Selected Answers

From Peter Bock of Performance Polymers Americas LLC on March 29, 2016:
Warren Brand, B. Brown and Ms. Lydia have technica ...read more Warren Brand, B. Brown and Ms. Lydia have technically correct answers, but they are looking at the question from a failure analysis point of view rather than a real-life operating point of view. They may be missing the forest while looking at the blisters on the trees. Taking a large tank in a refinery or chemical plant out of service, cleaning it and making it safe for entry is a million-dollar-plus project and is typically done on a seven- to ten-year cycle. If you remove a few blisters (as Warren suggests) and find ANY corrosion product or any steel loss on such a large tank, the cost of re-lining will probably be less than the potential cost of replacing steel seven years later (or a leak before then), if you did not properly identify and mitigate the root cause of the blisters. Right now, the tank is already out of service and clean, so a large part of the re-lining project cost (time and money) has been paid. Well-managed refineries where I have worked in the past actually had the in-house lining contractor on stand-by, so that after the owner's rep, third party inspector and lining supplier's rep inspected a tank lining and found it lacking, the contractor could immediately jump in and start the re-lining project.

From Lydia Frenzel of Advisory Council on March 25, 2016:
When I saw B Brown's statement,"oxygen molecule is ...read more When I saw B Brown's statement,"oxygen molecule is about half the size of a water molecule," I thought that something was wrong. I did a little research. The molecular volume of O2 molecule is 1.56 times the molecular volume of H2O water molecule. Another site verified this ; the oxygen molecular volume is 270 (cubic bohr)/mole; the water molecular volume is 159 (cubic bohr)/mol. Water migrates through a coating "easier" or "faster" than oxygen. As with B Brown, I agree with both prior answers.

From Mark Lewis of East Bay Municipal Utility District on March 24, 2016:
Another driver in this situation is whether or not ...read more Another driver in this situation is whether or not the lining is under warranty.

From B Brown of Self on March 22, 2016:
Another source of blistering in a water tank could ...read more Another source of blistering in a water tank could be a lining subjected to cathodic protection (CP) over-protection potential. CP is sometimes installed in municipal water storage. Problem is that the systems are rarely maintained, something that was not mentioned or defined in the question. Most coating manufacturers recommend a maximum potential of about -1 volt with reference to a copper/copper sulfate reference cell. When CP is the cause, the water inside the blister will be of high pH. One common cause is movement of the anodes where the anode ends up positioned too close to the coated surface. With rectified CP systems, failure of the CP controls is another common cause. With galvanic anode-applied CP, a magnesium anode placed too close to the surface to be protected can produce a potential more negative than -1 volt. Leaving the blisters alone is good advice if they are unbroken because the blister will be oxygen-deficient, and it is not uncommon to find the water inside clear and the substrate undamaged. You see, an oxygen molecule O2 is about half the size a water molecule H2O. I tend to agree with both answers. Leave them alone but check one or two to figure out what's going on. If the substrate is black, contaminants may be present and careful collection of water inside the blister may give you the answer. Does water wick away from portions of the surface? Oil may be present. Examining the surface with a short wavelength black light, 450n/m to look for surface fluorescence is not a fool-proof check for oil contamination but is a positive indication if it does glow.

From Jim Johnson of CHLOR*RID International Inc. on February 1, 2016:
It would be usual to break a few of the blisters j ...read more It would be usual to break a few of the blisters just to see and determine what is going on at the actual substrate. Assuming the blisters are caused by salt contamination, corrosion will be ongoing until the salts are removed. If the surface is not decontaminated and a new coating is applied, it will simply blister again, and corrosion will continue. Determining when to repair or recoat the surface requires far more information than is provided here.

From Warren Brand of Chicago Corrosion Group on January 27, 2016:
Short answer: leave them alone. Assuming these ar ...read more Short answer: leave them alone. Assuming these are osmotic blisters, once equilibrium has been reached across the semipermeable membrane (the coating), osmosis and corrosion beneath the blister should stop. There will most likely be some corrosion directly under the blister; but, again, if the corrosion has stopped, and the corrosion is not a structural issue, they can be left alone. If it makes you feel better, sure, pop a few. Of course, there are many other variables to consider. First, if getting access to the tank is very costly, then remedial actions may be prudent. But if you can inspect the tank easily and readily, then the blisters should be left alone.