API 653 Tank Calculations: What the Exam Tests

By Sandeep Bandyopadhyay, Senior aerospace engineer · PMP, MBA, Six Sigma

TL;DR

The calculations section of the API 653 body of knowledge lists 15 calculation types, all aimed at existing tanks, not new ones. The corrosion rate, remaining life, maximum fill height and bottom plate formulas are supplied in the exam. What you carry unaided is judgment: which thickness goes into which variable, which units apply, and the tables marked with an asterisk.

Original questions, not exam dumps. Every CertCrafter question is written from the published body of knowledge and the referenced codes. None is copied from a live exam, and CertCrafter is independent: not affiliated with or endorsed by any awarding body.

A row of large oil storage tanks standing in an industrial yard
Photo: Oil storage tanks - geograph.org.uk - 3912626 by John Myers (Wikimedia Commons), CC BY-SA 2.0, via Wikimedia Commons.

Open the API 653 body of knowledge and the first domain is calculations. It lists 15 types, from corrosion rates to lap patch plates, and it says plainly that the questions are about existing tanks, not new ones. That one sentence cuts the workload. You are not designing a tank. You are checking whether a tank that has been in the ground for thirty years can keep holding product.

This page covers the concept behind the math, how inspectors use it on a site, and where candidates lose marks.

What does the calculations domain of API 653 cover?

It is domain I, titled Calculations and Tabular Evaluations for Evaluating Thickness Measurements, Weld Sizes, and Tank Integrity. API’s list, in its own order:

Category What you determine
Corrosion rates and inspection intervals Metal loss, corrosion rate, remaining life, remaining corrosion allowance, interval
Joint efficiencies Joint type, radiography extent, the factor E from the tables
Maximum fill height Hydrostatic test height, localized or whole shell course
Weld sizes for shell and roof openings Sizes and spacing, read from API 650 and API 653 figures and tables
Hot tapping Requirements and minimum spacing to an existing nozzle
Settlement evaluation Edge, near-shell and remote bottom settlement
Number of settlement points Survey points for the settlement check
Impact testing Design metal temperature, material group, whether testing is required, whether values pass
Existing shell minimum thickness Allowable stress, E, liquid height, minimum acceptable thickness
Reconstructed shell minimum thickness Design and hydrostatic test conditions
Corroded area Minimum thickness, critical length, average thickness
Pitting Maximum pit depth, pit length in any 8 in. of vertical length
Bottom plate minimum thickness Minimum remaining thickness, maximum period of operation
Replacement plates Minimum dimensions
Lap welded patch plates Minimum thickness, weld size, allowable patch size

Notice how many rows are lookups. Joint efficiency, weld sizes, impact testing, settlement points: those are tables and figures, not algebra. The domain is as much a reading test as a math test.

How do corrosion rate and remaining life work?

Start with the idea. Steel thins. If you measured the same spot twice, the difference over the years between the readings tells you how fast it is going. Divide what is left above the minimum allowable thickness by that speed and you have the years remaining.

API’s body of knowledge gives the two formulas. Corrosion rate is t previous minus t actual, divided by the years between them. Remaining life is t actual minus t minimum, divided by the corrosion rate. The catch sits in the definition of t previous: it must be the reading taken at the same location as t actual. Two readings from two different spots tell you nothing about a rate.

Video: API 653 — Settlement, thickness, interval — one calculation, one number. API 653 — Settlement, thickness, interval — one calculation, one number. Watch on YouTube · CertCrafter
Image: video thumbnail, source: CertCrafter on YouTube (@certcrafters).

If you prefer to watch it done, the channel has a short on the settlement, thickness and interval calculation.

Thickness against time at one location: t previous, then t actual, with the slope as the corrosion rate, the gap between t actual and t minimum as the remaining corrosion allowance, and the distance to the t minimum line as remaining life
The two formulas on one picture. Illustration: CertCrafter, from API 653 Body of Knowledge, March 2027.

Worked example: remaining life of a shell course

These figures are made up for practice. Take one location on a shell course. The earlier reading was 0.500 in., the new reading is 0.440 in., and the two inspections were 6 years apart. The minimum allowable thickness for that location came out of your shell calculation at 0.300 in.

