SURVEY UPDATES AMINE STRESS CORROSION CRACKING DATA

The final report by National Association of Corrosion Engineers (NACE) task group T-8-14 has been published, revising and expanding the information on stress corrosion cracking of carbon steel in diethanolamine and diisopropanolamine service.

Three members of this task force have summarized its conclusions. These members are Chairman Ara J. Bagdasarian, a group leader for Unocal Science & Technology Corp., Brea, Calif.; Vice-chairman Cathleen A. Shargay, a staff corrosion engineer for Arco Products Co., Carson, Calif.; and Secretary John Coombs, a technical advisor in corrosion engineering for Arco Products Co. in Anaheim, Calif.

Reports of stress corrosion cracking (SCC) of carbon steel in amine solutions date back to the early 1950s. Cracking is mainly associated with the weld heat affected zone (HAZ) or the welding-related residual stress zone adjacent to the HAZ. Amine cracking is predominantly intergranular and oxide filled.

A few major events, including the rupture of an amine absorber and the consequent disastrous fire, prompted NACE's T-8 committee to set up a task group to survey existing amine plants. (It was later determined that the amine absorber failed from hydrogen-induced cracking rather than amine SCC.1)

The task group was formed in September 1984, It completed the survey in 1986, with the results presented at NACE Corrosion 87 in San Francisco. 2

A major conclusion of the survey was that cracking frequency was more prevalent in monoethanolamine (MEA) than in other amines. For example, the total reported cases of cracking for diethanolamine (DEA) was 22, as compared to 78 for MEA.

Further examination of the DEA data indicated that some units were previously in MEA service and the reported cracks were actually associated with that period. A detailed follow-up review of the DEA data also revealed that some cases were caused by processes other than amine cracking. In many cases, further inspection or testing had been done after the original survey was submitted. The revised data were published in 1989 (OGJ, June 5, 1989, p. 45).

SURVEY RESULTS

The task group has remained active and has solicited additional data on SCC incidences in amine units other than MEA. Data have been received from Europe and Asia, as well as from U.S. sources.

The new data do not alter the original NACE survey's conclusion--that SCC is more prevalent in MEA than DEA systems--but there are a significant number of reported cracking cases in both DEA and ADIP (diisopropanolamine) service.

A summary of the number of cracked vs. noncracked cases in DEA and diisopropanolamine (DIPA) service is shown in Table 1. Because the new data were collected to emphasize cracking and its causes, only the numbers for cases of cracking have been revised.

DEA

Table 2 summarizes 37 cases of DEA cracking, including both new and confirmed old data. This table shows the type of equipment, the age of the equipment, and its stress-relief status, for each case. It also shows that 11 cracks caused leaks, and 7 cracks were verified by metallography to be amine SCC.

In the cases comprising this data base, SCC occurred at normal operating temperatures of 110-265 F., with eight cases at or below 140 F. Fig. 1 shows cracking vs. normal operating temperature for the entire DEA data base.

DIPA/ADIP

Table 3 summarizes 42 cases of DIPA cracking, including both new and old data. All the cases of DIPA cracking are from ADIP units where DIPA concentration is between 15 and 20%. No case of cracking has been reported for Sulfinol plants that typically use 50% DIPA.

Some cracks led to leaks, and five were verified by metallography to be SCC. Temperatures ranged from 90 F. to 260 F. It was not clearly stated whether these were maximum or normal operating temperatures, but in most cases, they appeared to be the latter. A bar graph of ADIP cracking vs. normal operating temperatures is shown in Fig. 2.

ANALYSIS

Although amine SCC is still much more prevalent with MEA than with other amines, these data clearly show a significant number of cracks with both DEA and ADIP.

One major objective of continuing to obtain data on "other" amines was to help determine the minimum temperature for cracking. In DEA, 6 cases out of 37 reported were below 140 F. In these cases, cracking may have resulted from short-term exposure to higher temperatures.

The most common high-temperature exposure occurs in "steam-outs" during shutdowns Some companies require water washing before steam-out to wash away residual amine before exposing the equipment to steam. If the water washing is carried out properly, it should eliminate cracking during the short-term exposure to high-temperature steam.

Information is lacking on whether the cases of cracking between 110 and 140 F. were caused by short-term high temperature exposures. However, for three DEA cracks at 140 F. which were metallurgically confirmed, high-temperature steam has been ruled out as the culprit, based on survey responses. Hence, there are now three examples of confirmed DEA cracking at 140 F.

It has been reported that the minimum temperature for cracking varies based on solution chemistry such as the amine concentration and the acid gas loading.3 4 Hence, 140 F. may not be applicable to all units.

With one exception, all reported cracks were in nonstress relieved equipment, confirming an original survey conclusion that stress relief is usually an effective deterrent to SCC.

In establishing an inspection program for amine units with DEA or ADIP, attention should be given to the operating history including temperature upsets and steam-outs, the report says. Proper water wash prior to every steam-out will reduce the chance of cracking in equipment operating below 140 F. For new construction in DEA service, the lower cost of not stress-relieving equipment operating at low temperatures should be compared with the need for water wash prior to steam-out and the probability of higher-frequency inspection.

ADIP

With ADIP, cracking occurred at normal operating temperatures as low as 90 F. (Fig. 2). At this point, with the limited data base, the task group believes that the conclusions for DEA also apply to ADIP. In addition, as a minimum, the recommendations for DEA should be implemented for ADIP.

It should be emphasized that these conclusions and recommendations are for the prevention of amine SCC. Cracking by other mechanisms can also occur in certain sections of amine units. For further discussion on these subjects, the task group recommends referring to API Recommended Practice 945. This is the final report by NACE task group T-8-14. the American Petroleum Institute subcommitte on materials and corrosion has set up a permanent task group to continue to compile and analyze data related to amine cracking. All further data on amine cracking should be sent to the attention of this task group at API.

REFERENCES

McHenry, H.I., et al., "Examination of a Pressure Vessel that Ruptured at the Chicago Refinery of the Union Oil Company on July 23, 1984," National Bureau of Standards Report NBSIR 86-3049, March 1986.
Richert, J.P., Bagdasarian, A.J., and Shargay, C.A., "Stress Corrosion Cracking of Carbon Steel in Amine Systems," Paper no. 187, Corrosion 87, NACE, Mar. 9-13, 1987, San Francisco.
Schutt, H.U., "New Aspects of Stress Corrosion Cracking in Monoethanol amine Solutions," Materials Performance, Vol. 27, No. 12, December 1988, p. 53.
Lyle, F.F. Jr., "Stress Corrosion tracking of Steels in Amine Solution Used in Natural Gas Treatment Plants," Paper No. 158, Corrosion 88, NACE, Mar. 21-25, 1988, St. Louis.