News
BLOG – 10.21.20

Automation for Field Development: Top Tips

We caught up with Agan Balangalibun who works out of our Perth office in our Production Assurance team. Agan has been with Xodus for two and a half years, prior to joining Xodus he completed a graduate program in production engineering with Shell in Malaysia.

Agan is one of our young engineers leading our progress in automation and data science. Most of Agan’s coding and data science skills have been self-taught and he has used them to apply automation techniques to improve efficiencies on several projects. We caught up with him on how he has been applying these skills to Field Development.

Why does Field Development lend itself to automation?

Field Development, like many parts of the oil and gas industry, has the potential to be improved through the application of automation and data science. Here at Xodus, we have applied automation techniques to model a large variety of potential concepts, sometimes orders of magnitude larger than what is commonly attempted. This allows us to not only assess the edge of the design and operating envelopes but understand the scale of influence each parameter has on a system. We have visualised these results using tools that provide deep and intuitive insights. Already, the application of these tools in an Australian development we are working on has quickly identified the opportunities and limitations of various concepts that have been considered.

The rapidity that these tools give us allows us to consider a greater breadth of concepts and test more sensitivities, for the same amount of time spent on the project. This has driven up the value of the work that we can deliver to our clients.

What are your top tips?

Start with great domain knowledge

Our key resource at Xodus is our technical professionals; subject matter experts who know their areas of work inside-out. These experts, having worked problems from many angles often know the limitations of the current methodologies. Perhaps there are limited resources to do a certain amount of work or the analysis of results takes a long time. Maybe the results shown are important but difficult to communicate, or struggle to catch the attention of their audience. By sitting down with the experts to understand the current limitations in the industry, a wish-list of opportunities to be seized can be generated and then we can start to look at which of these lend themselves to being automated.

Employ and develop the automation skills that you need in the team.

Learning to code is no longer the domain solely of computer science majors or software developers. With buzzwords like “Data Science”, “Machine Learning” and “Artificial Intelligence” becoming common place, coding ‘bootcamps’ and internet courses teaching all levels of coding are more accessible than ever. While technical managers should have a cursory understanding of what coding is and how to leverage it, there is not a need for them to learn how to code and automate these processes themselves; after all, they add value through managing and providing insights.

Junior Engineers on the other hand are often the staff members grinding out manual tasks. As such, they are well positioned for automating these processes. They are often joining the workforce with coding skills in hand, and even if they are not, training them to reach a basic level of coding is a worthwhile investment.

Developing an automation culture takes active direction from leadership to empower the younger engineers to challenge ‘the old way’. What goes hand-in-hand with training is direction from above to solve and automate processes. There is no better teacher than necessity and having an actual real-life project to apply classroom skills is often the best motivation to learn a new skill.

There are many options to choose from when it comes to programming languages. There are pros and cons to each of the languages, but by using an open source language such as Python or R, one gains access to a huge resource, which leads me to the next point:

Leverage the open source community

It takes talent and practice to write elegant and robust code. To expect that engineers can be made into accomplished software developers by doing a bit of training is unrealistic and does a disservice to both software developers and your engineers. Instead, by learning some coding fundamentals, you can leverage the excellent work done by the open source community and apply specific modules and libraries to solve problems. Here is an example:

Is there a manual labour problem: does this task take too long to do by hand? Is it a repetitive task that can lead a person into a lapse in attention and thus cause errors? Are we likely to need to repeat this exercise once more mature base data is available? Then consider using regular expressions to parse large volumes of text to extract key information. Consider using pandas to efficiently organize and manipulate tabulated data. In particular, learning to structure data this way makes for easier integration with other coding modules that allow for machine learning or visualisation of that data.

And finally, what do you think the blockers to field development projects taking up the opportunity for automation?

Individuals not realising the value of challenging old ways of doing things or thinking that doing so would be a difficult and costly exercise. Automation and data science could very well bring a faster and more detailed turn around in field’s development.

The adoption of automation and data science in Xodus’ field development team is continuous and reflects Xodus’ overall strategy to bring innovation into all of its work. And I think the concept of data science to facilitate energy projects is really exciting and can’t wait to see what challenge the industry lays down for us next that we can apply this new way of thinking to.

‹ BACK
Similar Articles

Interested in talking to the experts? Enter your details below to arrange a meeting.

    Contact Us

    Contact Form