Erosion damage of blast joint of dump flood system at a petroleum field

Abstract

After 222 days of normal operation of dump flood system at a petroleum field, one of its blast joints, which are pin-box type, suffered leakage due to internal erosion damage at the joint threaded portion. The main problem was related to improper fitting of the blast joint due to unequal axial threaded distances of its pin and box parts. This improper fitting created severe turbulent flow condition and subsequently severe impact by water jets. In order to minimize such failure in the future, the fitting of the blast joint must be of a higher quality, which can be achieved by having equal axial threaded distances for its pin and box parts.

Keywords

Blast joint axial threaded distance improper fitting erosion damage

1. Background

The dump flood is a mechanism conceived to inject water from a reservoir with an active aquifer (formation water) into a depleted reservoir using a drilled well with marginal production or closed due to low pressure. In this case, a single well is used as a source of water supply and inject domain for enhancing oil recovery. The well is completed by an advanced ESP to produce and inject in the same wellbore across two set of perforations. Blast joints are thick-walled tubing joints installed as part of the tubing string and are placed adjacent/opposite to the well perforations to minimize erosion damage from the high velocity jetting action at the perforations that also may carry abrasive debris from the well formation. The blast joints are made of high strength steel tubes having 3.5” inner diameter, 20 mm thickness and are connected in a series with a total length of 115 feet, above pump intake at 6100ft depth. The normal operation conditions of the subject well are 2490 psi intake pressure, 4050 psi discharge pressure, 1050 BFPD flow rate, 68821 PPM salinity as NaCl, 76730 PPM total dissolved solids, 65417 PPM sodium and potassium chlorides salts, 6337 PPM calcium chloride salts. After 222 days of operation, the lowest blast joint suffered a leakage at its threaded portion. The failed blast joint was dismantled and subjected to different non-destructive and destructive investigations for failure analysis.

2. Investigations

The failed blast joint was carefully examined before sectioning it for destructive investigations. General and enlarged views of the failed blast joint after disassembling its pin and box parts are shown in Fig. 1. Visual inspection showed that leakage occurred at the joint’s threaded zone where its wall thickness was gradually reduced until it reached zero at a localized zone on both pin and box sides. Another important thing we noticed was that the existence of an external mechanical damage started at the front edge of the pin's threaded portion and then extended 500 mm along the pin's non-threaded length; parallel to the flow direction. Figure 2 shows the enlarged views from two opposite sides of both outer and inner surfaces of the leaked zone. Note that the threaded portions on both pin and box sides were damaged. The worst damage was localized at the same zone on both pin and box sides where leakage occurred. Away from leaked/damaged zone, no further external or internal damage was observed. Figure 3 shows close-up views of outer and inner surfaces of the pin side leaked zone. The most important thing we noticed was the damage of the pin's external threaded portion (Fig. 3-a) while no internal damage was observed (Fig. 3-b). Figure 4 shows close-up views of outer and inner surfaces of the box side leaked zone. Except leaked zone, no external damage was observed (Fig. 4-a). Another important thing we noticed was the severe damage of most of the box's internal threads, in particular at the front edge of its threaded portion (Fig. 4-b). In other words, only few rear threads of the box threaded portion were undamaged.

Fig. 1.

General (a) and enlarged (b) views of the failed blast joint after disassembling its pin and box. Wall thickness was gradually reduced until it reached zero at a localized zone on both pin and box sides. Note that the external damage started at the front edge of the pin's threaded portion and then extended along the pin length.

Fig. 2.

Enlarged views from two opposite sides of both outer and inner surfaces of the leaked zone showing damage of threaded portions of both pin and box. The worst damage is localized at the same zone on both pin and box sides where leakage was occurred.

Fig. 3.

Close-up views of outer (a) and inner (b) surfaces of the pin side leaked zone. Note that the pin's external threaded portion was damaged while no internal damage was observed.

Fig. 4.

