Engineering intelligence for resilient energy systems

Physics-based Risk Engineering.

Coupled aeroelastic analysis, asset reliability and lifetime modelling, site-specific risk assessment, high-fidelity simulation — physics-based, asset-detailed, and built to defend in a room.

Explore our work Start a conversation
6 yrs NREL — OpenFAST, FAST.Farm, AMR-Wind
TC88 International wind-design standards committee
Aero-hydro-servo-elastic Coupled multiphysics specialty
30+ Peer-reviewed publications
The Gap

The energy transition is being engineered around physics the industry has outgrown.

Wake models date to the 90s. Atmospheric stability and complex terrain are bolted on as corrections. Floating platforms surge and pitch, and the wake follows. Component reliability is collapsed into a safety factor. Extreme weather is treated the same way. Standard workflows weren't built around any of this. We know — we helped build some of them, and we work with the committees writing the standards that will replace them.

01

Wake physics is treated as a 1990s problem

Industry-standard wake models still rest on assumptions that pre-date utility-scale farms — uniform inflow, neutral stability, no inter-farm interaction. AEP forecasts and layout decisions inherit hidden bias as a result. Onshore, offshore, and floating.

02

Coupled loads aren't a post-processing step

Aero, structural, control, and — offshore — hydro and mooring response interact non-linearly. Decoupled toolchains miss fatigue drivers and extreme-load combinations that compound cost and certification risk.

03

Extreme weather isn't a design margin — it's the design

Tropical cyclones, atmospheric icing, breaking-wave events, complex-terrain flows. These define survivability and need mesoscale-to-microscale coupling, not a safety factor bolted on at the end.

How we work

The Parametrica chain

One analytical pipeline, from atmosphere to certification. Executed in-house, on tools we know down to the source code.

  1. 01

    Mesoscale weather

    Site-specific atmospheric conditions and extreme-event statistics with WRF and reanalysis-driven downscaling.

    WRF · ERA5
  2. 02

    Microscale & wake

    Farm-scale LES with motion-aware turbines. Plant-level wake interaction calibrated against reference cases.

    AMR-Wind · OpenFOAM · FAST.Farm
  3. 03

    Coupled aero-elastic loads

    Turbine, controller, and — for offshore — platform, hydro and mooring resolved together. Fatigue spectra and extreme-load envelopes built from first principles.

    OpenFAST · MoorDyn · HydroDyn
  4. 04

    AI-enhanced surrogates

    Machine-learning models that bring high-fidelity physics into design loops fast enough to actually use — not just publish.

    PyTorch · JAX · custom
  5. 05

    Certification & design verification

    Translation into the documentation certifiers, developers, and lenders need — IEC 61400 compliance support, design verification, technical due diligence.

    IEC 61400-3-1 · DNV · TDD
Workflow enablement

High-fidelity physics, inside your engineering loop.

LES, coupled aero-elastic simulation, and mesoscale weather modeling normally live in research labs — they're slow, HPC-dependent, and don't fit a design schedule. Parametrica's specialty is packaging them so they actually run inside the workflows your team already uses.

That's the missing layer between research-grade physics and bankable engineering. Most consultancies sit on one side or the other. We're built for the bridge.

  • 01

    AI surrogates of high-fidelity models

    Neural and reduced-order surrogates trained on LES, coupled aero-elastic, and tropical-cyclone simulations — design-loop speed with high-fidelity physics underneath.

  • 02

    Automated simulation pipelines

    Mesh generation, case setup, run orchestration, and post-processing turned into repeatable pipelines — so engineers iterate, not babysit jobs.

  • 03

    HPC orchestration & cloud delivery

    Production-grade execution on national-lab HPC, on-prem clusters, and cloud — including the data infrastructure to move results back into your tools without friction.

  • 04

    Integration with industry tools

    Surrogates and outputs wired into FLORIS, OpenFAST, and the downstream layout, AEP, and loads tooling your team and certifiers already trust.

