at the lower outlet temperature. Therefore, polysulphide spray nozzles should be located at the outlet of the exchangers. Water sump residence time should be 30 - 60 minutes to minimize the amount of polysulphide needed, by supplying a significant amount of time for the cyanide to react to SCN-. The water residence time also provides for good phase separation if the system is well designed, and also provides some buffer for HCN production swings. The relatively long residence time is necessary because the majority of the S5 and S4 polysulphide injected at the condensers likely will have reacted, leaving only the lower reactivity S3 and S2 species. If the excess polysulphide concentrations are increased in the first step, it is possible to do a good job of scrubbing with less recycle wash water and less sump residence time. The negative side of high polysulphide levels in the purge water is possible deposition of sulphur in the sour water stripper system. (See further discussion in the Sour Water Stripper Section). Therefore, it is not good practice to purge large excesses of polysulphide to the sour water stripper in an effort to remove cyanide using once through water washing. Any polysulphide use at the interstage of the WGC requires the separator to do a very good job of demisting the exiting gases to the second stage of the WGC. Polysulphide solution becomes very corrosive as it evaporates inside a compressor. Shell has experienced significant corrosion in the second stage of four WGCs due to water entrainment, over the past several years
4.3
AFTERSTAGE OF THE WGC
A polysulphide wash following the second stage of the WGC should be utilized for added scrubbing of the wet gas, prior to sending the gas to the RA column. The higher pressure at the afterstage aids the removal of HCN from the wet gas, which raises scrubbing efficiency, while the lower pH of the water at this stage reduces undesirable HCN ionization, which lowers scrubbing efficiency. The combined impact is a net benefit for scrubbing at the afterstage. About half the wash water rate at the afterstage as compared to the interstage is necessary to achieve the same fraction of HCN removal from the wet gas. For maximum cyanide scrubbing, it is recommended to have an afterstage polysulphide scrubbing configuration similar to the interstage design discussed earlier; see Figure 5. As in the interstage, the quantity of makeup water for washing should be chosen so as to keep the ammonium hydrosulphide concentration below 2%w. As an example of combined efficiency, if the interstage wash is 60% effective at removing HCN from the wet gas, and the afterstage wash is 60% effective at removing the remaining HCN, the combined HCN removal is 84%. Using multiple scrubbing points provides the same benefits to HCN removal as theoretical stages in a column. 19
CONFIDENTIAL
pH, Pressure, and Addition of Base The pH of the afterstage separator water is normally 7.5 to 7.8, as compared to the interstage separator water which typically has 8.5 - 9.0 pH. The lower pH is due to the increased absorption of acid gases (CO2 and H2S) in the afterstage condenser, due to the higher operating pressure at the afterstage, and the fact that the majority of the NH3 has been scrubbed in the main fractionator and WGC interstage. The effect on pH of CO2, H2S and NH3 absorption at different pressures has been validated by equilibrium modelling and field testing. Addition of ammonia or caustic to raise the pH is not recommended. This will only result in increased scrubbing of CO2 and H2S from the wet gas, with only a slight increase in pH, because separator water is highly buffered by carbonates and hydrosulphides. The increased NH4HS concentration results in higher corrosion rates, since concentration increases greatly overshadow the slight increase in pH. An added problem is that carbonate salts may drop out of solution as the concentrations are increased, which results in system plugging problems.
4.4
RECTIFIED ABSORBER (RA) COLUMN
In the gas plant, a water (condensate) wash should be used to prevent salt precipitation in the rectified absorber column. Shell site 1 Refinery CCU Gas Recovery RA and Debutanizer column configuration is shown in Figure C2, in Appendix C, and is typical of most RA and Debutanizer systems. No polysulphide should normally be added to the wash, for several reasons: First, there is concern that polysulphide solution getting into the reboil area of the RA column would result in the formation of elemental sulphur, leading to high corrosion rates and fouling. Second, polysulphide injected in the RA column will result in excess NH3 leaving with the dry gas, and being scrubbed out in the DEA absorber. When the DEA from the absorber is stripped in the DEA regenerator, the NH3 concentrates in the overhead condenser area, and can result in high corrosion rates due to the formation of ammonium hydrosulphide. NH3 is also an additional process load on a Claus sulphur plant. Third, any polysulphide carryover to an amine treater will result in degradation of the amine due to the possible formation of glycolates, acetates or oxylates. However, if there is good water removal (and therefore residual polysulphide removal) in the RA column, the addition of polysulphide to the wash water may be justified. This should be considered on a case-by-case basis. The condensate wash water should be oxygen free and of sufficient quantity to dilute the corrosive species, which are normally ammonium hydrosulphide and ammonium carbonate. Monitoring ammonium ferrous cyanide content in the wash water removed at the RA column water draw tray should be done. If ammonium ferrous cyanide is detected, it may indicate periods of high corrosion, probably further upstream in the process (e.g., at the WGC). RA column condensate wash rates may need to be higher, or more polysulphide may need to be added at the interstage or afterstage of the WGC.
20
CONFIDENTIAL
The condensate wash should normally be added to the cool absorbing liquid stream before it enters the top of the column (see Figure C2), and flow should be continuous. The water rate should be sufficient to dilute corrosive salts, and is typically 0.75 to 2.27 litres [0.2 to 0.6 US gallons] per barrel [1 barrel = 159 litres] of CCU feed. This rate will vary depending on the amount of entrained water in the hydrocarbon feed to the RA column. Entrained water is indicative of poor phase separation in the main fractionator overhead accumulators and WGC separators. The wash water rate usually is found by experience based on performance of an individual unit. Hydrogen probe activity and reboiler fouling should be monitored to determine this rate. HCN Partitioning in the RA and Debutanizer Product Streams, and Importance of Upstream Treatment for Cyanides Hydrogen cyanide is not entirely removed by polysulphide scrubbing in the WGC train, and some cyanide should be expected in the RA column. Hydrogen cyanide will tend to be recovered with the C2 and C3 streams (vapours from RA column, debutanizer overhead and depropaniser overhead).∗ HCN going with the C2 and C3 streams can lead to fouling and corrosion in the debutanizer and depropaniser systems. The C2 and C3 streams are typically treated with DEA to recover H2S. Cyanide can react with the DEA, forming corrosive salts (usually referred to as heat stable salts in amine treating terminology). These salts are normally acidic, and can be neutralized with caustic or removed from the DEA system by purging DEA and adding fresh DEA, or by bringing in a reclaiming company. Both options are expensive. Corrosion Concerns in the Debutanizer Column due to Inadequate Water Removal in the RA Column When designing the water removal system for the RA column, it is important to include sufficient trays between the reboiler and the water draw to remove all the water from the feed to the debutanizer. High corrosion rates in the overhead of the debutanizer occur when that system is allowed to operate wet. The presence of cyanide may make the corrosion problem worse. If the system can be kept dry, no corrosion will occur. If a water phase is present in the overhead of the debutanizer, it is important to keep the water draw system unplugged and actually to remove the water. If this water phase is allowed to reflux back into the column and/or go to the depropaniser, the problems compound and become much worse. For example, plugging of the Shell site 1 Refinery Gas Recovery depropaniser has occurred when the Gas Recovery debutanizer water phase has gone to the depropaniser.
