Quantum computing for power-grid optimization: Junction Brief – November 2026

Acentric Research is produced with AI tools under human review. The authors are not financial professionals, and nothing here is investment advice or a recommendation to buy or sell any security. AI analysis can be wrong; check every claim against its cited source.

This Brief is an initial scan, not a full investigation. The method and scores behind it are in Energy × Quantum in the Landscape Report.

Overview

Power grids must decide, hour by hour, which plants run so that forecast demand is met at the lowest cost. The problem is known as Unit Commitment, and a trade report says it grows more complicated as grids add solar and wind power, which are not always predictable. [1] This Brief covers attempts to use quantum computers, together with ordinary computers, on grid problems such as this one.

The sources opened date from 2025 to 2026: two trade-press reports, one relaying statements by IonQ and Oak Ridge National Laboratory and one summarising three preprints that were not opened, and one preprint abstract. Peer-reviewed results were not opened. Later developments were not searched.

Current state

What the technology does. A hybrid quantum-classical method splits a problem between a quantum computer and an ordinary one. In July 2025 IonQ said it had applied its approach to the Unit Commitment problem, using its 36-qubit Forte Enterprise machine with classical computers to manage 26 power generators across 24 time intervals. [1] IonQ said this showed the feasibility of using quantum devices for grid scheduling. [1] A preprint on the same Unit Commitment problem describes a variational quantum algorithm combined with a classical heuristic. It tested systems of 3, 10 and 26 generating units, and says convergence was shown on IonQ's Forte system "for select time periods". [2]

Who is working on it. The work is part of GRID-Q, a multi-year Department of Energy project led by Oak Ridge National Laboratory with academic and private partners; IonQ, a maker of trapped-ion quantum computers, is one of two quantum firms involved. [1] The trade report says IonQ is also involved in a $22 million project with EPB, a Chattanooga utility. [1] It says IonQ and Oak Ridge's earlier collaborations covered quantum algorithm development, noise-tolerant computing and hardware design, and that IonQ's machines are available on major cloud platforms. [1] In October 2026 Quantum Computing Report summarised three preprints by EPB and IonQ, funded under a $4 million NIST grant, on reconfiguring distribution networks, the switches that decide how power flows through local lines. It says EPB's own staff ran the methods on an IonQ Forte Enterprise system installed at the EPB Quantum Center in Chattanooga, on networks of up to 417 buses. It says the three methods aim to reduce line losses and allocate switches in real time, and that on an IonQ run of a deep-reinforcement-learning method the result was the exact optimum for a 33-bus network and within 0.30% of the optimum for a 69-bus network. [3]

How far it has come. Demonstrated on small test problems. The July 2025 demonstration was a case study; Oak Ridge said it showed feasibility on a trapped-ion device. [1] The October 2026 report describes methods benchmarked on networks of up to 417 buses and executed on the IonQ system for networks that include 33 and 69 buses, with quantum sampling guiding classical power-flow software. [3] No source opened reports a grid operator using a quantum computer in operations.

Opportunities

Oak Ridge's lead for its part of the work said that research is intended to test how application performance can gain a quantum advantage as the device scales. [1] Quantum Computing Report says that on the 417-bus network, in one of three workflows, quantum-derived hints cut the median time a classical solver needed to come within 1% of target from 1,415.47 seconds to 34.71 seconds, with a fixed budget of 29 qubits. The same report says that on smaller networks total run time was dominated by overhead and that the end-to-end gap with unaided classical solvers only began to narrow on the 417-bus network; it does not report a net end-to-end speedup. [3]

Challenges

  • Size of the test cases. The 2025 demonstration used 26 generators and 24 periods on a 36-qubit machine. [1] Oak Ridge said research is intended to test how performance can gain a quantum advantage as the device scales. [1]
  • Total run time. Quantum Computing Report says that on small and medium networks, total run time was dominated by quantum circuit overhead and classical processing, and that the gap with unaided classical solvers only began to narrow on the 417-bus system. [3]
  • Part of the problem on hardware. The preprint's abstract says the full method was tested on three system sizes and its convergence on the quantum machine was shown for select time periods. [2]
  • Device noise. Quantum Computing Report says that in two workflows, sampling guided classical power-flow software to low-loss configurations "despite device noise". [3]

Data sources

The sources below are where a deeper investigation would look; those marked "listed" were not searched for this Brief.

  • arXiv, for the EPB–IonQ preprints and similar papers. [4]
  • DOE Office of Science and Oak Ridge programme pages, for GRID-Q and related projects. [5]
  • NIST grants pages, for the EPB award. [6]
  • SEC EDGAR, for filings of the quantum computing companies involved. [7]
  • No registry of quantum demonstrations on power systems was found.

Changelog

  • 2026-11-19: First published.

Sources

  1. The Quantum Insider, "IonQ Partners with Oak Ridge National Laboratory, Demonstrating Quantum Power Grid Optimization", 2025-07-31 (trade press relaying IonQ and Oak Ridge statements). thequantuminsider.com(accessed 2026-11-19)
  2. Aboumrad, W. et al., "A New Hybrid Quantum-Classical Algorithm for Solving the Unit Commitment Problem", arXiv:2505.00145, submitted 2025-04-30 (preprint by IonQ and Oak Ridge authors; abstract page). arxiv.org(accessed 2026-11-19)
  3. Quantum Computing Report, "EPB and IonQ Deploy Hybrid Trapped-Ion Algorithms on Forte Enterprise to Optimize Power Grid Topology", 2026-10-28 (trade press summarising three preprints; the preprints were not opened). quantumcomputingreport.com (accessed 2026-11-19)
  4. arXiv.org, preprint server. arxiv.org (listed as a data source; not searched for this Brief)
  5. Oak Ridge National Laboratory. ornl.gov (listed as a data source; not opened for this Brief)
  6. National Institute of Standards and Technology, grants and awards. nist.gov (listed as a data source; not opened for this Brief)
  7. SEC EDGAR, full-text search. sec.gov (listed as a data source; not run for this Brief)