  1. Metal loss: 0.500 - 0.440 = 0.060 in.
  2. Corrosion rate: 0.060 / 6 = 0.010 in. per year.
  3. Remaining corrosion allowance: 0.440 - 0.300 = 0.140 in.
  4. Remaining life: 0.140 / 0.010 = 14 years.

The arithmetic took ten seconds. What the exam checks is the setup. A question might hand you five readings and two thicknesses and ask which pair belongs in the formula. Another may give a corrosion allowance and ask you to add or subtract it before you start, which the body of knowledge says to do “based on requirements from the exam problem.”

And fourteen years is not the interval. API 653 Section 6 sets the rules for inspection intervals, and those rules, not your division, decide when the tank gets looked at again.

What do inspectors actually do with these numbers?

On a site, the math starts weeks before anyone opens a book. A crew takes ultrasonic thickness readings across the shell courses and the bottom. Somebody has to decide which readings are comparable, because a rate is only as honest as the two points behind it. A reading taken on a new grid next to a reading on the old grid is not a pair.

Then comes the part the exam compresses into one question: the inspector picks the governing location. A shell course with plenty of metal on average can still fail on one thin patch, and the patch decides the outcome. That is why the corroded area category asks for t2, the critical length, and the average thickness t1 of the area, and why the pitting category treats pits separately from general loss.

Joint efficiency is the field detail most likely to be skipped. The factor E depends on the joint type and the extent of radiography done when the tank was built. If the file is missing, as in the old tanks many owners run, the inspector has to establish what can be supported from the records rather than assume the best. The free bottom lap patch question in the sample set tests a related field call, where two separate findings are rejectable.

Oil storage tanks seen across open ground, with pipework and walkways at the base
Photo: Oil storage tanks - geograph.org.uk - 6448277 by N Chadwick (Wikimedia Commons), CC BY-SA 2.0, via Wikimedia Commons.

Which formulas does the exam give you?

The body of knowledge is specific. The maximum fill height formula, for a localized corroded area or an entire shell course, will be provided, and the inspector is not expected to rearrange the t min formula to get it. The bottom plate formulas for minimum remaining thickness at the next inspection and for maximum period of operation are provided too.

Corrosion rate and remaining life are the two to know cold; API says the formulas and the interval rules “may be closed-book.” The API 653 tables marked with an asterisk, among them the joint efficiency tables 4.2a and 4.2b and the bottom thickness tables, must be answerable without the documents.

So split your study. Memorize two short formulas and the starred tables. Practice the rest on screen, where the references appear as PDFs, because the exam is 110 closed-book questions followed by 60 open-book ones, and the open-book half is where table and figure lookups live.

Five steps in order: take t actual at the location, find t previous at the same location, work out the corrosion rate, work out the remaining life, then set the inspection interval from API 653 Section 6
The chain every corrosion question follows. Illustration: CertCrafter, from API 653 Body of Knowledge, March 2027.

Where do candidates lose marks on tank calculations?

Units come first. The body of knowledge notes that tanks are built in US customary or SI units, that you must work in the unit the question uses, and that some tables are printed in only one of them, API 653 Tables 4.1a and 4.1b among them. The two sets are parallel, not exact conversions of each other. One of the free questions turns this into a scenario: field readings in millimeters, a table in inches. The answer is to convert to the table’s unit and finish the whole job in that system.

Second, the wrong input. Remaining life uses t actual and t minimum. Corrosion rate uses t previous and t actual. Put t minimum into the rate formula and you get a clean number that is wrong, and the distractor options are usually built from exactly that slip.

Third, studying the wrong standard. API 653 leans on API 650 for construction details, and the March 2027 sheet lists API 650 in its 14th Edition. Notes written for API 510 will send you to ASME Section VIII calculations that are not on this exam.

Fourth, time. At 110 questions in 2 hours 45 minutes, a closed-book calculation gets a minute and a half. Know the shape of the answer before you pick up the calculator.

What should you study first?

Work through the eight exclusions so you stop worrying about wind loads and venting. Then take the two formulas and do ten corrosion problems with deliberately messy inputs: mixed units, a corrosion allowance to add, a previous reading from the wrong spot. After that, rehearse the lookups: joint efficiency, impact testing, settlement points.

When you want a graded check, the full API 653 bundle has the calculation items marked closed-book or open-book as the exam splits them, and the free quiz is a quick sample. The API 653 product page holds the exam facts, and the guides hub lists the rest. For the editions your calculations must match, read the March 2027 changes and the free practice questions.

