Ultrasonic Testing for Spherical Tank Shell Welds

SIUI NDT

Abstract: The structure of spherical tanks, common flaws of shell welds, and methods of conventional ultrasonic testing (UT), Time-of-Flight Diffraction ultrasonic testing (TOFD) and phased-array ultrasonic testing (PAUT) for shell welds as well as their advantages and disadvantages are described in this study.

Keywords: spherical tank, shell weld, ultrasonic testing, TOFD ultrasound, phased-array ultrasound testing 

1. Common Flaws of Spherical Tank Structure and Shell Welds

1.1 Structure of spherical tanks

Spherical tanks, due to the symmetry of geometric shapes and uniform load carrying capacity, have the highest bearing capacity compared with pressure vessels of other shapes of the same wall thickness, and have the smallest surface area compared with other pressure vessels of the same volume. Therefore, when storing the materials of the same pressure and volume, using a spherical tank will save the most steel material and occupy the smallest area [1]. For this reason, spherical tanks have received wide application. As statistics shows, there are more than 5800 spherical tanks with a volume of over 500m3 in China[1].  

A spherical tank is composed of two parts: pedestal and spherical shell. The form of pedestal includes two types: pillars and plinths. A pedestal is for small volume support, and a pillar is for a volume greater than 50m3. Fig. 1 shows the form of pillar. A spherical shell is welded by multiple spherical petal steel plates, the general nominal volumes are 50, 120, 200, 400, 1000, 2000, 5000 and 10000 m3, and the common materials are 20R, 16MnR, 15MnVR and so on, with wall thickness in the range of 20~60mm[2]. According to the size of the volume, spherical shells are classified as 3-band, 4-band, 5-band, 6-band, and 7-band, which are subdivided into F, D, B, A, C, E and G zones, representing the upper polar zone, upper cold zone, upper temperate zone, equatorial zone, lower cold zone, lower temperate zone and lower polar zone respectively. See Fig. 1. The upper and lower polar zones have the same structure and are welded by 7 arc plates, as shown in Fig. 1(f). The number of other zones and the distribution of spherical petal-shaped plates are determined according to the volume of the spherical shell. See Table 1[2].

1.2 Common Flaws in Weld of Spherical Tank Shell

Spherical tanks are generally pre-assembled in the factory and then transported to the site for overall installation. The spherical shell welding of spherical tanks is also completed on site. Affected by welding conditions and environmental factors, weld flaws such as porosity, slag inclusion, incomplete fusion, incomplete penetration and cracks may occur, which not only affect product quality and service life, but even endanger personal lives and property safety. Therefore, during the construction of spherical tanks, the welds of spherical shells must be subjected to nondestructive testing. In addition, spherical tanks are mostly used to store chemicals such as oil and natural gas. In long-term use, due to the influence of the medium, the original tiny flaws in the welds may continue to expand or derive new flaws, therefore the welds of the spherical tank shells in service shall be inspected regularly. The main testing methods are X-ray and ultrasound. X-ray testing is effective, but it may be easily affected by factors such as component thickness and flaw orientation, and the testing efficiency is low. It requires high protection measures for inspectors, which is not comparable to the advantages of ultrasonic testing. For this reason, the methods of UT, TOFD and PAUT are introduced for testing spherical tank shell welds.

2. Distribution of Spherical Tank Shell Welds for Testing

For the sake of description, the shell weld of a 1000m3 spherical tank in a petrochemical plant was used as the test object. The spherical shell of the spherical tank has three zones, including upper and lower polar zones and equatorial zone. The distribution of steel plates and welds in each zone is shown in Figs 2 and 3. The upper and lower polar zones were welded by 7 steel plates respectively, and 11 welds were produced respectively. The equatorial zone was welded by 16 steel plates, and 16 welds were produced. The total number of steel plates is 30, and the number of welds is 38. The steel plate material is 16MnR, thickness 48mm, and the welding process was manual arc welding, X-shaped groove, width of upper and lower grooves 30mm. When using ultrasonic testing for the welds, considering the large curvature radius of the testing surface, they can be regarded as flat welds.

3. Ultrasonic Testing

3.1 Implementing Standard

The ultrasonic testing of spherical tank shell welds was used by implementing the standard NB/T47013.3 “Nondestructive testing of pressure-bearing equipment-Part 3: Ultrasonic Testing”.   

