Where Sucker Rods Break: Body, Pin or Coupling, and What Each Tells You

Where do sucker rods break?

Sucker rods break in one of three places: the rod body, the pin (including the forged upset behind it) or the coupling. Each location points to a different cause. Body breaks usually trace to corrosion pits, handling damage or bending; pin breaks to make-up; coupling breaks to looseness, wear against the tubing or corrosion. ¹⁵⁶

A sucker rod is a steel bar with a forged upset and a threaded pin at each end, and two rods are joined by a rod coupling. Every joint has three candidate failure points.

By design, the body should be where most failures happen. The pin undercut and the coupling have more cross-section than the rod body, and correct make-up locks a prestress into the pin, so in theory a string should fail in the body. ⁵ In practice many operators fight connection failures instead, which is why the location of a break is worth recording.

This guide covers where a rod parting happens. For the causes themselves see why sucker rods part; for a joint that unscrewed instead of breaking, see sucker rod back-off.

Body breaks: corrosion, damage and bending

A rod body break usually starts at a stress raiser on the surface: a corrosion pit, a nick, a wrench cut or a bend. In a 1975 field census of rod parts, corrosion and bending accounted for all but one of the body failures. Wear from rubbing the tubing is the third body pattern. ¹⁶⁹

One 1975 study inspected rod parts from a six-county area of West Texas and southeast New Mexico. Of its 92 body failures, 54 were due to corrosion, 37 to bending and 1 to damage. ¹ A pit acts as a stress raiser; a bent rod has higher stress on the concave side, where a fatigue crack eventually appears. Bends came mostly from handling and worn rod elevators. ¹

Any nick, scratch or wrench cut is a stress riser too; mechanical damage ranks second only to corrosion as a cause of failure. ⁶⁹ A wear break looks different: a flat worn on one side, usually running several feet in both directions from the break. ⁶

One 2017 operator database found its body failures were corrosion, most likely between 1,000 and 4,000 ft, at an average of 250 days to fail. ⁴ That is one asset's pattern, not a rule.

Pin and upset breaks: a make-up problem first

A pin break is usually a connection that lost its clamp. The number one cause of pin failures is looseness under load, from under-makeup, damaged shoulders or operating loads. Too much make-up breaks pins too. Breaks at the upset point to elevator damage or bending concentrated just outside the stiff connection. ¹³⁷⁹¹⁰

When the pin shoulder and coupling face separate, the pin rocks in the coupling, and that eccentric loading acts as a stress raiser at the last full thread root of the pin. ¹ In the 1975 study, 24 of 25 pin failures were due to looseness. ¹ A metallurgical study of failed field returns found the same spot: in every pin, failure started in the first thread fully engaged in the coupling, and none started in the shoulder region. ⁸

Insufficient make-up, a damaged shoulder face, or wide load ranges from pounding fluid can all separate the faces. ¹ Over-torque is the opposite mistake: a pin made up close to yield yields further under well load, loses its prestress and goes loose. ⁶ A field in southern Argentina that fought pin failures for 20 years concluded that its remaining pin cracks came from too much torque, not too little. ³

Before the API adopted the undercut pin, joint trouble was largely broken pins; by 1968 one author wrote that undercut pins had greatly reduced pin failure, "if not practically eliminated" it. ⁷ Make-up by circumferential displacement is covered in the back-off guide.

The upset, the forged transition from body to pin, has two culprits of its own: worn or misaligned elevators damage the upset taper, ⁹ and buckling concentrates bending stress just outside the stiffer connection. ¹⁰

Coupling breaks: looseness, wear and corrosion

Couplings break from the same looseness that breaks pins, from corrosion, and from wear. A coupling is the widest part of the string, so it is the first part to touch and rub the tubing. A coupling worn flat on one side points to rod-on-tubing contact from buckling or a crooked hole. ¹⁴⁶

In the 1975 study, couplings were the largest group: 111 failures, of which 56 were due to corrosion, 40 to looseness, 10 to wear and 5 to damage. ¹ A loose joint cracks the coupling opposite the last full pin thread, ¹ and in failed field returns those cracks grew from the inside out. ⁸

Wear is the coupling's own failure mode. Couplings show abrasion first because they have the largest diameter, ⁶ and many coupling failures happen in crooked holes, where the coupling is pulled against the tubing. ¹ In the 2017 database the primary coupling failure mode was wear from tubing contact, and 74% of tubing failures traced to contact with a rod coupling on an unguided section. ⁴

