Steel Had Its Run. Portugal Is Pouring the Future.

BayWa r.e. and Etermar Energia begin building a DNV-certified concrete semi-submersible floating wind prototype at Portugal's Port of Setúbal

20 May 2026

Construction began quietly at Portugal's Port of Setúbal in early March 2026. What's being built there has never existed before in Europe: a full-scale reinforced concrete semi-submersible floating wind foundation.

The platform is BayFloat, developed by BayWa r.e. and built in partnership with Etermar Energia. Funded through Portugal's Recovery and Resilience Programme, the prototype targets completion in July. DNV has already certified the design for 22 MW turbines, but paper certification and physical proof are different things. This build closes that gap.

Steel dominates floating offshore wind today. BayWa r.e. is betting on concrete instead, arguing it can be fabricated using existing port infrastructure and local supply chains, cutting the persistent cost pressures that shadow most floating wind projects. It's a credible argument, if the construction data holds up.

Dense instrumentation will feed live measurements into a real-time digital twin throughout fabrication, launch, and wet storage. Structural behaviour, geographic position, and environmental interaction will all be tracked continuously. Once the foundation is anchored, a two-year monitoring phase begins.

Ricardo Rocha, BayWa r.e.'s offshore wind technical director, has made the case plainly: certification alone cannot tell you how a platform actually performs across a full construction and deployment cycle. That knowledge only comes from doing it.

Portugal is well-placed to find out. Its deep coastal waters, established port network, and experienced maritime construction workforce make it one of the few places in Europe where a build like this makes logistical sense. Those aren't incidental advantages, and that's exactly why Setúbal was chosen.

The floating wind industry needs platforms that are affordable, financeable, and replicable before 2030. Two years of structural data from a real concrete foundation, in real Atlantic conditions, will do more to settle the material debate than any model ever could.