Acoustic emission monitoring during hydrotesting of a mounded LPG storage vessel of petrochemical industry

Abstract

One mounded LPG storage vessel of petrochemical plant got suddenly exposed to cryogenic nitrogen and developed cracks in the shell plate adjacent to the bottom nozzle and was subsequently repair welded. The structural integrity of the repair weld region was assessed by NDT examinations. The integrity of the repaired weld region was confirmed by conducting hydrotest. Acoustic emission (AE) monitoring during the hydrotest was also carried out. The acoustic emission signals generated during the hydrotest indicated that the signals were due to structural noise and microyielding of the vessel material. During the repressurizing cycle of the hydrotest, negligible AE signals were generated and this confirmed the structural integrity of the repair welded region of the vessel monitored by acoustic emission technique (AET).

Keywords

Acoustic emission hydrotest mounded LPG storage vessel non destructive testing petrochemical industry

1. Introduction

This paper reports the use of acoustic emisison technique (AET) towards assessing structural integrity of one mounded LPG storage vessel of petrochemical industry. The vessel was installed in 1998. The internal inspection of the vessel was carried out in 2004 and reinspected and hydrotesting was carried out in 2009. The storage vessel was taken regularly for statutory inspection at 5 years interval. The vessel was being prepared for recommissioning after a planned statutory outage. While preparing for recommissioning, it got exposed to cryogenic nitrogen which was let in inadvertently. The nitrogen supply was stopped immediately. Internal examination of the vessel was carried out and this showed the presence of cracks in the shell plate adjacent to the bottom nozzle through which cold liquid/gas nitrogen was introduced in the vessel. Detailed inspection by non destructive testing (NDT) was carried out and this revealed the presence of cracks extending to about 740 mm longitudinally from bottom nozzle towards the far side dished end and about 530 mm from the edge of bottom nozzle on either side circumferentially. The damaged portion was found to be confined to an area of 740 mm × 1317 mm around the nozzle. The rest of the vessel including all weldments did not show any sign of damage or presence of crack when tested by different NDT techniques. It was inferred that the vessel could be brought back to normal working condition by cutting out the defective portion and by inserting and welding a new plate of same specification. It was decided that, after carrying out various NDT examinations including X-radiography, ultrasonic and dye penetrant of the new weld joints, hydrotest of the vessel would be conducted at the original test pressure. It was also decided to monitor acoustic emission (AE) during the hydrotest at the location of the repair, in order to ensure the absence of any growing discontinuity in the weld region during the hydrotest and confirm the structural integrity of this part of the vessel.

Acoustic emission technique is a nondestructive testing tool used for assessing the structural integrity of pressure vessels and pipelines [1]. AET can be advantageously used in preference to other NDT techniques to locate and evaluate growing (dynamic) discontinuities in an entire structure at one time rather than selectively testing localized areas for static discontinuities as done by other NDT techniques. Acoustic emission is the class of phenomenon whereby transient elastic waves are generated by rapid release of energy from localized sources like places of transient relaxation of stress and strain fields in a material [1]. An AE sensor coupled to a sample undergoing dynamic changes detects a part of the elastic energy emitted in the form of elastic waves and provides the information about the nature of dynamic changes taking place in the sample. A few examples of AE emitting phenomena relevant to structural integrity monitoring are deformation, crack growth and fracture, corrosion, leaks and cavitations in fluid systems. Being highly sensitive, acoustic emisison technique can be used for detection of any micro-crack initiation, propagation of an existing micro-crack, and also a fluid leakage at inaccessible locations.

Hydro and pneumatic testing are employed for qualification and in-service assessment of industrial components operating under pressure [2]. While this testing can indicate that there are no through and through leaks, it would not indicate the presence of harmful defects in the components, that would have grown during hydrotest. On-line AE monitoring during hydro and pneumatic testing enables detection of such growing (dynamic) discontinuities in an entire structure at one time rather than selectively testing localized areas for static discontinuities. By employing multi sensor approach, it is also possible to locate the growing discontinuity, if any, during the hydrotesting [3]. AE technique was used for structural integrity assessment of a number of components [4–8]. AE was used to monitor the structural defects of a repaired storage tank [4]. AE signals recorded during loading of the tank was due to weld like defects and this was verified by subsequent radiographic testing [4]. AE was used to detect environmental cracking in an HF storage sphere at an early stage, without the need for frequent internal inspections [5]. AET has been used for online inspection of storage tank floors and for evaluation of corrosion and leakage areas in the tank floors, and in turn for assessment of structural integrity of the tank [6]. In another investigation [8], the results of AE monitoring during hydrotesting of exchange towers of a heavy water plant have been reported and based on these results, different regions of the towers monitored by AE are reported to confirm the structural integrity.

