Digital technology will become an important tool for drilling safety

Drillers are increasingly using artificial intelligence, machine learning, and computer vision to inform personnel of performance limitations and keep drillers safe

Patterson-UTI’s Rules Engine Exchange (REX) alerts drillers to potential issues that could limit rig performance so they can take action. The company plans to complete pilot installations of the system on 10 rigs by the end of this year.

As oil wells become more complex and new automation is added to the drilling process, drillers are increasingly relying on alarm systems both to protect people from hazards such as falling objects and to help them manage Massive amounts of data from surface equipment and various sensors in the wellbore.

With the advent of safety-related alerts, recent advances have brought computer vision to the forefront, primarily through continuous, dynamic monitoring of areas where heavy drilling equipment operates, and where the risk of serious injury or death is highest. By combining artificial intelligence and machine learning with cameras installed in hazardous areas, these systems help reduce potential blind spots and errors made by humans monitoring rigs.

Richard McConomy, manager of Seadrill’s PLATO digital platform, said: “Our main objective is to keep workers safe and we will use every possible tool i.e. video analytics, artificial intelligence, machine learning to help us achieve this.” Help companies leverage big data sets to make better decisions. “AI-powered video analysis is an effective tool to detect and manage safety issues in real time by sending alerts to prevent accidents.”

For more alerts related to equipment and operational performance, companies are focusing on leveraging advanced data analytics, which may involve offloading data processing from the cloud to the rig site for early warnings.

“We want to be notified when we’re putting performance on the table, not when we realize we’re limited,” said Trevor Olson, drilling optimization manager at Patterson-UTI. “If the mud pump goes offline and we If we can’t get enough flow in one section, then we’re not as productive as we’d like. We want to be able to see this happen and be proactive in addressing these constraints. ”

Figure 2 Patterson UTI’s REX system verified and provided 83 alerts. Another 30 updated alerts, focusing more on safety and dangerous events, are being validated.

Last year, Patterson-UTI began rolling out its Rules Engine Exchange (REX), a cloud-based real-time alerting system for monitoring devices and maximizing performance. It uses data from electronic well loggers, rig control systems, morning reports and real-time models calculated on the rig site to send operational event alerts in near real-time via text message, email and online interface.

Users can tailor the system to their needs, subscribe to specific types of alerts, and even create new alerts for themselves. Three categories of alerts are available: simple operational events, such as a rig starting or completing a trip or reaching the total depth of a well; equipment alerts, which notify the user when specific equipment enters a fault or warning state; and data quality alerts, which notify users when a given rig It will tell the user when a sensor returns an incorrect value. as shown in picture 2.

Patterson-UTI completed the first phase of REX development last year, namely building its backend infrastructure and user interface, installing it on a cloud computing network, and completing field trials to verify the effectiveness of the system. Since then, the company has been working on the next phase of development, which includes adding alerts to cover additional performance-limiting scenarios and equipment issues, as well as installing the REX engine directly onto the servers on the Patterson-UTI rig. The company plans to have the system fully installed on all rigs by the end of this year.

Mr Olson said building the infrastructure for the REX system to run directly on the rig would allow the company to deploy alerts locally rather than through a cloud-based infrastructure. By installing the system on the company’s CORTEX KEY edge server, which is installed on each rig and connected to rig control systems and electronic drilling recorders, the company believes it will be able to reduce processing data and alerts to rig crews the time required.

“The main obstacle with cloud computing right now is that we can only process as fast as the slowest data we receive,” Olson said. “We basically have rig control system data in real time, but the EDR we get from service providers The data is in 5 minute packets, so we have to run the rig alert system after at least 5 minutes. Then it takes a few minutes to run the code and deliver the alert. By storing all this in a CORTEX KEY, it will take no time to fetch the Data. Instead of five to seven minutes, we’ll be able to get notifications within a minute.”

Reducing delays is important to Patterson-UTI as it integrates more alerts for the potential for loss of well control into the REX system, Olson said.

The company currently has 83 alerts validated and available on the REX system, with an additional 30 alerts in the process of being validated. The new alerts are more focused on safety and dangerous event detection, tend to be more complex than the initial alerts developed for the system, and must be validated for efficiency with a set of sample data sets. Examples of new alerts include identifying potential rig outages, mud pump consumptive failures, stuck pipes and compliance with well control procedures.

New alerts like this one are a key driver of getting REX installed directly on rig servers, Olson said. “For almost all alerts on REX, latency doesn’t necessarily affect the quality of the system. It doesn’t, because in most cases the alerts aren’t for things that you need to know when they happen. But with We have a deep understanding of things like well control alerts, and we want to know about things the first time they happen so we can act quickly.”

Patterson-UTI said it expects to complete the pilot installation of REX on 10 rigs by the fourth quarter of 2023, before installing it on all rigs by the first quarter of 2024.

H&P’s PDS Redzone computer vision system can break down individual camera footage taken around the red zone of the pipeline system, and when PDS is activated, it will sound an audible alarm if it detects people in the red zone.

Like others in the drilling industry, alert systems are nothing new to Helmerich & Payne (H&P). The company integrates a number of custom alerts to measure various performance goals for clients. It has an alarm system that monitors engine overuse and is designed to conserve engine power on the rig. It also has an alarm system to avoid downhole shaft collision.

However, some of the company’s more recent technologies and systems have been developed around reducing the risk of falling objects. An upcoming technology from H&P aims to keep people off the Pipeline Delivery System (PDS) during high-risk activities: PDS Redzone, shown in Figure 3.

