WINDWARD WINDBREAK EFFECTS ON AIRFLOW

Transcription

1 WINDWARD WINDBREAK EFFECTS ON AIRFLOW IN AND AROUND A SCALE MODEL OF A NATURALLY VENTILATED PIG BARN A. Ikeguchi, G. Zhang, L. Okushima, J. C. Bennetsen ABSTRACT. Wind tunnel tests of 1:20 scale windbreak models placed in a windward position from naturally ventilated pig barns were performed to examine the capability of windbreaks to trap air contaminants emitted from livestock buildings. The airflow surrounding the windbreak and the building was measured. Dispersion of air contaminants was predicted with the proposed parameters P U, P V, and P W, which are related to airflow momentum in the leeward direction and to horizontal and vertical plume sizes, respectively. Windbreaks included a solid wall, a net screen, and another pig barn. With a solid wall as a windbreak, air emitted from the building was exhausted in the windward direction. Compared to a net screen or another barn, the solid wall resulted in the lowest airflow momentum. A high concentration odorous area was predicted to be located between the windbreak and the building. Sprinklers in this area could possibly trap the air contaminants. When a building was placed in the windward position as a wind barrier, the dispersion of odorous air behind the leeward building was predicted to be the smallest compared to a solid wall and a net; moreover, a smaller plume width was seen. When a net screen was used, the airflow between the wall and the building moved only in the leeward direction. Keywords. Natural ventilation, Odor dispersion, Pig barn, Windbreak, Wind tunnel. Naturally ventilated, open type pig barns have the advantage of energy saving and low noise (Strom and Morsing, 1984). As such, this type of building has gained wide acceptance in Europe and Japan. However, there is concern about dispersion of air contaminants from the building. The dispersion property was reported by Ikeguchi and Okushima (2001); they noted that an open livestock building allowed air contaminants to disperse widely toward the leeward side of the building. One possible control measure is a windbreak. Windbreaks are used in agriculture to minimize incidences of wind erosion related to land, waste, and construction. The functions of windbreaks are to change the direction of approaching wind and to create a wake. Much research with respect to windbreaks has been carried out to date; however, research efforts regarding control of dispersion of air contaminants from livestock buildings have been limited. The effect of tall barriers in reducing odor emissions from lagoons was investigated using numerical simulation (Liu et al., 1996); odor emission was reduced by 26% to 92%. Bottcher et al. (1999) reported the effect of windbreak walls on contaminant air dispersion from a tunnel ventilated swine building; the walls deflected the airflow upward and caused air plumes from the fans to flow higher above the surface of a downwind lagoon. In another study, Bottcher et al. (2000) reported that windbreaks did not reduce the odor emissions substantially but dispersed them upwards, thereby reducing the concentration of odors. As mentioned earlier, windbreaks at the leeward side of a fan ventilated livestock building will divert the upward flowing odorous air. It is important to prevent the contaminated air from dispersing to the outside farm. It is necessary to develop equipment that will trap the emitted air from livestock buildings. One option is to place a windbreak at the windward side of the building, thereby changing the airflow pattern inside and around the building. This study was conducted to determine the effect of windbreaks on the airflow pattern inside and around a naturally ventilated pig barn and to predict the degree of dispersion of air contaminants from the barn. Article was submitted for review in July 2002; approved for publication by Structures & Environment Division of ASAE in March The authors are Atsuo Ikeguchi, ASAE Member, Senior Researcher, National Agricultural Research Organization Headquarters, Tsukuba City, Japan; Guiqiang Zhang, Senior Research Scientist, Department of Agricultural Engineering, Danish Institute of Agricultural Sciences, Research Centre Bygholm, Horsens, Denmark; Limi Okushima, Senior Researcher, National Institution for Rural Engineering, Tsukuba City, Japan; and Jens Christian Bennetsen, Research Scientist, Ranboll Consulting, Virum, Denmark. Corresponding author: Atsuo Ikeguchi, National Agricultural Research Organization Headquarters, Kannondai 3 1 1, Tsukuba City, Ibarali ken Japan; phone: +81 (0) ; fax: +81 (0) ; e mail: ikeguchi@ affrc.go.jp. MATERIALS AND METHODS MODEL PIG BARN The model pig barn was a 1:20 scale geometric representation of a full scale naturally ventilated building. Full scale dimensions were as follows: length = 30 m, depth = 11 m, eave height = 2.4 m, and ridge height = 5.2 m. The sidewall opening height was 0.9 m. The geometric dimensions and coordinates for the scale model are shown in figure 1. This model was similar to that used by Morsing et al. (2002), who investigated the effect of inlet size and position on the air velocity in the animal occupied zone. Transactions of the ASAE Vol. 46(3): American Society of Agricultural Engineers ISSN

