SAER-5942

SAUDI ARAMCO ENGINEERING REPORT

SAER-5942

CONSULTING SERVICES DEPARTMENT TECHNOLOGY ITEM CSD-TI-01/99-J

Ammonium Bisulfide Corrosion in Hydrocracker and Refinery Sour Water Service

January 1, 2004

Robin D. Tems Materials Engineering and Corrosion Control Division

Consulting Services Department Dhahran

SAER-5942 Page 3

EXECUTIVE SUMMARY SAER-5942 is the Final Report on a Joint Industry Program (JIP) to investigate ammonium bisulfide corrosion in hydrocracker reactor effluent air cooler systems and similar sour water systems. The field partner for this project is Ras Tanura Refinery. The title of the Technology Item CSD-01/99-J was “Prediction and Assessment of Ammonium Bisulfide Corrosion under Refinery Sour Water Conditions.” Experimental work was performed by Shell Global Solutions (US) Inc. and InterCorr International. Through this project, the first systematic corrosion study in the industry for this system has been completed. Iso-corrosion curves have been generated, and the results have been directly applied at Ras Tanura Refinery and Riyadh Refinery. A corrosion prediction program has been developed based on the data. The program, when used in combination with experiential rules-of-thumb, provides improved understanding of system corrosivity. Further refinement of the program will be achieved during phase II of the JIP program. Direct savings from this project are on the order of $36,000 per year. Cost avoidance savings are of far greater significance and range from $400,000 per downtime-day to $50 million, depending upon the scenario postulated. A new Technology Item is recommended to expand the experimental database developed in this project.

SAER-5942 Page 5

TABLE OF CONTENTS

EXECUTIVE SUMMARY 3 1. SUMMARY 7 2. OBJECTIVE 8 3. SCOPE & METHODOLOGY 9 4. RESULTS 11 4.1 Iso-corrosion curves 11 4.2 Velocity effect 11 4.3 Ammonium bisulfide effect 15 4.4 Hydrogen sulfide partial pressure effect 15 4.5 Temperature effect 15 4.6 Chloride effect 15 4.7 Ammonium polysulfide effect 15 4.8 Imidazoline inhibitor effect 16 4.9 Hydrocarbon effect 17 5 COMPUTER PROGRAM: PREDICT–SW VERSION 1.02 18 6. APPLICATION TO RAS TANURA REFINERY 20 7. APPLICATION TO RIYADH AND RABIGH REFINERIES 23 8. INTERIM API GUIDELINE 24 9. FINANCIAL ANALYSIS 24 10. FUTURE WORK 24 11. LICENSE RESTRICTIONS CONCERNING DATA 25 12. REFERENCES 25 ATTACHMENT: CONTRACTORS FINAL REPORT

SAER-5942 Page 7

AMMONIUM BISULFIDE CORROSION IN HYDROCRACKER AND REFINERY SOUR WATER SERVICE

1. SUMMARY The results of this project are:

(1) Development of systematic data for the selection of materials for new hydrocracker effluent systems.

(2) Definition of critical parameters in the selection of materials for hydrocracker systems.

(3) Development of preliminary data on the effects of imidazoline inhibitor and ammonium polysulfide (APS) additions, and hydrocarbon type.

(4) Development of a Windows based computer program to predict corrosion rates in these systems.

(5) Development of opportunities for informal networking with experts from other major refining companies.

Figure 1 presents a relative ranking of the alloys evaluated in the test program. The object of successful design is to select an adequately corrosion resistant material to ensure safety and serviceability while minimizing the life-cycle cost of the system.

Figure 1: Relative corrosion resistance of materials in the test environments

Leas

t Res

ista

nt

Mos

t Res

ista

nt

Carbon SteelAlloy 400

410 SS304 SS316 SS

Alloy 600

Alloy 825Alloy 625

Alloy 2507

Alloy C-276

Alloy 800Alloy 2205

Alloy 20Cb-3

Alloy AL-6XN

Leas

t Res

ista

nt

Mos

t Res

ista

nt


410 SS304 SS316 SS

Alloy 600

Alloy 825Alloy 625

Alloy 2507

Alloy C-276

Alloy 800Alloy 2205

Alloy 20Cb-3

Alloy AL-6XN

SAER-5942 Page 8

The results have already been applied in Saudi Aramco through:

(1) Consultation provided to Ras Tanura in the selection of metallurgy for replacement of the South Refinery J-80 hydrocracker piping resulting in greatly improved system reliability.

(2) Consultation provided to Ras Tanura in the definition of inspection programs for the new Alloy 825 pipework resulting in a significant reduction in the inspection effort.

(3) Consultation to Riyadh Refinery and Inspection Department.

(4) Consultation to Rabigh Refinery project teams as needed.

(5) Presentation of early results to refinery corrosion engineers at the Downstream Discussion Meeting in Riyadh in 2001.

The contractors’ final report is presented as an attachment to this SAER and provides full details of experimental work and results. 2. OBJECTIVE Ammonium bisulfide forms in sour water systems where both ammonia and hydrogen sulfide are present. Ammonium bisulfide causes severe flow-enhanced corrosion damage and also is responsible for under-deposit corrosion attack. Flow-enhanced or erosion-corrosion can occur at impingement points such as elbows or Ts, or downstream of flow disturbances such as welds or flow control valves. In low-flow areas, under-deposit corrosion or interface corrosion with pitting in a line is also reported. This corrosion attack has been hard to predict. Existing industry guidelines have been built around collected field experience1,2 and there has been no systematic investigation of corrosion in these systems. The situation becomes of great importance in hydrocracker effluent systems where the pipe systems contain the corrosive agents plus flammable hydrogen and naphtha and high concentrations of toxic hydrogen sulfide at very high pressure. The ability to predict corrosion attack in these systems is of critical importance because the consequences of failure are significant. Unexpected high corrosion rates have sometimes been found in Saudi Aramco systems. For example, between late 1993 and 1995, corrosion rates ranging between 2.5 to 4 inches per year necessitated urgent repairs to critical parts of the Riyadh Refinery hydrocracker reactor effluent air coolers (REAC). The objective of this work is to provide a systematic study of the corrosion of a range of construction materials often used in hydrocracker effluent systems, to examine the variables that affect the corrosion rate, and to develop a Windows based computer program predict materials performance.

SAER-5942 Page 9

3. SCOPE & METHODOLOGY The project evaluated the performance of fourteen materials ranging from carbon steel to Alloy C-276 in a range of sour water environments simulating hydrogen-sulfide dominated, alkaline sour water. The list of materials is presented in Table 1. The experimental program only included base material. No welds or heat affect zones were included in the samples. Several of these materials are normally welded with over-alloyed fillers equivalent to Alloy 625 which performed exceptionally well. Future experimental designs would benefit from the ability to investigate welded material.

Table 1: Materials Evaluated

AISI 1018 carbon steel Alloy 400 63-70 Ni / 2.5 max Fe / bal Cu, Monel AISI 410 12 Cr martensitic stainless steel AISI 304 18-20 Cr / 8-10.5 Ni, austenitic stainless steel AISI 316 16-18 Cr / 2-3 Mo / 10-14 Ni, austenitic stainless steel Alloy 2205 21-23 Cr / 2.5-3.5 Mo / 4.5-6.5 Ni, duplex stainless steel Alloy 800 19-23 Cr / 30-35 Ni / bal Fe Alloy 600 14-17 Cr / 72 Ni / 6-10 Fe Alloy 20Cb-3 19-21 Cr / 32-38 Ni / 3 Cu / bal Fe Alloy 825 19.5-23.5 Cr / 2.5-3.5 Mo / 38-46 Ni / 1.5-3 Cu / bal Fe Alloy 625 20-23 Cr / 8-10 Mo / 5 max Fe / 3.15-4.15 Cb / bal Ni Alloy 2507 24-26 Cr / 3-5 Mo / 8 Ni / bal Fe AL-6XN 20-22 Cr / 23.5-25.5 Ni super austenitic stainless steel Alloy C-276 14.5-16.5 Cr / 4-7 Fe / 2.5 max Co / bal Ni, Hastelloy C-276

The program evaluated weight-loss corrosion. The study did not include an evaluation of environmental cracking or pitting from contaminants such as chloride. Some of the materials listed can experience polythionic stress corrosion cracking if inadequate shutdown and lay-up procedures are followed. The industry reports many cases of polythionic cracking of Alloy 800 and Alloy 800H due to sensitization during welding. Alloy 800 also has experienced pitting in field operations. Low-grade austenitic stainless steels such as Type 316 have experienced external pitting and chloride stress corrosion cracking in instrument tube service in many locations within the Ras Tanura J-80 plant. Type 316 stainless steel was not used for piping in the main hydrocracker circuit. Consideration should also be given to sulfide stress cracking and related mechanisms. Intermediate materials such as duplex stainless steels have been reported to experience sulfide stress cracking in hydrocracker REAC environments. Type 410 stainless steel would also be expected to be prone to sulfide stress cracking. The effects of key variables on corrosion rate were determined: ammonium bisulfide concentration, hydrogen sulfide partial pressure, velocity, temperature, and chloride

SAER-5942 Page 10

concentration. Sub-tasks evaluated the effects of chemical treatments and the effects of the presence of hydrocarbon. The variable ranges examined are presented in Table 2. Not all variables were examined at all conditions. Rather, baseline data were obtained for all materials and then specific tests performed to establish rules for other variables.

Table 2: Variables Evaluated

Ammonium bisulfide concentration, weight % 1 - 30 Hydrogen sulfide partial pressure, psia 30, 50, 100, 150 Velocity, feet per second 1 - 80 Temperature, °F 130, 190, 250 Chloride, ppm 0, 100, 1000 Inhibitor, imidazoline, ppmv 0, 50, 500 Inhibitor, ammonium polysulfide, ppmv 0, 50, 500 Hydrocarbon 0-100 %, light and heavy

The velocity variable reported is the nominal velocity of single phase liquid in the test autoclave system. For the purposes of developing the field prediction program, Predict-SW, these data were converted into sheer stress. As can be seen from Figure 2, linear velocities of a single phase fluid in a small bore pipe create higher shear stresses than the same linear velocities in a larger bore pipe. Shear stress analysis is further complicated when aspects such as multiphase flow, bends, and Ts are considered such as will be found in a real field system. To accommodate these factors, stress amplification factors were applied. For example, a weld protrusion was assigned an amplification factor of 3.5.3,) Laboratory experiments and program development were performed by Shell Global Solutions (US) Inc. and InterCorr International, both of Houston, Texas. The experimental apparatus employed two autoclave systems both with internally mounted pump and flow loop. Tubular coupons (0.5-inch OD, 0.75-inch long, with a 0.15 inch bore) were mounted in the flow channel which could accommodate six electrically-isolated coupons each test. In addition, each test accommodated six coupons in a static zone environment where there was only limited flow. The autoclave loading procedure required meticulous attention to detail to prevent oxygen contamination. Advanced environmental modeling was required to define the test solutions and gases loaded to each experiment in order to achieve the required experimental conditions, mimicking true field conditions within the limitations of laboratory equipment. The ionic modeling used a package developed by Aspen Technology, Inc.—ASPEN /OLI Flowsheeter. The modeling was also able to predict the pH of test environments at room temperature and at test temperature. pH values become more alkaline with an increase in ammonium bisulfide concentration or a decrease in hydrogen sulfide partial pressure. For 1 weight percent NH4HS at 77°F with 50 psia H2S, the predicted pH was 6.9. For 30 weight percent NH4HS at 77°F with 50 psia H2S, the predicted pH was 9.11.

SAER-5942 Page 11

Figure 2: Comparison between shear stress and velocity for various pipe sizes Shear Stress vs. Velocity Correlation

100% Liquid Flow; 1 - 20 wt% NH4HS; 130 F; 50 psia H2S; Straight Pipe Sections

10

100

1000

10000

0 10 20 30 40 50 60 70 80 90

Velocity (fps)

Shea

r Str

ess

(Pa)

0.15" ID Pipe (Lab Apparatus)2" ID Pipe4" ID Pipe6" ID Pipe

4. RESULTS

4.1 Iso-corrosion curves

Baseline iso-corrosion curves were determined for all materials to provide systematic data for the selection of materials for new hydrocracker systems. The effects of variables were determined for all materials. Figure 1 presents a simplified overview of material performance in the range of test environments. Alloy 400 and Alloy 2205 performed worse than expected while AL-6XN and Alloy 2507 performed considerably better than expected. Figure 3 presents baseline iso-corrosion curves for carbon steel. Figure 4 presents data for Alloy 2205, Figure 5 for Alloy 825, Figure 6 for Alloy 625, Figure 7 for Alloy AL-6XN, and Figure 8 for Alloy C-276. Note that different contour line values are used on the carbon steel chart compared to the alloy charts. Full data sets of parametric effects for variables are presented in the attached vendor’s report.

4.2 Velocity Generally, increasing velocity results in an increase in corrosion rate. This effect is most marked for carbon steel, and progressively decreases with higher alloys. See Figures 3 through 8, below.

SAER-5942 Page 12

Figure 3: Baseline iso-corrosion curves for carbon steel

ISOCORROSION DIAGRAM FOR CARBON STEEL(48-hr laboratory tests: P[H2S]=50 psia, T=130 F)

0

10

20

30

40

50

60

70

80

90

0 2 4 6 8 10 12 14 16

NH4HS Concentration, %w

Velo

city

, ft/

sec

< 5 mpy 5-20 mpy 20-50 mpy 50-100 mpy 100-200 mpy 200-300 mpy > 300 mpy

<5 mpy

5-20 mpy

20-50 mpy

50-100 mpy

100-200 mpy 200-300 mpy

>300 mpy

CONFIDENTIALInterCorr/Shell Sour Water JIP

Figure 4: Baseline iso-corrosion curves for Alloy 2205

ISOCORROSION DIAGRAM FOR ALLOY 2205(48-hr laboratory tests: P[H2S]=50 psia, T=130 F)

0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec

< 2 mpy

2-5 mpy

5-10 mpy

10-20 mpy

> 20 mpy< 2 mpy 2-5 mpy

5-10 mpy

10-20 mpy

> 20 mpy


SAER-5942 Page 13



0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec

< 1 mpy

1-2 mpy

2-5 mpy

5-10 mpy

> 10 mpy

< 1 mpy

2-5 mpy

5-10 mpy

> 10 mpy

1-2 mpy




0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec

< 1 mpy

2-5 mpy

< 1 mpy

2-5 mpy

1-2 mpy


SAER-5942 Page 14

Figure 7: Baseline iso-corrosion curves for Alloy AL-6XN

ISOCORROSION DIAGRAM FOR ALLOY AL6XN(48-hr laboratory tests: P[H2S]=50 psia, T=130 F)

0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec

< 1 mpy

1-2 mpy

< 1 mpy

1-2 mpy


Figure 8: Baseline iso-corrosion curves for Alloy C-276

ISOCORROSION DIAGRAM FOR ALLOY C-276(48-hr laboratory tests: P[H2S]=50 psia, T=130 F)

0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec

< 1 mpy

< 1 mpy


SAER-5942 Page 15

4.3 Ammonium bisulfide Increasing ammonium bisulfide concentration results in increased corrosion. Even at 2 weight percent NH4HS, corrosion rates can be from 20 to 50 mils per year (mpy) under certain shear stress conditions. At 8 weight percent NH4HS, carbon steel corrosion rates can range from a low of 5 mpy under low shear stress conditions up to 200 mpy for high shear stress conditions. See Figures 3 through 8 above.

4.4 Hydrogen sulfide partial pressure An increase in hydrogen sulfide partial pressure increased the corrosion rate; this increase was accentuated by higher velocities and higher ammonium bisulfide concentrations. The effect was far more extreme for the least corrosion resistant materials, carbon steel, Monel 400, and Type 410 stainless steel. For the other more corrosion resistant materials, the effect was still present but was an order of magnitude smaller. Detailed graphs are presented in Attachment 3 of the contractors final report.

