mechanical design of tank structure

Note: The source of the technical material in this volume is the ProfessionalEngineering Development Program (PEDP) of Engineering Services.

Warning: The material contained in this document was developed for SaudiAramco and is intended for the exclusive use of Saudi Aramco’semployees. Any material contained in this document which is notalready in the public domain may not be copied, reproduced, sold, given,or disclosed to third parties, or otherwise used in whole, or in part,without the written permission of the Vice President, EngineeringServices, Saudi Aramco.

Chapter : Vessels For additional information on this subject, contactFile Reference: MEX20303 J.H. Thomas on 875-2230

Engineering EncyclopediaSaudi Aramco DeskTop Standards

Evaluating MechanicalDesign of Tank Structures

Engineering Encyclopedia Vessels

Evaluating Mechanical Design of Tank Structures

Saudi Aramco DeskTop Standards

MODULE COMPONENT PAGE

INTRODUCTION................................ ................................ ................................ ..................... 1

DESIGN FACTORS THAT ARE CONSIDEREDIN THE MECHANICAL DESIGN OF STORAGE TANKS ................................ .................. 2

Metal Temperature ................................ ................................ ................................ .................... 2

Pressure ................................ ................................ ................................ ................................ ..... 3

Specific Gravity of Stored Liquid ................................ ................................ ............................. 3

Corrosion Allowance ................................ ................................ ................................ ................. 5

Other Loads ................................ ................................ ................................ ............................... 7

Settlement ................................ ................................ ................................ ................................ 10

DETERMINING WHETHER CONTRACTOR-SPECIFIEDSHELL THICKNESSES ARE CORRECT ................................ ................................ ............ 16

API-650 Requirements ................................ ................................ ................................ ............ 16

One-Foot Method ................................ ................................ ................................ .................... 19

Variable-Design-Point Method ................................ ................................ ............................... 25

Hydrostatic Test Case.........................................................................30Design Case........................................................................................ 35

DETERMINING WHETHER CONTRACTOR-SPECIFIED WIND GIRDERREQUIREMENTS FOR OPEN-TOP TANKS ARE CORRECT ................................ .......... 39

Pertinent Sections of API-650 ................................ ................................ ................................ . 41

General Wind Girder Requirements ................................ ................................ ........................ 42

Top Wind Girder Design Calculations ................................ ................................ .................... 43

Intermediate Wind Girder Design Calculations ................................ ................................ ....... 48

DETERMINING WHETHER CONTRACTOR-SPECIFIED DETAILSFOR OPENING DESIGN ARE ACCEPTABLE ................................ ................................ ... 54

General ................................ ................................ ................................ ................................ .... 55

Reinforcement and Welding ................................ ................................ ................................ .... 55

Thermal Stress Relief ................................ ................................ ................................ .............. 56

Manholes, Nozzles, and Flush-Type Cleanout Fittings ................................ ........................... 57




DETERMINING WHETHER CONTRACTOR-SPECIFIEDDESIGN DETAILS FOR TANK ROOFS ARE ACCEPTABLE ................................ .......... 63

Cone Roofs ................................ ................................ ................................ .............................. 63

Supported Cone Roof ................................ ................................ ................................ . 63Self-Supporting Cone Roof ................................ ................................ ........................ 65

Self-Supporting Dome Roof ................................ ................................ ................................ .... 65

Internal Floating Roof ................................ ................................ ................................ ............. 66

External Floating Roofs ................................ ................................ ................................ ........... 66

Single-Deck Floating Roof ................................ ................................ ........................ 67Double-Deck Floating Roof ................................ ................................ ....................... 68Special Considerations for External Floating Roofs ................................ .................. 68

Saudi Aramco and API Design Requirements ................................ ................................ ........ 69

32-SAMSS-005 Requirements ................................ ................................ .................. 71API-650 Requirements ................................ ................................ ............................... 71

Sizing Inlet Diffusers ................................ ................................ ................................ ............... 76

DETERMINING WHETHER CONTRACTOR-SPECIFIEDDESIGN DETAILS FOR TANK BOTTOMS ARE ACCEPTABLE ................................ .... 78

Minimum Thickness ................................ ................................ ................................ ................ 78

Cone Up or Down ................................ ................................ ................................ ................... 78

Annular Ring ................................ ................................ ................................ ........................... 78

Water Withdrawal................................ ................................ ................................ .................... 81

Saudi Aramco and API Design Requirements ................................ ................................ ........ 81

SUMMARY ................................ ................................ ................................ ............................ 82

WORK AID 1: PROCEDURE FOR DETERMINING WHETHER CONTRACTOR-SPECIFIED SHELL THICKNESSES ARE CORRECT ........................... 83

Work Aid 1A: Procedures (One-Foot Method) and AdditionalInformation for Calculating the Required Shell Thicknessfor Atmospheric Storage Tanks ................................ ...................... 83

Work Aid 1B: Procedure (Variable-Design-Point Method) and AdditionalInformation for Calculating the Required Shell Thickness forAtmospheric Storage Tanks ................................ ............................ 89




Calculation of Bottom Shell Course Thickness ................................ ...................... 89Calculation of the Second Shell Course Thickness ................................ ................ 90Calculation of Third and Higher Shell Course Thicknesses ................................ ... 92Conclusion ................................ ................................ ................................ ................ 94

WORK AID 2: PROCEDURE FOR DETERMINING WHETHER CONTRACTOR-SPECIFIED WIND GIRDER REQUIREMENTS FOR OPEN-TOPAPI-650 TANKS ARE CORRECT ................................ ............................ 95

Top Wind Girder Evaluation ................................ ................................ ................... 95

Intermediate Wind Girder Evaluation ................................ ................................ .... 105

WORK AID 3: PROCEDURE FOR DETERMINING WHETHER CONTRACTOR-SPECIFIED DETAILS FOR OPENING DESIGN AREACCEPTABLE ................................ ................................ ......................... 111

Shell Manholes ................................ ................................ ................................ ...... 112

Shell Nozzle and Flange ................................ ................................ ........................ 112

Flush-Type Cleanout Fittings ................................ ................................ ................ 117

Flush-Type Shell Connections ................................ ................................ ............... 117

Roof Manholes................................ ................................ ................................ ....... 119

Roof Nozzles ................................ ................................ ................................ ......... 119

WORK AID 4: PROCEDURE FOR DETERMINING WHETHER CONTRACTOR-SPECIFIED DESIGN DETAILS FOR TANK ROOFS AREACCEPTABLE ................................ ................................ ......................... 120

General................................ ................................ ................................ ................... 120

Frangible Fixed Roof ................................ ................................ ............................. 120

General Fixed Roof Tanks ................................ ................................ ..................... 123

Supported Cone Roof Tank ................................ ................................ ................... 124

Self-Supporting Cone Roof ................................ ................................ ................... 125

Self-Supporting Dome Roof ................................ ................................ .................. 126

External Floating Roof ................................ ................................ ........................... 127

Internal Floating Roof ................................ ................................ ............................ 130

WORK AID 5: PROCEDURE FOR DETERMINING WHETHER CONTRACTOR-SPECIFIED DESIGN DETAILS FOR TANK BOTTOMS AREACCEPTABLE ................................ ................................ ......................... 131

GLOSSARY................................ ................................ ................................ .......................... 138



Saudi Aramco DeskTop Standards 1

DESIGN Factors THAT ARE Considered in the Mechanical Design of StorageTanksThis section discusses the primary factors that are considered in the mechanical design of storage tanks. Thesefactors are as follows:

• Metal temperature

• Pressure

• Specific gravity of the stored liquid

• Corrosion allowance

• Other loads

• Settlement

Specifically, this section discusses the effect that each of the above-listed design factors could have on tankreliability if the design factor is not properly considered in the mechanical design of the storage tank.

Metal TemperatureThe metal temperature of storage tank components is determined by the operating requirements of the storedliquid, and by the ambient temperature at the tank location. The operating requirements and operatingconditions of the stored liquid are determined by process engineers. The mechanical design of storage tankcomponents must consider both the highest and the lowest temperatures to which the tank can be exposed. Asdiscussed in MEX 203.02, the maximum operating temperature is the highest temperature that must beconsidered in the tank design, and the design metal temperature is the lowest temperature that must beconsidered in the tank design. Both of these temperatures are specified on either Saudi Aramco Drawing 2696or the Storage Tank Data Sheet (API-650 Appendix L).

The maximum operating temperature determines the allowable stress that is used for the mechanical design ofstorage tank components. The allowable stress of each specific material is constant for all temperatures up to93°C (200°F); however, the allowable stress of each material decreases for temperatures that are above 93°C(200°F). API-650 Appendix M contains additional design criteria that must be followed for tanks that havemaximum operating temperatures over 93°C (200°F). The tank could experience a permanent deformation or aductile fracture if the design requirements of API-650 Appendix M are not followed.

As discussed in MEX 203.02, the design metal temperature affects the fracture toughness characteristics of thematerial and has a significant influence on tank material selection. If the design metal temperature that isspecified is higher than it needs to be, based on the specific tank application, the tank may experience a brittlefracture. The design metal temperature is based on ambient conditions for most storage tanks. As discussed inMEX 203.02, SAES-D-100 specifies the appropriate design metal temperatures for Saudi Aramco storagetanks.




PressureThe internal pressure at which a storage tank will operate determines which API standard is to be used for themechanical design of the tank and its associated components. API-650, Welded Steel Tanks for Oil Storage, isthe design standard for tanks that operate at internal pressures approximating atmospheric pressure. API-650may also be used for tanks that operate at internal pressures up to 17 kPa (ga; 2.5 psig); however, additionaldesign requirements that are contained in API-650 Appendix F must be followed if the internal pressureexceeds atmospheric pressure. This course will only consider API-650 tanks that operate at atmosphericpressure.

API-620, Design and Construction of Large, Welded, Low-Pressure Tanks, is the design standard that is usedfor tanks that operate at pressures that exceed the pressure limits of API-650 and are less than 103 kPa (ga; 15psig). When tanks have internal design pressures over 103 kPa (ga; 15 psig), they are designed as pressurevessels in compliance with Section VIII of the ASME Code. Tanks that are designed in accordance with API-620, or Section VIII of the ASME Code, are not discussed in this course. Saudi Aramco has a small number ofAPI-620 tanks, and pressure vessels are discussed in MEX 202.

The internal operating pressure is determined by process engineers based on the operating requirements of thestored liquid. The internal operating pressure is specified on either Saudi Aramco Drawing 2696 or the StorageTank Data Sheet (API-650 Appendix L). If the operating pressure is specified incorrectly, the mechanicaldesign of the tank will not be correct. For example, if the actual internal pressure for a particular applicationexceeds 17 kPa (2.5 psig) but is specified to be less than that, API-650 would be used for the mechanical designof the tank instead of API-620. The use of API-650 for the mechanical design of the tank would be incorrect inthis case. Depending on the magnitude of the actual pressure, this specification error could result in shell androof sections that are too thin or nozzles that are not adequately reinforced for the actual applied pressure. Theconsequences of this error could be permanent deformations or a ductile fracture of tank components.

Specific Gravity of Stored LiquidThe specific gravity of the liquid that is being stored, (G), in conjunction with the depth of the liquid,determines the hydrostatic pressure of the liquid, as illustrated in Figure 1. The total hydrostatic pressure at agiven elevation in a tank must be considered in determining the required thickness of the tank shell. Thespecific gravity of the stored liquid must be specified on either Saudi Aramco Form 2696 or the Storage TankData Sheet (API-650 Appendix L).




Figure 1. Hydrostatic Pressure in a Storage Tank

Storage tanks must be designed for the specific gravity of water (1.0), because the tanks are filled with water fortesting purposes after they are constructed. If the specific gravity of the liquid that is to be stored exceeds 1.0(water), the tank must be designed for the higher specific gravity. Most Saudi Aramco storage tank applicationsstore hydrocarbons with specific gravities that are less than 1.0. As discussed later in this module, the requiredtank shell thicknesses are calculated separately for the hydrotest case and the design liquid case using differentallowable stresses for each case.




A later section of this module discusses the procedures that are used to determine the required thickness of thetank shell. The required shell thickness is directly proportional to the specific gravity of the stored liquid. Ifthe specific gravity is not correctly specified, the calculated shell thickness will be incorrect. In extreme cases,the shell can become permanently deformed if it is too thin, and a ductile fracture may occur.

It may be desirable for operational reasons to change the liquid that is being stored after the tank has been inservice for some period of time. In situations where the stored liquid is changed, the tank must be evaluatedbased on the specific gravity of the new liquid. MEX 203.08 discusses the evaluation of existing storage tanks.

Corrosion AllowanceAs discussed in MEX 203.02, the components of a storage tank may lose metal due to corrosion that is causedby the stored liquid. To compensate for this metal loss, a "corrosion allowance" (CA) may be added to themetal thickness that is required for strength, as illustrated in Figure 2. This "corrosion allowance" offsets theexpected deterioration of the tank components while they are in service. When needed, the corrosion allowanceis typically added to the calculated required thicknesses of the shell, internal components, and structuralmembers that may be used to support a fixed roof. A corrosion allowance is typically not added to the requiredthicknesses of the roof itself or the bottom.




Where:

t = Minimum required shell thickness

CA = Corrosion allowance

T = Total required shell thickness

Figure 2. Corrosion Allowance in Tank Shell




Corrosion will sometimes occur on the exterior of the tank shell or roof sections. However, this corrosion isnormally caused by factors such as deterioration of the external paint system or poor local drainage. Acorrosion allowance is not added to the metal thickness to account for external corrosion. Periodic maintenancemust be performed to ensure that external corrosion does not become a problem.

As discussed in MEX 203.02, SAES-D-100 specifies the corrosion allowance requirements for Saudi Aramcostorage tanks. In brief, a corrosion allowance is normally not specified for storage tanks that are in crude oilservice; however, a corrosion allowance may be specified by the proponent of the project for tanks that are incrude oil or hydrocarbon services based on actual experience with other tanks in similar services. Corrosionallowances are specified for the shell and internal structures of storage tanks that are in water service if there isno internal cathodic protection system. A corrosion allowance, when required, must be specified on eitherSaudi Aramco Drawing 2696 or the Storage Tank Data Sheet (API-650 Appendix L). MEX 203.02 contains atable that summarizes Saudi Aramco corrosion allowance requirements.

If a storage tank shell corrodes during operation, and a corrosion allowance was not properly specified, thestresses in the shell will increase and may eventually exceed the allowable stress. In extreme cases, the shellcan become permanently deformed (or hole through) if it becomes too thin, and a ductile fracture may occur.MEX 203.08 discusses the evaluation of existing storage tanks that have experienced corrosion duringoperation.

Other LoadsThe mechanical design of a storage tank must also consider loads other than pressure. These other loadsinclude wind and earthquake, loads that are imposed by connected piping systems (and other attachments) onnozzles, and rainwater accumulation on external floating roofs. Figure 3 illustrates the application of wind andearthquake loads on a storage tank.




Where:

Fw = Base shear force due to wind

M = Overturning moment due to wind or earthquake

Figure 3. Wind and Earthquake Loads




32-SAMSS-005 does not specify any requirements with respect to other loads. SAES-D-100 specifies thefollowing loading and analysis requirements for the design of storage tanks:

• All tanks must be designed for a wind velocity of 137 km/h (85 MPH) in accordance withANSI A58.1, Exposure C. Tanks must also be checked for stability against overturningmoments that are caused by wind pressure.

A tank shell may become out-of-round if it is not adequately designed for the expected windvelocity. Excessive shell out-of-roundness could cause binding of floating roofs and/or theformation of cracks at shell welds.

The overturning moment requirement is most relevant for relatively small diameter storagetanks, because the tank weight might not be sufficient to resist the wind load. Anchor boltsmight be needed to keep the tank from turning over.

• The minimum required thickness of the tank shell for the design liquid, excluding corrosionallowance, must be used to determine wind girder requirements for external floating rooftanks. One or more wind girders are used to prevent tank shell out-of-roundness that could becaused by wind loads. Wind girder requirements are discussed later in this module.

• Seismic loads are to be considered, and the applicable seismic zone (to be found in SAES-M-100) must be specified on either Saudi Aramco Drawing 2696 or the Storage Tank Data Sheet(API-650 Appendix L). A tank rupture can occur if the tank is not designed for the requiredseismic loads.

• Nozzles must be designed for the static liquid load and loads that are applied by connectedpiping.

Special design considerations are not required to account for the static liquid load as long asstandard API-650 design details are used for the nozzles. Loads that are applied by piping thatis connected to a tank nozzle cause additional local stresses in the nozzle neck, the adjacentshell, and the associated attachment welds. These additional stresses, if they are too high, cancause the formation of local weld cracks in the nozzle or shell. Therefore, it must beconfirmed that the loads that are applied by the pipe are not excessive.

• For nozzles with diameters greater than 150 mm (6 in.) that are located in the first shell courseof tanks greater than 20 m (65 ft.) in diameter, a stress analysis of the tank shell-to-pipingintersection must be performed to assure adequacy of the design. The analysis must besubmitted to the Consulting Services Department for review prior to fabrication. This analysisis done to ensure that the local stresses at the nozzle-to-shell junction are not excessive.




