Why forecasting is both tempting and tricky

WORLDWIDE: Years ago, when the initial plans for a 100MW Dutch near-shore wind farm were unveiled, either 100 1MW turbines or 80 1.25MW units were to be employed. Years later, with project implementation still pending, an insider linking the ongoing project delay with the pace of wind-turbine upscaling joked: "If we wait long enough one 100MW turbine will do the job."

Google Translate

Today, while 1.5-2.5MW turbines represent the bulk of the onshore market, they are gradually increasing towards 3MW. Offshore, the workhorse is currently 3.6MW, but the market is now seeing 5-7MW and higher.

Generally, upscaling in modest increments has been a successful ongoing process since the start of the modern wind industry over 30 years ago.

Despite continuous proof that the challenges posed by bigger turbines can indeed be successfully met, there have always been concurrent voices forecasting the limits of turbine size. But what such predictions — made by inventors, engineers and scientists alike — share in common is that they have always been proven wrong.

I recall reading an article, years before I became involved with modern wind power, about a supposed 500kW power limit and the complex failure statistics used to prove the theory.

Another story that comes to mind is about an expert who was interviewed by a specialised journal in the late 1990s. He forecast that a 2MW turbine with an 80-metre rotor diameter was the optimum design from both a technological and an operating-economics point of view. At that time I already knew of several larger-scale onshore and offshore turbine projects either ongoing or planned. My immediate thought was to ask what drove him to make those not-very-clever statements and that the publication would certainly approach him again soon. Two years later he was indeed confronted by the past and unsuccessfully tried to salvage his reputation by arguing his earlier forecast was only meant for onshore turbines.

Turbine upscaling no doubt faces many challenges and one of the key issues is dealing with scale effects. When a solid component geometrically doubles in size, the area increases by a factor of four but volume and mass for an unchanged material increase by a factor of eight. This phenomenon is known as the square cube law.

Several factors have so far contributed towards successfully curbing the negative effects of the law and have enabled the continuous wind-turbine upscaling achievements we see today. The development of superior computer modelling and design tools, advanced turbine controls, the availability of new high-strength materials, and the use of optimised structural shapes have all played essential roles. Examples of the latter category include load-optimised cast main chassis design shapes and hollow, instead of solid, structural components aimed at combining high strength and stiffness with favourable mass characteristics.

Present constraints

An overall limitation of turbine upscaling predictions is that existing technology status, supported by short-term forecasting on main development trends, is the only confirmed basis for future extrapolation. Necessary longer-term input variables and uncertainties lie hidden in the future and cannot be analysed and incorporated. Analyses, therefore, provide imperfect outcomes whereby ignorance, conservatism and personal perceptions can be the main obstacles to conclusions based on knowledge, insight and verifiable facts.

A European Upwind project partner released a report late last year for a 20MW offshore wind turbine design using a 5MW reference turbine as a baseline starting point to upscale from. A choice had to be made between two 5MW German offshore wind turbines. "The Multibrid M5000 machine has a significantly higher tip speed than typical onshore wind turbines and a lower tower-top mass than would be expected from scaling laws previously developed," said the report. "In contrast, the Repower 5M machine has properties that are more expected and conventional. For this reason, we decided to use the specifications of the Repower 5M machine as the target specifications for our baseline model."

Based on the 5M’s 126-metre rotor size and by maintaining an unchanged ratio of 0.40 (in kW/m²) between power rating (P) and rotor swept area (A), the 20MW Upwind’s rotor diameter would be 252 metres.

Remarkably, several new 5-8MW offshore turbines that appeared in Windpower Monthly last year have a much lower specific head mass compared with the reference turbine. They also typically have higher tip speeds and are fitted with larger rotors per megawatt (P/A = 0.32-0.33).

Forecasting on future turbine size and technology development remains a tricky affair littered with uncertainties and the only sensible and safe answer for me on such questions is: "Let’s wait and see!"

Have you registered with us yet?

Register now to enjoy more articles
and free email bulletins.

Sign up now
Already registered?
Sign in