In the energy transition conversation, discussions tend to center on where the energy of the future will be sourced and on how to deal with future emissions – but data is vital to any progress in these areas and therefore machine learning (ML) and artificial intelligence (AI) are key elements of the energy transition. According to the International Energy Authority, reaching net zero by 2050: “requires nothing short of a total transformation of the energy systems that underpin our economies” and effective use of data will play a crucial role in this transformation.
We know, for example, that the new energy world will be much more flexible than it is at present, with demand being fed from a wide variety of sources. The ability to switch between these sources quickly and efficiently will require the intelligent use of large quantities of digitized information, varying from weather reports to mechanical data, in order to maximize yields from wind and solar farms and balance generation and consumption locally.
As many oil and gas companies have realized, digital technologies will be key to achieving their net zero targets. These range from using predictive analysis to ensure the efficiency of their operations, to developing more accurate subsurface models to make maximum use of resources and to facilitate emission-reducing strategies like carbon capture and storage. As an example of progress already made, one supermajor recently described how an algorithm introduced to reduce emissions has resulted in improved production data, justifying the investment in AI.
Digitalization and Standardization
At the forefront of the movement towards the more efficient use of data is the need to transfer as much of it as possible into digitalized formats that will be accessible to and readable by a range of users. The oil and gas industry has always generated large quantities of data, but often in formats attached to specific applications and organizations. The pressure for standardization is therefore increasing, with companies and organizations collaborating globally to create common data standards. The Open Subsurface Data Universe Forum is a case in point. It comprises over 200 organizations from all over the world, ranging from supermajor oil companies and giant consultancies like Accenture and Ernst & Young to established subsurface service organizations and smaller cross-industry tech businesses, as well as data specialists. Its aim is to deliver an “open source, standards based, technology-agnostic data platform for the energy industry that stimulates innovation, industrializes data management and reduces time to market for new solutions.”
It is important that the oil and gas industry makes full use of its existing data as well as gathers information from new sectors, as ML and AI need access to good quality digitized data, historical as well as current – but this can create huge data storage issues. Governmental and regulatory organizations like the UK’s Oil and Gas Authority (OGA) are increasingly becoming involved by collecting data and making it available so companies no longer need to store it. Teams and partners can thus all access the same data without having to download and make multiple copies. Collaborative data can also be used to look at the wider energy sector to understand how energy assets can be more integrated, by, for example, using nearby wind farms to power off shore installations and thus reduce that installation’s carbon footprint.
Trust and Openness: the New Normal
As this demonstrates, a key factor in this new digitized energy world is the need for collaboration across a diverse range of companies and organizations. The energy industry cannot fully embrace digitalization alone and nor can individual companies; even a company as big as Shell admits it can afford neither the cost nor the time of ‘going it alone’ in this complex field, so new ways of working both between companies and across industries are paramount. At the same time, collaborating is not always easy; systems have to protect both business and financial interests, whilst ensuring that important data that could, for example, have safety implications, is accessible to all who need it.
Sourcing data and applications through the Cloud encourages collaboration and has proved to be cost efficient for the oil industry, as companies can access computer resources as they need them, rather than purchase heavy duty software without knowing their future requirements. Accessing data in this manner usually proves more efficient, as the geoscientists need to spend less time organizing and managing data and more time analyzing it.
Using the Right Data
With all this data-gathering and sharing, it will be important to know that we are collecting both the right data and the right amount of data for our decision-making. There is always a possibility that using digital technology simply generates yet more data, without it necessarily being useful, but AI can be used to extract patterns and help show which data is best at finding the most efficient methods, not just operationally but also in digitalization, helping organizations make better decisions. With limited time to make the changes needed to achieve net zero, it is important that we focus on identifying and working on those areas that can make the most impact in the energy transition.
One of the areas in which AI can significantly help O&G companies to reach their net zero targets is emissions information. Traditionally, this data has been collected at the end point in facilities, but it is better to gather data throughout the process, which will require installing additional sensors to look at a wider range of information. In this way, instead of just concentrating on production data when considering emissions, we would be able to interrogate the whole process and include additional data such as maintenance, downtime and access to spare parts. Manipulating that data will not only give us a clearer picture of the emissions involved; it will also point to ways in which those emissions can be quantified and controlled throughout the process. The resultant data needs to be standardized, so it can be read across a variety of applications and comparisons can be more easily made.
Oil companies with ambitious net zero targets are using AI and ML to look beyond just production as they gather huge amounts of data in a wide range of areas to analyze emissions across the range of their activities. For example, they will need to include improving emission levels from ships carrying their products across the globe; or if they decide on an EV fleet of road tankers, how will the electricity to power them be sourced and used and what is the carbon footprint of the construction of the vehicles?
Major technological changes in any industry are never easy and using AI and ML in the energy transition is no different. Many questions are being asked, particularly around the speed, complexity and cost of the process. A much greater range of data than has been traditionally used will be generated in vast quantities – and even with AI, much more time will be spent in analyzing the data.
One of the most important issues is the huge amount of energy that will be needed just to power the giant computers needed for digitalization and AI; could we actually be making the problem worse? Certainly, training complex AI models consumes a lot of energy, but the savings generated from optimizing processes should outweigh the consumption. The process of creating a digital twin of a facility, for example, will obviously result in higher emissions, but the evidence so far suggests that aggregate emissions for the project will be lower.
To make any change effective, we need to have the right people with the best training available and also to ensure that everyone is involved in the process and understands the importance of what is happening and how to use the new technologies to make better decisions. Skills are a key part of the success of this transformation; not only ensuring that new people come into the O&G industry with the relevant expertise and a different mindset, but also helping existing employees gather the skills they will require for the future.
One area for concern is that increased collaboration and remote access heightens the potential for cyberattacks and data leaks, requiring increased security, which must be taken into account when considering the cost of digitalization.
The Way Forward
The need for digitalization, ML and AI in the petroleum industry has developed a new urgency as operators and the services sector strive to meet their ambitious net zero targets. However, one unexpected advantage of the Covid-19 pandemic is that it has accelerated the speed and lowered the cost of digitalization and AI throughout the world. It has also proved the veracity of the old adage ‘necessity is the mother of invention’; technological innovations have made great strides and many of these will be applicable to helping the oil and gas industry move through the energy transition.