OGUK’s recent UKCS Data & Digital Maturity Survey Report 2020 is a much needed snapshot on where the UKCS currently sits in terms of attitudes and actions towards digitalisation. Attitudinal surveys are a great way to understand how well industries are evolving towards digitalisation, as a lot of the work is delivered away from the limelight and value is rarely communicated. Unlike first oil, which can be empirically measured and celebrated, digitalisation initiatives in O&G provide value which is proprietary, difficult to quantify and typically only delivered in the medium-to-long-term.
An interesting quote from the survey report is as follows:
“With an organisation that has a coherent digital strategy supported by leadership, robust and accessible data and a workforce that fosters innovation and collaboration, the foundations for technology to provide its full value are in place”.
This is undoubtedly true, but ever since digitalisation became a viable option for UKCS operators in the early part of this decade, market conditions and industry structures have conspired against the long-term stability and collaboration needed for executing digital strategies.
The right assets in the right hands?
The report also highlighted the need for operators to drive change. Whilst this may work in the Norwegian Sector, where Equinor, AkerBP and Vår Energi see long-term supplier collaboration and digitalisation as key to creating value, the operator landscape in the UKCS is increasingly skewed towards lean, Private Equity backed asset holders, who use traditional approaches to create short-term returns as the basin further matures. Likewise, in the Gulf of Mexico, although the likes of Cantium Energy have used digitalisation to create a differentiated business model, operator strategies are similar to the UK but with the support of higher production volumes.
In addition, following recent asset rationalisations, it seems that the operators who have invested most heavily in digital initiatives are largely leaving the UKCS, taking some of their important know-how with them, but leaving a lot of it behind in the minds of production engineers, geologists and asset staff who are now working for less digitally inclined operators as contractors. This means in practice that everybody is using digital tools, but they are not necessarily used efficiently in well-designed workflows. The cohort of young “digital natives”, who it was hoped were going to bring fresh digital blood to Aberdeen have unfortunately seen their graduate programs chopped and headed off to different industries – probably never to return.
The death of end-to-end digital transformation
Operational digital transformation has remained an elusive goal, requiring a wholesale overhaul of strategy, data management, technology and workflows. It often hits some KPI’s on a management teams’ scorecards and needs significant spend on consultants. It could be argued that these two reasons, rather than practical gains, are why it received so much interest in the almost halcyon days of 2016-2019 when digital streams at conferences were well attended and the industry collectively committed to becoming data companies, not oil companies.
Back office digital transformation has been a success story for the UK E&P industry, as advanced ERP solutions, process automation and (semi) big data approaches suitable for manufacturing industries and large banks are relatively easy to implement in operators and supply chain companies. Vendors providing back office digital transformation solutions are orders of magnitude better resourced and capitalised than many operators, so have been able to absorb the development costs necessary to ensure that UKCS operators are broadly comparable to similar sized companies in other industries. Unfortunately, there are very few similar vendors with the products, know-how and knowledge necessary to bring operational digital transformation to UKCS operators, who have hugely diverse topside and subsea infrastructure, and hence challenges. iPads have been distributed to everybody, with the lucky drilling superintendent now able to check the progress of his cement job whilst reading his children their bedtime story, but no end-to-end digitalisation. The reality has been largely digitisation on the platforms, digitalisation for the back office - we should not confuse the two at our peril.
It is therefore fair to say that, in the current and future E&P environment, UKCS operators have neither the resources nor the intention to move from safely and efficiently operating mature assets to becoming global beacons of oilfield digital expertise through wholesale digital transformation. But there is another way that results can be realised.
A project-by-project approach for operational digital transformation
Perhaps luckily for the sanity of operator and supply chain staff the UKCS is now hopefully heading for the beta release of “digitalisation 2.0”, a more pragmatic approach where the focus will be on finding tangible value from data using readily available tools; with IT, asset and management staff all working together on a project-by-project journey to digitalisation using traditional project management approaches and state-of-the-art data management and analytics.
Projects focus on the most pressing challenges, and are funded accordingly
Spend on digitalisation competes with spend on production optimisation, reliability and exploration. From a portfolio analysis point of view, a large digital transformation programme will not necessarily compare favourably with a smaller digital project which aims to solve a specific asset challenge. A proof of concept study for using data to better manage corrosion can create easily implementable solutions and bring the right data to the right quality to the right people for more advanced analysis. It will also determine where workflows and processes can be streamlined and optimised – all tangible benefits at relatively low cost. Viewed from this perspective, making large upfront investments in big data lakes before the value is known is replaced by having the concept for the data lake in place, and filling the lake one project at a time.
Projects are defined, short and measurable
A six-month digital project is sufficient to implement Robotic Process Automation (RPA), create machine learning add-ins to existing software or to gain results from a supply chain data partnership. Tangible payback from short projects such as these create the building blocks for larger projects, and also the necessary understanding for more ambitious digital transformation programmes.
Projects build know-how, confidence, understanding and experience
Digitalisation projects require considerable involvement from all stakeholders. IT teams may not understand machine learning, drillers may not understand RPA and HSE teams may not understand natural language processing (NLP). Increasing this knowledge for everyone can only be achieved through experience during projects, as competing priorities will always put non-critical path digital training at the bottom of the to-do list. GE’s CEO stated in 2016 that he wanted every new hire to learn to code; but coding without purpose is a very inefficient and unguided way to build a digitally aware company. From a management perspective, learning to code is very different from learning what data analytics can achieve and what it requires.