5.0
POLYSULPHIDE USE IN COKING
The coking process produces cyanides. Polysulphide should typically be used in the gas compression train of coking units, similar to the use in cat cracking units. Polysulphide is used at the Shell site 2 Refinery Flexicoker. Although the flexicoking process is unique, it is useful to summarize the special polysulphide flexicoking applications. See Appendix E. The polysulphide treatment of flexigas may have future applications to coal or coke gasification.
∗ HCN in hydrocarbon liquid behaves as a non-ideal solution; if the systems showed ideal behavior, HCN would be recovered with the C4 and C5 streams. Actual values measured at Shell site 2 Refinery indicate that the HCN partitions at about the same place as H2S, which distils into the C2-C3 fraction. 21
CONFIDENTIAL
6.0
POLYSULPHIDE USE IN SOUR WATER STRIPPERS
Polysulphide addition to sour water strippers is a special case of polysulphide addition. Generally, principles of polysulphide addition described earlier apply; however, there are some special considerations.
6.1
INTRODUCTION
Polysulphide is used to control corrosion, blistering, and wet H2S cracking in sour water systems either by eliminating cyanides or protecting the metal surface. Also, cyanide limits may be placed on the discharge of SWSs for environmental reasons. In this case, reaction of cyanides to thiocyanates is desired. Thiocyanates are treatable in biotreaters, whereas cyanides are not. Whenever possible, cyanides should be controlled at the source, rather than at the sour water stripper. Polysulphide should be added at the SWS only if necessary to control cyanides that are not effectively treated upstream, or as a backup measure. Usual sources for cyanide are cat crackers and cokers. The main corrosion problem area is the sour water stripper overhead air coolers or water coolers. High concentrations of ammonium hydrosulphide are usually present (nearly 40%w NH4HS has been measured at Shell site 2 Refinery), due to the nature of sour water stripping. Corrosion even on alloy materials is potentially high in the SWS overhead condensing and reflux system because of high ammonium hydrosulphide concentrations and high velocity. Cyanide may act to destabilize weak sulphide films on tube surfaces. From experience at Shell site 2 Refinery and Wilmington, 50 ppmv HCN in the SWS overhead vapour can aggravate problems in 316L and Alloy 800 air coolers at high ammonium hydrosulphide concentrations, and polysulphide has been used successfully to mitigate the corrosion. Alloys such as Alloy 825 or Alloy 20 would also be suspect at high concentrations of NH4HS, or high pH, in the presence of cyanides. Without cyanide control, Alloy C-276 is the material of choice, basis success at SCGP-1 Sour Slurry Stripper overhead air cooler. Cyanide in SWS feed water is present as the “free” cyanide ion (CN-), and present as complexed Fe-CN and Ni-CN (Ni-CN especially from coking operations). Any cyanides in the SWS feed will be concentrated in the overhead, typically 10-20 times. There is some evidence of partial decomposition of cyanide complexes in the SWS column, depending on pH and residence time. Residual polysulphide from upstream treating, or polysulphide added directly to the SWS feed, will decompose to elemental sulphur at the high temperatures of the stripper bottoms and also will form elemental sulphur if any region of low pH exists in the SWS column (see Figure 1). This could cause column fouling with elemental sulphur, and will contribute to the biotreater sulphur loading. Typical constraints on residual polysulphide to SWS are 25 ppmw maximum active polysulphide. Appendix F is further details of polysulphide use at Shell site 2 Refinery SWS.
22
CONFIDENTIAL
6.2 GUIDELINES FOR POLYSULPHIDE ADDITION TO SOUR WATER STRIPPERS 1.
Polysulphide should be added to the feed to the SWS.
2.
Allow at least 30 minutes residence time between polysulphide addition and the SWS column entry. Polysulphide addition should be made sufficiently upstream of the SWS surge vessel, so there is as much mixing time as possible in the lines before entering the surge vessel. The rather long residence time is recommended to allow the polysulphide to mix with the water and react with “free” HCN in the feed to form thiocyanates. Normally the SWS feed enters the column at the upper trays, so polysulphide spiked feed has a possibility of reacting with any HCN liberated from complexed cyanides further down the column. Polysulphide addition will not remove all complexed cyanides, however.
3.
Maximum temperature in the SWS feed line should be 110°C [230°F]. Above 110°C [230°F], polysulphides slowly decompose to thiosulphates, and molten sulphur will deposit. See Figure 1.
4.
For maximum polysulphide effectiveness, pH of SWS feed should preferably be 9-10, with pH 8 as a lower limit. Below pH of about 7, polysulphide decomposes into elemental sulphur. See Figure 1. The higher pH also means more HCN is ionized to CN-. Open literature reaction kinetic studies with cyanide ion in solution indicate the reaction rate is enhanced at higher pH (Reference 7).
5.
Diversion of a portion of the SWS feed, whether polysulphide treated or not, as a direct wash of the SWS overhead condenser, may be necessary as further protection of the overhead condenser. A separate wash of the overhead condenser can also be used. If diversion of a portion of feed is used, recycling the condensed accumulator water back to the feed should be considered. If the overhead condenser is washed, 50% of injected water should remain as liquid after equilibrium flash, and efforts should be made to ensure that the water gets evenly distributed to the tubes of the condenser or air cooler. Also, the other vapour phase injection guidelines apply, such as an alloy section around the injection points (if the water is polysulphide spiked), and use of a spray nozzle.
