# AERMOD Tech Guide

## 6.4.3 Component of Dispersion Coefficients due to Downwash_{}

In ISC3, the primary effects of building downwash are on the plume growth (y and z ) for both the Huber-Snyder (H-S) (Huber and Snyder, 1976 and 1982) and Schulman-Scire (S-S) (Schulman and Scire, 1980) algorithms and on the plume rise for the S-S algorithm. These effects are also present in AERMOD, with some changes due to the fundamental difference in the model formulation, as described below.

In AERMOD as in ISC3, the decision as to whether a plume is affected by downwash is determined by comparing the plume height due to momentum rise at 2 building heights downwind to the Good Engineering Practice (GEP) (Code of Federal Regulations, 1995) height of the building. Direction-specific building dimensions are used in the same manner in ISC3 and AERMOD. For stack heights at least 1.5 Lb (where Lb is the lesser of the building height and width for the specific direction being considered), the H-S algorithm is invoked if downwash effects are to be considered. For stack heights less than 1.5 Lb , the S-S algorithm is used.

In both ISC3 and AERMOD, no concentration calculations are made for receptors less than 3 Lb from the source. This is the cavity region that is currently accounted for in the model SCREEN3 (Environmental Protection Agency, 1995). For receptors between 3 Lb and 10 Lb downwind, both ISC3 and AERMOD compute the same building-induced y and z and compare these to the values of y and z due solely to ambient turbulence (which are not the same in the two models and will lead to differences in predictions). The larger of the values of the two sets of y and z are chosen for concentration calculations. One complication for AERMOD is that in convective conditions, only the direct plume is assumed to be affected by downwash conditions. The indirect and penetrated plumes are assumed to escape the effects of downwash. For the direct plume in AERMOD, the average value of y and z for the two components of the direct plume are used for comparison to the building downwash-induced y and z values.

For receptors beyond 10 Lb downwind in AERMOD, the added enhancement in y and z due to the building effects (if positive) is "frozen" at the value attained at 10 Lb , and is added to the effects of turbulence, plume buoyancy, etc., in quadrature (the total variance is the sum of the squares of the components of ambient turbulence, buoyancy, and the excess due to downwash see eq. (85)). The ISC3 treatment is different in that the building-induced enhancement in y and z at 10 Lb is used to determine a virtual source location as if ambient turbulence was the only factor in the plume growth up to the 10 Lb distance. Due to the complicated nature of the ambient turbulence calculations in AERMOD, the virtual source treatment is not feasible. For the S-S algorithm in both models, the buoyant plume rise is depressed due to increased entrainment from the building-induced turbulence of ambient air into the buoyant plume. In AERMOD for convective conditions, this condition only affects the direct plume. The following sections summarize the specific enhancements made to both the lateral and vertical dispersion coefficients by AERMOD to account for building downwash effects.

*Momentum Plume Rise Equations for Use in Determining Applicability of Downwash*

The application of enhanced dispersion due to building downwash is determined by comparing the plumeâ€™s height after momentum rise (Hem) with the building height. The momentum plume rise equations used by AERMOD are as follows:

For convective conditions,

For stable conditions,

*Enhancement of the Lateral Dispersion Coefficient to Account for Downwash*

Enhancement of horizontal plume spread (y) is assumed to occur when the convective direct y plume height hed = 1.2 hb or when the stable plume height hes = 1.2 hb , where hb is the building height.

*Downwind Distance Range Between 3 and 10 Building Heights*

For downwind distances, x, such that 3Lb x < 10Lb,

For convective cases, ya = yl and yd is not used (see eq. **(85)**). Note that only the direct plume is adjusted for building downwash in this manner. The indirect plume and penetrated plume are not changed. The injected plume is treated the same as the stable plume for building downwash calculations. Similarly, for stable cases, ys = yl and yd is not used (see eq. **(87)**).

*Downwind Distance Greater Than 10 Building Heights*

For downwind distances x > 10Lb , yl is assumed to be a constant equal to its value at x=10Lb . Then for convective conditions yd in eq. **(85)** is calculated as

where ya is calculated from eq. **(88)**. For stable cases yd in eq. **(87)** is calculated as

where yas is calculated from eq. **(90)**. yas

*Enhancement of the Vertical Dispersion Coefficient to Account for Downwash*

Enhancement of vertical plume spread (z ) is assumed to occur when the plume height, He , calculated as the sum of the physical stack height and the momentum plume rise, is less than or equal to hb + 1.5 Lb .

*Downwind Distance Between Three and Ten Building Heights*

For downwind distances, x, such that 3Lb = x < 10Lb the vertical spread from the combined effects of ambient turbulence and building downwash zl is taken from ISC3 as

For the domain in which the Huber-Snyder algorithms (Huber and Snyder, 1982) apply, i.e. for (hb + 0.5Lb) H (hb + 1.5Lb) the coefficient A in eq. **(112)** is set equal to 1.0. For effective plume heights which are less than hb + 0.5Lb the Schulmann-Scrie (Schulmann and b bScrie, 1980) algorithms apply and the coefficient A in eq. **(112)** is given as follows:

Then for convective conditions zaj in eq.**(85)** is set equal to zl from eq. **(112)**. Similarly, for stable conditions zas in eq. **(87)** is set equal to zl .

*Downwind Distances Greater than Ten Building Heights*

For all downwind distances greater than 10Lb zl is first calculated for a downwind distance of 10Lb Then for convective conditions the building downwash component of the total vertical plume spread calculated after eq. **(85)** as

And for stable conditions zd is determined from eq. **(87)** as zd