The testing technical grade B was selected, i.e. using bilateral double-sided testing with a K-value probe (used in this example). If the conditions are restricted, K-value probe bilateral single-sided testing with 2 K-value probes is used instead. [3].

Quality gradeⅠwas selected, which requires that the length of single flaw less than or equal to 50mm when the reflection amplitude of flaw falls between DAC Curve E (evaluation curve) and Curve S (statistics curve); or the length of single flaw less than or equal to 1/3 wall thickness (less than or equal to 16mm when the reflection amplitude of flaw exceeds DAC Curve S [4].

3.2 Testing Equipment

Testing equipment: digital ultrasonic flaw detector Smartor (SIUI) , AFP2.5-1414-60L angle probe; all performance indicators of the instrument and the probe complied with the technical requirements of NB/T47013.3[5]; standard test block CSK-ⅠA, and comparison test block CSK-ⅡA.

3.3 Setting

The equipment setting steps were as follows:

• Adjusting the horizontal depth of the instrument according to the thickness of the test object and calibrate the velocity and zero point;
• Testing the probe flank and K value: use a CSK-1A test block to test the probe incident point and K value;
• Making DAC curves: the probe was placed on a CSK-ⅡA test block, and DAC curves were drawn by using φ2* 60mm transverse holes with depths of 10, 20, 30, 40 and 50mm as reflectors respectively. The sensitivities of Curve E, Curve S and Curve R were φ2* 60mm-14dB, φ1* 60mm-8dB and φ2* 60mm +2dB[6] respectively. The schematic diagram of DAC curves is shown in Fig. 4. In actual testing, for surface compensation, the sensitivity is generally increased by 4dB during testing

3.4 Testing and Assessment

According to the technical grades, a bilateral double-sided primary wave reflection method with an angle-beam probe placed vertically on the testing surface of weld centerline was used in this example for zigzag scanning. See Fig. 5. The scanning area was greater than or equal to 1.25P (P=2KT=2*2*48= 192mm);1.25P=1.25*192= 240mm; scanning speed no greater than 150mm/S; the coverage of adjacent areas for each scan greater than 10% of the probe ceramic width; the scanning area polished for rust removal; chemical paste for coupling agent.

During scanning, any flaw signals with the amplitude above DAC Line E should be evaluated, and recorded for their positions, peak amplitude, DAC curve zone and flaw length.

The flaw position shall be subject to the position where the peak reflected wave of flaw is obtained; the peak amplitude is the amplitude of the peak reflected wave signal and the DAC curve area where it is located shall be determined; the flaw length shall be based on the peak amplitude point, and the probe shall be moved left and right to measure the distance between two points where the peak amplitude is reduced by 6dB, i.e. the flaw length;　when the flaw reflected wave has only one high point and is located at or above the DAC Line E, the amplitude shall be reduced to 80% full screen height (FSH), and the length shall be measured by 6dB method; When the peak value of flaw reflection wave fluctuates, there are many high points and it is above DAC Line E, lower the amplitude to 80% FSH, and measure its length with the endpoint 6dB method.

3.5 Advantages and disadvantages of UT

The equipment for the inspection o spherical tank shell welds with UT is light weight, easy to carry and simple to operate. It is effective to test common flaws in welds by zigzag scanning perpendicular to the centerline of welds. However, UT also has its disadvantages:

• The probe shall have zigzag scan on both sides and both ends of the weld to achieve full coverage of the weld by beams, with low testing speed and low  efficiency;
• Echo signals displayed on the screen includes various true and false flaw signals such as weld excessive height, endpoint, deformed waves, which is demanding on testing personnel, inspectors with long-term practical experience for effective distinguishing and identification are preferred;
• No inspection can be recorded with UT, which is non-traceable;
• Scaffolding is required for manual UT of spherical tank shell welds, high cost and inconvenient for operation.

4. TOFD (Time Of Flight Diffraction)

4.1 Implementing standard

The TOFD ultrasonic testing of spherical tank shell butt welds was used by implementing the standard NB/T47013.3 “Nondestructive testing of pressure-bearing equipment-Part 10: TOFD ultrasonic testing”.   