“We’re addressing this where we are most vulnerable, where is our biggest opportunity? Around the PDS and the drill floor. That’s our biggest opportunity,” said Mike Lennox, senior vice president of H&P Americas.

To reduce the potential for serious injury in the PDS area, existing procedures are in place: use of appropriate buffer zones and barricades, equipment inspections, and thorough pre-job planning to ensure that all equipment components are in good working order and that safety program. Still, there is a risk of serious injury or death (SIF) in the red zone (defined by H&P as a 45-foot radius from the PDS rack).

The PDS Redzone computer vision system uses neural networks to break down individual frames of incoming video into smaller parts and detect essential features. Neural networks are trained on large datasets containing images or videos annotated with the locations of critical points, such as the top of a PDS or features of a human body. It learns to detect and associate these keypoints with specific body parts or features, allowing it to estimate the position and motion of objects or people in new, unseen images.

When the PDS is activated, and operating conditions meet the specified logic criteria for detecting a person or persons, the system automatically triggers an audible alarm and flashing lights, alerting the individual to enter the restricted area, and simultaneously stops the PDS hydraulic system.

The system was refined through field testing on a drilling rig in the Permian Basin. After refactoring the PDS Redzone code, the frame rate has increased by 150%, the field of view is wider, and the number of pixels per image has increased by 4 times. The camera system was also relocated from below the rig floor to near the rear winch of the PDS, with a focus on the top of the PDS.

In addition, H&P aims to establish an extended management protocol for restricted areas outside the red zone above the PDS, in line with existing practices using physical barricades and buffer zones. “We want to set it up to have a buffer zone on all sides of the pipeline system,” Mr Lennox said. “We’re not just monitoring the area near the PDS, we want to go further because the exposure of the pipeline is still very high.” The company plans to roll out the app across its entire team sometime next year, it added.

H&P’s Caliper Calculator app is also designed to reduce dropped object incidents. In 2011, a bushing fall due to a mismatched elevator size pointed to the need to create a better system for the task. The Caliper Calculator sits on the rig’s tablet and links directly to its FlexChecks electronic pre-job planning application, which alerts rig crews when ID and OD verification is required. It is designed to automatically determine if measurements are within allowed tolerances, thereby reducing the risk of using a wrongly sized lift.

Rough Drilling, another H&P app, also aims to reduce the risk of falling objects when vibration levels exceed expected levels as the drill string moves downhole.

The company defines “Rough Drilling” in terms of the standard deviation of hook load and torque. The app calculates these deviations in real time: if it detects a deviation severe enough to cause severe vibration—a threshold for “Rough Drilling” that depends on the rig and the formation being drilled—it displays a red flag on the monitor in the driller’s cabin and Sound an alarm to remind the driller to clean up the drill floor. If that alert exceeds a predetermined period of time, another alert will call for an inspection of the traveling equipment and mast to identify any potential problems requiring intervention.

“While we may not be able to eliminate rough drilling, we can prevent exposure by making sure employees are aware of the hazards and stay away from the rig floor during these activities,” Lennox said. “We must keep employee safety in mind and eliminate these hazards.”

Launching in 2021, the Rough Drilling app is currently available on all H&P rigs.

Figure 4 Seadrill has partnered with Brazilian technology developer Altave to install the Altave Harpia PPE detection and red zone monitoring system on three of its drillships operating offshore Brazil, including the West Tellus. Using cameras installed in high-traffic areas such as pipe decks, the captured video is processed by artificial intelligence and machine learning algorithms within the system to detect a lack of personal protective equipment among drilling crews. If the system detects such a deficiency, an alert is sent to the driller.

In the past year, Seadrill has deployed three drillships to Petrobras’ Búzios field offshore Brazil for exploration work: West Carina and West Tellus in September 2022, and West Jupiter in February this year. For all three rigs, the operator and contractor agreed to install Altave Harpia, a Brazilian-developed personal protective equipment detection and red zone management system to alert drillers of potential accident risks in red zones, as shown in Figure 4 .

The technology uses computer vision to interpret data collected from six cameras connected to a terminal, a hardware system that handles data input and output. Cameras are installed on pipe decks, drill floors and riser decks, which are high traffic areas with lots of moving equipment. Artificial intelligence and machine learning algorithms built into the software are used to detect whether people within the red zone lack personal protective equipment. If this is detected, the software sends an alert to the driller, who can then decide on the appropriate course of action. In addition to the pop-up alarm in the driller’s cabin, the alarm can also be seen on monitors mounted on the bridge, company personnel office, tool driver’s office, deck driver’s office, HSE office and on each crane located on the platform.

Camera-based systems like Alave Harpia allow Seadrill to have better granularity in its analysis than previous monitoring systems. McConomy said this could improve monitoring efficiency because it minimizes errors from human interpretation of the data. He said: “Because the camera system is driven by artificial intelligence, it is an effective tool for detecting and managing safety issues in real time and sending alerts to prevent accidents. Machine vision with artificial intelligence allows certain patterns to be automatically detected, such as differentiating wear and tear. gloved hands versus non-gloved hands, or to determine if a person is in a red zone. Tracking and trending are common alerts that give us simple opportunities to improve operational planning.”

While Seadrill has yet to release any safety metrics for its Brazilian rigs, Mr McConomy said he believes computer vision technology will be an important tool in a driller’s safety toolbox. “I am a firm believer that cameras will play a huge role as the sensors of the future. We are only scratching the surface of their capabilities and technological advancements, not only in terms of camera quality but also in terms of processing power. Brazil is currently our test bed for this particular technology, But the quality of the system is improving and we can benefit from standardization.”

Digital technology will become an important tool for drilling safety