6.
If polysulphide spiked water is added to the overhead condenser, polysulphide addition rate should be such that there is a slight excess of active polysulphide (HCN conversion activity) in the overhead accumulator water. Samples of the accumulator water should have a residual polysulphide of about 15 ppm active polysulphide sulphur, and should be light yellow in colour. Higher excess active polysulphide sulphur in accumulator water runs the risk of increasing column sulphur load and aggravating sulphur precipitation at lower pHs.
7.
Excess polysulphide addition to SWS feed should be avoided, because of the risk of precipitating elemental sulphur that can cause fouling in the bottom of the SWS column, its reboiler, and other associated equipment and piping. Also, although sulphur is a nutrient for biotreating, the bugs cannot tolerate wide swings in sulphur in biotreatment feed water. If necessary consult location effluent treating experts as to the sulphur limitations on biotreating.
23
CONFIDENTIAL
8.
Sodium polysulphide and ammonium polysulphide are equally effective for the same active polysulphide sulphur to cyanide ratio.∗ Ammonium polysulphide is preferred, because its yellow colour makes any excess readily identifiable. However, sodium polysulphide is simpler to handle (less volatile and less susceptible to freezing than non-morpholine treated ammonium polysulphide). With sodium polysulphide, the impact of the extra sodium load on sour water stripping should be considered. There may be sodium restrictions on SWS effluent. There are circumstances where the additional dissociated ion load (of Na+) with sodium polysulphide needs to be avoided; e.g., high cycle re-used sour water systems, such as Shell site 2 Refinery SWS#6. Also, if ammonium polysulphide is used for CCU or coker cyanide treating, it is probably not cost effective to have a separate handling system for sodium polysulphide.
9.
A tee from the top or side of the line is preferred for polysulphide addition to SWS feed, if the line is liquid full. An available bleeder at the bottom of the line for temporary use is also acceptable. An injection quill, discharging polysulphide in the centre of the pipe in the bulk flow direction, can also be used, if polysulphide systems are kept clean and free of particulates. The injection quill should be retractable, in case of plugging (by, e.g., sulphur). With a quill injection, polysulphide addition rate should be such that injected stream velocity is 1-2 times the free stream velocity. Minimum discharge hole size in the quill is 6mm [1/4”] and there should be an upstream strainer, with hole size approximately 1/3 the size of the quill opening. Carbon steel is acceptable metallurgy for these systems. Stainless steel is also acceptable and should be used for the quill. It is not necessary to pre-dilute the commercial polysulphide if injected into sour water, although pre-dilution may help dispersion of the active sulphur when contacted with sour water. Pre-dilution may be preferred for ease in controlling addition rate, and for occasions where there is not much mixing time available before a surge vessel.
6.3 SPECIAL CONSIDERATIONS ON MONITORING IN SOUR WATER STRIPPER SYSTEMS Since cyanide is concentrated in SWS overhead vapours, particular attention should be paid to regular inspection of overhead piping and the overhead condenser. Experience indicates that the overhead condenser (usually an air cooler) is the first piece of equipment to show signs of cyanide-enhanced ammonium hydrosulphide corrosion, probably because of the thin wall tubes and high NH4HS concentration in the condensed water. pH and soluble Fe in the overhead accumulator water should be monitored routinely as an indicator of corrosive attack. Cyanides in the feed, bottoms, and overhead accumulator water should also be monitored. The amount of excess (active) polysulphide in accumulator water should be monitored, as discussed above.
7.0
MONITORING POLYSULPHIDE EFFECTIVENESS
Two monitoring schemes are recommended; chemical analysis of the sour water and direct measurement of hydrogen activity. The two methods are discussed below. Typical sour water sampling locations and typical locations for hydrogen probes in a CCU gas plant are indicated in Figure 6.
∗ This is based on results at SCGP-1 at Shell site 2 Refinery, and experience at Shell site 1 Refinery SWS#4. 24
CONFIDENTIAL
7.1
SOUR WATER ANALYSES
7.1.1
BASE-SOLUBLE IRON
With a polysulphide treatment program in place at a location, routine checks for residual cyanide activity are recommended, coupled with occasional analysis by qualified staff to perform a detailed cyanide profile of the particular unit. The recommended routine check for residual cyanide activity is measurement of “basesoluble iron”. Appendix A is a typical procedure for determining base-soluble iron. Each location should adjust the procedure according to their particular needs and QA protocol. Base soluble iron has been shown to be free from interferences, and to give a reliable analysis of residual cyanide activity. The observed iron is proportional to active cyanide that is untreated by the active polysulphide sulphur present in the unit. Briefly, the sour water sample is adjusted to a high pH, pH >13, by addition of NaOH pellets. The only iron species that are soluble in a sample at that high a pH are complexes of iron with cyanide. The other iron species, e.g., the iron sulphides and iron chlorides, form insoluble FeOH dominated complexes. The pH adjusted sample is then filtered, and total iron determined with Inductively Coupled Plasma or Atomic Absorption. With ICP or AA, it is possible to determine both Fe and Ni, and deduce both Fe and Ni complexed cyanides. Nickel cyanides are usually not important except in coking operations.∗ The pH needs to be adjusted within two hours after the sour water sample is taken, as contact of sour water with air will form polysulphides and give cyanide readings that are too low. Alternatively, NaOH pellets (approximately 5 pellets per litre of sample) can be placed in the bottle before taking the sample, and then the pH checked within two hours. If the pH adjusted sample is stored for a period of time before analysis, it should be refrigerated to minimize decomposition. Base-soluble iron should be determined weekly. For special circumstances (e.g. effluent treating control) the base soluble iron may need to be determined daily. When high levels of base soluble iron are detected, checks should be made of the metering pump for polysulphide, water dilution system, flow measurement, and the quality of the polysulphide itself, whether purchased or generated in-situ. If all appears to be in order and the base soluble iron is high, “non-routine” cyanide analyses may be justified, either by Shell Global Solutions or plant analytical staff. These analyses are briefly discussed below (see 7.1.7). The level of base soluble iron that is cause for concern varies from unit to unit, but numbers in the range of 10 ppmw Fe are typical “high” values.
