By Robynn Andracsek, P.E., Burns & McDonnell and Contributing Editor

One question I often am asked is, “How tall should the exhaust stack be?” It’s a seemingly simple question, but the only way to get an exact answer is to run an air dispersion model. That takes time, money and a qualified consultant. Not every circumstance allows, or requires, this level of effort. Here is some basic information you may use as a guide.

Air dispersion modeling is required for most Prevention of Significant Deterioration (PSD) permits and some state construction permits. One regulatory requirement is that the modeled concentration, plus the background pollution level, must not exceed the National Ambient Air Quality Standards (NAAQS) or PSD Increment.

Air dispersion modeling using the Environmental Protection Agency’s AERMOD algorithm is based on equations with many stack parameters, including stack height, diameter, velocity, temperature and emission rate. Added to this basic description of every stack are its base elevation, where it is located and how it is situated in relation to nearby buildings. Clearly, no single parameter dictates the optimum combination that minimizes ground level pollution concentrations.

As a result, 500 pounds per hour of NO₂ emissions out of a 200-foot-tall stack might have a negligible impact, but 5 pounds per hour of NO₂ emission out of a 20-foot-tall stack with a rain cap could result in concentrations several times the NAAQS or other modeling threshold. Likewise, putting a scrubber on a coal-fired unit does not guarantee 1-hour SO₂ NAAQS compliance since ground level concentrations are directly, but not solely, related to the unit’s emission rate.

The first thing to remember when designing a stack for air dispersion modeling compliance is that rain caps often are problematic. Initial dispersion comes from the heat of the exhaust (known as “buoyancy”) and how fast it exits the stack (“momentum”). In modeling, a rain cap is considered to be any kind of obstruction on the end of the stack. (The same effect can also be achieved through a horizontal exhaust.) Only a vertical, unobstructed stack is allowed to account for the momentum component of initial dispersion. If rainwater must be prevented from entering the stack, a “tractor flap” type of exhaust will prevent water from entering when the unit is not in operation and allow the exhaust to exit freely when it is.

A common misconception is that good engineering practice (GEP) stack height is a minimum, or recommended, stack height. Not true. GEP dictates not the minimum but the maximum stack height you may take credit for in the model. This prevents constructing mile-high stacks. The solution to pollution is not dilution and most stacks are built to a height shorter than GEP. The GEP stack height is dictated by calculating the stack height in relation to each nearby building’s height and crosswind profile. The dominant building is often not even the one on which a stack is located.

GEP stack height is related to what’s known as building “downwash.” If you’ve ever been in an alley between two tall buildings on a windy day, you’ve likely seen leaves swirling around. They are caught in eddies created by wind flowing around the tall structures. The same concept helps explain downwash. The presence of an obstruction changes both the wind’s direction and velocity. Every structure will form an aerodynamic wake on its downwind side. The size and intensity of this “downwash zone” will increase in proportion to the wind speed. Stack exhaust discharged into this zone can be pulled down to the ground before it can disperse the way it would in the absence of the building. A rule of thumb is that GEP is 2.5 times the closest, largest building’s height, but it is much more complicated. Constructing each stack at GEP height would be an expensive eyesore.

One of the first modeling tools to employ is the ability to restrict a source’s operation to certain times of day. Winds are calmer at night and less dispersion occurs. If a source operates from 8 am to 6 pm, consider taking a permit limit if necessary that officially restricts its operations to that time period. You then can eliminate nighttime hours from the dispersion modeling for the particular source. This arrangement would not affect operational flexibility since the facility is already following the limitation and it allows the calmer nighttime hours to be excluded from the model. In terms of shorter averages (other than annual), such a limitation might resolve a modeling exceedance.

Air dispersion modeling is more art than science. As with all art, skill and experience determine the value of the result.