Historically, yield criteria alone have been used to design wells for sour service, based upon the assumption that the steel is robustly qualified and so proven relatively immune to the influence of hydrogen sulfide. Recently work has been initiated where use of higher strength steel known to be partially sensitive to hydrogen sulfide is qualified in reduced H2S environments by either crack propagation or crack initiation testing per ISO 15156/MRO1751  (NACE TMO1772 ). This paper recommends instead that three separate, stress-based criteria are needed to design such higher strength low alloy carbon steel (LACS) production casing (and tubing) for service in H2S-sour, HPHT wells. The paper is being written to foster discussion of this recommendation. The idea is that design against yield alone is not sufficient to quantify the performance of LACS pipe in a sour environment. In addition to (1) yield, it is proposed that separate design criteria must be met to prevent failure by (2) crack propagation and (3) crack initiation. The second criterion, crack propagation design, determines the limit of pipe performance assuming overload propagation of a pre-existing crack which is below inspection equipment detection thresholds. The crack propagation criterion addresses overload of cracks initially present in the pipe; similar to what is observed in the NACE Method D DCB test. The third criterion, crack initiation design, determines the limit of pipe performance based on the pressure and axial load necessary to cause crack initiation (e.g. sulfide stress cracking, SSC) in a sour environment. The crack initiation criterion corresponds to creation and stable growth of cracks which initially do not exist in the pipe; similar to what is observed in the NACE Method A tensile test. The failure criteria are expressed in terms of critical states of stress and measurable pipe material properties from yield tests and from NACE Method-A and Method-D environmental toughness tests. Both of these fracture design criteria are proposed to be functions of the pipe hardness.

Overlapping pipe failure envelopes are shown that correspond to the three different pipe failure criteria. Example design calculations are presented which show that the different design criteria (yield, crack propagation, or crack initiation) dominate the design of the casing depending on the different pipe diameters, grades, and well operating conditions. The examples show that the use of any one or two of these criteria is not sufficient to ensure that the other design criteria are satisfied: all three criteria need to be separately applied and satisfied. A case is made that a design factor needs to be applied to each of the three design formulas because of (a) load uncertainty; (b) scatter in pipe mechanical properties; (c) scatter in pipe sour material properties; and (d) potential interaction between these different mechanisms. Because of these uncertainties, a framework is outlined for a probabilistic design approach to avoid sour failure.

You can access this article if you purchase or spend a download.