∗ The chemistry of what happens is believed to be the following. Sour water systems contain phenolics, SCN-, OH- (and in some cases Cl-, formate, etc.), ligands which will coordinate with Fe. These are not as strong a ligand as CN-, but when CN- is present at <1 ppm and the other coordination species are at 100+ ppm, the law of mass action creates mixed complexes. The FeCN dominated complexes (greater than 2 CN- groups in the complex) are soluble due to the nature and strength of the FeCN bond. When a Fe complex is in solution with a very high excess of OH-, the OH- will displace most ligands quickly, and form a cyanide/OH- mixed complex which is soluble. The insoluble, non-cyanide, compounds such as Fe(OH)2 form precipitates or flocculates and are removed in the filtration step. The strength of the FeCN bond slows the displacement by the OH-, but CN will eventually be displaced also; after several days at pH>13, the iron cyanides will precipitate as Fe(OH)2. The high pH also holds the CN- ion in solution, supporting its equilibrium. 25
CONFIDENTIAL
7.1.2
COLOUR TEST FOR RESIDUAL POLYSULPHIDE
This is a quick test for residual polysulphide that can easily be done in the field. It involves visually matching the colour of a sour water sample with the colour of standard solutions. The colour of polysulphide solutions is only stable for 20-40 hours, so polysulphide solutions cannot be used as standards. Instead, various concentrations of potassium dichromate, K2Cr2O7, which has been found to have the same colour range as that of polysulphide solutions, are used in the visual comparison. Potassium dichromate solutions are stable in colour for much longer periods of time, on the order of months. Carefully made up polysulphide solutions have to be “calibrated” versus the potassium dichromate solutions, but this is only an occasional effort. The sour water sample, once taken, must be compared within 5 minutes with the standard solutions because of colour degradation of the sour water (contact of sour water with air will form polysulphides, and the colour will change). According to a paper by Hickson Kerley, Inc (Reference 9), a faint yellow colour indicates around 10 ppm (NH4)2Sx. A dark yellow colour may be 80 ppm or greater. The system is well controlled when the draw water has a faint yellow colour, with only a possible haze from hydrocarbons present. Gray or black draw water indicates the polysulphide wash system is ineffective, and may be indicative of the presence of iron cyanide. The colour test is recommended as a daily or once/shift check for the presence of active polysulphide. It should be considered as optional (location preference) if base soluble iron tests are being performed daily. The colour test is not designed, nor should be used, as a quantitative measure of active polysulphide. It should be use as a “relative” test to indicate whether the commercial polysulphide is still going in at about its intended rate, and has not been decomposed by temperature. Turbidity of the sample should be used as an indication of temperature decomposition (turbidity may be due to sulphur precipitates). With insitu generated polysulphide, yellow colour can be indicative of the less reactive S2 and S3 species, but with commercial polysulphide solutions, the species distribution is not as much a factor.
7.1.3
IRON CYANIDE “SPOT TEST”
The spot test involves placing a small volume of sour water sample on filter paper, and adding acidic ferric chloride solution, which causes ferric thiocyanate and ferrous/ferric ferrocyanide to form. Upon addition of distilled water, the thiocyanate in the sample then migrates beyond the blue precipitated ferrous/ferric cyanide spot, forming a reddish brown ferric thiocyanate ring. This test is not recommended because it is very subjective and has been shown to give false positives and false negatives. A major problem with this test is that the sour water sample already contains FeCN and thiocyanate (from any reaction of polysulphide with CN-). As the test looks for the thiocyanate that is created, the test can be very misleading.
7.1.4
TOTAL SOLUBLE IRON (UNFILTERED)
Measurement of total soluble Fe in an unadjusted-pH, unfiltered sample of sour water may be an indication of gross corrosive attack on steel, but can be misleading as an indication of polysulphide activity. Total Fe is quite variable, in part due to the irregular sluffing off of iron sulphide scale. Its use is considered optional (location preference).
26
CONFIDENTIAL
7.1.5
pH
pH of the sour water should be between pH 8 and pH 10 for optimum polysulphide effectiveness, as discussed earlier. pH is also a prime variable in corrosive attack by ammonium hydrosulphide. However, CCU and coker accumulator waters are often highly buffered by NH4+, S--, HS-, CO3--, and often the pH in these waters cannot readily be changed. pH can be affected by process changes which may cause unwanted effects. For instance, a recycle stream high in NH3 can add enough ammonia that carbonate salt precipitation and fouling can occur, due to ammonium carbonate [(NH4)2CO3] or ammonium carbamate [NH4CO2NH2] formation. pH measurements should be taken at least weekly at appropriate points in the system. Above 1 barg [15 psig] these measurements should be taken “closed cup” (see Section 3.2).
7.1.6
CHLORIDE, AMMONIA, SULPHUR
These measurements, although very useful for other purposes, are not specific for polysulphide activity.
7.1.7
NON-ROUTINE CYANIDE ANALYSES
These are established analytical procedures which are used by Shell Global Solutions staff when doing a cyanide profile of a unit. In addition, location analytical staff may have the capability of doing some or all of them. Some steps involve pretreatment of solutions with cadmium, which is a highly toxic material. Detailed procedures are available from Shell Global Solutions. In summary they are: 1.
Measurement of total cyanides displaced by EDTA. HCN is stripped from a buffered, acidified sample by distillation and N2 sparging. Digestion and stripping with EDTA (and no oxygen) in a buffered solution of pH 4.5 displaces cyanides off the metal complexes at conditions that avoid conversion of SCN- to cyanides. If only a very low pH (say about pH 2) without EDTA is used, the thiocyanate would be converted to cyanides. The stripped HCN gas is absorbed in a scrubber containing ammonium polysulphide solution, where free CN- is converted to SCN-. When scrubbing is completed, the polysulphide solution is treated with cadmium carbonate to remove excess polysulphide sulphur and hydrosulphides from solution, which would otherwise interfere with thiocyanate (SCN-) analysis. Finally, the treated solution is analyzed for thiocyanate by ion chromatography.
2.
WAD (Weak Acid Dissociable) cyanides. Similar to above, but uses only slightly acidified (pH 4.5) sample without EDTA. This method does not quantitatively recover CN- from strong metal complexes.
