The Detrimental Effects of Wet Insulation in the CUI Range
In a previous article, we shared our view that in the oil & gas industry, it is often not well understood how drastically water-absorbent insulation impacts corrosion under insulation in the CUI temperature range. Here we will delve a bit deeper into this, especially with respect to hot insulation on carbon steel.
In the hot insulation world of outdoor refineries, chemical plants and offshore facilities, water-absorbent insulation becomes the vehicle that harbors wetness, resulting in long-term contact with the steel, which is the ultimate cause of corrosion under insulation (CUI).
Most hot insulation materials have a lot of open space or porosity in them. When water wicks into and fills the insulation material sitting against the steel surface of equipment operating at a temperature that is not hot enough to quickly drive away wetness, any unprotected carbon steel will experience CUI because of these repeated long intervals of wetness. (For background reading about CUI, see Corrosion Under Insulation: The Challenge and Need for Insulation.)
Unprotected carbon steel refers to any area, large or small, where an immersion grade hot/boiling-resistant coating is not properly adhered to the steel. The unprotected area can be an entire process vessel, a tiny spot, a through-film scratch, defect or pin hole.
Long-term contact of the steel surface with hot or boiling water can result from wet insulation or containment of water that cannot readily drain or escape.
Figure 1: CUI induced by wet insulation.
Long-duration wetness of the steel is a function of the type of insulation, the temperature and the volume of water absorbed. After a rain event, days or weeks may pass during which the equipment is not hot enough to overcome the drastically degraded thermal properties of the insulation in order to vaporize water at the steel surface. If the water does vaporize, it must find its way out, and this can be quite a tortuous path when water has saturated significant amounts of the open porosity in the insulation.
The corrosion rate at any unprotected area of steel (barring any unplanned presence of some other invading spilled chemical) will be a function of:
- The steel temperature
- The electrolytic nature of the water (affected by soluble contaminants either from the insulation or those that find their way in from the outside environment, or both)
- The size of the area of the exposed steel vs. the size of the surrounding protected (properly coated) area
- The frequency of wet/dry cycles
That is the primary CUI mechanism for hot insulation. What do we know from our operating history?
- We know that it is the lower temperature equipment (usually less than 300°F/150°C), where the water is not driven off the steel quickly enough, which experiences CUI.
- We know that most of the CUI experienced on equipment installed prior to the advent of hot/boiling water-resistant coatings occurred under water-absorbent insulation.
- We know that where water repellent insulation has been used, and when water is not held up by equipment design flaws, the incidence of CUI has been nil to non-existent even on equipment that was installed long prior to the advent of hot water-resistant coatings technology.
How do we know these things? Let's take a trip back in history.
A Long Case History – How to Drastically Minimize CUI
For 30+ years, I observed the progress of a 40-year case history occurring in two massive Gulf Coast refinery complexes. That history allows us to demonstrate this fact:
Wherever water repellent insulation was installed—at a point in history when CUI coatings did not yet exist—CUI did not occur if water was allowed to freely drain and was not held up by equipment design flaws, or by water-absorbent insulation.
Back in 1974, after years of struggling with CUI, the refineries switched completely to expanded perlite, which is water repellent (throughout its thickness) to a temperature well above the CUI range (~450°F). Expanded perlite became the only approved hot insulation material for these facilities, and it was supplied to them by a North American manufacturer.
Epoxy phenolics were not commercialized until the late 1980s, when they became the new hybrid epoxy coatings that allowed higher temperature resistance and greater chemical resistance. They were welcomed by the industry because this allowed us to protect more of our equipment, especially from the effects of water on hot steel (think CUI).
It wasn't until 1991 that the corporate coating standard used by these refineries was updated to incorporate the new epoxy phenolics. Prior to that, the standard designated one coat of inorganic zinc or one coat of epoxy primer (depending on the operating temperature) as the coating for under insulation. These were intended to be shop primers that would keep steel from flash rusting on the trip from the fabrication shop to the job site where the insulation was installed.
So in the period between 1975 and 1991, any new or existing carbon steel to be insulated at these facilities received only expanded perlite insulation over virtually unprotected carbon steel. Remember that "unprotected" carbon steel was the normal course of business for any facility that had not yet specified or embraced what were then the "new" technology coatings. The big difference at these two refineries was that water repellent insulation was the standard, while most other industrial facilities around the world were still using water-absorbent calcium silicate or mineral wool insulation materials.
What continues to be the result at these two refineries some 40 years later?
There have been a vast number of opportunities to view the “unprotected" steel coated with only shop primers and insulated with expanded perlite. Partial removal of insulation occurred most often to accomplish either mechanical work or CUI inspections.
The photo below (Figure 2) shows an example of one of the new construction projects that occurred in that time frame when expanded perlite insulation was the standard hot insulation, but a proper CUI coating was not yet in the corporate coating standard.
The consistent discovery to date is this: Wherever expanded perlite was used and the equipment was properly designed to drain water, there was no corrosion.
A rare exception to this would be where temperature excursions above the water repellence limit of the insulation must occur. For these thermally cyclic scenarios, the use of rigid insulation of any type is not recommended, and other types of insulation along with coatings—which are resistant to thermal shock—should be considered. And, of course, keeping water out of insulation systems is even more crucial on this type of equipment.
