API 653 Exam Chapter 6

API 653 Exam Chapter 6 – Evaluation of Corroded Tanks

6.1 Introduction

Most of the calculations in API 653 reside in section 4: Suitability for Service. This is about what to do with tanks that are corroded or damaged in some way. It provides practical methods of determining whether a corroded tank complies with the well-established requirements of API 653, giving a viewpoint on whether it is safe for continued use.

For both practical and API 653 exam purposes it is important to understand the relationship between three codes involved in calculations of tank integrity: API 653, API 650 and API 579. Figure 6.1 shows the situation. API 650 is a pure construction code so its calculations and the parameters

Figure 6.1 Relationships between tank assessment codes
Figure 6.1 Relationships between tank assessment codes

they use are for new tanks. You can think of the objective of API 650 as being to produce a tank that will successfully pass its initial hydrostatic test, before being put into service. At this point, its job is done and API 653 takes over.

In dealing with in-service tanks that have undergone some degradation, API 653 overrides some of the integrity-related requirements of API 650. You can think of API 653 in three ways:

  • As an ‘out-of-design code’ assessment.
  • As a more realistic assessment, based on the realities of tank operation in use.
  • As a less conservative assessment, using up some of the margins hidden away in the design code that are not required, once it has passed its initial hydrostatic test.

All of these are correct, in their own way. A common technical thread running through all three, however, is the way that API 653 sees the issue of tank failure. Although not entirely unconcerned with leaks and environmental issues, the calculation routines of API 653 centre mainly around the objective of preventing structural failure and collapse of the tank. Leaks are undesirable, but they can be contained and/ or repaired. This same priority is shown by other API in service inspection codes, and those from other bodies also, so it is probably correct.

In practice it is easily possible for a tank with heavy and unsightly corrosion to pass an API 653 calculation assessment. Some people would therefore offer the view that API 653 is more interested in integrity rather than cosmetic appearance. If a tank fails the corrosion evaluation methods in API 653 the question of what to do gets wider. The options are:

  • Repair it to API 653/650.
  • Rerate it (by lowering the fill height of the tank) to API 653.
  • Do a fitness-for-service (FFS) assessment to API 579.
Figure 6.2 The section numbers of API 579
Figure 6.2 The section numbers of API 579

Looking at Fig. 6.2 you can see the context of API 579. It is a large detailed document containing many sections of detailed assessment methods for corrosion, bulging, weld problems and various other damage mechanisms that can affect tanks. Its use is not limited to tanks, but it fits well with the mechanical characteristics of atmospheric tanks of straight forward construction. An API 579 assessment is an out of (design) code assessment so, as such, carries an element of risk and even technical controversy, even though its methods are fairly robust and well-proven.

The detailed context of API 579 is, thankfully, not included in the API 653 exam body of knowledge. We will therefore concentrate on the evaluation methods of API 653 section 4: Suitability for Service.

6.1.1 The anticipated failure mechanisms

The evaluation methodologies of API 653 section 4 are almost exclusively concerned with preventing ductile failure. This is the main failure mode of corroded tanks in service. Do not confuse this with the competing mechanism of brittle fracture – related to material properties at low temperatures. This is primarily of interest for new tanks under hydrotest and is covered in a different section of API 653 (section 5).

6.2 The contents of API 653 section 4: suitability for service

Figure 6.3 shows the breakdown of the content of section 4: Suitability for Service and its partner Annex B: Evaluation of Bottom Settlement. Taken together, these provide good coverage on how to assess a corroded or damaged tank. Remember the key points about the calculation methods employed:

  • They are based on the design calculations of the tank construction code API 650 but include various overrides.
  • Areas in which API 653 clauses override the design requirements of API 650 are a fertile source of exam questions, as they are at the centre of the actual idea of API 653.
  • Only the simpler parts of Annex B: Evaluation of Bottom Settlement ever appear as exam questions. Some of the topics listed in the BOK never appear.

6.3 Tank roof evaluation

This comprises little more than a list of checkpoints. A few that arise as exam questions are:

Figure 6.3 Tank evaluation: the breakdown of API 653 section 4
Figure 6.3 Tank evaluation: the breakdown of API 653 section 4

Roof plates containing holes must be repaired or replaced. . Roof plates corroded to an average thickness of less than 0.09 in in any 100 in2 must also be repaired or replaced.