During hydrotesting of any component, micro yielding of material of the component occurs and residual stresses may be relieved. This micro yielding can be a source of acoustic emission. It should be noted that any subsequent pressure test would only stimulate AE from defects or other damage that had occurred in the component during service. By selectively placing AE sensors at different locations, growth of defects can be monitored. The objective of the present investigation was thus to monitor acoustic emission during hydrotesting of the mounded LPG storage vessel and to ascertain that no defect is growing in the repaired weld region during pressurization. AE monitoring was carried out for part of the vessel at one end covering the weld region, where repair work was done and the results are discussed in this paper. The AE monitoring during hydrotesting of the vessel was carried out jointly by Indira Gandhi Centre for Atomic Research (IGCAR), Kalpakkam and National Metallurgical Laboratory (NML), Jamshedpur.

2. Method

2.1. Storage vessel

Figures 1a–1b show the mounded LPG storage vessel which was subjected to hydrotesting. The vessel was constructed as per the code BS 5500(1994) & SMPV Rules 1981. The overall length of the vessel is 79,050 mm and it has a storage capacity of 1430 MT of LPG. The vessel is made up of carbon steels. Geometric details of the vessel are as follows:

Shell thickness: 27–30 mm

Corrosion allowance: 1.5 mm

Internal diameter: 16 mm

WL/WL length: 72,000 mm

Erection weight (MT): 396

Operational weight (MT): 2037

The vessel was painted with a 450 μm thick black epoxy paint, to prevent corrosion from soil. In addition, cathodic protection was also done. The design pressure of the vessel is 14.55 kg/cm² and can withstand a hydrostatic pressure of 18.7 kg/cm².

Fig. 1a.

Mounded LPG storage vessel.

Fig. 1b.

Mounded LPG storage vessel (close view).

2.2. Hydrotesting

Hydrotests are periodically carried out on industrial pressure vessels which are in-service to re-qualify the vessels for their continued operation. The test pressures and procedures are decided as per the respective ASME codes. The hydrotest of the storage vessel in this investigation was carried out by increasing the pressure from zero to 18.7 kg/cm² in the first cycle. The pressure was increased in steps and by periodic holding at different pressures. After the hydrotest, the pressure was reduced to 12 kg/cm². A repressurizing cycle was carried out immediately after the first cycle of the hydrotest. The repressurizing cycle was carried out up to 18.7 kg/cm² pressure. Details of the hydrotest of the vessel are given in Table 1.

Table 1
Details of the hydrotest of the vessel

Pressure range, kg/cm² Pressure holding conditions

Pressure, kg/cm² Time (min)

1–4 4 15

4–8.6 8.6 25

8.6–12 12 10

12–15 15 10

15–18.7 (repressurizing cycle) 18.7 30

Pressure range, kg/cm²	Pressure holding conditions
1–4	4	15
4–8.6	8.6	25
8.6–12	12	10
12–15	15	10
15–18.7 (repressurizing cycle)	18.7	30

2.3. Recording of acoustic emission signals

Two separate AE systems (named as System-A and System-B) were used for recording the acoustic emission signals generated during hydrotesting. This also helped to enhance the reliability of the inspection. For AE monitoring of the weld region using system A, six sensors (S1 to S6) were used in a rectangular configuration. Three sensors were placed in the front side and three sensors were placed in the backside. The positions of the sensors are shown in Fig. 2a. The sensors were of 150 kHz resonant frequency type each. Preamplifiers of 40 dB gain and compatible filters (100–300 kHz) were used along with the sensors. A threshold of 30 dB was optimized. AE monitoring was carried out using 16 channel DiSP-4 model system supplied by Physical Acoustic Corporation, USA. A distance of 1.8 m along the axial direction and 7.5 m along the circumferential direction between the sensors was maintained. All the sensors were connected to the acoustic emission testing system. For AE monitoring of the same region by using system B, four sensors (resonant type with center frequency of 125 kHz and built-in preamplifiers) were mounted at four different locations as shown in Fig. 2b for monitoring the entire region of interest. DISP AE win equipment of M/s Physical Acoustics Corporation, USA was used to record the signals. The gain in the preamplifier was fixed at 40 dB. AE sensors were grouped #I (1–2–3), #II (2–3–4) and #III (1–2–4) to cover the complete weldment. The sensors and the AE systems were calibrated using pencil lead break source. The location of the AE sources was also calibrated using the pencil break source. Acoustic emission signals generated during pressurizing, pressure hold and repressurizing cycles of the vessel were recorded and analysed.

