What can a Digital Oilfield do for you ?
Steve Smart
Tech Innovation for Mining | Applied AI | Targeted Innovation | Process Improvement | Integrated Production Operations
Anyone who has worked in the Oil & Gas industry knows that we have used advanced electronics and computing to improve operational efficiency for a long time, particularly since the 1970s when the first real-time control systems were introduced.
This adoption has not been uniform however, and even today it is not uncommon to find software and systems of varying maturity levels in different functions within the same organisation. Indeed we often see essentially the same work being duplicated by different groups using different applications, not for cross-verification purposes but, simply because we are not familiar with (or confident in) the specifics of what other teams are doing.
Communication and collaboration between the large numbers of people involved in E&P continues to be a major management challenge. When correctly implemented, Digital Oilfield solutions have proven to be one of the most effective ways of addressing these issues.
But Digital Oilfield solutions need to be constructed in such a way that they get IT "out of the way", and allow subsurface teams to focus directly on the real reservoir, geology and petroleum engineering problems at hand.
What is a Digital Oilfield?
The Digital Oilfield relates to the use of advanced software and data analysis techniques to improve the profitability of oil & gas production operations.
The primary goal of a modern Digital Oilfield solution is to provide the subsurface teams responsible for production operations with an analysis and decision support environment that fosters collaboration and promotes operational efficiency through the seamless integration of data, models and workflows - particularly across functional teams.
What should my Digital Oilfield platform do for me?
While there are many proprietary and in-house/bespoke digital oilfield systems currently in use, the general capabilities that make a Digital Oilfield solution useful are:
1) Integration
The foundational element of a Digital Oilfield is a comprehensive data and system integration platform, that provides a means for easy access to all relevant data stored across disparate business systems. Integrated data commonly includes real-time field data, allocated production and production forecasts.
A user may simply choose to display this data in dashboards or reports, or they may use the data as inputs into engineering modelling calculations - possibly as part of a business workflow.
A tell-tale sign that a data integration platform is sub-optimal ("clunky") is where as an engineer you are in a situation where you need to get data from one application to another, and to do this all you can do is a manual export/import of data files (such as CSVs). This demonstrates inefficiency as it involves expert engineers effectively doing clerical IT tasks.
The real question we should ask is; why is the data not already available in our application or - at the very least - we should ask why we can't just import the data we need directly into our application. Users of the Digital Oilfield solution should expect IT to handle this kind of data integration for us behind the scenes.
2) Engineering Modelling
These capabilities of a Digital Oilfield offer computation functionalities, from simple unit conversion calculations through to the execution of dynamic reservoir models, steady-state and transient flow calculations through wellbore and surface flow networks, and PVT fluid equation-of-state calculations.
Engineering modelling capabilities maybe provided by 3rd party modelling applications, or through simpler mechanisms such as structured spreadsheets. With either approach it is important to have process in place to manage models and engineering computation versions, such that any changes can be audited and modelling results can be re-calculated at a later date - even if the model and/or computation version have subsequently changed.
In practice, connecting proprietary 3rd party modelling applications into Digital Oilfield platforms can prove difficult due to the unwillingness of vendors to fully expose their functionality through APIs or the lack of agreement on standard industry protocols - presumably for commercial reasons.
3) Workflow Engine
A Digital Oilfield platform's workflow engine provides a means of automating analysis or engineering modelling tasks. Workflows can either be driven by an event (for example, automatically perform an analysis when an anomaly is detected in the field) or on a schedule (for example, a business report may require the calculation of a daily aggregation of field data).
A particularly useful workflow component is a pattern recognition algorithm (aka 'Event Detection') that can be configured to identify significant events - typically leading indicators of changes in the reservoir or production system. This pattern recognition can be either be configured manually by subsurface teams, or with the aid of predictive analytics algorithms that assist with the identification of the leading indicators that best predict significant field events.
4) Analytics
An analytics package can provide sophisticated mathematical capabilities to a Digital Oilfield system. As well as advanced statistical analysis, these components can provide business intelligence capabilities (aggregations, roll-ups, drill downs), and 'Big Data' functionality (such as the application of machine learning algorithms on huge data sets)
5) User Interface
The final element of a modern Digital Oilfield platform is the User Interface. This provides the user with both advanced visualisation tools as well as a user-friendly way of interacting with the other elements of the platform (such as to access analytics capabilities or to visually design workflows).