Note also these related safety points from API RP 575 (2.9):

  • When walking on a tank roof it is safer to walk on the weld seams than between them.
  • If planks are used as roof walkways they must span at least two rafters.
  • Watch out for blocked roof drains. Corrosion tends to be worse where water gathers in pools.
  • Tank vent valves fitted on the roof need to be checked for blockage.

There are generally no calculation questions related to roofs in the exam.

6.4 Shell evaluation

Evaluation of corroded shells is a major part of the API 653 body of knowledge. There are enough variations on the method to make for a fairly wide choice of exam questions. It also combines a useful group of principles that, in themselves, also turn up regularly in exam questions. Figure 6.4 shows these principles and we will look at them individually now.

6.4.1 Material strength

For tanks, this is straightforward. Construction steels are divided into well-established designations (‘grades’), each of which has a minimum specified yield strength (SY) and ultimate tensile strength ST (sometimes just abbreviated to Y and T). Remember that these are the minimum values the material has to meet to qualify for its designated grade – in practice it will almost certainly be stronger. Now look at Fig. 6.5, to see how these strength values are actually incorporated into tank design and repair.

Look at the principles highlighted in this figure. Although the rated yield (Y) and ultimate tensile strength (T) for any material are fixed, the percentage of this strength that is used to decide the value of allowable stress (S) to be used in the calculations varies depending on four things:

1. Whether it is a new-build/reconstructed tank or an existing one being assessed in its corroded condition.

2. Will the tank (in the future) only see its (lighter than water) petroleum product or will it be filled with water during a hydrotest?

Figure 6.4 Tank shell evaluation: the principles
Figure 6.4 Tank shell evaluation: the principles

The location of the actual shell course under consideration.

4. Whether it is an elevated temperature tank, designed for operation above 200 o F (for low carbon steel).

Referring to the table in Fig. 6.5 you can see how these parameters all affect the allowable stress value to be used. Note the key principles involved in deciding the S value:

Figure 6.5 Tank shell material strengths
Figure 6.5 Tank shell material strengths

It is normally the lowest of a percentage of Y (yield) or T (tensile) strength. There is no fixed rule as to which takes priority; you have to calculate both options and use the lowest value.

A higher S value is allowed if a tank is to be hydrotested. This is acceptable on the basis that the tank shell will only see the increased stress once, so any increased risk is small.

A higher S value is allowed for existing tanks (API 653) than for new or reconstructed tanks (to API 650). Again, a little more risk is acceptable with an existing tank because it has proved itself by passing its original hydrotest and has not fallen down since.

For existing tanks only, a higher S value is allowed in the upper areas of the shell than in the bottom two courses. This is because the upper courses are under less stress from the product weight, so are a bit less important from an integrity point of view.

If a tank is likely to see elevated temperatures (above 200 ℉) then a bit of extra caution is required. This takes the form of a stress reduction factor to allow for the fact that materials get weaker as temperature rises.

6.4.2 The principle of stress averaging

The stress on a tank shell is very simple. It is almost entirely pure hoop stress acting on the vertical plane, attempting to split the tank open from top to bottom. Simplistically, for an uncorroded tank, the hoop stress is calculated using the simple formula below:


Hoop stress S =______________


or rearranging to express it in terms of minimum required thickness (t)


t =______________



p = pressure (from the head of liquid)

d = diameter of the tank

S = allowable stress

E = weld joint efficiency (a type of safety factor

Using this equation it is easy to find the stress (S) or minimum required thickness (t) at any vertical position in the tank shell. The difficulty comes when you start to apply this

Figure 6.6 Tank hoop stress and averaging
Figure 6.6 Tank hoop stress and averaging

to a tank shell that is corroded along the vertical plane. Figure 6.6 shows the situation. In an uncorroded shell plate of uniform thickness the hoop stress is resisted by the uniform ‘thickness’ area of the plate and averaging is not an issue. Once the plate is corroded then the area resisting the hoop stress starts to vary from place to place; thinner areas of plate have less area to resist the force so the hoop stress value is higher. Some method is therefore needed to decide how to deal with this varying stress level.

The problem of averaging The danger with simple arithmetic averages is that they can be misleading. Look at the two examples in Fig. 6.6. Profile (a) is clearly more likely to fail than profile (b) but both have the same calculated arithmetic average.

API 653 attempts to neutralize this weakness by defining a calculated averaging length. This gives a length over which it is assumed (simplistically) that the hoop stresses on a corroded plane ‘average themselves out’. Figure 6.6 shows the calculation. Note how it is related to tank diameter (D) and minimum (spot) wall thickness (called, rather confusingly, t2), and how the square root sign reduces the sensitivity of L to large changes in D and t2.