Try these free questions

  1. A thickness table in the governing standard is published only in US customary units, while the field readings on TK-1204 were logged in millimetres. How should the evaluation be carried out?

    1. Round the millimetre readings to the nearest sixteenth of an inch first, since the table is built on fractional inch steps and interpolation is not permitted.
    2. Convert the readings to the units of the published table and keep the whole evaluation in that one system, since the two unit sets are independent and must not be mixed.
    3. Apply the table as printed to the millimetre values, because the numerical steps in each system are close enough for an acceptance decision.
    4. Treat the absence of an SI table as a gap and use the corresponding SI table from the construction standard in its place rather than converting any of the field data.
    Show the answer and explanation

    Answer: B. API 653 and API 650 publish US customary and SI values as parallel but independent sets, and the SI figures are rounded rather than exact conversions of their customary counterparts. Mixing a millimetre measurement with an inch table therefore produces an evaluation that belongs to neither system, and the correct handling is to convert the measurements into the units of the table being used and complete the whole calculation in that system. Both unit systems are examinable, and some tables exist in only one of them, so this situation is routine rather than exceptional. Option one adds rounding to fractions, which throws away real precision in the readings and has no basis in the standard. Option three reads inch values against millimetre numbers, an error of more than a factor of twenty five that would accept a badly thinned plate. Option four substitutes a table from a different context instead of doing the conversion the situation actually calls for.

  2. [Open book] The approved repair procedure for the TK-1204 bottom specifies a 1/4 in. A36 lap patch, lapped at least 1 in. onto sound plate, with a continuous fillet on the exposed side. The lap measures 3/4 in. along one edge and the fillet is stitched on 4 in. centres. What is rejectable?

    1. Only the stitched fillet, because the lap dimension is a fabrication preference rather than a requirement.
    2. Only the short lap, because a stitched fillet weld is normally accepted on bottom plate repairs.
    3. Both findings, since the lap must meet the stated minimum and the fillet must be continuous.
    4. Neither finding, provided a vacuum box test over the fillet weld shows no leakage at all.
    Show the answer and explanation

    Answer: C. Both findings are rejectable against the procedure the owner approved. A lap-welded patch plate has to overlap sound plate by at least the dimension the procedure states, here 1 in., because the overlap carries load into the parent plate and gives the fillet something to sit on, so 3/4 in. along one edge is short. The fillet also has to be continuous, since the joint has to be liquid tight and a stitched weld leaves open crevices on the product side of the bottom that will corrode and then leak. The first option treats the lap dimension as a preference when it is a stated requirement. The second option accepts stitch welding and gives up tightness. The fourth option offers a vacuum box test as a substitute for the weld detail, but a passing leak test shows only that the weld is tight today, not that the joint has the strength and the corrosion resistance the detail was designed to give.

Take the full free 10-question sample with scoring →

Frequently asked questions

Do you have to memorize the API 653 formulas?

Fewer than most candidates expect. The body of knowledge says the maximum fill height formula, and the bottom plate formulas for minimum remaining thickness and maximum period of operation, will be provided in the exam. It also says the corrosion rate and remaining life formulas may be closed-book, so learn those two. The API 653 tables marked with an asterisk must be known without the documents.

Are the calculations for new tanks or existing tanks?

Existing tanks. The body of knowledge states that calculation questions are oriented toward existing tanks, not new tanks. Reconstructed tank shells appear as their own category, but designing a new tank from scratch does not.

Which calculations are not tested?

The body of knowledge lists eight: required thickness for wind, earthquake and small internal pressures; nozzle calculations for external loads; flange calculations; brazing requirements; venting requirements; ladder, stairway and other structural calculations; bottoms supported by grillage; and variable point method calculations. Skip them and spend the hours elsewhere.

Official sources

  1. API 653 Body of Knowledge, March 2027 — accessed 16 October 2026
  2. API 653 Publications Effectivity Sheet, March 2027 — accessed 16 October 2026
  3. API 653 Aboveground Storage Tank Inspector certification page — accessed 16 October 2026

Exam rules change. Check the awarding body's official page before you apply. CertCrafter is independent and is not affiliated with or endorsed by any awarding body.

Was this helpful?

Comments

    Comments are read by a person before they appear. How we moderate.