Designing a wind farm is a highly complex optimisation challenge. At its core are two critical and tightly intertwined disciplines: turbine layout and cable routing. Both are essential for minimising the Levelized Cost of Energy (LCoE). The best turbine layout maximises power production, while the optimal cable routing minimises infrastructure cost and electrical losses. Yet traditionally, these two tasks have been handled separately, often by different teams using different tools. This siloed approach can limit performance and slows decision-making.
‍

Why co-optimisation matters

Integrating layout and cable optimisation means updates to turbine positioning automatically recalculate cable lengths, electrical losses, and total energy yield. This creates a fast, responsive feedback loop between the yield engineers and the electrical engineers, resulting in smarter decisions being reached faster.

This co-optimisation is particularly beneficial because in the project design phase, changes are constantly being made to both layout and electrical design. Wind resource assessments evolve. Turbine models get updated. If layout and electrical design are managed in separate tools, it's easy for inputs into each model to fall out of sync. 

‍

How it works

Vind AI’s digital platform lets you design and optimise the turbine layout, as well as cable routes, connection to grid, cable types and substation placement, based on real-world constraints such as exclusion zones, cable chain limits, bathymetry and topography. The overall yield calculations give you a full overview of the expected energy production, taking into account the wake effects, turbine losses and electrical losses.

‍

When both are housed in the same digital platform—and importantly, are designed to work in an integrated manner within that platform—any change in project assumptions, like wind resources or the turbine power curves, immediately reflects in cable routing, losses, and cost, keeping the entire system coherent and up to date. When a turbine is slightly moved, the effect on the cable length is automatically updated, and you will see the effect on the electrical losses and overall yield. Even better, if you want to compare two different turbine layouts, you can see the difference in both yield and electrical factors side by side.
‍

‍

Integrated financial analysis

Maximising power production and minimising losses by co-optimising layout and electrical design is only half the story. A truly integrated optimisation of your project means you also want costs to be integrated into your analysis from the start. 

Vind AI lets you specify unit costs on all elements in your project, making it possible to consider the total CAPEX and installation costs at the same time as the yield and loss assessment. This financial integration gives users a true system-level view of wind farm performance. You can simulate multiple design scenarios side by side, compare different cable type strategies, and understand the trade-offs in LCoE across the entire energy export chain. And if supply costs change, you can, of course, easily make changes.

‍

Algorithms + human design expertise

While digital platforms can now automate most of the heavy lifting of co-optimising your layout and electrical systems, we know that expert designers have expertise and knowledge that cannot be outsourced to algorithms.  In Vind AI, manual overrides and detailed adjustments are always available.

‍

“The park and array cable layout optimisation features in Vind AI provide a great starting point for the initial energy production estimates and electrical system modelling.”

_{Charlotte Elmelid, Electrical Systems Manager, Vårgrønn}

‍

And since the platform is web-based and collaborative, multiple team members—across disciplines—can work together simultaneously. Changes are reflected instantly across the design, ensuring a single source of truth that’s always current, with no need to email files or sync across shared drives.

‍
‍

The new way of working

‍

“Optimising individual components of wind farms separately is becoming an outdated way of working. New optimisation methods are powerful enough to optimise the whole system. When choosing a digital platform for wind farm assessment and design, developers should make sure it supports both turbine layout and cable routing in a single tool, so that they can take advantage of the most advanced optimisation methods to design smarter parks, faster.”

_{Jan-Tore Horn, CTO, Vind AI}

‍

By combining detailed electrical modelling with spatial, geotechnical, and cost considerations, integrated design tools are moving closer to the reality of wind development: a highly coupled, multi-variable optimisation problem where no decision exists in isolation.

‍

‍

How electrical modelling works in Vind AI: 
‍

• Optimise for minimal cable length: Optimising substation placement and cable routing for a given turbine layout, to minimise total cable length.
  
  
• Customise your components: Customisable catalogue of inter-array cables, export cables and substations.
  
  
• Instant estimates of electrical losses: State-of-the-art power flow analyses, where current and associated electrical losses are calculated all the way from the turbines to the grid connection, including the landfall.
  
  
• Granular loss calculations:  By integrating electrical loss estimation directly with the yield model, losses are computed on a time-resolved production profile. This enables hourly resolution of losses across each system component, leveraging 30 years of high-fidelity wind resource data to ensure statistically robust results.
  
  
• Pinpoint exactly where your electrical losses occur: Both inter-array cable losses and export cable losses are modelled, including turbine transformer losses, cable losses (with optional dielectric losses) and onshore and offshore substation transformer losses.
  ‍‍
• Rapid assessment of transmission options: For offshore wind farms, both HVAC and HVDC transmission options can quickly be assessed, with options for reactive power compensation if needed. 

‍