6.4.3 Weld joint efficiency (E)

Weld joint efficiency (symbol E) is a concept found in several API and ASME codes. Crudely, it acts as a design margin that increases the minimum required shell thickness. In other design codes, joint efficiency is affected by the weld location, its configuration (single groove, double groove, etc.), amount of NDE and a more general, somewhat hidden, consideration of construction ‘quality’. For storage tanks the situation is simpler, being mainly governed by the design code and year edition to which it is built. You can see this in Table 4-2 of API 653.

Is joint efficiency (E) real?

Yes, as long as you think of it in the following way:

E is mainly for use in new tank design. Its purpose is to make sure a tank is strong enough to pass its initial hydrotest.

Once a tank is corroding in use its E is still relevant, however; if the corrosion is not near a weld it can be ignored (by making E = 1). This is a low-risk assumption acceptable to API 653.

Figure 6.7 summarizes the joint efficiency information contained in API 653 Table 4-2. Note how E is much lower

Figure 6.7 Shell joint efficiencies E
Figure 6.7 Shell joint efficiencies E

for tanks built to older design codes. This reflects the uncertainties and variability of material control and uncertain manufactured ‘quality’ prevalent in earlier times.

Joint efficiency for riveted shells

There are a lot of older riveted tanks still in use in the USA, so you can expect a couple of API exam questions about them. The principles are fairly straightforward:

For corroded areas near riveted seams, E varies from 0.45 to 0.92; the latter for a total of 12 rows of rivets (6 either side of the joint centreline). See API 653 Table 4-3.

As with welded seams, if corrosion is located well clear (6 inches away) from the nearest rivets the joint efficiency can be relaxed to E = 1.

6.4.4 Hydrostatic head consideration

An atmospheric storage tank shell is a good example of a very straightforward stress regime. If we conveniently ignore wind loading, the tensile stress on any point of the shell (at least above the bottom few inches where it meets the bottom annular plate) is an almost pure example of hoop stress. In turn, this hoop stress at any vertical location is directly proportional to the height of liquid (h) above it. We can see this from the hydrostatic head equation:

6.4.4 Hydrostatic head consideration
6.4.4 Hydrostatic head consideration

For a tank shell of uniform plate thickness over its full height, the hoop stress varies directly with height, as shown in Fig. 6.8. This will also apply if we just consider a single shell course. In practice, shell plate thickness normally

Figure 6.8 Hoop stress and height
Figure 6.8 Hoop stress and height

increases from top to bottom, modifying the situation slightly.

In assessing corroded shell plates, API 653 uses simple adaptations of the hoop stress equation. This changes slightly in practice although, importantly, not in principle, depending on the amount of corrosion in the specific plate course being considered. This can be explained by referring to it as the socalled ‘one-foot rule’.

Figure 6.9 The one-foot rule
Figure 6.9 The one-foot rule

6.4.5 The one-foot rule Figure 6.9 shows the equation to be used to calculate the minimum acceptable thickness. There are two options:

The corroded area is sufficiently large to warrant conducting the assessment of a full shell course plate. .

The corroded area is much more localized within a plate.

Note the difference between the equations used; they differ only by their use of either the term (H) or (H – 1) within the equation (hence the one-foot rule name). In practice, it can be difficult to decide exactly which to use, as people’s opinions will vary on whether a corroded patch is extensive enough to warrant a full course assessment or not. For API 653 examination purposes the question will tell you which assessment to use. Looking at the example in Fig. 6.9 you can also anticipate that the answers obtained from both methods do not vary by very much, particularly for larger tanks. The same principle is used when rerating tanks or calculating a safe full height for a tank that has a corroded shell.

6.4.6 What do we do about pitting?

In the world of API, pitting is seen as a very different thing to wall thinning. Although they may (and generally do) occur together, they are seen as separate for assessment purposes. The principles of dealing with pitting are common to several API in-service inspection codes. The points listed below summarize the approach of API 653 ( to pitting:

As a matter of principle, widely scattered pits in a tank shell can be ignored. This is because they are unlikely to result in structural failure of the tank.

The definition of widely scattered is (practically at least) defined by the content of Fig 6.10 taken from API 653 Fig. 4-2. Note the reference to 2 inches in 8 inches – cumulative lengths of pitting greater than this are considered too concentrated to be isolated and so must be assessed as a corroded area instead. Strangely, no depth of this pitting is defined – so pitting is pitting.