Fig. 2a.

Location of new cut out plate along with the positions of the AE sensors, S1 to S6 for system A.

Fig. 2b.

Location of new cut out plate along with the positions of the AE sensors, 1 to 4 for system B.

3. Results

3.1. System-A

The acoustic emisison signals generated were analysed using location softwares. Figures 3a–3c show the location of the AE sources in the rectangular region for pressurization of the vessel in different pressure ranges i.e. from 0–4 kg/cm² (Fig. 3a), 4–12 kg/cm² (Fig. 3b) and 12–18.7 kg/cm² (Fig. 3c). The variation of total count vs. time for different sensors for one of the above ranges of pressurizations (4–12 kg/cm²) is shown in Fig. 4. The amplitudes of the signals generated for different sensors were analysed. Typical amplitude distribution plot for pressurization 4–12 kg/cm² for one of the sensors (S1) is shown in Fig. 5. Peak amplitude of AE signals generated during any particular event can be related to the intensity of the source in the material producing acoustic emission and thus can be used for source characterization. The results from all the sensors and for different pressure ranges are given in Table 2.

Fig. 3a.

Location of AE sources during pressurization 0–4 kg/cm² for system A.

Fig. 3b.

Location of AE sources during pressurization 4–12 kg/cm² for system A.

Fig. 3c.

Location of AE sources during pressurization 12–18.7 kg/cm² for system A.

Fig. 4.

Total count of AE signal vs time during pressurization 4–12 kg/cm² for system A. S1 to S6 indicates sensors 1 to 6.

Fig. 5.

Distribution of peak amplitude of AE hits generated for sensor 1 (S1) during pressurization 4–12 kg/cm².

It is seen that during pressurization in the range 0–4 kg/cm² (Fig. 3a), AE signals are generated at a few locations in the rectangle made by the sensors 1 to 6. During pressurization at higher pressures i.e. from 4–12 kg/cm² (Fig. 3b) and 12–18.7 kg/cm² (Fig. 3c), signals are generated at a few newer locations in the rectangle. During different pressurization cycles, AE signals are mainly generated at the central region of the rectangle and can be understood by the fact that the vessel exerts more pressure in the central region due to expansion during hydrotesting. It is seen from Fig. 4 that the AE counts are generated during the pressurization cycles only and no AE is generated during the pressure holding periods. Similar results were obtained during the other pressurization ranges.

The amplitude of AE signals (Fig. 5 and Table 2) show that, in the lower pressure range (0–4 kg/cm²), signals are generated with peak amplitude up to 56 dB. In the pressure range from 4–12 kg/cm², signals are generated with peak amplitude up to 68 dB. In the pressure range from 12–18.7 kg/cm², signals are generated up to 58 dB. Thus, a few hits of higher peak amplitude (up to 68 dB) which are generated during pressurization from 4–12 kg/cm², are not generated in the higher pressure range (12–18.7 kg/cm²). The spike like signals of amplitude >60 dB for the pressurization from 4–12 kg/cm² were also not generated continuously. After the first cycle of pressurization, repressurization (from 12 to 18.7 kg/cm²) of the vessel was carried out. The location results during repressurization are shown in Fig. 6 which shows absence of AE generation. These observations indicate that the acoustic emission signals generated in the rectangular weld repair region during hydrotesting are due to expansion of the vessel and noise by support structure and also due to microyielding of the material of the vessel.

3.2. System-B

The AE results obtained by using system-B are shown for some of the pressure ranges. Figures 7a–7c show the location results for pressurization from 0–8.6 kg/cm² (Fig. 7a), 8.6–12.0 kg/cm² (Fig. 7b) and 15–16 kg/cm² (Fig. 7c). The variation of total count vs. time for one of the pressurizations (8.6–12 kg/cm²) is shown in Fig. 8. The amplitude distributions of the signals for different pressure ranges are shown in Figs 9a–9c.

Table 2
Maximum peak amplitudes of AE signals for different sensors for system-A

Pressure range, kg/cm² Maximum peak amplitude of AE signals for different sensors, dB

S1 S2 S3 S4 S5 S6

0–4 56 56 52 53 56 50

4–12 68 67 68 68 67 68

12–18.7 58 58 58 56 58 58

Pressure range, kg/cm²	Maximum peak amplitude of AE signals for different sensors, dB
0–4	56	56	52	53	56	50
4–12	68	67	68	68	67	68
12–18.7	58	58	58	56	58	58

Fig. 6.

Location of AE sources during repressurization 12–18.7 kg/cm² for system A.

Fig. 7a.

Location of AE sources during pressurization 0–8.6 kg/cm² for system B.

