DTU investigates power flow with quantum computing
Researchers from the Technical University of Denmark (TDU) claim a world first in using quantum computing for power flow modelling.
The focus of this work of the researchers from DTU’s Department of Wind and Energy Systems was on exploring the current capabilities of quantum computing for power flow studies.
Previously to the best of their knowledge, while quantum power flow algorithms have been proposed, these have been tested with simulated rather than real quantum computers.
For their investigations the researchers set up new algorithms based on the ‘fast decoupled load flow’ method.
Have you read?
New energy sector quantum computing use cases
E.ON to manage decentralized energy system using quantum computing
These were then tested on five different quantum computers from IBM, including four publicly available models and a fifth larger one to corroborate the results and evaluate scaling of the method – this latter important for future implementations of full-scale where the benefits of the computational power of quantum computing are expected to be more apparent.
In all cases the quantum power flow converged to the same solution as the classical method, although due the noise in the current quantum hardware, required over 30 iterations compared with the five for classical convergence.
Moreover, the larger system required a greater number of iterations due to a greater impact from the noise, indicating the need for the development of low noise quantum hardware for computations on larger systems, the researchers say.
“We have shown that current hardware is capable of performing a power flow for small test systems, but scalability is currently a major issue,” they write in their presentation.
While lower noise quantum computers are anticipated in years to come, activities for power system researchers in the meantime include the development of different approaches to the setting up of the algorithms and to the power flow computation to take advantage of the characteristics of quantum computing, eg around handling probability distributions.
Encouraged by their results the researchers intend to investigate how quantum computers can provide real-time analyzes of the stability of an electricity grid based on 100% renewable energy with large numbers of intermittent wind and solar PV generation systems.
They also intend to investigate what processes are best suited to quantum computing as a complement rather than replacement to traditional computing methods.
“The development is moving fast and I am sure that in the future we will perform calculations of the electricity system with a strong contribution from quantum computers,” says Brynjar Sævarsson, who carried out the calculations.
“They can do something that ordinary computers can’t. This means that we can develop the tools needed to safely and stably operate an electricity system based on renewable energy, and I am now starting that effort.”