On-site data is the first option in such evaluations. But they rarely exist going into a study, while the distribution of meteorological sites throughout the nation is "fairly dense," said Scire, a vice president of Massachusetts-based TRC Environmental Corp., who testified on behalf of SWEPCO.
That opens the door to off-site data, which are used so long as both sites are "adequately representative" in terms of geography, climate, manmade and other factors that can influence wind patterns, Scire said.
Upon studying such factors, Scire found conditions at Shreveport Regional Airport sufficiently similar to the site for the John W. Turk Jr. power plant 15 miles northeast of Texarkana.
"That is not meant to say they are the same in any way," said Scire, who described how urban areas in Shreveport and wooded areas near the plant site have similar effects on prevailing wind flow from the south and southeast.
"But they are similar enough for purposes of [air] modeling."
Several factors that led to Scire's conclusion rebutted statements by Michael Hunt, an engineer and consultant who testified on behalf of the Sierra Club and National Audubon Society - groups that want to halt the plant's construction.
Hunt said that SWEPCO's reliance on five years of airport data - instead of an optional, one-year test at the plant site - failed to demonstrate that the Turk plant would not violate state or federal air rules.
He testified that airport data such as wind speed, wind direction and the frequency of such measurements are less detailed than what instruments designed for emissions studies could produce in an on-site review.
Concerning wind speeds, Hunt said they appear calmer on average than they actually are because airport instruments do not account for speeds less than 59 inches per second.
Scire countered that such "calms" are addressed by mathematical formulas approved by the Environmental Protection Agency. He added that missing data from Shreveport falls within a 10 percent threshold allowed by the EPA.
Similar steps are taken with wind direction, Scire said. Hunt testified that wind direction at airports is reported to the nearest 10 degrees, which can produce errors of 20 percent to 70 percent in emission concentrations.
While Hunt said single-degree increments are preferred, Scire said measurements are calculated to account for such variances. In written testimony, Scire also said that the EPA allows the use of 10-degree data, although 1-degree data is "preferred."
As to why SWEPCO relied on airport data rather than perform a one-year plant site study, Scire said the additional cost and delay to the $2.1 billion project were not justified.
Costs for the plant alone are estimated at $1.63 billion. Other expenses include $89 million for power lines, substations and other upgrades to link the plant to the electric grid. They also include a $343 million deal that SWEPCO struck with state and federal officials last year to help secure the Turk plant's air permit. It calls for SWEPCO to reduce emissions from a Texas plant to offset "visibility impacts" from the Turk plant in the Caney Creek Wilderness Area near Mena.
Project delays have pushed the plant's projected June 2012 startup to October 2012. Further delays could occur in obtaining an environmental permit for areas controlled by the U.S. Army Corps of Engineers, SWEPCO has said.
Along with the savings of using airport data, Scire cited other benefits.
"If off-site data are representative of conditions at the site, a five-year data set may produce a more accurate representation of rare, worst-case conditions than a shorter, one-year set," he said.