Reading the fracture face

Most rod breaks are fatigue breaks: a flat face at right angles to the rod, a smooth zone with curved "oyster shell" or beach marks spreading from a pit or nick, and a rough final zone opposite it. A necked-down, stretched break means tensile overload, which is rare. ¹⁶⁸⁹¹¹

A classic fatigue face has three features: ⁶

1. The origin. A small nick or pit at the surface, at the eye of the curved marks. It started the crack.
2. The beach marks. The "oyster shell" bands, formed by pauses in crack growth over many strokes.
3. The final zone. Jagged, rough edges opposite, where the remaining steel could no longer carry the load.

The rough zone makes the steel look brittle, and such breaks are sometimes wrongly called crystallization. ⁶ On pins and couplings the fatigue zone is often rubbed smooth by the crack faces working against each other. ⁸

A tensile overload looks different. The load concentrates at the weakest point in the string, the rod necks down around its circumference, and it breaks where the cross-section is reduced. ⁹ A stuck pump is the usual cause, and it is rare; ¹¹ the 1975 field study found no tensile failures at all. ¹

For a cause-by-cause table of break faces, see why sucker rods part. A connection with intact threads and no fracture at all is a back-off, not a break.

Top or bottom: where in the string breaks cluster

A rod can part anywhere, but the top and the bottom fail for different reasons. The top of each taper section carries that section's highest stress. The bottom can go into compression on the downstroke and buckle, which bends rods and connections and wears couplings against the tubing. ^78101112

A rod part "can occur anywhere along the length of the sucker-rod string," one diagnostics patent says. ¹¹ But the load is not the same all the way down.

The top of each section. The upper end of a rod string must support more weight than the lower end. ¹² That is why strings are tapered, and the preferred taper keeps the stress at the top of each section about equal; tapers that run the lower, smaller sections at maximum allowable stress are a common mistake. ⁷ For taper design, see sucker rod grades and string design.

The bottom of the string. On the downstroke, forces at the pump can push the bottom of the string into compression, and a compressed rod bows into the tubing. A 1993 study of more than 3,000 wells noted that when a string buckles, most commonly in the 3/4 in rods, the connections take a bending moment and the tubing takes friction and banging, which can also unscrew a connection. ⁸ A 2024 paper lists deep body failures from bending and connection failures from vibration and shock among the harm buckling does. ¹⁰ Pounding fluid bends rods as far as the tubing will permit. ⁷ As one 1962 author put it, operating rods in compression causes "large unpredictable and disastrous stresses." ²

A deep part is also harder to spot: little rod weight is lost, so the dynamometer card changes less. ¹¹ For why the bottom buckles, see sucker rod buckling.

Body, pin or coupling: what the published splits say

Published splits disagree. A 1975 field study found more coupling failures than body failures and few pin failures. A 1962 laboratory survey found pins dominating in high-tensile rods. A 1985 field history reported pins at half of all rod failures. A 2017 operator database put couplings at about a quarter of rod-system failures. ¹²³⁴

The McGee paper prints 46.4% for body failures, but that figure doesn't match its own count; the counts (111, 92 and 25 of 228) are the safer read. ¹

These numbers can't be averaged into one industry split. They measure different things in different eras:

• Rod design changed. The 1962 survey came just as the API adopted undercut pins and rolled threads; ² by 1968 undercut pins had greatly reduced pin failure. ⁷ A pin-heavy split from that era describes the rods of that era.
• The sample decides the split. A lab sees the failures somebody sent in; a field census like the 1975 study sees every part in an area; the 2017 database groups pin and body together.
• Local practice decides it too. The Argentine field's pins were made up on threads already damaged on earlier jobs. After it changed its make-up procedure, pin failures fell from a peak of 25 a month to fewer than 5 within a year. ³
• Well conditions shift it. Corrosive fluids drive body and coupling pits; crooked holes and buckling drive coupling wear; pounding fluid and poor make-up drive pin and coupling looseness. ¹⁶⁷

What the sources agree on is the causes inside each location: corrosion and bending in the body, looseness at the pin, and looseness, corrosion and wear at the coupling. ¹ And a correctly made-up connection should be stronger than the body it joins. ⁵ If your connections fail more often than your rod bodies, the make-up and the downhole loads are the first places to look.

