Proceedings of the Third International Conference on Metals & Hydrogen F03

Towards a better understanding of hydrogen embrittlement testing of steel in sour service conditions by coupling H diffusion to fracture mechanics

Krista Van den Bergh (1)1 , Zinedine Zermout (1)1 , Patrick Goes (1)1

  • (1) 1

    ArcelorMittal Global R&D Gent, OCAS N.V., President John F. Kennedylaan 3, 9060 Zelzate, Belgium


Pipelines are the most cost-effective method of oil or gas transportation over long distances. However, the transport of sour gas imposes severe restrictions on the use of oil and gas pipeline steels as a result of H2S corrosion cracking, which leads to a high risk for the pipeline integrity.

The present study summarizes the investigations performed in the framework of the development of Oil Country Tubular Goods (OCTG) sour resistant steel grades at AM Global R&D Gent. Techniques used to study the sour resistance are hydrogen permeation measurements and double cantilever beam (DCB) testing in H2S environment.

The influence of the environmental parameters on the sulphide stress cracking (SSC) is studied. It is seen that the SSC test severity, as a function of the hydrogen sulphide partial pressure and pH at the interface, will influence the hydrogen permeation process. The formation of ferrous sulphide corrosion products on the steel surface will also change the hydrogen permeation rates.

Also the influence of the steel-making process on the SSC behaviour of the steels is evaluated. It was found that by changing steps in the steel-making process for steels with same chemistry, the SSC performance can be improved.

To explain the decrease in KISSC (sulphide stress cracking threshold stress intensity factor) as the thickness of the DCB specimen is decreased, an attempt has been made to correlate the SSC resistance rating with the hydrogen bulk concentration on the crack plane. For this, a finite element method (FEM) model describing the hydrogen diffusion with realistic boundary conditions was elaborated. A first finding is that KISSC does not depend solely on the bulk hydrogen content in the crack plane.  A possible improvement in explaining the observed features is to take into account the crack propagation dynamics. For this purpose, a novel experimental device was designed and implemented. It already shows differentiating crack propagation features for the two DCB specimen sizes. As an outlook, the crack propagation features will be used to better quantify the hydrogen content at the crack tip. This information will be key to establish the conditions (hydrogen content, local stresses) under which fracture occurs for a given material.