7.2
DIRECT MONITORING OF HYDROGEN ACTIVITY
As discussed earlier, atomic hydrogen generated by the corrosion process can penetrate the metal, and combine to form molecular hydrogen in discontinuities within the metal (blisters). The permeation rate of atomic hydrogen can be monitored directly using hydrogen probes. There are two types of hydrogen detectors, “insertion” probes, and “patch” probes. Location corrosion specialists should be consulted for assistance in selecting, locating, and data interpretation of hydrogen probes.
27
CONFIDENTIAL
7.2.1
INSERTION PROBES
Insertion probes measure the flux of atomic hydrogen into a carbon steel probe inserted into the process, and are used to assess hydrogen permeation in steel. The concept is sketched in Figure 7. Atomic hydrogen, from the corrosion reaction of Fe with HS[equation 4], diffuses through the thin steel tube of the probe, which is of known surface area. At the inner surface of the steel tube, the atomic hydrogen recombines into molecular hydrogen, and accumulates in the sealed probe cavity of known volume. The pressure of molecular hydrogen in the sealed cavity is measured by a pressure gauge connected to the probe cavity. The rate of pressure increase is directly related to the hydrogen flux (usually expressed as cm3/in2-day [1 in2 = 6.45 cm2]) passing through the thin steel tube. Because a very small leak from the probe assembly may be interpreted as absence of hydrogen activity, a probe should always be slightly pressurized before use, i.e., with nitrogen to about 0.34 barg [5 psig]. After hydrogen build-up, the probe should not be depressurized below 0.34 barg [5 psig]. Any constant decrease in probe pressure indicates a leak. Basis considerable studies in the 1950’s (mostly by Shell), the following guidelines were developed to correlate hydrogen flux with hydrogen blistering/cracking damage of conventional pressure vessel steels: a)
A hydrogen flux less than 0.025 cm3/(in2 day) is acceptable long term.
b)
A hydrogen flux greater than 0.025 cm3/(in2 day), but less than 0.125 cm3/(in2 day), is acceptable short term, but can result in some blistering/cracking with long term exposure.
c)
A hydrogen flux greater than 0.125 cm3/(in2 day) is cause for immediate concern, as significant hydrogen blistering/cracking could result within a few months.
7.2.1.1 SHELL PROBE Shell has for many years successfully used a calibrated probe developed in the 1950’s, of a particular surface area and volume, herein called the “Shell probe” (see Figure 8). The following guidelines pertain to the Shell probe: a)
Pressure increases less than 0.14 bar [2 psi] per week on the Shell probe are acceptable long term.
b)
Pressure increases of 0.14 to 0.7 bar [2 to 10 psi] per week on the Shell probe are acceptable short term, but can result in some blistering/cracking with long term exposure.
c)
Pressure increases greater than 0.7 bar [10 psi] per week on the Shell probe are cause for immediate concern, as significant hydrogen blistering/cracking could result within a few months.
7.2.1.2 OTHER INSERTION PROBES Other insertion probes are available from several manufacturers, more information can be obtained from Shell Global Solutions. Manufacturer’s recommendations should be followed for probes from other sources. The allowable pressure increase rate for other manufacturer’s probes will vary depending on the probe surface area and interior volume. A probe-specific calibration chart should be developed that relates that probe’s pressure increase rate to hydrogen flux. This can then be used to relate that probe’s pressure increase to the hydrogen flux guidelines in 7.2.1, much the same as was done for the “Shell probe” in 7.2.1.1. Figure 9 is a sample calibration chart. It was developed for probe applications in the MMC CFH. Note that it is not a universal calibration chart (except for the “Shell probe”). 28
CONFIDENTIAL
7.2.2
PATCH PROBES
The second type of probe is a patch attached to the outer surface of the vessel where hydrogen penetration is suspected. The patch probe is sealed to the vessel wall, and this sealing is the area in which most problems lie. In contrast to insertion probes, patch probes measure hydrogen flux through actual equipment; however, they respond less quickly than do insertion probes. The patch probes can be placed at specific locations where insertion probes are not feasible. There are three common types of patch probes electrochemical, pressure, and vacuum.
7.2.2.1 ELECTROCHEMICAL With electrochemical patch probes, the space between the outer vessel surface and the interior of the patch is filled with an electrolyte such as concentrated sulphuric acid or caustic. When atomic hydrogen arrives at the outer surface of the vessel, it reacts electrochemically with the electrolyte. The electrical current resulting from this reaction is measured, and is directly related to the hydrogen flux. This type of probe has been used successfully in laboratory investigations, but Shell has made very limited use of it in processing units. It is cumbersome to handle in the field and requires handling of the concentrated sulphuric acid or caustic.
7.2.2.2 PRESSURE Pressure type patch probes have a contoured patch of steel set and sealed against the outer vessel surface. A pressure gauge is attached to a hole penetrating the patch. As with insertion probes discussed above, the rate of pressure increase in the small cavity is measured. These probes can be very sensitive because the cavity volume is very small, but this creates difficulties in calibrating hydrogen flux.
7.2.2.3 VACUUM Vacuum patch probes are similar to pressure patch probes except that a vacuum is pulled on the cavity between the vessel exterior wall and the metallic patch, and a vacuum gauge instead of a pressure gauge is used. Hydrogen arriving at the outer vessel surface is collected in the small cavity, and the rate of loss of vacuum is measured. Vacuum type probes, as well as pressure type probes, require a perfect seal between the sides of the metallic patch and the vessel wall. This has been a problem in many applications. One vacuum type probe, known as the “Betafoil”, is a thin metallic foil sealed to the vessel at the edges using an epoxy adhesive.
7.2.3
PLACEMENT OF PROBES
In general, placement of hydrogen probes should be where hydrogen damage is expected. Typical placement is shown in Figure 6. Traditionally, most instances of hydrogen damage in CCUs are associated with handling of wet gas (light ends plus H2S plus cyanides) at elevated pressure. This typically includes the aftercoolers and separators following gas compression, rectified absorber column, and overhead condensing systems of the debutanizer and perhaps the depropaniser. In some instances hydrogen damage has occurred in lower pressure sections of the plant, such as the main fractionator overhead condensing systems. With regard to placement of the probes on equipment, probes should be placed where hydrogen damage is most likely or has occurred. Probes should normally be located where there is a separated sour water phase present, such as the bottoms of a condenser shell or a separator vessel (or boot), or where aqueous sour water can condense on vessel walls in contact with the vapour. Probes should not normally be located where single phase liquid hydrocarbon streams exist. 29
CONFIDENTIAL
Placement of probes is very important to avoid misleading indications that the system is under control. An example of this is when inhibitors are added that can protect the sour water phase, but are not volatile and therefore cannot protect vapour areas unless sprayed (or otherwise contact) the vessel walls in the vapour space.
7.2.4
READING OF PROBES
In general, probes should be read and the readings recorded at a frequency of once per week to once per month. The location (or regional) corrosion specialist should establish the frequency of probe reading based on likelihood of hydrogen damage, history of hydrogen damage, process changes, etc. These readings can normally be taken by operators, but may also be taken by inspectors, corrosion engineers, or chemical vendor representatives. Periodic readings should be reviewed by the location (or regional) corrosion specialist for trends.
30
CONFIDENTIAL
References 1.
W.F. Giggenbach, “ Equilibria Involving Polysulphide Ions in Aqueous Sulphide Solutions up to 240o”, “Inorganic Chemistry”, Vol 13, No 7, p 1724-1730 (1974).
2.
Z.A. Foroulis, “Role of Solution pH on Wet H2S Cracking in Hydrocarbon Production”, Corrosion Prevention and Control, 84-89, Aug 1993.
3.
J.B. Sardisco and R.E. Pitts, “Corrosion of Iron in an H2S-CO2-H2O System Composition and Protectiveness of the Sulphide Film as a Function of pH”, Corrosion (NACE), Vol 21, 350-354, November 1965.
4.
NACE Corrosion Data Survey, National Association of Corrosion Engineers, Houston, Texas.
5.
T.J. Edwards, G. Maurer, J. Newman, and J.M. Prausnitz, “Vapour-Liquid Equilibria in Multicomponent Aqueous Solutions of Volatile Weak Electrolytes”, AIChE J, Vol 24 (6), 966-976 (1978).
6.
Luthy GL, Bruce SG, Walters RW, Nakles DV, “Cyanide and Thiocynate in Coal Gasification Wastewaters”, Journal WPCF, Vol 51, No 9, p 2267-2282 (1979)
7.
Luthy GL, Bruce SG, “Kinetics of Reaction of Cyanide and Reduced Sulphur Species in Aqueous Solution”, Environmental Science and Technology, Vol 13, No 12, p 1481-1487 (1979).
8.
J. Gutzeit, “Process Changes for Reducing Pressure Vessel Cracking Caused by Aqueous Sulphide Corrosion”, Materials Performance, pp 60-63, May 1992.
9.
Hickson Kerley, Inc, “How to Use Ammonium Polysulphide to stop corrosion and hydrogen blistering”, unpublished paper, date unknown.
10. S. Licht, G. Hodes, J. Manassen, “Numerical Analysis of Aqueous Polysulphide Solutions and Its Application to Cadmium Chalcognide/Polysulphide Photoelectrochemical Solar Cells”, Inorg Chem, Vol 25, 2486-2489 (1986).
31
CONFIDENTIAL
32
CONFIDENTIAL
33
CONFIDENTIAL
34
CONFIDENTIAL
35
CONFIDENTIAL
36
CONFIDENTIAL
37
CONFIDENTIAL
38
CONFIDENTIAL
39
CONFIDENTIAL
40
CONFIDENTIAL
APPENDIX A.
ANALYTICAL PROCEDURE
BASE SOLUBLE IRON EQUIPMENT: 1) ICP (Inductively Coupled Plasma) instrument or AA (Atomic Absorption) instrument 2) 250 ml [8 US oz.] high density polyethylene bottle 3) disposable beakers 4) whatman 40 filter paper 5) pH paper strips or pH meter 6) .2µ syringe filters 7) disposable funnels 8) disposable 10ml syringes
CHEMICALS: 1) Sodium Hydroxide pellets
PROCEDURE: 1) Collect sample in a high density polyethylene bottle. 2) Pour about 150ml of sample into a disposable beaker. 3) Check pH. Want pH to be >13. 4) Add 3 pellets of Sodium Hydroxide to sample in the beaker. 5) Check pH. Be sure it is >13. 6) Let sit in beaker for 1 hour. 7) Filter into another beaker thru whatman filter paper. 8) Filter approximately 50mls of filtered sample from beaker in #7 above through a .2µ syringe filter into a container that is acceptable for the ICP analysis. 9) Cap and label the sample. 10) Analyze sample by ICP or AA method.
THEORETICAL CN USING METALS CONCENTRATION Theoretical CN = (Base-Soluble Ni ppm)(1.8) + (Base-Soluble Fe ppm)(2.8) basis the assumption of Ni(CN)4 and Fe(CN)6 molar ratios.
41
CONFIDENTIAL
APPENDIX B.