How to log a break so the pattern shows

Record every rod part with its depth, rod size and taper, the failed member (body, upset, pin or coupling), what the face looks like, and the cause. Keep the broken piece. A record that says only body, pin or coupling, with no cause, can't tell you what to fix. ¹⁴

The 1975 study started because field records held the depth and the failed member but no cause, so remedies couldn't be tailored to the problem. ¹ With root causes logged, the 2017 operator could see that its coupling failures were wear and its body failures corrosion in a known depth band. ⁴ A practical minimum for each break:

1. Where. Depth, rod size, taper section, and distance from the nearest taper change.
2. What part. Body, upset, pin or coupling. For a pin, which thread; for a body, any flat or bend nearby.
3. What face. Fatigue (flat, beach marks, origin) or overload (necked down). Photograph the origin.
4. What else. Pits, wear flats, wet joints, a coupling worn on one side, and the dynamometer card before the failure.

Patterns across a field (same depth, same taper, same part) matter more than any single break. Repeated breaks or wear near the bottom of the string point to compression; see rod failure prevention.

Where SRABS fits

Location tells you which fix to look at. Corrosion programs, handling and make-up stay with your rod program; no pump fixes those. SRABS targets the breaks that cluster at the bottom of the string, where compression bends rods and drives couplings into the tubing.

The SRABS Pump is "a drop-in anti-buckling rod pump that keeps the string in tension where it would otherwise buckle." For the approach see how it works and the system. To estimate what fewer rod failures would mean for your fleet, run the ROI calculator. Its default 100-well scenario is modelled, not a field result.

The bottom line

Where a rod breaks tells you what to fix. Body breaks point to corrosion, handling and bending; pin breaks to make-up; coupling breaks to looseness and wear. Read the face, log the depth, and look for the pattern. If the breaks and the wear keep landing in the bottom of the string, the cause is compression.

Test that on one well.

FAQ

Sources

1. McGee, J. M., "Evidence of the Cause of Sucker Rod Failures," Southwestern Petroleum Short Course, 1975. https://www.swpshortcourse.org/paper/evidence-cause-sucker-rod-failures
2. Meripol, R. B., "Sucker Rod Failures Can Be Reduced," Southwestern Petroleum Short Course, 1962. https://www.swpshortcourse.org/paper/sucker-rod-failures-can-be-reduced
3. Giacomel, R., "A 20 Year Fight Against Sucker Rod Pin Failures," Southwestern Petroleum Short Course, 1985. https://www.swpshortcourse.org/paper/20-year-fight-against-sucker-rod-pin-failures
4. Clarke, F. W., "Sucker Rod Pump Root Cause Failure Analysis," Southwestern Petroleum Short Course, 2017. https://www.swpshortcourse.org/paper/2017001-sucker-rod-pump-root-cause-failure-analysis
5. Hein, N. W., Jr., Malone, S., "Sucker Rod Connection Failures: Why They Occur and What to Do About Them," Southwestern Petroleum Short Course, 2011. https://www.swpshortcourse.org/paper/sucker-rod-connection-failures-why-they-occur-what-do-about-them
6. Morris, R. A., "Sucker Rod Failure Analysis," Southwestern Petroleum Short Course, 1971. https://www.swpshortcourse.org/paper/sucker-rod-failure-analysis
7. Hardy, A. A., "Common Causes of Failure in Sucker Rod Strings," Southwestern Petroleum Short Course, 1968. https://www.swpshortcourse.org/paper/common-cause-failure-sucker-rod-strings
8. Waggoner, J. R., Buchheit, R. G., "Analysis of Sucker Rod and Sinkerbar Failures," Southwestern Petroleum Short Course, 1993. https://www.swpshortcourse.org/paper/analysis-sucker-rod-sinkerbar-failures
9. Knous, J., Bava, B., "Why Inspect Sucker Rods," Southwestern Petroleum Short Course, 2010. https://www.swpshortcourse.org/paper/reasons-inspect-sucker-rods
10. Anderson, J., Oliva, E., "Sinker Section Design to Reduce Buckling Related Failures," Southwestern Petroleum Short Course, 2024. https://www.swpshortcourse.org/paper/2024034-sinker-section-design-reduce-buckling-related-failures
11. US Patent 4,509,901, "Method and apparatus for detecting problems in sucker-rod well pumps" (1985). https://patents.google.com/patent/US4509901
12. US Patent 8,113,277, "Sucker rod end fittings and method of using same" (2012). https://patents.google.com/patent/US8113277