SettlementSettlement of the tank bottom or shell is caused by compression or movement of the soil under the tank or thetank foundation. Settlement can be the result of improper foundation design, unusual flooding or high tides, orfrom slowly flowing soil in marshy or swampy locations.

When the foundation settles evenly around the tank periphery, the resulting settlement causes only minorproblems. However, nonuniform settlement causes tank shell out-of-roundness. Excessive shell out-of-roundness can cause flat spots or buckles on the tank shell and roof of fixed roof tanks, and seal damage,binding, and/or excessive gaps between the roof and shell in floating roof tanks. Figure 4 illustrates differentialshell settlement.




Figure 4. Differential Shell Settlement

A tank bottom may also experience differential settlement of several different types, as illustrated in Figures 5through 7. Settlement, depending on its type and severity, can overstress the bottom plates or bottom-to-shelljunction, or buckle the bottom shell course. This overstress can cause the bottom plate welds to crack and leak.If the leaks are large enough, portions of the foundation may become unstable. This instability could lead tothe loss of local bottom support, further increase the stress in the bottom plates, and eventually result in asignificant tank failure.




Figure 5. Center-to-Edge Bottom Settlement




Figure 6. Bottom Edge Settlement




Figure 7. Bottom Settlement or Bulges




Soil samples are typically taken from the area where a tank is to be erected. This sampling is done to determinethe composition of the soil, and whether there are any significant variations in soil composition over the areaupon which the tank will rest. From this soil composition data, civil engineers are then able to determine themaximum amount of settlement that is expected during the life of the tank, and whether this settlement will berelatively uniform. If the civil engineers determine that the predicted settlement is excessive, a soilsimprovement program is undertaken in order to reduce the amount of predicted settlement, and itsnonuniformity, to acceptable values.

The predicted settlement values are not specified on either Saudi Aramco Drawing 2696 or the Storage TankData Sheet (API-650 Appendix L). The predicted settlement is not directly considered in the mechanical designof a new storage tank because it is assumed that the tank foundation has been stabilized to the extent that isnecessary to keep settlement within an acceptable limit.




DETERMINING WHETHER CONTRACTOR-SPECIFIED Shell thicknesses ARECORRECTThis section first reviews the requirements that are in API-650 for shell thicknesses of storage tanks. Thissection then explains the following methods that are used to calculate shell thicknesses:

• The one-foot method

• The variable-design-point method

The discussion of each calculation method is accompanied by a sample problem that illustrates the applicationof the method.

The normal job function of Saudi Aramco engineers is to review the drawings and calculations for storage tankshells that are provided by tank manufacturers in Contractor Design Packages. For example, the shell platethicknesses that are specified by the contractor must be reviewed to determine if they are acceptable. The SaudiAramco engineer is not responsible for making the shell thickness calculations himself for the initial tankdesign. Situations may arise during the review of Contractor Design Packages when the Saudi Aramcoengineer must do check-calculations himself in order to verify the contractor’s work. The Saudi Aramcoengineer must know how these calculations should be done to be able to verify that the contractor-specifiedthicknesses are acceptable, to confirm that the contractor’s calculations are done correctly, and to performcheck-calculations himself as required.

API-650 RequirementsFigure 8 shows the relevant paragraphs of API-650 that specify requirements for tank shell thickness.




Topic Paragraph

Shell Design - General

Allowable Stress

Calculation of Thickness by the One-Foot Method

Calculation of Thickness by the Variable-Design-Point Method

Calculation of Thickness by Elastic Analysis

3.6.1

3.6.2

3.6.3

3.6.4

3.6.5

Figure 8. API-650 Tank Shell Thickness Requirements

Work Aid 1 contains the specific procedures to be used to calculate shell thicknesses using either the one-footmethod or the variable-design-point method. Tank shell thicknesses will normally only be calculated by thetank manufacturer. However, the contractor may make preliminary estimates of the shell course thicknessesusing the one-foot method for tanks that are no more than 61 m (200 ft.) in diameter. The sections that followsummarize general shell thickness requirements that are contained in API-650 and elaborate on several of theprocedural items that are contained in Work Aid 1. API-650 requires that an elastic stress analysis be done fortanks that cannot be designed by either of the other two thickness calculation methods. Elastic stress analysisof storage tank shells is not discussed in this course.

API-650 contains several general requirements that relate to shell thickness. Several of these requirements arehighlighted as follows:

• The required thickness of a storage tank shell must be calculated for two separate cases:

- Case 1 considers the specific gravity of the stored liquid. The shell thickness that results fromthis case is the “design shell thickness,” td.




- Case 2 considers the specific gravity of the water (1.0) which will be used for the hydrostatictest of the tank. The shell thickness that results from this case is the “hydrostatic test shell thickness,”tt.

• td for a tank is computed on the following basis:

(1) The assumption that the storage tank is filled to its design capacity with a liquidhaving a specific gravity specified by the purchaser.

(2) The design allowable stress (discussed in MEX 203.02).

(3) Any corrosion allowance specified by the purchaser.

The shell thickness computed on this basis is for the normal, maximum design loading for thetank.

• tt for a tank is computed on the following basis:

(1) The assumption that the storage tank is filled to its design capacity with water.

(2) The hydrostatic test allowable stress (discussed in MEX 203.02).

(3) No corrosion allowance.

• The tank shell is to be checked for stability against buckling from the design wind velocity aspreviously cited from SAES-D-100, (i.e., 137 km/h [85 MPH]). Intermediate wind girders,increased shell-plate thickness, or both, are to be included in the design in order to stiffen theshell, if required for stability. The most common approach is to weld circumferential windgirders to the shell. The use of wind girders will be discussed later in this module.

• The tank manufacturer is to furnish the purchaser with a drawing with the following data foreach shell course:




- Required shell thicknesses for both the design condition (including corrosion allowance) andthe hydrostatic test condition.

- Nominal thickness used.

- Material specification.

- Allowable stresses.

• Isolated radial loads on the tank shell, such as those caused by heavy loads on platforms andelevated walkways between tanks, are to be distributed to avoid excessive local shell stresses.

One-Foot MethodThe one-foot method is based on limiting the approximate membrane stress to the allowable stress at a locationthat is 1 ft. above the bottom of the course being considered. The required shell thickness is then determinedbased on that stress. A distance of 1 ft. above the bottom of the course is assumed to be the location ofmaximum membrane stress.

A step-by-step procedure for the one-foot method of API-650 for calculation of the required shell thickness foratmospheric storage tanks is provided in Work Aid 1A. This is the most commonly used method for thecalculation of shell thicknesses because it is the simplest method. The one-foot method is not valid for tanksthat are over 61 m (200 ft.) in diameter.

The following briefly summarizes the overall calculation approach of the one-foot method and severaladditional points. Complete procedural details are contained in Work Aid 1A. Use Figure 9 as a reference.




Figure 9. Tank Shell Courses




• A tank shell is constructed of individual rows of plates. Each individual plate row is called a“course” (see Figure 9), and each course is of a specified height. The total height of the tank shell isthe sum of the individual course heights.

• The required thickness of each individual course is calculated separately. The requiredthickness considers the depth of liquid (H) measured from the bottom of the course to the maximumliquid level.

• The hydrostatic pressure that is imposed by the stored liquid is a maximum at the bottomcourse and a minimum at the top course. If the same plate material specification is used for all thecourses, the required plate thickness to resist the hydrostatic pressure decreases in going from thebottom course to the top course.

• Different plate material specifications may be used for different courses. For example, ahigher-strength material specification may be used for the bottom course than is used for the uppercourses. Use of a higher strength material for one or more lower courses minimizes the required platethicknesses for the courses that are subjected to the highest hydrostatic pressures.

• The required thickness is calculated for each course for both the design liquid and for water.The minimum acceptable course thickness based on hydrostatic pressure considerations is the higher ofthe two calculated thicknesses (i.e., td or tt).

• The actual course thickness that is used for each course must also be greater than minimumvalues that are specified in API-650, based on tank diameter. In addition, no course may be thickerthan the course that is under it.

• Sample Problem 1 illustrates the application of the one-foot method.




Sample Problem 1: Calculation of Shell Thickness Requirements Using the One-FootMethod

Figure 10 is an excerpt from a Contractor Design Package for an atmospheric storage tank. Calculate therequired shell thicknesses for this storage tank using the one-foot method.

Tank Size:

Diameter - 100 ft. Total Height - 40 ft.

Shell Courses: 5-courses, each 8 ft. high

Maximum Design Liquid Storage Height: 38 ft.

Liquid Specific Gravity: 0.85

Corrosion Allowance: 1/16 in.

Shell Plate Material:

A516 Grade 70 for bottom course

A516 Grade 60 for all other courses

Bottom Plate Material:

Annular Plate: A516 Grade 70

Rest of Bottom: A36

Roof Plate Material: A36

Roof Type: External Floating

Figure 10. Sample Problem 1 Data




Solution

Work Aid 1A is used to solve this problem.

td =2.6D H −1( )G

Sd+ CA

tt =2.6D H −1( )

St

• For the first (bottom) shell course (H = 38 ft.):

For the A516, Grade 70 steel, Sd = 25 300 psi, and St = 28 500 psi (Table 3-2 of API-650).

td =2.6 100( ) 38− 1( ) 0.85( )

25 300+ 0.0625

td = 0.386 in.

tt =2.6 100( ) 38−1( )

28 500tt = 0.338 in.

The design condition is the governing case for the first course; therefore, the shell must be at least0.386 in. thick. In practice, the next larger readily available, standard plate thickness will be orderedby the vendor for each course.




• For the second shell course (H = 30 ft.):

Note that the shell plate material has changed to A516 Grade 60 for this course and all

higher courses. For A516 Grade 60, Sd = 21 300 psi and St = 24 000 psi.

td =2.6 100( ) 30−1( ) 0.85( )

21300+ 0.0625

td = 0.363 in.

tt =2.6 100( ) 30−1( )

24 000tt = 0.314 in.

The design condition is the governing case for the second course; therefore, the shell must be at least0.363 in. thick.

• For the third shell course (H = 22 ft.):

td =2.6 100( ) 22− 1( ) 0.85( )

21300+ 0.0625

td = 0.28 in.

tt =2.6 100( ) 22−1( )

24 000tt = 0.23 in.

The design condition is the governing case for the third course. Therefore, the shell must be at least0.28 in. thick.

If the calculations were continued, they would show that the minimum permissible thickness is the governingfactor for the fourth course; therefore, the shell must be at least 0.25 in. thick.

Since the minimum permissible thickness determined the required thickness of the fourth course, the fifthcourse must be at least 0.25 in. thick also.




The required minimum shell thicknesses are summarized as follows:

• First course: 0.386 in.

• Second course: 0.363 in.

• Third course: 0.28 in.

• Fourth course: 0.25 in.

• Fifth course: 0.25 in.

If the Contractor Design Package for this tank included shell course thicknesses, you must confirm that thecontractor-specified thicknesses are at least these minimum values.

Variable-Design-Point MethodRecall from the previous discussion that the one-foot method cannot be used if the tank diameter exceeds 61 m(200 ft.). The variable-design-point method of API-650 is normally used to calculate the required shellthicknesses for these larger-diameter tanks.

A step-by-step procedure for calculation of the required shell thickness for atmospheric storage tanks by thevariable-design-point method is provided in Work Aid 1B. The paragraphs that follow describe the overallapproach.

The variable-design-point method calculates the required thickness of each shell course at an elevation that iscloser to the actual point of maximum stress than the one-foot method assumes. The variable-design-pointmethod may be used when the purchaser does not require use of the one-foot method, and when the followingequation based on tank geometry is satisfied:

SI Units English Units

0.268 DtH

≤ 26DtH

≤ 2

Where: D = Tank diameter, m (ft.)

t = Bottom shell course thickness, mm (in.)

H = Maximum design liquid level, m (ft.)

If this inequality is not satisfied for a particular tank geometry, the shell thicknesses must be calculated byelastic stress analysis.




The variable-design-point method is an iterative calculation procedure. The calculation is begun by picking athickness for the shell course, and by then performing a calculation using that thickness in order to calculate therequired thickness. Normally, the starting thickness for the first iteration is the thickness that is determined bythe one-foot method. The calculation results in a revised required thickness for the shell course. Thecalculation is then repeated using the result of the first calculation as the initial estimated thickness. Theprocedure continues until the starting and calculated thicknesses converge. Calculations are made for both thedesign and hydrotest cases, as in the one-foot method.

The variable-design-point method for calculation of the required tank shell thickness has the followingcharacteristics:

• It is a more sophisticated design calculation than the one-foot method and usually results in athinner tank shell. The reduction in shell thickness is more important for larger diameter tanksbecause it results in a greater reduction in the amount of material that is needed to fabricate theshell. The reduction in shell thickness then results in a greater cost saving for large diametertanks.

• It is much more time-consuming than the one-foot method if the calculations are donemanually. However, tank suppliers will typically use a computer program for thesecalculations.

• It is a valid method for calculation of tank shell thicknesses for tanks that are over 61 m (200ft.) in diameter that satisfy the previously stated equation based on tank geometry. Thismethod may also be used for smaller diameter tanks. However, it normally is not worthwhileto employ this method for small diameter tanks.

• It cannot be used when the one-foot method is specified by the Purchaser. Saudi Aramco doesnot specify the use of one calculation method over another.

Refer to Tables K-1 through K-3 in API-650 Appendix K. These tables summarize the results of shell thicknesscalculations based on the variable-design-point method for a variety of tank diameters and shell heights. Thesetables may be used to help make an initial assessment of shell thickness results that are provided in a ContractorDesign Package when the variable-design-point method is used. These tables should be used cautiously for thefollowing reasons:




• The tables are based on the use of just three specific allowable stresses (i.e., theresults are only accurate for the material specifications that correspond to these allowablestresses). Therefore, their results are not correct for any other allowable material stresses, andmust be adjusted.

• The tables summarize the results for only the hydrostatic test case calculations. Therefore, theacceptability of the thicknesses for the design case must still be verified.

• The tables assume that all of the shell courses are fabricated using the same materialspecification. Therefore, the thicknesses are not correct if more than one material is used inthe fabrication of the shell. It is common for the shell of a large-diameter tank to be fabricatedusing at least two material specifications.

• The tables assume that each shell course is 2.4 m (96 in.) high. Therefore, the thicknesses arenot correct if the courses above the first course are not of this height.

• The tables are based on specific tank diameters and shell heights. The thicknesses must beadjusted for other tank diameters and heights.

Because of the above limitations, Tables K-1 through K-3 should at most be used to make an initial estimate ofthe shell thicknesses that are required based on the variable-design-point calculation method. Any finaldecision regarding the acceptability of contractor-specified shell thicknesses should be based on the specifictank geometry and material specifications and must also consider the calculations for the design case.




Sample Problem 2: Calculation of Shell Thickness Requirements Using the Variable-Design-Point Method

Figure 11 is an excerpt from a Contractor Design Package for an atmospheric storage tank. Calculate therequired shell thicknesses for this tank using the variable-design-point method. Work Aid 1B is used to solvethis problem.

Tank Size:





Corrosion Allowance: 0.05 in.

Shell Plate Material:

A573 Grade 70 for all courses



Rest of Bottom: A36




Solution:

Since the tank exceeds 61 m (200 ft.) diameter, Work Aid 1B must be used.For the A573 Grade 70 steel, Sd = 28 000 psi, St = 30 000 psi.

tpd =2.6D H−1( )G

Sd

+ CA




tpd =2.6 280( ) 64 −1( ) 0.85( )

28 000( ) + 0.05

= 1.442 in.

tpt =2.6D H−1( )

St

tpt =2.6 280( ) 64 −1( )

30 000( )= 1.529 in.

First Shell Course:

t1d = 1.06 −0.463D

HHGSd

2.6HDG

Sd

+ CA

t1d = 1.06 −0.463 280( )

64( )64( ) 0.85( )28 000( )

2.6 64( ) 280( ) 0.85( )

28 000( )

+0.05

= 1.423 in.

t1t = 1.06 −0.463D

HHSt

2.6HD

St

t1t = 1.06 −0.463 280( )

64( )64( )

30 000( )

2.6 64( ) 280( )

30 000( )

= 1.501 in.

The required bottom shell course thickness is 1.501 in.




Hydrostatic Test CaseThe hydrostatic test case calculations will first be done for all remaining courses, then the design conditioncalculations will be done.

Second Shell Course:

h1

rt1=

12 8( )[ ]6 280( ) 1.501( )

= 1.912

Since this is between 1.375 and 2.625, t2 is calculated by the equation that follows (after calculating t2a).

t2 = t2a + t1 − t2a( ) 2.1−h1

1.25 rt1

Use the "upper course" procedure to first calculate t2a.

tpt2 =2.6 280( ) 64 − 8( ) −1[ ]

30 000( )= 1.335 in.= tu

tL = 1.501 in. (from the earlier calculations).