IN-SITU GENERATION OF POLYSULPHIDE
Although the purchase of commercially available ammonium polysulphide is strongly preferred today, and is further justified because of better system control and enhanced reactivity due to the increased chain length of the molecule, there remain in operation some installations of in-situ polysulphide generation. Ammonium polysulphide (APS) is produced in-situ by passing alkaline sour water through a bed of elemental sulphur sticks or flowers of sulphur. Guidelines for in-situ produced APS are similar to those for commercial APS solutions; that is, maximize the production of S5-2 and control the decomposition of S5-2 to lower order Sx-2. The water should have a pH between 9.0 - 10, with target 9.5. At the relatively high pH of the water and thus of the APS solution, the higher order and more reactive S5 and S4 species are preferentially formed over the S2 and S3 species (Reference 1). Also, high pH aids in HCN ionization, especially important if the in-situ generated solution is to be used directly in a staged polysulphide wash column (as at Shell site 2 Refinery). For production of in-situ APS, the temperature of the alkaline feed water should be between 65°C and 110°C [150 and 230°F]. The solubility of the rhombic sulphur (S8) increases with temperature. Temperatures below 65°C [150°F] favour the stability of S5-2 once produced, but it is necessary to promote solubility of sulphur. Above 110°C [230°F], polysulphide decomposes. Significant sulphide (S-2) and bisulphide (HS-) are also necessary; as is seen by the reactions for dissolution of elemental sulphur and equilibrium of polysulphide ions (Reference 10): S8 + 2 S-2 <=> 2 S5-2 HS- <=> S-2 + H+ x Sx-2 + H2O <=> (x-1) Sx+1-2 + HS- + OH- (x = 2 - 4) In the production of in-situ APS, it is often necessary to operate at higher NH4HS concentrations than 2%. This is acceptable in carbon steel equipment, as long as there is a very high level of active sulphur in the polysulphide. For more information, see the description of the Shell site 2 Refinery CCU polysulphide wash column in Appendix D. Major equipment usually includes two sulphur vessels, a sour water surge vessel, and circulating pumps. Polysulphide circulation in this case takes place in a semi-closed loop. The polysulphide wash solution is sufficiently dilute to be pumped directly from the surge vessel into the process stream. It is subsequently separated from the hydrocarbon phase in a separator and returned to the surge vessel. A small slip stream of the pump discharge is passed through one of the sulphur vessels and back to the surge vessel to replenish the polysulphide in the system, by replacing that which is depleted, for instance, through forming a sulphide film on internal equipment surfaces or by reaction with cyanide. The amount of polysulphide in the circulating stream is controlled by the flow rate of the slip stream through the sulphur vessel. A small amount of circulating solution is blown down to control the thiocyanate concentration. The blowdown should be performed at a location containing the weakest polysulphide solution. Typically, this is the boot of the separator.
42
CONFIDENTIAL
Make-up water to the system can be provided by introducing suitable sour water from a source such as the main fractionator overhead accumulator. The sulphur vessels are sized to have a sulphur capacity sufficient for several weeks of operation. When the supply of sulphur in the active drum decreases to a predetermined low level, flow is switched to the spare sulphur vessel and the depleted vessel is reloaded.
43
CONFIDENTIAL
APPENDIX C. POLYSULPHIDE USE AT SHELL SITE 1 REFINERY CAT CRACKING UNIT Polysulphide use at Shell site 1 Refinery Cat Cracker gas plant is shown in Figure C1. Figure C2 is a sketch of the Shell site 1 Refinery RA and Debut columns, but the general lineup is similar at other locations. The polysulphide injection systems follow the principles elaborated earlier, but are tailored to the specific process equipment available. 1.
Polysulphide Injection at Main Fractionator Overhead
There was polysulphide injection for a short time at the main fractionator overhead, upstream of the overhead condensers, as part of the WGC revisions in 1994. Polysulphide injection at this point proved to be ineffective and resulted in elemental sulphur being formed, with sulphur particles getting into product and fouling injection quills/strainers. Ionic equilibrium calculations showed that when the wash water containing ammonium polysulphide was mixed with the main fractionator overhead gas at 110°C [230°F], the water pH dropped from 8.6 prior to injection, down to 7-7.5. At this low pH elemental sulphur was produced from the ammonium polysulphide, and the sulphur dropped out. This injection was such that 25% of water remained as liquid. Calculations showed that if significantly more injected polysulphide solution remained as liquid, on the order of 75%, that not as much NH3 would be desorbed and not as much sulphur would precipitate. Currently there is a sour water wash with no polysulphide at the main fractionator overhead. 2.
Polysulphide Injection at Interstage WGC
a. Polysulphide Wash of Condenser At Shell site 1 Refinery, the condenser wash water rate, including the polysulphide addition, is 20 gpm. This is a ratio of 0.44 US gal/bbl [1 barrel = 159 litres] CCU feed, and falls in the recommended range of 0.3 - 1.0 US gal/bbl [1 barrel = 159 litres]. b. Separator Water as an Additional Vapour Scrub At Shell site 1 Refinery, this system was designed to be a total of 200 gpm, via four countercurrent spray injections of approximately 50 gpm each. Countercurrent injection results in generation of fine drops, and aids in overcoming the mass transfer resistance of vapour HCN into the liquid drops. Also, countercurrent injection results in generation of more overall surface area because the drops produced are finer, further enhancing HCN absorption into the liquid phase. The recirculation system has not been constructed at Shell site 1 Refinery as of 1/96. It is currently in the 1997 shutdown plan. 3.
Polysulphide Scrubbing in Afterstage
The polysulphide wash of the WGC afterstage at Shell site 1 is shown in Figure C1. At Shell site 1, the afterstage compressed gases are at 138 to 143°C [280 to 290°F], which is too hot for polysulphide to survive at the injection point. Instead, a water wash with no polysulphide is provided upstream of the air cooler to force the dew point at the injection location.
44
CONFIDENTIAL
Polysulphide is injected downstream of the air cooler and upstream of water coolers, where the process temperature is much lower. The afterstage polysulphide wash is a once-through system, with the wash waters being routed directly to the SWS feed surge vessel. There is no water circulation back to the inlet of the afterstage separator, as was done at the interstage separator. Reasons for this are first, that the separator boot is not large enough to allow for physical separation of a large amount of water and liquid hydrocarbon. Also, the boot is not large enough to allow sufficient residence time for much reaction of the cyanide ion to thiocyanate before the water would be recirculated back. There is a surge vessel in the system, which was part of the former in-situ generation of ammonium polysulphide. The surge vessel may provide sufficient residence time for reaction to thiocyanate in a recirculation scheme, but the problem remains of too small a HP Separator boot for good water/oil separation. Also, the current surge vessel piping would expose recirculated polysulphide solution to high temperatures and decomposition at the air cooler inlet. The combined interstage and afterstage polysulphide wash at Shell site 1 also keeps the polysulphide within guidelines at the SWS (<25 ppmw active sulphur). The interstage has only a small rate of net polysulphide-containing sour water going to the SWS feed surge vessel, and there is no polysulphide injected into the main fractionator system. This allows the polysulphide sulphur content of the afterstage sour water to vary depending on the cyanide production in the CCU reactor.
