Avoiding Corrosion Related Leaks in Processing Facilities

Introduction

The purpose of this article is to stimulate some thinking about how the vast network of consultants, academics, technology providers and industry organizations can help materials and corrosion subject matter experts (SMEs) with our challenges in assuring the mechanical integrity of aging plants. I will try to do this by talking about one company’s journey as they implement a corrosion (thinning, cracking & metallurgical deterioration) review process aimed at avoiding corrosion-related leaks in aging oil and gas refining and petrochemical facilities.

But first, let’s start with an example to illustrate what I think some of the types of challenges that we as materials & corrosion engineers face. The first failure analysis assignment that I had as a new materials engineer out of school was to investigate strange failures of tank insulation bands - the stainless steel bands that are put around large storage tanks to help hold insulation in place. These bands are positioned, pulled tight in tension, and then crimped in place with a clamping device. They were made of an 18-8 (304 ss) type stainless steel. Traditionally, the annealed and pickled bands had been used. These had a dull matte finish and had been used successfully for many years. But then, one day some new bright shiny bands were purchased for a tank insulation job, and the insulation crew installed them as normal.

When they left at the end of the work day, everything seemed normal with the bands installed. But the next morning when the crew came back to continue work, the bands were loose and falling off the tank. They brought samples of the broken bands into our materials laboratory to find out what happened, since this had now gone on a few days and rework was needed. The obvious answer was to buy dull straps, not the new shiny ones. But why?

An Investigation found that the shiny straps were made shiny by bright annealing in hydrogen. When they were crimped, there was a transformation to martensite. The shiny straps were saturated with hydrogen and fractured from delayed cracking after crimping. The dull strap also transformed to martensite, but they didn’t have hydrogen present and didn’t crack. What’s the significance? Seemingly minor changes can cause unexpected failures.

So, how are we as materials engineers supposed to prevent these types of issues that result from things possibly out of our control? Would you expect to use a Management of Change (MOC) process for changing tank insulation straps from dull to shiny?

To understand, we need to go full circle to the beginning of this story. If we go back about 5 or 6 years, we had a few major leaks caused by corrosion - both thinning and cracking phenomena - which our management viewed as “unexpected” or “unpredicted.” A group of materials and corrosion SMEs from within the company was tasked with making a review of our corrosion management practices and their implementation, and to propose some improvements. At the time, there was a tendency to attribute the issues to poor inspection practices.

A first lesson that I learned in this process was the value of using data to make the problems more obvious. We used a continuous improvement type of approach and started by “go see and assess.” Opinions were not worth much. We needed data and facts to understand where the breakdowns were.

We organized our efforts in the areas of Process, People, Plant and Performance.