The testing technical grade is set as Level A, that is, the blind areas of single side scanning, scanning surface and bottom surface shall be less than or equal to 1 mm. The testing method is mainly aimed at longitudinal flaws, excluding transverse flaws[7] .

The quality grade shall be Grade I, that is, for 48mm wall thickness, the length of surface opening or near-surface flaw shall not be greater than or equal to 20mm and the height shall not be greater than or equal to 3mm; the length of buried flaw shall not be greater than or equal to 20mm and the height shall not be greater than or equal to 5mm[8] .

4.2 Testing equipment

The TOFD ultrasonic testing equipment includes an instrument, a crawler with an encoder, transmitting and receiving probes, and a test block for calibration. According to NB/T 47013, when the workpiece thickness is less than or equal to 50mm, adopt a broad-band narrow-pulse probe with T/R for single-sided testing. The functions and performance indicators of the testing equipment comply with NB/T47013[9] .

4.3 Setting and calibration

The setup and calibration steps are as follows:

• Probe spacing settings

The two probes were symmetrically arranged relative to the centerline of the weld, and the center distance PCS was set according to the 2/3 rule, i.e. the intersection point of the beams of the two probes located at the position[10]  of 2/3 wall thickness of the workpiece to be tested, as shown in Fig. 7. The specific value is calculated according to Formula (1), where the coverage of beams is the best.

PCS=2S=2*2/3T*tanθ…… （1）

Where: PCS-center distance between two probes;

       T-workpiece wall thickness; θ-probe refraction angle.

Substitute the wall thickness T=48mm, tanθ=tan600 =1.73 into Formula (1), and set PCS according to the calculated value;

• Time window settings

The A-scan time window start position is set to 0. 5μs or above, and the end position

is 0. 5μs or above after a deformation wave on the bottom surface of the workpiece;

• Sensitivity settings

When the sensitivity is set by artificial flaws on the comparison block, the amplitude of diffraction signal is set between 40~80% FSH, and the surface coupling compensation of about 4dB is carried out in actual scanning.

When the sensitivity setting is based on the amplitude of the lateral wave on the workpiece, the amplitude height of the lateral wave is set between 40 and 80% FSH.

• Scan Increment Setting

When the workpiece thickness is in the range of 12mm≤T≤150mm, the maximum scanning increment is 1.0 mm.

• Encoder calibration

When the crawler moves 500mm on the test block or workpiece, the error between the displacement value displayed by the instrument and the actual displacement value shall be less than 1%;

• Depth calibration

Measure the signal of the reflector with known thickness on the tested workpiece or test block and obtain the depth value, with the error not greater than 1% or 0.5mm of the actual thickness.

4.4 Testing and assessment

According to the weld distribution diagram in Figs 3 and 4 above, before testing, all the weld excessive heights around T and Y joints were grinded to be flush with the surrounding parent metal, and the grinding length was 250mm outward from the intersection center, so as to ensure that the crawler could pass through the joint smoothly and the probe contact was good.

Non-parallel scanning was used for this testing. The crawler moves along the weld length direction, and the scanning direction is perpendicular to the beam direction, as shown in Fig. 7. The maximum speed shall not exceed the calculated value of Formula (2). During scanning, the transmitting probe generates non-focused longitudinal beams incident on the tested workpiece at a certain angle, some of which propagate along the near surface and is received by the receiving probe, which are called lateral waves, and some of which are reflected by the bottom surface and received by the probe, which are called bottom waves. When there are flaws in the tested workpiece, the diffraction signals received by the receiving probe through the tip of the flaw are called upper and lower endpoint diffraction waves. These signals are synthesized with the probe position signals from the encoder to form D scan diagram, as shown in Fig. 9.

Vmax= PRF·ΔX / N  [11]   …… （2）

Where: Vmax-maximum scanning velocity, mm/ s; PRF-repetition frequency, Hz;

ΔX-Set scanning increment value, mm; N-Average number of instrument signals (generally set 1~4).

During the scanning process, if any suspected flaws are found, they can be evaluated in real time, or the data can be stored for later evaluation.

If it is determined that the testing data is valid and defective, the fitting arc cursor method is generally used to determine the position of the front and rear endpoints along the X axis direction. The flaw length is calculated from the position of the front and rear endpoints. The depth is the position of the upper endpoint, and the flaw height is the distance between the upper and lower endpoints. If the length and height exceed the set quality limits, it is judged as failed.  

4.5 Advantages and disadvantages of TOFD

Compared with ultrasonic testing methods, TOFD has the following advantages and disadvantages:

4.5.1 Advantages of TOFD

• Regardless of beam angles, flaws in any direction within the beam coverage can be detected, with higher detection rate of 80% or above usually;
• Higher quantitative accuracy, height quantitative error for linear or volumetric flaws less than 1mm;
• Fast and simple testing, light weight, easy handling, real-time observation of testing results, timely analysis of testing results, higher testing efficiency;
• Test results can be recorded throughout the process, stored for a long time and traceable;
• Scaffolding is required for manual TOFD ultrasonic testing of spherical tank shell welds, high cost.

4.5.2 Disadvantages of TOFD

• There are near-surface and near-bottom blind spots, flaws located near the surface may be submerged by lateral waves and un-detected, and flaws near the bottom may be submerged by bottom waves and missed detection;
• Characterizing flaws is difficult and requires experienced inspectors to perform testing;
• Benign flaws might be exaggerated, such as small porosity, and the quantification of porosity flaws is inaccurate.
• Scaffolding is required for manual TOFD ultrasonic testing of spherical tank shell welds, high cost and inconvenient for operation.

5. PAUT

5.1 Implementing Standard

Refer to NB/T47013 Nondestructive testing of pressure-bearing equipment-Part 15: PAUT for PAUT of spherical tank shell welds. 

The testing technical grade is B, which requires phased-array beams to achieve at least two full coverage of the testing area.

The scanning mode can be realized from bilateral double-side or bilateral single-side [12] . The following is an example of bilateral single-side scan.

The quality classification shall be Grade I. When the amplitude of flaw signal is between Curve E and Curve S, the length of single flaw shall not exceed 50 mm. When the amplitude of flaw signals exceeds Curve S, the flaw length shall be less than or equal to 13 of the wall thickness, i.e., for 48mm wall thickness, the flaw length shall not exceed 16mm[13] .

5.2 Testing equipment

The PAUT equipment consists of an instrument, a crawler with an encoder, probes, and a test block for calibration.

Referring to NB/T 47013 standard, when detecting plate welds, for workpieces of technical grade B and wall thickness of 40≤T≤100, bilateral single-side primary wave and secondary wave sector scanning can be adopted, and each lateral shall have greater than or equal to 2 probe positions[14] ,i.e., when one scanning can complete full coverage, two probes are required on each end, which are respectively placed at different positions in the weld center, and 4 probes are required on both ends.

For this reason, 4 probes of model 5.0L32-0.6-10 were selected, with frequency of 5MHz, 32 elements and a 20N55S wedge. The equipment selected is SIUI integrated ultrasonic testing system ALPHA 1006, an automated four-wheeled scanner with built-in phased array + TOFD module, which can be controlled with a remote controller and realize real-time wirelessly data transmission with a laptop. The simulation test block was welded with steel plates of similar material and wall thickness to the test object and artificial flaws were made. CSK-IA was used as the standard test block, PRB-II was used for the comparison test block. The appearance of the equipment is shown in Fig. 10. Its functions and performance indicators complied with NB/T47013 [15] .

5.3 Setup

The setup steps are as follows:

• Workpiece management: input information such as workpiece thickness, groove type, weld width, and construct the structure diagram of the tested workpiece;
• Probe and wedge selection: input the selected probe and wedge model (including frequency, number of elements, element size, wedge size and angle);
• Scan and echo type selection: select sector scan for scan type, TW for echo type ;
• Set the distance from the flank of two sets of probes (including wedge) to the weld center separately (positions of two sets of probes are different) and steering angle range (steering angles of the two groups of probes are different), wherein probes 1 and 2 are one group, and the same focusing law is adopted for incidence from both sides, and the lower half of the weld is mainly tested by the primary wave; the probes 3 and 4 are the other group, and the same focusing law is adopted for incidence from both sides, and the upper half of the weld is mainly tested by the secondary wave; and then, the beam simulation function of the instrument is adopted, and by adjusting the distance between the two groups of probes and the weld center and the steering angle range, Make the beams generated by the two groups of probes fully cover the testing area and overlap each other by more than 25% in the depth range of adjacent sections at the upper and lower parts, as shown in Fig. 11.