Projects will overlap, reducing repetition in data management
The painful and manpower intensive process of data management, data cleansing and data preparation famously requires 80% of a project’s effort and should only be undertaken when absolutely necessary. A project creates this necessity and allows data to be partitioned into “project data” and “other”. Focusing on project data reduces the effort required to a minimum. Once a project has been successfully executed, the “other” data pool can be much more efficiently managed using the workflows and approaches developed under the first data project. Digital projects should spread data quality across an organisation like ink blots on a page. Eventually the result is that data transformation is achieved when all the individual project ink blots join up.
A balanced project portfolio builds into a knowledge-based digitalisation strategy
Every digitisation strategy should be built on a concrete knowledge of the value of digitalisation. This knowledge will only come from understanding your data, your capabilities and how additional value can be created. Projects give you this understanding. Only when projects have tangibly demonstrated what does and doesn’t work can an informed strategy based on your strengths can be developed. It is very likely that digital projects may already have taken place or are underway. A retrospective analysis of previous projects, putting them in a portfolio context using common performance metrics, can be a very useful exercise to understand how far “hidden” digitalisation is already happening in your organisation.
Project collaboration, data sharing and communication
Data sharing is essential for operational digitisation projects to be successful, and data sharing requires collaboration within an organisation and between organisations. The technical solutions to share data are mature and affordable, but the biggest challenge when sharing data is the uncertainty of outcome. A six-month project may result in nothing, or it may result in the IP for a digital service which could be commercialised into an industry leading company. Trying to contractually capture the range of possible outcomes pre-project would delight the law firms of Aberdeen but is more likely to end up in the potential project landing in the “too hard to do” legal bucket. The second largest perceived problem is data security, where the risks are self-evident but often over-stated.
Standard agreements are sufficient to start most data projects
Existing frameworks and supplier relationships can be used to deliver digital projects, where the risk/reward structure is clear. IP can be a sticking point but can be managed through realistic discussions on the IP value of what will be created in a project. Models and algorithms only have value when applied on a specific and well-understood data set, and IP can be protected with licensing agreements. Standard agreements, such as those used by the OGTC, are invaluable time savers which can make the difference between a project starting or not.
For most data projects IP will only start being generated after a number of months of data preparation and investigation. This allows a staged approach to IP to be taken – the first stage of a project should be scoped to finish at the point where IP will start to be generated, whilst accepting that although know-how has potential commercial value, it is difficult to legally protect. This allows projects to start and to pass through stage gates whilst the ultimate IP uncertainty is gradually being reduced. Once the IP implications are better understood, subsequent stages can be sanctioned based on an equitable IP agreement.
Data security needs flexibility based on understanding the project objectives
For a collaborative project between an operator, a well services contractor and a data company, an operator will only have data on their assets whilst their well services contractor may have data on 40 other operators. The data company will have internal IT infrastructure but will use a network of vendors to deliver their services. Digitalisation projects will often start with the exchange of Excel spreadsheets but will quickly progress to data subsets and data analysis outputs being exposed to 3rd party systems.
Much like with IP, having a staged project approach to data security allows all participants to track and agree to what data is shared, when it is shared and whom it is shared with. Trying to capture all eventualities at project start is very daunting, hence a progressive approach should be taken based on each participants data risk threshold. Data anonymisation and normalisation is an achievable task when dealing with data carefully chosen to achieve project objectives, but virtually impossible if applied to an exhaustive data set. For example, consider the data management approach for the entire National Data Repository (NDR) versus a subset of 1000 well logs. Analysis and processing can often be done on metadata – data about data – rather than proprietary source data, which can present a lower risk to data security. Understanding data flows as the project progresses is therefore key to managing data security.
Don’t hide data projects away
When anyone in the UKCS completes a data project, the results should be shared as widely and as openly as possible.
Endila’s AIForeSight™ database contains hundreds of case studies for where AI has been applied to investigate or solve E&P operational challenges by operators and researchers around the world. The knowhow developed in these projects was hard won and can be a very useful foundation for determining how a data driven project should be executed, and what approaches are most likely to work. Many of these case studies were reported in SPE papers but have now been largely forgotten. For every AI project which has an SPE paper there are undoubtedly another two which remain completely under the radar. If the industry fails to learn from other projects, then we will be locked in a cycle of constantly re-inventing the wheel and starting from a blank page.
Step-by-step to a digital future
We have to accept that the Norwegians have set a very high bar for the digital oilfield, and the UKCS is not going to achieve the same results with the same approach - our industry dynamics are too different. There is however a real and low-cost opportunity for a more basin-aligned stepwise approach through joining up smaller digitalisation projects which solve tangible operational challenges.
Although the Technology Leadership Board can push the UKCS in the right direction, the OGUK can encourage collaboration and the OGTC can provide fantastic resources to assist with project execution, the project concepts and willingness to attempt innovative digital projects must come from hard pressed asset teams and product line technology managers in the supply chain. Developing these ideas, sharing them across silos and discussing them with existing vendors creates the seeds which can be turned into projects. As the cliché goes, “the best time to plant a tree was 20 years ago – and the second-best time is now”. The UKCS operational and data community needs to spend time now planting enough seeds before worrying about how to manage the forest.
Through our sister data science company Endila, we have helped operators and OFS companies tackle digitalisation challenges. Our work has included the development of AI models for drilling reporting data, developing reporting dashboards for real-time data and analysing the value-add of digitalisation for supply chain companies.