2 Figure 1. Wind tunnel and dimensions of windbreak, scale model. EXPERIMENTAL DESIGN The experimental factor was a windward windbreak, consisting of a net screen, a solid wall, or a building, that is, a scale model of a similar pig barn, located windward at 3 H (representative height = ridge height, 260 mm in model scale). The net screen and the solid wall both had a height of 260 mm and a length of 4 m. The net was made of polyethylene, which had a porosity of 59.5%. The solid wall was made of vinyl chloride board. The windbreak building was similar to the model barn. The positions of the windbreak and the measured pig barn are shown in figure 1. WIND TUNNEL The experiments were performed in a wind tunnel at the National Research Institute of Agricultural Engineering, Japan, which was also used by Morsing et al. (2002) and Ikeguchi and Okushima (2001). The wind tunnel was 60 m long, and the test section was 20 m long with a 3 4 m cross section. The three dimensional air velocity traversing system was controlled by a computer. Further details are given by Ikeguchi and Okushima (2001). MEASUREMENT Air velocities were measured using three dimensional hot film anemometry (Model IF300, TSI Inc., St. Paul, Minn.) with an accuracy of m/s. A traversing system with a displacement accuracy of 1 mm carried the sensor to each measuring point. Velocity measurements were made at a sampling rate of 200 Hz, and each sampling lasted 41.9 s. These were selected by considering the smallest eddy of Kolmogorov and the larger eddy caused by the fluctuation of fan rotation. The number of the measurement points outside the building was 94, 94, and 88 for the net, solid wall, and building, respectively. Inside the building, there were 790 TRANSACTIONS OF THE ASAE

3 Figure 2. Velocity measurement points. Grid intersections are measuring points, and the scale is the scale model (all units in cm). 12 measurement points. Figure 2 shows the location of the measurement points. At the reference height, which is halfway between the ridge and the eaves, a reference wind velocity (U R ) of 2.71 m/s was defined in order to obtain a wind profile corresponding to that of the full scale building. AIR VELOCITY PROFILE IN WIND TUNNEL The air velocity profile in the wind tunnel is shown in figure 3. The measurements were made at the center of the turntable without the scale building (fig. 1). As in a previous study (Ikeguchi and Okushima, 2001), the wind profile was expressed by means of the logarithmic law: z U = 11.9ln ( R 2 = 0.996, S.E. = 0.041) (1) 0.56 where U = mean air velocity in the main flow direction (x direction) (m/s) z = height (mm) 0.56 = roughness length for scale model (mm) S.E. = standard error (m/s). The roughness length of 0.56 corresponded to a full scale roughness length of 11.2 mm, which corresponded to the roughness of a typical grass field (Tsuboi, 1977, p. 103). The vertical turbulence intensity profile in the main flow direction (fig. 4) can be expressed by the following equation (Japan Architecture Center, 1996, pp ): Figure 3. Wind profile in the wind tunnel without the barn, 15 measured points z I I ( R2 u = ur = 0.65, S.E. = 0.044) (2) zr where I u = turbulence intensity (dimensionless) I ur = turbulence intensity at reference height (dimensionless), 0.09 z = height (mm) z R = reference height (mm), 190 mm in the scale model. Turbulence intensity, which expresses the degree of turbulence of flow, was defined as: σ I u u = U (3) where u is the standard deviation of instantaneous air velocities measured at 200 Hz in the x direction (m/s) EVALUATION PARAMETERS Shimogata et al. (1974) reported that mean air velocity in the main flow direction was related to contaminant concentration, and turbulence intensity was associated with plume width. Therefore, Ikeguchi and Okushima (2001) proposed the following parameters to express the degree of dispersion of an air contaminant: Height in wind tunnel, z(mm) z I u = I ur z R (R 2 =0.65) Turbulence intensity,i u (dimensionless) Figure 4. Profile of turbulence intensity in the wind tunnel without the barn. Vol. 46(3):

