Feb-2021

Advances in damage modelling (TIA)

High temperature hydrogen attack (HTHA) is a serious damage mechanism that can affect hydroprocessing, catalytic reforming, and hydrogen production units in the refining industry.

Dave Dewees
Becht

Article Summary

HTHA damage occurs in stages; starting with grain boundary voids and progressing to fissures and eventually cracks that can compromise the integrity of pressurised equipment. Although the mechanisms of HTHA have been known for over 100 years, managing equipment in HTHA service has remained extremely challenging because:
-    No reliable model existed for the rate of damage formation and progression (as a function of temperature, hydrogen partial pressure, and stress).
-    Without an understanding of how fast isolated voids turn into dangerous cracks, inspection has focused on detecting small scale damage with mostly frustrating results.

While the simplest (and most costly) solution is to ’alloy up’ to more resistant materials, the quantity of affected equipment is large and such an approach typically takes many, many years. Additionally, in the revision to API RP 941, a new as-welded carbon steel curve was added (see Figure 1); as can be seen, equipment that had been comfortably below the original Carbon Steel Nelson Curve (and therefore a lower priority for replacement) suddenly becomes unacceptable by an unknown margin.

The (SA-106 Gr. B) CCR piping in Figure 1, operating at 625°F and 75 psi hydrogen partial pressure, is a real world example that owners are currently facing. Without a time basis, there is no direct solution other than to shutdown immediately, which is obviously not realistic (or typically even warranted).

For this reason, Becht has in developed and validated a time based HTHA damage model. Time based results using this new model are shown in Figure 2 for the example CCR piping which has been selected as a worst-case scenario. That is, the example piping would be a significant failure risk (indeed, several failures have been documented in this same operating range in less time as indicated in the figure). However, even in the limiting case, since there is a quantitative basis, expedited inspection and replacement plans could begin without the extreme uncertainty that currently accompanies such difficult decisions. Further, while this example indicates short term action, in fact many cases lead to replacement consideration, one, two or even more turnarounds in the future.

The key to such decisions is generalising the single Nelson Curve into a time based family of curves (see Figure 3). The complete basis for these curves (as well as other families of curves for common cases) can be found here:
Part 1: Introduction
Part 2: Through-Wall Damage Modelling
Part 3: C-0.5Mo Dataset and Calibration
Part 4: Carbon Steel Dataset and Calibration
Part 5: Time-Dependent Nelson Curves for Carbon and C-0.5Mo Steels

When used in combination with materials, inspection, and process expertise, it is hoped that this technology will be part of a useful and impactful multidiscipline solution to the HTHA reliability issues that challenge industry.

This short case study originally appeared in PTQ's Technology In Action feature - Q1 2021 issue.

For more information: Ddewees@becht.com