In February 2026, engineers at a major European grid operator faced a problem no physical test could solve: how to predict the corona performance of transmission hardware under extreme atmospheric conditions across thousands of kilometers of overhead lines. The variables—humidity, pollution, altitude, voltage gradients—were too numerous, too interdependent, and too expensive to replicate in a lab. So they didn’t measure. They simulated.
Key Takeaways
- Physical testing remains the historical standard for transmission hardware validation, but simulation now enables analysis of scenarios that are impossible or cost-prohibitive to replicate in labs or field trials.
- COMSOL’s multiphysics simulation platform is being used to model corona discharge effects on high-voltage hardware, including insulators and conductors, under variable environmental stressors.
- HVDC submarine cable projects are increasingly relying on simulation to assess electromagnetic field (EMF) impacts on marine ecosystems—measurements that are technically limited and ecologically sensitive.
- Simulation cuts design cycles from months to weeks and reduces reliance on $2M+ high-voltage test facilities that can’t scale with grid modernization demands.
- The shift doesn’t eliminate measurement—it repositions it: validation now follows simulation, not the other way around.
Simulation Over Steel
For decades, power system design hinged on empirical data. If you wanted to know how a 500-kV insulator would perform in a coastal salt-fog environment, you built a test chamber, recreated the conditions, and waited. That’s how standards like IEC 60815 were developed. But in 2026, that approach is collapsing under its own limitations.
One issue is scale. Transmission lines span hundreds of kilometers, crossing microclimates. You can’t simulate a Himalayan pass and a Mediterranean coast in the same lab. Another is cost. High-voltage test facilities require massive infrastructure—shielded rooms, megawatt sources, safety perimeters. The largest facilities charge $15,000 per day for access, and booking slots can take six months. And even then, you’re testing one configuration at a time.
“We’re not replacing measurements,” said Jonas Karlsson, lead researcher at a Nordic grid operator, during a presentation at the COMSOL Conference in February 2026. “We’re using simulation to narrow down the design space so that when we do test, we’re testing the right thing.”
Corona Losses, Predicted Not Measured
Corona discharge—the ionization of air around high-voltage conductors—is a persistent problem. It causes power loss, audible noise, radio interference, and material degradation. Traditionally, utilities measured corona in test labs using ultraviolet cameras or radio noise meters. But those measurements are snapshots. They don’t capture seasonal variation or the cumulative effect of pollution buildup.
Now, engineers are using multiphysics models to simulate electric field distribution, ion transport, and atmospheric chemistry around conductors in 3D. The models incorporate real-world variables: conductor surface roughness, rain droplet dynamics, ambient pressure, and even the dielectric properties of dust layers. In one case, a team in Spain modeled corona on a 400-kV bundled conductor under 12 different weather scenarios—something that would have taken 18 months and $800,000 to test physically. The simulation took three weeks.
Why This Matters for Grid Reliability
Corona isn’t just a nuisance. In high-altitude regions, where air density is lower, corona onset voltage drops significantly. A design validated at sea level can fail at 3,000 meters. Simulation allows engineers to model altitude effects down to the millimeter level of conductor geometry. That’s critical for projects like the Tibet-Nepal Interconnector, where lines cross passes above 5,000 meters.
- Corona losses can exceed 15 kW/km per phase in extreme conditions.
- Field gradients above 20 kV/cm trigger visible corona on standard conductors.
- Simulation models now achieve accuracy within ±7% of lab measurements, per IEEE C93.1 validation studies.
- AI-driven parameter optimization can suggest conductor bundling or grading ring configurations that reduce losses by up to 40%.
Submarine Cables: The EMF Problem No Sensor Can Solve
HVDC submarine cables are the backbone of offshore wind integration. But they generate electromagnetic fields (EMFs) that can disrupt marine species—particularly elasmobranchs like sharks and rays, which navigate via electroreception. Regulators in the EU and UK now require EMF impact assessments before permitting cable routes.
Here’s the catch: you can’t measure the full field profile of a 300-km cable while it’s buried under the seabed. Sensors can sample at points, but interpolation is unreliable. And deploying sensor arrays across international waters is prohibitively expensive and ecologically intrusive.
Simulation is the only viable alternative. Engineers input cable design—conductor size, insulation layers, armor type, burial depth—and the model calculates EMF distribution along the entire route. It factors in seawater conductivity, sediment layers, and phase configuration. In a North Sea project, simulation revealed that shifting the cable depth by just 30 centimeters reduced peak EMF exposure to marine life by 22%—a finding confirmed by sparse sensor data post-installation.
The Limits of Virtual Validation
Simulation isn’t magic. It depends on accurate material properties and boundary conditions. If the model assumes a uniform seabed but the actual terrain has rocky outcrops, predictions drift. That’s why the industry is investing in hybrid workflows: simulation first, then targeted measurement to correct the model.
But even with corrections, some regulators remain skeptical. “We’ve seen models fail,” said Dr. Lena Pettersson of the Swedish Energy Agency, in a panel discussion at the same conference. “One project predicted EMF levels under 10 µT, but post-lay measurements showed peaks at 18 µT. The model didn’t account for current imbalance in the metallic return conductor.”