4 P U R = U dz G D (4) P V R = I dz G V (5) P W R = I dz G W (6) U U D = (7) U R σ I V V = U (8) σ I W W = U (9) where P U = integrated U D (dimensionless) P V = integrated turbulence intensity in the y direction (dimensionless) P W = integrated turbulence intensity in the z direction (dimensionless) G = ground level R = ridge height dz = normalized small section by ridge height (H) in the z direction. I V, I W = turbulence intensity in the y and z directions, respectively (dimensionless) U D = normalized air velocity in the x direction (dimensionless) U = measured mean air velocity in the x direction at each measuring point (m/s) U R = reference air velocity (2.71 m/s) V = standard deviation of instantaneous air velocity measured in the y direction (m/s) W = standard deviation of instantaneous air velocity measured in the z direction (m/s). The equivalent algebraic equations of 4, 5, and 6 were expressed as follows: where i j = 5 z j PU UD i i,j j H = 5 z j PV IV i i,j j H = 5 z j PW IW i i,j j H (10) (11) (12) = horizontal measured point number (i = 1 to 14; see fig. 2) = vertical measured point number up to ridge height (j = 1 to 5; see fig. 2) z j = distance between vertical measured points at j 1 and j (m); z 1 = 3.0, z 2 = 3.0, z 3 = 6.0, z 4 = 7.0, z 5 = 5.0 cm (in the scale model). P U indicates the momentum of the main flow direction and expresses the degree of the dispersion in the x direction. A larger absolute value means that the contaminant concentration is low at the actual position. P V and P W are related to plume sizes in the y and z direction, respectively. Dispersion of contaminants was evaluated in terms of the above parameters in two areas; one area was between the windbreak and the building, and the other was the leeward position of the building. In the former area, measured lines were i = 3, 4, and 5, and in the latter area they were i = 11 to 14. In each area, analysis of variance with respect to parameters was performed with the number of measured lines as replicate. RESULTS AND DISCUSSION AIRFLOW PATTERN The airflow patterns for normalized airflow vectors for each windbreak are shown in figure 5. With a building as a windbreak, the return flow behind the leeward side of the building was the smallest compared to the other windbreaks (fig. 5a). The return flow (i.e., airflow directed to the left) only occurred up to 0.5 H; airflow above the eave height and between the buildings was directed windward. Outside air entered the building through the leeward inlet, and inside the building, the air in the upper part of the building flowed in the leeward direction. For the solid wall, a return flow at the leeward side of the building occurred up to a distance of 4 H (fig. 5b). The solid wall resulted in the largest wake behind the building. The flow in the area above the eave height and between the wall and the building was likewise directed toward the wall. However, the flow near ground level was directed leeward. The air velocities between the solid wall and the building were smaller than those for the other windbreaks. Inside the building, airflow near the floor was directed to the leeward side of the building; this was the same as for the building used as a windbreak. However, the airflow near the roof was directed to the windward side. For the net screen, the wake reached a level up to 2 H (fig. 5c). The flow pattern for this case allowed the flow between the net screen and the building to be directed to the leeward side and the air velocities to be higher than those observed for the other windbreaks. The net screen seemed to act as a rectifying board. The net screen acted as a porous media, reducing the airflow velocities through it. No recirculation was seen between the net screen and the building. Inside the building, the air flowed in the windward direction. DISPERSION PARAMETERS The average values of the parameters P U, P V, P W, in the area between the windbreak and the building are listed in table 1. With respect to P U, there were significant differences (F = 36.2, (2.8) P = 0.00) among the windbreaks. For the building, the solid wall, and the net screen, the average P U values were 0.23, 0.06, and 0.43, respectively. These values indicate that, in the case of the building and the solid wall, air contaminants will likely disperse in the windward direction, and in the case of the net screen, they will likely disperse more easily in the leeward direction than in the other cases. Because the value of P U for the net screen was positive 792 TRANSACTIONS OF THE ASAE

