Resumo:
Among the artificial lift methods for oil extraction, Sucker Rod Pumping is one of the
most widely used worldwide. In this system, a volume of fluid is raised to the surface
through an alternating motion maintained by an induction motor associated with a me-
chanical structure. Downhole conditions pose significant challenges in beam pumping
systems, representing a concern for oilfield professionals. Although effective, especially
for shallow wells (less than 300 meters), this pumping system requires periodic analysis
of dynamometric cards, gross production tests, and monitoring of the dynamic fluid le-
vel in the annulus of the well between the production tubing and the casing to diagnose
failures and optimize the process. However, obtaining these cards often relies on invasive
instrumentation, increasing costs and system complexity. This dependence on field ins-
trumentation can be problematic, as any sensor element failure can result in pumping unit
shutdown, negatively impacting production. To address this issue, studies have sought
alternative or redundant field measurement system solutions to minimize the economic
impacts associated with pumping unit shutdowns and primary measurement element re-
placement. This study aims to implement an estimator for surface dynamometer cards
using measurements associated with the induction motor power supply bus and data from
the beam pumping system’s mechanical structure. The research presents a methodology
that integrates physical measurements of voltage, current, slip, and velocity in the in-
duction motor, serving as input variables for a state observer of this motor. Estimation
does not occur in real-time but is recorded in an electronic file for subsequent compu-
tational processing. The output of this observer, combined with a model of the motor’s
equivalent circuit, is used to infer the torque produced on the motor shaft. The velocity
measured at the motor shaft is directly linked to the angular velocity of the pumping
unit’s crankshaft due to the interconnection between pulleys and belts involved in this
process. The crankshaft’s angular position angle can be obtained from the angular speed;
this angle helps determine the polished rod position. To calculate the polished rod load,
an estimator needs to be fed with input variables, including the torque on the motor shaft
and the crankshaft angle. The relationship between load and polished rod position results
in the composition of the dynamometer card. The results present dynamometer cards un-
der various simulated operational conditions in an experimental oil well, such as normal
operation, operation with gas presence in the well, operation with leakage in the stuffing
box valve, and operation with insufficient fluid supply, which align more closely with real
conditions. This practical application offers an alternative method for acquiring surface
dynamometer cards, which can function as a less costly and complementary approach in
fault diagnosis compared to current methods, contributing to maintenance decisions for
the pumping unit.