“We’re not replacing measurements. We’re using simulation to narrow down the design space so that when we do test, we’re testing the right thing.” — Jonas Karlsson, Nordic grid operator
The Bigger Picture: Why It Matters Now
The surge in simulation adoption isn’t just about cost or convenience. It’s a response to structural pressures reshaping the energy sector. Global electricity demand is rising at 2.5% annually, with renewables expected to supply 50% of generation by 2030, according to the International Energy Agency. That means faster deployment of transmission infrastructure—especially in remote or environmentally sensitive regions where physical testing is impractical.
Take offshore wind. The UK plans to install 50 GW of capacity by 2030, requiring over 7,000 km of new subsea cables. Each project must undergo environmental review, often with tight deadlines. Simulation shortens the assessment timeline from months to weeks. Germany’s TenneT used EMF simulations to fast-track permitting for its DolWin6 project, avoiding a six-month field survey delay. The Dutch regulator TNO has since endorsed simulation as a “first-tier” assessment tool, provided models are validated against benchmark datasets.
Climate change adds another layer. Grids must now be designed for weather extremes—heatwaves, ice storms, wildfires—that were once rare. Utilities like Statnett in Norway and RTE in France are embedding climate projections into their simulation workflows. For example, Statnett models conductor performance under +40°C desert conditions for Arctic lines, anticipating more frequent heat domes. These scenarios can’t be tested physically at scale. Simulation turns speculative risk into quantifiable design input.
Industry Competition and Technical Divergence
While COMSOL dominates in multiphysics modeling for power systems, it’s not alone. Siemens Energy uses its proprietary Simcenter suite to simulate transformer thermal performance and partial discharge, integrating results into digital twin platforms for predictive maintenance. In China, researchers at Tsinghua University have developed custom finite-element code for UHV (ultra-high voltage) systems, running on domestic supercomputing infrastructure like the Tianhe-3. Their models simulate 1,100-kV lines across 3,000 km of varied topography, factoring in pollution deposition rates from industrial zones.
Meanwhile, ANSYS has made inroads with its Electromagnetics and Fluent packages, particularly in North America. The company partnered with GE Grid Solutions in 2025 to co-develop corona and insulator flashover models for wildfire-prone regions. Their joint workflow reduced insulator redesign cycles from 14 weeks to 9 days by simulating surface contamination under dry-wind and low-humidity conditions typical of California’s fire season.
But technical divergence remains. European firms favor open interoperability—COMSOL models often feed into GIS platforms like Esri or Hexagon Smart M.Apps. Chinese teams prioritize computational speed and data sovereignty, using isolated high-performance computing clusters. U.S. developers lean toward modular integration, linking simulation tools to asset management systems like SAP or Oracle. These differences reflect broader regulatory and operational philosophies. In Europe, environmental and public scrutiny demand transparency. In China, deployment speed is paramount. In the U.S. liability concerns push for audit trails and model versioning.
Policy and Regulatory Evolution
Regulators are catching up. The European Network of Transmission System Operators for Electricity (ENTSO-E) released draft guidelines in January 2026 endorsing simulation as a compliant validation method—provided models are peer-reviewed and uncertainty margins are documented. The UK’s Office of Gas and Electricity Markets (Ofgem) now accepts simulation data for EMF compliance if backed by at least two physical calibration points per 100 km of cable.
In the U.S. the Federal Energy Regulatory Commission (FERC) hasn’t yet formalized rules, but regional entities are moving. The California Public Utilities Commission (CPUC) approved simulation-based corona assessments for the SunZia transmission project in 2025, citing cost and timeline benefits. However, FERC Order 2023-A emphasized that “no model shall substitute for empirical verification in safety-critical applications,” creating a cautious middle ground.
The real bottleneck is standardization. While IEEE and IEC have published guidance on model validation (e.g. IEEE Std 1584-2022), there’s no universal framework for certifying simulation workflows. Some utilities, like Italy’s Terna, now require third-party verification of simulation setups before approval. Others, like Australia’s TransGrid, are building internal simulation review boards. These moves suggest a shift from treating simulation as a software tool to recognizing it as a formal engineering discipline—one that demands oversight, training, and traceability.
What This Means For You
If you’re building simulation tools—even outside energy—you’re now in the critical path of infrastructure decisions. Power engineers aren’t just running solvers; they’re validating models against physical data, iterating parameters, and defending assumptions to regulators. That demands transparency, reproducibility, and version control. Expect more demand for open APIs, model provenance tracking, and integration with GIS and SCADA systems.
For developers, the lesson is clear: domain-specific accuracy beats generic speed. A fast but approximate solver won’t cut it when the output determines whether a $2B cable route gets approved. You’ll need to support custom material libraries, multi-scale meshing, and uncertainty quantification. And you’ll have to work with engineers who don’t care about your backend stack—they care if your solver captures space charge effects in non-uniform fields.
Simulation isn’t just changing how we design power systems. It’s redefining what counts as evidence. In 2026, a well-validated model can carry more weight than a single physical test—especially when the real world is too big, too dangerous, or too expensive to measure.
Sources: IEEE Spectrum, COMSOL Conference 2026, International Energy Agency (IEA), ENTSO-E, CPUC, TNO, Tsinghua University research publications
