Businessman using digital interface with holograms 3D rendering

Can digitalisation and AI accelerate the energy transition?

We now know that the energy system will need to change substantially in the coming decades. But what role will digital technology and AI play? And how do we get there faster?

By Dan Jeavons, Vice President for Digital Innovation and Computational Science on Jun 24, 2021

I believe that the next 10 years will be shaped by two mega trends –the energy transition and digitalisation.

The world faces an urgent challenge. How does it tackle climate change and move to a net-zero emissions energy system while also meeting the growing demand for energy? Fifty-one billion tonnes of greenhouse gas are added to the atmosphere each year. Climate science is clear that in order to stabilise climate change, CO2 emissions need to fall to zero by 2050.

Digital technologies can make it possible to design and operate entirely new energy systems at the device, plant and regional scales – transforming the way we manage the carbon footprint of industrial processes; digital technologies can provide the tools and mechanisms for optimising the energy efficiency of operations and enabling the sharing economy; they can enable more accurate greenhouse gas emissions tracking and transparent reporting across supply chains and can also enable more effective monitoring of carbon offsets. 

I believe the next wave of digital technology will be even more disruptive than what we have seen so far:

  • Sensor based technology is generating vast data sets which can now be processed in real-time using the ever-growing cloud capabilities. This is enabling us to observe and understand the physical world in new ways.
  • The development of augmented and virtual reality technology is allowing us to visually represent these vast datasets in the context of the physical reality that is being observed – creating digital twins of physical objects.
  • The development of AI technology allows us to interrogate these data sets, to simulate previously unforeseeable scenarios, to optimise processes, predict anomalies, identify objects and draw meaning from disparate data sources. Increasingly these models, rather than being purely data driven, are infused with scientific constraints. All of this enables us to enrich the digital twin.
  • Furthermore, the ability to provide verification and trust through distributed ledgers is making new collaborations possible.

Let me share a few examples of how these technologies are already making an impact for Shell and our customers.

Over decades we have built up deep knowledge of our industrial processes. We have aggregated our process data sets into cloud-based data stores – enabling us to use this data to develop solutions which optimise these processes. In our own plants, we have shown that optimisation technology can reduce the CO2 emissions of one of our LNG facilities by as much as 130 kilotons per year – the equivalent of taking 28,000 US cars off the road1 for a year.

We are also using these optimisation capabilities to accelerate research in clean energy technology. For example, in research into low carbon fuels we are using data-driven simulation effort combined with physics-based models to optimise efficiency and yields, reduce capital expenditure, and reduce time to market. We used this approach to demonstrate that sustainable aviation fuel concepts would work at scale. In 2020 we produced 500 litres of synthetic kerosene from carbon dioxide, water and renewable energy to replace conventional hydrocarbon feedstock. In a world first, the synthetic kerosene was blended with conventional jet fuel to power a KLM airlines passenger flight from Amsterdam to Madrid in early 2021.

We are working with companies, sector by sector to cut carbon emissions together. For example, we are working with Dalmia Cement in India, to identify and optimise pathways to reach net-zero emissions at one of their cement plants. With data-driven simulations, we can replicate current plant conditions to analyse the systemic impact of changes in technology, fuels or input materials. This was instrumental in quantifying the technological and financial impacts of each emissions mitigation option for their plant.

Digital and AI can also reduce emissions in the way we move around. Shell is working with customers and partners in the shipping industry to help accelerate decarbonisation towards a net-zero emissions future for shipping. In collaboration with the University of Southampton in the UK, we have developed a low cost, quick to deploy optimisation software that reduces greenhouse gas emissions and fuel costs for marine vessels: Shell’s Just Add Water System. It has been deployed on more than 50 of our ships so far, yielding up to a 7% reduction in emissions per ship. We are now working with Kongsberg Maritime to scale this up making the solution available across the industry.

The same principles can be applied to electric vehicles. Smart charging refers to the monitoring, management and control of electric vehicle charging stations with the goal of optimising energy consumption. We are bringing charging data together, to better understand user charging profiles, essentially helping us understand how customers charge their vehicles. Our smart charging algorithm helps to maintain a stable and balanced power grid by spreading charging demand, also enabling customers to charge at a lower cost when energy demand is lower. This aggregation of power demand enables the system to maximise the use of renewable energies when they are available. Today we are already deploying these smart charging capabilities for our customers in North America.

