PNWIS Cross-border Air and Waste Solutions The effect of vegetative buffers on wind and dispersion of particulate matter around poultry barns Andreas Christen (1), Daryll Pauls (2), Shabtai Bittman (3) (1) Department of Geography, University of British Columbia, Vancouver, BC, Canada (2) Atmospheric Science Program, University of British Columbia, Vancouver, BC, Canada (3) Agri-Food Canada, Agassiz, BC, Canada

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Motivation Controlling emissions of particulate matter from poultry facilities is desirable to mitigate negative impacts on natural resources (air and water pollution), agricultural plots, animal and human health. Emissions from poultry barn visualized using smoke (Photo: A. Christen, UBC) Vegetative buffers can possibly ‘filter’ emitted particulate matter and reduce total emissions leaving a property.

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Density of facilities in the Lower Fraser Valley

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Airflow of a typical 2-barn layout ExhaustAir intake

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Without vegetative buffer Flow is accelerated between barns. The emitted particulate matter is quickly dispersed and carried away. With vegetative buffer Wind speed is slowed and mixing is reduced. This causes locally higher concentrations. High concentrations of particulate matter means higher probability of deposition on leaves and ground. However, boundary layer resistances are also higher, reducing deposition rates.

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Research questions  What is the potential of vegetative buffers to remove particulate matter before it leaves a property?  What would be most efficient buffer layouts that promote the removal of particulate matter - arrangement, height and density?

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Layout considerations for vegetative buffers In-fill of alleyway + Vegetation removes particulate matter where concentrations are highest, i.e. close to source. + Moderate costs and maintenance. - Complicates operation (access to alleyway) Perimeter - Removes particulate matter downwind of facility, where concentrations are presumably lower. - Substantial costs and water need. + Acts also as noise and visibility barrier. Property line

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Methods  Numerical model runs using the microclimate model Envi-met (Version 3.1b5, Bruse 2010)  Envi-met is a 3-d Eulerian CFD model that is designed to simulate interactions between surfaces, buildings and plants, and the air flow which may contain particulate emissions (PM10).  Modelled domain: 180 x 120 x 30 grid cells at a grid resolution of 1.5 x 1.5 m horizontally.

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 ReferenceIn-fill of alleywayPerimeter R Neutral stability 1.7 m/s wind at 1.5 m 6 PM10 sources of 10 mg s -1 on each side of alleyway at 1.5 m height I1 2m tall staggered shrubs with a plan area density of 13% (300 m 3 vegetation) P1 6m tall coniferous hedge, 15m away from edges of barn (6,000 m 3 vegetation) I2 Same as I1 with only 1/3 leaf area density (300 m 3 vegetation) P2 Same as P1 but 30m away (8,000 m 3 vegetation) I3 Same as I1 but 6m tall (900 m 3 vegetation) P3 Same as P2 but 15m tall (24,000 m 3 vegetation) Model runs +

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Effect of perimeter buffer on wind speed (P1) Cavity behind hedge Acceleration in alleyway

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Effect of in-fill buffer on wind speed (I1) Cavity behind hedge Reduction of wind in alleyway

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Effect of buffers on wind speed Wind P1 I1

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Modelling PM10 Concentrations 2 x 6 Sources at 1.5 m height with 10 mg s -1

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Effect of buffer on PM10 concentrations Wind P1 I1

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Change in PM10 concentration in alleyway Increase Decrease Change to reference (R) In-fill I1I2I3P1P2P3 Perimeter In-fill

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Effect of leaf area density Leaf area density (m 2 m -3 ) Wind (m s -1 ) Leaf area density (m 2 m -3 ) Concentration (µg m -3 ) In-fill only, average values in alleyway at 1.5 m above ground In-fill

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Total deposition of PM10 on ground, vegetation and roofs Deposition on leaves Less deposition in lee

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 PM10 removal by vegetation Fraction of total emissions removed by leaves Perimeter In-fill I1I2I3P1P2P3

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Vents Scenario R (with no buffer) ground 97% roofs 3%

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Scenario P3 (15m high trees) Additional deposition on leaves Less deposition on ground in cavity ground 88% roofs 3% vegetation 9%

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Scenario I1 (2m in-fill) Additional deposition on leaves Less deposition on ground (behind shrubs) ground 88% roofs 3% vegetation 9% Scenario R (with no buffer)

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Fraction of PM10 emitted that is removed before leaving the property % of emitted PM10 PerimeterInfillNo buffer

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Flow over buffer causes shelter effect Accumulated snow reduced This is well-known from snow-fences - Solid fences generate strong lee eddies behind and reduce snow trapping ability.

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Summary  As expected, vegetative buffers decrease wind and consequently increase concentrations (desired).  Perimeter layouts show little to no impact. More efficient to modify wind and dispersion is an infill of the alleyway.  In the best case, up to ~10% of emitted PM10 was deposited on the buffer.  In none of the scenarios, total PM10 leaving the property was significantly reduced. In some cases the enhanced deposition on leaves was offset by much reduced deposition on ground behind buffer elements.

PNWIS 2011 Christen, Pauls and Bittman / Nov 10, 2011 Further research  Under different atmospheric conditions (direction of flow, stability)  Revise or re-evaluate microphysics in model. Is deposition on ground and leaves properly described.  Size distribution of aerosols has larger effect on settling. With larger sizes effects of buffers are increased, with smaller size, reduced.