At the heart of every wind resource assessment is the wind flow model. It provides an estimate of the wind speed variability across the project area and serves as the guide for designing the turbine layout. If the wind flow model is flawed or biased then the wind farm won’t produce as expected and renewable energy that could have been generated will be lost. If the wind flow model is accurate then the actual wind farm production won’t be a surprise, the investors will be happy and there will be less explaining to do in general. With accurate models, our industry will gain more credibility and the uncertainty of project performance will reduce, giving us more of an advantage as we compete against other sources of energy production.
The problem with existing modeling tools
When I started working as a wind consultant for Jack Kline back in 2010, there were only a handful of commercially-available wind flow modeling softwares and they could be categorized as either linear or CFD. At that time, the issue of accuracy with linear models was becoming more and more well-documented and the difficulty and complexity of CFD was being realized.
Was there a way to reduce the estimate error and eliminate bias in wind flow modeling using a method less complicated than CFD? Could high-quality models be produced more quickly and without requiring an expert knowledge level so that wind flow analysts could spend more time doing what they love, which is innovate and explore different techniques to ultimately improve the methodologies used in resource assessment? Couldn’t we do better?
Thinking differently about wind flow modeling
It was Jack who introduced me to the idea of using terrain exposure to model wind speed and it made so much intuitive sense to me. Once the wind speed distribution has been measured at a site, the variations in the wind speed should be predictable based on how the terrain changes within the project area. On top of that, time and time again, Jack and I would observe strong correlations between exposure and wind speed. And, since the only two options for wind flow modeling software fell at the two extremes (overly simple versus overly complex), there was an obvious gap in the technology and I believed that exposure-based modeling could fill this void.
When I started developing the software that is now Continuum, we called it RAMWind and it worked by generating linear regression equations between exposure and wind speed which were then used as predicting tools and, overall, it worked quite well. The main problem with this approach, however, was that a minimum of three met sites were required and it was based purely on empirical data.
Since exposure-based modeling seemed to work at all sorts of different sites, I thought that there had to be some commonality between the models and that some sort of universal model must exist. This led to me to take on a research project where I scrutinized the model coefficients from over a dozen different sites across North America. After analyzing the models from every possible angle that I could think of, it finally became clear that the model coefficients were a function of the level of terrain complexity and this marked the beginning of developing Continuum.
Innovating to help wind flow analysts
In May 2014, we introduced the first version of Continuum (1.0) and, while it offered a model that could be used with a single met site unlike RAMWind, it was still based on empirical data and I was pressed by some early users about the physics behind the model. This led me to dust off my old fluid mechanics textbook and get back to basics. Eventually, I figured out how to derive the equations used in Continuum from the Navier-Stokes conservation of momentum equation – the same set of equations used in CFD.
By combining a method that originated from empirical observation with first principles of physics, Continuum has transformed into a highly accurate model that circumvents the complexity of CFD models. Also, it was designed to be user-friendly and quick to generate results and to incorporate all of the met data simultaneously, taking full advantage of the entire data set and letting the measurements speak for themselves.
With a tool that produces high quality wind flow models in a fraction of the time, wind analysts can now spend more of their time innovating and advancing the state-of-the-art in resource assessment which will lead to better wind farm performance, lower project uncertainty and improve the overall state of our rapidly-growing industry. And that’s how and why we decided to develop the Continuum wind flow modeling software. Learn more about how it works or start a free trial now. www.cancalia.com