AI as a tool to offset electrical power scarcity
Written by Jens Eriksvik & Johan Hallberg Szabadváry
Sweden's major population centers, including Gothenburg, Stockholm, and Malmö, are faces with a looming threat of power shortages due to capacity constraints in the national grid. Property owners, transportation sectors, and heavy industries will face challenges to drive their business. AI is part of the toolbox to solve this - but getting started is key.
"The role of AI in meeting Sweden's energy challenges is essential. Reliable forecasting of future needs allows for optimal use of the available resources. Predictive maintenance of local and national grids will be essential for reliable energy supply. AI empowers stakeholders to make optimal decisions under uncertainty to build a more robust and efficient energy system. Innovative business models such as dynamic tariff models and local flex can incentivize active energy management using AI to reliably predict and avoid load spikes. However, using AI comes with its own energy demands. Properly choosing model size and complexity to solve the problem at hand will be essential to ensure a sustainable energy future."
- Johan Hallberg Szabadváry, Data scientist at Algorithma, Industrial PhD student
The increasing demand for electricity, driven by electrification and industrial transformation, is putting additional strain on an already constrained power supply. This situation, coupled with the risk of prolonged outages and price volatility, creates a need for innovative solutions to manage and optimize power usage across these sectors. AI-powered solutions are emerging as a vital tool to help navigate this complex energy landscape and ensure a stable and cost-effective power supply.
Main drivers of electrical power scarcity
Grid capacity constraints: The primary cause of power scarcity in major Swedish cities is not local grid limitations but rather bottlenecks in the overarching electricity network. The transmission grid, which moves electricity from production centers in the north to consumption centers in the south, is currently insufficient to meet the growing demand. Efforts are underway to expand and renew the transmission infrastructure to alleviate these bottlenecks (Holmberg & Tangerås, 2023). Additionally, there is an urgent need for significant investments in grid infrastructure to accommodate the increasing load from electrified transport systems, requiring coordinated efforts between public and private sectors (Volvo Group).
Increased demand from electrification: As Sweden transitions towards a more electrified society, the demand for electricity is rising significantly across multiple sectors:
Property: Ensuring stable power supply for buildings and tenants amid shortages.
Transportation: The electrification of transport, such as EVs and electric road systems, is increasing the load on the grid (Smart City Sweden).
Heavy industry: The production of green hydrogen for industries like steel manufacturing could increase electricity demand by up to 55 TWh per year—40% of Sweden's current consumption (Holmberg & Tangerås, 2023).
Risk of electricity shortages: Throughout the 2020s, there is a risk of electricity shortages in Sweden's major cities, particularly Stockholm and Uppsala. These shortages can arise from various factors, including low hydro power reservoir levels, disruptions in nuclear power generation, and limitations in transmission lines (Holmberg & Tangerås, 2023). Additionally, the closure of nuclear reactors and increased reliance on intermittent renewable sources like wind and solar, without sufficient backup, have made Sweden more vulnerable to electricity shortages and high costs, particularly in southern Sweden. Integration with the continental grid has further compounded these risks by exposing Sweden to external supply fluctuations (Engelsberg Ideas).
Price volatility: The ongoing energy crisis has led to increased electricity costs for consumers, including property owners, transportation sectors, and heavy industries. The shift toward variable renewable energy sources, such as wind and solar, contributes to this price volatility due to their intermittent nature (Holmberg & Tangerås, 2023).
As Sweden accelerates its transition toward electrification, several sectors face unique challenges in managing their energy needs. Property owners must secure reliable power supplies for their buildings while controlling costs and adjusting to a changing energy environment. Meanwhile, the transportation sector's rapid shift to electric vehicles intensifies grid demand, necessitating innovative solutions like smart charging and grid integration. Heavy industries are also confronted with rising electricity needs, such as for green hydrogen production, all while striving to maintain operational efficiency and stability (Holmberg & Tangerås, 2023; Smart City Sweden).
AI as (part of) a solution
Addressing Sweden's energy challenges requires a comprehensive solution that combines several key components. These include targeted investments in grid infrastructure, fostering public-private partnerships, aligning policies with technological advancements, and developing innovative business models.