4.5 Temperature Increasing temperature caused an increase in corrosion rates; however, the effect was small when compared to the effect of NH4HS, velocity, or H2S partial pressure. The greatest effects were seen for carbon steel and Type 410 stainless steel at lower ammonium bisulfide concentrations and higher velocities. Detailed graphs are presented in Attachment 4 of the contractors final report.

4.6 Chloride Chloride concentrations up to 1000 ppm were determined to have no appreciable effect on the baseline corrosion rate. If anything, increasing chloride concentration resulted in lowering the corrosion rate, especially under very high ammonium bisulfide loadings. No significant solution changes were perceived that might help to explain this result. However, experiences reported by other operators were that ammonium chloride deposits can cause pitting in corrosion resistant alloys. Of greatest risk is the first row of horizontal air cooler tubes where the deposits could drop out and remain on the surface. Even higher alloys such as Alloy 825 could be prone to this type of attack. Other operators report corrosion rates up to 40 mpy for Alloy 825 under moist ammonium chloride deposits. This corrosion mechanism is an acid chloride attack. This type of attack is best corrected by adequate water wash.

4.7 Ammonium polysulfide (APS) The effects of APS additions on the corrosion of carbon steel were examined over the range 50 to 500 ppmv. At 500 ppmv and a test velocity of 20 feet per second, 92 percent efficiency was obtained. APS was sensitive to velocity, and efficiency reduced to about 73 percent at 500 ppmv treatment level and 80 feet per second. Results are shown in Figure 9.

SAER-5942 Page 16

Figure 9: Inhibition efficiency of ammonium polysulfide additions APS

0

10

20

30

40

50

60

70

80

90

100

0 100 200 300 400 500 600 700 800 900 1000

Dosage (ppmv neat chemical)

Effic

ienc

y (%

)

80 ft/sec

20 ft/sec

APS PREDICT-SW RulesFor velocities < 20 ft/sec, use 20 ft/sec curveFor velocities > 80 ft/sec, use 80 ft/sec curveFor velocities between 20 and 80, interpolate between the respective curves


4.8 Imidazoline inhibitor The efficiency of imidazoline oil soluble corrosion inhibitor in reducing the corrosion rate of carbon steel was examined at 100 ppmv and 500 ppmv. Because the tests involved an oil soluble inhibitor in an aqueous test environment, the test results were

Figure 10: Inhibition efficiency of imidazoline Imidazoline

0

10

20

30

40

50

60

70

80

90

100

0 100 200 300 400 500 600 700 800 900 1000


Effic

ienc

y (%

)

Most Conservative

Least Conservative

Imidazoline PREDICT-SW RulesFor Horizontal-Stratified, Horizontal-Wave, Horizontal-Laminar and Vertical-Laminar Use Efficiency = 0For Horizontal-Annular Mist and Vetrical Annular, use most conservative curveFor all other flow conditions, use least conservative curve


SAER-5942 Page 17

skewed by the efficiency of delivering the inhibitor to the metal test surface. Experiments where the inhibitor was more efficiently delivered to the surface, that is, high velocity tests at 500 ppmv of neat inhibitor, resulted in an inhibition efficiency of 97 percent. Tests where the inhibitor was less efficiently delivered to the metal surface resulted in reduced efficiencies. The field consequence of this is that in turbulent multiphase flow, imidazoline should provide improved corrosion control. In separated flows such as highly stratified flow or a single phase water line downstream of a separator, inhibition would not be effective. In very high velocity flows, there would be a dynamic balance between continual supply of inhibitor to the surface and removal by the stream. As the velocity increases further, it is expected that the inhibitor film would be removed at a faster rate and corrosion would continue. A summary of imidazoline data are presented in Figure 10, which shows the correction factors included in the Predict-SW program.

4.9 Hydrocarbon

The corrosion rate of carbon steel was determined with various amounts of light hydrocarbon (API 40) or heavy hydrocarbon (API 20) present. Significant effects were seen with the addition of 25 percent hydrocarbon which reduced experimental corrosion rates by about 85 percent. For Type 410 stainless steel, significant effects were seen with the addition of 10 percent hydrocarbon which reduced experimental corrosion rates by about 90 percent. For Type 304, and Alloy 2205 stainless steels, significant effects were seen with the addition of 10 percent hydrocarbon which reduced experimental corrosion rates by about 75 percent. With other alloys, data were more scattered, and the effect was less significant, a reduction of the order of 50 percent. The results for carbon steel and the rule programmed into Predict-SW are presented in Figure 11.

Figure 11: Influence of hydrocarbon on the corrosion rate of carbon steel

Influence of Hydrocarbon - Carbon Steel

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70 80 90 100

Hydrocarbon Content (vol%)

Effic

ienc

y (%

)

Heavy HydrocarbonLight HydrocarbonRule


SAER-5942 Page 18

5. COMPUTER PROGRAM: PREDICT-SW The data generated in this JIP have been incorporated into a Windows-based corrosion prediction program, Predict-SW Version 1.02. Under the license agreement, Saudi Aramco has a single user copy of the program that is available through Consulting Services Department. Contact Robin Tems on 03-874-6130 or [email protected] for further information. The program’s predictive capabilities were evaluated versus experimentally derived data. Most predictions were within 10 percent of the measured value, though a few were more scattered, within approximately 25 percent of the measured value.

Table 3: Predict-SW input data for J80-P-0137

Line / vessel number J80-P-0137(1) J80-P-0137(2)

Environment Total pressure, psig 2125 psig 385 psig Hydrogen sulfide, mol % 0.71 mol % 0.71 mo l % Temperature, degrees F 125 F 125 F Ammonium bisulfide concentration, weight % 2 wt % 2 wt % Oil phase present (API-20, API-40 or none) None None Chemical additives if any (APS, Imidazoline, or none) None None Process Stream Conditions Pipe configuration (straight, 3-D bend, 90-degree elbow, weld protrusion)

Weld Protrusion

Weld Protrusion

Type of flow (Horizontal, vertical up, vertical down) Horizontal Horizontal Application Pipe INTERNAL diameter, inches 1.687 in 1.939 in Corrosion allowance, inches 0.125 in 0.125 in Design life, years 20 year 20 year Condition of pipe (new, lightly corroded, heavily corroded) Corroded Corroded Process Stream Flow Rates & Properties Vapor Properties --- --- Flow rate, MMSCFD --- --- Specific gravity (if not available, default values will be used) --- --- Viscosity, cp (if not available, default values will be used) --- --- Sour Water Properties Flow rate, bbls/day 158.5 gpm 158.5 gpm Density, kg/m3 (if not available, default values will be used) 61.7 lb/ft3 61.7 lb/ft3 Viscosity, cp (if not available, default values will be used) 0.53 cp 0.53 cp Liquid Hydrocarbon Properties Flow rate, bbls/day --- --- Density, kg/m3 (if not available, default values will be used) --- --- Viscosity, cp (if not available, default values will be used) --- --- Corrosion rate Measured in field at failure, mpy 66 70 Predict-SW, mpy <1 <1

SAER-5942 Page 19

The program’s predictive capabilities versus field experience reported by the sponsors could not be evaluated because insufficient data were provided in the final report compared to the required program inputs. In future projects, it is recommended that Saudi Aramco attend all project meetings including wrap-up meetings to ensure that the company’s needs are met. The program was used to evaluate a recent failure in Ras Tanura J-80 plant. The sour water line from the high pressure separator, D-130, corroded at two welds in a reduced diameter portion near a control valve. One failure occurred in a weld in a reducer from 4-inch pipe to 2-inch pipe. Adjacent pipe was undamaged, corroding at about 1 mpy. The other failure occurred in the 2-inch pipe at the first weld downstream of the valve. Again adjacent pipe was virtually undamaged, corroding at 1 mpy. The 4-inch line was undamaged, corroding at 1 mpy or less. The input data for calculation of the corrosion rate are presented in Table 3, above. Some data are known from system measurements, whereas other data are based on original design data or on generally available data for the unit as a whole. These uncertainties result in uncertainty in the calculation. Based on the data in Table 3, above, the corrosion rate estimates by Predict-SW were less than 1 mpy. The measured rate in the field at two different failure points was 65 to 70 mpy. Thus, the program models the bulk of the pipe successfully, but would appear to need further development in modeling severe turbulence locations. The calculation is also very sensitive to the concentration of hydrogen sulfide present in the line, which was based on generic data for the unit rather than specific measurements

Figure 12: Effect of H2S concentration on calculation of corrosion rate, J80-P-0137(1); line pressure 2125 psig

0

20

40

60

80

100

120

140

0 2 4 6 8 10 12mol % H2S

Cor

rosi

on ra

te, m

py

SAER-5942 Page 20

for the line. Calculations of corrosion rate versus hydrogen sulfide concentration for the high pressure portion of the line (J80-P-0137(1)) are shown in Figure 12. This shows that if there is a small error in the actual mol percentage of hydrogen sulfide present, the predicted corrosion rate can be significantly different. We can explore this possibility further by examining the same effects for the low pressure portion of the line. These results are shown in Figure 13. Here the effect is less significant, due to the lower partial pressure of hydrogen sulfide at the reduced system pressure.

Figure 13: Effect of H2S concentration on calculation of corrosion rate, J80-P-0137(2); line pressure 385 psig

0

20

40

60

80

100

120

140

0 2 4 6 8 10 12mol % H2S

Cor

rosi

on ra

te, m

py

The conclusion is that while the program can be used to indicate differences in the general severity of the environment, the program needs further refinement before it is able to fully analyze each and every piece of piping because insufficient data may exist for specific lines. Design of new systems should therefore use a combination of experiential rules of thumb in addition to Predict-SW calculations. 6. APPLICATION TO RAS TANURA REFINERY Plant J-80 at Ras Tanura is a hydrocracker that was started-up in 1999 under BI-3717. The project selected Alloy 825 for the reactor effluent air coolers (REAC) and the inlet piping to the air coolers, while carbon steel was selected for the high pressure pipework downstream of the REAC. This downstream pipework carries a mixture of hydrogen, sour water, hydrogen sulfide, and naphtha at approximately 2,225 psi and a temperature of about 130°F. The designers had opted for a so called “C” design header which

SAER-5942 Page 21

provided a balanced inlet to the REAC but an unbalanced outlet flow distribution pattern with a complex series of multiple elbows. Historically, unbalanced flow distribution patterns have been one of the critical factors that can result in severe erosion-corrosion in REAC systems. Additional wash water was added to the design to dilute the ammonium bisulfide levels in the system and reduce the corrosion that was expected due to high velocities and unbalanced header design. To support these measures, an intensive ultrasonic thickness measurement was established surveying 30,000 points. These various UT surveys indicated that corrosion was proceeding at high rates and that the useful life of the system may be less than five years. Based on early inhibitor results from the JIP, the use of corrosion inhibitors was recommended to extend the useful life of the system. The planned treatment was a dosage of 6-10 ppm (based on total hydrocarbon flow) of a typical crude unit overhead filmer such as Nalco 5186. For ease of application, this product can be blended with the water wash upstream of the high pressure water wash pumps, thereby eliminating the need for extensive high pressure system additions. However, in 2001, after 5 months operational service and three complete UT surveys, the opportunity arose to upgrade the system. Based on the JIP data plus experience with the reactor effluent air coolers, Alloy 825 was selected for use in the new downstream pipework. Further, the piping design was modified to provide a balanced flow system. The only concern with the selection of Alloy 825 was the possibility of ammonium chloride under-deposit corrosion. Based on JIP experiential data, the most likely place to find ammonium chloride under-deposit corrosion is in the first few rows of a horizontal air cooler. Therefore, the Alloy 825 REAC was opened and examined for ammonium chloride attack. No ammonium chloride damage was found. Opening the REAC was no easy task as this was a fully welded construction due to the nature of the service: high pressure hydrogen plus reaction products. When the old pipework was removed, detailed examination was possible, though some damage occurred during removal. A pyrophoric iron sulfide fire occurred and the pipework was liberally flooded with water. Based on the ultrasonic thickness measurements performed in the field and at CSD’s Materials Engineering and Test Facility, the corrosion rate of the original carbon steel pipework was found to range between 66 and 180 mpy. These data rely heavily on the baseline UT data. Should there be any errors in these, the measured corrosion rates would be unreliable. Visually, much of the pipework was unremarkable, with minor pitting present in some places. API5 reports that the expected manifestation of corrosion in these pipe systems is a swirled groove indicating the path of water or erosion corrosion at impingement points. The cap from the horizontal header had considerable corrosion products present and apparently some significant pitting. However, the thick wall forgings were somewhat irregular in dimension from the original manufacturing process

SAER-5942 Page 22

so there is even some uncertainty associated with these measurements. Figure 14 provides an example of ultrasonic data.

Figure 14: Thinnest reading versus baseline data, Elbow Z

0.800

0.900

1.000

1.100

1.200

1.300

1 2 3 4 5 6 7 8 9 10 11 12

Data points around pipe

Thik

ness

, inc

hes

Minimum thickness found in lab investigation after 5 months run-time, 2.5 yr total time. Run-time corrosion rate 180 mpy. Remaining life to Tmin = 2.9 yr

The series of lines represent different bands of data around the pipe, the individual points representing specific locations on each band. Estimation of the expected corrosion rate was performed using an early version of the Predict-SW program. Assuming a pipe velocity of 20 feet per second, 1 percent ammonium bisulfide, and 150 psi hydrogen sulfide, a corrosion rate of 70 mpy was predicted. This is in-line with the range of data derived from the UT data. Figure 15 shows the inside of the south end cap from the header. Figure 16 presents the north end cap after cleaning. Consultation was provided to Ras Tanura in the definition of inspection programs for the new Alloy 825 pipework resulting in a significant reduction in the inspection effort. Inspection requirements on the old carbon steel pipework system required an inspection team of two men full time plus engineering support from Ras Tanura Inspection Unit. Based on the low corrosion rates expected with Alloy 825 plus experiential data gained through the JIP, a recommendation was issued to inspect Alloy 825 only once every 5 years, with a commensurate savings in direct costs.

SAER-5942 Page 23

Figure 15: Inside of south end cap

Figure 16: Inside of south end cap

7. APPLICATION TO RIYADH REFINERY AND RABIGH REFINERIES Data from the JIP have been applied to several requests from Riyadh Refinery concerning materials selection under differing hydrocracker scenarios. Consultation has also been provided for potential projects at Rabigh Refinery.

Figure 17: Corrosion rates on nozzle from R215-E3

4

2

0

In c h e sp e r

y e a r

J a n J a n A u gA u g1 9 9 4 1 9 9 5

SAER-5942 Page 24

The most notable challenges with the Riyadh system occurred before data from this project were available. Nozzles in the reactor effluent air coolers experienced sporadic corrosion damage at rates up to 4 inches per year. Figure 17 shows data for R215-E3 (B) between December 1993 and August 1995. Note that the x-axis is not linear, but records specific data sets. The y-axis is the corrosion rate in inches per year. The different colors show data from different air coolers within the set of air coolers. The data set shows the sporadic nature of ammonium bisulfide corrosion, which has been partially attributed to mal-distribution of wash water, mal-distribution of flow, and velocities up to 27 feet per second through the nozzles. In this instance, corrosion was successfully alleviated by the application of Alloy 625 weld overlay on the nozzles. 8. INTERIM API GUIDELINE Concurrent with the present Technology Item, API has released the interim results of a new survey of hydrocrackers and hydrotreating facilities. API 932-A was issued in September 2002.5 Out of 24 units in the survey:

• 25% had experienced corrosion-related fires • 21% had no corrosion problems • 50% had substituted alloy for carbon steel. • 8% of plants used inhibition, but 50% of those still had unscheduled outages • Both hydrocrackers and hydrotreaters have comparable failure behaviors.

