Deepwater Digital: A Tale of Digital Twins in the Offshore Oilfield

Can you fathom the complexity behind a 100,000+ ton ship drilling a 16-inch drill bit down through 12,000 feet of water and then 20,000 feet of seabed? Offshore oil and gas operations and many other industrial processes and products have a few things in common – the need to optimize processes to capture value. Even before the advent of digital technology, optimization often remained an incredibly daunting task in the offshore oil industry. Oil exploration platforms operate in remote environments, have a very low risk/failure tolerance, involve many stakeholders, and require access to vast amounts of often compartmentalized data. Various stakeholders ranging from the Chief Engineer and his trusty staff (see photo below), the regional vice president of operations onshore, and the original equipment manufacturer in Houston all have one common goal: optimizing their assigned responsibilities and duties. 

Chief Engineer and 2nd Engineer (me). Background: West Mira – a mobile offshore oil and gas exploration rig.

Now enter from stage right, the digital twin. Digital twins are virtual and made up of sensor data fed in real-time from various processes or things – in our case, things or processes on an oil rig – used to create a virtual representation of a physical object, process, or even people. The digital twin is then used to analyze or contextualize data to drive decisions and actions.

The specific benefits of digital twins are varied based upon their particular use cases. For example, an oil and gas platform could operate multiple digital twins (DTs) or one for a specific piece of equipment or process. Each of these DTs could have different use cases, say the vice president of operations, engineering team, human resources, and marketing. Funneling various sensor data from the oil platform into a digital twin allows multiple people from multiple backgrounds to perform their analysis of diagnostic feedbacks (the present time) or prognostics “feed-forwards” (the future possibility). The oil and gas industry provides two great current digital twin examples and their use cases: blowout preventer monitoring ( a product) and well construction process automation (a process).

Blowout Preventer

Blowout preventers (BOPs) are essential pieces of mechanical equipment found on onshore and offshore oil and gas drilling operations. In short – they provide oil exploration platforms a means of controlling fluid pressures within the drilling operation – aka stopping oil spills. Think of a large cup of diet coke from McDonald’s. Your goal is to extract all of the soda found below the layers of ice. Now apply thousands and thousands of pounds of pressure upon the soda in the cup and attempt to drink it without spilling/erupting/losing control. in this instance the BOP would sit at the top of the lid, just as the straw comes out of the cup. They’re incredibly complex devices made up of high-pressure hydraulic, electronic, piping, and control systems that allow a means of controlling ‘oil well pressure,’ or in our case, diet soda. By utilizing a digital twin, those responsible for the operation and maintenance of the BOP can 1) feed data to a DT accessible by the platform and the client onshore 2) analyze diagnostic data 3) conduct predictive analysis (determine when parts will fail). DT’s have driven value for owner-operators by limiting the amount of time spent pulling the BOP all 10,000 feet up off the ocean bottom to check a single valve and remove human bias. When considering the impetus for eliminating human bias in BOP maintenance and operation, one can look no further than the 2010 oil spill in the Gulf of Mexico.

Well Construction Automation Process

When general contractors build a home, they rely on detailed plans. When oil and gas contractors drill an exploratory oil well for, say ExxonMobil, or any other oil company, they use a well plan. A DWOP or ‘drilling well on paper’ was and likely remains the first significant step that convenes all parties involved to discuss a step-by-step plan of operations well beforehand. This process is immensely complex and involves a critical time path that must consider weather conditions, regulatory requirements, personnel considerations, geologic conditions, platform selection, support capabilities and configurations, and many other attributes. The DWOP is an ideal situation with complex and disaggregated data for applying a digital twin, and a firm named AnyLogic did just that and saved both time and money in doing so.

Well Construction Process Simulation
Simulation Based Digital Twin for Well Construction Process Optimization – AnyLogic Simulation Software

Digital twin technology success is not beholden to the oil and gas industry. While digital twins’ optimization benefits are seen mostly in industrial settings, their abilities are also being explored in healthcare, real estate, and many other industries. The critical threads amongst all those considering digital twin tech hinge not on the complexity of the process or industry, but, as noted below in a 2019 Boston Consulting Group white paper on the proper identification of “high-value use cases” and the understanding that success depends heavily on changing the way people work.

