Results of a four-year research program at the Yale University Metallurgy Department are presented. After a review of the problems associated with spontaneous fracture of steels in sour gas condensate wells an account of the investigation is given as follows:
Part 1—Hydrogen sulfide embrittlement was compared with that caused by cathodic charging. An investigation was made of spontaneous fracture under conditions of constant deformation, as dependent on alloy composition, heat treatment, plastic deformation and chemical environment.
Part 2—Hydrogen absorption by iron-nickel alloys from hydrogen sulfide was compared to that from sulfuric acid.
Part 3—Permeation rate of hydrogen through austenitic, martensitic and ferritic states of iron-nickel alloys was measured.
Part 4—Time dependence of static fracture of iron-nickel alloys as a function of applied stress, structure and hydrogen concentration was evaluated.
The following conclusions are given: Basic cause of failure is absorption of hydrogen. Principal factors in service failure are internal stress, magnitude of applied stress and hydrogen content. Both martensitic and ferritic structures fail when the stress level and hydrogen content are sufficiently high. Trace amounts of carbon, oxygen and nitrogen in 10 percent nickel steel have no detectable effect on failure. Adoption of limiting strength or hardness below which fracture does not occur in constant deformation tests is unreliable in predicting service life.
Because it is impractical to lower operating stress significantly below apparent “endurance limit” the remaining alternative is reduction of hydrogen concentration in the steel. Deadload tests in the environment must be made before safe operating procedures can be defined. 3.2.2
