Few in the energy industry need to be convinced that artificial intelligence is changing the way in which the sector operates and offers the potential to speed up the energy transition – attendees at this month's Future Digital Twin & GenAI conferences in Amsterdam and Abu Dhabi will testify to the growing buzz around the topic.
But how to get the most out of AI, especially in the short-term, remains a matter for discussion. DNV says it could be a while before the full effects are felt, but they are likely to significant when they do come.
“We believe that AI is strongly subject to Amara’s law: its impact is overestimated in the short term and underestimated in the long run,” the consultancy says in its Energy Transition Outlook 2024, published in October.
This is seen as being the case, in particular, for generative AI and large language models, whose applicability in the energy sector, DNV notes, is currently limited to low-risk applications, such as productivity-enhancing tools for staff like MS copilot and experimental work.
More significant for oil and gas companies in the short-term will be so-called discriminative AI – applications like computer vision, forecasting, predictive maintenance and design optimization.
“This is in operation today, and much more is expected in the future. As methods evolve to deal with the black box issues and hallucinations, generative AI will increasingly be used in industrial applications,” DNV says.
Barriers to AI uptake, such as resistance to change and technical and cyber security issues, must still be surmounted. DNV also notes that the application of AI across oil and gas operations, where every site is geologically unique, can be more complex than for the renewables industry, where data and algorithms from one production site are often applicable to others.
Energy transition not on track
There seems little doubt that large-scale adoption of AI in operating energy systems is at what DNV calls an inflection point. That could feed into a speeding up of the energy transition if it helps to make operations of the oil and gas sector and energy intensive industries more efficient and reduce carbon emissions.
But AI is likely to play an even more significant role in terms of the energy transition through its application in power grids, energy storage and maximising output from and siting of renewable energy sources. For example, operation and maintenance of grid infrastructure handling a plethora of power inputs and outputs is likely to be transformed by AI given the predictive abilities of AI-driven applications when allied with other developments, such as digital twins.
It is an impact on the pace of energy transition that cannot come soon enough, given the DNV report’s downbeat headline figures on the future trajectory of global emissions. The consultancy says that 2024 is likely to be the year of peak energy emissions, as the world switches to solar and wind power and electric vehicle usage increases. But, DNV says a low decline in post-peak emissions are set to make key climate goals unattainable.
It forecasts global energy-related CO2 emissions will fall by only 5% by 2030, compared with 2023 levels, and that global CO2 emissions in 2050 will still be 17 gigatons, missing UN climate change targets and nowhere near the objective of hitting net zero global emissions by 2050.
“Cumulative emissions drive a global temperature increase, and we are most likely heading towards 2.2°C of global warming in 2100,” DNV says.
The findings chime with those of the International Energy Agency in its recent published World Energy Outlook 2024.
“Based on today’s policy settings, global carbon dioxide emissions are set to peak imminently, but the absence of a sharp decline after that means the world is on course for a rise of 2.4 °C in global average temperatures by the end of the century, well above the Paris Agreement goal of limiting global warming to 1.5 °C,” the IEA says.
(Photo: Shutterstock)
