Transforming oil and gas turnarounds using digital

4B1EF826 F928 490E 8E53 8F5200CFF63B e1510538899661

Transforming oil and gas turnarounds using digital

Turnarounds will continue to be a critical part of the oil and gas industry for the foreseeable future. How could turnarounds be re-imagined in a more digital world?


Why turnarounds need to improve


Recently, I had a breakfast meeting with a senior planner for an oil and gas outfit. His role was to improve turnarounds at his company. Turnarounds and shutdowns are the practice of carefully shutting down a continuous operation like oil and gas production, carrying out a range of repairs and mechanical adjustments, and then carefully bringing the equipment back into service.


A turnaround is a complex undertaking – the range of repairs and adjustments that are included in the typical turnaround is significant.  Everything from a valve rebuild all the way to recycling catalyst in a processing unit gets carefully planned out and scheduled during the year (or years) ahead of the turnaround event, and detailed in a plan that spans thousands of lines. Thousands of contract workers descend on the plant and await instructions to carry out their assigned tasks. Parts and equipment arrive in a steady stream to massive lay down yards ahead of use. A fleet of rental equipment (power units, cranes, hoists) stands by awaiting job assignment. Scaffolding and other temporary structures jump into place to enable safe work.

Management, being mostly rational, recognise that a plant only makes money when it’s running, and so the goal is to minimize the amount of time and cost it takes to carry out the turnaround. Managers are therefore motivated to double and triple order the number of workers and pieces of rental kit to reduce delays in getting work started and completed. Mobilising a big inventory of these resources on site is a costly way to do business, particularly when turnarounds take place in remote hard to reach places like Canada’s far north oil sands and east coast off shore.  

It makes sense, therefore, that improving turnarounds is important. But plants are all different, creating a barrier to copying good turnaround practices from one plant to the next. Turnarounds are infrequent affairs, perhaps an annual undertaking, and institutional memory of what works and didn’t work fades particularly during times of high staff turnover (like now). It can be hard to justify investing in improvements to turnaround efforts, when a turnaround might only last 15-20 days.



The cost of digital, however, has fallen dramatically, whereas the cost of just about everything else (people, time, regulatory compliance, equipment, emissions, consumables), has gone up. Almost every worker on site has a supercomputer in their pocket (we call them smartphones, but that’s what they are). It therefore begs the question – how might digital solutions be applied to the challenges of executing cheaper and better turnarounds?


What problems are we trying to solve?


There’s little point in throwing technology at a poorly understood problem. Here’s my starter list of turnaround challenges.


Garbage in, garbage out.


Planning and executing an efficient turnaround is based on a good understanding of the plant. If tags don’t match equipment, if equipment records are out of date, if diagrams don’t reflect actual installed equipment, then the plan is based on poor quality data, and will struggle to be efficient. The biggest pay off from digital on a turnaround is to fix the data before the turnaround even starts.


Managing turnarounds as projects and not as processes.


The management tools typically used for running turnarounds assume management is dealing with a project and not a process. Therefore turnarounds show the kinds of cost and productivity improvements typical of projects, and not the more dramatic and impactful gains from the world of manufacturers. Turnarounds could leverage the kinds of tools we would see in manufacturing.


Plans with hidden problems.


Complex plans built by humans will have inevitable errors in them. Every turnaround planner can tell you of their personal colossal boo-boo, when they ordered a crane for a task only to discover the crane exceeded the space limits. Digital could help reveal these problems, and make plans more resilient.


Can’t find people, assets and equipment.


Once a turnaround gets in motion, just finding people, equipment and tools on site can be a nightmare. Many a turnaround features urgent text messages offering bounties to the first person able to locate a missing tool (although perversely this also encourages tools to go walkabout so as to trigger the bounty payment). Digital could help make the invisible visible.


Inexperienced resources.


A large contingent and inexperienced workforce creates its own challenges. They need quick training on task, and above average levels of communication to address unforeseen issues in the moment. Digital could accelerate their productivity.

Top 10 technology enabled improvements


Digital tools are not going to change the cultural challenges of turnarounds – field rivalry, misaligned performance measures,  organisation models that block knowledge sharing and promulgation of lessons, and NIMBY-ism.


However, digital tools can still have a big impact. Here’s my top 10 list of candidate solutions.