To restrict the allowable depth of pits, API 653 ( prohibits any pit from being so deep that it leaves less than 50% of the calculated tmin for that location in the shell. In assessing this, make sure to add any future corrosion allowance (the amount that may suddenly or gradually disappear before the next scheduled inspection).

Figure 6.10 Pitting evaluation: API 653 ( b)
Figure 6.10 Pitting evaluation: API 653 ( b)

The end result of applying these criteria to pitting is either that it can be ignored or that it is serious enough to be treated as a corroded area, using the main corrosion assessment method. We will look at this next.

6.4.7 Corrosion assessment – the big principle

Figure 6.11 shows the big principle of API corrosion assessment. This appears, with minor variations, in most API assessment codes for tanks, vessels and pipework and in API 579 – the advanced fitness-for-service (FFS) code relevant to them all. Note the key issues:

Figure 6.11 API 653 ( pitting assessment
Figure 6.11 API 653 ( pitting assessment

It is a double-barrelled assessment. meaning .

There are two separate assessments, and you have to pass both to be acceptable. and .

One assessment relates to the average thinning (on the assumption that this is what causes failure). 

The other assessment relates to the thinnest ‘spot’ reading (as long as it is not isolated pitting of course), under the assumption that this is what is going to cause the problems.

Looked at individually, each of these assessments is perfectly capable of providing a misleading picture of the integrity of the shell. Taken together, however (when you have to pass both), they provide a much better and balanced assessment. Figure 6.12 shows a sample (simplified) shell assessment.

6.5 API 653 (4.4): tank bottom evaluation

Tank bottoms cause most of the problems with the integrity of storage tanks. Over time, they are likely to suffer from a variety of problems of settlement, corrosion or even cracking, leading to leaks. API 653 provides a full list of causes of bottom failure in section 4.4.2. Figure 6.13 shows the major causes; note how most of them relate, either directly or indirectly, to corrosion of some sort.

6.5.1 Release prevention systems (RPSs)

API 653 (4.4.3) goes into some detail about so-called release prevention systems (RPSs). This generic term refers to any method (either a physical feature or an action that can assist in maintaining the integrity of the tank bottom, i.e. preventing a leak from happening) They are:

  • Internal inspection of the tank bottom
  • A leak detection system
Figure 6.12 Typical (simplified) shell assessment
Figure 6.12 Typical (simplified) shell assessment

Leak testing (to find problems in advance)

Cathodic protection of the underside of the tank bottom

Internal lining of the tank bottom

API 653 (4.4.3) does not go so far as recommending which of these is likely to be the best method – it simply reports that they exist, with a brief description of what they are. Do not expect lots of exam questions on this. In contrast to RPSs, release prevention barriers (RPBs) are used to contain or otherwise mitigate a bottom leak once it

Figure 6.13 Causes of tank bottom failure: API 653 (4.4.2)
Figure 6.13 Causes of tank bottom failure: API 653 (4.4.2)

has happened. Tank bunds, earthwork liners and drainage channels are the classic RPBs. There is a lot more detail in API 650 appendices I1 and I2, which are in the API 653 BOK.

6.5.2 Bottom evaluation – general principles

Tank bottom evaluation is subdivided into the three main elements that make up the tank bottom. .

The bottom plates themselves (4.4.5) are the main plates, normally overlapped and lap welded inside the tank.

The annular ring (4.4.6) is a thicker ring of fairly narrow plates, butt welded together in an annulus located under (and welded to) the lower shell plate course. The annular ring therefore supports almost all of the steelwork weight of the shell and its attachments. Older tanks may not always have this thicker annular ring but most modern ones do.

The critical zone is not a separate set of floor plates but simply the annular area extending 3 inches inward from the shell, all round the tank. You can think of it as just a particular critical region of the annular ring. It appears in API 653 ( and the definition section (3.10).

Not all of the technical points in these sections of API 653 get universal agreement. Some tank codes from other countries take different views on the risk of operating with quite thin bottom plates and annular rings, and so prefer to specify a greater minimum acceptable corroded thickness in preference to relying on an RPS/RPB as a last line of defence. Differences of opinion apart, API 653 does provide consistent and easy-to-follow acceptance levels, which many tank operators follow quite successfully.