Close-up views of outer (a) and inner (b) surfaces of the box side leaked zone. Except leaked zone, no external damage was observed, while the box's internal threaded portion was severely damaged.

Enlarged views of the inner surface of both leaked zone and its opposite side after longitudinal splitting and re-assembling are shown in Fig. 5. The most important thing we noticed were the undamaged threads of the box’s rear portion at both leaked zone (Fig. 5-a) and its opposite/facing zone (Fig. 5-b). The pin threaded portion included 21 threads within 70 mm axial distance while the box threaded portion included 27 threads within 100 mm axial distance (Table 1). This means that the shoulder of the pin was not in contact with the front edge of the box. In other words, the box rear threads were not mounted with pin threads. Away from the leaked zone, no internal damage was observed on the inner surface of either pin or box side. Except leaked zone, no reduction in wall thickness was observed either for pin or box sides. In other words, uniform wall thickness of 20 mm average value was obtained for both pin and box sides. It is confirmed that the damage or metal removal started on both the pin's external threaded portion and box's internal threaded portion where wall thickness was gradually reduced until it reached zero at the leaked zone. It should be reported that non-destructive testing using both dye penetrant test (PT) and magnetic particles test (MT) showed no indications for cracking either around or away from leaked zone.

Fig. 5.

Enlarged views of the inner surface of both leaked zone (a) and its opposite side (b) after longitudinal splitting and re-assembling. Note the undamaged threads of the box's rear portion at both leaked zone and its opposite/facing zone.

Table 1

Threaded axial distance of pin and box sides

	Threaded axial distance	No. of threads
Pin	70 mm	21
Box	100 mm	27

Specimens from leaked and non-attacked zones were taken for chemical analysis, metallurgical investigations and hardness measurements. Results of chemical analysis of the failed blast joint are shown in Table 2 where it is normal carbon manganese steel. Optical microscopic photographs with different magnifications of cross sections taken from leaked and non-leaked zones are shown in Figs. 6 and 7, respectively. It is clear that normal banded ferritic-pearlitic structure with no lamination or other internal defects was observed for both leaked (Fig. 6) and non-leaked (Fig. 7) zones. In other words, no distinct abnormality of microstructure was obtained for either leaked or non-leaked zones. In general, no significant difference could be observed in the microstructure of both leaked and undamaged zones.

Fig. 6.

Optical microscopic photographs with different magnifications of a cross section taken from leaked zone showing normal banded ferritic-pearlitic structure.

Fig. 7.

Optical microscopic photographs with different magnifications of a cross section taken from non-leaked zone showing normal banded ferritic-pearlitic structure.

Table 2

Chemical analysis of the failed blast joint’ material (wt%)

	Element
Material	C	Si	Mn	P	S	Cr	Ni	Cu	Al	Cu
Failed	0.17	0.46	1.38	0.013	0.008	0.10	0.16	0.17	0.02	0.15
joint

Survey of hardness measurements of leaked and non-leaked zones of the failed blast joint was carried out and the results are shown in Table 3. The given values are the average of five readings. Almost no considerable difference in hardness values of leaked and non-leaked zones was obtained. Average hardness values of 189HV and 187HV were obtained for leaked and non-leaked zones respectively. These hardness values are normal for the used ferritic-pearlitic carbon manganese steel.

Table 3

Results of hardness measurements of leaked and non-leaked zones of the failed blast joint

Zone	HV (10 kgf)
Leaked zone	189
Non-leaked zone	187

Note: The given hardness values are the average of five measurements.

3. Discussion

Chemical analysis, metallurgical examinations and hardness measurements revealed that the failed blast joint material is normal carbon manganese steel. It is confirmed that mechanical damage occurred on both the pin's external threaded surface and the box's internal threaded surface where wall thickness was gradually reduced until it reached zero at a localized zone. Away from leaked/damaged zone, no further external or internal damage was observed. The failure of the subject blast joint is related to internal erosion damage. In general, liquid erosion is one type of wear that is the progressive loss of original material from a solid surface due to mechanical interaction between the surface and a fluid in a very small area. In other words, the liquid erosion originates from the impact or impingement by liquid jets [1–4].

