It’s all in the yaw

So I simply ask, just for the sake of argument: would you rather have a hurricane bearing down on an array of wind turbines or zeroing in on a field of oil platforms? (GW)

Betting the farm: new model shows offshore wind farms at risk from hurricanes

By Kyle Niemeyer
ars technica
February 14, 2012

If we want to obtain 20 percent of our electricity from wind power by 2030, the US is going to need at least 50 gigawatts from offshore wind farms, according to the US Department of Energy. The National Renewable Energy Laboratory (NREL) estimated that this wouldn’t be a problem—we could provide four times our 2010 electricity generation capacity with offshore wind power alone. The US hasn’t actually built any offshore wind farms yet, although there are at least 20 in the planning stages.

As part of that planning, the Interior Department recently performed a review, concluding there would be no significant environmental or socioeconomic impacts from wind farms off the mid-Atlantic Coast. However, according to a paper published today in the Proceedings of the National Academies of Science, we should be worrying the converse: the impact of the environment on the wind farms, from hurricanes in particular. In certain risky offshore regions off the Atlantic and Gulf Coasts, there is a high probability that at least one turbine would be destroyed by hurricanes within 20 years, and a smaller chance that half the turbines in a farm would be wiped out.

The authors of the paper, a group from Carnegie Mellon, used a probabilistic model to estimate the number of turbines destroyed by a hurricane. They performed this analysis for four locations where either farm leases have already been signed or projects have been proposed: Galveston County, TX; Dare County, NC; Atlantic County, NJ; and Dukes County, MA. The differences in location affect the probability of hurricane occurrence and the maximum wind speed, which were obtained using historical data.

Turbines are at risk from hurricanes due to the high maximum wind speeds, which exceed the design limits and can cause blade loss and tower buckling. In 2003, two typhoons (Maemi and Dujuan) destroyed turbines in Japan and China.

It’s all in the yaw

In the current study, the researchers only considered buckling, since blades can be more easily replaced. The buckling is much more likely when the turbines cannot yaw, or turn into the direction of the wind. Most modern turbines have this ability, since it allows them to generate more power. Somewhat ironically, however, hurricanes usually knock out the external power necessary for yaw motors. Even with power, winds in a hurricane can change direction faster than the turbine can yaw.

Initially, the team looked at a 50 turbine farm and calculated how many turbines would be destroyed if the farm was hit by a single hurricane, assuming the turbines couldn’t yaw. If the farm encountered a Category 3 hurricane, meaning wind speeds at least 45 meters per second, up to 6 percent of the turbines could buckle. A Category 4 hurricane, with wind speeds at least 50 m/s, could buckle nearly half of the farm. For some perspective, every state on the Gulf of Mexico and nine of the 14 on the Atlantic coast has been hit by Category 3 or higher hurricanes since 1856.

Next, they calculated the probability of turbine buckling in the four specific locations over a period of 20 years. In this case, they performed separate analyses of turbines that could and could not yaw.

The team found that Galveston and Dare Counties are the riskiest locations (of the four considered) for offshore wind farms. In Galveston County, without the ability to yaw, there is a 60 percent probability that at least one turbine tower would buckle in 20 years, and 30 percent chance that half the farm would be destroyed. Dare County is little better, with the same 60 percent chance for one turbine and 9 percent for more than half.

Atlantic and Dukes counties, on the other hand, were much safer. Without the ability to yaw, there was a 15 percent probability in Atlantic and 10 percent probability in Dukes that at least one turbine would buckle in the 20 year period. In both locations, there was a less than one percent probability that more than half the farm would buckle.

Giving turbines the ability to yaw significantly decreases the probability of buckling. The chances of a single turbine buckling in Galveston and Dare Counties drop to 25 percent and 15 percent, and the probabilities of more than half the farm being destroyed drop to 10 percent and less than one percent, respectively.

If the turbines in Atlantic and Dukes Counties can yaw, there is little chance that even a single turbine would be destroyed in the 20 year period.

Further emphasizing the importance of yaw, the researchers found that without yaw, there is a one in ten chance that an entire 50 turbine farm could be destroyed in Galveston County.

What can we do?

The authors suggest three reasonable ways to avoid hurricane damage: 1) increase the maximum wind speed that designs can handle; 2) make sure the turbine can quickly turn into the direction of rapidly changing winds; and 3) build most offshore wind farms in areas with lower risk.

Most modern turbines are rated to withstand maximum wind speeds of around 42 m/s. These designs are typically based on conditions in northern Europe and the North Sea, the location of most existing offshore wind farms. Some estimates put the cost of increasing the safety factor at 20-30 percent for turbines on land, but the fraction would likely be smaller for offshore turbines. Most of the cost there goes into transportation, installation, and maintenance, rather than the construction of the turbine itself.

Adding battery backups for yaw motors could add up to $40,000 to the turbine cost and a metric ton (literally) to the weight. But allowing a turbine to yaw during a hurricane significantly reduces the probability of buckling. In the safer locations, this risk is practically eliminated when the turbines can yaw. Batteries could potentially buffer the farm’s power output when the weather isn’t threatening.

There are a couple limitations of this study. The authors emphasize that the results in their paper do not represent all possible offshore wind farm locations—hurricane occurrence and intensity vary wildly from location to location. In addition, the entire analysis is based on historic data. Any predictions would be thrown off if climate change affects hurricane behavior, and the relationship here is unclear at best.

In fact, there is a good chance the probabilities presented here are underestimates. The model only considered tower buckling, but falling blades can also destroy the turbine. In addition, the buckling analysis didn’t account for the forces of heavy waves, another common occurrence during hurricanes.

This type of analysis could potentially be performed as part of a potential offshore wind site evaluation. Even without redesigning turbines to withstand higher winds, simply avoiding more dangerous locations could reduce the risk of turbine damage significantly.