5 U D = (a) Building U D = (b) Solid wall U D = (c) Net screen Figure 5. Airflow pattern with normalized air velocity (U D ) for each windbreak. Measured air velocity in the scale model was 2.71 U D Table 1. Average parameters (and standard deviations) between windbreaks and the building (n = 3). Building Solid Wall Net Screen P U [a] ** 0.23 (0.137) 0.06 (0.092) 0.43 (0.048) P [b] V ** 0.24 (0.005) 0.22 (0.030) 0.16 (0.021 P [c] W ** 0.29 (0.005) 0.27 (0.029) 0.15 (0.014) ** Significant difference less than 1% among the effects of three windbreaks. [a] Integration value in the z direction (height) with respect to normalized air velocity (U D ), dimensionless (eq. 7). [b] Integration value in the z direction with respect to turbulence intensity in the y direction, dimensionless (eq. 8). [c] Integration value in the z direction with respect to turbulence intensity in the z direction, dimensionless (eq. 9). in this area, air contaminants are not expected to flow from the building toward the windward side. In the case of the solid wall, the concentration of contaminants is predicted to be highest because the value of P U was smallest in this area. Significant differences between windbreaks with respect to P V (F = 12.1, (2,8) P = 0.01) and P W (F = 54.6, (2.8) P = 0.00) were found in the area between the windbreak and the building. The P V and P W values for the building were larger than those for the solid wall. This indicates that air contaminants from the leeward side building are predicted to disperse widely. Comparing the net screen and the solid wall in terms of P V and P W, the values were lower in the first case than in the latter. This result was also shown by Borrelli et al. (1989), when a porous windbreak reduced the turbulence intensity close to the windbreak. The net screen could have resulted in large amounts of small scale turbu Table 2. Average parameters (and standard deviations) at leeward side of the building (n = 4). Building Solid Wall Net Screen P U [a] * 0.49 (0.089) 0.09 (0.166) 0.25 (0.226) P [b] V 0.29 (0.014) 0.28 (0.021) 0.28 (0.018) P [c] W * 0.19 (0.017) 0.25 (0.019) 0.24 (0.031) * Significant difference less than 5% among the effects of three windbreaks. [a] Integration value in the z direction (height) with respect to normalized air velocity (U D ), dimensionless (eq. 7). [b] Integration value in the z direction with respect to turbulence intensity in the y direction, dimensionless (eq. 8). [c] Integration value in the z direction with respect to turbulence intensity in the z direction, dimensionless (eq. 9). lence due to its porosity and thereby reduced the air velocity between the net screen and the building. The average P U, P V, and P W (from 0.5 to 4 H) at the leeward side are listed in table 2. There were significant differences among windbreaks in terms of P U (F = 5.60, (2.11) P = 0.03) and P W (F = 4.7, (2.11) P = 0.04). The air contaminant concentration in the leeward direction was the smallest for the building, and largest for the solid wall. P W was the smallest for the building, indicating that the plume width in the z direction is predicted to be smallest for the building. No significant difference was observed in terms of P V. For the solid wall, a higher concentration of air contaminants is predicted to exhaust from the building on both the windward and leeward sides. The concentration in the area between the windbreak and the building is expected to be larger than that on the leeward side of the building. It Vol. 46(3):

6 would be possible to trap the air contaminants in those areas because the airflow momentums of both sides were smaller compared to the other windbreaks. Therefore, it is possible for the solid wall to inhibit dispersion of air contaminants to the outside. On the other hand, with the building, it could be difficult to trap the air contaminants because the airflow momentum was larger and the concentration of contaminants is predicted to be the lowest. The values of P U, P V, and P W at each horizontal position are shown in figures 6, 7, and 8, respectively. For all measuring points in the case of the solid wall, the absolute values of P U were smallest. This indicates that the concentration of air contaminants is predicted to be higher for the full area than for the other windbreaks. From airflow vectors (fig. 5b) and from this result, the air contaminants are predicted to disperse in the windward direction from the building through the windward opening and to create a high concentration area between the windbreak and the building. In the same area for the net screen, the concentration is predicted to be the lowest compared to the other windbreaks because the absolute values of P U were higher than those in the other cases (fig 6). In the leeward area behind the building, the concentration of contaminants is predicted to be lowest in the case of the building as a windbreak. The plume widths in the y and z directions between the windbreak and the building are predicted to be larger in the order of the net screen, the solid wall, and the building. In the leeward area of the building, the plume widths are predicted to be larger in the order of the building, the net screen, and the solid wall. The differences in plume widths in the y direction were small in this area. However, the plume in the z direction was wider for the solid wall. Because the wake of the solid wall was the largest compared to the other windbreaks, the wake lifted up the air contaminants. PROPOSED INHIBITIVE METHOD OF DISPERSING CONTAMINANTS A windward windbreak influenced the airflow patterns around and within the building. There are alternative ways to control dispersion of contaminants from the building. From this study, the best result could be achieved by trapping the air contaminants in the area between the windbreak and the building, where the airflow velocity was lower, and the air contaminants from the building flowed toward the windbreak when the solid wall was located at the windward side. This combination introduced a large recirculating zone between the solid wall and the building. The wind was diverted over the solid windbreak and over the building, making the air enter the building through the leeward side and exit from the windward side. Air contaminants may thus be trapped between the windbreak and the building, as shown in figure 9. Ikeguchi (2002) and Ikeguchi and Xin (2001) proved that sprinkling could reduce dust and ammonia concentrations in an enclosed layer house. Therefore, sprinkling is expected to reduce the concentration of air contaminants even when used outside. For example, if sprinklers are installed between the windbreak and the building, they can be used to wash the air contaminants from the building at slow speed. The sprinklers can be adjusted according to the actual weather conditions, such as wind PU (dimensionless) Building Solid Net 0.5H 2H 0.5H 2H 4H Distance from the windbreak in the scale model (m) Figure 6. Parameter P U, the integration value in the z direction with respect to normalized air velocity (U D ). PV (dimensionless) Building Solid Net 0.5H 2H 0.5H 2H 4H Distance from the windbreak in the scale model (m) Figure 7. Parameter P V, the integration value in the z direction with respect to turbulence intensity in the y direction (I V ). PW (dimensionless) Building Solid Net 0.5H 2H 0.5H 2H 4H Distance from the windbreak in the scale model (m) Figure 8. Parameter P W, the integration value in the z direction with respect to turbulence intensity in the z direction (I W ). direction, wind speed, etc. Windward windbreaks seem to have an advantage in trapping the air contaminants more easily than leeward windbreaks because the velocity of trapped air contaminants is lower. 794 TRANSACTIONS OF THE ASAE

