The roles of conceptual and mathematical models in drilling

The roles of conceptual and mathematical models in drilling

This is a topic page to show an overview of a sub field of Integrated operations, describing the knowledge developed by the IO Center

0
No votes yet

users have rated this content. We would love to have your vote as well. Log in and rate!



A discussion of the process of going from work as imagined, i.e. from conceptual and mathematical models developed before an operation, to work as done based on experience and skills of the operational team to handle small and large deviations from models. How can models better follow all operational phases?
Content

A "drilling model" is an ideal representation of the well and the drilling operation. It may refer to the driller's mental model of the well, a computer simulation calculating the fluid dynamics of the wellbore, a more statistical description of the parameters, or even the drilling plan, which succinctly captures the drilling operation as we imagine it will proceed. But whether the model is on a computer or in the heads of the drilling team, some challenges remain the same. Namely:

  • When receiving new information, to update your model fast enough to influence decision-making and in a way that actually brings it closer to reality
  • Dealing with the fact that your model will always be a simplification of reality, with inaccuracies and blind spots.

In our and related case studies, we have seen how an experienced drilling crew master the above challenges. When we interview the crew, drilling is often described as a streamlined and standardized activity with clearly defined processes. The success of a drilling program is said to depend on planning and strict compliance. We refer to this streamlined ideal, which is tied to the drilling plan and procedures, as "work as imagined". But in practice, experienced personnel have a rich toolbox of alternative strategies that they employ to succeed while deviating from the plan or going beyond it. This mode we call "work as done". This is not better or worse than "work as imagined", rather they are two modes of thinking that the crew alternate between as the situation demands it.

This switching is perhaps best observed when the crew faces the model-related challenges mentioned above. In normal operations, the model, be it a mental or a computer model, can incorporate new data easily. For instance, an increase in bottomhole pressure may be observed, but the driller expects it and the computer calculations confirm that it can be fully explained by the recent increase in pump rate. The data is explained in the framework of the model. Other times, the crew will encounter data which defy their expectations. If a gas kick is encountered, there will be a discrepancy between what the computer model predicts and what is actually observed, to the extent that the driller reject his mental model of the well, replacing it with one describing a kick event. In the language of sensemaking theory we say that the crew enters a "re-framing" mode, where the data no longer fits the frame of the main model, and a new model must be constructed.

Re-framing also occurs in less dramatic situations. In one case (Haavik 2010)  a drilling crew discovered a persistent but non-critical discrepancy between the expected and measured mud pit volume. While the drilling continued, a video conference was convened with the company's expert centre. The centre members came up with several interpretations. What's interesting here is that the interpretations or alternative models of the operation did not result from a checklist or pre-determined procedure, but was formulated by people across the drilling organization that improvised with their different sets of information, tools and experience. Some drew up a comparison with a nearby well that suffered similar problems, others conjectured that the mud was too dense and was leaking into the formation, yet others suggested a chemical reaction between mud and formation. For a short while, the full complexity of the well and the operation was showcased for everyone involved. In the end, the mud logger inspected the shale shakers and found that they failed to efficiently separate mud and cuttings. The cause had been identified. Inspecting the shale shaker was not the primary work of the mud logger, but neither was the re-framing clearly confined within the borders of people's work tasks.

In short, when the operation no longer runs smoothly, the team steps outside the confinements of their streamlined procedures and roles. Temporarily, the complexity of the information picture rises as the group tries to make sense of the situation and construct a new model of reality. The competence shown in this re-framing process belongs to what we call "work as done", and is a crucial capability of the team, despite not being enshrined in procedures or organizational charts.

