The following is a conversation between a Washington state regulator and Cal Chapman about corrosion protection for underground fiberglass tanks.
CP testing of Fiberglass Underground Storage Tanks – a Washington State regulator in field, asking for help in data interpretation:
I'm baffled here folks. I was inspecting an UST yesterday and found they were behind on CP testing. I opened the sump and found what appeared to be an Owens Corning fiberglass UST. The last test was good. I tested for continuity and protection. All meet criteria. Has anyone tested a fiberglass tank and got acceptable CP results? What is the cause of this? Most likely there is a metal strike plate inside the tank, which is where I made contact. Thanks.
Possible Answers and Other Info from Cal Chapman:
Jason, you've already had a lot of good comments offered. Here are some other, fairly general points to take into account. Before 1978 or so, UST's were just about always steel with simple coating on the outside, unless they were true Owens-Corning fiberglass tanks which were brought out in early 1970's, I believe. As has been pointed out, fiberglass tanks must have a steel striker plate, or taking stick readings for fuel levels would, sooner or later, punch a hole in the tank. That striker plate, though, is usually encased in fiberglass; you can get a magnet to stick to it, but you're not supposed to get a metal-to-metal contact. And even if striker plate is exposed, the rest of the tank's fiberglass body is completely non-conductive, so with a fiberglass tank, you won't ever get a "structure-to-soil" CP reading from that tank structure to the surrounding soils.
Around 1978, too, the first "Steel Tank Institute P3" (STI-P3 design) tanks came out. They were steel inner shell as the real structure and then had a complete fiberglass external "wrap" for coating, that was supposed to be at least 20 mils thick. Then they had a couple of magnesium anodes attached at factory, which were supposed to be un-bagged and put in native soil contact during tank installation (installers sometimes forgot to take the plastic off, and these anodes never were allowed to give protection to tank!). Those mag anodes would then give you continuity to electrolyte/soil, for any CP measurements done over life of this kind of tank. Even if the anodes were completely consumed, you had copper wire touching off to soil. CP readings on STI-P3 tanks are usually easy to get and fairly easy to interpret. You either get/got really negative voltages, meaning magnesium was doing its job and very little of the tank metal was in electrolytic contact.
Reviewing your later points and Derek's questions and points, here are a few more things to consider. And, yes, I've been around UST systems since late 1988, first doing leak detection and then assessment/cleanup, and then CP. If you think you're getting an actual tank reading, it's not the tank -- you're proving up that tank structure is fiberglass. But all manners of grounding structures, metal conduits for electrical runs, possibly steel vent line, the power brought to sub pump (consider both AC ground which could connect back to other metal in area, AND AC neutral wire that can be cross-connected to ground in on or more places), even rebar in or under concrete slab can give you electrical connections you don't expect. And you're right to be considering copper or brass tight-fill adapter as a separate metal from steel. Mixed-metal contacts measured in comparison to the CSE won't always give us the answers we figure we should see. I think you may have a few metal structures electrically contacting one another. Main question you and the UST owner/operator face: what fuel system components are susceptible to corrosion AND routinely contain fuel? They require CP.
Interestingly, the regulators in 1984 through present wrote a dumb approach into the rules (my opinion). They required electrically isolating steel pipe risers from tank structure, during tank system installation. For STI-P3 tanks, or the later "ACT 100" design tanks (after about 1990 or so, the fiberglass wrap over steel tank was made at least 100 mils thick, and then no CP was required on the tank long-term), this meant the tank installer did not keep steel tank metal and the metal riser structures electrically continuous -- I would much prefer that these pieces of steel are kept continuous at installation. What we have seen countless times is that risers need CP, and the tank metal needs CP. But they are not electrically connected. And there are no inexpensive ways to bond these various risers and tank metal together, to apply CP to an entire metal unit -- whether that's one tank and appurtenances, or several tanks, with risers, submersible pumps, etc. So to protect all components which routinely contain fuel (the way the rule reads), you have to decide whether each riser really needs and gets CP.
What this means is that you might measure "structure to soil" voltage between half-cell and one riser on tank, and get a particular reading. Then you measure on a different riser, and get a different reading. You can measure on vent pipe, which may be steel, and get yet another reading. Now, does this mean all these pieces of steel are screwed into a steel STI-P3 or ACT-100 tank, or into a fiberglass tank? You don't know yet.
To truly determine type of tank is tough. If you can figure out tank diameter, often done by pulling drop tube and then measuring both total depth to tank bottom, and the depth of riser to tank top. If you get a diameter that belongs to fiberglass tank, usually much different from typical steel tank construction dimensions, then you know what you have. Or you can get magnet and bond wire stuck to bottom inside of tank, and apply some current (safely, please) to look for voltage shift on all metal "touch points" accessible on outside of tank/tank system. If you get shift, you are electrically conductive and tank must be steel, AND it must have conductive path of that steel to soil side.
Much to talk about, certainly. And my one question is this: you say that in sub pump sump, you could see that the Owens-Corning fiberglass "ribs" were present -- I presume that's what you took to mean the tank was fiberglass, for sure. But the striker plate would not give you a metal-to-metal contact with anything on outside of tank. As we see from your photos just provided, the metal pipe risers for fill and ATG are both in contact with soil on the outside. These will give CP readings compared to half-cell, regardless of whether tank composition is metal or fiberglass. It looks like you are testing risers only. And there may not be any other metal involved.
Regarding the water across bottom of tank being conductive and tying other steel into CP circuit, it would have to be a little on the salty side to be conductive enough. If the product in tank is gasoline with ethanol, then the water on tank bottom has significant ethanol dissolved in it. Ethanol much prefers to be in soluble combination with the water and preferentially moves from the gasoline (non-polar liquid) to the water fraction. This should mean the water/ethanol conductivity is much greater than fresh water. So that would make it possible for your copper wire and screw on bottom of your stick to have electrical contact over to the ATG probe. But probe grounding contact is then somehow connecting back to everything else . . . that I'm not wrapping my head around; the ATG probe should generate a DC signal which is hard-wire-sent back to ATG console. That DC ground is likely same as electrical ground for the ATG system -- which needs robust and separate grounding, in addition to the AC-related grounding.