C =

tLtu

tLtu

−1

1+ tLtu

1.5

C =

1.501( )1.335( )

1.501( )1.335( ) − 1

1+1.501( )1.335( )

1.5

= 0.060

Calculate the maximum stress point, x.




x1 = 0.61 rtu + 3.84CH

x1 = 0.61 6 280( )[ ]1.335( ) + 3.84 0.060( ) 64 − 8( ) = 41.79 in.

x2 = 12CH

x2 = 12 0.060( ) 56( ) = 40.32 in.

x3 = 1.22 rtu

x3 = 1.22 6 280( ) 1.335( ) = 57.78 in.

x = 40.32 in.

tdx =2.6D H−

x12

St

t2a =2.6 280( ) 56( )−

40.32( )12

30 000= 1.277 in.

Since the calculated value of t2a differs significantly from the initial value assumed (1.335 in.), repeat theprocedure using tu = 1.277 in.

C =

1.501( )1.277( )

1.501( )1.277( )

−1

1+1.501( )1.277( )

1.5

= 0.084

x1 = 0.61 6 280( )[ ]1.277( ) + 3.84 0.084( ) 64 −8( ) = 46.24 in.

x2 = 12 0.084( ) 56( ) = 56.19 in.

x3 = 1.22 6 280( ) 1.277( ) = 56.51 in.

x = 46.24 in.

t2a =2.6 280( ) 56( )−

46.24( )12

30 000= 1.265 in.

This is much closer, but do one more iteration.




C =

1.501( )1.265( )

1.501( )1.265( )

−1

1+1.501( )1.265( )

1.5

= 0.089

x1 = 0.61 6 280( )[ ]1.265( ) + 3.84 0.089( ) 64− 8( ) = 47.18 in.

x2 = 12 0.089( ) 56( ) = 59.57 in.

x3 = 1.22 6 280( ) 1.265( ) = 56.24 in.

x = 47.18 in.

t2a =2.6 280( ) 56( )−

47.18( )12

30 000= 1.263 in.

Since 1.265 in. and 1.263 in. are very close, no further interations are required. Use the value of 1.263 in. fort2a.

Finish calculation of the second shell course thickness.

t2 = 1.263( )+ 1.501( ) − 1.263( )[ ] 2.1−12 8( )

1.25 6 280( ) 1.501( )

= 1.399 in.




Third Shell Course:

tpt3 =2.6 280( ) 64 −16( )− 1[ ]

30 000( ) = 1.141 in.

C =

1.399( )1.141( )

1.399( )1.141( ) −1

1+1.399( )1.141( )

1.5

= 0.106

x1 = 0.61 6 280( )[ ]1.141( ) + 3.84 0.106( ) 64 −16( ) = 46.25 in.

x2 = 12 0.106( ) 48( ) = 61.06 in.

x3 = 1.22 6 280( ) 1.141( ) = 53.41 in.

x = 46.25 in.

tt3 =2.6 280( ) 48( )−

46.25( )12

30 000( ) = 1.071 in.

Iterate again.

C =

1.399( )1.071( )

1.399( )1.071( ) −1

1+1.399( )1.071( )

1.5

= 0.140

x1 = 0.61 6 280( )[ ]1.071( ) + 3.84 0.140( ) 64− 16( ) = 51.68 in.

x2 = 12 0.140( ) 48( ) = 80.64 in.

x3 = 1.22 6 280( ) 1.071( ) = 51.75 in.

x = 51.68 in.

tt3 =2.6 280( ) 48( )−

51.68( )12

30 000( ) = 1.060 in.




Iterate one last time.

C =

1.399( )1.060( )

1.399( )1.060( )

−1

1+1.399( )1.060( )

1.5

= 0.146

x1 = 0.61 6 280( )[ ]1.060( ) + 3.84 0.146( ) 64− 16( ) = 52.65 in.

x2 = 12 0.146( ) 48( ) = 84.10 in.

x3 = 1.22 6 280( ) 1.060( ) = 51.48 in.

x = 51.48 in.

tt3 =2.6 280( ) 48( )−

51.48( )12

30 000( ) = 1.061 in.

Since 1.060 in. and 1.061 in. are very close, no further iterations are required. Use 1.061 in. for the thickness ofthe third shell course.

The minimum acceptable shell thickness for a 280 ft. diameter tank is 0.375 in. Therefore, calculation of thefourth course thickness is required. The calculations that are required for the fourth and higher courses use thesame procedure as was used for the third course. Therefore, only the final values for the intermediateparameters and course thicknesses are shown in the following table:




CourseNumber C x1 x2 x3 x tt

4 0.0864 36.96 41.47 47.39 36.96 0.896

5 0.126 36.46 48.50 41.96 36.46 0.703

6 0.162 32.93 46.73 35.96 32.93 0.516

7 0.237 28.94 45.5 28.77 28.77 0.330

(0.375)*

8 - - - - - 0.375*

* Minimum permitted thickness is 0.375 in. for a 280 ft. diameter tank.

Proceed with the calculations for the design condition.

Design CaseThe design case calculations will now be done. Again, only the final iteration is shown through the third shellcourse, and only the final values are shown for the higher courses.

Second Shell Course:

tL = 1.423 in.

Assume tu = 1.203 in.

Note that the corrosion allowance is first subtracted from the values of tL and tu that are used in the followingintermediate calculations, and then the corrosion allowance is added back in for the final thickness calculation.




C =

1.423 − 0.05( )1.203 − 0.05( )

1.423 − 0.05( )1.203 −0.05( )

− 1

1+1.423− 0.05( )1.203− 0.05( )

1.5

= 0.0902

x1 = 0.61 6 280( )[ ]1.203 −0.05( ) + 3.84 0.0866( ) 56( ) = 46.25 in.

x2 = 12 0.0902( ) 56( ) = 58.2 in.

x3 = 1.22 6 280( ) 1.203− 0.05( ) = 53.7 in.

x = 46.25 in.

t2a =2.6 280( ) 56 −

46.25( )12

0.85( )

28 000= 1.152 in.

Since (1.203 - 0.05) = 1.153 in. checks, no further iterations are needed.

t2 = t2a + t1 − t2a( ) 2.1−h1

1.25 rt1

+ CA

Note that the value for t1 that is used in this equation is the bottom shell course thickness that was calculated forthe design case minus corrosion allowance.

t2 = 1.152( )+ 1.423 − 0.05( )− 1.152( )[ ] 2.1−12 8( )

1.25 6 280( ) 1.423 − 0.05( )

+ 0.05

= 1.312 in.

Third Shell Course:

tL = 1.312

Assume tu = 1.02




C =

1.312− 0.05( )1.02 − 0.05( )

1.312 − 0.05( )1.02− 0.05( )

− 1

1+1.312− 0.05( )1.02 − 0.05( )

1.5

= 0.139

x1 = 0.61 6 280( )[ ]1.02 − 0.05( ) + 3.84 0.12( ) 48( ) = 50.18 in.

x2 = 12 0.12( ) 48( ) = 79.86 in.

x3 = 1.22 6 280( ) 1.02− 0.05( ) =~ 49.25 in.

x = 49.25 in.

td3 =2.6 280( ) 48( ) −

49.25( )12

0.85( )

28 000( ) + 0.05( ) = 1.02 in.

Since the calculated and assumed thicknesses match, no further interations are needed. Use 1.02 in. as thethickness for the design case of the third shell course.

The minimum acceptable shell thickness for a 280 ft. diameter tank is 0.375 in. Therefore, calculation of thefourth course shell thickness is required. The calculations that are required for the fourth and higher coursesuse the same procedure as was used for the third course. Therefore, only the final values for the intermediateparameters and course thicknesses are shown in the following table for courses four through eight:

Course

NumberC x1 x2 x3 x td

4 0.0875 36.075 42.011 45.262 36.075 0.868

5 0.126 35.53 48.41 40.09 35.53 0.692

6 0.157 31.70 45.11 34.53 31.70 0.522

7 0.237 28.30 45.44 27.52 27.52 0.353(0.375)*

8 - - - - - 0.375*

* Minimum permitted thickness is 0.375 in. for a 280 ft. diameter tank.




Figure 12 summarizes the results of both sets of calculations (i.e., hydrostatic test case and design case) and theresulting minimum thickness that is required for each shell course.

SHELLCOURSE

HYDROSTATICTEST

THICKNESS, in.DESIGN

THICKNESS, in.

MINIMUMREQUIRED

THICKNESS, in.

1 1.501 1.423 1.501

2 1.399 1.312 1.399

3 1.061 1.020 1.061

4 0.896 0.868 0.896

5 0.703 0.692 0.703

6 0.516 0.522 0.522

7 0.375 0.375 0.375

8 0.375 0.375 0.375

Figure 12. Required Shell Thicknesses

Note from this summary that the hydrostatic test condition governs some course thicknesses, and that thedesign condition governs other course thicknesses. No shell course has a thickness greater than the shell coursebeneath it. If the Contractor Design Package for this tank included shell course thicknesses, you must confirmthat the contractor-specified thicknesses are at least these minimum values.




DETERMINING WHETHER CONTRACTOR-SPECIFIED Wind GirderRequirements for Open-Top Tanks ARE CORRECTThis section discusses the wind girder requirements for open-top tanks, and demonstrates how wind girders aredesigned.

An open-top tank is essentially a vertical cylinder that is open at the top and closed at the bottom. As Figure 13illustrates, this cylinder can be forced out-of-round by wind pressure that acts against it, unless adequatestiffness against deformation is provided by the shell alone or by other means. If excessive, shell out-of-roundness could prevent free vertical travel of the floating roof, or could cause the formation of cracks in shellwelds.

Figure 13. Shell Out-of-Roundness Caused By Wind




Theoretically, there are two ways to provide adequate stiffness:

• The tank shell can be made sufficiently thick to provide all the needed stiffness, or

• Some additional method of stiffening the shell can be provided.

In most cases, it is not economical to make the shell thick enough to provide all of the necessary stiffness.Therefore, additional stiffness to resist shell deformation is provided by welding circumferential stiffening ringsaround the outside of the tank. These stiffening rings are referred to as wind girders. Figure 14 shows severaltypical configurations for wind girders, and Figure 15 illustrates the general placement of both a top and anintermediate wind girder on a tank shell.

Figure 14. Typical Wind Girders




Figure 15. Wind Girder Placement on Tank Shell

Pertinent Sections of API-650All API-650 requirements that are related to the top and intermediate wind girders are contained in Para. 3.9.Several of the specific paragraph references that are in that section and their general content are as follows:

• Para. 3.9.3 - Restrictions on Stiffening Rings. Basic minimum size requirements.

• Para. 3.9.4 - Stiffening Rings As Walkways. Size and location requirements if a stiffeningring is used as a walkway.

• Para. 3.9.5 - Supports for Stiffening Rings. Criteria for determination of the need and thesize of stiffening ring supports.




• Para. 3.9.6 - Top Wind Girder. Sizing requirements for a top wind girder.

• Para. 3.9.7 - Intermediate Wind Girder. Criteria for determination of the need, location, andsize of intermediate wind girders.

The sections that follow discuss several of the general wind girder requirements, along with the designcalculations that are required for top and intermediate wind girders. Participants are referred, for other windgirder requirements, to API-650.

General Wind Girder RequirementsAPI-650 requires that all open-top tanks be provided with stiffening rings (i.e., wind girders), that the stiffeningrings be located at or near the top of the top course, and that they preferably be located on the outside of thetank shell. Stiffening rings will typically be made of standard structural sections, or will be formed from platesections that are welded together and then welded to the shell.

The general approach to wind girder design consists of determining the following:

• The minimum required section modulus for the top wind girder.

• Whether a second, intermediate wind girder must also be provided at some lower elevation onthe shell.

• The location of the intermediate wind girder (if one is needed).

• The minimum required section modulus of the intermediate wind girder(if an intermediate wind girder is needed).

Wind girder design calculations will be discussed shortly.

API-650 contains additional general requirements for wind girders that cover the following topics:

• Minimum sizes are specified for angles that are used as stiffening ring components or as a topcurb angle. The minimum plate thickness that may be used for built-up stiffening rings is 6.35mm (1/4 in.). The minimum angle sizes that are specified will provide a basic level ofstructural rigidity to the tank. The minimum angle sizes are specified separately from the topand intermediate wind girder evaluations.




• Wind girders must have drain holes to remove trapped liquid. Trapped liquid could causeaccelerated corrosion of both the wind girder and the adjacent portion of the tank shell.

• Size, location, and design details are specified for wind girders that are regularly used aswalkways. These details ensure that personnel safety is considered in wind girder design.

• Wind girder support requirements are specified. These requirements include criteria fordesigns where support is needed, along with maximum spacing limits between wind girdersupports.

• Design details are specified for the region of a wind girder where the tank stairway passesthrough the wind girder. The opening through the wind girder that is required to permitstairway access locally weakens the wind girder. Stiffening requirements are specified toreinforce this area.

Top Wind Girder Design CalculationsAs previously stated, API-650 requires that all open-top tanks be provided with a stiffening ring that is locatedat or near the top of the tank. This stiffening ring is the top wind girder. The purpose of the top wind girder isto ensure that the top section of the tank shell is stiff enough to prevent deformation which may be caused bythe design wind velocity blowing across the entire tank shell.

The top wind girder must be sized to have a large enough section modulus to provide adequate shell stiffening.Top wind girder design calculations consist of determining the minimum required section modulus, and thenselecting a large enough wind girder section to provide this section modulus. API-650 contains an equation tocalculate the minimum required top wind girder section modulus. The required top wind girder sectionmodulus is based on the tank diameter and shell height. API-650 also contains a table which summarizes thesection module of various structural shapes and attachment configurations to the tank shell. API-650 permitsthe inclusion of a portion of the tank shell in the wind girder section modulus. This inclusion is accounted forin the section modulus table.




Saudi Aramco has simplified top wind girder selection by developing several standard wind girder designswhich meet API-650 requirements. These standard designs are contained on Standard Drawing AD-036211.

The easiest approach to take is to use a standard API-650 or standard Saudi Aramco top wind girder detail thathas at least the required section modulus. However, nonstandard wind girder details may be used in certainsituations. For example:

• The tank contractor may decide to use a nonstandard detail due to cost or scheduleconsiderations.

• The standard details include a portion of the tank shell in determining their section modulus,as permitted by API-650. However, the shell corrosion allowance is not considered in thesection modulus determination for the standard details. Saudi Aramco requires that shellcorrosion allowance be considered in the wind girder calculations. Therefore, if the tank has acorrosion allowance and a standard wind girder detail is specified, the section modulus mustbe calculated, based on the corroded shell thickness, to confirm that the standard wind girderis still acceptable.

Work Aid 2 contains a procedure for designing or evaluating the top wind girder, in accordance with SaudiAramco and API-650 requirements. Work Aid 2 also contains a procedure that may be used to calculate theactual wind girder section modulus, using its specified geometry.

Sample Problem 3: Calculation of Top Wind Girder Size

Calculate the top wind girder size for the atmospheric storage tank that was used in Sample Problem 1. Figure16 is an excerpt from the Contractor Design Package for this tank. The shell course thicknesses for Courses 1through 3 are slightly higher than the minimum values that were calculated in Sample Problem 1 because theseare now the as-ordered plate thicknesses. Work Aid 2 is used to solve this problem.




Tank Size:





Corrosion Allowance: 1/16 in.

Shell Plate Material/Thickness:

Course Material Thickness, in.

1 A516 Grade 70 7/16

2 A516 Grade 60 3/8

3 A516 Grade 60 5/16

4 A516 Grade 60 1/4

5 A516 Grade 60 1/4



Rest of Bottom: A36







Solution

ZT = 0.0001D2H2V

100

2

ZT = 0.0001 100( )2 40( ) 85100

2

ZT = 28.9 in.3

Detail 3 of Drawing AD-036211 with a dimension "A" of 20 in. should be considered first, based on the 100 ft.diameter of this tank. The section modulus of Detail 3 must now be checked even though it is a standard SaudiAramco design, because the shell corrosion allowance must be considered. Work Aid 2 contains the procedurefor calculating the wind girder section modulus.

The following summarizes the information that is needed for making the section modulus calculation for theselected wind girder:

Corroded Top Shell Course Thickness, Ts 0.1875 in.

Wind Girder Thickness, Tw 0.25 in.

Wind Girder Extension, A 20 in.

Wind Girder Height, Hw 6 in.

Wind Girder Lip, Lw 2.5 in.

The results of the intermediate calculations are summarized in Figure 17:




Section Area, A x Ax d I Ad2

1 1.172 0.094 0.110 11.116 0.003 144.82

2 5.000 10.188 50.94 1.022 166.667 5.22

3 1.500 20.063 30.094 -8.853 0.008 117.56

4 0.625 18.938 11.836 -7.728 0.326 37.33

∑ A ∑(Ax) ∑I ∑(Ad2)

Totals 8.297 92.979 167.003 304.93

Figure 17. Sample Problem 3 Intermediate Calculations

Determine the location of the centroid of the combined area including the wind girder itself and the portion ofthe shell that acts with the girder.

x = ∑(Ax)/∑A

x = 92.979/8.297 = 11.21 in.

Determine the moment of inertia of the combined area.