The report highlights the need for the generation of systematic laboratory data to explain the differences between successful and unsuccessful hydroprocessing operations. 9. FINANCIAL ANALYSIS The base cost of the Saudi Aramco JIP membership was $59,000, with another $825,000 being provided by other partners. Additionally, Shell Global Solutions released to the sponsors corrosion data and laboratory methodologies that were developed over several years prior to the initiation of this project. The funding of this JIP provided a leverage of well over 16 times Saudi Aramco’s base research investment. In addition to the base cost, Saudi Aramco incurred costs to participate in planning and progress meetings, resulting in a total cost of about $90,000, as approved for the project. Direct financial benefits come principally from improved reliability of the reactor effluent system in the hydrocracker, Unit J-80, Ras Tanura Refinery. The high pressure hydrogen/naphtha/sour water pipework downstream of the Ras Tanura J-80 reactor recorded significant wall loss during operation based on repetitive ultrasonic readings performed by the refinery and confirmed later by Inspection Department and CSD. Corrosion leaks in such high pressure systems carrying flammable and toxic gas can result in fires. A typical major hydrocracker fire can cost up to $50 million; the risk of

SAER-5942 Page 25

fire is minimized by the use of corrosion resistant alloy. The recent API survey5 found that 25% of hydroprocessing plants experience corrosion-caused fires. In addition to the risk of fire, a corrosion leak that shuts down the system results in a loss of approximately $0.4 million per day. The replacement of carbon steel pipework with Alloy 825 has eliminated the need for continual inspection of the pipework. This represents a direct saving of approximately $36,000 per year for contractor and Saudi Aramco manpower. The JIP was funded by BP, Chevron Research and Technology Company, ConocoPhillips, ExxonMobil Research and Engineering, Flint Hills Resources, Fluor Daniel, Idemitsu Kosan Company, Kuwait National Petroleum Company, Petrobras, Saudi Aramco, Shell Global Solutions (US), Sunoco, Syncrude Canada, TotalFinaElf, UOP, and Valero Energy Corporation. 10. FUTURE WORK As stated in the initial technology proposal, the present work (Phase 1) covers hydrogen sulfide-dominated alkaline sour water systems. Also of interest are sour water systems with a larger percentage of ammonia, an evaluation of the effects of cyanide, and the development of additional data on inhibitors. These are being studied in Phase 2 of the JIP program. Saudi Aramco sponsorship of Phase 2 is recommended. 11. LICENSE RESTRICTIONS CONCERNING DATA Use or release of data from this project is strictly governed by the JIP agreement. Shell Global Solutions retains all rights to its prior data. 12. REFERENCES 1. R.L. Piehl, “Survey of Corrosion in Hydrocracker Effluent Air Coolers”, Materials

Performance, Vol 15 (1), January 1976, pp 15-20.

2. C. Harvey, R. L. Piehl, A. Singh, “Corrosion of Reactor Effluent Air Coolers,” UOP, 1995.

3. B.V. Johnson, H.J. Choi and A.S. Green, “Effects of Liquid Wall Shear Stress on CO2 Corrosion of X-52 C-Steel in Simulated Oilfield Production Environments”, NACE Corrosion/91, Paper 573.

4. Keller H., 1974, Erosion Corrosion in Wet Steam Turbines, VGB Kraftwerkstechnik, No. 54, pg. 292.

5. API 932-A, “A Study of Corrosion in Hydroprocess Reactor Effluent Air Cooler Systems.”

SAER-5942 Page 26

FINAL REPORT ON PREDICTION AND ASSESSMENT OF AMMONIUM

BISULFIDE CORROSION UNDER REFINERY SOUR WATER SERVICE CONDITIONS

PHASE I

Prepared by:

Dr. Michael S. Cayard InterCorr International, Inc.

14503 Bammel-N. Houston, Suite 300 Houston, Texas 77014 USA

Mr. Richard J. Horvath Shell Global Solutions (US) Inc.

P.O. Box 1380 Houston, Texas 77251 USA

InterCorr Project No.: L993580TK June 10, 2003

1. Table of Contents

Executive Summary ............................................................................................................ 1 Introduction ......................................................................................................................... 2 Background ......................................................................................................................... 2

Previous Investigations ................................................................................................... 2 Predictive Capabilities..................................................................................................... 3

Program Overview .............................................................................................................. 4 Task 1.1 - H2S-Dominated Alkaline Sour Water Systems.............................................. 4

Subtask 1.1.1 – 50 psia H2S Partial Pressure Systems................................................ 4 Subtask 1.1.2 – Parametric Effect of H2S Partial Pressure ......................................... 4

Task 1.2. – Parametric Effects on Sour Water Corrosion ............................................... 5 Subtask 1.2.1 – Temperature Effects .......................................................................... 5 Subtask 1.2.2 – Chloride Effects................................................................................. 5

Task 1.3 - Effect of Hydrocarbon/Sour Water Mixtures................................................. 5 Task 1.4 - Performance of Chemical Treatments in Sour Water Environments............. 6 Task 1.5 - Development of Predict-SW .......................................................................... 6

Materials Evaluated............................................................................................................. 7 Experimental Procedures..................................................................................................... 7

Test Facility..................................................................................................................... 7 Corrosion Coupons.......................................................................................................... 9 Test Protocol ................................................................................................................. 10 Ionic Modeling .............................................................................................................. 11

Results and Discussion...................................................................................................... 12 Task 1.1 – H2S Dominated Alkaline Sour Water Systems ........................................... 12

Subtask 1.1.1 – 50 psia H2S Partial Pressure Systems. ............................................. 12 Subtask 1.1.2 – Parametric Effect of H2S Partial Pressure. ...................................... 15

Task 1.2 – Parametric Effects on Sour Water Corrosion .............................................. 19 Subtask 1.2.1 – Temperature Effects. ....................................................................... 19 Subtask 1.2.2 – Chloride Effects. .............................................................................. 24

Task 1.3 – Effect of Hydrocarbon / Sour Water Mixtures ............................................ 26 Task 1.4 – Performance of Chemical Treatments in Sour Water Environments .......... 33 Task 1.5 – Development of Predict-SW Software Model............................................. 36

Field Experience........................................................................................................ 38 Conclusions ....................................................................................................................... 42 References ......................................................................................................................... 45

LIST OF APPENDICES

Appendix I – Participating JIP Sponsors Appendix II – Baseline Isocorrosion Diagrams Appendix III – Corrosion Rate Correction Factors for H2S Partial Pressure Appendix IV – Corrosion Rate Correction Factors for Temperature Appendix V – Corrosion Rate Correction Factors for Hydrocarbon Appendix VI – Predict-SW Correlations with Isocorrosion Diagrams

Appendix VII – Field Experience Presented by the JIP Sponsors

1

2. EXECUTIVE SUMMARY This report presents the experimental results and findings from a joint industry program conducted by InterCorr International, Inc. entitled “Prediction and Assessment of Ammonium Bisulfide Corrosion Under Refinery Sour Water Service Conditions”. The program was conducted over the period from March 2000 to February 2003 and was jointly sponsored by a group of refining and engineering companies. The goal of this research program was to develop a quantitative engineering database and guidelines to predict corrosion in H2S-dominated alkaline sour water systems as a function of NH4HS concentration, velocity (shear stress), H2S partial pressure, temperature, chloride concentration, hydrocarbon content and chemical treatments. The program included five tasks. The first four involved data development using a specialized laboratory flow loop designed specifically to handle the environments and particulars associated with conducting experiments in NH4HS solutions. The final task involved the development of a comprehensive software tool to predict corrosion rates of a wide range of materials in NH4HS environments. The first objective of the program involved developing comprehensive baseline isocorrosion curves to enable determination of corrosion rates for fourteen materials ranging from carbon steel to Alloy C-276 as a function of flow loop velocity and NH4HS concentration. All data were obtained at 50 psia H2S partial pressure and 130 F. These isocorrosion curves, or “Horvath Curves” as they have come to be known, have already had a profound impact on the refinery industry. They represent the first comprehensive set of NH4HS corrosion data available to the industry for use in evaluating corrosion in, and selecting materials of construction for, various process units containing alkaline sour water. The next phase of the program involved experiments designed to investigate the role of several key process variables on the corrosion rate in NH4HS environments. Variables examined included H2S partial pressure, temperature, and chloride concentration. H2S partial pressure proved to be a major variable leading to significant increases in corrosion rate, particularly with increasing NH4HS concentrations. The effect of temperature was secondary to the effect H2S partial pressure. Corrosion rate increased with increasing temperature, but the increase was more pronounced at low than at high NH4HS concentrations. Corrosion rate was found to be unaffected by the presence of chloride at concentrations up to 1,000 ppm in the sour water. Additional testing was conducted to assess the benefit of hydrocarbon and two types of chemical treatments on reducing corrosion rates. The presence of light and heavy hydrocarbons was investigated. It was found that substantial protection from the hydrocarbon could be attained with hydrocarbon contents ≥ 25 vol% on carbon steel and with as little as 10 vol% on the balance of the alloys. Two chemical treatments were studied with carbon steel, namely ammonium polysulfide (APS) and imidazoline. Both treatments appeared promising, however since imidazoline is an oil soluble chemical, its ability to protect material exposed to the corrosive sour water environment was considered questionable for certain multi-phase flow regimes. The final program objective involved the development of a user-friendly software tool named Predict-SW. This software tool successfully incorporated the data obtained in this program and

2

combined these data with flow modeling calculations on plant tubing/piping configurations to predict the corrosion rates of the fourteen materials studied over a wide range of NH4HS concentration, H2S partial pressure, temperature, hydrocarbon content and chemical treatment. (Top)

3. INTRODUCTION Presented herein is the final report for a research program conducted by InterCorr International, Inc. (InterCorr) in collaboration with Shell Global Solutions (US) Inc. (SGSUS) entitled “Prediction and Assessment of Ammonium Bisulfide Corrosion Under Refinery Sour Water Service Conditions”. The program was conducted over the period from March 2000 to February 2003 and was jointly sponsored by a group of refining and engineering companies. The sponsoring companies are listed in Appendix I. This report contains a comprehensive summary of the program goals, facilities, experimental methods, results and discussion, and program findings. (Top)

4. BACKGROUND The subject of alkaline sour water corrosion in petroleum refineries over the past 25 years has been addressed in the literature [1]. However, despite the prevalence of sour water corrosion problems in refining operations, there was very little ammonium bisulfide corrosion data published in the open literature. Additionally, most of these studies did not take into account velocity / shear stress and had not investigated a wide range of exposure conditions. The approach used in many studies had been to focus on empirical findings heavily relying on evaluations of operational experience. Hence, there was a need for more precise and quantitative data on ammonium bisulfide corrosion for a variety of materials under simulated service conditions. These data were needed as a technical basis for improved prediction of ammonium bisulfide corrosion for use in materials selection, control of process unit operation, and assessment of chemical treatments. (Top)

PREVIOUS INVESTIGATIONS Perhaps the most notable article was Piehl’s paper [2] describing a survey conducted by the NACE Group Committee T-8 (now STG 34) covering corrosion in the reactor effluent air coolers and associated piping in 42 hydrocrackers and hydrotreaters. Analysis of the survey results established that sour water corrosion was mild to negligible when the concentration of ammonium bisulfide was 2 percent or less and the velocity was 20 ft/sec or less. This experience has been generally utilized by the industry for control of sour water corrosion in hydroprocessing unit reactor effluent air coolers, and has served the industry well when followed. Although an actual corrosion threshold was not identified by the survey results, it was noted that corrosion rates may be severe above 3 to 4 percent ammonium bisulfide concentration. Damin and McCoy [3] reported results of laboratory corrosion tests conducted in a stirred autoclave over the ammonium bisulfide concentration range of 10 to 45 percent. They measured low corrosion rates for carbon steel and AISI 316 stainless steel up to about 35 percent ammonium bisulfide concentration. Above that, the corrosion rate of both materials increased

3

rapidly to extremely high rates of attack. Their test results appeared to demonstrate the presence of a threshold ammonium bisulfide concentration at which the corrosion resistance of the materials changed dramatically. They postulated that this was the result of the formation of a metal ammonium complex that could act to strip the normally protective iron sulfide film from the metal surface. It should be noted that the 35 percent concentration threshold was observed at near-stagnant conditions (actual velocity in the stirred autoclave was estimated to be 1 to 2 ft/sec). The only laboratory corrosion data documenting the effect of velocity on ammonium bisulfide corrosion was reported by Scherrer et. al. [4] In their tests, conducted with 4.5 to 10 percent ammonium bisulfide, the corrosion rate of carbon steel increased by 40 to 64 percent when increasing the velocity from 11 to 21 ft/sec, the highest velocity tested. (Top)

PREDICTIVE CAPABILITIES It was readily evident that there was currently insufficient corrosion data to fully understand and predict the corrosiveness of ammonium bisulfide solutions over a wide range of concentration and velocity / shear stress. A recent API survey on corrosion in refinery sour water systems also indicated similar findings [5]. The effect of other parameters such as pH, temperature, partial pressures of H2S and NH3, and solution contaminants such as oxygen, chlorides, and cyanides were not quantified. Additionally, compared to carbon steel, there were even less corrosion data available for many alloys commonly used in this service. For example, no data were found for alloy 2205, which in recent years has been used in the higher concentration systems. Experience surveys have been conducted, but have restricted applicability due to limited availability of information. However, these studies have identified the extent of corrosion problems in process units handling alkaline sour water and the critical needs related to these units in terms of improving predictive capabilities and system reliability. Some operating companies and process licensers have developed their own procedures for controlling ammonium bisulfide corrosion of carbon steel based on operating experience. In many instances, these permit concentrations of ammonium bisulfide exceeding the 2 percent concentration recommended by Piehl, perhaps up to the 8 to 10 percent range, while maintaining the 20 ft/sec maximum velocity criteria. Furthermore, there are some hydroprocessing units with carbon steel reactor effluent systems that have actually operated for periods of time with ammonium bisulfide concentration in the 15 to 20 percent range. It has not been uncommon for hydroprocessing units, designed to handle a given ammonium bisulfide concentration, to be exposed to higher concentrations when the nitrogen level in the feed is increased without a corresponding increase in the injection rate of wash water. This usually results in increased corrosion and in some cases unit reliability problems and unscheduled shutdowns. Within the last 15 years, there have been several major incidents where ammonium bisulfide corrosion caused loss of containment in hydroprocessing units that resulted in damage/lost production on the order of 50 million USD. There have also been failures in the overhead systems of some sour water stripper columns that resulted in significant reliability impacts. Some of these involved rapid corrosion of alloys such as AISI 316 stainless steel and alloy 800, which were previously thought to be resistant in this service. (Top)

4

5. PROGRAM OVERVIEW The goal of this research program was to develop a quantitative engineering database and guidelines to predict corrosion in alkaline sour water systems as a function of NH4HS concentration, velocity (shear stress), H2S partial pressure, and various additional parametric variables. This information was used as a basis for the development of a more accurate and comprehensive software predictive tool including assessment methodologies for control of ammonium bisulfide corrosion of a wide range of materials of construction to help attain safe and reliable operation of process units handling ammonium bisulfide. The program included the following tasks.

