WELL CONTROL IS A DATA PROBLEM

Starting off, what is well-control?

Well-control is a situation where normal operations are halted due to an observation of irregular behaviour of the fluids in the well. Well-control situations start when individuals or systems perceive a deviation from expected behaviour, and can play out as a short check before operations return to normal, or a complete disaster.

Deviations from expected behaviour can originate from downhole conditions, rig equipment, sensors, or from the many people on the rig site. With its many possible origins and outcomes, mastering well-control is a lifelong study where everything is uncertain.

Well-control is managed by the same team that is responsible for operations, and well-control events kicks off at any time, disrupting any other process for the management. 

Well-control always gets the full attention of management, seniors are on call to mitigate. Often, drilling management is personally involved in well-control situations, discussing situations, scenarios and solutions. 

Well-control situations are fix-it-at-all-cost situations with top priority, often overriding procurement routines, decision structures, and operational procedures. 

For all these reasons, well-control is a major uncertainty in well investments because of the interruptions of operations, random outcomes, and the capacity it eats of any well organization.

So why is it important to automate well-control?

From an economic perspective, you invest in a well to produce hydrocarbons. The quality of a well impacts the bottom line of any operator, in good scenarios the well is paid off in a week, in bad scenarios, it is the end of your company. The risk of well-control in well investments is hard to estimate with its uncertain nature.

In this paper on larger well-control situations, well-control is categorized to human errors, organization and failing technology. Underlying root causes are frequently pointing to poor planning and risk understanding, and lack of communication. This means that well-control can largely be reduced by changing how we work in planning and how we communicate.

More recent publications show a general pivot to the pore pressure plot to be the solution. The pore pressure plot is a  planning tool to predict pressure distribution along the planned trajectory of a well.

Exactly how much time well management is spending on well-control issues is not easy to measure, but we can use an indicator that 20 percent of time spent on a rig is on average non productive, we can assume drilling management spends a significant fraction of their time on well-control prevention, training and execution. Time that arguably should be spent on improving planning and risk mitigation to improve well investments.

1 Well control, can you ignore it?

2 Handling alarms is a high responsibility, is it fair to place on individuals?

3 The drill floor roles were established in the 70's. Is the risk of the old higher than the risk of the new?

4 Is it fair to place all the responsibility of a well on a few individuals working shifts?

5 The responsibility is huge for a team that still communicates by phone and radio.

6 Do you have time to verify that all the knobs are in the same position?

7 How can we mitigate risk if we don't have data?

7 In a well control incident, people die and companies goes down.

9 Recognize the traditional solutions?

10 It is time to take well control and well delivery serious and implement quality assurance on data and workflows

What does it mean to automate well-control?

Good question, maybe a way to look at it is to evaluate what the world would look like when you have automated well-control? And then we can back-track from that scenario. When well-control no longer influences day to day operations and the overall risk picture of the well investment, the task has been automated.

The following processes needs to be influenced to achieve this:

• Planning properly predicts risks
• Decisions during operations are made based on plans and well response
• Well management does not spend time on problem solving during operations
• Rig technology ensures decision making without human interaction

Now hold out your hands while the R&D budget is pouring into these problems, and go build.

How would the automation of well-control influence the oil and gas industry?

Depending on the phase of operations an oil company is in, the well costs are 50-80% of their CAPEX, so implementing a significant new technology would alter the core structure of the company.

Here are some assumptions on how the data driven oil company would look like:

• Fewer layers of management are needed as there are fewer tasks to handle
• Planning teams are responsible for the outcome of operations and are therefore higher skilled at risk management
• With higher quality plans and risks, the R&D investments and technology selections can be sharper, directed at performance gaps
• The reputation of the oil and gas industry will restore its tech image of the early 2000’s