Io = ∑(Ad2) + ∑I

Io = 304.93 + 167.003 = 471.933 in.4

Determine the maximum distance to the outermost fiber of the combined area as the greater of c1 or c2.




c1 = (0.1875 + 20 - 11.21) = 8.978 in.

c2 = x = 11.21 in.

Therefore, c = 11.21 in.

Determine the section modulus of the combined area.

Z = Io/c

Z = 471.993/11.21 = 42.1 in.3

Since the actual section modulus of the combined area exceeds the required area modulus (i.e., 42.1 in.3 > 28.9in.3), the wind girder detail that was initially selected is acceptable.

Intermediate Wind Girder Design CalculationsSituations exist where just a top wind girder alone will not provide enough shell stiffness for a givencombination of tank height, tank diameter, and tank shell course thicknesses. Put in simple terms, the distancebetween the top wind girder and the tank bottom is too large, in these situations, to resist wind-induced shelldeformation. Installation of an intermediate wind girder at a location between the top wind girder and the tankbottom reduces the unstiffened length of the shell, and is required in order to prevent shell deformation in thesecases.

Intermediate wind girder design calculations in accordance with API-650 requirements consist of the followinggeneral steps:

• Determine if an intermediate wind girder is needed, based on design wind velocity, tankdiameter, and shell course thicknesses.

• Locate the intermediate wind girder.

• Calculate the minimum required section modulus of the intermediate wind girder and select astandard structural shape that provides this section modulus.




The ideal location of the intermediate wind girder is such that the portions of the tank shell between theintermediate wind girder and the top wind girder, and between the intermediate wind girder and the bottom ofthe tank, have approximately the same stiffnesses. It would be incorrect, however, to locate the intermediatewind girder at the mid-height between the top wind girder and the tank bottom. As we've seen from the earliershell thickness calculations, the tank shell thickness decreases in going from the bottom to the top course.Because the lower courses are thicker than the upper courses, the lower portion of the tank shell is inherentlystiffer than the upper portion of the tank shell. Therefore, if the intermediate wind girder was located at themid-height of the shell, the upper portion of the tank shell would not be stiffened enough.

The API-650 procedure for locating the intermediate wind girder considers the variation in shell coursethickness. As illustrated in Figure 18, the API-650 procedure mathematically converts the actual tank shellheight to a "transformed shell" height. As detailed in Work Aid 2, the shell transformation is done byaccounting for the actual individual course thicknesses. The transformed shell then has the same stiffnessthroughout its height. Locating the intermediate wind girder at the mid-height of the transformed shell resultsin equal shell stiffness both above and below the intermediate wind girder. The intermediate wind girder is thenlocated on the actual tank shell in the same course and in the same relative position within that course as it ison the transformed shell. Using this approach, the intermediate wind girder is located much higher than themid-height on the actual tank shell.




Figure 18. Transformed Shell and Intermediate Wind Girder

Work Aid 2 contains a procedure for determining the need for an intermediate wind girder, locating it, andsizing it, in accordance with Saudi Aramco and API-650 requirements.

Sample Problem 4: Need, Location, and Size of Intermediate Wind Girder

For the same tank as in Sample Problem 3, determine if an intermediate wind girder is needed. If anintermediate wind girder is needed, determine its required section modulus and locate it on the tank shell.Reference Figure 16 for the necessary tank data. Work Aid 2 is used to solve this problem.

H1 = 6(100t )100t

D

3 100V

2




H1 = 6 100 0.25 − 0.0625( )[ ] 100 0.25 − 0.0625( )100

3 10085

2

H1 = 12.6 ft.

The results of the transformed shell calculation are summarized in Figure 19.

Course WWtr = W

tuniform

tactual

5

1 88

1/ 4 −1/167 /16 −1/16

5

= 1.41

2 88

1/ 4 −1/163 / 8 −1/ 16

5

= 2.23

3 88

1/ 4 −1/165 /16 −1/16

5

= 3.90

4 88

1/ 4 − 1/161/ 4 − 1/16

5

= 8

5 88

1/ 4 − 1/161/ 4 − 1/16

5

= 88 Wtr −total 23.54 > 12.6 ft.

Figure 19. Sample Problem 4 Transformed Shell

Since the height of the transformed shell is greater than H1 (i.e., Wtr-total = 23.54 ft. > 12.6 ft.), an intermediatewind girder is required.




23.542

= 11.77 ft. ≤ 12.6 ft.

Since (Wtr-total/2) ≤ H1, only one intermediate wind girder is required.

Locate Intermediate Wind Girder - The intermediate wind girder should be located approximately 11.77ft. down from the top wind girder or up from the bottom of the tank on the transformed shell. The transposedwidths of the top two courses are the same as their actual widths. Therefore, the intermediate wind girdershould be located 11.77 ft. below the top wind girder because the widths of the top two courses are the same onboth the transformed shell and the actual shell (i.e., 16 ft. total width for both courses together). Since the topwind girder is located at or near the top of the tank, the intermediate wind girder is located approximately 3.77ft. below the top edge of the fourth course (i.e., 11.77 - 8 = 3.77 ft.). This location is more than 6 in. from thehorizontal joint between the courses and is therefore acceptable.

With the intermediate wind girder placed at 3.77 ft. below the top of the fourth shell course, the transposedheight of the shell between the intermediate wind girder and the top wind girder is 11.77 ft., and the transposedheight of the shell between the intermediate wind girder and the bottom of the tank is 11.77 ft. Therefore, nosegment of the transposed shell exceeds H1 (i.e., 12.6 ft.).

Size Intermediate Girder - The intermediate wind girder is sized by calculating its minimum requiredsection modulus.

ZI = 0. 0001D2H1V

100

2

ZI = 0. 0001 100( )2 11.77( ) 85100

2

= 8. 5 in.3

The corroded thickness of the fourth course (i.e., where the intermediate wind girder is located) is 3/16 in.Based on Table 3-22 of API-650, one angle with dimensions of6 x 4 x 3/8 in. that conforms to Detail C of Figure 3-18 of API-650 is acceptable because its section modulus is9.02 in.3 Selection of the required intermediate wind girder in this manner is illustrated in Figure 20, which isan excerpt from Figure 3-18 of API-650.




Shell Thickness (inches)

Member Size (inches) 3/16 1/4 5/16 3/8 7/16

One Angle: Figure 3-18, Detail c

2 1/2 x 2 1/2 x 1/42 1/2 x 2 1/2 x 5/164 x 3 x 1/44 x 3 x 5/165 x 3 x 5/165 x 3 1/2 x 5/165 x 3 1/2 x 3/8

1.681.983.504.145.536.137.02

1.792.133.734.455.966.607.61

1.872.233.894.666.256.928.03

1.932.324.004.826.477.168.33

2.002.404.104.956.647.358.58

6 x 4 x 3/8 9.02 10.56 11.15 11.59 11.93

Figure 20. Sample Problem 4 - Intermediate WindGirder Selection

To summarize, a top wind girder that matches Detail 3 with a dimension "A" of 20 in. is required. Anintermediate wind girder is also required. The intermediate wind girder must be located 11.77 ft. below the topwind girder, and it must have a section modulus of at least 8.5 in.3 A 6 x 4 x 3/8 in. angle provides thenecessary section modulus for the intermediate wind girder.




DETERMINING WHETHER CONTRACTOR-SPECIFIED DETAILS FOR openingDesign ARE ACCEPTABLEThe primary purposes that are served by openings into tank shells and roofs are as follows:

• Fill and empty the tank.

• Provide access to the tank interior.

• Permit tank cleanout.

• Provide connections for items such as instrumentation, mixers, heaters, and water drawoffs.

Fill and discharge nozzles are located as low on the tank shell as possible, consistent with maintainingacceptable spacing between adjacent welds. Locating these nozzles as low as possible maximizes utilization ofthe total tank volume.

Manholes are located in the tank roof and shell to provide access to the tank interior for inspection andmaintenance when the tank is taken out of service.

Accumulated dirt and sludge must be periodically cleaned out of the tank interior in order to maintainmaximum possible storage volume, to prevent localized preferential corrosion that could occur underaccumulated deposits, and to avoid possible floating roof damage that can be caused by uneven support whenthe roof is landed. Large-size cleanout fittings are typically installed in the tank shell and are designed to beflush with the tank bottom to facilitate cleanout.

Nozzle connections that are required for instrument connections, heaters, and mixers must meet the same designrequirements as are used for the filling and emptying nozzles.

This section discusses Saudi Aramco and API-650 requirements for shell openings and covers the followingtopics:

• General Requirements

• Reinforcement and Welding

• Thermal Stress Relief

• Manholes, Nozzles, and Flush-Type Cleanout Fittings




DETERMINING WHETHER CONTRACTOR-SPECIFIED DETAILS FOR TANKOPENING DESIGN ARE ACCEPTABLE, CONT'D

Work Aid 3 provides an overall procedure to use to determine whether contractor-specified design details foropenings meet Saudi Aramco and API requirements. The sections that follow elaborate on several of theserequirements and briefly discuss several other points.

GeneralSeveral general API-650 requirements that relate to openings are as follows:

• Attachments must be made by full-penetration welds except for insert-type reinforcement thatmeets API-650 details. Full-penetration welds develop the full strength through the thicknessof the attachment and are thus better able to resist any loads that are applied by connectedpiping.

• Connections and appurtenances that meet the requirements of API-620 are acceptable. Theopening design procedure that is contained in API-620 is much more detailed than theprocedure that is contained in API-650. Therefore, although acceptable, an API-620 nozzledesign approach is only used in special cases for an API-650 tank.

• Sheared or oxygen cut surfaces on manhole necks, nozzle necks, reinforcing plates, and shell-plate openings must be uniform and smooth, with rounded corners, except when fully coveredby attachment welds. These surface condition requirements and corner requirements minimizelocal stress concentrations that could act as crack-initiation points during operation.

Reinforcement and WeldingAPI-650 requires that the cross-sectional area of the reinforcement at openings must equal or exceed theproduct of the vertical diameter of the opening that is cut in the shell, and the nominal plate thickness.However, API-650 simplifies the nozzle design process by specifying standard nozzle sizes and locations,together with associated reinforcing pad sizes and thicknesses, nozzle neck thicknesses, and attachment weldsizes. These specified nozzle details will satisfy the reinforcement requirements for most nozzle installations.Typically, design calculations are only required for cases where high loads are applied to a nozzle (such as fromconnected piping), an API-650 Appendix F tank (designed for a small internal pressure in addition tohydrostatic head), or if nonstandard nozzle design details are required.




API-650 specifies minimum spacing requirements between welds that are made at connections, and other weldsin the tank. These weld spacing requirements are specified to ensure that weld shrinkage stresses at one welddo not result in unacceptably high residual stresses at another weld. Excessive residual stresses could reduceweld reliability by making the weld more prone to the formation of cracks upon application of the design loads.The weld-spacing requirements are based on the size and type of weld, the thickness of the shell plate, andwhether thermal stress relief has been done. Para. 3.7.3 in API-650 specifies weld-spacing requirements.

Thermal Stress ReliefAPI-650 contains specific stress relief requirements that are not related to weld spacing requirements atopenings. These postweld heat treatment (PWHT) requirements include minimum temperatures and hold times.API-650 requires PWHT in the cases that follow:

• All flush-type shell connections and flush-type cleanout fittings. The entire nozzle assemblymust be stress-relieved, and the stress relief must include the bottom reinforcing plate (or annular plate)and the flange-to-neck weld. This stress-relieved assembly is then welded into the shell and bottom.These flush-type connections and fittings are discussed further in the next section.

Flush-type nozzle assemblies have more reinforcement and welding than other nozzles that are ofcompatible size. These assemblies are also installed into the bottom-to-shell junction area of the tank.This area is critical for tank integrity and is already subject to a complicated combination of loads andstresses. Therefore, stress-relieving the assembly prior to welding it into the shell eliminates theintroduction of more stresses into this area.

• When the shell material is in Material Group I, II, III, or IIIA, all openings of 305 mm (12 in.)or larger in diameter in shell plates of 25.4 mm (1 in.) or more in thickness must be prefabricated intothe shell plate and the prefabricated assembly stress-relieved. Flange-to-neck welds and other nozzle-neck and manway-neck attachments may be excluded if specified conditions are met.




• When the shell material is Group IV, IVA, V, or VI, all openings that require reinforcement inshell plates more than 12.7 mm (1/2 in.) thick must be prefabricated into the shell plate andthe prefabricated assembly stress-relieved. Flange-to-neck welds and other nozzle-neck andmanway-neck attachments may be excluded if specified conditions are met.

Manholes, Nozzles, and Flush-Type Cleanout FittingsStandard-size shell and roof manholes are specified in API-650. Standard shell manholes range in size between508 mm (20 in.) and 900 mm (36 in.) in diameter. Standard-size roof manholes are typically 508 mm (20 in.)and 610 mm (24 in.) in diameter. Shell and roof manholes are illustrated in Figures 3-4A and 3-13 respectivelyin API-650.

API-650 specifies standard design details for three basic nozzle configurations: "regular-type," "low-type," and"flush-type." The low-type nozzle is located lower on the tank shell than the regular-type. Any reinforcementthat is required for a low-type nozzle will typically extend to the tank bottom in order to avoid weld spacingproblems with the bottom-to-shell junction weld, and to achieve a more uniform local stress distribution. Theflush-type nozzle is another type of shell connection that is used for filling and emptying large diameter tanks.In the flush-type nozzle, the bottom of the nozzle is even, or flush, with the tank bottom. This configurationmaximizes tank volume utilization. Shell nozzles are illustrated in Figures 3-5 and 3-11 in API-650. Figure 21illustrates regular- and low-type nozzles, and Figure 22 illustrates a flush-type nozzle.

In the flush-type cleanout fitting, the bottom of the nozzle is also even, or flush, with the tank bottom. Thisconfiguration simplifies personnel and equipment entry into the tank, and the removal of sludge and debriswhich have built up. A standard flush-type cleanout fitting is illustrated in Figure 3-9 in API-650.




Figure 21. Typical Regular- and Low-Type Nozzle Details




Figure 22. Typical Flush-Type Nozzle Details




32-SAMSS-005 requirements supplement those that are contained in API-650. Several of these requirementsare highlighted as follows:

• Reinforcing pads must be welded to the shell for all pipe supports, gauges, and sample lines.The addition of reinforcement spreads the applied local load over more of the tank shell, andreduces the probability of excessive local stresses.

• All attachments to the shell must be seal-welded all around. The continuous seal weldingprevents water or dirt from getting between the pad and shell and causing accelerated localcorrosion.

• Reinforcing pads and all other external pads must have rounded corners of at least 50 mm (2in.) radius. Rounded corners reduce the local stress concentration effects that could result inlocal crack initiation.

• Nozzle reinforcing pads and pads that cross shell seams must have a 6.35 mm (1/4 in.) tappedhole. This hole is used for a compressed air-soapsuds test of the fillet welds attaching the padto the shell.

Sample Problem 5: Opening Evaluation

The tank that was used in Sample Problems 1 and 3 requires a flush-type cleanout fitting. Design informationfor this tank is contained in Figure 16 that was used for Sample Problem 3. The cleanout fitting size wasspecified to be 24 in. high by 24 in. wide.

The contractor's proposal for this cleanout fitting specifies that the opening will be cut in a standard bottomcourse plate, and will be reinforced with a reinforcing plate. The cleanout fitting is located away from all shell-seam welds. The reinforcing plate that was proposed by the contractor is illustrated in Figure 23. Thecontractor also specified that the completed assembly will be given PWHT in accordance with API-650.Determine if the contractor's proposal for this cleanout fitting is acceptable. If the proposal is not acceptable,state why.




Figure 23. Reinforcing Plate for Sample Problem 5




Solution

Work Aid 3 is used to evaluate this proposal

The design is deficient in the following areas:

• The width at the base of the reinforcing plate must be 72 in. instead of60 in. This width is based on Table 3-11 of API-650. This width would widen the flat lengthat the top of the plate to 14 in.

• The thickness of the shell plate that contains the opening must be 1/2 in. instead of thestandard shell thickness for the bottom course of 7/16 in. This thickness is based on Table 3-13 of API-650.

• The thickness of the reinforcing plate must be 1/2 in. instead of the 1/4 in. that was specifiedby the contractor. This thickness is based on Table3-13 of API-650.




DETERMINING WHETHER CONTRACTOR-SPECIFIED Design details For TankRoofs ARE ACCEPTABLEThe subjects that are discussed in this section are as follows:

• Design considerations for tank roofs based on roof type

• Saudi Aramco and API design requirements

• Sizing inlet diffusers

API-650 provides rules and guidelines to achieve roof designs that are within specified allowable stress andload limits. API-650 also provides minimum requirements for the fabrication of tank roofs and their connectionto the tank shell. The various types of tank roofs that will be discussed were illustrated in MEX 203.01.