Task 1.1 – H2S Dominated Alkaline Sour Water Systems Subtask 1.1.1 – 50 psia H2S Partial Pressure Systems Subtask 1.1.2 – Parametric Effect of H2S Partial Pressure

Task 1.2 – Parametric Effects on Sour Water Corrosion Subtask 1.2.1 – Temperature Effects Subtask 1.2.2 – Chloride Effects

Task 1.3 – Effect of Hydrocarbon / Sour Water Mixtures Task 1.4 – Performance of Chemical Treatments in Sour Water Environments Task 1.5 – Development of Predict-SW Software Model

A description of each of these tasks is provided below. (Top)

TASK 1.1 - H2S-DOMINATED ALKALINE SOUR WATER SYSTEMS The first task in the Phase I program involved the evaluation of H2S-dominated alkaline sour water systems having moderate to high H2S partial pressures resulting in generally low to intermediate pH (7 to 9). These conditions are common in hydrocracker and hydrotreater applications. SUBTASK 1.1.1 – 50 PSIA H2S PARTIAL PRESSURE SYSTEMS. This work focused on the development of corrosion rate data in H2S-dominated systems with PH2S = 50 psia and a temperature of 130 F (55 C). Data were obtained at NH4HS concentrations ranging from 1% to 30 % and velocities in the lab flow loop ranging from 0 to 80 ft/sec. The results from this task provided the baseline data that were subsequently used for comparison to the balance of the program data that focused mainly on parametric effects. These data were obtained by SGSUS prior to the initiation of the program and were licensed to the program sponsors as part of the program participation agreements. (Top)

SUBTASK 1.1.2 – PARAMETRIC EFFECT OF H2S PARTIAL PRESSURE. THIS WORK INVOLVED A SERIES OF TESTS DESIGNED TO OBTAIN CORROSION RATE DATA IN H2S-DOMINATED SYSTEMS (AS IN SUBTASK 1.1.1) BUT AT ADDITIONAL LEVELS OF H2S PARTIAL PRESSURE INCLUDING PH2S = 30 PSIA, 100 PSIA AND 150 PSIA. NOTE DATA WERE

5

ALREADY AVAILABLE AT PH2S = 50 PSIA FROM SUBTASK 1.1.1. IN THIS SUBTASK, THE TEST TEMPERATURE WAS FIXED AT 130 F (55 C). THE NH4HS CONCENTRATIONS RANGED FROM 1% TO 20%.

TASK 1.2. – PARAMETRIC EFFECTS ON SOUR WATER CORROSION Task 1.1 was designed based on relatively simple matrices. This allowed for maximum experimental control to clearly establish trends in corrosion rate as an independent function of a limited number of variables, namely concentration, velocity and H2S partial pressure. This approach produced specific baseline conditions for H2S-dominated sour water conditions. Once this task was completed, and the fundamental trends identified at the baseline conditions, efforts were placed on defining the influences of other key variables that would be expected to impact alkaline sour water corrosion. Based on the existing knowledge of corrosion in sour water environments, the two key parametric effects that needed to be addressed for H2S-dominated systems were: (1) temperature and (2) chloride concentration. SUBTASK 1.2.1 – TEMPERATURE EFFECTS. This subtask examined the influence of solution temperature on the corrosion rate of selected materials in combination with velocity and NH4HS concentration. The two temperatures evaluated in this subtask were 190 F and 250 F (88 C and 121 C). Data for a third temperature (130 F, 55 C) were already available from the previous task. (Top) SUBTASK 1.2.2 – CHLORIDE EFFECTS. This subtask examined the influence of chloride concentration in the sour water on the corrosion rate of selected materials in combination with velocity and NH4HS concentration. Two chloride concentrations were evaluated in this subtask, namely 100 ppm and 1,000 ppm. Data for 0 ppm chloride were already available from the previous tasks. The chloride ion concentration was attained by the addition of HCl. (Top)

TASK 1.3 - EFFECT OF HYDROCARBON/SOUR WATER MIXTURES Oftentimes, refinery liquid hydrocarbon systems can become contaminated with sour water carryover. Additionally, sour water environments can also contain substantial amounts of hydrocarbons resulting in the presence of multiphase environments (oil / gas / water). Hence, this task examined the role of hydrocarbons in effectively inhibiting corrosion among the materials investigated in this program. (Top) The test matrix for this task incorporated both light and heavy hydrocarbons. Experiments covered hydrocarbon contents in the range of 10 to 98 vol%, with data for 0 vol% available from Task 1.1. All testing was conducted at a fixed temperature of 130 F (55 C) at two H2S partial pressures, 50 and 100 psia. The effect of shear stress on the corrosion performance was also examined with the heavy hydrocarbon. (Top)

6

TASK 1.4 - PERFORMANCE OF CHEMICAL TREATMENTS IN SOUR WATER ENVIRONMENTS This task evaluated the performance of two chemical treatments on ammonium bisulfide corrosion, namely ammonium polysulfide (APS) and imidazoline. Testing was conducted in selected environments to spot check their ability to reduce the corrosion rates, as well as compare the performance between the two chemical treatments evaluated. NH4HS concentrations ranged from 2 to 20 wt% with the majority of the testing at 8 wt%. The testing was conducted at 130 F (55 C), with a few tests conducted at 190 F (88 C) to evaluate the impact of temperature. Chemical treatments were conducted based on ppmv neat chemical and the dosages ranged from 50 to 500 ppmv. (Top)

TASK 1.5 - DEVELOPMENT OF PREDICT-SW

THE DATA COLLECTED FROM THE PROGRAM WERE USED TO DEVELOP A USER-FRIENDLY WINDOWS-BASED SOFTWARE TOOL CALLED PREDICT-SW. PREDICT PROVIDES A DATA SCREEN TO INPUT THE ENVIRONMENT, APPLICATION AND PROCESS STREAM VARIABLES. THIS INFORMATION IS USED TO DETERMINE THE EFFECTIVE SHEAR STRESS FOR THE PROCESS FLOW CONDITIONS. THIS SHEAR STRESS IS FURTHER USED WITH THE ENVIRONMENT VARIABLES TO PREDICT THE CORROSION RATE OF ALL MATERIALS EVALUATED IN THIS PROGRAM.

6. MATERIALS EVALUATED Fourteen materials were evaluated in this program, ranging from carbon steel to Alloy C-276. A list of the materials evaluated and their nominal compositions is provided in Table 1.

TABLE 1 Materials Evaluated and Nominal Compositions

Material %C %Cr %Mn %Mo %Ni %Cu %Fe %Cb %Co AISI 1018 0.15-0.20 - 0.6-0.9 - - - bal - -

Alloy 400 0.3 max - 2.0 max - 63-70 bal 2.5 max - -

AISI 410 0.015 max 11.5-13.5 1.0 max - - - bal - -

AISI 304 0.08 max 18-20 2.0 max - 8-10.5 - bal - -

AISI 316 0.08 max 16-18 2.0 max 2-3 10-14 - bal - -

Alloy 2205 0.03 max 21-23 2.0 max 2.5-3.5 4.5-6.5 - bal - -

Alloy 800 0.1 max 19-23 1.5 max - 30-35 0.75 max bal - -

Alloy 600 0.15 max 14-17 1.0 max - 72 min 0.5 max 6-10 - -

Alloy 20Cb-3 0.07 max 19-21 2.0 max 2-3 32-38 3-4 bal - -

Alloy 825 0.05 max 19.5-23.5 1.0 max 2.5-3.5 38-46 1.5-3.0 bal - -

Alloy 625 0.1 max 20-23 0.5 max 8-10 bal - 5.0 max 3.15-4.15 -

Alloy 2507 0.03 max 24-26 1.2 max 3-5 6-8 - bal - -

AL-6XN 0.03 max 20-22 2.0 max 6-7 23.5-25.5 - bal - -

Alloy C-276 0.02 max 14.5-16.5 1.0 max 15-17 bal - 4-7 - 2.5 max

7

Note: P, S and Si compositions not shown.

These materials were procured in round bars for machining of the flow-through coupons and in plate or sheet for preparation of the flat corrosion coupons evaluated in the static zone of the flow loop. (Top)

7. EXPERIMENTAL PROCEDURES Several aspects of the experimental procedures are described including a general description of the test facility, description of the flow-through and static corrosion coupons and ionic modeling conducted to determine the quantity of chemical ingredients to simulate the intended test conditions.

TEST FACILITY A clever test technique was required in this program due to the sensitivity of sour water environments to contamination by O2, and the high velocities required to obtain meaningful results that could be applied to the field. The test methodology, originally conceived by SGSUS, incorporated an internal gear pump driven by a magnetically coupled spinning attachment. The design allowed for continuous flow velocities up to 80 ft/sec through the flow-through corrosion coupons. In addition, since the fluid was constantly re-circulated within the test autoclave, the risk of O2 contamination was minimal. A schematic of the test facility is provided in Figure 1. In addition to the internal gear pump and drive mechanism, the test facility incorporated an internal thermocouple, water cooling coil to maintain constant temperature, gas inlet / outlet ports, and solution inlet and outlet ports. (Top)

8

Figure 1 – Test facility containing an internal gear pump.

9

CORROSION COUPONS Two types of mass-loss corrosion coupons were utilized in this program. The flow-through coupons subjected to the fluid passing through the pump consisted of ½-inch OD by ¾-inch long coupons with a 0.15-inch ID bore that was exposed to the fluid. Six flow-through coupons were evaluated in each test. Each coupon was electrically isolated from the others and from the coupon holder assembly using PTFE or PEEK washers that also contained a 0.15-inch diameter bore. The OD of the flow-through coupons was shielded from the environment with heat shrink tubing and further contained in an Alloy C-276 tubular coupon holder. A schematic of the flow-through corrosion coupon is provided in Figure 2. (Top)

Figure 2 – Flow-through corrosion coupon.

Corrosion coupons were also mounted in two stacks of three coupons underneath the pump to provide static (low flow) mass-loss corrosion rate results. These static zone corrosion coupons were also useful to verify the corrosion rate trends observed on the flow-through coupons. The static zone coupons were 2.0-inch long by 0.5-inch wide by 0.125-inch thick. A schematic of the static zone corrosion coupon is provided in Figure 3. (Top)

Figure 3 – Static zone corrosion coupon.

10

TEST PROTOCOL Solution preparation and subsequent saturation with H2S was conducted external to the test autoclave in a glass mixing vessel. There were several advantages to mixing the solution in a glass vessel. First, the solution could be monitored visually to ensure that it did not become cloudy, which was an inherent sign it was contaminated with O2. Another advantage of mixing the solution in a glass vessel was the ability to monitor the stages of H2S saturation. At the initiation of the H2S purge, the gas bubbles would be completely absorbed in the solution before they reached the surface of the solution. At saturation, the H2S gas bubbles floated to the surface without being absorbed. Another method used to gage the extent of H2S saturation was to monitor the solution expansion; again an advantage in the glass mixing vessel. Based on the ionic modeling, described below in more detail, the prescribed amount of distilled water was placed in the glass mixing vessel and purged with N2 for a minimum of 48 hours. Reagent grade NH4OH was then added to the mixing vessel in the prescribed quantity. Dilute NaCN solution (10 – 20 ppm) was injected into the mixing vessel to react with the remaining trace of O2. The solution was then saturated with ultra-pure H2S gas (< 1 ppm O2). The clarity of the solution and solution volume change were monitored to ensure that no O2 contamination occurred and the solution was completely saturated with H2S prior to transfer to the test autoclave. Earlier in the testing protocol, the flow-through coupons and flat coupons for the static zone were cleaned, measured, and pre-weighed. The stack of six flow-through coupons was assembled and all coupons were stored in a dessicator until being mounted onto the pump prior to the testing. The required pump speed (rpm) to attain the desired test velocity was determined by calibration with water. After the water calibration, the pump, autoclave head and associated fixturing were thoroughly dried. The test coupons (flow-though stack and flat coupons) were then mounted onto the pump and the autoclave head with pump assembly was assembled onto the autoclave body. The autoclave was pressure tested with N2 and then evacuated under vacuum / backfilled with N2 five times to deaerate the autoclave and pump assembly. The N2 was evacuated under vacuum and the autoclave was backfilled with H2S. A slight H2S pressure was put on the mixing vessel containing the H2S-saturated NH4HS solution to transfer the solution to the autoclave. H2S was added to the autoclave, and the autoclave was heated to test temperature. The pump flow was initiated slowly while the H2S partial pressure was adjusted to the desired pressure and the system reached equilibrium with the H2S. The pump speed was subsequently adjusted to the required rpm to attain the respective test velocity. Solution pH was monitored at test initiation, midway through the test, and at test conclusion. In the tests that contained oil fractions and chemical treatments, an impeller positioned in the bottom of the autoclave was used to mix the solution in the autoclave during the test. This was particularly important at high oil fractions and / or low test velocities to maximize the mixing of the solution. Each flow loop experiment was conducted for 48 hours at test conditions. After the final pH measurement, the pump and heat were switched off, the autoclave head pressure was bled down through a vent scrubber containing caustic, and the test solution was pushed out into a caustic

11

scrubber under N2 pressure. Caustic solution was pushed into the test autoclave and circulated through the pump at a low rpm. The caustic was pushed out and water was pushed into the autoclave to rinse. Once the water was removed, the autoclave was opened and the corrosion coupons were retrieved for cleaning and corrosion rate determination. A water calibration check was performed to verify the pump flow rate versus the rpm. The autoclave and pump assembly were then cleaned thoroughly prior to running the next test. (Top)

IONIC MODELING Ionic modeling was conducted to determine the appropriate amounts of distilled water and NH4OH required to achieve the desired NH4HS concentration in equilibrium with the desired H2S partial pressure and total vapor pressure at test temperature. This ionic modeling was conducted by SGSUS. The ionic package used was developed by Aspen Technology, Inc. – ASPEN / OLI Flowsheeter. The model contains a special “electrolytes package” and allows calculation of the phase behavior of an aqueous NH4HS solution in equilibrium with its vapor as a function of temperature, pressure and composition. It was critical to the design of suitable experiments to model actual process unit conditions, within the constraints of available laboratory equipment. To model a test at prescribed temperature, NH4HS concentration and H2S partial pressure, it calculates the amount of distilled water, concentrated NH4OH and the total pressure required. The model further predicts the pH at both ambient temperature and at the test temperature. In the experiments containing hydrocarbons, the modeling became significantly more complex, but was successfully used to design the experiments. The ionic modeling was used to study the influence of NH4HS concentration and H2S partial pressure on pH. The results are shown in Figure 4. As expected, the pH increases with an increase in NH4HS concentration and a decrease in H2S partial pressure. (Top)

Calculated pH for Variable NH4HS Concentrations(PpH-Calculation = Flash Pressure , TpH-Calculation = 77 F )

6.5

7.0

7.5

8.0

8.5

9.0

9.5

0 5 10 15 20 25 30

NH4HS Concentration (wt%)

pH

P* = 10 psia (Flash Pressure ~12.3 psia)P* = 20 psia (Flash Pressure ~22.3 psia)P* = 30 psia (Flash Pressure ~32.3 psia)P* = 40 psia (Flash Pressure ~42.3 psia)P* = 50 psia (Flash Pressure ~52.3 psia)

Figure 4 – Predicted pH using the ionic model.

12

The predicted pH values were compared to measured values of pH over the range of 1 to 30 wt% NH4HS at an H2S partial pressure of 50 psia. The measured pH values were in excellent agreement with the model. The results of this comparison are provided in Table 2. (Top)

TABLE 2 Predicted vs Measured pH

NH4HS Conc.

