Tank integrity testing: an umbrella term, and the four different tests hiding under it

Tank integrity testing is not a defined examination. It is a phrase that covers at least four distinct activities, and the measured cost per click on it, the highest in this batch, says buyers are using it as a shopping term rather than a specification. The useful first step is to translate it into the actual question being asked.

Question one: is it leaking right now

That is tightness or leak testing, and for underground tanks it is a regulated activity with prescribed methods and certified testers under a state program. It is not aboveground storage tank work, and the separation is explained in the article on underground versus aboveground tanks.

Question two: how much steel is left

That is a thickness and condition assessment: ultrasonic readings on the shell, floor scanning, visual examination of welds, roof and appurtenances. It answers what the metal has lost, not whether a leak exists today, and it is the core of an in-service examination. The distinction catches people out after an incident: a condition assessment carried out last year does not establish that a tank was sound yesterday, and a tightness test carried out yesterday says nothing about how much steel is left.

Question three: is it fit to keep operating

That is a fitness-for-service evaluation, and it is an engineering calculation rather than a measurement: remaining thickness against a calculated minimum, a corrosion rate, a projected date and a set of repairs ranked by urgency. It consumes the output of question two. The page on tank integrity assessment describes what such an evaluation has to contain to be usable.

Question four: will it hold what we now want to put in it

A change of service, a denser product or a higher fill height re-opens the original design calculation. Nothing about the steel has changed, but the required minimum thickness has. This is the question owners ask last and should ask first, because the answer can make the other three moot.

Why the wrong test gets bought

Because all four are sold under the same words, and a quotation for one can look cheap beside a quotation for another. An owner comparing a leak test against a condition assessment is comparing a morning against a week. The only defence is to write the question into the request rather than the product name.

Who is asking, and why it changes the answer

The request rarely originates with the person who owns the tank. It comes from an insurer renewing a policy, from a lender or a buyer during a sale, from a regulator following an inspection of the site, or from an internal audit finding. Each of those parties asks for a different document and has a different tolerance for conditional language. Tracing the request back to its source usually names the test on its own, and it prevents the common outcome where a technically correct report fails to satisfy the party that triggered it.

What a usable deliverable looks like

The method used and its limitations stated plainly, the acceptance criteria it was judged against, the raw data in a form that can be compared next cycle, the qualifications of whoever performed it, and a conclusion that answers the question that was asked. Anything that concludes only that the tank “passed” has not said against what. It is also worth checking who is expected to rely on the document. A report written to satisfy an internal audit, one written for an insurer and one written to support a regulatory filing are not interchangeable, and the difference is decided before the work starts rather than at the review meeting afterwards.