Tests pinpoint sucker-rod failures

F. Elshawesh, A. Elhoud, E. Elagdel
Petroleum Research Center
Tripoli, Libya

The 7 to 8-in. rod had failed after a few months of service while the 1-in. rod failed after 1 year. Both rods had been used in a sweet-oil environment.

Both rods failed by corrosion fatigue because of repeated loads during operations. Pitting because of the presence of chloride ions (Cl-) and carbon dioxide (CO2) was initiated on the rod surface, which in turn acted as a crack origin from which the fatigue crack initiated and propagated during operations.

The pitting was on the external surface. These pits were large and penetrated through the rod cross-section.

Fatigue cracking is initiated at the bottom of the pit where high stress concentration is expected and propagated because the rods were subjected to the alternating stresses during operation.

The extent of the fatigue crack varied in the two examined rods because of the difference in the rod heat treatment and microstructure.

Fatigue

While pumping an oil well, sucker rods are subjected to cyclic loading. Therefore, they can fail by fatigue or corrosion fatigue that originates from localized corrosion such as pitting on the rod surface.

The principal cause of sucker-rod body failure is overstressing. Fatigue failure is affected by the maximum stress, range of stresses, and the number of reversals or cycles. Fig. 1 [36317 bytes] shows a typical sucker rod installation.

For sucker rods, these factors are represented by peak polished-rod load, range of loading, and speed.

Nevertheless, the safe working stress may be influenced by the system's shock or vibration loads.¹

Most engineering components fail from fatigue when they are subjected to repeated or cyclic loading. Fatigue cracks generally initiate at a surface attacked by corrosion such as pitting, crevices, etc., or at a discontinuity which usually acts to raise stress such as notches or inclusions. However, failures may in some cases initiate at a subsurface defect such as a casting defect, inclusion, or quench or hydrogen-assisted crack.^1-4

The nominal stresses at which fatigue cracks can form may well be below the material yield strength, but on a microscale the local stresses may be much higher as a result of the stress concentrations around defects and inclusions. This leads to highly localized plastic deformation that is not apparent on a normal engineering scale.²⁵

Macroscopically, a fatigue fracture is brittle in nature and can be considered as a special type of brittle fracture. Fracture surfaces of components with fatigue failures usually exhibited three distinct areas: crack initiation, crack propagation, and final rupture. These occur when the crack reaches a critical length at which it propagates rapidly to cause sudden failure during a single tensile-load application (overload).3

The apparent ductility or brittleness of the overload fracture will vary depending on the strength, ductility, and toughness of the material, and the temperature. Macroscopically, there will be no evidence of necking associated with the fracture.

Visual examination

Fig. 2 shows the two failed rods on which a detailed visual examination was carried out. The examination covered the fracture surface as well as the external rod surface.

The sucker rods were fractured at 90° and indicated a brittle fracture. Three distinct zones can be easily differentiated.

The fatigue fracture initiates at the external surface from the severe localized corrosion known as pitting (Fig. 3 ).

These pits seemed to have been initiated because of the presence of chlorides and CO2 in the producing streams.

Macroscopic examination

Macroscopic examination of the fracture surfaces revealed three distinct zones: a crack-initiation zone which includes the pits on the external surface, a second zone known as the crack propagation zone, and a third which is the final rupture zone (Fig. 4 and Fig. 5 ).

The fracture surface in the second zone revealed radial marks indicating the crack origin and direction.

The extent of the second zone was found to be different in the two fracture surfaces. This may be due to differences in the fracture toughness because of variances in the rod microstructure.

The third zone is the final fracture which shows shear lips, an indication of a ductile fracture (overload).

Microscopic examination

Samples from the failed rods were metallurgically examined to determine the microstructure.

The 7/8-in. rod microstructure was not in accordance with the specifications and consisted of white areas of ferrite and dark areas of pearlite instead of a martensitic structure (Fig. 6 ).

This type of microstructure usually is obtained by heating in a furnace followed by air cooling and has a low hardness (Table 1 [9543 bytes]).

In contrast, the microstructure of the 1-in. rod was in accordance with specifications and consisted of white areas of ferrite and dark areas of martensite (Fig. 7 ). This microstructure has a high hardness because of martensite precipitation caused by fast cooling after the high temperature oil bath followed by tempering at low temperatures.

Rod properties

Table 1 shows that the 1-in. rod has a higher hardness compared to the 7/8-in. rod (348 HB compared to 293 HB). The hardness variation may be attributed to the difference in the heat treatment.

Both rods have the same chemical composition and meet specifications (Table 1). This composition indicates that the rods are of alloy steel.

The carbon increases the strength and hardness, and it lowers the ductility and corrosion resistance of the alloy. The addition of chromium tends to increase the hardness and greatly enhances the corrosion resistance. The addition of molybdenum increases the hardness.

Irrespective of strength criteria, steel should be chosen for corrosion resistance in accordance with the following published data.⁵⁶

• Carbon steel for a mildly corrosive environment

• Ni-Mo steel for an H2S environment

• Ni-Cr steel for a strongly corrosive brine environment.

Future operations

To prevent localized corrosion and hence improve the fatigue strength, a high-alloyed steel such as 13 Cr would be better in the operating environment.

Quality control is also important, as seen from one of the rods not meeting the required specifications. Quality control should include chemical composition, microstructure, and mechanical properties. Heat treatment needs to be carefully monitored.

Organic corrosion inhibitors, batched down the annulus, could also prevent pitting which started the failure.

References

1. Zara, J., and Doherty, W.T., Practical Petroleum Handbook, 5th edition, Gulf Publishing Co., Houston.

2. ASME Handbook, Failure Analysis and Prevention, 8th Edition, Vol. 1, ASME International, Materials Park, Ohio, 1986.

3. Siebert, O.W., Failure Anaysis Process Industries: Corrosion Theory and Practice, Moniz, B.J., and Pollack, I.W. editors, Nace Publications, 1986.

4. Lefebvre, B., et al, "Behavior of Carbon Steel and Chromium Steels in CO2 Environments, Advances in CO corrosion," Symposium on Corrosion by CO2 in the oil and Gas Industry, Corrosion 84, NACE, 1985.

5. Day, J.J., and Byrd, J.B., Beam pumping: design and analysis, The Technology of Artificial Lift, Brown, K.E. editor, PennWell Books, Tulsa.

6. Szilas, A.P., Production and Transportation of Oil and Gas, Elsevier Scientific Publishing Co., 1975.

The Authors