(%w) Predicted pH

at 77 F Measured pH

at RT 1 6.90 6.87, 6.92, 6.92, 6.92 2 7.18 7.13, 7.17, 7.22, 7.16 5 7.57 7.66, 7.59, 7.66, 7.63 8 7.83 7.85, 7.89, 7.87, 7.87 10 7.98 8.01, 8.05, 8.05, 8.08 15 8.32 8.36, 8.36, 8.38, 8.41 20 8.60 8.64, 8.64, 8.67, 8.63 30 9.11 9.13, 9.20, 9.03, 9.21

Note: 50 psia H2S partial pressure

8. RESULTS AND DISCUSSION

TASK 1.1 – H2S DOMINATED ALKALINE SOUR WATER SYSTEMS Task 1.1 included two subtasks, namely Subtask 1.1.1 – 50 psia H2S Partial Pressure Systems and Subtask 1.1.2 – Parametric Effect of H2S Partial Pressure. A summary of each subtask and respective results are presented in the following sections. (Top) SUBTASK 1.1.1 – 50 PSIA H2S PARTIAL PRESSURE SYSTEMS. The data development conducted under Subtask 1.1.1 was the most significant to the overall program, as it established the baseline corrosion rates from which data from the remaining tasks were compared. All of the data in this subtask were collected at 50 psia H2S partial pressure and at a temperature of 130 F (55 C). NH4HS concentration and flow-through coupon stack velocity were varied to produce isocorrosion diagrams of test velocity as a function of NH4HS concentration. Corrosion rates measured on the static corrosion coupons at the lowest test velocity were used to represent the corrosion rates at 0 ft/sec on the isocorrosion diagrams. Due to the profound impact these diagrams have already had on the refining industry, several of the program sponsors have been referring to these diagrams as the “Horvath Curves”, giving credit to the individual who conceived their creation, Richard J. Horvath. The test matrix for this subtask involved NH4HS concentrations that varied from 1 to 30 wt% and test velocities that varied from 10 to 80 ft/sec. H2S partial pressure and temperature were fixed at 50 psia and 130 F (55 C), respectively. A total of 32 independent tests were conducted and included six materials in each test. Obviously with the inclusion of fourteen materials in this program, decisions regarding which materials to include in a particular test were required.

13

Selection of the materials was based on the performance of the materials in other tests. For instance, once the material exhibited high corrosion rates at a particular NH4HS concentration and/or velocity, it was replaced with a more resistant material in tests conducted at more severe conditions. The resulting isocorrosion diagram for carbon steel is provided in Figure 5. The results indicate there are three discrete corrosion regimes. At low NH4HS concentrations, low corrosion rates are observed at low velocity. In this regime, corrosion rates increase only marginally with increased test velocity. At intermediate NH4HS concentrations, low to moderate corrosion rates are observed. However, these corrosion rates increase markedly with an increase in test velocity. At high NH4HS concentrations, moderate to high corrosion rates are observed at low velocity. As with the intermediate NH4HS concentrations, the corrosion rates also increase markedly with an increase in test velocity. (Top)


0

10

20

30

40

50

60

70

80

90

0 2 4 6 8 10 12 14 16


Velo

city

, ft/

sec


<5 mpy

5-20 mpy

20-50 mpy

50-100 mpy

100-200 mpy 200-300 mpy

>300 mpy


Figure 5 – Isocorrosion curve developed for carbon steel. The isocorrosion curves for the balance of the materials are provided in Appendix II. Trends in the isocorrosion curves were similar to those obtained on carbon steel. However, these curves gradually shifted upwards and towards the right with a corresponding increase in material resistance. Additional data obtained in this subtask can be obtained on the program website at the following URL: http://www.intercorr.com/multi/sourwater/download/finaldata.exe Several observations were made regarding alloy performance. Alloy 400 was not much better than carbon steel over the range of NH4HS and velocities investigated. Alloys 304 and 316 exhibited similar performance. Despite the current belief that Alloy 2205 was suitable to mitigate corrosion in REAC systems, it was found to corrode at intermediate and high NH4HS concentrations, especially at high velocities. Alloy 825 was also found to corrode at intermediate and high NH4HS concentrations, especially at high velocities. Alloy 625 was significantly better

14

than Alloy 825 at intermediate and high NH4HS concentrations. Alloy C-276 was found to be resistant to NH4HS at all conditions evaluated in this subtask. When using these data, it is important to note that the velocity plotted on the isocorrosion diagrams corresponds to the velocity through the 0.15-inch ID bore of the flow-through coupons. This velocity does not equate to the velocity of the multiphase fluids passing through typical REAC systems or other associated equipment. In order to make the appropriate comparison, the wall shear stress for the system in service must be calculated and correlated with a corresponding laboratory flow loop velocity that produces the same value of shear stress. For a 100% liquid system, the wall shear stress was calculated for 2-inch, 4-inch and 6-inch straight piping and compared to the shear stress in the 0.15-inch laboratory flow-through coupon. The results of these analyses are provided in Figure 6. (Top)

Shear Stress vs. Velocity Correlation100% Liquid Flow; 1 - 20 wt% NH4HS; 130 F; 50 psia H2S; Straight Pipe Sections

10

100

1000

10000

0 10 20 30 40 50 60 70 80 90

Velocity (fps)

Shea

r Str

ess

(Pa)

0.15" ID Pipe (Lab Apparatus)2" ID Pipe4" ID Pipe6" ID Pipe

Figure 6 – Correlation of shear stress and velocity in 100% liquid flow.

As shown in the figure, the shear stress produced in 2-inch to 6-inch piping can be achieved at much lower velocities in the small diameter flow-through coupons utilized in the laboratory flow loop. For example, the shear stress related to 17 ft/sec flow in a 4-inch pipe can be achieved at 10 ft/sec in the 0.15-inch ID coupon used in the laboratory study. Holding velocity constant produces a much higher shear stress in the laboratory coupon than in the 2 to 6-inch piping. While these relationships discussed above apply to 100% liquid flow, differences are also present in multiphase flow conditions. The PREDICT-SW software package developed as part of this program allows this shear stress – velocity correlation to be made with ease. (Top)

15

SUBTASK 1.1.2 – PARAMETRIC EFFECT OF H2S PARTIAL PRESSURE. The data collected in Subtask 1.1.1 were developed at an H2S partial pressure of 50 psia. Subtask 1.1.2 was intended to investigate the role of H2S partial pressure both below and above the 50 psia H2S data already collected. The test matrix for this subtask involved NH4HS concentrations that varied from 1 to 20 wt%, test velocities that varied from 10 to 80 ft/sec, and H2S partial pressures of 30, 100 and 150 psia. The test matrix included 28 tests, coupled with the results of the 32 tests conducted at 50 psia H2S in Subtask 1.1.1. The results of this subtask proved that H2S partial has a significant effect on the corrosion rate in H2S-dominated sour water systems. The results obtained for carbon steel are provided in Figures 7 through 9. (Top)

CARBON STEEL CORROSION RATE vs P[H2S] at Stagnant Conditions(48-hr laboratory tests: Velocity=stagnant, T=130 F)

0

20

40

60

80

100

120

140

160

180

200

220

240

260

280

300

20 30 40 50 60 70 80 90 100 110 120 130 140 150 160

P[H2S], psia

Cor

rosi

on R

ate,

mpy

1% NH4HS 2% NH4HS 5% NH4HS 8% NH4HS 10% NH4HS 20% NH4HS


Figure 7 – Influence of H2S partial pressure at near stagnant conditions on carbon steel. When the data are plotted as corrosion rate versus partial pressure H2S for varying NH4HS concentrations, the data yield a fan shaped appearance. Corrosion rate increases with a corresponding increase in H2S partial pressure. The increase is more pronounced at higher NH4HS concentrations. This is particularly evident for the near stagnant data shown in Figure 7. The results at 20 and 80 ft/sec are shown in Figures 8 and 9, respectively. Similar conclusions can be drawn from these data as were drawn from the stagnant data. (Top)

16

CARBON STEEL CORROSION RATE vs P[H2S] at 20 ft/sec(48-hr laboratory tests: Velocity=20 ft/s, T=130 F)

0

50

100

150

200

250

300

350

400

20 30 40 50 60 70 80 90 100 110 120 130 140 150 160

P[H2S], psia

Cor

rosi

on R

ate,

mpy

1% NH4HS 2% NH4HS 5% NH4HS 8% NH4HS 10% NH4HS


Figure 8 – Influence of H2S partial pressure at 20 ft/sec on carbon steel.

CARBON STEEL CORROSION RATE vs P[H2S] at 80 ft/sec(48-hr laboratory tests: Velocity=80 ft/s, T=130 F)

0

50

100

150

200

250

300

350

400

450

500

550

600

650

700

20 30 40 50 60 70 80 90 100 110 120 130 140 150 160

P[H2S], psia

Cor

rosi

on R

ate,

mpy 1% NH4HS

2% NH4HS 5% NH4HS 8% NH4HS 10% NH4HS


Figure 9 – Influence of H2S partial pressure at 80 ft/sec on carbon steel. The differing behavior between the near stagnant, 20 ft/sec and 80 ft/sec conditions produced a rule set incorporating three curves to correct the corrosion rate for the effect of H2S partial

17

pressure. The following equation can be used to correct the corrosion rate derived from the isocorrosion curve for the effect of H2S partial pressure. (Top)

( )5022−+= SHppFCRCR SHppiso

where: CR – calculated corrosion rate (mpy)

CRiso – predicted corrosion rate from the baseline isocorrosion curve

(50 psia H2S, 130 F with units of mpy)

SHppF2 – correction factor for H2S partial

pressure (mpy / psia) pp H2S – H2S partial pressure of environment (psia)

The H2S partial pressure correction factor for carbon steel is provided in Figure 10. The rules for the choice of curve are as follows:

• For velocities ≤ 1 ft/sec, use the 1 ft/sec curve • For velocities > 1 ft/sec and ≤ 20 ft/sec, interpolate between the respective curves • For velocities > 20 ft/sec and ≤ 80 ft/sec, interpolate between the respective curves • For velocities > 80 ft/sec, use the 80 ft/sec curve

CORRECTION FACTOR F{pp} for CARBON STEEL

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

0 5 10 15 20 25 30


F{pp

}, m

py/p

si

1 ft/sec 20 ft/sec 80 ft/sec


Figure 10 – H2S partial pressure correction factor for carbon steel. The H2S partial pressure correction curves for the balance of the materials are provided in Appendix III. As the alloy resistance improves, the shape of the 1, 20 and 80 ft/sec curves become similar. Hence, the only difference between the curves is the magnitude of the corrosion

18

rate correction. Additional data obtained in this subtask, such as the data presented in Figures 7 -9 for the remaining materials can be obtained on the program website at the following URL: http://www.intercorr.com/multi/sourwater/download/finaldata.exe. Several observations were made regarding the influence of H2S partial pressure. H2S partial pressure has a significant effect on the corrosion of carbon steel and all the alloys tested. The corrosion rates of carbon steel, 12 Cr, 304 stainless steel and Alloy 2205 at 100 and 150 psia H2S were 2 to 4 times their respective corrosion rates at 50 psia H2S. Alloys 2205, 800 and 600 have comparable corrosion resistance that is only marginally better than 304 stainless steel at high H2S partial pressure. This raised concern about the useful resistance of these alloys at high H2S partial pressure (100 to 150 psia) for NH4HS concentrations greater than 5 wt%.(Top) Alloys 825, 20Cb-3 and 625 are somewhat more corrosion resistant than Alloys 2205, 800 and 600 at high H2S partial pressure. The measured corrosion rates for Alloys 825, 20Cb-3 and 625 show a dramatic increase over the very low corrosion rates measured at 50 psia H2S. This raised concern about their useful resistance at high H2S partial pressure (100 to 150 psia) for NH4HS concentrations greater than 8 wt%. Alloy 20Cb-3 is more corrosion resistant than Alloy 825 at high H2S partial pressure, which was a reversal of performance observed at 50 psia H2S. Alloys 2507 and AL-6XN are far more corrosion resistant than Alloy 2205, and the previously mentioned nickel alloys at high H2S partial pressure. Measured corrosion rates for Alloys 2507 and AL-6XN were < 2 mpy up to 10 wt% NH4HS at high H2S partial pressure, but showed a marked increase in corrosion rate at 20 wt% NH4HS. Alloy C-276 was the best of class, with measured corrosion rates < 2 mpy up to 20 wt% NH4HS at 150 psia H2S partial pressure. Using the data obtained as part of Task 1.1, a general ranking of alloy resistance could be developed. This relative ranking was affectionately referred to as the “pony plot” in the program sponsor meetings. The materials were ranked individually from least resistant to most resistant. This “pony plot” is presented in Figure 11. (Top)

19

Leas

t Res

ista

nt

Mos

t Res

ista

nt


410 SS304 SS316 SS

Alloy 600

Alloy 825Alloy 625

Alloy 2507

Alloy C-276

Alloy 800Alloy 2205

Alloy 20Cb-3

Alloy AL-6XNLe

ast R

esis

tant

Mos

t Res

ista

nt


410 SS304 SS316 SS

Alloy 600

Alloy 825Alloy 625

Alloy 2507

Alloy C-276

Alloy 800Alloy 2205

Alloy 20Cb-3

Alloy AL-6XN

Figure 11 – Alloy resistance to NH4HS corrosion.

TASK 1.2 – PARAMETRIC EFFECTS ON SOUR WATER CORROSION Task 1.2 included two subtasks, namely Subtask 1.2.1 – Parametric Effect of Temperature and Subtask 1.2.2 – Parametric Effect of Chloride. A summary of each subtask and respective results are presented in the following sections. (Top) SUBTASK 1.2.1 – TEMPERATURE EFFECTS. The data collected in Subtask 1.2.1 were developed to investigate the role of temperature on the corrosion rate in H2S-dominated NH4HS environments. These data were used to develop rules for adjusting the corrosion rates obtained from the isocorrosion curves developed as part of Subtask 1.1.1. The test matrix for this subtask involved NH4HS concentrations that varied from 1 to 8 wt%, test velocities of 20 and 80 ft/sec, H2S partial pressures of 50 and 100 psia, and temperatures of 190 and 250 F (88 and 121 C). The matrix included 16 tests, coupled with the results of the tests conducted at 130 F (55 C) from Subtasks 1.1.1 and 1.1.2. The results of this subtask verified increased temperature produced a corresponding increase in corrosion rate as expected. The results also quantified the effect of temperature such that rules could be developed. The results obtained for carbon steel are provided in Figures 12 through 17. (Top)

20

EFFECT OF TEMPERATURE ON CORROSION RATE OF CARBON STEEL (48-hr laboratory tests: P[H2S]=50 psia, Velocity=20 ft/sec)

0

10

20

30

40

50

0 1 2 3 4 5 6


Cor

rosi

on R

ate,

mpy

130 F 190 F 250 F


Figure 12 – Corrosion rate versus NH4HS as a function of temperature at 20 ft/sec

and 50 psia H2S.