The very best User Interfaces also enable team collaboration, and transparently handle information management issues such as user access, data security and versioning.
What do you do with a Digital Oilfield platform?
The astute reader will realise that everything described previously just provides a 'blank slate'. To go from a Digital Oilfield platform to a Digital Oilfield solution requires the user interface and visualisations to be designed, the necessary engineering models to be prepared, and business processes automated through workflows.
The fastest way to do this is through 'pre-canned' visualisation and workflow templates which are available commercially. An alternative approach to get started quickly is to base these on team experience or even classical charts and workflows in engineering textbooks. There are also workflows that are in the public domain. (for example there are closed-loop 'real-time production optimisation' workflows that are described in detail in SPE papers).
Ultimately a Digital Oilfield solution best achieves its objectives when visualisations and workflows are tailored to the real field context in an ongoing fashion. This allows subsurface teams to capture organisational learning and adapt their visualisations and workflows over time as the needs of field management change across the lifecycle.
To conclude there is a key caveat with Digital Oilfields: Just like all of the other software and devices we use - implementation is everything. A Digital Oilfield solution can tick all the boxes on paper, but the gold standard is how easily and seamlessly it enables Subsurface teams to do their jobs effectively.
Tech Innovation for Mining | Applied AI | Targeted Innovation | Process Improvement | Integrated Production Operations
9 年Hi César. I pretty much agree 100% with all the points you make. In particular, I agree that decoupling PE roles from IT is a terrible idea. In my experience, Digital Oilfields work best when they are led by PEs with strong IT skills - with IT Department staff acting in a supporting role when required. I also agree with you that Instrumentation is critically important. Indeed quality instrumentation is a pre-requisite. Finally, I agree that it is best practice to understand how a car works to drive one. Digital Oilfield systems should be transparent, and Engineers should ALWAYS be able to get to the raw data, manipulate it, and be able to see exactly how workflows and calculations execute. The point about "clerical IT tasks" was more about the 'default' mode of operation. I favor something similar to a modern car's automatic gearbox: Most of the time you drive in Auto, but if you want more control you can always change gears manually. I find the problem with clunky CSV export/import tasks is they are really the offer *the most basic form* of interoperability - and impede straightforward automation of workflows. To move to the Digital Oilfield of the future, we need to expect a lot more out-of-the-box from all of our vendors.
Tech Innovation for Mining | Applied AI | Targeted Innovation | Process Improvement | Integrated Production Operations
9 年Thank you for your kind words Fahd. And I'm glad I've caused some debate. I agree with you about the important contribution that Operational Technologies are making to our industry. Digital Oilfields and Operational Technologies certainly do work hand-in-hand. While the term Digital Oilfield could be expanded to include Operational Technologies, I believe it is useful for us to make a distinction. Operational Technologies tend to focus on discrete challenges within oilfields, whereas the term 'Digital Oilfield' as used in the literature is more about bringing all of the discrete parts together to facilitate collaboration and decision support; through the use of advanced software & data analysis, workflows, and integration across technologies and systems. The closest we have to an industry definition of the term can be found on the SPE site: https://petrowiki.org/Digital_oilfields
Data, Technology, Innovation, Markets | BSc Industrial Engineering. MSc Mechanical, EHS & Petroleum Engineering | Not a leader, nor a follower
9 年Good intro Steve. I have to agree with Fahd, though, the Digital Oilfield is far more than what we are currently advertising; he rightly talks about the core role of instrumentation, and I'd like to make a point about the IT role, as this is inevitably an IT-heavy discipline. Trying to decouple Petroleum Engineering from IT doesn't seem a good long-term strategy; integration and cross-pollination is a better bet. I've been told the usual "you don't need to understand how a car works to drive one"..well, you'd better do if you are going to be 24/7 on the road. It is certainly unfortunate that, as a discipline, we are less IT-literate than other engineering disciplines; don't know the reasons, but it's a fact. Current PE-IT decoupling approach only increases this skills' gap. While I agree it's not the most efficient use of expertise, those clerical IT tasks keep you in touch with the real data and all the associated transformation processes, developing or maintaining the sharp eye required to spot issues/errors in the process. Most users are told that DOF will enable them to skip those steps completely, and that's -in my opinion- both a mistake and a lie.