Fig. 7b.

Location of AE sources during pressurization 8.6–12 kg/cm² for system B.

Fig. 7c.

Location of AE sources during pressurization 15–16 kg/cm² for system B.

Fig. 8.

AE cumulative counts during pressurization 8.6–12 kg/cm² for system B.

Fig. 9a.

AE amplitude distribution during pressurization 0–8.6 kg/cm² for system B.

Fig. 9b.

AE amplitude distribution during pressurization 8.6–12 kg/cm² for system B.

Fig. 9c.

AE amplitude distribution during pressurization 15–16 kg/cm² for system B.

It can be seen from Figs 7a–7c that the AE signals are mostly generated from the repaired welded zone. In the pressure range of 8.6–12.0 kg/cm², the vessel is at the normal operating pressure. The cumulative count for the pressurization 8.6–12 kg/cm² (Fig. 8) indicates that the AE counts are generated during the pressurization cycles only and no AE is generated during the pressure holding periods. Similar results were obtained from other pressurization ranges. During holding period at 12 kg/cm² pressure also, the AE response was silent which indicated no growing defect during this hold time (encircled in the figure).

The amplitude distribution results showed that the amplitudes of the signals for different pressure ranges are below 60 dB and mostly below 50 dB. The amplitude distribution in the pressure range 8.6–12.0 kg/cm² (Fig. 9b) indicates that the peak amplitudes are discrete in nature. A few high-energy burst type signals were generated from the weld zone shown in this pressure range (Fig. 9b). This is similar to the results of higher peak amplitude (up to 68 dB) signals in the pressure range from 4–12 kg/cm² observed by system-A. But this was not generated continuously. In high pressure ranges i.e. in the pressure range from 12 kg/cm² to 15 kg/cm², no significant activities are observed and most of the signals are below 60 dB amplitude. During the pressure range 15 to 16 kg/cm², some of the events generated were due to continuous rainfall combined with scrolling down of loose rubble. In still higher pressure ranges (16.1–18.7 kg/cm²), a very few high energy AE signals were observed, but these were very random in nature and most of the signals were below 65 dB amplitude. No sustained AE activities could also be seen in these stages. During repressurization, negligible AE was observed. This is similar to that observed for system A. Most of the signals during repressurization were also below 60 dB.

4. Discussion

Acoustic emisison during hydrotesting of a vessel could occur due to expantion of the vessel as well as due to any crack/flaw if it is growing. During hydrotesting of the storage vessel under investigation, majority of the AE events generated were random in nature. They were not also generated continuously thus ruling out the possibility of any continuous process. During the pressure holding periods, the AE response was silent. The above observations indicated that there was no dynamic activity in the vessel at any pressurization even at maximum operating pressure. In the pressure range of 8.6–12 kg/cm², the signals captured by both the systems were of higher amplitude. But such high amplitude signals were not generated in the subsequent higher pressure cycles. The higher amplitude signals generated during 8.6–12 kg/cm² could be attributed to the noise. During hold time at hydrotest pressure (18.7 kg/cm²), a few high energy bursts were observed but they also did not indicate any damage progression, as they were random in nature and were of very short duration events. Such events are attributed to mechanical noise.

It is also seen that AE is generated only during the pressurization cycles. During pressurization of a vessel, AE could be generated due to deformation of the material of the vessel and/or due to growth of any defect. The generation of AE due to the occurrence of deformation in metallic materials has shown that the generation and motion of dislocations are the main activity giving rise to AE signals and the rate of the AE generation is maximum during yielding of materials [9]. The AE signals generated in the storage vessel during hydrotesting are attributed to the micro yielding of the material of the vessel and to the structural noise. The negligible AE generated during repressurization cycle is due to ‘Kaiser effect’ of acoustic emission generation [1]. The generation of AE during deformation in metallic materials is characterized by the ‘Kaiser effect’ which states that, unless the previous maximum stress is exceeded, no emission would be generated during loading of the material in the subsequent cycle of loading. This means that AE during repressurization of the vessel up to a pressure of previous maximum pressure can only occur if there has been any formation or growth of any discontinuity. AE generated during repressurization up to the pressure of previous maximum can thus occur only when there is growth of any defect which is produced during in-service or during previous pressurization following the guideline governing Dunegan Corollary [10]. The “Dunegan Corollary” states that the acoustic emission experienced during proof testing reveals damage occurred during the previous operational period. In case a structure suffers no damage during a particular operational period, then on the subsequent proof test no acoustic emission is observed. However, in the event of a significant change of state within a working period, during the subsequent proof loading AE is observed, as the discontinuity is subjected to a higher stress than previously. The Dunegan Corollary is widely used for the inspection of pressurized systems [11]. In the case of the present vessel, acoustic emission captured by both the systems during the repressurization cycle was negligible. This confirmed that there is no growing discontinuity at any locations of the vessel monitored by AET during the hydro testing.