CARBON STEEL CORROSION RATE vs TEMPERATURE at 20 ft/sec(48-hr laboratory tests: P[H2S]=50 psia, Velocity=20 ft/sec)

0

10

20

30

40

50

60

70

80

90

100

120 130 140 150 160 170 180 190 200 210 220 230 240 250 260

Temperature, oF

Cor

rosi

on R

ate,

mpy

1% NH4HS 2% NH4HS 5% NH4HS


Figure 13 – Corrosion rate versus temperature as a function of NH4HS at 20 ft/sec and 50 psia H2S. (Top)

21

CARBON STEEL CORROSION RATE vs TEMPERATURE at 20 ft/sec

(48-hr laboratory tests: P[H2S]=100 psia, Velocity=20 ft/sec)

0

25

50

75

100

125

150

175

200

225

250

275

300

120 130 140 150 160 170 180 190 200 210 220 230 240 250 260

Temperature, oF

Cor

rosi

on R

ate,

mpy

2% NH4HS 8% NH4HS


Figure 14 – Corrosion rate versus temperature as a function of NH4HS at 20 ft/sec

and 100 psia H2S. (Top)

EFFECT OF TEMPERATURE ON CORROSION RATE OF CARBON STEEL (48-hr laboratory tests: P[H2S]=50 psia, Velocity=80 ft/sec)

0

10

20

30

40

50

60

70

80

90

100

0 1 2 3 4 5 6


Cor

rosi

on R

ate,

mpy

130 F 190 F 250 F


Figure 15 – Corrosion rate versus NH4HS as a function of temperature at 80 ft/sec


22

CARBON STEEL CORROSION RATE vs TEMPERATURE at 80 ft/sec

(48-hr laboratory tests: P[H2S]=50 psia, Velocity=80 ft/sec)

0

10

20

30

40

50

60

70

80

90

100

120 130 140 150 160 170 180 190 200 210 220 230 240 250 260

Temperature, oF

Cor

rosi

on R

ate,

mpy

1% NH4HS 2% NH4HS 5% NH4HS


Figure 16 – Corrosion rate versus temperature as a function of NH4HS at 80 ft/sec and 50 psia H2S. (Top)

CARBON STEEL CORROSION RATE vs TEMPERATURE at 80 ft/sec(48-hr laboratory tests: P[H2S]=100 psia, Velocity=80 ft/sec)

0

25

50

75

100

125

150

175

200

225

250

275

300

325

350

375

400

120 130 140 150 160 170 180 190 200 210 220 230 240 250 260

Temperature, oF

Cor

rosi

on R

ate,

mpy

2% NH4HS 8% NH4HS


Figure 17 – Corrosion rate versus temperature as a function of NH4HS at 80 ft/sec


23

Additional data obtained in this subtask, such as the data presented in Figures 12 - 17 for the remaining materials can be obtained on the program website at the following URL: http://www.intercorr.com/multi/sourwater/download/finaldata.exe. Several observations were made regarding the influence of temperature on the corrosion in H2S-dominated NH4HS environments. First, the corrosion rates of carbon steel and all alloys investigated in this program increased with increasing temperature. Second, the effect of temperature on the corrosion rate of carbon steel and 12 Cr was greatest at low NH4HS concentrations, and diminished as the NH4HS concentration increased. Lastly, temperature appeared to have less effect on corrosion than NH4HS concentration, velocity, and H2S partial pressure. These data provided in Figures 12 – 17 were used to develop a rule set to correct the corrosion rate of carbon steel for the effect of temperature. Equation 1 which was used to correct for the effect of H2S partial pressure was appended to account for the effect of temperature. (Top)

( ) ( )1305022−+−+= TempFSHppFCRCR TempSHppiso

where: TempF – correction factor for temperature (mpy / °F)

Temp – temperature of the environment (°F)

The temperature correction factor for carbon steel is provided in Figure 18. The rules for the choice of curve are as follows:

• For velocities ≤ 1 ft/sec, use the 1 ft/sec curve • For velocities > 1 ft/sec and ≤ 20 ft/sec, interpolate between the respective curves • For velocities > 20 ft/sec and ≤ 80 ft/sec, interpolate between the respective curves • For velocities > 80 ft/sec, use the 80 ft/sec curve

24

CORRECTION FACTOR F{temp} for CARBON STEEL

0.000

0.050

0.100

0.150

0.200

0.250

0.300

0.350

0.400

0 5 10 15 20 25 30


F{te

mp}

, mpy

/deg

F



Figure 18 – Temperature correction factor for carbon steel. The temperature correction curves for the balance of the materials are provided in Appendix IV. As the alloy resistance increases, the peak in the correction curve observed at low NH4HS concentration for carbon steel moves progressively to higher NH4HS concentrations. For Alloys 2507, AL-6XN and C-276, this peak (if present) has moved beyond the 30 wt% NH4HS concentration evaluated in this program. (Top) SUBTASK 1.2.2 – CHLORIDE EFFECTS. The data collected in Subtask 1.2.2 were developed to investigate the role of chloride on the corrosion rate in H2S-dominated NH4HS environments. The test matrix for this subtask involved NH4HS concentrations that varied from 1 to 15 wt%, test velocities of 20 and 80 ft/sec, an H2S partial pressure of 50 psia, temperature of 130 F (55 C), and chloride concentrations of 100 and 1,000 ppm from additions of HCl. The test matrix included 16 tests, coupled with the results of the tests conducted with 0 ppm chloride from Subtask 1.1.1. (Top) The addition of HCl to attain the desired chloride concentration was anticipated to result in a decrease in pH that might result in increased corrosivity. However, these NH4HS solutions are highly buffered and the pH change caused by the addition of HCl was negligible. Table 3 provides the results of the ionic modeling predictions over the range of 100 to 1,000 ppm chloride. (Top)

TABLE 3 Predicted pH Change with HCl Additions

NH4HS Conc.

(wt%) Cl− Concentration

(ppm) Predicted pH

at 77 F pH Change from 0 ppm Cl−

(ppm) 1 0 6.90 -

25

1 100 6.90 0 1 1,000 6.89 0.01

Note: Predictions based on 50 psia H2S partial pressure At 5 wt% NH4HS and higher, no pH change was predicted up to 1,000 ppm Cl−.

The results obtained on carbon steel at 20 and 80 ft/sec are provided in Figures 19 and 20, respectively. (Top)

EFFECT OF CHLORIDE ON CORROSION RATE OF CARBON STEEL (48-hr laboratory tests: P[H2S]=50 psia, Velocity=20 ft/sec)

0

50

100

150

200

250

0 5 10 15


Cor

rosi

on R

ate,

mpy

0 ppm Cl- 100 ppm Cl- 1,000 ppm Cl-


Figure 19 – Effect of Cl− on the corrosion rate of carbon steel at 20 ft/sec.

The data obtained at 100 ppm chloride were compared to those obtained at 1,000 ppm chloride for carbon steel and the remaining alloys investigated in this subtask. No differences in corrosion rate were observed that could be attributed to the presence of chloride in the sour water. Comparison of these data at 100 and 1,000 ppm chloride with data collected in Subtask 1.1.1 containing 0 ppm chloride agreed in most cases up to 10 wt% NH4HS. However, at 15 wt% NH4HS, the 0 ppm chloride sour water exhibited higher corrosion rates than sour waters containing 100 and 1,000 ppm chlorides, particularly at the test velocity of 80 ft/sec. A number of tests were repeated to confirm the corrosion rates determined in the presence of chlorides. These tests yielded similar results. In many of the cases, the corrosion rates exceed 100 mpy on carbon steel giving confidence that O2 was not present in the environment that would lead to the formation of polysulfide, which tends to inhibit the corrosion rates. Hence, based on the above findings and group discussion at the meetings, the corrosion rate determined in PREDICT-SW will not be adjusted for the effect of chloride. (Top)

26

EFFECT OF CHLORIDE ON CORROSION RATE OF CARBON STEEL (48-hr laboratory tests: P[H2S]=50 psia, Velocity=80 ft/sec)

0

50

100

150

200

250

300

350

400

450

500

0 5 10 15


Cor

rosi

on R

ate,

mpy

0 ppm Cl- 100 ppm Cl- 1,000 ppm Cl-


Figure 20 – Effect of Cl− on the corrosion rate of carbon steel at 80 ft/sec.

Another point of interest in this subtask was the behavior between AISI 304 and AISI 316 in the presence of chlorides. No differences were observed in the corrosion behavior in the base data with 0 ppm chloride. This was also observed in this subtask. Neither alloy appeared more favorable from the standpoint of corrosion in the presence of chlorides. (Top) Additional data obtained in this subtask, such as the data presented in Figures 19 and 20, for the remaining materials can be obtained on the program website at the following URL: http://www.intercorr.com/multi/sourwater/download/finaldata.exe.

TASK 1.3 – EFFECT OF HYDROCARBON / SOUR WATER MIXTURES The data collected in Task 1.3 were developed to investigate the role of hydrocarbon on the corrosion rate in NH4HS environments. It is known that the presence of hydrocarbon can have an inhibiting effect, thereby reducing the corrosion rate. The decrease in corrosion rate results from reducing the time the sour water corrosive phase contacts the metal surface, as the wetability of the hydrocarbon restricts the ability of the sour water to contact and corrode the metal surface. The presence of hydrocarbon may act to increase the corrosion rate as a result of increasing the shear stress produced at the metal surface. This behavior obviously becomes a greater possibility as the density and viscosity of the hydrocarbon increase. To investigate these phenomena, the test matrix for this task incorporated both a light and heavy hydrocarbon. Experiments covered hydrocarbon contents in the range of 10 to 98 vol%, with 0 vol% data available from Task 1 for comparison. All testing was conducted at a fixed

27

temperature of 130 F (55 C) at two H2S partial pressures, 50 and 100 psia. The effect of shear stress on the corrosion performance was also examined with the heavy hydrocarbon. Exxsol D80 was chosen as the light hydrocarbon to study in this task. Exxsol D80 is a dearomatized aliphatic performance fluid (solvent). The heavy hydrocarbon chosen for study was Tufflo 1200, which is a naphthenic process oil. Selected properties and distillation data for the two hydrocarbons are provided in Tables 4 and 5, respectively. (Top)

TABLE 4 Properties of the Exxsol D80 and Tufflo 1200 Hydrocarbons

Property Exxsol D80 Tufflo 1200 Distillation Range 407 – 455 F 557 – 1001 F

Flash Point 181 F 430 F Specific Gravity @ 60 F 0.798 0.934

API Gravity 40 20

TABLE 5 ASTM D-86 Distillation Data for Exxsol D80 and Tufflo 1200

Exxsol D80 Tufflo 1200 Initial Boiling: 407 F 557 F

5%: 412 F 726 F 10%: 414 F 777 F 50%: 423 F 862 F 90%: 442 F 932 F 95%: 448 F 950 F

Final Boiling: 455 F 1001 F In addition to the properties above, density and viscosity were measured over the temperature range of 100 to 250 F (38 to 121 C). These data were collected to provide accurate properties for use in the flow modeling conducted as part of the experiments and to provide a set of properties for use in PREDICT-SW, if desired by the user. These data are provided in Figures 21 and 22, respectively. (Top)

28

Specific Gravities of Hydrocarbons

700

750

800

850

900

950

1000

80 100 120 140 160 180 200 220 240 260

Temperature (F)

Spec

ific

Gra

vity

(kg/

m3 )

Exxsol D80Tufflo 1200


Figure 21 –Specific gravity of the light and heavy hydrocarbons as a function of temperature. (Top)

Absolute Viscosities of Hydrocarbons

0.001

0.010

0.100

1.000

10.000

80 100 120 140 160 180 200 220 240 260

Temperature (F)

Abs

olut

e Vi

scos

ity (P

oise

)

Exxsol D80Tufflo 1200


Figure 22 –Viscosities of the light and heavy hydrocarbons as a function of temperature. (Top)

29

The corrosion data collected for carbon steel using various volume fractions of the heavy hydrocarbon at 50 and 100 psia H2S are plotted in Figure 23. These tests were conducted with 8 wt% NH4HS and a temperature of 130 F (55 C). As shown, substantial protection from the heavy hydrocarbon was not achieved until the hydrocarbon content exceeded 25 vol%. A strong dependence on the H2S partial pressure was also observed in the data collected at 0 and 10 vol% heavy hydrocarbon. This further adds to the conclusion that increased H2S partial pressure results in a dramatic increase in the corrosion rate.

Effect of Heavy Hydrocarbon - CS(8% NH4HS, 130 F, ττττ = 510 Pa (or as noted) [50 ft/sec water], Tufflo 1200)

0

50

100

150

200

250

0 10 20 30 40 50 60 70 80 90 100

Hydrocarbon Content (vol.%)

Cor

rosi

on R

ate

(mpy

)

50 psia H2S

100 psia H2S

τ = 1,063 Paτ = 1,754 Paτ = 1,645 Pa


Figure 23 – Effect of heavy hydrocarbon on the corrosion rate of carbon steel. (Top)

Similar data collected with the light hydrocarbon are shown in Figure 24. The data point collected at 40 vol% and 100 psia H2S appeared higher than the expected trend. Unfortunately, there were no other data points between 10 and 40 vol% to better illustrate the probable relationship. Despite these drawbacks in the data set on the light hydrocarbon, the findings still support the effect of the increased H2S partial pressure and inhibiting effect of the hydrocarbon at increased hydrocarbon contents. Due to the higher density and viscosity of the heavy hydrocarbon, the effect of shear stress could also be studied. The shear stress associated with 100% sour water at 50 ft/sec was 510 Pa. With the inclusion of the heavy hydrocarbon in the liquid, shear stress equivalent to four times this value could be attained (i.e. 2,040 Pa). The results obtained on carbon steel are provided in Figure 25. In all cases examined, the higher shear stress produced a higher corrosion rate at a given set of environmental conditions. The magnitude of the corrosion rate difference was also a function of the corrosivity of the environment. At lower hydrocarbon contents, the environment

30

is primarily sour water and the increased shear stress has a greater influence on increasing the corrosion rate. At higher hydrocarbon contents, less sour water is available and the corrosion rate difference between the two shear stresses was small. At 98 vol% hydrocarbon, the difference in corrosion rate was negligible. (Top)

Effect of Light Hydrocarbon - CS(8% NH4HS, 130 F, ττττ = 510 Pa [50 ft/sec water], Exxsol D80)

0

50

100

150

200

250

0 10 20 30 40 50 60 70 80 90 100


Cor

rosi

on R

ate

(mpy

)

50 psia H2S

100 psia H2S


Figure 24 – Effect of light hydrocarbon on the corrosion rate of carbon steel. (Top)

Effect of Shear Stress - CS(8% NH4HS, 130 F, 100 psia H2S, Tufflo 1200)

0

50

100

150

200

250

300

350

400

0 10 20 30 40 50 60 70 80


Cor

rosi

on R

ate

(mpy

)

510 Pa2,040 Pa


Figure 25 - Effect of shear stress on the corrosion rate of carbon steel. (Top)

Suspect Point

31

Additional data for the remaining materials can be obtained on the program website at the following URL: http://www.intercorr.com/multi/sourwater/download/finaldata.exe. To model the influence of the hydrocarbon, efficiencies of protection were determined for each data point. This is a similar treatment used to evaluate inhibitor effectiveness. Using the data collected for carbon steel, irrespective of the test environments or light versus heavy hydrocarbon, efficiencies were calculated as referenced to the 0 vol% hydrocarbon results. As shown in Figure 26, efficiency of protection increased rapidly up to 25 vol% beyond which the efficiency increased gradually until approaching almost complete protection at 98 vol% hydrocarbon.


0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70 80 90 100


Effic

ienc

y (%

)



Figure 26 – Influence of hydrocarbon on the corrosion rate of carbon steel. (Top)

A similar analysis was performed on the balance of the materials. These results are provided in Appendix V. The results obtained for 12 Cr are also shown in Figure 27. The main difference observed between carbon steel and 12 Cr (as well as the balance of materials) was that substantial protection from the presence of hydrocarbon was experienced with 12 Cr at hydrocarbon contents of only 10 vol% as compared to 25 vol% with carbon steel. Another interesting finding became evident from the test results with 12 Cr, the balance of the alloys and to some extent carbon steel. That is, the corrosion rate data for mixtures with the light and heavy hydrocarbons essentially overlayed each other and separate trends were not observed.