To do all this requires a different mindset and culture. To accelerate collaboration open source, standards and interfaces are becoming increasingly important. Too often in the energy industry, proprietary systems are the norm and data sharing is rare, this reduces the opportunities for smaller start-ups to participate in the ecosystem and also limits the ability of partners to collaborate easily across organisational boundaries. It is for this reason that Shell has pushed so hard – leading the industry in developing common data standards and platforms such as OSDU, Open Footprint Forum and the Open AI Energy Initiative. Using open data standards mean that businesses can reduce the time and effort required to collect and store their data. Instead, they can focus on analysing the data and developing new solutions, including to reduce their overall environmental footprint.

Shell’s deep knowledge of the energy system combined with some of the technology we are already developing gives me confidence that we are only scratching the surface of what is possible. I am excited by the impact our projects are already having. And most of all, I am excited about what comes next…

To hear more watch this keynote and interview below of Dan Jeavons with Bernard Marr, CEO & Co-Founder, Bernard Marr & Co at CogX 2021.

1 This is an illustrative conversion based on the conversion rate of the US Environmental Protection Agency.

The Planet & Smart Cities: Can digitalisation and AI accelerate the energy transition?

The Planet & Smart Cities: Can digitalisation and AI accelerate the energy transition?

The energy industry is going through a double transformation: one is decarbonisation, the other is digital transformation. We now know that the energy system will need to change substantially in the coming decades. But what role will digital technology and AI play? And how do we get there faster? Featuring: Dan Jeavons - VP Digitalisation and Computational Science - Shell Bernard Marr - CEO & Co-Founder - Bernard Marr & Co. (Moderator) #CogX2021

Transcript

[Moderator]

Hello everybody welcome to CogX day three and we are just about to have a very interesting discussion on digitalisation and the green transition. We're going to start off with a keynote from Dan Jeavons the VP of Digitalisation and Computational Science at Shell as part of the planet and smart cities track. Dan handing over to you.

[Dan Jeavons, VP Digitalisation and Computational Science]

Thank you so much for having me. It's an absolute delight to be back at CogX and it's great to be taking part in a hybrid event and to be here in person. I'm also absolutely delighted that I'm joined by Bernard Marr. Bernard and I had a conversation around ai and energy a couple of years ago at CogX. I'm really excited to be updating this conversation. As we look at the challenges that we face in the energy sector, I think it couldn't be more apt that today the theme of CogX is around getting the next 10 years right.

The reason I think that is such an exciting theme is that it's never more true that we need to get the next 10 years right than in the energy sector. It can't have escaped any of our notice that the climate challenge is real. Coming out of the Covid-19 pandemic the focus on solving some of the challenges that we have in the energy sector and the broader industrial sector couldn't be more in focus. To put that in context at present the emission the global emissions are about 51 billion tons of CO2 per annum and when we stopped flying during the f pandemic, and more or less stopped driving during 2020, we saw those emissions levels drop by around six percent. I think that puts in stark focus the challenge that faces all of us to truly transform the energy sector and the growing sense of urgency that those of us working in the sector feel to try to accelerate the pace of change.

I think it couldn't be more abundantly clear that part of the solution to this problem as well as behavioral change has to be technological innovation. We recognize that although there are going to be many aspects to solving this problem, which is going to involve collaboration of all parts of society, technology innovation is going to be a key part of driving the innovation and the change in the fundamental systems that power our society today. That's going to be necessary to achieve some of the net zero targets that we've set out.

Within that, I also want to focus on the role of digital technology. For me it's very clear that we can't argue that digital technology is going to solve energy transition. However, I do believe that digital technology has a very key role to play in helping accelerate the global pathway to net zero. During the rest of this talk I'm going to try and unpack a little bit of why I think digital technology is so important. But really, it has three key roles to play.

  1. Digital technology is helping us to design new energy and industrial processes that ultimately are going to operate with much lower CO2 footprints.
  2. Digital technology is also going to help us to optimise the existing energy system and to reduce its emissions; and perhaps most importantly
  3. It's going to help us to track the global footprint, and also to monitor and create transparency around the emissions that we all generate and the footprints that we're all responsible for.

So why do I have such faith in digital technology? I think the key point is right now in the digital domain we see the convergence of digital technology like never before which is further accelerating digital change. In particular I see that happening in the following key areas:

Sensor data is being generated all around us all the time whether that be through smart devices in your home or an increasing amount of sensors within our industrial facilities all of that data can now be aggregated in near real time in the cloud and made accessible in such a way that massive queries on full history of data can be run relatively easily on commodity hardware.

Alongside that we see the emergence of digital twin technology which is starting to leverage technology from games and also from artificial reality and virtual reality where effectively we are bringing together all of the data in context to create a digital representation of the physical asset and that the combination of virtual reality aspect in conjunction with the real data provides a true opportunity for AI where we can start to simulate and look at exceptions and create unknown conditions. As a result, we can start to optimise much more effectively the overall operations of a particular industrial facility.

