Engineering Design & Analysis, Ltd.

Copyright © 2024 All Rights Reserved

  Fatigue Analysis of Pumpjack Pitman Arm


Pumpjacks
Pumpjacks for oilfield service are designed to document API 11E Pumping Units. The standard states the requirements and guidelines for the design and rating of beam pumping units for use in oil & gas field service. Portions of the standard are very detailed, other portions leave it to the discretion of the designer to consider and follow through.

The determination of structural loads from gearbox output shaft to attachment of the equalizer bar to walking beam can be easily determined but the impact of cyclic loading is routinely overlooked leading to loss in reliability and fatigue failure.

Design for Fatigue in PumpJacks
API 11E does not explicitly address fatigue in specific structural members which transmit torque from the gearbox through the pitman arm linkage assembly and through the walking beam to enable the reciprocating motion driving the subsurface piston pump.

According to the standard, the design of structural members is based on simple stresses. The standard advises to consider stress risers and concentration factors when stress risers occur. Hence, the OEM will only provide simplified calculations using an allowable stress based on a fraction of material yield strength, specifically 0.3·Sy. This appears to be generous as limits for allowable stresses for pressure vessels are in the region of 2/3·Sy.

However, there is a unique distinction between pumpjack operation and many pressure vessels (in refinery service), pumpjacks experience millions of cycles of loading; many pressure vessels only experience a few hundred cycles over their lifetime. Hence, pumpjack design needs to always account for cyclic loading. Since pumpjacks of some design have been in use since the 18th century, it would be assumed that design procedures have been well documented. As with all things done "modern", this means the drafting and design tools often are used by "software" operators rather than experienced personnel; this leaves designs that are designed by the software and not the designer.

Figure 1    A Typical Oil Field Pumping Unit
API 11E pumpjack


The sketch to the left shows a typical pumpjack assembly and the two Pitman Arms (10), each connecting the gearbox output shaft to the equalizer bar (20).

In the case being presented, fatigue failures occurred at the upper connecting joint between Pitman and equalizer bar.

The origin of the failure was at the fillet weld attachment of the structural beam to bracket connector. (This is not clear in the graphic).

While many Owner's may react and demand that FEA analyses be provided, these may not provide the needed information if the designers are not versed in fatigue analysis and is compounded by designers and suppliers appearing to be averse to making this analysis.

While many pumpjacks operate for decades, it appears modern-day design treatments rely on software-based designs that likely include "optimized" designs which rely on FEA optimization routines based on static design principles.

The result may be failure well before any reasonable operating life expectancy, say 25 years. A pumpjack battery of 24 or more pumpjacks that fails prematurely can be extremely costly in terms of structural repair and production loss.

Figure 2    Premature Fatigue Failure of a Pumpjack
Fatigue Failure


Figure 2 depicts a fatigue analysis based on manual, closed-form calculations for a specific pumpjack design which failed after only a few couple of years in medium duty rating service.

A "bounding" approach was used to define the possile range of expected life. The closed-form calculation depicts the bounding determinations; since the total cycle count intersected the lower bound estimate within a "good" range i.e., -1· σ to μ·Nf, the lower bound determination provides confidence that the unit was underdesigned for fatigue loading. The upper bound curve also provides information that informs a possible alteration to the equipment and a deterministic opportunity to ensure service life to a target value such as 50 or more years of service.

If you are interested in these real examples of applying engineering to the evaluation of service integrity and reliability issues, contact us for more details.

Please feel free to contact the author with your critique or suggestions for adding value to our tech briefs.