Introduction to Snow and Ice Accumulation on Overhead Conductors

Winter storms pose a significant threat to overhead power lines. The accumulation of snow and ice adds substantial weight, increases aerodynamic loads, and can lead to conductor galloping, flashovers, or even catastrophic structural failure. Understanding the physical mechanisms behind ice accretion is essential for designing resilient transmission and distribution networks. Computational Fluid Dynamics (CFD) using ANSYS Fluent offers a robust platform to simulate the complex multiphase flow, particle deposition, and phase-change phenomena that govern ice buildup. This article provides an authoritative, step-by-step guide to modeling snow and ice accumulation on power lines with ANSYS Fluent, covering everything from geometry setup to result interpretation and practical infrastructure improvements.

The Physics of Ice Accretion on Power Lines

Ice accumulation on overhead conductors occurs through several distinct mechanisms, depending on meteorological conditions:

  • In-cloud icing: Supercooled water droplets in fog or clouds freeze on contact with the cold conductor surface, forming rime or glaze ice.
  • Precipitation icing: Freezing rain or wet snow adheres to the line and freezes, often producing dense, heavy glaze ice.
  • Atmospheric icing from fog or drizzle: Similar to in-cloud icing but at ground level, often leading to rapid accretion rates.

The rate and shape of ice growth depend on wind speed, temperature, droplet size distribution, liquid water content, and the conductor’s surface roughness and temperature. Snow accumulation follows a different dynamic: dry snow typically sheds easily, while wet snow can stick and build up, especially when wind-driven. Accurate CFD modeling must capture these nuances, including the transition between rime (dry growth) and glaze (wet growth) ice.

Why Use ANSYS Fluent for Power Line Icing Simulations?

ANSYS Fluent provides a comprehensive set of tools for multiphase flow, particle tracking, and heat transfer that are critical for icing simulations:

  • Eulerian-Lagrangian approach: The continuous air phase is solved with the Navier-Stokes equations, while discrete snow or water droplets are tracked as Lagrangian particles.
  • Eulerian-Eulerian multiphase model: For high droplet concentrations, the mixture or Eulerian model can treat both phases as interpenetrating continua.
  • Phase change modeling: Built-in solidification/melting models simulate the freezing of supercooled droplets and melting of ice.
  • User-Defined Functions (UDFs): Custom code can implement empirical accretion laws, ice density correlations, or de-icing heat flux profiles.

\ANSYS Fluent is the industry standard for such analyses because of its validated physics and extensive post-processing capabilities, enabling engineers to visualize ice shapes, thermal histories, and aerodynamic loads. The software’s scalability allows simulations of single conductors, bundled conductors, and even entire transmission line sections.

Setting Up the Geometry and Mesh

Creating the Conductor Model

Begin by building a 3D model of the power line segment. In many cases, a 2D cross-section is sufficient to capture ice accretion profiles because the line is long and the flow is predominantly two-dimensional around a cylinder. However, for bundled conductors or complex terrain effects, a full 3D domain is necessary. Typical steps:

  1. Create the conductor as a cylinder with the actual diameter (e.g., 30 mm to 50 mm for typical overhead lines).
  2. Define the computational domain extending several conductor diameters upstream and downstream (at least 10D) to avoid boundary effects.
  3. Use a symmetry plane if the flow is symmetric (e.g., horizontal wind perpendicular to the line).
  4. Apply a fine mesh near the conductor surface with y+ values appropriate for the turbulence model (y+ ≈ 1 for k-ω SST).

Boundary Conditions and Material Properties

Assign velocity inlet and pressure outlet boundaries. For icing simulations, the inlet must also carry discrete droplets or a secondary phase. Key material properties include:

  • Air: Incompressible ideal gas or constant density; viscosity and thermal conductivity defined at ambient temperature (~-5°C to 0°C).
  • Water droplets: Density 1000 kg/m³, diameter typically 10–50 µm (median volume diameter).
  • Ice: Density for rime ice (200–600 kg/m³) or glaze ice (900 kg/m³); thermal conductivity ~2.2 W/(m·K).

