Knut Steinar Bjorkevoll

Successful Use of Real Time Dynamic Flow Modelling to Control a Very Challenging Managed Pressure Drilling Operation in the North Sea

An advanced dynamic flow and temperature model was used to optimize and control MPD operations in real time on the Gullfaks field in the North Sea. The well to be drilled only had a 7 bar window between the pore and fracture pressure according to prognosis. However, drilling objectives were eventually fulfilled aided by very accurate downhole pressure control. This paper addresses the model specific challenges, analyzes the differences between model calculations and downhole pressure data, and discusses how to bring hydraulic modelling further in accordance with future operational needs. Challenges related to how to tune the system efficiently and accurately, data quality issues, displacement operations, etc., are described and enlightened by downhole memory data made available when the string was back on surface. Ideas on how to build a more robust and easy to use system without sacrificing the advantages of having a high fidelity model in the real time loop are discussed. The experience and ideas described contribute to the development of a very accurate and reliable MPD system, which is capable of automatic pressure control during the whole sequence of drilling, tripping, circulation, displacements etc. An important goal for the future will be to reduce the offshore crew dedicated to the modelling function to a minimum, with the provision of onshore support during operations.

Statistical Method for Detection of Poor Hole Cleaning and Stuck Pipe

The drilling of an oil or gas well is an expensive undertaking. Hence, it is not surprising that mistakes and accidents during drilling incur a high cost. Accidents could result in the loss of expensive equipment and subsequent delays setting back the operation for days or weeks and thus running up large bills on rig-time and personnel hours. Some types of accidents also pose a risk to the personnel or the environment. In this dissertation we study alarm systems which could give the driller an early warning of upcoming problems, and thus provide time to avoid these accidents. We explore alarm systems which combine advanced physical models of the well and drilling process with artificial intelligence and time series analysis. Finally, we determine the advantages as well as the challenges of this approach. It is our hope that this dissertation is accessible to both practitioners in machine learning and control engineering, as well as to petroleum engineers with a passing familiarity with machine learning. Hence this dissertation starts with a quick introduction to drilling problems and some terms from time series analysis and machine learning. We then briefly describe the theory of observer-based fault detection and isolation. Theories of supervisory control systems are also introduced, as these concern both the choice of algorithms and how AI-based alarm systems integrate with the rest of the operation. From chapter 6 and onward, the challenges to fault detection in drilling are discussed. We focus on clarifying what restrictions the available training data put on our choice of machine learning methods. In chapter 8 and 9, we propose ways to combine machine learning and observer-based fault detection. Experimental results are presented in chapter 10, before we end with concluding remarks in chapter 11. Our main conclusion, reflected in our experimental results, is that physical models and artificial intelligence can be combined to produce hybrid alarm systems that are better than what we could have achieved using these approaches separately. When using artificial intelligence we treat fault detection in drilling as a machine learning problem. In the course of our work we find that this problem domain differs in important respects from textbook examples of machine learning problems. Determining the distinctive characteristics of this problem domain is crucial in designing the alarm system. Drawing on examples from different fields we determine these characteristics and propose novel alarm system architectures that build on recent developments in machine learning.

Spotting A False Alarm. Integrating Experience And Real-Time Analysis With Artificial Intelligence

The drilling of an oil or gas well is an expensive undertaking. Hence, it is not surprising that mistakes and accidents during drilling incur a high cost. Accidents could result in the loss of expensive equipment and subsequent delays setting back the operation for days or weeks and thus running up large bills on rig-time and personnel hours. Some types of accidents also pose a risk to the personnel or the environment. In this dissertation we study alarm systems which could give the driller an early warning of upcoming problems, and thus provide time to avoid these accidents. We explore alarm systems which combine advanced physical models of the well and drilling process with artificial intelligence and time series analysis. Finally, we determine the advantages as well as the challenges of this approach. It is our hope that this dissertation is accessible to both practitioners in machine learning and control engineering, as well as to petroleum engineers with a passing familiarity with machine learning. Hence this dissertation starts with a quick introduction to drilling problems and some terms from time series analysis and machine learning. We then briefly describe the theory of observer-based fault detection and isolation. Theories of supervisory control systems are also introduced, as these concern both the choice of algorithms and how AI-based alarm systems integrate with the rest of the operation. From chapter 6 and onward, the challenges to fault detection in drilling are discussed. We focus on clarifying what restrictions the available training data put on our choice of machine learning methods. In chapter 8 and 9, we propose ways to combine machine learning and observer-based fault detection. Experimental results are presented in chapter 10, before we end with concluding remarks in chapter 11. Our main conclusion, reflected in our experimental results, is that physical models and artificial intelligence can be combined to produce hybrid alarm systems that are better than what we could have achieved using these approaches separately. When using artificial intelligence we treat fault detection in drilling as a machine learning problem. In the course of our work we find that this problem domain differs in important respects from textbook examples of machine learning problems. Determining the distinctive characteristics of this problem domain is crucial in designing the alarm system. Drawing on examples from different fields we determine these characteristics and propose novel alarm system architectures that build on recent developments in machine learning.