Suspect Point in Figure 24

32

Influence of Hydrocarbon - 12 Cr

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70 80 90 100


Effic

ienc

y (%

)



Figure 27 – Influence of hydrocarbon on the corrosion rate of 12 Cr. (Top)

These curves of efficiency versus hydrocarbon content were used to correct the sour water only corrosion rates for the presence of hydrocarbon/sour water mixtures in PREDICT-SW. The flow regime was also an important consideration. The benefit of the presence hydrocarbon was only taken for turbulent flow regimes that would cause the pipe/tube wall inner surfaces to be continuously wetted by hydrocarbon. For laminar, stratified or wave flow regimes, some portion of the pipe/tube internal surface would be wetted by the sour water phase alone, and not by the hydrocarbon phase. In these cases, no benefit for the presence of the hydrocarbon was taken. To account for the hydrocarbon, Equation 2 was further modified as follows:

( ) ( )( )( )hcTempSHppiso TempFSHppFCRCR η−−+−+= 11305022

(3)

where: hcη – efficiency of protection from the hydrocarbon content

hcη = 0 for Horizontal – Stratified Flow = 0 for Horizontal – Wave Flow = 0 for Horizontal – Laminar Flow = 0 for Vertical – Laminar Flow

33

Since only six materials were tested in this task, efficiencies for the influence of hydrocarbon on the remaining eight materials were decided as shown in Table 6. The selections were based on material class as well as behavior observed in the balance of the program data. (Top)

TABLE 6 Efficiency Rule Assignments for Effect of Hydrocarbon

Task 1.3 Data Set Efficiency Rule Assignment Carbon Steel Carbon Steel

12 Cr 12 Cr AISI 304 AISI 304 and 316

Alloy 2205 Alloy 2205 Alloys 2507 and 800 Alloys 400, 800, 600, 825, 625, 2507, 2507, AL6XN, C-276

TASK 1.4 – PERFORMANCE OF CHEMICAL TREATMENTS IN SOUR WATER ENVIRONMENTS The data collected in Task 1.4 were developed to investigate the role of chemical treatments on reducing the corrosion rate in NH4HS environments. Two chemical treatments were investigated, namely ammonium polysulfide (APS) and imidazoline. Testing was conducted in selected environments to spot check the ability to reduce the corrosion rates, as well as compare the performance between the two chemical treatments. The test matrix incorporated NH4HS concentrations that ranged from 2 to 20 wt% with the majority of the testing at 8 wt%. The testing was conducted at 130 F (55 C), with a few tests conducted at 190 F (88 C) to evaluate the impact of temperature. Chemical treatment dosages were based on ppmv neat chemical, and the dosages ranged from 50 to 500 ppmv. Similar to the analysis of the hydrocarbon results discussed earlier, all of the data collected in this task were reviewed on the basis of an efficiency of protection. In terms of the data collected with the ammonium polysulfide (APS), efficiency was observed to vary with test velocity. As shown in Figure 28, data collected showed that APS offered more protection at 20 ft/sec than at 80 ft/sec. The curves were extrapolated to 1,000 ppmv for use in PREDICT-SW, however no additional benefit beyond 500 ppmv was given. These results indicate that the APS can form a more stable protective film on the metal surface at lower velocities. At higher velocities (shear stress), this film becomes unstable resulting in reduced protection. (Top)

34

APS

0

10

20

30

40

50

60

70

80

90

100

0 100 200 300 400 500 600 700 800 900 1000


Effic

ienc

y (%

)

80 ft/sec

20 ft/sec

APS PREDICT-SW RulesFor velocities < 20 ft/sec, use 20 ft/sec curveFor velocities > 80 ft/sec, use 80 ft/sec curveFor velocities between 20 and 80, interpolate between the respective curves


Figure 28 – Efficiency of protection versus APS dosage. (Top)

The following rules were established to determine the efficiency of protection when using APS:

• For velocities ≤ 20 ft/sec, use the 20 ft/sec curve • For velocities > 20 ft/sec and < 80 ft/sec, interpolate between the respective curves • For velocities ≥ 80 ft/sec, use the 80 ft/sec curve

The data collected using imidazoline were analyzed in a similar fashion. These data did not follow the same trend with test velocity that was observed with APS. Although many of the tests conducted at 80 ft/sec resulted in higher efficiencies of protection than those tests run at 20 ft/sec, trends were not as clear as the those observed with APS. This behavior was explained on the basis of mixing. Recall that imidazoline is oil soluble and these tests were conducted in the absence of hydrocarbon. The inhibitors were diluted with 0.5 vol% isopropyl alcohol to increase the solubility, but still did not make them water soluble. Hence, tests conducted at 80 ft/sec provided better mixing of the environment and assisted in increasing the contact of the inhibitor with the metal surfaces. At lower test velocities, separation of the inhibitor probably occurred and resulted in decreased inhibition. For this reason, the upper and lower bounds to the data (regardless of velocity) were used to represent a most and least conservative guideline (see Figure 29). Use of these curves would be based on the applicable flow regime, considering the ability of the imidazoline to wet the entire internal surface of the pipe/tube. The choice of curve was set as follows: (Top)

• Horizontal – Stratified Flow, Horizontal – Wave Flow, Horizontal – Laminar Flow and Vertical – Laminar Flow use Efficiency = 0

• Horizontal - Annular Mist Flow & Vertical – Annular Flow, use most conservative curve • Other flow regimes, use least conservative curve.

35

Imidazoline

0

10

20

30

40

50

60

70

80

90

100

0 100 200 300 400 500 600 700 800 900 1000


Effic

ienc

y (%

)

Most Conservative

Least Conservative

Imidazoline PREDICT-SW RulesFor Horizontal-Stratified, Horizontal-Wave, Horizontal-Laminar and Vertical-Laminar Use Efficiency = 0For Horizontal-Annular Mist and Vetrical Annular, use most conservative curveFor all other flow conditions, use least conservative curve


Figure 29 – Efficiency of protection versus imidazoline dosage. (Top)

To account for chemical treatment, Equation 3 was further modified as follows:

( ) ( )( )( )(hcTempSHppiso TempFSHppFCRCR η −−−+−+= 111305022

(4)

where: chemη – efficiency of protection from either APS or imidazoline

Since carbon steel was the only material evaluated in this task, the ability of the chemical treatments to provide corrosion protection on the remaining 13 materials is unknown. Hence, only the corrosion rate of carbon steel is corrected in Predict-SW for the presence of APS or imidazoline. (Top)

36

TASK 1.5 – DEVELOPMENT OF PREDICT-SW SOFTWARE MODEL

THE FINAL TASK OF THIS PROGRAM WAS TO DEVELOP A WINDOWS-BASED SOFTWARE TOOL CALLED PREDICT-SW. PREDICT PROVIDES A DATA SCREEN FOR INPUT OF THE ENVIRONMENT, APPLICATION AND PROCESS STREAM VARIABLES. THIS INFORMATION IS THEN USED TO ESTIMATE THE CORROSION RATE ON THE FOURTEEN MATERIALS EVALUATED IN THIS PROGRAM USING THE FOLLOWING SEQUENCE:

Calculate effective shear stress

from process flow conditions

⇓⇓⇓⇓

CONVERT THE FIELD SHEAR

STRESS INTO AN

EQUIVALENT LAB FLOW LOOP VELOCITY

⇓⇓⇓⇓ Using lab velocity

and NH4HS concentration,

predict corrosion rates

for all materials using the respective

isocorrosion diagrams

⇓⇓⇓⇓ Correct corrosion

rates for the effect of H2S partial pressure,

temperature, hydrocarbon content

37

and chemical treatment using Eq. 4

The input screen is shown in Figure 30. The environment section provides input for the total pressure, H2S content, temperature, NH4HS concentration, oil type, chemical type and dosage. The choices for oil type include user defined, light and heavy. If user defined is selected, the user must enter values for the density and viscosity of the hydrocarbon in the process stream flow section at the bottom right section of the input screen. If light or heavy is selected, density and viscosity are calculated from the Exxsol D80 and Tufflo 1200 properties as a function of the temperature (recall Figures 21 and 22). These values are updated in the process flow section and are blocked from further input, since they are calculated values. (Top) The choices for chemical treatment type are none, APS and imidazoline. If none is selected, the chemical dosage input section is removed from the input data screen. If either APS or imidazoline are selected, the user must input the ppmv of neat chemical based on the volume of sour water in the process stream. Even though the imidazoline is oil soluble, data collected in this program was in 100% sour water and hence the rules were developed using this assumption. The user then completes the application section, which includes input for tube / pipe inner diameter, corrosion allowance, design life and pipe roughness. Choices for pipe roughness include new, lightly corroded and heavily corroded, corresponding to roughness values of 0.0015, 0.01 and 0.04 inches, respectively. A custom roughness dialogue box is provided if the user desires to incorporate a custom roughness value. (Top)

38

Figure 30 – Data input screen for Predict-SW. Next, the user inputs the process stream conditions including the piping configuration and type of flow. Configurations include straight, 3-D bend, 90° elbow and weld protrusion. These configurations are used to amplify the effective shear stress determined in the flow modeling, which assumed straight sections. These amplification factors, derived from the literature [6, 7], are provided in Table 7. (Top)

39

TABLE 7 Shear Stress Amplification Factors for Various Piping Configurations

Configuration Shear Stress Amplification Factor Straight 1

3-D Bend 1.5 90° Elbow 2.6

Weld Protrusion (5 mm) 3.5 In the event the user desires to conduct in-house analysis to determine the effective shear stress, the custom shear stress box can be selected. This disables the variables used by Predict-SW to determine shear stress and opens two additional boxes for input of the shear stress and flow regime. Irrespective of the shear stress option chosen, the user must input the process flow rates for gas, sour water and liquid hydrocarbon in the process stream flow section of the input screen. These values are used to determine the liquid hydrocarbon content used in Equation 4 to adjust the corrosion rate. If a custom shear stress is used, the calculate button can now be clicked. If the user desires Predict-SW to calculate the shear stress, the user must first input the specific gravity of air, densities of sour water and hydrocarbon as well as the viscosities of air, sour water and hydrocarbon. If light or heavy was chosen as part of the environment input, the density and viscosity of the hydrocarbon is calculated, and user input in these boxes is restricted. Once these values are input, the calculate button can be clicked and Predict-SW will run the analysis. This JIP effort demonstrated the importance of understanding the mechanical forces in plant systems where multiphase environments and various flow regimes are involved. The flow modeling embedded in Predict-SW incorporates accurate mapping of different flow regimes and characterization of corresponding hydrodynamic parameters for multiphase flowing systems. The model utilizes widely known flow maps from Taitel-Dukler and Mendhane et al and provides the end user the ability to assess pressure drops, liquid hold up, dimensionless factors and wall shear stress for both vertical and horizontal, single-phase and multiphase fluid systems [8-20]. The results window shown in Figure 31 displays the corrosion rates for the fourteen materials evaluated in this program. Adjacent to each corrosion rate value is a box colored either green or red. A green box indicates the material’s corrosion rate was deemed acceptable based on the user inputs for corrosion allowance and design life. A red box indicates the corrosion rate exceeds a rate allowable for the corrosion allowance and design life provided. In addition to the corrosion rate results, the window displays the flow regime, calculated shear stress, superficial liquid and gas velocities and equivalent lab flow loop velocity (100% sour water). A comments section is also provided to document the user assumptions, scenarios, etc. These comments are then recorded for printing or saving of the consultation. (Top)

40

Figure 31 – Results screen for Predict-SW. The results from Predict-SW were compared to the data on the isocorrosion curves for velocities of 1, 20 and 80 ft/sec. Since the isocorrosion diagrams provide the reference point to any subsequent scaling of the corrosion rate conducted, it was extremely important that this starting point was accurate. These data comparisons are provided in Appendix VI. (Top) In addition to the baseline isocorrosion curve predictions, Predict-SW was checked for several of the experiments where H2S partial pressure and temperature varied from the baseline 50 psia H2S and 130 F values. The Table 8 presents the measured and Predict-SW calculated values for these

41

experiments. As shown, the Predict-SW calculated corrosion rates were in good agreement with the experimental results.

TABLE 8 Predict-SW Estimates on Tests with Varying H2S and Temperature

Case

Material

NH4HS (wt%)

H2S (psia)

Temp. (F)

Velocity (ft/sec)

Measured CR (mpy)

Predicted CR (mpy)

1 CS 1 50 250 80 53 52 2 CS 1 100 130 80 48 48 3 304 1 100 130 80 2.8 3.4 4 304 1 150 130 80 6.7 6.3 5 CS 2 100 190 80 80 84 6 CS 2 30 130 20 2 2.5 7 CS 2 100 130 20 58 43 8 CS 2 100 130 80 71 65 9 CS 2 150 130 20 73 72

10 CS 2 150 130 80 104 103 11 304 2 150 130 80 11 10.3 12 CS 5 30 130 10 5.6 7.1 13 CS 5 100 130 20 83 76 14 304 5 30 130 80 0.8 <1 15 304 5 150 130 80 19 18.8 16 CS 8 100 190 80 332 358 17 CS 8 100 190 20 200 178 18 CS 8 100 130 20 194 170 19 CS 10 100 130 20 306 246 20 CS 10 100 130 80 438 572

FIELD EXPERIENCE. During the course of the program, the program sponsors were requested to bring field experience to the meetings for discussion. This experience was extremely beneficial in assisting to understand and confirm the effect of the independent variables studied in the course of the program. Ten field cases were presented from the user group over the course of the program. The data from these cases were sanitized to remove company information. The cases were consolidated and summarized in Appendix V. (Top)

42

9. CONCLUSIONS The following conclusions were drawn based on the results of this program.

1. Based on the isocorrosion diagrams developed, there are three discrete corrosion regimes. (a) At low NH4HS concentrations, low corrosion rates are observed at low velocity that increase only marginally with increased test velocity. (b) At intermediate NH4HS concentrations, low to moderate corrosion rates are observed. However, these corrosion rates increase markedly with an increase in test velocity. (c) At high NH4HS concentrations, moderate to high corrosion rates are observed at low velocity. As with the intermediate NH4HS concentrations, the corrosion rates also increase markedly with an increase in test velocity.

2. Alloy 400 was not much better than carbon steel over the range of NH4HS and velocities investigated.

3. Alloys 304 and 316 exhibited similar performance.

4. Despite the current belief that Alloy 2205 was suitable to mitigate corrosion in REAC systems, it was found to corrode at intermediate and high NH4HS concentrations, especially at high velocities.

5. Alloy 825 was also found to corrode at intermediate and high NH4HS concentrations, especially at high velocities.

6. Alloy 625 was significantly better than Alloy 825 at intermediate and high NH4HS concentrations.

7. Alloy C-276 was found to be the most resistant material to NH4HS corrosion evaluated in this program.

8. H2S partial pressure has a significant effect on the corrosion of carbon steel and all the alloys tested. The corrosion rates of carbon steel, 12 Cr, 304 stainless steel and Alloy 2205 at 100 and 150 psia H2S were 2 to 4 times their respective corrosion rates at 50 psia H2S.

9. Alloys 2205, 800 and 600 have comparable corrosion resistance that is only marginally better than 304 stainless steel at high H2S partial pressure.

10. Alloys 825, 20Cb-3 and 625 are somewhat more corrosion resistant than Alloys 2205, 800 and 600 at high H2S partial pressure. The measured corrosion rates for Alloys 825, 20Cb-3 and 625 at higher H2S partial pressures show a dramatic increase over the very low corrosion rates measured at 50 psia H2S.

11. Alloy 20Cb-3 is more corrosion resistant than Alloy 825 at high H2S partial pressure, which was a reversal of performance observed at 50 psia H2S.

43

12. Alloys 2507 and AL-6XN are far more corrosion resistant than Alloy 2205 and the previously mentioned nickel alloys at high H2S partial pressure. Measured corrosion rates for Alloys 2507 and AL-6XN were < 2 mpy up to 10 wt% NH4HS at high H2S partial pressure, but showed a marked increase in corrosion rate at 20 wt% NH4HS.

13. Alloy C-276 was the best of class, with measured corrosion rates < 2 mpy up to 20 wt% NH4HS at 150 psia H2S partial pressure.

14. The corrosion rates of carbon steel and all alloys investigated in this program increased with increasing temperature.

15. The effect of temperature on the corrosion rate of carbon steel and 12 Cr was greatest at low NH4HS concentrations, and diminished as the NH4HS concentration increased.

16. Temperature appeared to have less effect on corrosion than NH4HS concentration, velocity, and H2S partial pressure.

17. The pH change of the NH4HS environment caused by the addition of HCl to achieve chloride concentrations in the range of 100 to 1,000 ppm was negligible. Furthermore, no differences in corrosion rate were observed which could be attributed to the presence of chlorides in the sour water.

