Artificial Intelligence and the current Energy Crisis

As the world is emerging from COVID-19 lockdowns it is now faced with congested supply chains and an emerging energy crisis. Energy firms are dealing with the complexity of a changing energy mix, shifting demands and volatile politics making the future less linear than in the past. Artificial Intelligence (AI) offers immense potential to respond to market volatility but is faced with adoption challenges. This article outlines practical approaches organisations can do now to prepare.  

Market Overview

The global price of gas started to soar during last year's long winter, with gas fields producing at maximum capacity. Meanwhile, renewable energy production lowered, as the weather was less windy over the summer causing more gas to be burned for electricity. As gas prices soar so does overall energy prices as gas is a significant source of electricity - a commodity marginally priced. Similarly, oil prices are now hovering around US$100 for the first time since 2014, while renewable energy capacity is increasing, intermittence remains a significant challenge to balance supply with demand.

Supply-chain transportation bottlenecks and the sharp rise in carbon and energy costs have significantly driven up the cost of scarce metals that have low base inventory levels such as steel, copper, nickel, aluminium, and lithium that are critical components in the energy transition. This all occurred at a time when demand increased as the world came out of successive lockdowns as businesses try to recover.

During all this volatility global pressure from consumers, institutional investors and governments increase on businesses to reduce their carbon footprint. Energy prices have always had volatility from geo-politics, but the conflict in Ukraine is creating further uncertainty as European energy demand is critically dependent on gas supply difficult to substitute in the short term.

Energy will remain strategically critical, and states, companies and people are positioning themselves for a more volatile and complex future which in turn is increasing pressure on the scarce resources required to produce, deliver, and consume energy.

The Artificial Intelligence Opportunity

The increasing use of electric vehicles and renewable, fluctuating power sources, such as wind and solar energy, has made it more challenging to forecast the demand on scarce metals, logistics requirements and energy prices. AI algorithms (such as machine learning and deep learning) enable valuable insights that can enable Energy providers and regulators to make more accurate forecasts, advanced scenarios, and stronger energy policies. AI solutions will provide more resilient energy strategies, save money, reduce emissions, and improves safety for the company and its customers.

AI algorithms can identify patterns, provide greater granular insights within large sets of historical and real-time data, and predict outcomes given certain data inputs that are self-learning, leveraging feedback loops to improve accuracy.

This enables energy companies and regulators to better anticipate/simulate energy consumption, consumer energy demand, weather conditions, material availability, carbon emissions and logistics constraints for optimal scenario projections. They also increase accuracy in the short-term by improving production decision-making, enhancing dispatch efficiency, and reducing required operating reserves.

Consumers can benefit from AI in the ability to reduce energy bills through the optimisation of solar and battery systems, while also shifting loads to periods when grids are less congested.

Integrated AI systems within workflows enables prescriptive insights; not just for forecasting and scenario planning but for advising on specific tactics such as when to buy/sell/source energy, metals, materials, and services.

These are just some of today’s benefits, and when looking ahead it is expected that more AI algorithms are converged with autonomous energy systems, that are more tightly integrated with smart city infrastructure for unprecedent efficiency, creating a more dynamic market that balances between favourable and unfavourable conditions. 

While AI offers immense potential it does come with its own adoption challenges.

AI Adoption Challenges

AI is known to accelerate a world with less energy volatility, yet 90 percent of organisations have difficulty scaling AI across their enterprises, due to three crucial areas.

• Lack of in-depth knowledge: It is difficult to find organisations with the required expertise to build holistic AI-powered software solutions that have real practical industry value.
• Mistrust: in AI from sceptical users at the organisation and individual level, due to their inherent “black-box” nature.
• Vulnerability: to cyber-attacks that creates inherent data security and governance concerns leading many AI systems to be flagged as high-risk.

AI Adoption Recommendations

Energy organisations need to begin preparing today for a future of rapid volatility, ensuring they are well placed to leverage the value of AI. Below are three key areas to start with.

• AI Vision: Incorporates AI within an organisation’s energy transition vision, leveraging approaches that focus on demand driven forecasts, decarbonisation, and supply-chain optimisation. A well-crafted vision should be compelling, clear and provides complete redefinition of operating models across the entire energy value chain.
• Explainable AI: As AI advances towards autonomous systems that perceive, learn, decide, and act on their own. The effectiveness of adoption will depend on building AI trust. This requires systems that convey their understanding to users on the rationale for decision making and deep industry and AI domain expertise.
• Digital Architecture: “Building block” architectures are required due to the agile nature of AI systems. They are required to ensure that systems are reliable, robust, scalable, and secure via the appropriate access control, data protection, integrations, governance, and secure code.