Automatic Real-Time Drilling Supervision, Simulation, 3D Visualization, and Diagnosis on Ekofisk

eDrilling is a new and innovative system for real time drilling simulation, 3D visualization and control from a remote drilling expert centre. The concept uses all available real time drilling data (surface and downhole) in combination with real time modeling to monitor and optimize the drilling process. This information is used to visualize the wellbore in 3D in real time. eDrilling has been implemented in an Onshore Drilling Center in Norway. The system is composed of the following elements, some of which are unique and ground-breaking: • An advanced and fast Integrated Drilling Simulator which is capable to model the different drilling sub-processes dynamically, and also the interaction between these sub-processes in real time. • Automatic quality check and corrections of drilling data; making them suitable for processing by computer models • Real time supervision methodology for the drilling process using time based drilling data as well as drilling models / the integrated drilling simulator • Methodology for diagnosis of the drilling state and conditions. This is obtained from comparing model predictions with measured data. • Advisory technology for more optimal drilling. • A Virtual Wellbore, with advanced visualization of the downhole process. • Data flow and computer infrastructure eDrilling has been implemented in an Onshore Drilling Center on Ekofisk in Norway. The system has been used on several drilling operations. Experiences from its use will be summarized and presented. This paper has main focus on utilization of an advanced flow model for real time supervision and control of ECD and ECD related effects. Introduction The southwestern part of the Norwegian continental shelf, called the Ekofisk Area, contains eleven major chalk fields. The Ekofisk field is the first and main discovery, discovered in 1969 and put on production in 1972. The fractured chalk reservoir lies at a depth of 9500 – 10700 feet and is approximately 11.2 x 5.4 kilometers in area, with production coming from two zones Ekofisk and Tor. It is one of the North Sea Giants with a STOIIP of 7 MMBO! Currently there are 4 fields in production, 4 fields abandoned with current production around 325,000 bbls per day of oil and 350 scf of gas per day. Water injection is currently used to maintain reservoir pressure, and approximately 900,000 bbls of water are injected each day. There are over 150 wells that have been drilled on the Ekofisk, and due to the complexity of the field, with its numerous faults and fracture networks, location of injected water, and pressure uncertainties, all result in well placement challenges. New wells are being drilled both as injectors and for production from newer facilities, and from jack-ups. Much of this drilling is supported from the Onshore Drilling Centre (ODC) located onshore about 280 Km from the field. It is connected to

e-Drilling: A System for Real-Time Drilling Simulation, 3D Visualization and Control

eDrilling is a new and innovative system for real time drilling simulation, 3D visualization and control from a remote drilling expert centre. The concept uses all available real time drilling data (surface and downhole) in combination with real time modelling to monitor and optimize the drilling process. This information is used to visualize the wellbore in 3D in real time. The system is composed of the following elements, some of which are unique and ground-breaking:

Automatic Prediction of Downhole Pressure Surges in Tripping Operations

A simplified dynamic model of the tripping operation is used together with an ensemble Kalman filter to predict transient pressure surges when running the drillstring in or out of the hole. Dynamic downhole pressure measurements from a tripping operation with mud circulation are used as input to the Kalman filter. Such data can be achieved by mud pulse telemetry at the field just before the tripping operation starts. The model is automatically adapted to the particular situation (well, bit-depths, drilling mud, etc.). This is important since exact values of some downhole parameters, like viscosity of the drilling mud, might be unknown and/or changing with time. We show by comparison with filed measurements, that the automatically updated model is capable of reproducing the transient pressure surges in consecutive runs of the string without mud circulation.

eDrilling used on Ekofisk for Real-Time Drilling Supervision, Simulation, 3D Visualization and Diagnosis

eDrilling is a new and innovative system for real time drilling simulation, 3D visualization and control from a remote drilling expert centre. The concept uses all available real time drilling data (surface and downhole) in combination with real time modelling to monitor and optimize the drilling process. This information is used to visualize the wellbore in 3D in real time. eDrilling has been implemented in an Onshore Drilling Center in Norway.The system is composed of the following elements, some of which are unique and ground-breaking: • An advanced and fast Integrated Drilling Simulator which is capable to model the different drilling sub-processes dynamically, and also the interaction between these sub-processes in real time. The Integrated Drilling Simulator is used for automatic forward-looking during drilling, and can be used for what-if evaluations as well. • Automatic quality check and corrections of drilling data; making them suitable for processing by computer models • Real time supervision methodology for the drilling process using time based drilling data as well as drilling models / the integrated drilling simulator • Methodology for diagnosis of the drilling state and conditions. This is obtained from comparing model predictions with measured data. • Advisory technology for more optimal drilling. • A Virtual Wellbore, with advanced visualization of the downhole process. A new generation visualization system designed to integrate all participants involved, will enable enhanced collaboration of all drilling and well activities in a global environment. • Data flow and computer infrastructure eDrilling (Ref. 1) has been implemented in an Onshore Drilling Center on Ekofisk in Norway. The system has been used on several drilling operations. Experiences from its use will be summarized and presented; both related to technical and work process issues. The supervision and diagnosis functionalities have been useful in particular. The system has given very early warnings on ECD and friction related problems. This paper will present the eDrilling system used on a specific Ekofisk wells with focus on experiences from its use. Introduction The southwestern part of the Norwegian continental shelf, called the Ekofisk Area, contains eleven major chalk fields. The Ekofisk field is the first and main discovery, discovered in 1969 and put on production in 1972. The fractured chalk reservoir lies at a depth of 9500 – 10700 feet and is approximately 11.2 x 5.4 kilometers in area, with production coming from two zones Ekofisk and Tor. It is one of the North Sea Giants with a STOIIP of 7 mmbo! Currently there are 4 fields in production, 4 fields abandoned with current production around 325,000 bbls per day of oil and 350 scf of gas per day. Water injection is currently used to maintain reservoir pressure, and approximately 900,000 bbls of water are injected each day.