 

In drilling, computer models show their strength in the normal mode of work. Computer models may rapidly absorb torrents of data and draw conclusions about e.g. downhole pressure from real-time pressure, flow rate, and temperature readings, as long as the model assumptions are accurate. Kicks can be detected earlier when a computer is used in real time to predict all the mundane variations in flow, caused by expanding mud or displacement by the drill string. But when a kick occurs or the drill string gets stuck, the computer model does not internally represent the situation. The discrepancy between model and measurements creates a strong signal for the crew to enter their re-framing mode, but the computer model itself does not follow suit. The computer model, which now represents a past state of the well, is ignored as the crew juggles interpretations and debate solutions. Here we see an integration gap between the team and the computers.

How to bridge this gap is still an open question, but there is a lesson from reservoir engineering. In that field there is an industry drive towards automatic updating of the reservoir models as new data keeps ticking in. An automatic re-framing of the model if you like. However, interviews show that the personnel is often negative to this. The cause appears to be that the humans feel left behind with their now outdated mental models of the reservoir. Even worse, the engineers' understanding of the reservoir was made through familiarizing themselves with the old model, knowing how it would respond and knowing its limitations. They were well integrated with it and did not in fact have an independent and 100% mental model. So when the computer model was updated, their understanding was in a sense not only outdated but lost.

It's easy to imagine that a crew with a drilling model which automatically re-framed itself would end up in a similar situation. Solving the man-machine integration gap is difficult, but holds the promise of realizing a great unclaimed potential in the personnel that has been there all along.

 

Further reading:

The Overlooked drilling hazard - Decision Making From Bad Data

Haavik, "Making drilling operations visible: the role of articulation work for organisational safety."  Cognition, Technology & Work, Volume 12, Issue 4 , pp 285-295. DOI: 10.1007/s10111-010-0139-2

Almklov, Haavik, "The role of models in operative decision making processes in the subsurface domain – A pragmatic perspective" , 2009 IO center report for project 2.1

Nybø, "Anomaly detection and sensemaking in time series interpretation", IO center interal report 7020138/01/12 for project T1/IO3

Nybø, Frøyen, Lauvsnes, Korsvold, Choate "The overlooked drilling hazard: Decision making from bad data" SPE150306, presented at 2012 Intelligent Energy conference.

Korsvold, Bremdal: "Collective learning as a principal mean to safer and efficient drilling", 2013 IO center report for IO center project IO3, SINTEF F23988.

Other key information

304 results
Below, you will find related content (content tagged with same topic(s) as this topic page)
Content type: Report

The Integrated Operations Maintenance and Modification Planner (IO-MAP) The first usability evaluation – study and first findings

The Integrated Operations Maintenance and Modification Planner (IO-MAP) is a software test bed, which is being designed in an iterative process.

5
Content type: Ongoing activity

A benchmark for integration of reservoir engineering and geophysics

This activity facilitates the collaboration between reservoir engineers and geophysicists

0
Content type: Publication

A Derivative-Free Approach for the Estimation of Porosity and Permeability Using Time-Lapse Seismic and Production Data

This is a scientific publication written in collaboration with the IO Center.

0
Content type: Publication

A Lagrangian-Barrier Function for Adjoint State Constraints Optimization of Oil Reservoirs Water Flooding

This is a scientific publication written in collaboration with the IO Center.

0
Content type: Report

A Method for Function Analysis and Allocation in IO-driven MTO Systems: Challenges and Prospects

Describes use of the IO MTO method in functional analysis and allocation

0
Content type: Report

A new telemedicine practice on-the-go

Development and demonstrations of a telemedicine prototype that enables the key features of the “Virtual Examination Room - On the go”.

0
Content type: Publication

A Robust Scheme for Spatio-Temporal Inverse Modeling of Oil Reservoirs

This is a scientific publication written in collaboration with the IO Center.

0
Content type: Publication

A Sparse Basis POD for Model Reduction of Multiphase Compressible Flow

This is a scientific publication written in collaboration with the IO Center

0
Content type: Publication

A structured approach to improved condition monitoring

A paper describing a systematic approach to select and implement appropriate condition monitoring for systems, structures and components.

0
Content type: Presentation

Acceptance criteria for dynamic risk analysis

Presentations at IO Center Work Shop Sept 2013

0

Pages