National Oilwell Varco rep in Baku talks unified AI-based platform for drilling operations

Oil&Gas Materials 23 November 2023 14:11 (UTC +04:00)
National Oilwell Varco rep in Baku talks unified AI-based platform for drilling operations
Kamran Gasimov
Kamran Gasimov
Read more

BAKU, Azerbaijan, November 23. National Oilwell Varco (NOV), a company specializing in equipment and components for oil and gas drilling and production, has developed a single standardized artificial intelligence (AI)-based platform for drilling operations, NOV's business development manager, completions and production solutions Torbjørn Hegdal said, Trend reports.

He spoke at the SPE Caspian Technical Conference in Baku.

"When you go from understanding algorithms to building a platform, it takes a lot of effort. So, the first step is to digitize and then collect data and systematize it. I would like to go a little bit deeper into the macro picture and focus on some of the things that we do in our company, specifically drilling operations, process operations, both surface and subsea, and maybe business development. We believe that at the heart of automation is the automation of drilling operations," Hegdal noted.

"It's the use of technology to process data, look at it and learn from it, and it's all an automated process. We see a lot of benefits when it comes to energy efficiency, chemical use, etc. to really reduce the impact of our operations. It promotes remote working. It kind of gets people out of unsafe places, increases economic efficiency, etc.," he said.

According to him, a lot of it has to do with condition monitoring, predictive maintenance, specifically with early detection and getting ahead of operations.

"Briefly explained, it's about platform development. There are a huge number of possible applications within our business. That is why we have developed a single standardized platform. There is a lot of data collected from wells, production, operations, drilling operations, and the platform was created precisely to systematize this data and to be able to learn from it. One of the things we're doing is analyzing the data on site, in real time, so that it's not affected by communications and it doesn't go to the cloud. But it's about the way we store the data and also use it remotely. But again, this is an attempt to move to a more standardized platform," he noted.

Hegdal mentioned that the platform is designed and used primarily for drilling operations, but technologically, the company is looking to take the same approach.

"So it's about collecting data, systematizing it and moving it into some kind of unified space and onto a unified platform. And it's very much about creating and developing a digital twin to work on. And then the use of different levels of artificial intelligence, from machine learning to working deeper with the data and making faster decisions in operations. So, in drilling, it's very much about automating drilling and identifying that function," he added.

Torbjørn Hegdal added that there are two sides to this. One is identifying any outliers, whether that's a risk or actually optimizing performance. It's about achieving repeatable operations. There are other applications, such as vessel condition monitoring. It is used on more powerful vessels to truly control this part of the operation as well.

According to Hegdal, many analyses are only performed on-site, but operations can also be performed remotely.

"When it comes to workflow, it's about predictive maintenance to minimize downtime and process issues. So, I am just going to dive into a couple of examples because I think it's a good way to look at a micro level of what's actually happening in algorithms and how we can implement it in business," he emphasized.

Hegdal said that the example relates to a water treatment process, whether it's produced water or seawater that is processed for pumping purposes.

"It's a case study, but it's really to understand the performance of membranes and try to be more proactive. If it is done manually, just observing deviations, it's going to take time and end up causing more downtime. But in this case, machine learning allows to track patterns in equipment performance and detect a typical problem - biofouling - at a much earlier stage," he said.

"Of course, this has a direct impact on uptime, but there are other benefits, such as reduced chemical consumption, and overall it's a more sustainable solution. I would take it one step further, and that is to do with the underwater part. This is to visualize what happens underwater when the work process is done above. It's all about reliability, uptime and minimizing maintenance. So we're going to take the same approach here as well to really understand all the data that's coming in," Hegdal explained.

"So when we talk about water treatment, we use electrolysis, all the chemicals on the seafloor, and we're continuously monitoring that. We get a lot of data, but again, it's about understanding and seeing that data and learning from it," he said.

Stay up-to-date with more news at Trend News Agency's WhatsApp channel