One. Fix the data


Data is the great enabler in turnarounds. High quality data unlocks most of the benefits in the other areas. The year before the turnaround, I would set out to correct the equipment tags, clean up and rationalize the engineering diagrams, integrate the various systems that create asset data so that the business has just one version of the truth about the plant.


Two. Build a digital twin


Using the clean data, I would build a digital twin of the asset or plant undergoing the turnaround, along with the tools, equipment, and people that will work together on the turnaround. The digital twin enables engineers to model and “play” with the turnaround using planning tools, artificial intelligence and machine learning to see how the turnaround executes.


Three. Apply AI to the plan


Next, I would subject the turnaround plan to a bout of machine learning and artificial intelligence. An AI engine could put the plan through millions of iterations of possible scenarios, much like how AI learns to play GO or chess. AI could test out weather scenarios, the impacts of incidents, and other hard to anticipate challenges. AI would help uncover hidden plan issues and constraints, and identify the optimal use of people and resources. Engineers could then apply more humanistic insight to get to an even better plan.


As the plan is worked, I would apply game engine technology to marry the plan to the digital twin. Engineers could then watch the turnaround play out as a game or video, using services like Real Serious Games. This would create a better mechanism for aligning contractors and workers to the tasks to be executed.


Four. Deploy cloud computing


Next, the actors in the turnaround (people, tools, assets, equipment, consumables, inventories), need access to a network to be visible. I would stretch a cloud computing environment over the site, and leverage all those in-pocket supercomputers to enable visibility. The cloud will feed all the actuals back to the planners who will use AI and its scenario ability to constantly course correct.


Five. Sensorise the inanimate


Many of the inanimate objects used in the turnaround (tools, rental equipment), will need to be equipped with light-duty beacons to broadcast their presence. As soon as one of the supercomputers comes within range of a beacon, it can handle the heavy lifting of broadcasting the location of the beacon to the cloud.


Six. Use AR and VR to train up resources


We need to enable the field teams to execute the turnaround. This is a great role for augmented reality and virtual reality. I would equip the human team with Google’s 3D cardboard viewer (or, for the rich outfits, Hololens), and a set of apps that presents the digital twin and the activities in the plan for their task for the day or week. Project engineers could annotate the visuals with alerts, cautions, task clarifications, lessons from other turnarounds.


Seven. Use social tools to communicate to teams


We need to be able to communicate to the field in the moment – alerts, broadcast announcements, safety reminders, specific task changes, and so on. I would deploy some of the modern social tools like Slack to this job. It’s easy to use, and training is plentiful on line.


Eight. Orchestrate on the day using service tools


While engineers need to see the whole plan, teams just need to see their specific tasks. I would deploy best in class cloud solutions like ServiceNow and SalesForce for managing daily work and services. These solutions have readily available apps for the smartphones and integrate already with the technologies that underpin the digital twin, enabling delivery of diagrams, specs, instructions, BOMs and all the other key data directly to the worker. Smartphones also allow for markup of diagrams, capture of work records, photographs, video, time tracking, and so on.


Nine. Use AR for problem solving in the moment


Field teams will often need access to some expertise in the moment – dealing an unfamiliar pump or a diagram that doesn’t match. I would selectively deploy a hands-free AR tool such as from FieldBit to enable a team to communicate with a resident expert back in a control room, which would keep the team moving and avoid having a senior engineer waste time traveling to a site for what could be something quite trivial.


Ten. Bring in the robots


Some jobs were never meant for people. I would use robots for key hazardous jobs such as inspections of tanks (yes, there are swimming robots that can dive into a full tank and check for repair needs), inspections at height (flying drones), and in pipelines (smart pigs). These jobs will be controlled by humans, but the robots will be on the cloud, using the digital twin.



Turnarounds need not be the huge time and cost problem that they have become. Using modern, low cost and available digital ltechnologies could turn turnarounds around. And wouldn’t that be great.


  • øystein stray
    Posted at 05:38h, 14 November Reply

    I like your article!
    I am Øystein Stray, representing Visco. We provide a digital twin solution called VCOG, that is designed for Owner-Operator, utilizing the power of visualization, deeply integrated with business processes and relevant data.
    We are in the process of building a Turn-Around app that cover most of your highlights, but we would like to anchor the solution based upon SME input.
    Are you interested?

    • Geoffrey Cann
      Posted at 11:02h, 10 December Reply

      I am interested to learn more but I’m probably not your guy to advise on app features for turn arounds.

Post A Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.