Multiphase Flow and Particle Tracking

Discrete Phase Model (DPM) for Snow and Droplets

ANSYS Fluent’s DPM is ideal for dilute flows (volume fraction < 10%). Inject particles from the inlet with a specified mass flow rate, velocity equal to the wind, and random size distribution following a Rosin-Rammler curve. The solver calculates particle trajectories considering drag, gravity, and turbulent dispersion. When a particle hits the conductor surface, a UDF or built-in sticking model determines whether it adheres. For supercooled droplets, the heat balance equation predicts the freezing fraction:

Freezing fraction = (latent heat released) / (convective heat removed) – if the fraction equals 1, all water freezes (rime ice); if less than 1, liquid film forms (glaze ice).

For snow, the sticking efficiency depends on the snow’s wetness and the surface temperature. Dry snow (temperature < 0°C) typically does not stick, while wet snow (>0°C) can adhere and build up. A UDF can incorporate empirical correlations from the literature, such as those from the National Renewable Energy Laboratory (NREL).

Eulerian Multiphase Model for Dense Spray

When the droplet volume fraction exceeds 1–2%, the DPM becomes computationally expensive and less accurate. Instead, use the Eulerian multiphase model, which solves a separate set of continuity and momentum equations for the air and water phases. This approach is better suited for high liquid water content (LWC) conditions typical of freezing rain. The ice accretion model can still be applied via a UDF that computes the ice mass from the water phase flux hitting the wall.

Heat Transfer and Phase Change Modeling

Solidification/Melting Model

ANSYS Fluent includes a solidification/melting model based on an enthalpy-porosity formulation. This is essential for simulating the freezing of droplets after deposition. Enable the model in the multiphase or DPM setup, and define the phase-change temperature range (e.g., -0.5°C to 0°C for water). The model tracks the liquid fraction in each cell; a liquid fraction of 1 means liquid water, 0 means ice. For power lines, the conductor can be modeled as a wall with a specified temperature (often near ambient) or with a heat generation term to represent ohmic heating from the current.

Conjugate Heat Transfer for Conductor Heating

In reality, the conductor temperature is influenced by the electric current (Joule heating) and the ambient conditions. To capture this, create a solid zone inside the conductor with volumetric heat generation (I²R losses). Then, enable conjugate heat transfer at the fluid-solid interface. This coupling is vital because even a few degrees of conductor heating can prevent icing or cause melting, especially in high-current lines. The simulation reveals whether the conductor stays above 0°C, delaying or preventing accretion.

Modeling Ice Accretion Shape and Growth Over Time

Ice does not form a uniform cylinder; it grows into complex shapes with horns, feathers, and ridges that profoundly alter the aerodynamic behavior. To simulate shape evolution, ANSYS Fluent can be combined with mesh morphing or re-meshing techniques. The workflow:

  1. Run the flow solution and particle tracking for a short time step (e.g., 1 minute of real time).
  2. Compute the ice mass deposited on each wall face using the particle flux and freezing fraction.
  3. Update the geometry by moving wall nodes outward according to the ice volume and density.
  4. A smooth mesh using dynamic mesh layering or a smoothing method.
  5. Repeat the process until the desired total accretion time (e.g., 1 hour) is reached.

This iterative approach, often implemented via journaling scripts or UDFs, produces realistic ice shapes that can be validated against wind tunnel experiments. The resulting geometry can then be used for structural load analysis or for simulating galloping induced by aerodynamic instabilities.

Aerodynamic Loads and Galloping Instability

Once ice has accreted, the line’s cross-section becomes asymmetric, leading to time-varying lift and drag forces that can cause large-amplitude oscillations known as galloping. ANSYS Fluent can compute the aerodynamic coefficients (C_L, C_D, C_M) of an iced conductor as a function of angle of attack. These coefficients are then used in a separate structural dynamic solver (e.g., ANSYS Mechanical or a custom MATLAB script) to predict the onset and severity of galloping. Key parameters:

  • Critical wind speed: The speed at which aerodynamic damping becomes negative.
  • Ice shape symmetry: Even slight irregularities dramatically alter the lift slope.
  • Natural frequency of the span: Determined by tension, mass per unit length, and span length.

