The offshore drilling industry has maximized utilization, stressing personnel and equipment to improve productivity. Component failure causing unplanned shutdowns is a major problem. Multiple technologies including sensors, data collection and analysis with expert software mitigates the risk of costly unplanned downtime.

Condition Based Monitoring (CBM) continuously collects relevant, near real-time and periodic, data from critical equipment components. It applies predictive analytics using decision tree analysis to build models depicting in advance the potential for failure and helps establish root cause. The purpose is to increase uptime, improve performance and reduce maintenance costs on offshore drilling rigs.

A Confluence of technologies are used, including:

  • -

    Sensor engineering to extract relevant data

  • -

    Data collection/processing

  • -

    Expert software for Predictive, Decision Tree modeling

  • -

    Actionable information distribution through internet/intranet

It is a proactive approach replacing reactive or calendar/task based systems. It is proven and well established in other industries and recognized as a "best practice".

There are many challenges in the driller environment.

Offshore drilling is remote, widely dynamic operationally, manned with over burdened personnel. Machine condition data does not exist. Some operational data is collected but it is not analyzed and trended for the purpose of eliminating downtime. The CBM/PdM technology reduces downtime with limited personnel involvement.


Sensor technology exists that can be applied to equipment components such as bearing sets. Detailed data collected from the lubricants can be trended and analyzed. Multiple variables (predictors) are used to establish equipment profiles, define normalcy and measure variant, deviation and trends. Regression analysis can produce decision trees forming predictive models about potential problems, well in advance of any incidents.

How a lubricant/bearing decision tree works -

Sensors are installed and collect relevant data from the bearing sets in near real-time with time/date tags. Data (content ppm, viscosity, temperature) is collected from the lubricants and is tagged, analyzed and correlated with the bearing data.

The data is fed into the predictive analytics engine (software), which performs a complex analysis and builds a decision tree that models the data.

A great advantage of decision trees over classical regression and neural networks is they are easy to interpret into actionable items with a clear understanding of how and why a downtime incident is avoided.

You can access this article if you purchase or spend a download.