API-650 classifies roofs by their shape, operation, and support as follows:

• Cone roof

- Supported

- Self-supporting

• Self-supporting dome roof

• Internal floating roof

• External floating roof

Cone Roofs

Supported Cone RoofA supported cone roof gets most of its support either from rafters that are positioned on top of girders andcolumns, or from rafters that are positioned on top of trusses (either with or without columns). Rafters areoriented radially from the tank shell toward the center, and girders are oriented as circumferential chords aroundthe tank. Columns support the roof structural members from the tank floor. The number and size of the rafters,girders, and columns are based on the tank size and external loading. Figure 24 illustrates the relationshipamong the rafters, girders, and columns.




Figure 24. Rafters, Girders, and Columns in Cone Roof Tank




Some special design considerations for supported cone roofs are as follows:

• Lateral bracing may be required for rafters or other roof support members if the supportmembers are not adequately restrained by the roof or by the inherent stiffness of their design.

• Roof plates are welded to each other and to the tank top angle.

• Roof plates are not attached to the rafters in order not to detract from the design intent of thefrangible joint between the roof and shell. The frangible joint is discussed later in thismodule.

• When pipe is used for support columns, the pipe is sealed or is designed such that it may bedrained and vented. This design measure minimizes the likelihood that corrosion will occurinside the pipe columns.

• Any required roof corrosion allowance must be added to the minimum nominal roof platethickness. As previously noted, a corrosion allowance is normally not specified for the roofplate, but a corrosion allowance may be specified for roof support members.

Requirements are also specified for the slope of the roof, top angle size, and rafter spacing. These requirementsare provided in Work Aid 4.

Self-Supporting Cone RoofA self-supporting cone roof is formed to the approximate shape of a right circular cone, and is only supported atits periphery by the tank shell. Requirements are specified for the slope of the roof, roof thickness, and theminimum area of the roof-to-shell junction region. These requirements are provided in Work Aid 4.

Self-Supporting Dome RoofA self-supporting dome roof is formed approximately to a spherical shape, and is only supported at itsperiphery by the tank shell . API-650 specifies requirements for the roof radius of curvature, roof thickness,and the minimum area of the roof-to-shell junction region. These requirements are provided in Work Aid 4.




Internal Floating RoofInternal floating roofs are often installed inside existing fixed roof tanks for specific product storageapplications in order to minimize vapor losses and/or oxygen entry (e.g., MTBE service). An internal floatingroof has some of the same design considerations as an external floating roof. However, since an internalfloating roof is shielded by a fixed roof, an internal floating roof does not have to resist the same externalenvironmental conditions (e.g., rainfall).

Since internal floating roofs are not exposed to rainfall, they do not require drain systems and do not have to bedesigned for the weight of rainwater. Consequently, the internal floating roof may be made from materials thatwould be unsatisfactory for an external floating roof, such as aluminum, stainless steel, or plastic. Although aninternal floating roof that is made from these materials may be less rigid, it will still provide the requiredreliability. Also, internal floating roofs do not have to be designed for the 122 kg/m2 (25 lb./ft.2) roof live loadrequirement of an external roof. However, the internal floating roof must be able to support a reasonablepersonnel and equipment load without damaging the roof or causing product leakage onto the roof. In addition,the roof supports and attachments must be designed for a uniform live load of 61 kg/m2 (12.5 lb./ft.2) unlessthe roof has drains or other means to automatically prevent liquid accumulation.

API-650 Appendix H contains design requirements for internal floating roofs. These requirements focus on thefollowing general areas:

• Material specifications for the roof itself and the peripheral seal

• Flotation requirements

• Design for the joints between the roof components

• Minimum thickness for roof components based on the material that is used

• Peripheral seals

• Roof penetrations and supports

• Fabrication, erection, welding, inspection, and testing

External Floating RoofsExternal floating roofs may be either single-deck type or double-deck type and present special designchallenges. External floating roofs must be free to rise and lower as the liquid level inside the tank changes,must also resist the same environmental conditions as closed-top tanks, and must provide a reasonably tightseal at the junction between the roof and shell. Some of the primary design considerations for external floatingroofs are as follows:




• The roof must float properly under all design load conditions. The design load conditionsmust consider the density of the stored liquid, the roof weight, and the external loads that areapplied to the roof. Rainwater accumulation and its associated weight is a major external loadthat must be considered in the roof design.

• The roof must not bind as it rises and lowers. Binding could cause the roof to tilt. This tiltedcondition would allow stored liquid to get on top of the roof and cause an unbalanced load.An unbalanced load could damage or sink the roof.

• The liquid cannot push the roof off the top of the tank. There must be enough tank shellheight (freeboard) to permit the liquid level to rise to its maximum design level while keepingthe roof entirely contained within the shell.

• The design must permit the tank to be safely emptied and filled completely.

• Tank appurtenances cannot be damaged as the roof rises and lowers through its entiremovement range. There can be no internal projections that would hamper roof movement ordamage the peripheral seal.

• The roof drain system must work properly and not be damaged as the roof rises and lowersthrough its entire movement range. This system must also prevent stored liquid from gettingon top of the roof should the drain hose or pipe become damaged. A check valve is installedin the drain line and is located inside the roof sump in order to prevent the stored liquid fromgetting on top of the roof in case the drain line fails.

Single-Deck Floating Roof

Single-deck floating roofs are equipped with pontoons, as discussed in MEX 203.01. Thepontoons float on the stored liquid and support the floating roof deck plate. The mostcommon design uses a continuous pontoon structure at the roof periphery, with the roof deckplate welded to the pontoon structure. Other designs use both the peripheral pontoon and acentral pontoon. The pontoon system is designed as a compartmented structure. With thecompartmented-structure design, a local leak at a cracked weld or at a punctured plate will notresult in complete flooding of the pontoon, and the roof will still maintain some buoyancy.




Double-Deck Floating Roof

Double-deck floating roofs have both a top and bottom steel deck, as discussed in MEX203.01. Thus, the entire roof is really a pontoon structure. The area between the two decks isdivided into compartments. The compartment system is designed so that some of thecompartments can fail and the roof will still maintain some buoyancy. The double-deck typefloating roof will typically be used for larger diameter tanks over (90 m [300 ft.]) in order toprovide sufficient buoyancy and stability.

Special Considerations for External Floating RoofsAppendix C of API-650 contains specific design requirements that apply only to external floating roofs. Thesedesign requirements focus primarily on roof flotation and structural details. The material that follows presents ageneral discussion of these areas. Further details are provided in Work Aid 4. 32-SAMSS-005 also containsrequirements that supplement Appendix C of API-650.

Flotation - A floating roof must float on the liquid that is contained within the tank. The lift or buoyancy thatis provided by the liquid is affected by the liquid's density (specific gravity). The buoyancy that is provided bya liquid increases as its density increases. The tank could store a variety of liquids over its design life.Therefore, the roof is designed to provide flotation with a reasonably light liquid (Specific Gravity = 0.7) toensure that the roof will float on top of any liquid that the tank is likely to store.

The roof can suffer a minor mishap that would damage it and allow liquid to enter its pontoon compartments.The roof is designed so that a failure of any two adjacent pontoon compartments will not sink the roof. Theroof is also designed so that any penetrations through the roof will not permit the stored liquid to flow on top ofthe roof. For example, the support leg sleeves must be long enough so that liquid will not exit through themwhen the deck sags under the load due to accumulated rainfall.

External floating roofs are particularly susceptible to damage due to the accumulation of rainwater. If too muchwater accumulates on the roof, the roof can be damaged and sink under the weight of the water. Externalfloating roofs are equipped with sumps and drains to collect rainwater and remove it safely from the roof. Thedrain system is designed to operate automatically with the roof at any level. Care is taken in the design of thedrain system to ensure that the drain system is not damaged as the roof rises and lowers while the tank is beingfilled or emptied. The drain system is designed to drain the roof under the heaviest anticipated rainfalls for thetank's location.




The roof must be designed to float with a reasonable amount of rainwater accumulation on it. This provisionallows for a possible temporary failure of the drainage system, or a reasonable delay in the ability of personnelto drain the accumulated rainwater off of the roof. The usual design allowance for a pontoon-type roof is for arainfall of 254 mm (10 in.) in a 24 hour period. Double-deck type roofs may be designed for a lesser watervolume, if emergency drains are installed to ensure that the rainwater level is kept to a lower design value.

Structural Requirements - Floating roof tanks are equipped with rolling ladders that provide access from aplatform at the top of the shell to the top of the roof. These ladders must adjust automatically to changes in rooflevel and remain functional as the tank fills and empties.

Floating roofs are equipped with support legs that can be adjusted to two positions from the top of the roof.The low-leg position will hold the roof at a safe height above the bottom when the tank is empty to permit theentry of personnel. The high-leg position permits the roof to travel lower during operation to maximize tankvolume use, and to keep the roof from traveling below the filling and emptying nozzles, manways, and otherappurtenances that are located below the tank roof. Special attention is given to the points where the legs attachto the roof to ensure that the stresses induced in the roof when the legs are supporting the roof are properlydistributed, to avoid damaging the roof. Adequate protection must also be provided for the tank bottom toprevent it from being damaged by the legs with a full weight load. The tank bottom is typically protected bysteel pads that are welded to the bottom at the locations where the legs will rest when the roof is landed.

The tank and roof system is provided with a mechanism to keep the roof centered and prevented from rotation.If a roof were to move off-center, it could bind on the tank shell while the tank is being filled and emptied.This binding could damage the tank shell, roof, or roof seal, and cause the roof to sink. If a roof were to rotate,it could damage the roof access ladder, fire-fighting systems, roof seal, and other appurtenances. Also, whenthe roof lands, the roof support legs would not be aligned with their corresponding bottom reinforcing pads, andthe legs could damage the bottom.

Saudi Aramco and API Design Requirements32-SAMSS-005 supplements the roof design requirements that are contained in API-650, especially for externalfloating roofs. Work Aid 4 contains procedures that may be used to determine if contractor-specified roofrequirements are acceptable.

Figure 25 summarizes the locations within 32-SAMSS-005 and API-650 where information that is related totank roof design may be found.

Standard Topic Location

32-SAMSS-005 Seal welding of self-supported roofs

Gauging requirements

Top-angle attachment for self-supportingroofs

External Floating Roofs

Para. 3.10.2

Para. 3.10.2

Para. 3.10.7

Para. C.3




API-650 General requirements Para. 3.10.2

Frangible joint Para. 3.10.2.5.1

Allowable stresses Para. 3.10.3

Supported cone roofs Para. 3.10.4

Self-supporting cone roofs Para. 3.10.5

Self-supporting dome andumbrella roofs

Para. 3.10.6

Top-angle attachment for self-supportingroofs

Para. 3.10.7

External floating roofs Appendix C

Internal floating roofs Appendix H

Figure 25. Roof Design Requirements




32-SAMSS-005 RequirementsRoof requirements that are contained in 32-SAMSS-005 relate to the topics that are listed in Figure 25. Refer tothe referenced paragraphs in Figure 25 and Work Aid 4 for specific details.

API-650 Requirements

General - API-650 roof requirements relate to the topics that are listed in Figure 25. Refer to the referencedparagraphs in Figure 25 and Work Aid 4 for specific details.

Frangible Roof-to-Shell Joint - A fixed roof atmospheric storage tank may experience a higher thanexpected internal pressure during operation or emergency conditions. This high pressure might occur in spiteof all design and operational precautions. If this pressure is high enough, a tank failure could occur. If thefailure occurs at a shell seam that is below the liquid level, a significant and possibly complete loss of the tankcontents could take place. API-650 requires that one of the following two design precautions be taken:

• Tank emergency venting is provided and designed in accordance with API-2000, VentingAtmospheric and Low-Pressure Storage Tanks. The emergency vents are designed to relievethe excess pressure, and the tank is then adequate for any internal pressure that can occur.

• A "frangible joint" is provided at the roof-to-shell intersection.

Figure 26 shows typical attachment details between a cone roof and the top angle on the shell. When a highinternal pressure occurs, the pressure will tend to lift the roof up, which in turn pulls the top portion of the shellinward. The roof-to-shell intersection can be designed to be the "weak link" in the tank structure and fail beforeany other shell joint or the shell-to-bottom joint. This preferential failure is achieved by using a relatively smallsize fillet weld to attach the roof to the top shell angle, and by limiting the overall strength of the roof-to-shelljunction region so that it cannot resist the inward buckling load that is caused by a high internal pressure. Theroof-to-shell junction is called a "frangible joint" when it is designed as the weak link in the tank structure.




Figure 26. Cone Roof Attachment to Top Angle




When the frangible joint fails, a local region of the roof-to-shell junction will buckle inward and a portion of theroof-to-shell weld will open. This additional opening provides more venting capacity. Thus, the excesspressure is vented and failure of a lower tank seam is prevented. Since the failure occurs above the liquid level,the stored liquid will not be released in this situation. The damaged portion of the roof and shell will requirerepair, but this repair will typically not be a major job and a significant tank spill is prevented.

Work Aid 4 contains the required evaluation procedure for a frangible joint.

Sample Problem 6: Frangible Joint for an Atmospheric Storage Tank

You are reviewing the design of an atmospheric fixed cone roof storage tank. The contractor has said that theroof-to-shell junction is a frangible joint and no emergency venting capacity is required. Is he correct?

The roof-to-shell junction details are shown in Figure 27. Additional tank design information is as follows:

Shell and framing weight: 332 777 kg

Corrosion allowance: 1.5 mm




Figure 27. Sample Problem 6 - Frangible Joint Details




Solution

Work Aid 4 is used to solve this problem.

It can be immediately stated that the junction is not a frangible joint because the roof attachment fillet weldexceeds a 4.76 mm size. Therefore the fillet weld size must be reduced to 4.76 mm. However, the otherfrangible joint criteria must also be checked to see if any additional changes are needed.

The roof slope cannot exceed 1:6, or tan θ ≤ 0.167. Based on Figure 28, the roof slope is 8°.

tan 8° = 0.141

Therefore the roof slope is acceptable.

A ≤217 W

30 800 tanθ=

217 × 332 77730 800 tan8°

A ≤ 16 682 mm2 to be frangible

Rc = 48 800× 0.5 = 24 400 mm

R2 = Rc / sinθ = 24 400 / sin8° = 175 321mm

wh = 0.3 R2th = 0.3 175 321× 6.35 = 316.5 mm

However wh can only be a maximum of 305 mm

wc = 0.6 Rctc = 0.6 24 400× 8 = 265.1mm

A roof = thwh = 6.35× 305 = 1 937

Ashell = tcwc = 8 × 265.1 = 2 121

Aangle = 9.5 75 + 75 − 9.5( )[ ] = 1 335

Available area = 5 393 mm2

Since the available area is less than the maximum permitted area, the junction meets all the requirements for afrangible joint, except for the fillet weld size. The fillet weld must be reduced to 4.76 mm size in order for thisto be a frangible joint.




Sizing Inlet DiffusersLiquid that is being pumped into a tank can damage or sink a floating roof (external or internal), and possiblyforce liquid onto the roof if the liquid flow velocity is too high. The possibility of such damage is increased ifwater slugs are also present in the inlet flow. The damage is caused when the high velocity liquid stream hitsthe relatively thin roof plates and structural members. Therefore, it is common to install a diffuser on the inletpipe inside of the tank. Such an inlet diffuser reduces the velocity of the liquid that enters the tank and directsthe inlet flow streams away from the tank roof. Therefore, the diffusers prevent damage and possible liquid"bubble over" onto the floating roof.

Figure 28 illustrates a typical diffuser installation for a horizontal inlet pipe. Diffusers are designed using pipethat is the same size as the inlet nozzle, and that has slots cut in its bottom quadrant. These slots are sized sothat the velocity of the liquid entering the tank is approximately 1 m/sec (3 ft./sec). The diffuser is positionedon a radial line to the center of the tank, and it is supported from the tank bottom.




Figure 28. Inlet Diffuser




DETERMINING WHETHER CONTRACTOR-SPECIFIED Design DETAILS FORTANK BOTTOMS ARE ACCEPTABLEThis section discusses Saudi Aramco and API requirements for tank bottoms. It covers the following topics:

• Minimum thickness

• Cone up or down

• Annular ring

• Water withdrawal

• Saudi Aramco and API design requirements

For the tanks that are discussed in this course, the bottoms are continuously supported by the ring wall andfoundation pad. Therefore, stress in the bottom plates themselves is not a factor during initial tank design.Stress is a design consideration for bottom annular rings.

Minimum ThicknessWork Aid 5 provides the minimum thickness and other requirements for storage tank bottoms. Since stress isnot a major factor in the design of a storage tank bottom, minimum thickness requirements are primarily toensure that there is adequate allowance for bottom corrosion. The minimum thickness requirements for annularrings (discussed below) are based on both stress and corrosion considerations.

Cone Up or DownTank bottoms may be designed as either cone up or cone down. A cone-up configuration slopes up from thetank periphery to the center of the tank. A cone-down configuration slopes down from the periphery to thecenter of the tank. For flat-bottomed tanks, SAES-D-100 requires that a coned-down bottom with a 1:120radial slope be used for tanks that are in services where water draw-off is required, unless otherwise stated inthe Tank Data Sheet. Otherwise, 32-SAMSS-005 requires a cone-up bottom with a 1:120 radial slope. A cone-up bottom permits less foundation grading and provides for future foundation settlement, which will be greaterin the center of the tank.