Engineers can use these results to install anti-galloping devices like detuning pendulums or interphase spacers, or to increase the tension in the conductor.

De-Icing and Anti-Icing Strategies: Validating with CFD

CFD simulations are not only for predicting accumulation but also for testing mitigation strategies:

Conductor Heating (Joule Effect)

A common method to prevent icing is to deliberately increase the current load in the line, raising the conductor temperature above freezing. ANSYS Fluent can model this by setting a constant wall temperature (e.g., 5°C) or using conjugate heat transfer with variable electrical heating. The simulation shows how quickly the ice melts or whether the heat is sufficient to keep the surface dry.

De-Icing Fluid Application

For some distribution lines, anti-icing fluids (like propylene glycol) are sprayed. A multiphase simulation with a species transport model can track the fluid film thickness and its freezing point depression effect. The model can optimize the flow rate and application timing.

Passive Ice Shedding

Simulations can also assess how ice sheds under wind or thermal cycles. By modeling the ice-conductor interface adhesion strength (typically via a UDF that predicts detachment when shear stress exceeds a threshold), engineers can predict when and where ice will fall, helping to mitigate risks below the line.

Case Study: High-Voltage Transmission Line in Mountainous Terrain

Consider a 230 kV transmission line crossing a mountain pass prone to freezing fog. Using ANSYS Fluent, engineers modeled a 100 m span with a 40 mm diameter ACSR conductor. Key conditions: wind speed 15 m/s, ambient temperature -8°C, LWC 0.5 g/m³, median droplet diameter 30 µm. After simulating 2 hours of accretion, the ice mass was predicted at 3.2 kg/m – a 150% increase over the bare conductor weight. The ice shape was asymmetric, with horns forming on the upstream side. Aerodynamic coefficients revealed a negative lift slope between -5° and +10° angle of attack, indicating a high risk of galloping. The line operator used this data to install interphase spacers and to schedule a current-boosting de-icing procedure during storms. CIGRE guidelines on icing loads recommend similar CFD-based assessments for critical lines.

Validation and Limitations of CFD Models

While CFD is powerful, it must be validated against experimental data or field measurements. Some limitations include:

  • Turbulence models: The standard k-ε may mispredict separation points behind iced cylinders; scale-resolving models (LES, DES) are more accurate but computationally expensive.
  • Ice phase complexity: Changes in density, porosity, and mechanical properties during accretion are not fully captured by simple solidification models.
  • Computational cost: Iterative shape evolution over many time steps can be time-consuming; reduced-order models might be needed for real-time engineering applications.

Despite these challenges, ANSYS Fluent remains the most practical tool available for detailed icing studies. For a comprehensive overview of validation benchmarks, see the NREL wind turbine icing research, which shares similar physics with power line icing.

Future Directions: Machine Learning and Digital Twins

The next frontier in power line icing modeling is the integration of CFD results with machine learning to create fast surrogate models. By training neural networks on thousands of CFD simulations spanning wind speed, temperature, LWC, and conductor diameter, utilities can generate real-time icing forecasts. These surrogates feed into digital twin systems that monitor weather data and predict ice loads across the entire grid, enabling proactive maintenance. ANSYS Fluent’s ability to automate parametric studies via journaling makes it an ideal engine for generating the training dataset.

Conclusion

ANSYS Fluent provides engineers with a detailed, physics-based platform to model snow and ice accumulation on power lines. By combining multiphase flow, particle tracking, heat transfer, and mesh deformation, simulations yield realistic ice shapes, aerodynamic loads, and thermal histories. This information is directly applicable to designing stronger conductors, planning de-icing operations, and preventing galloping failures. As computing power increases and surrogate models mature, CFD will become an even more integral part of winter weather resilience for electrical infrastructure. For any utility operating in cold climates, investing in these simulation capabilities translates directly into fewer outages, lower repair costs, and safer, more reliable power delivery.