5. Conclusion

Acoustic emission monitoring of one repaired mounded LPG storage vessel which showed the presence of cracks during NDT examinations and subsequently repair welded, was carried out during hydrotesting. The acoustic emission signals generated during the hydrotest indicated that the signals were due to structural noise and microyielding of the vessel material. The negligible acoustic emission during the repressurizing cycle confirmed that there was no growing discontinuity in the repair welded region of the vessel monitored by AE during the hydrotesting.

Footnotes

Acknowledgements

Authors are thankful to Mr. S.C. Chetal, Former Director, Indira Gandhi Centre for Atomic Research (IGCAR), Kalpakkam, Dr. P.R. Vasudeva Rao, Director, IGCAR, and Dr. S. Srikanth, Director, National Metallurgical Laboratory, Jamshedpur, for constant encouragement and support. The authors thank Dr. B.P.C. Rao, Head, Non Destructive Evaluation Division (NDED), IGCAR, for useful discussions. The authors thank Mr. S Arunkumar, NDED, IGCAR, for his help during the above work. The authors would like to thank Mr. S.N. Kumar, Mr. J.B. Torne and Mr. D. Das, LPG Recovery Plant USAR, Raigad, Maharashtra, India for many valuable support and help during the work. The authors are also thankful to Dr. V.R. Krishnan and Mr. A. Amin, Engineers India Ltd, New Delhi, India, for many useful discussions during the course of the above work.

Conflict of interest

The authors have no conflict of interest to report.

References

Testing

A.E.

, Nondestructive Testing Handbook, 3rd edn, Vol. 6, American Society for Nondestructive Testing, 2005.

Peacock

M.J.

, Acoustic emission monitoring of a large pressure vessel during a pneumatic re-qualification test, Journal of Acoustic Emission 8 (1989), 11–19.

Guang

Ting

X.Y.

W.A.

Z.B.

and Xue

H.W.

, AE testing of a low-alloy steel pressure vessel, NDT&E International 26 (1993), 291–294. doi:10.1016/0963-8695(93)90003-D.

Kwon

J.R.

Lyu

G.J.

Lee

T.H.

and Kim

J.Y.

, Acoustic emission testing of repaired storage tank, International Journal of Pressure Vessels and Piping 78 (2003), 373–378. doi:10.1016/S0308-0161(01)00038-2.

Peacock

, Acoustic emission testing of an HF acid sphere, Insight 46 (2004), 461–467. doi:10.1784/insi.46.8.461.39371.

Liu

Guo

and Jin

, Comprehensive inspection and evaluation technique for atmospheric storage tanks, Nondestructive Testing and Evaluation 25 (2010), 45–59. doi:10.1080/10589750902795358.

Raj

Mukhopadhyay

C.K.

and Jayakumar

, Frontiers in NDE research nearing maturity for exploitation to ensure structural integrity of pressure retaining components, International Journal of Pressure Vessels and Piping 83 (2006), 322–335. doi:10.1016/j.ijpvp.2006.02.025.

Mukhopadhyay

C.K.

Haneef

T.K.

Rao

B.P.C.

and Jayakumar

, Condition monitoring of exchange towers of heavy water plant during hydrotesting by acoustic emission, Insight 53 (2011), 29–33. doi:10.1784/insi.2011.53.1.29.

Heiple

C.R.

and Carpenter

S.H.

, Acoustic emission produced by deformation of metals and alloys – Review: Part I, Journal of Acoustic Emission 6 (1987), 177–204.

10.

Harris

D.O.

and Dunegan

H.L.

, Verification of structural integrity of pressure vessels byacoustic emission and periodic proof testing, in: Testing for Prediction of Material Performance in Structures and Components Shane

R.S.

, ed., ASTM Special Technical Publication, Philadelphia, PA, 1972, pp. 158–170. doi:10.1520/STP34682S.

11.

ASTM, Standard test method for examination of liquid filled atmospheric and low pressure metal storage tanks using acoustic emission, in: Annual Book of ASTM Standards, v.03.03: Nondestructive Testing, American Society for the Testing of Materials, Philadelphia, PA, 1997, E1930-97.

Pressure range, kg/cm²	Pressure holding conditions

	Pressure, kg/cm²	Time (min)
1–4	4	15
4–8.6	8.6	25
8.6–12	12	10
12–15	15	10
15–18.7 (repressurizing cycle)	18.7	30