From Sensors To Models To Visualization - Handling The Complex Data Flow

The paper describes the underlying infrastructure of an advanced real time integrated drilling simulator that is under development. Focus will be on handling the complex data flow and meeting the requirements regarding simultaneity when designing such a compound system. A synthesis of multiple transient and steady state models for drilling sub-processes with links to 3D visualization software and measured data is being built on the technology described in this paper. Possible applications include: • personnel training • testing and comparing drilling plans • real-time diagnosis: warnings, optimization and forward looking • process control • post analysis

Advanced Dynamic Training Simulator For Drilling As Well As Related Experience From Training Of Drilling Teams With Focus On Realistic Downhole Feedback

The paper presents a highly advanced training simulator that combines an advanced top-side simulator with a dynamic downhole simulator with an advanced transient integrated hydraulics and thermal wellbore model and a dynamic torque and drag model. The simulator is aimed at drilling and well operations, and is able to handle most of the normal operations, including high pressure high temperature (HPHT) wells, through-tubing rotary drilling (TTRD), extended reach drilling (ERD) and managed pressure drilling (MPD). The underlying simulator technology is modular, allowing for new modules to be added at a later stage. For instance, MPD control systems can easily be added to the simulator, allowing for training on an MPD operation with both the drilling crew and the MPD supplier. The simulator is able to use pre-programmed scenarios, replay, fast forward and rewind to facilitate efficient training and review sessions Moreover, the simulator is designed to provide realistic personnel training on emergency procedures and operations such as well control in a safe environment, thereby limiting the human factor in critical operations as well as possibly improving the procedures by frequent use and revision. Also, the simulator allows for integration of HSE in early well planning through simulator training on the actual well to be drilled. The main innovation is to use dynamic models verified in real-time operations together with an advanced top-side drilling equipment simulator for training on well specific scenarios. The value added for the industry is to give the drilling and/or engineering teams a possibility to verify and train on identified risk elements prior to drilling a well, as well as retrain during the operation on a “true” virtual copy of the well. So far more than 60 drilling teams have experience from training, and the feedback has been very positive. The paper will present the simulator as well as experiences from typical training cases on challenging wells. Introduction A highly advanced training simulator that combines an advanced top-side simulator with a high complexity and dynamic down-hole simulator has been developed. The simulator is aimed at drilling and well operations, and is able to handle most of the normal operations, including high pressure high temperature wells, through-tubing rotary drilling, extended reach drilling and managed pressure drilling. Moreover, the simulator is designed to provide realistic personnel training on emergency procedures and operations such as well control in a safe environment, thereby limiting the human factor in critical operations as well as possibly improving the procedures by frequent use and revision. Also, the simulator allows for integration of HSE in early well planning through simulator training on the actual well to be drilled. Such simulator training will also reduce the time that is used for training offshore, increase operational understanding and improve communication between driller and other personnel. The underlying simulator technology is modular, allowing for new modules to be added. For instance, managed pressure drilling (MPD) control systems can easily be added to the simulator, allowing for training on an MPD operation with both the drilling crew and the MPD supplier. The simulator is able to use pre-programmed scenarios, replay, fast forward and rewind.

Limiting Factors for the Ability to Achieve Accurate Pressure Control in Long Wells

Extended-Reach Drilling (ERD) with narrow pressure margin or uncertain geo-pressure is a challenge with respect to accurate pressure control. The back-pressure Managed Pressure Drilling (MPD) method has been widely used in drilling operations with the aim of controlling annulus pressure within safe bounds, and may also be applicable for ERD wells. However, the ability to control the pressure accurately is limited by several factors. Some of which are related to back-pressure MPD operations in general and some of which are more specific to ERD wells. In this paper, a study is presented on how pressure control is affected and sometimes limited by the actual data availability and quality, equipment, hydraulic models, control algorithms, and downhole conditions during an MPD operation in an ERD well. By using a transient well flow model, the theoretically obtainable MPD performance can be simulated. The benefit by utilizing real-time downhole pressure measurements transmitted by a wired drill pipe is demonstrated by simulations. It is shown quantitatively how variations in delay of measurement and bandwidth will influence the ability to control downhole pressure accurately in an ERD well. Benefit by this approach is a more accurate prediction of what is obtainable with MPD and how various factors may influence the ability to control downhole pressure.