45
CONFIDENTIAL
46
CONFIDENTIAL
47
CONFIDENTIAL
APPENDIX D. CCU
POLYSULPHIDE USE AT SHELL SITE 2 REFINERY
The overview below is an illustration of in-situ polysulphide use, and is not a complete description of polysulphide use at Shell site 2 Refinery CCU. Shell Global Solutions should be contacted for further information. At the Shell site 2 Refinery CCU, there is no WGC interstage, and a trayed column is used to scrub cyanides between the afterstage condenser and the rectified absorber column. The lineup of the polysulphide wash column is shown in Figure D2. The sump of the column acts as the high pressure separator. This is a good location for a scrubbing column, since the gas volume is reduced at the high operating pressure, allowing the column to be smaller in diameter. Polysulphide solution is circulated over ten trays, which typically removes greater than 90% of the cyanide entering the column. Shell site 2 Refinery feeds a crude oil which produces a relatively large amount of cyanide, and this is one of the primary reasons for using a column for scrubbing. Shell site 2 Refinery generates ammonium polysulphide in-situ, by procedures described in Appendix B. Very high levels of active sulphur, 5000 - 7000 mg/L, are necessary to effectively remove HCN in the column. This high level of active sulphur can only be achieved at high pH (optimum pH 9.5) and high levels of bisulphide ion HS-, to support the reactions occurring in the sulphur drum. When the system operates successfully, the natural bisulphide ion concentration seems to be in the 23% range (NH4HS 4-6% range), basis its own chemistry and the gas concentrations of NH3/H2S/CO2. However, the NH4HS concentration varies; levels of NH4HS of 10-15% are not uncommon. Essential to the success of the Shell site 2 Refinery system is the relatively high level of NH3, about 0.7% vol, in the vapour feed to the polysulphide wash column. The very high excess active sulphur seems to provide acceptable protection of carbon steel lines and equipment.
48
CONFIDENTIAL
49
CONFIDENTIAL
APPENDIX E. FLEXICOKER
POLYSULPHIDE USE AT SHELL SITE 2 REFINERY
At the Shell site 2 Refinery Flexicoker, polysulphide is used in the main fractionator gas compression train, and in flexigas treating. The polysulphide in the gas compression train is generated in-situ, and used directly in a trayed column after the WGC second stage. This is similar to the lineup for polysulphide use at the Shell site 2 Refinery cat cracker. The other polysulphide use is in treatment of flexigas, which contains fine coke particles. Figure E1 is a simplified drawing highlighting the use of polysulphide in this application. Even though flexicoking is a unique process, one reason for discussing polysulphide use in flexicoking is that it may have more general applicability to solids processing and the prevention of cyanide contamination of effluent waters. At the Shell site 2 Refinery Flexicoker, cyanide production is such that at the Coke Gas Cooler (CGC), sour water absorbs HCN from the flexigas, producing cyanide concentrations in the CGC sour water of 20-50 ppm (wt /vol). Polysulphide is added in two places, to the makeup water from the CGC to the venturi, and at the sour water bleed to the SWS (see Figure E1). The sour water bleed polysulphide addition is a typical instance of SWS feed treatment, but the circulating water polysulphide addition merits further discussion. When the flexigas containing HCN is contacted in the venturi scrubber, cyanide in the flexigas is converted to Ni(CN)4-2, as there is much more Ni than Fe in the flexigas fines. Polysulphide active sulphur will not treat complexed Ni or Fe cyanides, so the strategy is to have a polysulphide wash in the venturi scrubber as the solids are contacted. Treatment for cyanides is part of the overall effluent treatment program. The polysulphide active sulphur is believed to function mostly as a sulphiding agent for the Ni on the surface of the ash particles. This is believed to hinder the reaction of the HCN with the Ni, to produce the Ni-CN complex. As Figure E1 shows, sour water with added polysulphide is injected into a vapour stream at 193°C [380°F], significantly above the 121°C [250°F] recommendation for maximum vapour phase injection temperature. In this case what is desired is to inject the polysulphide solution at high temperature, to have the polysulphide, and/or the sulphur at breakdown, sulphide the surface of the particle, so as to form insoluble nickel-sulphide rather than soluble nickel-cyanide. This is done solely for effluent treating purposes. Sodium polysulphide is used rather than ammonium polysulphide. It was found with ammonium polysulphide that deposits, probably ammonium carbonate or increased calcium carbonate deposits, formed in the venturi, the coke gas cooler circulating water piping, and the wet fines strippers (flexigas contains about 5% mol CO2). To avoid a separate handling system for ammonium polysulphide, sodium polysulphide is also used for treating the sour water bleed to the SWS.
50
CONFIDENTIAL
51
CONFIDENTIAL
APPENDIX F. CORROSION OF AIR COOLERS AND POLYSULPHIDE USE AT SHELL SITE 2 REFINERY SWS
Shell site 2 Refinery has separate sour water strippers for treating sour waters from flexicoking (SWS#3), and sour water from cat cracking (SWS#4,5). Both systems have a history of cyanide aggravated ammonium hydrosulphide corrosion, which was effectively mitigated by polysulphide addition. At the Shell site 2 Refinery SWS that treats Flexicoker sour water, 316L tubes in the air cooler averaged 0.36 mm/yr [14 mpy] corrosion rates, and had to be renewed on a 12-22 month cycle, from 1986-1990. Polysulphide treatment at the Flexicoker was started in 1989-1991, and since the retube in 1990 corrosion rates have averaged 0.06mm/yr [2.5 mpy]. A portion of the SWS feed water containing polysulphide is diverted as an air cooler wash. In the Shell site 2 Refinery SWSs that treat CCU sour water, Alloy 800 air cooler tubes have suffered cyanide-enhanced NH4HS corrosion. About 0.25mm/yr [10 mpy] corrosion was observed from 1986-1992, and a tube failure occurred in 1992. Ammonium hydrosulphide concentrations in the SWS overhead system, measured in 1992, ranged from 24% NH4HS to 38% NH4HS. Polysulphide treatment of the SWS feed was started shortly after the tube failure, with a portion of polysulphide-treated feed diverted to overhead wash of the air coolers. When polysulphide treatment at the CCU was started, SWS feed addition was stopped, so residual polysulphide from cat cracking treats the SWS feed water. Corrosion rates are apparently much less since 1992 (on the order of 0.025mm/yr [1 mpy]). With polysulphide treated sour water at the CCU, typical HCN concentrations are 1-1.5 ppmv at the SWS overhead air coolers.
52