• Display mode selection: There are a variety of display modes to choose from, A+B+C shall be selected generally;

Through the steps above, the setup of the equipment has been completed for the workpiece to be inspected before starting the calibration.

5.4 Calibration

CSK-ⅠA was used as calibration standard block, PRB-Ⅱ was used as comparison block.

Calibration includes velocity, delay, AGC, TCG, and encoder calibration as follows:

• Velocity calibration: to obtain accurate velocity of the workpiece under testing, CSK-ⅠA test block calibration can be used;
• Delay calibration: to correct that the distance from each A scan line to the same acoustic path reflector is equal and the display zero is consistent with the surface of the workpiece under testing. CSK-IA test block calibration can be used.
• AGC calibration: to correct that the energy of each A-scan line reaching the reflector of the same sound path is equal, which is conducive to the subsequent TCG calibration. CSK-IA test block calibration can be used.
• TCG calibration: prepare TCG curves for flaw quantification, so that the echo amplitudes of reflectors with the same size at different sound paths detected by beams at the same angle are the same, and the depth range of calibration shall include at least 48 mm. According to NB/T47013 standard, the φ2 transverse holes of PRB-II contrast test block with different depths are used as reflection reference points for calibration, and the sensitivity of Curve E is φ2 - 14dB, Curve S is φ2 - 8 dB and the Curve R is φ2+ 2 dB.
• Encoder calibration: When the crawler is moved 500mm on the test block or workpiece, the error between the displacement value displayed by the instrument and the actual displacement value should be less than 1%.

5.5 Testing and Assessment

5.5.1 Testing

According to the weld distribution diagram in Figs 3 and 4 above, before testing, all the weld excessive heights around T and Y joints were grinded to be flush with the surrounding parent metal, and the grinding length was 250mm outward from the intersection center, so as to ensure that the crawler could pass through the joint smoothly and the probe contact was good.

The testing shall be conducted by combining longitudinal vertical scanning with electronic sector scanning. The crawler shall scan along the weld length direction, and the direction shall be vertical to the beam trend. The deviation between the probe position and the weld center position shall not be greater than 5%, as shown in Fig. 13. The maximum speed shall not exceed the calculated value of Formula (3)[16] .

Vmax= PRF·ΔX / N ·M [9]   …… （3）

Where: Vmax-maximum scanning velocity, mm/ s; PRF-repetition frequency, Hz;

ΔX-set scanning increment value, mm; N-Average number of instrument signals (generally set 1~4).

M-Number of sector sweep steps.

The probe used for PAUT is different from the probe for UT. The piezoelectric ceramic (PZT) is an array composed of multiple independent small PZT units. A computer is used to control each PZT to transmit and receive ultrasound according to a certain delay rule. The probe does not need to move back and forth. It can cover the entire weld section pointed by beams at one time, realizing the steering and focusing of the beams in the inspected area. The testing results can generate A, S, B, C and other scanning patterns, with higher flaw detection rate, resolution, measurement accuracy and testing reliability.

During the scanning process, if suspected flaws are found, they can be evaluated in real time, or the data can be stored for later evaluation.

5.5.2 Assessment

During the scanning process, when flaws are found, measure "flaw depth and horizontal distance" on image A, "flaw length" on image C, and "flaw height" on image B. This is shown in Fig. 13.

After measuring flaws, the quality of the tested workpiece can be evaluated according to the selected quality level. If the flaw length exceeds the selected quality level requirements, it is failed.

5.6 Advantages and Disadvantages of PAUT

5.6.1 Advantages of PAUT

Compared with UT and TOFD, PAUT has the following advantages:

PAUT adopts multi-angle scanning, and the probe can cover the entire weld section pointed by beams at one time without moving back and forth;

• Visual images with A/S/B/C scans can measure length, depth, height and horizontal position of flaws, which assist inspectors to determine the nature of flaws;
• Fast and simple testing, light weight, easy handling, real-time observation of testing results, timely analysis of testing results, higher testing efficiency;
• Test results can be recorded throughout the process, stored for a long time and traceable.

5.6.1 Disadvantages of PAUT

• Flaw orientation and beam incidence angle affect echo energy intensity, resulting in errors in flaw equivalent measurement due to the ever-changing shape of natural flaws;
• The testing rate of single porosity is low;
• Scaffolding is required for manual PAUT of spherical tank shell welds, high cost and inconvenient for operation.

