The Oil and Gas operators today are progressively focused on maximizing efficiencies in land drilling operations in unpredictable resource plays. Many in the business and industry are turning to “factory drilling” concepts to enhance performance and thus reduce nonproductive time. The ultimate goal of this new way of conducting drilling operations is to minimize the construction times of wellbore, thereby minimizing total well cost, through real-time data, and smooth integration of component technologies, and multitask operations.
Even though the drilling industry has recognized that historic data can be used to improve and optimize the drilling operations, methods to systematically exploit historic data for this purpose are relatively scarce. Typically, vast amounts of data are collected before and during the drilling process. In the past, it has been impossible to account for all of the data when performing optimization techniques. What is needed, therefore, is a method for remotely performing drilling optimization methods based on the available data. Hence, a need for an automated solution which will help the drilling team in the backend office to make a decision while watching and comparing patterns in the current well, similar to the ones from the historical data of the previously drilled wells.
The push for speed
While the continuous nature of this drilling procedure helps remove and address wellbore-related nonproductive time issues and also takes into account the optimal and faster drilled wells through continuous learning, the push for speed is placing greater stress on drilling equipment, especially downhole tools. The more organizations turn to these drilling principles, and as the total volume of wells drilled in a factory mode continues to rise, supply chain concerns are demanding steps to alleviate any damage to downhole equipment/parts components such as mud motors.
An automated solution is to be developed for wells drilled for identifying efficient drilling, through which the drilling parameter settings weight on bit (WOB), drill string or drill bit revolutions per minute (RPM), differential pressure (∆P), and pump rate (total pump output; TPO) settings linked to the highly efficient drilling epochs are extracted from the historical well data and stored in a remote data store. These settings are to be formation- and depth-aligned to the new well to be drilled in real time, and they are smoothed and displayed on the electronic drilling recorder (EDR) in real-time as drilling parameter roadmaps( source: Use of Historic data to improve drilling efficiency: A pattern recognition method and trial results, SPE -178901-MS ,by Sean D Kristjansson, Adam Neudfeldt, Stephen W. Lai and Julian Wang,March 2016).
Pattern Recognition and Extraction
The operators are thus turning to predictive analytics technology to address these issues, such as case-based reasoning (CBR) which will provide an effective solution for mitigating risks and reducing nonproductive time.
Case Based Reasoning uses integrated machine learning to accomplish real-time computerized intelligent behavior. Pattern recognition is designed in a manner in which the data streams are examined to find meaningful patterns. The pattern recognition engine is designed to understand when the surface and downhole parameters from a real-time data stream are trending outside an operating window defined by a rule-based structure developed through a series of statistical algorithms. It uses the past experiences to understand, recognize and solve new problems, and when applied in real time to the decision-support workflow while drilling, it creates a productive and powerful combination of automation and experience to help operators and administrators recognize early indications of bigger potential problems or risks and take appropriate and suitable measures sooner. The software designed would have the capability to automatically recall and present relevant cases from its knowledge base of drilling experience and best drilling practices, thus providing the drilling engineer in the backend office with recommendations for corrective action when a potential issue has been first recognized.
The potential of rigs, component systems and downhole tools are developing at an incredible pace to optimize drilling performance. The technologies and software will help enabling improved drilling efficiencies in onshore and include ever more adept levels of automation and real-time data management and analysis.
Mapping occurrences of each Pattern
Once the pattern recognition engine has processed real-time drilling data and identified events, the next process, will be the engine comparing current event patterns to the patterns captured in historical cases. This oil and gas well logging determined by a combination of other characteristics, which includes tool string geometry, lithology information, hole geometry, and mud system information provides a system for collecting real time and historic data concerning the drilling operations of an oil and gas well which are then processed and correlated to match real time data at a particular elevation with historic data and then displayed on a screen for immediate utilization in the drilling operation. The data are processed and correlated at preselected increments of well elevation and provide important information.
The knowledge and expertise captured and stored in the Case Based Reasoning process provides support to the engineers and decision makers across an Oil and Gas organization. Fast interpretation and fast access to recommendations based on best practices and lessons learned makes it possible to avoid drilling problems, reduce risks and nonproductive time thereby increasing safety and drilling efficiency.
The process would provide monetary benefits to the Oil and Gas Operators in view of achieving the following objectives
Increase in reliability of Equipment
Reliability of equipment determines the life of equipment (example; drill bit). If proper data is present of the amount of drill bit dulling/wearing out that has already occurred in historical datasets and how it has had an effect on the optimized parameters. Then amount of drill bit dulling that has occurred may be estimated based on current well data for those portions of the formation that have already been drilled, as well as data related to such things as WOB, TOB, RPM, mud flow rate, drilling pressure, and data related to measurements of the drill bit properties while drilling. A matched pattern/result obtained allows timely troubleshooting of the problem. And hence reliable operations can be expected.
Major loss a company faces is at the time of shutdown of drilling, due to breaking of a drill bit. This would lead to the next procedure of removal of the drill string and changing the drill bit which has a very high cost. Moreover breakdown cost is much higher than the preventive maintenance cost. With the help of Pattern recognition, life of the drill bit can be increased. Prediction or forecast can be done with proper data.
Improves Process efficiency
If the parameter settings for the formations and well depth alignment for the new well is displayed on the Electronic Drilling Recorder (EDR) of the back office drilling teams, in Real time as the drilling Parameter Roadmaps, distinctive levels of efficiency can be achieved.
Oil and Gas organizations with well-established drilling operations looking for continuous improvement are starting to realize the power of this approach, and are driving the change in using predictive analytics to stay ahead of downhole tool damage or outright failures. Other organizations will also follow suit in coming years to recognize the benefits of this enabling technology, as high-efficiency drilling techniques become the prevalent process to reduce time-based costs, or in any applications where downhole tool damage is a prevailing issue.
- ‘Unconventional Data-Driven Methodologies Forecast Performance In Unconventional Oil and Gas Reservoirs’ by Keith R. Holdaway, Louis Fabbi, and Dan Lozie, SAS Institute Inc (Paper 1910-2015)
2.’Technologies to improve drilling Efficiency and Reduce Costs’ ,OG21 Technology group, published on October 15,2014.
3.’Use of Historic data to improve drilling efficiency: A pattern recognition method and trial results’, SPE -178901-MS ,by Sean D Kristjansson, Adam Neudfeldt, Stephen W. Lai and Julian Wang,March 2016
4.’System for optimizing drilling in Real Time’, US 7142986 B2, Nov 28,2006
Try deep learning using MATLAB