In the photo shown below, the installation contractor used bulk mineral wool to stuff the gap immediately above the support ring, rather than take time to cut the necessary small pieces of expanded perlite. The vertical white lines are dried insulation powder, which was deposited by quickly drying water running down behind the water repellent insulation. Note the areas of coating in excellent condition and the areas that exhibit rusting are the same steel with the same coating (3 mils epoxy primer), operating at the same temperature, experiencing water entry. The draining/flowing water actually cools the steel.
Figure 2: Vertical tower with properly designed insulation support rings allowed water drainage until it met with absorbent insulation. New tower installed: 1989. 130–180°F temperature range. CUI inspection: 2001.
In the pictured example, the location where coating failure occurred was only where steel was in contact with the water-absorbent mineral wool insulation. Corrosion occurred where the amount of water absorbed was such that it could not be evaporated quickly enough before the water got hot. The water held by the fibrous insulation increased in temperature as it was held against the hot steel, and caused the 3 mil coating to fail as it penetrated and undercut a coating film that was truly never intended to be resistant to hot water.
What are the key lessons learned from this 40-year case history?
- Outdoor equipment operating in the CUI range, with water-absorbent insulation and without a high-quality application of hot water-resistant coating, is at a high risk of CUI.
- In addition to CUI risk, we learned that water absorbed into the insulation drastically impacts the insulation's ability to function as insulation.
- If the insulation on CUI-range equipment is expanded perlite, and the equipment does not have the benefit of a high-quality application of a hot water-resistant coating, the risk of CUI will remain wherever equipment design details or insulation installation errors prevent water drainage and cause long-term water holdup.
Other pertinent industry experiences include:
- 5-year-old floating production storage and offloading (FPSO): Severe CUI on vessels without proper coating, insulated with glue-adhered cellular glass. Severe CUI on piping without proper coating, insulated with mineral wool.
- 1-year exposure of newly installed outdoor equipment (in southern U.S.) insulated with mineral wool: After several months of startup and lineout attempts, the insulation was found in many areas to be water-saturated and preventing proper functionality of the equipment.
Design Flaws and Application Errors
If we did find CUI under expanded perlite, it was in places where liquid was held up, such as at insulation support rings that were continuously welded to a vertical vessel (design flaw), or on vertical pipes where someone had sealed the bottom termination with a generous application of caulk that prevented water drainage (applicator or specification error).
Silicone caulk should be used to seal around protrusions at the top and on the sides of equipment to keep water from getting in, but never at the bottom where gravity naturally allows water to flow out.
In my personal experience, as late as 2008, new construction vessel drawings provided by some big engineering firms included improperly designed insulation support rings. Even though the corporate standard for insulation required (as it had for decades) a support ring design that allowed free drainage of water, we nonetheless received drawings with this erroneous design detail—a horizontal flat bar continuously welded to a vertical pressure vessel.
Vertical vessels are prone to CUI at such improper insulation support rings because:
- Support rings act as heat sinks resulting in cooler steel temperature near the ring, thus longer water contact, and
- Preventing quick water drainage from behind water repellent insulation means any unprotected steel sees a greater risk of CUI.
Properly designed support rings for vertical vessels consist of horizontal flat bars, sized to provide a 1/4 to 3/8-inch space between the ring and the vessel shell to allow for water drainage. The flat bar is supported on small rectangular tabs of steel welded with vertical orientation to protrude 90 degrees from the vessel shell—the exact design seen in the photo below. For piping, support for rigid insulation should similarly allow for water drainage. Welded horizontal tabs, spaced about 3 inches apart, have been successfully used.
Rigid, water-repellent expanded perlite is molded to fit the radius of the vessel. However, as with all rigid insulation, there is always a small interstitial space between the insulation and the vessel (i.e. a place for water to run in). So when water gets into such an insulation system, gravity will take over, but the water needs a place to exit.
Figure 3: CUI near continuously welded insulation support ring or stiffener.
Why do such mistakes in designs and practices continue? There are several reasons:
- The expertise needed for standards revisions is dwindling and often not located within the owner companies. This general lack of experts puts a tremendous burden on the remaining experienced personnel.
- There is more and more "combining" of jobs, and one person is often responsible for multiple disciplines and/or multiple plant sites.
- The less-than-popular job of standards revision can mean, in some companies, many years between updates.
- No one likes to have his or her name on a "revised" standard.
Sometimes new (or even old) products or methods prevent us from seeing and understanding all of these mechanisms and how they relate directly to CUI. (Learn about CUI Detection Techniques for Process Pipelines in this two part series.) Sometimes new products and devices are based on incorrect assumptions about CUI, and they can actually work against us. When we're talking about industrial facilities, which must safely operate year after year, small issues can make a big difference.
None of this is new knowledge, but it is a difficult thing for companies to change their corporate standards, and it is common for engineering companies to just pick up the old standard or specification. Small details like the type of insulation, coating imperfections or improperly designed insulation support rings can impact, in a catastrophic way, the ability of the equipment to operate safely.
And, with each new generation of personnel coming into corrosion management, these lessons must often be relearned, especially if industry standards do not provide a better basis for understanding and managing the actual causes of CUI.