Annular RingThe outer ring of bottom plates on which the tank shell rests is called an annular ring when the bottom platesare joined by butt-welded construction. Otherwise, this region of the bottom is called a sketch plate if normallap-welded bottom construction is used in this region. Figure 29 illustrates lap-welded bottom plates under thetank shell. Work Aid 5 provides the minimum requirements for use of an annular ring.




Figure 29. Lap-Welded Bottom Plate Under Shell

A butt-welded annular ring is typically required when the stress in the bottom shell course is relatively high,which is indicative of a high bending moment in the bottom-to-shell junction region. The butt-welded annularring provides a more reliable construction to resist a high bending moment than lap-welded sketch plates. API-650 also specifies the minimum required width and thickness of the annular ring. As the loads in this regionincrease, the minimum values for both of these dimensions also increase in order to keep the local stresses inthe bottom-to-shell junction region within acceptable limits.

SAES-D-100 requires that the predicted tank settlement be reviewed by the Consulting Services Department todetermine if additional requirements for tank bottom design or soil improvement are needed. High predictedtank settlement could be another reason to use an annular ring, or to use an annular ring that is wider or thickerthan API-650 would require.

32-SAMSS-005 requires that a 12 mm (1/2 in.) thick, asphalt-impregnated board be installed on top of thefoundation ring wall under the annular ring, as illustrated in Figure 30. This board helps to minimize thelikelihood of corrosion on the underside of the annular ring. Corrosion of the annular ring would reduce itseffectiveness. The board also helps accommodate foundation imperfections, uneven weld surfaces, and backingstrips.




Figure 30. Asphalt-Impregnated Board Under Annular Ring




Water WithdrawalWater that may be contained in the liquid being stored tends to settle out to the bottom of the tank. The slopeof the tank bottom causes the water to flow to one general location in the tank, where it collects in a sump andcan be periodically withdrawn. Work Aid 5 contains sump and water drawoff pipe requirements foratmospheric storage tanks.

Saudi Aramco and API Design RequirementsFigure 31 summarizes the primary topics that are related to tank bottom design, and the locations within 32-SAMSS-005 and API-650 where information that is related to these topics may be found.

Standard Topic Location

32-SAMSS-005 Bottom plates

Annular bottom plates

Para. 3.4

Para. 3.5

API-650 Bottom plates

Annular bottom plates

Para. 3.4

Para. 3.5

Figure 31. Bottom Design Requirements




Work Aid 1: PROCEDURE FOR DETERMINING WHETHE R CONTRACTOR-SPECIFIED shell thicknesses ARE CORRECT

This Work Aid is to be used with the copy of API-650 that is in Course Handout 1 and with the copy of 32-SAMSS-005 that is in Course Handout 2. Work Aid 1A is used when the One-Foot Method is used todetermine the required tank shell thicknesses. Work Aid 1B is used when the Variable-Design-Point Method isused to determine the required tank shell thicknesses. Shell course thickness information that is contained inthe Contractor Design Package must be verified using the appropriate procedure.

Work Aid 1A: Procedures (One-Foot Method) and Additional Information for Calculatingthe Required Shell Thickness for Atmospheric Storage Tanks

Note:

This method is not to be used for tanks that are over 61 m (200 ft.) in diameter. Refer to Work Aid 1B forlarger diameter tanks. This procedure must be modified for application to API-650 Appendix F (small internalpressure) and Appendix M (elevated temperature) storage tanks. Appendix F and Appendix M tank types arenot discussed in this course.

1. Using the following formula, calculate the required shell thickness for the design case:


td =4.9D H − 0.3( )G

Sd+CA td =

2.6D H −1( )GSd

+ CA

Where: td = Design shell thickness, mm (in.)

D = Nominal tank diameter, m (ft.)

H = Design liquid level, m (ft.). The design liquid level is the height from the bottom ofthe course under consideration to the top of the shell, including the top angle, ifany, or to the bottom of any overflow that limits the tank filling height, or to anyother level specified by the Purchaser, restricted by an internal floating roof, orcontrolled to allow for seismic wave action.




G = Design specific gravity of the liquid to be stored

Sd = Allowable stress for the design case, MPa (psi). To be obtained from Table 3-2 of API-650 for the specified course shell plate material. An excerpt fromTable 3-2 is provided in Figure 34.

CA = Corrosion allowance, mm (in.). To be obtained from the Tank Data Sheet orfrom other provided information.

A 283A 285A 131A 36A 131

A 442A 442A 573A 573A 573

A 516A 516A 516A 516

CC

A, B, CS-

EH 36

5560586570

55606570

ASTM30,00030,00034,00036,00051,000

30,00032,00032,00035,00042,000

30,00032,00035,00038,000

Specifications55,00055,00058,00058,00071,000a

55,00060,00058,00065,00070,000a

55,00060,00065,00070,000

20,00020,00022,70023,20028,400

20,00021,30021,30023,30028,000

20,00021,30023,30025,300

22,50022,50024,90024,90030,400

22,50024,00024,00026,30030,000

22,50024,00026,30028,500

Figure 34. API-650 Table 3-2 (Excerpt)

PlateSpecification Grade

Minimum YieldStrength

Minimum TensileStrength

Product DesignStress Sd

HydrostaticTest Stress St




2. Using the following formula, calculate the required shell thickness for the hydrostatic

test case:


tt =4.9D H − 0.3( )

Sttt =

2.6D H−1( )St

Where: tt = Hydrostatic shell thickness, mm (in.)

St = Allowable stress for the hydrostatic test case, MPa (psi). Obtainfrom Table 3-2 of API-650 for the specified course plate material.

D and H = As defined in Step 1.

3. The required minimum shell thickness for the course is the greatest of the following:

• The design shell thickness calculated in Step 1.

• The hydrostatic test shell thickness calculated in Step 2.

• The minimum shell thickness value based on the tank diameter given in Figure 35.

NOMINAL TANK DIAMETER MINIMUM PLATE THICKNESS

m ft. mm in.

<15.2 <50 6.35* 1/4*

15.2 to <36.6 50 to <120 6.35 1/4

36.6 to 61 120 to 200 7.94 5/16

>61 >200 9.53 3/8

* Minimum nominal thickness of 6.35 mm (1/4 in.) required by 32-SAMSS-005.

Figure 35. Minimum Tank Shell Plate Thickness

4. Repeat Steps 1 through 3 for each shell course.

5. Verify that no shell course is thicker than the one beneath it.




Work Aid 1B: Procedure (Variable-Design-Point Method) and Additional Information forCalculating the Required Shell Thickness for Atmospheric Storage Tanks

Note:

This method may only be used for tanks that satisfy the following equation:


0.268 DtH

≤ 26DtH

≤ 2

Where: D = Tank diameter, m (ft.)

t = Bottom course shell thickness, mm (in.)


Note:

Since the One-Foot Method of Work Aid 1A cannot be used for tanks that are larger than 60 m (200 ft.) indiameter, the variable-design-point method is the most common design approach that is used for large diametertanks. A detailed elastic analysis must be used for tanks that do not satisfy the above equation or for bandedtanks.

Note:

Figure 36 provides an overall flow chart of the variable-design-point method.




Figure 36. Variable-Design-Point Method Flow Chart




1. Use Work Aid 1A to calculate a preliminary thickness for the bottom shell course for the

design condition, tpd, and for the hydrostatic test condition, tpt, using the One-Foot

Method.

Note:

Throughout the rest of this procedure, calculations for both the design case and hydrostatic test case are carriedout independently. Do the calculations for either the design case or for the hydrostatic test case for all shellcourses. Then do the calculations for the other case for all shell courses. In addition, all subsequent equationsare based only on English units of measure because several contain constants that are based on English units.

Calculation of Bottom Shell Course Thickness

2. Use the following formula to calculate the bottom course thickness for the design case:

t1d = 1.06 −0.463D

HHGSd

2.6HDG

Sd

+ CA

Where: t1d = Minimum thickness of the bottom shell course for the design condition, in.

D = Tank diameter, ft.

H = Design liquid level, ft.

G = Design specific gravity of the liquid to be stored.

Sd = Allowable stress for the design condition, psi. Obtain from Table 3-2 of API-650(Reference Figure 34) for the specified shell course plate material.

CA = Corrosion allowance, in. Obtain from Tank Data Sheet or other providedinformation.




3. Use the following formula to calculate the bottom course thickness for the hydrostatic

test case:

t1t = 1.06 −0.463D

HHSt

2.6HD

S t

Where: t1t = Minimum thickness of the bottom shell course for the hydrostatic test case, in.

St= Allowable stress for the hydrostatic test case, psi. Obtain from Table 3-2 of API-650 (Reference Figure 34) for the specified course plate material.

H and D = As defined in Step 2.

4. Identify the greater of the shell thicknesses determined in Step 2 or Step 3. The required

bottom shell course thickness is the lesser of this value (i.e., the larger of Step 2 or Step

3) or the shell thickness that was determined by the One-Foot Method in Step 1.

Calculation of the Second Shell Course Thickness

5. Calculate the following ratio.

h1rt1

Where: h1 = Height of the bottom shell course, in.

r = Nominal tank radius, in.

t1 = For the design case, actual bottom shell course thickness minus any corrosionallowance, in.

= For the hydrostatic test case, actual bottom shell course thickness, in.




6. The procedure that is used to determine the required thickness for the second shell

course depends on the value of the ratio that was calculated in Step 5. Based on the

calculated ratio, select the appropriate choice from the selections that follow:

• If Ratio ≤ 1.375:Required second shell course thickness is the same as the thickness required for the first shellcourse. Proceed with the third shell course thickness calculation at Step 7.

• If Ratio ≥ 2.625:Required thickness of the second shell course is calculated using the same procedure as for thethird and higher shell courses. Proceed with the second shell course calculation at Step 7.

• 1.375 < Ratio < 2.625:Required second shell course thickness is calculated using the equation that follows:

t2 = t2a + t1 − t2 a( ) 2.1−h1

1.25 rt1

+ CA

Where: t2 = Minimum required thickness of the second shell course for the case beingconsidered, in.

t2a = Minimum thickness for the second shell course calculated using the method thatapplies to higher shell courses, in. See Step 7 to calculate this value. Then returnhere to complete the calculation of the second shell course thickness. For thedesign case, subtract the corrosion allowance first before entering values into thisequation.

t1, h1 and r = As previously defined.




Calculation of Third and Higher Shell Course Thicknesses

7. Use Work Aid 1A to calculate a preliminary thickness for the case being considered

(i.e., design or test) using the One-Foot Method.

8. Use the following formulas to calculate the three possible values (i.e., x1, x2, and x3) for

the point of maximum stress. The smallest of x1, x2, or x3 is the value for x that is used

in Step 9:

C =

tLtu

tLtu

−1

1+tLtu

1.5

x1 = 0.61 rtu + 3.84CH

x2 = 12CH

x3 = 1.22 rtu

Where: C = Constant used in the following expressions.

tL = Thickness of the course immediately below the course being considered, in. Forthe design case, subtract the corrosion allowance first.

tu = Thickness of the shell course being considered at the girth joint, in. Use thevalue that was calculated in Step 7 as the initial value for tu. For the design case,subtract the corrosion allowance first.

x1, x2, x3 = Three possible points of maximum stress in the shell course being considered,measured as the distance from the bottom of the shell course, in.

r and H = As previously defined.




9. Calculate the required shell course thickness for either the design or hydrostatic test

cases (depending on which case is being considered) using the equations that follow:

• Design case calculation:

tdx =2.6D H −

x12

G

Sd+ CA

Where: tdx = Minimum thickness for the shell course for the design case, in.

x = Smallest of the three values for the possible point of maximum stress ascalculated in Step 8.

D, H, G, Sd and CA = As previously defined.

• Hydrostatic test case calculation:

ttx =2.6D H − x

12

St

Where: ttx = Minimum required thickness for the shell course for the hydrostatic test case, in.

D, H, x, and St = As previously defined.

10. Use the value that was calculated for tx in Step 9 as the value for tu. Repeat the

procedure from Step 8 to this Step until there is little difference between successive

calculated values of tx. Two additional times through the calculations are usually

sufficient.




Work Aid 1B: Procedure (Variable-Design-Point Method) and AdditionalInformation for Calculating the Required Shell Thickness forAtmospheric Storage Tanks, cont'd

11. Once the thicknesses for each shell course for one condition (design or hydrostatic test)

have been obtained, repeat Steps 2 through 11 for the other condition. When the

thicknesses for each shell course for both conditions have been obtained, proceed to

Step 12.

Conclusion

12. For each shell course, the minimum required thickness is the greater of the minimum

required thickness for the design condition, or the minimum required thickness for the

hydrostatic test condition, and in no case less than the thickness specified in Figure 35.

13. Verify that no shell course is thicker than the shell course beneath it.




Work Aid 2: PROCEDURE FOR DETERMINING WHETHER CONTRACTOR-SPECIFIED wind girder REQUIREMENTS for open-topAPI-650 tanks ARE CORRECT

This Work Aid is to be used with the copy of API-650 that is in Course Handout 1, and Saudi Aramco DrawingAD-036211 that is in Course Handout 3. All needed tank design information should be obtained from theContractor Design Package.

Top Wind Girder Evaluation

1. Use the following formula to calculate the minimum required section modulus of the top

wind girder.


ZT = 0.0578 D2H2V

161

2ZT = 0.0001D2H2

V100

2

Where: ZT = Required minimum top wind girder section modulus, cm3 (in.3)

D= Nominal tank diameter, m (ft.)

H2 = Height of tank shell, including any freeboard that is provided above the maximumfilling height, m (ft.)

V = Design wind speed, mph Typically, this is 137 km/h (85 MPH) for Saudi Aramcoapplications.

2. Determine the appropriate standard top wind girder design details and dimensions to be

used for the size of tank from Figure 37 and Figure 38. Figure 38 is an excerpt from

Saudi Aramco Drawing AD-036211 that is in Course Handout 3 for reference. Use the

class reference copy of drawing AD-036211 for additional information that is not shown

in Figure 38.




Figure 38. Saudi Aramco Drawing AD-036211 (Excerpt)

3. If the contractor has not used a standard Saudi Aramco wind girder, refer to Figure 3-18

of API-650 to determine which API-650 standard design was used. Then refer to Table

3-22 of API-650 to determine the section modulus. An excerpt from Table 3-22 is

contained in Figure 39. Note that the corroded shell thickness is to be used to determine

the section modulus from this table.




Shell Thickness (inches)

Member Size (inches) 3/16 1/4 5/16 3/8 7/16

Curb Angle: Figure 3-18, Detail b

2 1/2 x 2 1/2 x 1/42 1/2 x 2 1/2 x 5/163 x 3 x 1/43 x 3 x 3/84 x 4 x 1/44 x 4 x 3/8

1.611.892.322.783.644.17

1.722.042.483.354.415.82

------------------

------------------

------------------

One Angle: Figure 3-18, Detail c

2 1/2 x 2 1/2 x 1/42 1/2 x 2 1/2 x 5/164 x 3 x 1/44 x 3 x 5/165 x 3 x 5/165 x 3 1/2 x 5/165 x 3 1/2 x 3/86 x 4 x 3/8

1.681.983.504.145.536.137.029.02

1.792.133.734.455.966.607.61

10.56

1.872.233.894.666.256.928.03

11.15

1.932.324.004.826.477.168.33

11.59

2.002.404.104.956.647.358.58

11.93

Formed Plate: Figure 3-18, Detail e

b = 10b = 12b = 14b = 16b = 18b = 20b = 22b = 24b = 26b = 28b = 30b = 32b = 34b = 36b = 38b = 40

------------------------------------------------

23.2929.2735.4942.0648.9756.2163.8071.7279.9988.5897.52

106.78116.39126.33136.60147.21

24.6331.0737.8845.0752.6260.5268.7877.3986.3595.66

105.31115.30125.64136.32147.35158.71

25.6132.3639.5347.1055.0763.4372.1881.3090.79

100.65110.88121.47132.42143.73155.40167.42

26.3433.3340.7848.6756.9965.7374.8984.4594.41

104.77115.52126.66138.17150.07162.34174.99

Figure 39. Section Moduli for Several Standard Wind GirdersFrom API-650




4. Verify that the section modulus for the selected standard API-650 wind girder is equal to

or greater than the required modulus that was calculated in Step 1. If it is not, a stronger

standard design should be used, or a "special" wind girder should be designed. This

latter approach should rarely be required.

5. In some situations, it may be necessary to verify or calculate the section modulus of the

wind girder that is actually used. This necessity may occur if a nonstandard design is

used, or to verify that a standard design is acceptable after shell corrosion allowance is

considered. The following procedure and equations are provided as an example of how

the section modulus is calculated for a wind girder geometry that looks like Detail 3 of

Saudi Aramco Drawing AD-036211. A similar approach is used for other wind girder

geometries. Refer to Figure 40 for the wind girder reference geometry.




Figure 40. Wind Girder Reference Geometry

Note

• Numbers indicate the separate areas to be combined to obtain section modulus.

• Effective length of shell included in calculation.