6.5.3 Bottom plate minimum thickness API 653 (4.4.5) and Table 4-4

This is one of the more difficult equations of API 653 to understand. There is nothing wrong with the principles behind it; it is simply written using odd symbols and in a strange way. Part of the problem comes from the fact that it tries to incorporate scenarios in which a bottom may have already been repaired and/or have had a lining applied to arrest corrosion before it is assessed.

Note a key point about this equation, which we will call the MRT equation:

The ‘MRT’ equation just tells you how to calculate the remaining thickness of the bottom plates at the next

inspection. It does not actually give you a minimum acceptable value. That is given in API 653 Table 4-4. Here is what the equation in looks like, with some slightly simplified English:

6.5.3 Bottom plate minimum thickness API 653 (4.4.5) and Table 4-4

If you can fight your way through the confusing wording (and ignore all the stuff on repairs which may or may not have been done) this actually makes reasonable sense. It simply says:

MRT = current thickness – (time to next inspection  corrosion rate)

Qualify this by the following couple of points and you can see the equation for what it really is – a complicated way of expressing a simple idea:

  • The corrosion rate consists of the internal corrosion rate plus the external corrosion rate (as the tank bottom has an inside and an outside).
  • If the tank has an internal lining, just assume that the internal corrosion rate = 0.
  • If this tank has cathodic protection (CP), just assume that the external corrosion rate = 0. 
  • If either the internal or external surfaces have been repaired, the original corrosion rate that necessitated the repairs must be assumed to be still in force, unless you have evidence that it has changed (e.g. gone to zero if a lining has been applied after the repairs).

6.5.4 Minimum acceptable bottom plate thickness

Once you have calculated the remaining bottom thickness at the next inspection, the idea is that you then compare the results with Table 4-4 (reproduced in Fig. 6.14). This allows a basic minimum thickness of 0.1 in (2.5 mm), but can be reduced to half that (0.05 in or 1.25 mm) if the tank has either an internal lining or some method of containment to catch leaks if they do occur. These thicknesses are quite low,

Figure 6.14 Bottom plate minimum thickness: from API 653 Table 4-4
Figure 6.14 Bottom plate minimum thickness: from API 653 Table 4-4

and of course can be overruled either way by an RBI assessment.

6.5.5 What about the critical zone?

The 3 in wide critical zone (defined in API 653 definition 3.10 remember) may be either part of the annular ring (if the tank has one) or the bottom plate, if it does not. Figure 6.15 shows this specific requirement, a hybrid limit of the lower of:

Figure 6.15 Tank bottom limits: critical zone API 653 (
Figure 6.15 Tank bottom limits: critical zone API 653 (

50% of actual original bottom thickness (excluding any corrosion allowance) or

50% of tmin of the lower-shell plate course but

It must not be less than 0.1 in (2.5 mm), excluding isolated pitting as usual.

If the tank does have an annular ring, then this restriction on the minimum thickness of the critical zone still applies; it just falls within the annular ring rather than the bottom plates.

6.5.6 Minimum thickness of the annular plate ring (4.4.6)

The minimum acceptable thickness of the annular ring needs to be greater than that of the bottom plates, as it is under more stress from supporting the weight of the shell (plus sometimes bending from foundation settlement or other sources). The thicknesses are shown in API 653 Table 4-5 and summarized in Fig. 6.16. Note how three additional factors (that did not affect the main bottom plates) come into play:

  • The thickness of the first shell course
  • The actual ‘product’ stress in the first shell course
  • Whether or not the specific gravity of the product is greater than 1 (heavier than fresh water)

Owing to the fact that the condition of a tank bottom is one of the most important life-limiting factors and the main reason for actually doing internal inspections, the above subjects appear regularly as both open- and closed-book exam questions. The calculations are normally straightforward and you can see some typical examples at the end of this chapter.

6.6 Foundation evaluation: API 653 (4.5)

This short section situated at the end of section 4 contains a few general points on the condition of concrete foundations

Figure 6.16 Bottom limits: annular ring API 653 Table 4-5
Figure 6.16 Bottom limits: annular ring API 653 Table 4-5

but nothing about the main foundation-induced problem, which is that of settlement. This is physical movement of part of or all of the foundations, causing stresses and distortion of the tank structure above it. This is covered in API 653 Annex B, which we will look at next. Although physically separated in different parts of the code, section 4: Suitability for Service and Annex B: Evaluation of Tank Bottom Settlement are closely related to each other and should be considered together when assessing tank bottoms.