"To address the power scarcity in Sweden, it is essential to start using AI today. AI is a powerful tool—it optimizes energy use, predicts future needs, and strengthens grid reliability. The key is to take action early, combining these AI capabilities with investments in infrastructure and new business models to navigate this complex landscape and ensure a stable energy supply for the future."
- Jens Eriksvik, CEO at Algorithma
A critical part of this strategy is AI, which enhances power management by optimizing energy use, predicting future needs, and improving grid reliability. AI also supports dynamic tariff models, decentralized grid management, and interactive simulations, helping stakeholders make informed decisions and build a more resilient and efficient energy system. AI-powered solutions can help these sectors navigate their respective challenges by:
Optimizing power usage: AI can analyze energy consumption patterns to identify opportunities for efficiency improvements, load shifting, and demand response, maximizing the use of available resources and managing decentralized energy systems by optimizing local renewable energy production and reducing dependency on the national grid (AI Sweden).
Predictive capabilities: AI can help anticipate future energy needs by forecasting energy requirements, enabling property owners, transportation planners, and industrial leaders to plan for infrastructure upgrades and investments in renewable energy sources (AI Sweden).
Smart grid maintenance: AI-powered systems, like Exeri's Smart Grid Surveillance, use smart sensors and algorithms to detect and locate faults in the grid, enabling efficient and proactive maintenance, reducing downtime, and improving grid reliability (Exeri).
Dynamic tariff models: AI supports the development of dynamic tariff models that reflect real-time grid conditions and customer behavior, incentivizing active energy management, benefiting both customers and the grid by reducing the need to oversize the network (Plexigrid).
Decentralized management: AI algorithms provide decentralized management of power grids, maintaining voltage levels within required limits and preventing fluctuations caused by integrating renewable energy and EVs, ensuring grid stability without centralized control (AI Sweden).
Interactive simulations: Tools like Behovskartan use AI for interactive simulations of the Swedish energy system, allowing stakeholders to visualize and understand electricity production and consumption, building a more resilient and flexible electrical system (AI Sweden).
At Jönköping Energi, Algorithma’s subject matter expert Johan Hallberg Szabadváry, participates in an Industrial research project called “Local flex”, where the aim is to avoid demand spikes in the grid by creating a market where customers can sell options to reduce their electricity usage for a limited time. Our task is to predict the future demand far enough ahead to be able to buy enough flexibility to keep the load at a manageable level. Electricity demand forecasting is well studied, and many machine learning (ML) algorithms exist to deal with the problem. In our problem setting, we would ideally like to predict the probability of seeing a demand that is below a threshold that defines the need to buy flex. These probabilities can then be used to make optimal decisions under the remaining uncertainty. Many ML algorithms can do this, but the challenge is that these probabilities are often poorly calibrated, in the sense that they do not reflect the actual outcomes. Our research is therefore aimed at producing well calibrated, reliable probabilistic forecasts that can be used for optimal decision-making.
In conclusion, by leveraging AI capabilities and strategic guidance, organizations can address complex challenges in energy management, improve efficiency, and optimize resource use across sectors.
Getting started
1. Understand your data landscape: Begin by assessing the quality, availability, and relevance of your data. Identify key areas where data can drive decisions, such as predictive maintenance or supply chain optimization (Algorithma: AI in Predictive Manufacturing).
2. Build the algorithmic business: Develop a clear AI and data strategy that aligns with your business objectives, defining use cases, establishing data governance, and fostering a culture of continuous improvement (Algorithma: Building the Algorithmic Business).
3. Machine learning and optimization: Integrate machine learning and optimization techniques into decision support systems to enhance data-driven decision-making and operational efficiency (Algorithma: Getting Value from Machine Learning and Optimization).
By following these steps, businesses can begin their journey toward becoming more data-driven and AI-enabled, ensuring they are well-prepared to handle future challenges and opportunities.
What we do
We build algorithmic businesses. Algorithma is a strategic advisory and AI consulting firm, empowers businesses across industries, incl. property, transportation, and heavy industry sectors to unlock transformative value through innovation, strategy, and technology-driven transformation. By leveraging our expertise in AI-driven optimization, digital transformation, and advanced analytics, Algorithma helps clients develop tailored AI solutions that optimize power usage, manage costs, and anticipate future energy needs. Our end-to-end support, from strategy to implementation, ensures that our clients can harness the power of technology for sustainable growth and success.
Visit Algorithma's website for more insights.
