Felipe Pimenta was stuck. A doctoral student in oceanography at the University of Delaware, he had to complete a final project for his offshore wind power class. His department required him to take an elective course outside his field of study, and he had settled on the wind power class because his mentor, Richard Garvine, was one of three professors teaching it in the spring of 2005. Now, he was faced with the final project: assessing the feasibility of an offshore wind farm southeast of Brazil.
Felipe Pimenta used NASA satellite data to estimate the potential for wind power generation off the Brazilian coast. (Photograph ©2008 by Felipe Pimenta.)
Settling on a topic for the project had not been difficult. A native Brazilian, Pimenta was attending the University of Delaware on a scholarship from the Brazilian government. He wanted to do something that could help his country, and he knew that finding new sources of electricity was a high priority in Brazil. A country that hosts both the world’s largest rainforest and the world’s largest river by volume, Brazil relied almost entirely on hydropower. Pimenta recalled the threat of blackouts in 2001 when the country’s reservoirs dropped to dangerously low levels after years of drought. When water levels in reservoirs behind the hydroelectric dams dropped, the country nearly halted, he says. After the electricity crisis, Brazil launched an effort to diversify its power supply. Because it is comparable in cost to fossil fuel power plants, Pimenta thought offshore wind power might be a good option.
Offshore wind power is a relatively untapped source of clean and renewable energy. (Photograph ©2007 by MrPhilDog.)
“I knew that winds were pretty strong in southern Brazil. That’s where I lived,” says Pimenta. Winds tend to be strongest over flat regions like large plains or the ocean, where no mountains or other surfaces impede their movement. Offshore wind farms are more expensive to install than onshore ones, but offshore locations have greater potential for accommodating large projects. Further, highly populated regions, where power is needed most, tend to be near the coast, both in Brazil and throughout the rest of the world. Brazil has a number of wind projects under development, but all are land based. “I don’t think people have really looked or thought about the magnitude of the resource that is available offshore,” says Pimenta. “The region is very close to populated centers of the country.”
This satellite map of city lights shows the dense populations along Brazil’s southeast coast. The brightly lit cities of Rio de Janeiro, São Paulo, and Porto Alegre stand out most prominently. (NASA image by Robert Simmon, based on MODIS Blue Marble and Defense Meteorological Satellite Program city lights data.)
So, the topic had been easy. The problem was finding data to support his hunch that offshore wind had great potential in Brazil. To determine how much wind energy can be generated over coastal waters, Willett Kempton and Richard Garvine (who taught the technical and scientific aspects of Pimenta’s class) had shown their students how to estimate wind speeds 80 to 100 meters off the surface of the ocean, where a turbine’s massive blades would spin. Starting with wind speeds recorded by meteorological buoys near the ocean’s surface, students extrapolated wind speeds using a mathematical equation that describes winds at different heights in the atmosphere.
From there, he intended to calculate how much energy could be produced with a General Electric 3.6s Offshore turbine and the REpower Systems 5M turbine. The companies had released detailed information about the amount of power each type of turbine would produce at a variety of wind speeds. To estimate how much electricity Brazil could expect to generate using existing technology, he just had to know the wind speed.
Pimenta calculated the energy-generating potential of the GE 3.6s (top) and REpower Systems 5M (lower) wind turbines. GE installed a demonstration wind farm off the coast of Arklow, Ireland, that generates 25 megawatts of power with 7 turbines. Six REpower turbines comprise the Thornton Bank wind power project off the coast of Belgium. (Photographs ©2006 by Sesse Lind/GE & ©2008 by Jan Oelker/REpower Systems.)
While the United States has a flotilla of buoys monitoring winds in its coastal waters, Pimenta could only find data for two buoys in southeast Brazil. Two data points to track the winds over roughly 80,000 square kilometers? Hardly sufficient. He brought his problem to a class discussion.
“Why don’t you use QuikSCAT?” a classmate asked after hearing Pimenta’s description of his project. NASA’s QuikSCAT satellite measures global wind speed and direction by sending pulses of microwave energy to the ocean and recording the energy that bounces back from the wind-roughened surface. The energy of the microwave pulses changes depending on wind speed and direction, giving scientists a way to monitor wind over the ocean.
NASA’s QuikSCAT satellite measures wind speed and direction over the ocean. This map shows the average winds in January. Pale blue indicates light winds, while red indicates strong winds. Arrows indicate direction. (NASA map by Robert Simmon, based on the Scatterometer Climatology of Ocean Winds.)
“QuikSCAT data is well known among physical oceanographers,” says Pimenta. Scientists at NASA’s Jet Propulsion Laboratory had been creating satellite-based maps of winds over the world’s oceans since QuikSCAT’s launch in 1999. At the same time that Pimenta was working on his offshore wind research, Timothy Liu, Wenqing Tang, and Xiaosu Xie, all of the Jet Propulsion Laboratory, were using QuikSCAT measurements to estimate the amount of energy that could, in theory, be generated from ocean winds. Wind data had even been part of a suite of NASA products that fed into RETScreen software, a free program used to estimate clean energy potential. (See Power to the People on the Earth Observatory.) But as far as he knew, no one had done a practical estimate of the amount of power that would be produced with existing turbine technology using QuikSCAT data.