And finally when you combine that with distributed ledger technology through blockchain you start to be able to create tracking across organizational boundaries in such a way that you can create the level of transparency that I believe is going to be needed in order to accelerate the sort of energy transition that I've outlined.

So, I see that convergence of technology as being absolutely fundamental but what i also want to do is talk about some tangible examples of things that are already happening that point us in this direction.

Over the past few years, we have been aggregating already our vast quantity of sensor data into an aggregated cloud environment that's now running at the multi-trillion row scale and we're using that to develop a whole series of new capabilities.

Within Shell, one example is that we're applying this to our traditional oil and gas processes and we've already seen the impact of that, for example within our LNG business. The deployment of one of these algorithms on an LNG train has demonstrated that we can reduce the CO2 footprint of that train by about 130 kilotons per annum through reduced flaring and that gives for example the impact of that is about the equivalent of taking 28,000 vehicles off the road. To put that in perspective, the deployment of this sort of AI, a single algorithm can have that sort of impact on CO2 emissions, but it goes beyond that.

We can also use these same approaches this combination of machine learning data driven simulation in context. Leveraging these vast quantities of data, also in conjunction with physics-based models in order to create new optimisation scenarios and to design new processes. Some of you may have seen recently that we were able to develop 500 litres of synthetic kerosene. This involved was removing CO2 from the air and turning it into sustainable aviation fuel that we used to fly a plane from Amsterdam down to Spain.

I think the key point there is that we're using now the AI technology to help us design and rapidly accelerate the deployment of technology like this at scale and in an affordable way to make these things commercially realisable.

A final example, we're also leveraging these capabilities more and more in the power space. As we build a power business through virtual power plants. And perhaps never more so than in our electric vehicles which we see cropping up all over our cities now and in the number of charge points 80,000 to date with access to about 225,000 that Shell is operating. We have ambitions to grow beyond that to about half a million charge points by 2025. So, if you get look at all of this is in the electric mobility space, we're going to see a rapid acceleration of these capabilities. Of course, all of that new data creates an opportunity for AI and we're starting to develop that to optimise the charging processes to ensure that when we plug in the vehicles are charged in the most energy efficient manner, leveraging the maximum amount of renewables and we're deploying that technology already to customers in the US today.

So, I've covered a lot of ground in a very short period of time. I'd love to talk about this for longer. In closing I will talk about the overall approach. I think we have to recognise that this is a radical change it's a radical change in culture for an organisation like ours.

It's also a radical transformation in the energy system and in the infrastructure that we operate all over the world. To do that we have to work differently and i think most importantly we have to work together.

One of the things that we've been driving in Shell is openness radical openness at new levels. We've led the industry in what we call the OSDU, which is an open energy data platform designed to allow easy exchange and to avoid proprietary formats. We recognise that that open data exchange is going to be critical to powering the capabilities that I've described today.

We also launched the Open AI Energy Initiative to allow the exchange of IP between major energy providers to accelerate the adoption of ai and in turn to accelerate the embedding of ai in the energy system. Of course, we're trying to also leverage this in the emissions calculation space through our Open Footprint Forum where we're trying to encourage open standards for the calculation of CO2 footprints.

I want to close with a degree of hope. I think we all have to recognise that solving the energy problem in a world where the demand for energy is increasing and where energy is part of assuring the quality of life that we all enjoy.

We have to recognise that transforming that system is going to take time it's going to be challenging and it's going to take collaboration and it's probably the largest problem that our generation faces. But I also feel that there is hope.

I see the level of collaboration growing within the energy industry to embrace the need to change. I also see the cultural shift that's needed to help to accelerate some of the transformations I've talked about and the adoption of technology that is going to be needed to transform our energy system.

I hope that today I've given you a flavour of what we're trying to do.

Now I'm going to hand over to Bernard because I'd love to hear his perspectives. 

Dan Jeavons is the Vice-President for Digitalisation and Computational Science. He has been a key contributor to Shell’s digitalisation transformation. 

In 2020 he was listed in the Constellation Research #BT150 and Truata’s Top 100 Data Visionaries.

 

 

Shell.ai

From machine learning to computer vision, deep learning to virtual assistants and autonomous vehicles to robotics, Shell has been focused on a range of technologies that have supported advances in AI.

Find out more

More in Digitalisation

Digital technologies in the energy industry

Shell is a pioneer in the development and deployment of many digital technologies.

News Room

Read the latest announcements and news to discover what our teams have recently accomplished.