6.7 Bottom settlement: API 653 Annex B

Exactly how much of Annex B is in the API 653 examination body of knowledge (BOK) is open to some interpretation. This annex contains a lot of quite detailed information that maps well on to the way that tank settlement assessments are actually done in the field. Some of it, however, is far too complicated to be suitable for the API 653 exam. The BOK partly addresses this by mentioning a few exclusions, but at first reading it is not particularly easy to translate this into which part of Annex B you need, or need not, study.

Fortunately, the reality is fairly straightforward – most of the content of Annex B does not appear as exam questions, either because it is too complicated or it does not fit well into the multichoice exam question format. This means that there are generally fewer settlement-related questions than perhaps you might expect, given its importance. They are mainly predictable open-book questions and not too difficult. We will look at the most popular subjects and then at some simple questions at the end of this chapter.

Annex B is set out along fairly logical guidelines as follows:

The different types of bottom settlement, categorized predominantly by the way that they affect the shell. The types are uniform, planar and differential.

Measurement of various types of bottom settlement. The three clear types included in the BOK are: –

Edge settlement (B-2.3)

Bottom settlement near the shell

Bottom settlement remote from the shell nearer the centre of the tank floor.

Evaluation of the types of settlement against acceptable limits, given in the form of graphs or simple linear formulae.

Decisions based on the results of the evaluation about performing additional NDE or repair. Details of the repairs themselves are not included in Annex B-2, as they are fully covered in API 653 section 9.

6.7.1 Types of settlement

The problem with storage tanks is that they are structures that have little rigidity. Most have no cross-bracings supporting the shell, leaving only the hoop strength of the thin shell to support the load exerted by the product liquid. Even the margins in material thickness are small, compared to those used in pressure vessels and pipework. To make things worse many older tanks were built on either simple rubble foundations or low quality concrete.

The result is that when tanks move (or ‘settle’) on their foundations, the uniform hoop stress regime in the shell is soon disturbed, leading to unpredictable stresses, strains and distortion. This can soon lead to cracking and leaks, or, in extreme cases, collapse of the tank. To try and avoid this tank owners should take regular measurements of tank settlements over time.

Distortion of a tank shell involves a rather complicated three-dimensional geometry that is not easy to measure, or even describe in simple terms. Measurement and analysis is therefore generally left to specialist contractors. Quite a bit of API 653 Annex B is devoted to describing these specialist techniques – the good news is that they are not really included in the BOK, and are too complicated to appear in the form of API exam questions. For exam purposes, you can think of the types of tank settlement as being simplied into three separate components as follows (see Fig. 6.17):

Uniform settlement. This is ‘sinking’ of the tank perfectly vertically downwards, with no tilting, twisting, buckling or any other type of distortion whatsoever. It rarely happens exactly like this in practice, but you can think of it as one of the components of any real settlement pattern.

Planar tilt. You can think of this as the tank shell and bottom assembly simply tilting over to one side. Again, it does this in a perfectly even and uniform manner – the shell remains perfectly circular without it buckling or kinking. Owing to the tilting there is a small, usually insignificant, increase in vertical liquid height, and therefore in the hoop stress at the bottom of the shell. Looking at Fig. 6.17 you can see how the shape of a uniformally tilted tank can be represented by a perfect cosine wave.

Figure 6.17 Tank settlement: API 653 Annex Fig B-3
Figure 6.17 Tank settlement: API 653 Annex Fig B-3

Non-planar differential settlement. This is the one that causes the real problems in use. The fact that a tank is a very thin shell structure means that as it settles and tilts, it nearly always distorts as well. This causes extra shell stresses, sticking of floating roofs and breakage of support columns, girders and connecting nozzles and pipework.

Is non-planar differential settlement in the exam?

Probably not. Quite a lot of API 653 Annex B is devoted to the measurement and description of differential settlement. It is done by measuring the difference between the actual shape of the settled tank compared to that of the planar-tilted ideal cosine curve. The greater the difference, the greater is the differential distortion and the larger the resulting stresses. You can see the principle explained in API 653 (2.2.4) and Figs B-4 and B-5 – described as the ‘least-squares fit method’. Fortunately, this is far too complicated to make an API exam question so the calculation will not appear in the exam. You can maybe expect a question on the principles of tilting and distortion and their effects but no calculations.

To repeat: you do not need to learn the specific details and equations of ‘out-of-plane’ differential settlement. They will not be in the closed-book exam.