Where:

Ts =Nominal top shell course thickness minus corrosion allowance, cm (in.)

Tw =Wind girder thickness, cm (in.)

A =Wind girder width, cm (in.)

Hw =Wind girder height, cm (in.)

Lw =Wind girder lip, cm (in.)

a. Calculate the area of each section:

A1 = (32Ts + Tw) Ts

A2 = ATw

A3 = HwTw

A4 = LwTw

b. Calculate the distance from the tank inside diameter to the centroid of each

individual area.

x1 = 0.5 Ts

x2 = (Ts + 0.5 A)

x3 = (Ts + A - 0.5 Tw)

x4 = (Ts + A - 0.5 Lw)




c. Use the results of Steps 5a and 5b and Figure 41 to locate the centroid of the

combined area, x , cm (in.).




Section A

(Step 5a)

x

(Step 5b)

Ax

1

2

3

4

∑∑A ∑∑Ax

x =Ax∑A∑

Figure 41. Centroid of Combined Area

d. For each section, calculate the distance between the centroid of the combined

area and the centroid of the individual section, d, cm (in.).

e. Calculate the moment of inertia of each area, I, about its centroid, cm4 (in.4).




I1 =32Ts + Tw( )Ts

3

12

I2 = TwA3

12

I3 =HwTw

3

12

I4 =TwLw

3

12




f. Use the results of Steps 5a, 5d, and 5e and Figure 42 to calculate the moment

of inertia of the combined areas, Io, cm4 (in.4).

Section A

(Step 5a)

d

(Step 5d)

Ad2 I

(Step 5e)

1

2

3

4

∑∑Ad2 ∑∑I

Io = Ad2 + I∑∑

Figure 42. Moment of Inertia of Combined Area

g. Determine the distance from the centroid of the combined area to each edge of

the combined area, c3 and c2, cm (in.).

c1 = Ts + A − x

c2 = x

h. Determine the maximum of either c, or c2 determined in Step 5g, c, cm (in.).




i. Calculate the section modulus of the combined area, Z, cm3 (in.3).

Z =Ioc




Intermediate Wind Girder Evaluation

6. Use the following formula to calculate the maximum height of the unstiffened shell:


H1 = 72t12tD

3 161V

2H1 = 6 100 t( ) 100 t

D

3 100V

2

Where: H1 = Vertical distance between the intermediate wind girder and the top angle of theshell or the top wind girder of an open-top tank, m (ft.)

t = Thickness of the top shell course excluding corrosion allowance, cm (in.)

D and V = As defined in Step 1.

7. Use the following formula to calculate the transposed width of each shell course:

Wtr = Wcourse

tuniform

tactual

5

Where: Wtr = Transposed width of each shell course, m (ft.)

Wcourse = Actual width of each shell course, m (ft.)

tuniform = Design thickness of the top shell course, excluding corrosion allowance,cm (in.)

tactual = Design thickness of the shell course for which the transposed width isbeing calculated, excluding corrosion allowance, cm (in.)

Add up the transposed widths of all the shell courses that were obtained from these

calculations to determine the height of the transformed shell, Wtr-total. The conceptual

relationship between the actual shell height and the transformed shell height is illustrated

in Figure 43. The table in Figure 44 may be used to summarize the results of the

transformed shell calculations.




Figure 43. Actual Shell and Transformed Shell




Course W Wtr = Wtuniform

tactual

5

1

2

3

4

5

6

7

8

Wtr −total = Wtr∑

Figure 44. Transformed Shell Calculation Summary

8. If the height of the transformed shell is less than or equal to H1 as calculated in Step 6,

no intermediate wind girder is required. If the height of the transformed shell is greater




than H1, an intermediate wind girder is required. If (Wtr-total/2) >H1, a second

intermediate wind girder is needed.

Note that H1 need only consider the distance between the intermediate wind girder and

the top wind girder (not the top of the shell). Therefore, the top wind girder should be

located first before concluding that a second intermediate wind girder is required. Only

one intermediate wind girder is usually required.




9. If an intermediate wind girder is required, it should be placed at the mid-height of the

transformed shell. The location of the wind girder on the actual shell should be at the

same course and same relative position as on the transformed shell. See Figure 42. This

location is determined using the thickness relationship that was used in Step 7.

The intermediate wind girder must not be placed within 150 mm (6 in.) of a horizontal

joint of the shell. When the preliminary location of an intermediate wind girder is

within this distance from a horizontal joint, the girder shall preferably be located 150

mm (6 in.) below the joint.

10. Verify that no segment of the transformed shell between any wind girder and another

wind girder or the bottom of the tank exceeds H1.

11. Calculate the minimum required section modulus of the intermediate wind girder using

the following formula:


ZI = 0 . 0577 D2H1

V

161

2

ZI = 0 . 0001 D 2H1

V

100

2

Where: ZI = Required minimum intermediate wind girder section modulus, cm3 (in.3)

H1 = Vertical distance between the intermediate wind girder and the top angle of theshell or the top wind girder of an open-top tank, m (ft.)

D and V=As previously defined.

12. A standard structural steel shape may be used for the intermediate wind girder. Refer to

Table 3-22 of API-650 (excerpted in Figure 39) to select a shape that has a section




modulus that is at least as large as that calculated in Step 11. In using Table 3-22, use

the corroded thickness of the shell course at which the intermediate wind girder is

located.




Note that the use of this table will typically be conservative. API-650 permits including

a portion of the tank shell for a distance of 1.47 Dt above and below the attachment of

the intermediate wind girder to the shell in determining the available section modulus.

Table 3-22 uses a smaller distance along the tank shell.

Section modulus calculations may be made for the geometry of the intermediate wind

girder that is actually used. These calculations may be done in situations where a

nonstandard intermediate wind girder is used, if the corroded shell course thickness

differs from values that are shown in Table 3-22, or if the total permissible portion of the

tank shell is considered. The general calculation approach that is described in Step 5

may be used to make this section modulus calculation.




Work Aid 3: PROCEDURE FOR DETERMINING WHETHER CONTRACTOR-SPECIFIED DETAILS FOR Opening design ARE ACCEPTABLE

This Work Aid is to be used with the copies of SAES-D-100 and 32-SAMSS-005 that are in Course Handout 2,and with the copy of API-650 that is in Course Handout 1. All needed tank design information should beobtained from the Contractor Design Package.

1. If the opening is not attached to the storage tank by full penetration welding, reject the

design. Failure to meet this requirement is not expected to be a problem for new tank

construction. The only exception to the requirement is that partial penetration welds

may be used for insert-type reinforcement that conforms to API-650 details.

2. If the opening is for a 50 mm (2 in.) or smaller standard weight coupling, no additional

reinforcement is required.

3. Determine which of the following categories matches the opening being reviewed and

proceed accordingly.

• Shell manhole, proceed to Step 4

• Shell nozzle and flange, proceed to Step 5

• Flush-type cleanout fitting, proceed to Step 6

• Flush-type shell connection, proceed to Step 7

• Roof manholes, proceed to Step 21

• Roof nozzles, proceed to Step 22

Except for flush-type openings and connections, the reinforcement must be located

within a distance above and below the centerline of the opening equal to the vertical

dimension of the opening in the tank shell plate. The reinforcement can be provided by

any one or any combination of the following:




• Attachment flange of the fitting

• Excess plate thickness beyond that required by Work Aid 1

• Portion of the fitting neck that may be considered as reinforcement

• Reinforcing plate

Detailed evaluation of nozzle reinforcement for only hydrostatic loads will normally not

be required as long as the nozzle design details and dimensions conform to a standard

API-650 design.

Shell Manholes

4. Verify that the manhole conforms to API-650, Figures 3-4A and 3-4B, and the

dimensional requirements contained in Tables 3-3 through 3-7 for the specified manhole

size. If no problems are detected, the design is acceptable.

Shell Nozzle and Flange

5. Verify that the nozzle and flange conform to API-650, Figure 3-5 and the dimensional

requirements contained in Tables 3-8 through 3-10 for the specified nozzle size. If no

problems are detected the design is acceptable. Excerpts from these portions of API-650

are contained in Figure 45 through Figure 48.




Figure 45. Shell Nozzles




Column 1 Column 2 Column 3 Column 4 Column 5 Column 6 Column 7c Column 8c Column 9c

Size ofNozzle

OutsideDiameterof Pipe

NominalThicknessof Flanged

Nozzle PipeWalla

t n

Diameter ofHole in

ReinforcingPlateDR

Length ofSide of

ReinforcingPlateb orDiameter

L=Do

Width ofReinforcing

PlateW

MinimumDistance

from Shellto Flange

FaceJ

Minimum Distance fromBottom of Tank to Center of

Nozzle Regular Low Typed Type HN C

Flanged Fittings4846444240

4846444240

ccccc

48 1/846 1/844 1/842 1/840 1/8

96 3/492 3/488 3/484 3/480 3/4

117112

107 1/4102 1/297 3/4

1616151515

5250484644

48 3/846 3/844 3/842 3/840 3/8

242220

1816141210

86432f

1 1/2f

242220

181614

12 3/410 3/4

8 5/86 5/84 1/23 1/22 1/81.90

0.500.500.50

0.500.500.500.500.50

0.500.4320.3370.3000.2180.200

24 1/822 1/820 1/8

18 1/816 1/814 1/812 7/810 7/8

8 1/46 3/44 5/83 5/82 1/2

2

49 1/245 1/241 1/2

37 1/233 1/229 1/2

2723

1915 1/4

1210 1/2

____

6055 1/450 1/2

45 1/440 1/4

3633

28 1/4

23 1/419 1/215 1/413 1/2

____

121111

10101099

887766

282624

2220181715

13119876

24 3/422 3/420 3/4

18 3/416 3/414 3/413 1/211 1/2

9 1/27 7/8

65 1/43 1/2

3Threaded Fittings

3g

2f

1 1/2f

1f

3/4f

4.002.8752.2001.5761.313

CouplingCouplingCouplingCouplingCoupling

4 1/83

2 1/81 11/161 7/16

11 1/4________

14 1/4________

__________

97654

5 3/83333

Figure 46. Dimensions for Shell Nozzles




Column 1 Column 2 Column 3 Column 4 Column 5 Column 6Thickness of

Shell andReinforcing

Platea

t and T

Minimum PipeWall Thickness

of FlangedNozzlesb

t n

MaximumDiameter of Holein Shell plate (Dp)

Equals outsideDiameter of Pipe

Plus

Size ofFillet Weld B

Size of Fillet Weld A Nozzles 2-, 1 1/2-, 1-, Larger Than and 1/4 inch 2 Inches Nozzles

3/161/4

5/163/8

7/161/2

9/165/8

11/163/4

13/167/8

15/161

1/21/21/21/21/21/2

1/21/21/21/2

1/21/21/21/2

5/85/85/85/85/85/8

3/43/43/43/4

3/43/43/43/4

3/161/4

5/163/8

7/161/2

9/165/8

11/163/4

13/167/8

15/161

1/41/41/41/41/41/4

1/45/165/165/16

3/83/83/8

7/16

1/41/41/41/41/4

5/16

5/165/165/165/16

5/165/165/165/16

Figure 47. Dimensions for Shell Nozzles, Pipe, Plate, and Welding Schedules




Figure 48. Dimensions for Shell Nozzle Flanges




Flush-Type Cleanout Fittings

6. Verify that the flush-type cleanout fitting and its support conform to API-650, Figures 3-

9 and 3-10. Verify that the dimensions conform to Tables 3-11 through 3-13. Verify

that PWHT in accordance with API-650 paragraph 3.7.4 will be provided on the fitting

assembly. If no problems are detected, the design is acceptable.

Flush-Type Shell Connections

7. Verify that calculations are provided to confirm that the design prohibits shell uplift at

the cylindrical-shell-to-flat-bottom junction.

8. Verify that calculations are provided to confirm that the vertical membrane stress in the

cylindrical shell at the top of the opening for the connection does not exceed one-tenth

of the circumferential design stress in the lowest shell course that contains the opening.

9. Verify that the maximum width of the flush-type connection does not exceed

914 mm (36 in.) and that the maximum height of the opening does not exceed 305 mm

(12 in.).

10. Verify that the thickness of the bottom transition plate in the assembly is at least 13 mm

(0.5 in.).

11. Verify that the shell plate that contains the flush-type shell connection is given PWHT in

accordance with paragraph 3.7.8.3 of API-650.




12. Verify that the flush-type shell connection conforms to API-650 Figure 3-11 and the

dimensional requirements that are contained in Table 3-14 for the specified opening

size.

13. The cross-sectional area of the reinforcement over the top of the connection must satisfy

the following equation:

A r ≥K 1ht

2

Where: Ar = Cross-sectional area of reinforcement, mm2 (in.2)

K1 = Area coefficient based on tank diameter and height, and opening height, fromAPI-650 Figure 3-8

h = Maximum height of the opening, mm (in.)

t = Thickness of the shell plate that contains the flush-type connection, mm (in.)

14. Verify that the shell plate that contains the flush-type connection is the proper thickness.

Except for 203 mm (8 in.) by 203 mm (8 in.) flush-type connection openings, the shell

plate that contains the flush-type connection must be at least 1.6 mm (1/16 in.) and not

more than 3.2 mm (1/8 in.) thicker than the adjacent plates in the bottom shell course.

For a 203 mm (8 in.) by 203 mm (8 in.) flush-type connection opening, the shell plate

that contains the flush-type connection may be the same thickness as the shell plates in

the remainder of the bottom shell course.

15. Verify that the reinforcing plate is the same thickness as the shell plate that contains the

flush-type connection.

16. Verify that the following equation is satisfied.




L ≤ 1.5h

Where: L = Height of the reinforcing plate above the bottom of the opening, mm (in.)

h = Height of the opening, mm (in.)

If L must exceed this limit in order for [(L – h) ≥ 150 mm (6 in.)], only the portion of

reinforcement that is within 1.5h can be considered effective.

17. Verify that the width of the tank bottom reinforcing plate at the opening centerline is at

least 254 mm (10 in.) plus the combined thickness of the shell plate in the connection

assembly and the shell reinforcing plate.

18. Verify that the thickness of the bottom reinforcing plate conforms to the following:


tb =h2

355 600+ 0.00584b H tb =

h2

14 000+

b310

H

Where: tb = Minimum thickness of the bottom reinforcing plate, mm (in.)

h = Maximum height of the opening, mm (in.)

b = Maximum width of the opening, mm (in.)


In no case shall the thickness of the bottom reinforcing plate be less than:

• 16 mm (5/8 in.) for H = 14.6 m (48 ft.)

• 17.5 mm (11/16 in.) for H = 17 m (56 ft.)

• 19.1 mm (3/4 in.) for H = 19.5 m (64 ft.)

19. Confirm that the minimum thickness of the nozzle neck and transition piece, tn, is 16

mm (5/8 in.). Externally applied loads may require this area to be thicker.




20. If no problems are detected in Steps 7 through 20, the flush-type connection design is

acceptable.

Roof Manholes

21. Verify that the roof manhole conforms to API-650 Figure 3-13 and the dimensional

requirements contained in Table 3-15 for the specified manhole size. If no problems are

detected, the design is acceptable.

Roof Nozzles

22. Verify that flanged roof nozzles conform to API-650 Figure 3-14, and the dimensional

requirements contained in Table 3-16 for the specified nozzle size. Verify that threaded

roof nozzles conform to API-650 Figure 3-15, and the dimensional requirements

contained in Table 3-17 for the specified nozzle size. If no problems are detected, the

design is acceptable.




Work Aid 4: PROCEDURE FOR DETERMINING WHETHER CONTRACTOR-SPECIFIED DESIGN DETAILS FOR TANK ROOFS AREACCEPTABLE

This Work Aid is to be used with the class reference copies of SAES-D-100 and32-SAMSS-005 that are in Course Handout 2, and the copy of API-650 that is in Course Handout 1. Allneeded tank design information should be obtained from the Contractor Design Package.

General

1. Verify that the roof plates have a minimum nominal thickness of 4.76 mm

(3/16 in.). Any corrosion allowance should be added to this nominal thickness for

supported roofs.

2. Verify that the roof and its supporting structure are designed for the dead load plus a

uniform live load of 122 kg/m2 (25 lb./ft.2) times the projected area.

3. Verify that all internal and external structural members have a minimum nominal

thickness of 4.3 mm (0.17 in.) plus corrosion allowance.

4. If the roof is a fixed roof and is specified to have a frangible joint, proceed to Step 5. If

the roof is not specified to be frangible and is a fixed roof, proceed to Step 9. If the roof

is an external floating roof, proceed to Step 28. If the roof is an internal floating roof,

proceed to Step 46.

Frangible Fixed Roof

5. Verify that the following formula is satisfied:


A ≤217W

30 800 tanθA ≤

0.153W30 800 tanθ




Where:

A = Area resisting the compressive force, mm2 (in.2). All members in the region of theroof-to-shell junction, including insulation rings, are considered to contribute to thecross-sectional area. Refer to API-650 Figure F-1 for determination of thecontributing areas of the roof, shell, and top angle. An excerpt from Figure F-1 iscontained in Figure 49.