18. The presence of hydrocarbon in the environment resulted in reduced corrosion rates. Substantial protection was achieved for carbon steel at hydrocarbon contents ≥ 25 vol%. The remaining alloys exhibited substantial benefit with hydrocarbon contents ≥ 10 vol%.

19. Equivalent protection was observed with the presence of light (API gravity 40) and heavy hydrocarbon (API gravity 20).

20. Higher shear stress produced higher corrosion rates at a given set of environmental conditions. The magnitude of the corrosion rate difference was further related to the corrosivity of the environment.

21. Both ammonium polysulfide (APS) and imidazoline were successful in reducing the corrosion rate of carbon steel in NH4HS environments.

22. At low velocity, APS formed a more stable protective film on the metal surface that led to greater protection. At higher velocity or shear stress, this film became less stable resulting in reduced protection.

23. The results of imidazoline, while promising, relied on sufficient mixing of the environment to ensure contact of the imidazoline with the metal surface. Imidazoline is oil soluble and hence more difficult to partition into the corrosive sour water phase. Thus, the potential reduction of corrosion when using imidazoline may not be realized with certain flow regimes, particularly stratified or laminar flow.

24. This program demonstrated the importance of understanding the mechanical forces in plant systems where multiphase environments and various flow regimes are involved.

44

25. A user friendly software tool, Predict-SW, was developed that successfully incorporated the data obtained in this program and combined these data with flow modeling calculations on plant piping configurations to predict the corrosion rates of the fourteen materials studied over a wide range of NH4HS concentration, H2S partial pressure, temperature, hydrocarbon content and chemical treatment.

(Top) Prepared by: Michael S. Cayard Date: June 10, 2003 Dr. Michael S. Cayard InterCorr International, Inc. Richard J. Horvath Date: June 10, 2003 Mr. Richard J. Horvath Shell Global Solutions (US) Inc. Reviewed by: Russell D. Kane Date: June 10, 2003 Dr. Russell D. Kane InterCorr International, Inc.

45

10. REFERENCES 6. ASM Handbook, Volume 13, Corrosion, ASM International. 7. R.L. Piehl, “Survey of Corrosion in Hydrocracker Effluent Air Coolers”, Materials

Performance, Vol 15 (1), January 1976, pp 15-20. 8. D.G. Damin and J. D. McCoy, “Prevention of Corrosion in Hydrodesulfurizer Air Coolers

and Condensers”, Materials Performance, Vol 17 (12), December 1978, pp 23-26 (see also NACE CORROSION/78, paper # 131).

9. C. Scherrer, M. Durrieu, and G. Jarno, “Distillate and Resid Hydroprocessing: Coping with High Concentrations of Ammonium Bisulfide in the Process Water”, Materials Performance, Vol 19 (11), November 1980, pp 25-31 (see also NACE CORROSION/79, paper # 27).

10. API Recommended Practice 932-A, “A Study of Corrosion in Hydroprocess Reactor Effluent Air Cooler Systems”, American Petroleum Institute, September 2002.

11. B.V. Johnson, H.J. Choi and A.S. Green, “Effects of Liquid Wall Shear Stress on CO2 Corrosion of X-52 C-Steel in Simulated Oilfield Production Environments”, NACE CORROSION/91, Paper 573.

12. Keller H., 1974, Erosion Corrosion in Wet Steam Turbines, VGB Kraftwerkstechnik, No. 54, pg. 292.

13. W. P. Jepson et. al., "Model for Sweet Corrosion in Horizontal Multiphase Slug Flow," CORROSION/97, Paper 11, NACE, Houston, TX, 1997.

14. K. D. Effird et. al., "Experimental Correlation of Steel Corrosion in Pipe Flow with Jet Impingement and Rotating Cylinder Laboratory Tests," CORROSION/93, Paper 81, NACE, Houston, TX, 1993.

15. A. M. K. Halvorsen and T. Sontvedt, "CO2 Corrosion Model for Carbon Steel Including A Wall Shear Stress Model for Multiphase Flow and Limits For Production Rate to Avoid Mesa Attack," CORROSION/99, Paper 42, NACE, Houston, TX, 1999.

16. G. Schmitt et. al., "Evaluation of Critical Flow Intensities for FILC in Sour Production," CORROSION/98, Paper 46, NACE, Houston, TX, 1998.

17. O. Baker, “Simultaneous Flow of Oil and Gas”, Oil and Gas J., 53, 185, July 1954. 18. B. A. Eaton, et. al., "The Prediction of Flow Patterns, Liquid Holdup and Pressure Loses

Occurring During Continuous Two-phase Flow in Horizontal Pipelines," JPT, pp. 815-828, Trans. AIME, 240 June 1967.

19. Al-Sheikh, et al., “Prediction of Flow Patterns in Horizontal Two-Phase Pipe Flows”, IBID, 48, 21, 1970.

20. Y. Taitel, and A. E. Dukler, "A Model for Predicting Flow Regime Transitions in Horizontal and Near-Horizontal Gas Liquid Flows", AIChE Journal, v22, p47, 1976.

21. Govier et al., “The Horizontal Pipeline Flow of Air-Water Mixtures”, Canadian J. of Chemical Engineering, No. 93, 1962

22. J. M. Mandhane, G. A. Gregory, and K. Aziz, "A Flow Pattern Map for Gas-Liquid Flow in Horizontal Pipes," Intl. J. Multiphase Flow, Vol. 1, pp. 537-553, 1974.

23. H. D. Beggs and J. P. Brill, "A Study of Two-Phase Flow in Inclined Pipes," J. Pet. Tech., pp. 607-617, May 1973.

24. A. E. Duckler et. al., "Frictional Pressure Drop in Two-Phase Flow: B. An Approach Though Similarity Analysis," AIChE J., pp. 44-51, Jan. 1964.

25. S. Srinivasan, " An Analytical Model to Experimentally Emulate Flow Effects in Multiphase CO2/H2S Systems," CORROSION/00, Paper 58, NACE, Houston, TX, 2000.

(Top)

11. Appendix I

PARTICIPATING SOUR WATER JIP SPONSORS

BP Chevron Research & Technology Company

ConocoPhillips, Inc. ExxonMobil Research and Engineering

Flint Hills Resources, L.P. Fluor Daniel Inc.

Idemitsu Kosan Company, Ltd. Kuwait National Petroleum Company

Petrobras Saudi Aramco

Shell Global Solutions (US) Inc. Sunoco

Syncrude Canada Ltd. TotalFinaElf

UOP Valero Energy Corporation

(Top)

12. Appendix II

BASELINE ISOCORROSION DIAGRAMS


0

10

20

30

40

50

60

70

80

90

0 2 4 6 8 10 12 14 16


Velo

city

, ft/

sec


<5 mpy

5-20 mpy

20-50 mpy

50-100 mpy

100-200 mpy 200-300 mpy

>300 mpy



0

10

20

30

40

50

60

70

80

90

0 2 4 6 8 10 12 14 16


Velo

city

, ft/

sec

< 2 mpy

2-5 mpy

5-10 mpy

10-20 mpy

> 20 mpy

< 2 mpy

5-10 mpy

10-20 mpy

> 20 mpy

2-5 mpy


ISOCORROSION DIAGRAM FOR 12Cr STEEL(48-hr laboratory tests: P[H2S]=50 psia, T=130 F)

0

10

20

30

40

50

60

70

80

90

0 2 4 6 8 10 12 14 16


Velo

city

, ft/

sec

< 2 mpy

2-5 mpy

5-20 mpy

20-50 mpy

> 50 mpy

< 2 mpy

2-5 mpy 5-20 mpy 20-50 mpy > 50 mpy


ISOCORROSION DIAGRAM FOR 304SS(48-hr laboratory tests: P[H2S]=50 psia, T=130 F)

0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec

< 2 mpy

2-5 mpy

5-10 mpy

10-20 mpy

> 20 mpy

< 2 mpy 2-5 mpy

5-10 mpy 10-20

> 20 mpy


ISOCORROSION DIAGRAM FOR 316SS(48-hr laboratory tests: P[H2S]=50 psia, T=130 F)

0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec

< 2 mpy

2-5 mpy

5-10 mpy

10-20 mpy

> 20 mpy

< 2 mpy

2-5 mpy

5-10 mpy 10-20

> 20 mpy



0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec

< 2 mpy

2-5 mpy

5-10 mpy

10-20 mpy

> 20 mpy

< 2 mpy 2-5 mpy

5-10 mpy

10-20 mpy

> 20 mpy



0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec < 2 mpy

2-5 mpy

5-10 mpy

10-20 mpy< 2 mpy 2-5 mpy

5-10 mpy

10-20 mpy

> 20 mpy



0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec

< 2 mpy

2-5 mpy

5-10 mpy

10-20 mpy< 2 mpy 2-5 mpy

5-10 mpy

10-20 mpy

> 20 mpy


ISOCORROSION DIAGRAM FOR ALLOY 20Cb3(48-hr laboratory tests: P[H2S]=50 psia, T=130 F)

0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec

< 2 mpy

2-5 mpy

5-10 mpy

10-20 mpy

< 2 mpy

2-5 mpy 5-10 mpy

10-20 mpy



0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec

< 1 mpy

1-2 mpy

2-5 mpy

5-10 mpy

> 10 mpy

< 1 mpy

2-5 mpy

5-10 mpy

> 10 mpy

1-2 mpy



0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec

< 1 mpy

2-5 mpy

< 1 mpy

2-5 mpy

1-2 mpy



0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec

< 1 mpy

1-2 mpy< 1 mpy

1-2 mpy


ISOCORROSION DIAGRAM FOR ALLOY AL6XN(48-hr laboratory tests: P[H2S]=50 psia, T=130 F)

0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec

< 1 mpy

1-2 mpy

< 1 mpy

1-2 mpy


ISOCORROSION DIAGRAM FOR ALLOY C-276(48-hr laboratory tests: P[H2S]=50 psia, T=130 F)

0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25 30


Velo

city

, ft/

sec

< 1 mpy

< 1 mpy


(Top)

13. Appendix III

CORROSION RATE CORRECTION FACTORS FOR H2S PARTIAL PRESSURE

CORRECTION FACTOR F{pp} for CARBON STEEL

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

0 5 10 15 20 25 30


F{pp

}, m

py/p

si



CORRECTION FACTOR F{pp} for Alloy 400

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

0 5 10 15 20 25 30


F{pp

}, m

py/p

si



CORRECTION FACTOR F{pp} for 12 Cr STEEL

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

0 5 10 15 20 25 30


F{pp

}, m

py/p

si



CORRECTION FACTOR F{pp} for 304 SS

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 5 10 15 20 25 30


F{pp

}, m

py/p

si



CORRECTION FACTOR F{pp} for 316 SS

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 5 10 15 20 25 30


F{pp

}, m

py/p

si




0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 5 10 15 20 25 30


F{pp

}, m

py/p

si



CORRECTION FACTOR F{pp} for Alloy 800 & Alloy 600

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 5 10 15 20 25 30


F{pp

}, m

py/p

si



CORRECTION FACTOR F{pp} for Alloy 20Cb3

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0 5 10 15 20 25 30


F{pp

}, m

py/p

si




0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 5 10 15 20 25 30


F{pp

}, m

py/p

si




0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 5 10 15 20 25 30


F{pp

}, m

py/p

si



CORRECTION FACTOR F{pp} for Alloy 2507 & AL6XN

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.1

0.11

0.12

0.13

0.14

0.15

0 5 10 15 20 25 30


F{pp

}, m

py/p

si



CORRECTION FACTOR F{pp} for Alloy C-276

0

0.005

0.01

0.015

0.02

0.025

0.03

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30


F{pp

}, m

py/p

si



(Top)

14. Appendix IV

CORROSION RATE CORRECTION FACTORS FOR TEMPERATURE

CORRECTION FACTOR F{temp} for CARBON STEEL

0.000

0.050

0.100

0.150

0.200

0.250

0.300

0.350

0.400

0 5 10 15 20 25 30


F{te

mp}

, mpy

/deg

F



CORRECTION FACTOR F{temp} for ALLOY 400

0.000

0.050

0.100

0.150

0.200

0.250

0 5 10 15 20 25 30


F{te

mp}

, mpy

/deg

F



CORRECTION FACTOR F{temp} for 12 Cr STEEL

0.000

0.010

0.020

0.030

0.040

0.050

0.060

0 5 10 15 20 25 30


F{te

mp}

, mpy

/deg

F



CORRECTION FACTOR F{temp} for 304 & 316 SS

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

0.040

0 5 10 15 20 25 30


F{te

mp}

, mpy

/deg

F



CORRECTION FACTOR F{temp} for Alloy 2205

0.000

0.001

0.002

0.003

0.004

0.005

0.006

0.007

0.008

0.009

0.010

0 5 10 15 20 25 30


F{te

mp}

, mpy

/deg

F



CORRECTION FACTOR F{temp} for Alloy 800, 600, 20Cb3, 825 & 625

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.10

0.11

0.12

0 5 10 15 20 25 30


F{te

mp}

, mpy

/deg

F



CORRECTION FACTOR F{temp} for Alloy 2507 & AL6XN

0.0000

0.0005

0.0010

0.0015

0.0020

0.0025

0.0030

0 5 10 15 20 25 30


F{te

mp}

, mpy

/deg

F



CORRECTION FACTOR F{temp} for Alloy C-276

0.0000

0.0005

0.0010

0.0015

0.0020

0 5 10 15 20 25 30


F{te

mp}

, mpy

/deg

F



(Top)

15. Appendix V

CORROSION RATE CORRECTION FACTORS FOR HYDROCARBON


0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70 80 90 100


Effic

ienc

y (%

)



Influence of Hydrocarbon - 12 Cr

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70 80 90 100


Effic

ienc

y (%

)



Influence of Hydrocarbon - 304

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70 80 90 100


Effic

ienc

y (%

)



Influence of Hydrocarbon - 2205

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70 80 90 100


Effic

ienc

y (%

)



Influence of Hydrocarbon - 2507 & 800

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70 80 90 100


Effic

ienc

y (%

)



(Top)

16. Appendix VI

PREDICT-SW CORRELATIONS WITH ISOCORROSION DIAGRAMS

(Top)

17. Appendix VII

FIELD EXPERIENCE PRESENTED BY THE JIP SPONSORS

Field Experience Presented by the JIP Sponsors

Case

Mater

ial

NH4HS

(wt%)

H2S (psia)

Temp. (F)

Pipe I.D. (in)

Geome

try

Velocity

(ft/sec)

Vapor

(vol%)

Sour Wate

r (% of

Liquid)

Hydrocarbon

(% of Liquid)

Hydrocarbon

Type

Measured CR (mpy)

1 CS 3.9-5.5

67 150 4 Elbow 16.4 - 7.7 92.3 Hydrocracked

Waxy Distillate

12 to 32

2 CS 4.9-7.4

67 150 4 Elbow 16.4 - 6.3 93.7 Hydrocracked

Waxy Distillate

80 to 100

3 CS 14-18

82 110 5.76

Elbow 40 99.99

99.5 0.5 Recycle Gas from CHPS in Cat Feed

HT

550

4 CS 23-27

8-10

65-110

7.98

Elbow / Tee

50 99 25 75 Light Hydrocar

bon, Naphtha,

vapor

80-240

5 CS 8-10 55 130 9.06

Elbow 30 99+ 99+ <1 Gas Oil & H2

68

6 CS 8-10 55 130 9.06

Elbow 13 95 50 50 Gas Oil & H2

120

7 CS 8 60 130 1.69

Elbow / Tee

13 0 100 0 None 170

8 CS 3.5 49 109 8.30

Piping 9.8-16.4

99 100 0 Light (C1 – C5)

118

9 A351 CF8M

13 150?

130 1 Valve 12 0 100 0 None 90

10 304L 20 21 120 8.75

Tee 19 97.7 18 52 Naphtha Through Gas Oil

16

(Top)

Documents

SAER-5942