It is believed that internal erosion of the subject blast joint started at the circumferential contact zone between the front edge of the pin and the box's threaded portion. Basically, both pin and box threads are tapered/slightly conical threads where a positive seal between the threads to be created by thread deformation when they are tightened to the proper torque. However, improper fitting due to unequal axial threaded distances on both pin and box sides has resulted in severe turbulent flow and subsequently, impact or impingement by water jets at the circumferential contact zone of both pin’s front edge and box’s threaded portion. This in turn resulted in gradual mechanical removal of metal particularly, on the pin’s outer surface in comparison with the box’s inner surface. Then, thinning was progressed until leakage was finally occurred in the form of water jet that resulted in extending the damage to the external non-threaded length of the pin; parallel to the flow direction (Fig. 1) [5–8]. The rate of erosion damage is accelerated with high flow rate of water that may carry abrasive particles/debris. The damage is accelerated also by sharp particle shape, higher particle hardness, higher impact velocity and lower impingement angle.

4. Conclusions

Based on the results obtained in this investigation, it can be concluded that the failure of the subject blast joint is related mainly to internal erosion damage. The liquid erosion damage originates from the impact or impingement by liquid jets.

It is believed that internal erosion of the subject blast joint was started at the circumferential contact zone between the front edge of the pin and the box's threaded portion.

Improper fitting due to unequal axial threaded distance on both pin and box sides means that the box rear threads were not mounted with pin threads. This in turn created severe turbulent flow condition and subsequently, impact or impingement by water jets at the circumferential contact zone of both pin’s front edge and box’s threaded portion. This led to gradual mechanical removal of metal particularly, on the pin’s outer surface in comparison with the box’s inner surface. Thinning was progressed gradually until abnormal clearance induced between the pin’s outer surface and box’s inner surface that in turn created stronger impact effect and subsequently, accelerated erosion damage. This finally led to leakage in the form of water jet that resulted in extending the damage to the external non-threaded length of the pin.

5. Recommendations

In order to avoid similar failures in the future, liquid impingement/water jets condition should be avoided. In this concern, the fittings of both the pin and box threaded portions must be of a higher quality. In other words, the threaded portion's axial distance should be the same for both pin and box sides so that all the box threads are mounted with the pin threads.

Footnotes

Acknowledgements

The authors would like to thank Prof. David Taplin, UK/Canada for the deep and fruitful discussion.

References

ASM Handbook, Failure Analysis and Prevention, Vol. 11, ASM International, Materials Park, OH, 1996.

Metals Handbook, 9th edn, Failure Analysis and Prevention, Vol. 11, ASM, 1996.

Barer

R.D.

and Peters

B.F.

, Why Metals Fail, 6th edn, Gordon and Breach Science Publishers, New York, 1991.

Brooks

C.R.

and Choudhury

, Metallurgical Failure Analysis, McGraw-Hill, New York, 1993.

El-Batahgy

and Fathy

, Failure of a column reboiler in a MEG recovery unit, Materials Performance8 (2008), 2–5.

El-Batahgy

and Fathy

, Fatigue failure of thermowells in feed gas supply downstream pipeline at a natural gas production plant, Case Studies in Engineering Failure Analysis1 (2013), 79–84.

El-Batahgy

and Wheeler

, Fatigue failure of outlet pipe work of flare system at natural gas processing plant, in: 13th International Conference on Fracture (ICF13), 16–21 June, 2013, Beijing, China.

El-Batahgy

and Ageza

, Wear-fatigue failure of flash gas compressor at natural gas production plant, in: The 62nd Annual Assembly of International Institute of Welding (IIW), 12–18 July, 2009, Singapore.