6.7.2 Edge settlement

The various types of tank settlement that we have just looked at are not, in themselves, the problem. It is their distortion effects that are important. API 653 Annex B-2 divides these into two separate situations – settlement distortion under, or very near, the edge of the tank (‘edge settlement’) and that well away from the edge nearer the centre of the tank (‘bottom settlement’). Unlike shell distortions, these are easy to assess using simple graphs and calculations, and so appear in the API 653 BOK. Expect one or two exam questions to appear on this subject, but no more.

Edge settlement is when the tank bottom settles sharply around the edge of the tank often caused by ‘washout’ or crumbling of foundations. Looking at Fig. 6.18 (see also Fig. B-6 in API 653) you can see how this results in sharp deformations of both the bottom and shell steelwork, causing serious bending stresses. The API 653 Annex B assessment uses a simple ratio of the (vertical) length of edge settlement to its length (in the tank radial direction). It then divides this into two separate scenarios as follows:

If the edge settlement is an area where the tank bottom

Figure 6.18 Tank bottom edge settlement: (API 653 Annex B-2.3)
Figure 6.18 Tank bottom edge settlement: (API 653 Annex B-2.3)

plate welds run near-parallel (±20o ) to the shell, the extent of edge settlement is called Bew.

If the edge settlement is in an area where the tank bottom plate welds run near-perpendicular (±20o ) to the shell, the extent of edge settlement is called Be.

Note how both of the above cases refer to the vertical amount of settlement (B); they are simply renamed Bew or Be, depending on which way the local tank bottom plate welds are orientated.

Assessment against the edge settlement graphs Once the edge settlement measurements are available, the assessment is easy. Figure 6.19 summarizes the content of API 653 Figs B-11 and B-12.

To use the graph, simply enter the graph on the horizontal axis with the measured radius (R) of the settled area. Remember that this is not actually a true radius, as such, but the length of the settled area measured in the tank radial

Figure 6.19 Edge settlement assessment: API 653 B-11 and B-12
Figure 6.19 Edge settlement assessment: API 653 B-11 and B-12

direction. Then, using the relevant curve for the tank diameter in question, read off Bew (or Be) from the vertical axis. This is the maximum allowed dimension of Bew (or Be) acceptable to API 653. Any more than this needs repair or further specialized evaluation.

API 653 exam papers do not like to contain a lot of figures or graphs, so exam questions tend to be limited to code clauses that do not require use of the graphs. Note the following key points: 

Bew, when bottom welds are (±20° ) parallel to the shell, is more conservative than Be so it is normal to do this assessment first (B-2.3.4).

When Bew or Be are ≥75 % of their limit (and larger than 2 in) the welds in the region should be inspected with PT/ MT to check for cracking (API 653 Figs B-11 and B-12). . Any bottom plate exhibiting a strain (permanent plastic deformation) of more the 2–3 % should be replaced (B-4.2).

The settlement graphs were originally developed for 1 4 in thick tank bottoms but can also be applied with reasonable accuracy for thicknesses between 5/16 in and 3/8 in.

In general, settlement occurs fairly slowly over the first few years of service (B-3.4.5).

Watch out for the edge settlement clause B-3.4.6 (a to d). There are possible exam questions in here.

6.7.3 Bottom settlement

Remember that this is assessed differently depending on whether the settlement is near the shell or further away, towards the centre. The methods are very similar, using a simple equation based on the depth (BB) of the settlement (bulge) compared to its radius (R). Figure 6.20 shows the situation. Note how the simple linear graph in API 653 Fig. B-10 contains absolutely nothing new; it is simply the equation BB = 0.37R shown in a different way.

As with edge settlement, the limits of bottom settlement

Figure 6.20 Assessing bottom bulges: API 653 B-2.4 and B-2.5
Figure 6.20 Assessing bottom bulges: API 653 B-2.4 and B-2.5

are there to prevent the floor plates being bent too sharply, which would cause risk of cracking of the lap welds. It may still be technically possible to operate the tank with a settlement in excess of these limits, but an engineer’s assessment would be required. As bottom settlement limits can be determined without using graphs (i.e. using the BB = 0.37R equation) this subject can appear as an (open-book) exam question.

The presentation in the code is a little confusing, mainly because the combination of equations and graphs (meaning exactly the same) makes this subject more complicated than it actually is.

Figure 6.21 gives a summary of code formulae used in a tank evaluation.

Now try these practice questions.

