Assessing the Potential of Advanced Sensor Technologies and IoT in Predicting Downhole Drilling Issues

Fast detection of downhole drilling problems and drilling fluid property changes remains one of the critical
challenges requiring further attention from the drilling industry. One potential solution to this issue might
be the use of advanced sensor technologies installed at the bell nipple and the utilization of the Internet of
Things (IoT) as a gateway to transfer data between the rig site and the Real-Time Drilling Operations Centre
(RT-DOC). In this context, this paper discusses the development of a large-scale laboratory prototype that
mimics the bell nipple in field applications and presents the results of experiments conducted to evaluate
the potential of the sensors installed on the prototype to detect downhole problems.
Four sensors have been integrated into the prototype: an in-line density sensor, an in-line viscosity sensor,
a non-intrusive level sensor, and a Doppler ultrasonic velocity sensor. Real-time data recorded by these
sensors were collected and transmitted via the cloud to a data analytical station. Multiple experiments were
conducted using the developed prototype. The main aims of these experiments are to assess the accuracy of
the sensors, evaluate the response time of data transmission, evaluate the sensors response to rapid changes,
and their ability to detect downhole problems, particularly kicks and losses.
The main findings of the work can be summarized as follows: The five-layer system developed for
data acquisition and analysis, leveraging IoT technology, exhibited seamless functionality in continuously
and autonomously collecting sensor data, processing it, and transmitting it to a real-time monitoring
and analysis framework without human intervention. Furthermore, integrating fluid level measurements
alongside complementary indicators such as delta flow demonstrated promise in effectively detecting kick
and loss symptoms. The evaluated in-line density and viscosity sensors displayed exceptional accuracy and
responsiveness, enabling real-time monitoring of fluid properties and facilitating the prompt detection of
rapid changes. This capability offers significant advantages to rig-side mud engineers, allowing for swift
identification of variations compared to the traditional hourly measurement approach reliant on marsh funnel
viscosity and mud balance. While the ultrasonic velocity sensor exhibited satisfactory accuracy, its utility in
detecting loss and kick events cannot be fully realized using the delta flow method. However, better results
in detecting the loss and kick events are achieved by using advanced incident classification algorithms.