= Aroof + Ashell + Aangle

Aroof = Roof contributing area, mm2 (in.2)

= thwh

wh = Smaller of 0.3 R2th or 305 mm (12 in.)

Ashell = Shell contributing area, mm2 (in.2)

= tcwc

wc = 0.6 Rc tc

Aangle = Cross sectional area of top angle, mm2 (in.2)

R2 = Length of normal to the roof, mm (in.)

= Rc / sinθ

Rc = Inside radius of tank shell, mm (in.)

th = Nominal thickness of roof plate, mm (in.)

tc = Nominal thickness of top shell course, mm (in.)

W = Total weight of the shell and any framing, but not the roof plates, that are supportedby the shell and roof, kg (lb.)

θ = Angle between the roof and a horizontal plane at the roof-to-shell junction, degrees

tan θ = Slope of the roof, expressed as a decimal quantity

If the formula is not satisfied, the joint is not frangible.




Figure 49. Permissible Compression Ring Details




6. Verify that the shell-to-roof compression ring details conform to one of the details in

API-650 Figure F-1, details a through d (See Figure 49). If the details do not conform,

the joint is not frangible.

7. Verify that the continuous fillet weld between the roof plates and the top angle does not

exceed 4.76 mm (3/16 in.). If the fillet weld exceeds 4.76 mm (3/16 in.), the joint is not

frangible.

8. Verify that the slope of the roof at the top angle does not exceed 1:6. If the slope of the

roof exceeds 1:6, the joint is not frangible.

General Fixed Roof Tanks

9. For a fixed roof that is not frangible, verify that proper protection against excessive

pressure in accordance with API Standard 2000, Venting Atmospheric and Low-Pressure

Storage Tanks, is provided. If the protection is inadequate, the roof is unsatisfactory.

10. If lateral loads for the roof support columns have been specified, verify that the columns

are designed to satisfy the following formula. If the formula is not satisfied, the roof

support columns are unsatisfactory.

f aF a

+f bxFbx

+fby

Fby≤ 1

Where: fa = Computed axial stress, kPa (psi)

Fa = Allowable axial stress, if the axial force alone existed, kPa (psi)

fb = Computed compressive bending stress at the point under consideration, kPa (psi)

Fb = Allowable compressive bending stress, if the compressive bending moment aloneexisted, kPa (psi)




x and y = Axis of bending about which the stress applies, kPa (psi)

11. Verify that the seams of the roof plates are welded on the top side with continuous full-

fillet welds. If the seam welds are not continuous full-fillet welds on the top side, the

roof is unsatisfactory.

12. Verify that the roof plates are attached to the top angle of the tank with a continuous

fillet weld on the top side only. The fillet weld size shall be 4.76 mm (3/16 in.) unless a

smaller size is specified. If the roof plates are not attached in the prescribed manner, the


13. For a fixed roof tank with a supported cone roof, proceed to Step 14. For a fixed roof

tank with a self-supported cone roof, proceed to Step 18. For a self-supporting dome

roof proceed to Step 23.

Supported Cone Roof Tank

14. Verify that the slope of the roof is at least 1:16. If the slope is less than 1:16, the roof is

unsatisfactory.

15. Verify that the tank shell is equipped with a top angle that is at least the size given in

Figure 50. If the top angle is too small, the roof is unsatisfactory.

TANK DIAMETER, D MINIMUM TOP ANGLE SIZE

m ft. mm in.

D≤10.7 D≤35 50.8 x 50.8 x 4.76 2 x 2 x 3/16

10.7<D≤18.3 35<D≤60 50.8 x 50.8 x 6.35 2 x 2 x 1/4

D>18.3 D>60 76.2 x 76.2 x 9.53 3 x 3 x 3/8

Figure 50. Minimum Top Angle Size




16. Verify that the rafters that are located in the outer ring are spaced at no more than 1.91

m (6.28 ft.) center-to-center, and that the rafters in the inner rings are not more than 1.68

m (5-1/2 ft.) center-to-center. If the rafters are spaced too far apart, the roof is

unsatisfactory.

17. If a supported cone roof design has been satisfactory in all steps to this point, the roof is

satisfactory for the purposes of this Work Aid.

Self-Supporting Cone Roof

18. Verify that the top-angle sections are joined by butt welds that have complete

penetration and fusion. If the sections are not joined in the prescribed manner, the roof

is unsatisfactory.

19. Verify that the slope of the roof is at least 2:12 and no more than 9:12. If the slope of

the roof is outside the allowable range, the roof is unsatisfactory.

20. Verify that the roof thickness is at least the thickness that is calculated by the following

formula, but does not exceed 12.7 mm (1/2 in.). If the thickness is outside the specified

range, the roof is unsatisfactory.


troof =0.208D

sinθ≥ 4.76 mm troof =

D400 sinθ

≥ 3 /16 in.

Where: troof = Thickness of the roof, mm (in.)




D = Nominal diameter of the tank shell, m (ft.)

θ = Angle of cone elements to the horizontal, degrees

21. Determine the participating area at the roof-to-shell junction, based on API-650 Figure

F-1, and the nominal material thicknesses minus the corrosion allowance. See Step 5 for

calculation of the participating area. Note that the corroded roof and shell plate

thicknesses are used for this calculation. However, the nominal thicknesses are used for

the calculations done in Step 5. The amount of participating area must be at least the

area calculated by the following formula. If the participating area is less than the

amount specified, the roof is unsatisfactory.


Ap ≥2.314D2

sinθAp ≥

D2

3 000 sinθ

Where: Ap = Participating area at the roof-to-shell junction, mm2 (in.2)

D = Nominal diameter of tank shell, m (ft.)

θ = Angle of cone elements to the horizontal, degrees

22. If a self-supporting cone roof has been satisfactory in all steps to this point, the roof is


Self-Supporting Dome Roof

23. Verify that the top angle sections are joined by butt welds that have complete

penetration and fusion. If the sections are not joined in the prescribed manner, the roof

is unsatisfactory.




24. Verify that the radius of curvature of the roof satisfies the following formula. If the

radius of curvature does not satisfy the following formula, the roof is unsatisfactory.

0.8D ≤ rr ≤ 1.2D

Where: D = Nominal tank diameter, m (ft.)

rr = Roof radius, m (ft.)

25. Verify that the thickness is at least the thickness calculated by the following formula, but

does not exceed 12.7 mm (1/2 in.). If the thickness is outside the specified range, the



troof ≥ 0.4166r r ≥ 4.76 mm troof ≥rr

200≥ 3 /16 in.

Where: troof = Thickness of roof, in.

rr = Radius of roof, ft.

26. Determine the participating area at the roof-to-shell junction, based on API-650 Figure

F-1, and the nominal material thicknesses minus any corrosion allowance. See Step 5

for calculation of the participating area. Note that the corroded roof and shell plate

thicknesses are used for this calculation. However, the nominal thicknesses are used for

the calculations done in Step 5. The amount of participating area must be at least the

area that is calculated by the following formula. If the participating area is less than the

amount specified, the roof is unsatisfactory.





Ap ≥ 4.627DrrAp ≥

Drr

1500

Where: Ap =Participating area of the roof-to-shell junction, mm2 (in.2)

D =Nominal tank shell diameter, m (ft.)

rr =Roof radius, m (ft.)

27. If a self-supporting dome roof has been satisfactory in all steps to this point, the roof is


External Floating Roof

28. Verify that the top of the roof has a minimum slope of 1:64. If the slope is less than

1:64, the roof is unsatisfactory.

29. Verify that the deck of a pontoon type roof will be in contact with the liquid during

normal operation.

30. Verify that the deck plates are joined by continuous full-fillet welds on the top side.

Confirm that full-fillet welds that are at least 50 mm (2 in.) long on 250 mm (10 in.)

centers are used on the underside of any lap plates that are located within 300 mm (12

in.) of stiffening girders or support legs.

31. Verify that vendor calculations confirm the roof has sufficient buoyancy to remain afloat

when the tank contains a liquid with a specific gravity of 0.7 under the following

conditions:

• 250 mm (10 in.) of rainfall in a 24 hour period, with the roof intact. However, double-deckroofs with emergency drains that are designed to prevent the stored liquid from flowing ontothe roof may be designed for a lesser rainwater volume as determined by the drain sizing.

• In single-deck pontoon roofs, single-deck and any two adjacent pontoon compartmentspunctured with no water or live load.




• In double-deck pontoon roofs, any two adjacent compartments punctured with no water or liveload.

If sufficient buoyancy does not exist, the roof is unsatisfactory.

32. Verify that vendor calculations confirm that the pontoons of single deck roofs will not

be permanently distorted under the loads that are specified in Step 31. Also confirm that

the roof sag will not permit any stored liquid to get on the roof through any roof

penetrations (e.g., leg sleeves or vents).

33. Verify that each pontoon compartment has a liquid-tight manway. The top edge of the

manway necks must be at an elevation that prevents liquid entry into the compartment

under the conditions specified in Step 31. If the manways are not provided or are

inadequate, the roof is unsatisfactory.

34. Verify that each pontoon compartment has vents that protect against internal or external

pressure. The vents must be at a level that prevents liquid entry into the compartment

under the conditions specified in Step 31. If the vents are not provided or are

inadequate, the roof is unsatisfactory.

35. Verify that a ladder has been provided that automatically adjusts to any roof position. If

the ladder is a rolling ladder, verify that it has full-length handrails on both sides and has

been designed for a 454 kg (1 000 lb.) mid-point load with the ladder in any operating

position. If a ladder is not provided or is inadequate, the roof is unsatisfactory.




36. Verify that roof drains are adequate to prevent backflow of stored liquid in case of drain

pipe leakage, and to prevent the roof from accumulating a water level greater than

design at the maximum rainfall rate when the roof is floating at the minimum operating

level. Verify that a check valve is installed in the drain line at the roof sump to prevent

backflow. The drain size must be at least 76.2 mm

(3 in.) for roofs that are no larger than 36.6 m (120 ft.) in diameter, and at least 101.6

mm (4 in.) for roofs that are larger than 36.6 m (120 ft.) in diameter. If the drains are

not provided or are undersized, the roof is unsatisfactory.

37. Verify that adequate automatic bleeder vents and rim vents are provided to prevent

overstressing of the roof deck or seal membrane. Vents that are capable of evacuating

air and gases from underneath the roof during initial filling must be provided. If the

vents are not provided or are inadequate, the roof is unsatisfactory.

38. Verify that roof support legs are provided. The support legs must be capable of

supporting the roof and a uniform live load of 122 kg/m2 (25 lb./ft.2). The length of the

legs must be adjustable from the top of the roof. The low-leg position must provide at

least 2 m (6.5 ft.) clearance between the lowest portion of the roof and the tank bottom.

All tank appurtenances must be cleared by the roof in its lowest position. If the support

legs are not provided or are inadequate, the roof is unsatisfactory.

39. Verify that at least two roof manholes with a minimum inside diameter of at least 610

mm (24 in.) have been provided. Verify that all roof manholes that were specified in the




purchase specification are provided. If the manholes are not provided or are inadequate,

the roof is unsatisfactory.

40. Verify that an adequate sealing device is provided between the periphery of the roof and

the tank shell. If the device is not provided or is inadequate, the roof is unsatisfactory.

41. Verify that all bulkheads and rims inside the pontoon or double deck roofs are

continuously welded along their upper edges to the top deck plates.

42. Verify that inlet pipes have been equipped with diffusers, if specified in the purchase

specification.

43. Verify that roof centering and anti-rotation devices have been provided.




44. Verify that the additional requirements that are contained in Appendix C of

32-SAMSS-005 are satisfied.

45. If an external floating roof has been satisfactory in all steps to this point, the roof is


Internal Floating Roof

46. Verify that the design complies with API-650, Appendix H.




Work Aid 5: PROCEDURE FOR determining whether contractor-specifiedDesign details for Tank Bottoms are acceptable

This Work Aid is to be used with the copies of SAES-D-100 and 32-SAMSS-005 in Course Handout 2, andwith the copy of API-650 in Course Handout 1. All needed tank design information should be obtained fromthe Contractor Design Package.

1. Verify that the tank bottom has a minimum nominal thickness of 6.35 mm (1/4 in.) plus

any specified corrosion allowance. In general, any specified tank corrosion allowance

must be specifically stated as also applying to the bottom to be considered. A tank

bottom corrosion allowance is normally not used.

2. Verify that the tank bottom plates will project at least 25 mm (1 in.) beyond the outside

edge of the weld that attaches the bottom to the shell.

3. If the bottom shell course of the tank is designed based on the allowable stress for

materials in Group IV, IVA, V, or VI, proceed with Step 4. If the bottom shell course of

the tank is designed using other materials, proceed with Step 11.

4. Verify that butt-welded annular rings are used based on the criteria that follows:

• Butt-welded annular rings must be used if the material allowable stress governs the bottomshell course thickness. However, the use of butt-welded annular bottom plates are alwaysacceptable.

• Lap-welded annular bottom plates (i.e., sketch plates) are acceptable if the product stress forthe first shell course does not exceed 160 MPa (23 200 psi) and the maximum hydrostatic teststress for the first shell course does not exceed 172 MPa (24 900 psi).

5. Use the following formula to calculate the hydrostatic test stress in the first shell course:


Sh =4.9D H− 0.3( )

tSh =

2.6D H −1( )t




Where: Sh = Hydrostatic test stress in the first shell course, MPa (psi)

D = Nominal diameter of the tank, m (ft.)

H = Design liquid height, m (ft.)

t = Nominal thickness of the bottom shell course as constructed, mm (in.)

6. Verify that butt-welded annular rings have at least the thickness that is specified in

Figure 51.

Sh, psit, in. Š27 000 Š 30 000 Š 33 000 Š 36 000

t Š 0.750.75 < t Š 1.001.00 < t Š 1.251.25 < t Š 1.501.50 < t Š 1.75

1/41/41/45/16

11/32

1/49/32

11/327/161/2

9/323/8

15/329/165/8

11/327/169/16

11/16

Figure 51. Minimum Permissible Annular Ring Thickness

7. Verify that the butt-welded ring has a radial width that provides at least 600 mm (24 in.)

between the inside of the shell and any lap-welded joint in the remainder of the bottom

and at least a 50 mm (2 in.) projection outside the shell.

8. Verify that the butt-welded annular ring has a radial width that is at least that required by

the following formula:


r =215 tb

HGr =

390 tbHG




Where: r = Radial width of the annular bottom plate, mm (in.)

tb = Thickness of the annular bottom plate, mm (in.)


G = Design specific gravity of the liquid to be stored

If this formula results in a radial width that is less than the requirement of Step 7, Step 7

governs the radial width.

9. Verify that the ring of annular bottom plates has a circular outside circumference, and

either a circular or regular polygonal shape with the number of sides equal to the number

of annular plates inside the tank.

10. Verify that the bottom of the tank is a cone up (unless the Tank Data Sheet specifies a

cone down, in which case it must be a cone down) with a slope of at least 1:120.

11. Verify that the bottom-to-shell junction conforms to Para. 3.1.5.7 of API-650 (reference

Figure 52).




Figure 52. Bottom-to-Shell Junction Details




12. If water drawoff is specified, verify that the draw-off sump and outlet pipe comply with

the requirements of Figures 53 through 55 (taken from Saudi Aramco Drawing AD-

36500) and the dimensional requirements that are contained in Table 3-18 of API-650.

TANK DIAMETER WATER DRAW-OFF PIPE SIZE

m ft. mm in.

3 to 6 10 to 20 50 2

6 to 27 20 to 90 100 4

27 to 48 90 to 160 150 6

Figure 53. Required Water Draw-Off Pipe Size

13. Verify that the top of the foundation ring under the annular plate is provided with a 13

mm (1/2 in.) thick asphalt-impregnated board.

14. Verify that a foundation drip plate has been provided in accordance with Saudi Aramco

Standard Drawing AB-036050.




Figure 54. Water Draw-off for Tanks Over 6 m (20 ft.)Diameter




Figure 55. Water Draw-off for Tanks 3 m (10 ft.) to 6 m (20ft.) Diameter




glossary

annular ring A butt welded outer ring of bottom shell plates on which the tank shell rests.

cone roof A fixed roof formed approximately to the surface of a right cone.

dome roof A fixed roof formed approximately to a spherical surface.

frangible roof A fixed roof designed such that the roof-to-shell junction will fail due tohigh internal pressure prior to failure of any other shell or bottom seam.

NPS Nominal pipe size in inches.

obround A shape formed by two parallel sides with semicircular ends.

PWHT Postweld heat treatment is bringing the temperature of a metal object up to ahigh enough level for a prolonged duration of time to remove stress that iscaused during the fabrication process.

self-supporting roof A fixed roof designed to be supported only at its periphery at the tank shell.

sketch plates A lap-welded outer ring of bottom plates on which the tank shell rests.

supported roof A roof that has its principal support provided by rafters that are mounted ongirders and columns, or by rafters on trusses with or without columns.

mechanical design of tank structure

Documents

design factors

intermediate wind girder

technical material

point method

foot method

pertinent sections of

hydrostatic test case

additional information