Capabilities

Four pillars of physics-based engineering

From coupled aeroelastic design through asset reliability and site-specific risk to high-fidelity simulation — four places where the physics is hardest and the right analysis is worth the most.

Coupled aeroelastic loads & floating dynamics

Aero-elastic simulation for onshore and offshore turbines; aero-hydro-servo-elastic coupling and motion-aware wake interaction for floating platforms; site-specific design verification across the board.

See methodology →

Asset reliability, availability & lifetime engineering

Physics-based component damage and fatigue modelling, availability and downtime distributions, and load reassessment for lifetime extension and repowering of installed renewable assets.

See methodology →

Site-specific Risk Assessment

Physics-based hazard, exposure, and damage assessment for a named site, lease area, or portfolio — the output a developer, lender, insurer, or broker can defend in a room.

See methodology →

High-fidelity simulation, AI surrogates & certification

Farm-scale LES, mesoscale-to-microscale coupling, machine-learning surrogates, tropical-cyclone modelling, and IEC 61400 design verification and technical due diligence.

See methodology →
Toolchain

Open tools, owned end-to-end.

We work in the same codebases used at national labs and by leading OEMs — not in black-box commercial wrappers. When you ask why a number came out a certain way, the answer goes to the source.

OpenFAST

Aero-hydro-servo-elastic simulation. Co-developed at NREL.

FAST.Farm

Plant-scale wake and load interaction.

AMR-Wind

GPU-accelerated LES for utility-scale farms.

OpenFOAM

General-purpose CFD with custom solvers.

WRF

Mesoscale weather and extreme-event downscaling.

PyTorch / JAX

AI surrogates for high-fidelity physics.

Dr. Georgios (Yorgos) Deskos
Founder

Dr. Georgios (Yorgos) Deskos

Principal Consultant

Six years at NREL contributing to high-fidelity wind simulation tools — OpenFAST, FAST.Farm, AMR-Wind. Active member of the IEC 61400-3-1 standards committee. 30+ peer-reviewed publications across fluid mechanics, atmospheric science, and offshore engineering.

A combination of hands-on national-lab development, standards-committee involvement, and a peer-reviewed research record. Parametrica is built around it.

  • PhD · Imperial College London
  • 6 yrs · NREL Boulder
  • IEC 61400-3-1 Working Group
  • 30+ peer-reviewed publications
More about Yorgos →
Research output

The literature behind the consulting.

30+ peer-reviewed papers. Three active multi-institution research programs. The same physics we run for clients survives review in Journal of Fluid Mechanics, Journal of the Atmospheric Sciences, and BAMS.

  • STORM — Hurricane resiliency for offshore wind (NREL, ANL, PNNL, NCAR, JHU, Brown).
  • Weather-Hazard Engineering — Physics-based hazard characterisation for clean-energy infrastructure.
  • Cluster Wakes JIP — Inter-farm wake losses, integrated into FLORIS (NOWRDC).
Browse research & publications

One physics engine, end to end.

Coupled aeroelastic response, component-level reliability, and weather modelling run on a single in-house engine — built, validated, and version-controlled against published benchmarks.

Engineering tooling, owned end to end. Used to deliver every Parametrica engagement.

Selected thinking

Notes from the field

All writing →
Floating wind

What platform motion really does to your wake.

Why fixed-rotor wake models systematically mis-predict AEP for floating arrays — and what to do about it.

 Extreme weather

What hurricane loads really do to offshore wind.

Why simple safety factors don't capture tropical-cyclone hazard — and how mesoscale-to-microscale coupling changes the design conversation.

 AI & CFD

Surrogates that engineers actually use.

The gap between an ML model that fits a dataset and one that survives contact with an engineering design loop.

Have a hard problem in renewable engineering?

A wake study, a loads analysis, a reliability or lifetime assessment, an extreme-weather analysis, a certification submission — onshore, offshore, or floating, tell us what you're trying to engineer.

Start a conversation