6. Conclusion

The UT, TOFD and PAUT methods for spherical tank shell welds and their respective advantages and disadvantages are studied in this paper. Combining the characteristics of these testing methods, PAUT +TOFD method is recommended. PAUT pays more attention to flaws from upper and lower surfaces of welds, while TOFD mainly detects flaws from internal welds. The two ultrasonic imaging technologies are used to inspect welds simultaneously, and their respective advantages can be combined to make the testing results more intuitive and improve flaw detection rate.  

In addition, it can be seen from the article that no matter for UT, TOFD or PAUT, as long as it is manual testing, scaffolding is required, which is expensive and inconvenient to operate. With the continuous development of technology, the automated scanner with integrated ultrasonic module for spherical tank shell inspection will address the demand of weld testing of spherical tank shells in an effective solution.

References

[1] Chen Gang et al. Nondestructive Testing, 2005, 27(1): 27.

[2] Li Jie, Research on Positioning, Navigation and Control Technology of Spherical Tank Testing Robot [D], Southeast University, 2022,P 34.

[3] Energy Industry Standard of the People's Republic of China NB/T47013.3 Nondestructive testing of pressure-bearing equipment Part 3: Ultrasonic testing [S]. National Energy Administration, 2015, P116

[4] Energy Industry Standard of the People's Republic of China NB/T47013.3 Nondestructive testing of pressure-bearing equipment Part 3: Ultrasonic testing [S]. National Energy Administration, 2015, P131

[5] Energy Industry Standard of the People's Republic of China NB/T47013.3 Nondestructive testing of pressure-bearing equipment Part 3: Ultrasonic testing [S]. National Energy Administration, 2015, P90-91

[6] Energy Industry Standard of the People's Republic of China NB/T47013.3 Nondestructive testing of pressure-bearing equipment Part 3: Ultrasonic testing [S]. National Energy Administration, 2015, P125

[7] Energy Industry Standard of the People's Republic of China NB/T47013.10 "Nondestructive testing of pressure-bearing equipment Part 10: Time-of-flight diffraction ultrasonic testing [S]. National Energy Administration, 2015, P371

[8] Energy Industry Standard of the People's Republic of China NB/T47013.10  "Nondestructive testing of pressure-bearing equipment Part 10: Time-of-flight diffraction ultrasonic testing [S]. National Energy Administration, 2015, P381

[9] Energy Industry Standard of the People's Republic of China NB/T47013.10  "Nondestructive testing of pressure-bearing equipment Part 10: Time-of-flight diffraction ultrasonic testing [S]. National Energy Administration, 2015, P365~370

[10] Energy Industry Standard of the People's Republic of China NB/T47013.10  "Nondestructive testing of pressure-bearing equipment Part 10: Time-of-flight diffraction ultrasonic testing [S]. National Energy Administration, 2015, P373

[11] Energy Industry Standard of the People's Republic of China NB/T47013.10 "Nondestructive testing of pressure-bearing equipment Part 10: Time-of-flight diffraction ultrasonic testing [S]. National Energy Administration, 2015, P377

[12] Energy Industry Standard of the People's Republic of China NB/T47013.15 "Nondestructive testing of pressure-bearing equipment Part 15: Phased-array ultrasonic testing [S]. National Energy Administration, 2021, P30

[13] Energy Industry Standard of the People's Republic of China NB/T47013.15 "Nondestructive testing of pressure-bearing equipment Part 15: Phased-array ultrasonic testing [S]. National Energy Administration, 2021, P47

[14] Energy Industry Standard of the People's Republic of China NB/T47013.15 "Nondestructive testing of pressure-bearing equipment Part 15: Phased-array ultrasonic testing [S]. National Energy Administration, 2021, P36

[15] Energy Industry Standard of the People's Republic of China NB/T47013.15 "Nondestructive testing of pressure-bearing equipment Part 15: Phased-array ultrasonic testing [S]. National Energy Administration, 2021, P7~10

[16] Energy Industry Standard of the People's Republic of China NB/T47013.15 "Nondestructive testing of pressure-bearing equipment Part 15: Phased-array ultrasonic testing [S]. National Energy Administration, 2021, p13