7 Sprinkler Airflow Solid wall ÖÖÖÓÓÓ ÖÖÖ ÔÔÔÔÔÔ Polluted air Figure 9. Proposed implement to trap contaminant air from the building with windward windbreak. CONCLUSIONS A 1:20 scale model of a naturally ventilated open type pig barn and a windbreak in the windward direction at a distance of three times the ridge height of the barn were tested in a wind tunnel to investigate airflow patterns surrounding the windbreak and the building and the degree of dispersion of air contaminants from the building. The following conclusions were drawn: A windbreak consisting of a solid wall on the windward side caused the air to flow towards the windbreak. A higher reduction of the air momentum was observed for the solid wall compared to a net screen or another building. In the area between the windbreak and the building, the air contaminants were predicted to accumulate, compared to the leeward side of the building. Therefore it appears that a solid wall could trap air contaminants emitted from the building. Sprinklers installed between the windbreak and the building could be capable of removing these air contaminants. When another building was used as a windbreak, the airflow momentum was larger and the concentration of air contaminants was smaller at the leeward side of the building compared to the solid wall and the net screen. As such, it would be difficult to trap air contaminants emitted from the building. With a net screen, there was airflow towards the leeward side in the area between a windbreak and the building. As such, air contaminants are predicted to be exhausted only toward the leeward side behind the building. REFERENCES Borrelli, J., J. M. Gregory, and W. Abtew Wind barriers: A reevaluation of height, spacing, and porosity. Trans. ASAE 32(6): Bottcher, R. W., R. D. Munilla, K. M. Keener, K. E. Parbst, and G. L. Van Wicklen Windbreak walls and wet pad scrubbers for reducing odorous dust emissions from tunnel ventilated swine buildings. In Proc. International Symposium on Dust Control in Animal Production Facilities, Horsens, Denmark: Danish Institute of Agricultural Sciences, Department of Agricultural Engineering, Research Centre Bygholm. Bottcher, R. W., K. M. Keener, and R. D. Munilla Comparison of odor control mechanisms for wet pad scrubbing, indoor ozonation, windbreak walls, and biofilters. ASAE Paper No St. Joseph, Mich.: ASAE. Ikeguchi, A Ultrasonic sprayer controlling dust in experimental poultry houses. CIGR e Journal. Vol. IV. October Available at: ejournal.tamu.edu/volume4.html. Ikeguchi, A., and L. Okushima Airflow patterns related to polluted air dispersion on open free stall dairy buildings with different roof shapes. Trans. ASAE 44(6): Ikeguchi, A., and H. Xin Field evaluation of a sprinkling system for cooling commercial laying hens in Iowa. Applied Eng. in Agric. 17(2): Japan Architecture Center Guidebook of wind tunnel tests. Tokyo, Japan: Japan Architecture Center, Department of Publication (in Japanese) Liu, Q., D. S. Bundy, and S. J. Hoff The effectiveness of using tall barriers to reduce odor emission. In Proc. International Conference on Air Pollution from Agricultural Operations, Ames, Iowa: Midwest Plan Service. Morsing, S., A. Ikeguchi, J. C. Bennetsen, J. S. Strom, P. Ravn, and L. Okushima Wind induced isothermal airflow patterns in a scale model of a naturally ventilated swine barn with cathedral ceiling. Applied Eng. in Agric. 18(1): Shimogata, S., K. Sugawara, and O. Yokoyama Wind tunnel experiment of diffusion of exhaust gas emitted from roof ventilator of long factory building. Pollution Control 9(1 2): (in Japanese). Strom, J. S., and S. Morsing Automatically controlled natural ventilation in a growing and finishing pig house. J. Agric. Eng. Research 30(4): Tsuboi, Y., ed Agricultural Meteorology Handbook. Tokyo, Japan: Youkendo, Ltd. Vol. 46(3):

8 796 TRANSACTIONS OF THE ASAE