Digital twins are clearly a powerful tool and here to stay. What is the next step? Will they pivot out of their industrial wheelhouse? Tune in next time or dial into my twitter feed for more thoughts and insights on digital twins and …digital threads.


  1. Really great deep post about a topic not alot of people understand.

  2. Scott Siegler · ·

    I enjoyed reading this and learning a bit about a topic I had zero knowledge of beforehand. Given the importance of oil in day-to-day life, and also it’s profitability, I’m not surprised to learn that a cutting edge piece of equipment like a digital twin is the product of offshore drilling. However, this is a really cool example of how every company is becoming a “tech company.” It will be cool to see how this is adopted and utilized across industries.

    1. shaneriley88 · ·

      It will for sure! It’s amazing to learn about how seemingly black box-like neural networks are being applied to DTs. “We don’t know why the factory runs better, but we like it!”

  3. sayoyamusa · ·

    Excellent post, Shane! Digital twin technology was totally new to me, but your clear navigation was super helpful to understand this cool concept/tool. I also enjoyed your real work experiences!
    There seems a striking contrast between dynamic digital twin and static normal simulation. I think, as a marketer, my company can utilize DTs to test our new products or new advertisements and I feel excited that the new technology enables us to increase the probabilities of winning the market!
    Also, I agree that DTs have lots of possibilities for other industries. Health care is definitely the one as you mentioned. I’m positive this technology will bring several solutions for social problems such as mitigating damages of natural disasters.
    Your informative post has made me want to learn more and I’ve started to research other examples. I appreciate that you enlighten me on this new topic!

    1. shaneriley88 · ·

      Thank you! Your response warmed my heart so much I purchased a six-pack of Kirin Ichiban last night!

  4. changliu0601 · ·

    Excellent post!I like your diet soda example which is really helpful for me to understand the new knowledge.

    1. shaneriley88 · ·

      Thanks! I had to boil it down a bit to understand it myself.

  5. Great post Shane. I learned a lot from this post, really interesting technology and the various uses for it. I also appreciate the personal photos. Nice to see you out on the seas!

    1. shaneriley88 · ·

      Thanks, shipmate!

  6. courtneymba · ·

    Great article! I had never heard of a “digital twin” before the discussion yesterday. Not gonna lie, based on the name, I thought it was going to be more of a marketing exercise and finding my own personal twin in terms of profiles, preferences or looks lol! But this is cool too ha! As others mentioned, I also appreciate how you made the complex simple here and wove in your personal experience and relevant photo (it shows how massive and sophisticated the ships are!).

    1. shaneriley88 · ·

      Cheers! It was fun applying some new concepts to my old past.

  7. Chuyong Liu · ·

    Thank you Shane for educating me about Digital Twin. Although I have to admit it took me a long time to figure out how BOP and Well Construction works. But most importantly I enjoyed so much reading this blog, this is my first time hearing about digital twin. After some research, I found out that this technology is only starting to be noticed in 2017 in China in some article published by one of our best tech universities. This is a fantastic technology to be applied to complex operations and could save lives! Learning about digital twin really opened my horizon since I have never really looked at digital transformation in industries like oil drilling.

    I have a question that since digital twin is solving how businesses integrate their various operations and information technology, and it is analyzing data and making decisions. How is DT different from AI?

    1. shaneriley88 · ·

      Thank you and sorry for the delay!!!

      AI is more broad in nature. DT’s can be made up of the a handful of or many inputs (temp, weight, acidity..etc etc) – many of which might not be generated by an AI device. Most industrial DT’s rely on inputs from devices or things operated by humans (natural intelligence) Oddly enough, the oil and gas world is now using some pretty cutting edge DT’s with the latest AI (ex. neural networks) great quote from the attached article:

      “Even though the basic thermodynamics and chemical process have been understood for a long time…it’s this uncodified knowledge of how to do that,” . “They have intuition and certain heuristics that they apply to get it done. But no human could really ever take into account all 300 sensors and say, Okay, this is what should be done.”

%d bloggers like this: