Secure Power System Operation based on Quantum Computations

Power systems worldwide are undergoing a profound transformation, driven by the urgent need to reduce greenhouse gas emissions, rising fuel costs, and increasing reliance on renewable energy sources. Although renewable technologies such as wind and solar offer cleaner alternatives to fossil fuels, their intermittent and weather-dependent nature introduces significant operational challenges. These include reduced system inertia, altered dynamic behavior, and greater difficulty in ensuring grid stability and control.

Traditional methods for power system simulation and optimization, often based on simplified models and assumptions, are becoming increasingly inadequate. Moreover, power flow analysis under uncertain and stochastic conditions is becoming more complex due to growing electrification and distributed energy generation, and increasing variability in consumption patterns. As power systems operate closer to their physical limits, the need for scalable high-performance tools capable of evaluating millions of scenarios becomes critical. Blackouts are extremely costly and disruptive to society; power grids must therefore be resilient against faults, not just single faults but even combinations of several faults. Solving these kinds of combinatorial problems is challenging and will become even more difficult as the number of system components keeps increasing. Addressing the challenges facing future power grids requires a paradigm shift in computational approaches, pushing the boundaries of current hardware and algorithms to enable secure, efficient, and real-time power system operation.

Quantum computing is an emerging and rapidly developing field that offers novel computational paradigms. Unlike classical computers, which process data in bits, quantum computers use qubits that can exist in superposition and become quantum entangled, allowing them to explore vast solution spaces simultaneously. This inherent parallelism positions quantum computing as a potential game-changer for some of the most computationally intensive problems in power system analysis.

Although the use of quantum computing in power systems remains in its early research stages, this thesis aims to explore several key applications where quantum algorithms, those with demonstrated advantages over classical counterparts, can offer significant value.

This thesis introduces the use of quantum computations for power system applications and summarizes three research articles produced throughout the PhD study. In each article, we develop a quantum computing based application for power system analysis and demonstrate its potential through both simulated and real quantum computing.

The first article explores quantum computing to solve a power flow, the most fundamental algorithm used in power system analysis, and the developed application is, to the best of our knowledge, the first quantum AC power flow that goes beyond simulations and is successfully executed on real quantum hardware. We based this approach on the HHL quantum algorithm, which can potentially solve a system of linear equations exponentially faster than classical methods. However, a number of challenges are faced in this first approach, indicating that significant improvements in quantum hardware are required before this type of approach could become beneficial.

The second article, where instead of essentially attempting to run a classical power algorithm on a quantum computer, we explore how one of the unique features of quantum states could be used for power flow analysis. The article introduces the first quantum computing framework for Stochastic Quantum Power Flow analysis in power systems. Here we use the probabilistic nature of quantum states to represent uncertainty in power injections and develop an approach to estimate the risk of line overloading. The developed algorithm creates a quantum circuit which maps generator injection distributions to line flow distributions so that quantum Monte Carlo sampling, which has up to quadratic advantage over classical methods, can be applied.

Finally, we develop a hybrid quantum-classical approach to identify the most critical combination of line outages in a power grid. This N-k contingency assessment is formulated as a combinatorial optimization problem which is split into a binary master-problem and a continuous sub-problem using benders decomposition. The master-problem is then solved using a quantum computer, while the sub-problem is solved classically. Assessing every scenario in an N-k contingency assessment quickly becomes infeasible for large power systems as k increases. Here we propose the use of quantum computing, and our results show that quantum computing has the potential to solve larger problems than purely classical methods. While the preliminary results of this study are very promising, more research is needed in this area before any quantum advantage is claimed.

Collectively, these studies provide a foundation for future research in an emerging new research field, i.e. quantum-enhanced power system computation. Quantum based applications will be defined by hybridization, reformulation, and forward-looking experimentation. The studies reveal that achieving a quantum advantage in power system analysis will not come from transplanting classical algorithms into quantum form but from rethinking how power system problems are expressed, exploiting quantum phenomena like superposition, entanglement, and probabilistic encoding. While full-scale, fault-tolerant quantum hardware remains a future prospect, the groundwork is being laid today. Through these pioneering efforts, the contours of tomorrow’s grid computation landscape begin to take shape: one where quantum and classical computations interact seamlessly to model, analyze, and secure the increasingly complex and uncertain power systems of the 21st century.