6.8 API 653 section 4: evaluation: practice questions (set 1)


Q1. API 653: isolated pitting
How many pits of less than half the minimum required wall thickness, each of 0.25 in diameter, are allowed in an 8 in vertical line of corrosion on a tank?



Q2. API 653: joint efficiencies (unknown construction standard)
What is the weld joint efficiency factor for a tank built in 1950 with a single butt welded joint with a ‘back-up bar’? The construction standard is not known.



Q3. API 653: allowable stress S
If the yield strength (Y) of a material of construction is 35 000 psi and the tensile strength (T) is 60 000 psi, what is the value of the maximum allowable stress for the bottom two courses of a tank made of this material?



Q4. API 653: allowable stress for unspecified steel
What is the design strength of unspecified steel used in the top courses of a storage tank?



Q5. API 653: calculation of the averaging length
The minimum shell thickness t2 (exclusive of isolated pits) in a tank of 100 ft diameter is measured at 0.375 in. What is the critical length over which the thickness readings should be averaged out?



Q6. API 653: calculation of minimum acceptable shell thickness
A tank has the following dimensions:
Nominal tank diameter 150 ft
Total height of tank 30 ft
Maximum fill height 30 ft
Number of courses equally spaced 5
The specific gravity of contents 0.85
Construction code API 650 basic standard
5th edition
Material of construction A283 C
What is the minimum acceptable thickness tmin for the bottom
course? Use the ‘full course’ equation tmin = 2.6(H – 1)DG/SE



Q7. API 653: hydro test height (entire course consideration)
An API 12C 15th edition tank of diameter 150 ft has a bottom course made of A285C. The measured minimum average thickness for the entire bottom course is 0.2 in. What is the maximum allowable fill height for hydro testing this tank?



Q8. API 653: calculation of allowable t2 (including corrosion allowance)
A tank has an average minimum measured thickness (t1) of 0.5 in and has a design corrosion allowance of 0.125 in. What is the minimum allowable individual thickness (t2) of a corroded area anywhere, excluding any isolated pitting?



Q9. API 653: bottom annular plate thickness
If the stress in the first course of a tank is 26 000 psi and the course is 1/2 in thick, what is the minimum thickness for the bottom annular plate?



Q10. API 653: hydro test fill height for a complete tank
A tank is constructed to an unknown code but all the joints are butt welded. It is 120 ft in diameter and the material of construction is an unknown low carbon steel. There is localized corrosion on the bottom course and the minimum thickness is 0.25 in. What is the maximum fill height for hydro testing the complete tank?



6.9 API 653 appendix B: tank bottom settlement: practice questions (set 2)

Q1. API 653: types of settlement
Shell settlement can be made up of three types. What are they?



Q2. API 653: types of settlement
Which of these does not induce stresses in the tank structure (but may in the connections)?



Q3. API 653: shell settlement
What shape does differential settlement follow?


14. Q4. API 653: maximum allowable deflection
Maximum allowable out-of-plane deflection S (ft) is given by the equation:
S< (L2 – Yx11) [2(ExH]
L = arc length (ft) between measurement points = 20 ft
Y = yield strength (psi) = 20 000 psi
E = Young’s modulus (psi) = 24×106 psi
H = tank height (ft) = 50 ft

What is the maximum allowable out-of-plane deflection S?



Q5. API 653: shell settlement
In general, when is most settlement presumed to have occurred for (say) a 10-year-old tank that is showing settlement?



Q6. API 653: edge settlement repairs
It is acceptable to repair the bottom-to-shell weld without further investigation by an experienced engineer as long as the actual settlement is not greater than:



Q7. API 653: measured edge settlement
A tank which when new has the center of its bottom lower than its edges is called a:



Q8. API 653: settlement
What are the ‘least squares fit’ method all about?



Q9. API 653: settlement
What can be caused by rigid (planar) tilt?



Q10. API 653: settlement
A lack of circularity at the top of a tank is typically a feature of:


Click Here To Read Next API 653 Exam Chapter 7 –API 650: Tank Design

If You Want To Do API 653 Training course or Want Buy Latest Self Study Material Please feel free to contact us

BUY API 653 Learning Package Study Material
99% Passing Guarantee
Cost: $25.00
Email Us To Buy: upweld.org@gmail.com
Click Here To Contact On Whatsapp https://wa.link/4dn6ee
Ideally document for passing the examination. Free guidelines during learning.
We give you lot’s of study material Free of cost with this question bank.

Leave a Reply