Setting the Next Internal Inspection Date from Two UT Surveys
An 80 ft diameter x 48 ft high tank stores product of G = 1.0. Its shell butt welds were fully radiographed at construction, so E = 1.0, and the authorized allowable stress is S = 25,000 psi. The owner's procedure carries no additional corrosion allowance for the shell calculation. Two UT surveys of the bottom course are on file: minimum average 0.540 in. in 2015 and 0.504 in. in 2021. An MFL floor scan in 2021 gave a minimum average bottom thickness of 0.160 in. outside the annular ring, against 0.220 in. in 2015; the annular ring averages 0.230 in. and has not been re-surveyed since 2015. The tank must stay in service through the 2022 planning cycle, and the terminal wants the next internal inspection date set from these data.
1. Compute the bottom-course t_min and remaining life, and the bottom-plate remaining life. Which component governs?
Read the worked answer
Shell, bottom course:
- t_min = 2.6 x 80 x (48 - 1) x 1.0 / (25,000 x 1.0) = 9,776 / 25,000 = 0.391 in.
- Rate = (0.540 - 0.504) / 6 yr = 0.0060 in./yr
- Remaining life = (0.504 - 0.391) / 0.0060 = 18.8 yr (about 2040)
Bottom plates (0.10 in. acceptance floor):
- Rate = (0.220 - 0.160) / 6 yr = 0.0100 in./yr
- Remaining life = (0.160 - 0.100) / 0.0100 = 6.0 yr (2027)
Governing component: the bottom, by a factor of three. The shell is not close to being the limiter. Also flag the annular ring as an open item: 0.230 in. against a 0.1875 in. code floor looks fine, but there is no second data point, so its rate is unknown and it must be re-surveyed at the next opportunity (the annular ring may also need to be thicker than the floor where it carries shell loads, so check the API 650 basis rather than the floor alone).
2. Recommend the next internal inspection date and any mid-interval actions, with justification.
Read the worked answer
Calculation says 2027 (6 years from the 2021 survey). The code logic is that the interval must be short enough that no component reaches t_min before the next inspection, and the rate must be derived from at least two sets of measurements.
I would set 2026 and require a mid-interval look in 2024. Reasons:
- The 0.010 in./yr rate is a two-point average over six years. Two points give a rate, not confidence; the 2015 to 2021 period may not represent the next six years.
- "Minimum average" deliberately smooths local pitting. A floor that averages 0.160 in. can easily contain 0.09 in. pits today, and the average will not show them.
- The 2024 action need not be a full entry: an MFL or a dense UT grid in the areas that drove the 0.160 in. average is enough to confirm the rate and to decide whether a bottom repair can be scheduled at the 2026 shutdown rather than discovered during it.
Document the interval, the rate used, the data set it came from, and the inspector's approval. Recalculate after each survey; do not carry the old date forward.
3. Management wants to defer to 2029 to align with another unit's shutdown. What is your response?
Read the worked answer
On the present data, no. Deferring to 2029 means eight years past the 2021 survey:
- Projected average bottom thickness = 0.160 - 8 x 0.010 = 0.080 in. That is below the 0.10 in. floor, and the average understates the pitting, so a through-wall perforation before 2029 is a realistic outcome rather than a tail risk. A bottom leak in a product tank is a release, a containment cleanup, and an unplanned emergency entry, which is exactly the scenario the inspection interval exists to prevent.
What would have to be true before I would sign a deferral:
- A new MFL/dense UT survey that shows a materially lower rate or a higher true minimum than the 2021 data.
- A formal risk-based evaluation (API 580/653 basis) documenting the risk, the release-prevention and leak-detection measures that will be in place, and the contingency plan.
- Written acceptance by the owner's engineer and the authorized inspector, with the shortened interval on record if the data do not support the deferral.
The cost of an early internal inspection is small next to the cost of an unplanned bottom failure.