The height of the latest generation of wind turbines gives rise to effects that are unforeseen. The effects, in terms of performance, are positive in most cases: at these heights, powerful airflows in the lower atmosphere start playing a role and will enable a wind farm to harvest extra energy. University of Twente researchers present their work on this, in the Journal of Renewable and Sustainable Energy.

Strong airflows in the lower atmosphere, so-called ‘low-level jets’, have an impact on the performance of wind farms. The height at which this jet moves, makes the difference: does it flow above the turbines, at the level of the turbines, or even below? This determines if all turbines benefit from it, or just the front row, researchers of the University of Twente demonstrate in their paper in the Journal of Renewable and Sustainable Energy.

Wind turbines, making up a growing share of our ‘energy mix’, have increased in height over the years: the early generations had heights of 50 meter and below, the latest generation exceeds heights of 250 meter, with rotor blades of over 100 meter in length. Effects that are seen in boundary layers in the lower atmosphere, typically at heights between 50 and 1000 meters, start playing an important role. For smaller wind turbines, these effects typically take place above the wind turbine, while for the current sizes, they can play a role at the level of the turbine or even below that. The ‘rivers of air’, called low-level jets (LLJ’s), were observed at many places in the world including the North Sea region.

Researchers Srinidhi Gadde and Richard Stevens did extensive supercomputer simulations on a wind farm consisting of 40 turbines, four by ten, to study the effect of LLJ’s on performance. Earlier research showed that the wake of each turbine influences the next turbine in the row. This wake also plays an unexpected role in attracting the low-level jet.

In case the jet flows in the direction of the wind farm, at the level of the wind turbines, it is just the first row of turbines that benefits from it: it has no effect on the turbines further downstream. If the jet is above the wind turbines, however, the turbulent flow behind each turbine causes the jet to move down. Extra energy is harvested by the wind farm as a whole. The most remarkable effect takes place at jets flowing below turbine level: these jets are pushed upward by an effect called negative wind shear, resulting in higher levels of energy for turbines further downstream.

These new insights can help to design the wind turbines and to position them in a wind farm. Further research will have to shed light on effects like the transition from land to sea and several temperature effects.

The research was done in the Physics of Fluids group of the University of Twente. It is part of the Computational Sciences for Energy Research program of Dutch Research Council NWO and Shell.

The paper ‘Effect of low-level jets height on windfarm production’, by Srinidhi Gadde and Richard Stevens, is in the latest edition of the Journal of Renewable and Sustainable Energy (JRSE).

Rice computer scientists show benefits of bioinformatics with PlasmidHawk

Tracking the origin of synthetic genetic code has never been simple, but it can be done through bioinformatic or, increasingly, deep learning computational approaches.

Though the latter gets the lion’s share of attention, new research by computer scientist Todd Treangen of Rice University’s Brown School of Engineering is focused on whether sequence alignment and pan-genome-based methods can outperform recent deep learning approaches in this area.

“This is, in a sense, against the grain given that deep learning approaches have recently outperformed traditional approaches, such as BLAST,” he said. “My goal with this study is to start a conversation about how to combine the expertise of both domains to achieve further improvements for this important computational challenge.”

Treangen, who specializes in developing computational solutions for biosecurity and microbial forensics applications, and his team at Rice have introduced PlasmidHawk, a bioinformatics approach that analyzes DNA sequences to help identify the source of engineered plasmids of interest.

“We show that a sequence alignment-based approach can outperform a convolutional neural network (CNN) deep learning method for the specific task of lab-of-origin prediction,” he said.

The researchers led by Treangen and lead author Qi Wang, a Rice graduate student, reported their results in an academic journal. The open-source software is available here: https://gitlab.com/treangenlab/plasmidhawk.

The program may be useful not only for tracking potentially harmful engineered sequences but also for protecting intellectual property.

“The goal is either to help protect intellectual property rights of the contributors of the sequences or help trace the origin of a synthetic sequence if something bad does happen,” Treangen said.

Treangen noted a recent high-profile paper describing a recurrent neural network (RNN) deep learning technique to trace the originating lab of a sequence. That method achieved 70% accuracy in predicting the single lab of origin. “Despite this important advance over the previous deep learning approach, PlasmidHawk offers improved performance over both methods,” he said. 

 The Rice program directly aligns unknown strings of code from genome data sets and matches them to pan-genomic regions that are common or unique to synthetic biology research labs

“To predict the lab-of-origin, PlasmidHawk scores each lab based on matching regions between an unclassified sequence and the plasmid pan-genome, and then assigns the unknown sequence to a lab with the minimum score,” Wang said.

In the new study, using the same dataset as one of the deep learning experiments, the researchers reported the successful prediction of “unknown sequences’ depositing labs” 76% of the time. They found that 85% of the time the correct lab was in the top 10 candidates.

Unlike the deep learning approaches, they said PlasmidHawk requires reduced pre-processing of data and does not need retraining when adding new sequences to an existing project. It also differs by offering a detailed explanation for its lab-of-origin predictions in contrast to the previous deep learning approaches.

“The goal is to fill your computational toolbox with as many tools as possible,” said co-author Ryan Leo Elworth, a postdoctoral researcher at Rice. “Ultimately, I believe the best results will combine machine learning, more traditional computational techniques, and a deep understanding of the specific biological problem you are tackling.”

Rice graduate students Bryce Kille and Tian Rui Liu are co-authors of the paper. Treangen is an assistant professor of computer science.