Astudy of the air quality in the breathing zone in a room with displacement ventilation

Transcription

1 Building and Environment 36 (2001) Astudy of the air quality in the breathing zone in a room with displacement ventilation H. Xing a, A. Hatton b, H.B. Awbi c; a Department of Engineering Science, University of Oxford, Oxford, UK b Arvin Industries, Warton, UK c Department of Construction Management and Engineering, The University of Reading, Whiteknights, P.O. Box 219, Reading, RG6 6AW, UK Abstract This paper is concerned with the dierence in the air quality that is perceived by the occupants (breathing zone) and that existing in the occupied zone as a whole. An environmental chamber with displacement ventilation system has been used to carry out the measurements with the presence of a heated mannequin and other heat sources. Measurements of the age of air distribution, the air exchange index and the ventilation eectiveness were carried out at dierent points in the chamber for dierent room thermal loads. CFD simulations were also carried out for the purpose of ow visualisation as well as the calculation of air velocity, temperature and age of air distribution. In addition, CFD simulations were carried out to study the eect of changing the airow rate to the chamber and the position of air inlet to extend the range of parameters. The results from the CFD simulations were compared with those from measurements and good agreement was obtained in most cases. c 2001 Elsevier Science Ltd. All rights reserved. Keywords: Indoor air quality; Breathing zone; CFD; Displacement ventilation 1. Introduction Displacement ventilation (DV) is widely used in mainland Europe and is also gaining popularity in the UK. In DV systems, cool air supplied at low level is entrained by plumes rising from heat sources. In the case of a body plume, the air that is entrained by the plume rises to the occupant s head [1] and subsequently breathed by the occupant. As a result, the air quality at the breathing zone (the nose) is expected to be better than that in other parts of the occupied zone [2]. Although a considerable amount of research work has previously been done on DV systems, there has not been any major study of the air quality at the breathing zone. Current air movement design procedures focus on achieving an average air temperature and air quality index for the occupied zone without taking into consideration the eect of the thermal plume around the occupant s body and how Corresponding author. address: h.b.awbi@rdg.ac.uk (H.B. Awbi). this inuences the occupant s perception of the indoor environment. Because the perceived air quality is the only one that aects the well-being of the room occupant this ought to be considered in the design of air distribution systems. This paper presents results of the local mean age of air, temperature distribution and air velocity in an environmental chamber using three types of DV units. Heat loads for a small oce room, i.e. a heated mannequin, a heated box (a computer simulator), heated plates xed to the wall of the chamber (to represent areas of solar illuminance on the wall) and a uorescent light were placed in the chamber. For the air quality measurements SF 6 was used as a tracer gas and for the air movement measurements omni-directional anemometers and PRT sensors were used for measuring the air speed and temperature, respectively. The aim was to nd whether there was a correlation between the air quality at a person s breathing zone and that existing in the room. In particular, it was intended to know whether there was an improvement in the air quality at the breathing zone due to the body plume entraining fresh air supplied over the oor as in the case of displacement ventilation systems /01/$ - see front matter c 2001 Elsevier Science Ltd. All rights reserved. PII: S (01)

2 810 H. Xing et al. / Building and Environment 36 (2001) Nomenclature C p (t) concentration at a point p at time =t C(0) concentration at time =0 C e (t) concentration at the exhaust at time =t C s (t) concentration at the supply at time =t p local mean age of air at point p room mean age of air n nominal time constant = room volume=air ow rate 2. Experimental set-up and test procedure 2.1. Test chamber The experiments were carried out in the environmental chamber at the University of Reading. The chamber was ventilated by a single low-level wall DV unit or two circular oor units. The chamber has external dimensions 4:0 m 3:0 m 2:52 m ceiling height. It consisted of two compartments, a small compartment to accommodate the air handling unit and the test compartment for the measurements. The dimensions of the test compartment are 2:78 m (length) 2:78 m (width) 2:3 m (height). The internal heat gain was represented by a uorescent light (36 W) and a computer box (400 mm 400 mm 400 mm) with a 150 W light bulb tted inside. In addition, a desk and a chair were placed in the chamber. To investigate the conditions with high heat loads, 2 heated plates (producing 200 W) were used to represent areas of solar illuminance on a side wall. Fig. 1 shows a plan of the test compartment and its contents. Aheated mannequin was constructed from 1 mm aluminium sheet (Fig. 2). The overall dimensions of the mannequin were 1:725 m (standing) and 1:305 m (seated) in Fig. 2. The heated mannequin. Fig. 1. Aplan of test compartment of the Reading Chamber and its contents (unit of length is mm). height, 0:50 m in width (chest), 0:23 m in thickness with a total surface area of 1:60 m 2. Heating elements inside the body, head and legs of the mannequin were controlled to provide a surface temperature equal to that of a typical naked human [3]. The output of the heating elements was determined by the room condition which was typically about 100 W. Apolyurethane tube was attached to a copper tube (the nose) inside the head and fed through the torso and out to the gas sampler. This location represented the sampling point for the breathing zone. Throughout the tests the mannequin was unclothed to avoid possible interference of the clothes with the body plume. The eects of clothing and body movement will be interesting investigations in their own right, hence these were not considered here.

3 H. Xing et al. / Building and Environment 36 (2001) Table 1 The gas sampling points and sensor locations for the tests Stand no.= Concentration Velocity Height of location sample point and temperature sample point no. sample points (m) Fig. 3. The three types of DV units: (a) at wall unit, (b) semi circular unit, (c) swirl unit. The room ventilation load was determined for each test from the measured supply and exhaust air temperatures and the ow rate. In addition, the temperatures of the chamber surfaces and the mannequin were used as boundary conditions in the CFD simulations. Three types of DV units were used in the tests: DV unit 1, was a at faced wall unit xed on to the partition wall which separates the two compartments. The size of DV unit 1 was 0:5 (width) 0:5 m (height). DV unit 2, was a semi-circular wall unit located in the same position as unit 1. The unit was 0:5 m 0:5 m high with a radius of 0:25 m. Two DV unit 3, which were oor swirl units of 0:15 m in diameter, were xed on to the suspended oor of the chamber 0:7 m from the partition wall and 0:7 m from each of the side walls. Figs. 3a c show a schematic of the three DV units. The chamber was equipped with an open ventilation system that draws air from the laboratory and exhausts the contaminated air out of the building. Four-wire Platinum Resistance Thermometer (PRT) sensors (accuracy=±0:15 K) have been used to measure the air temperature and the inside and outside surface temperatures of the chamber. Other measuring devices used in the tests were an accurate Wattmeter, DANTEC omnidirectional velocity sensors and a Bruel and Kjaer SF 6 gas sampling system. The SF 6 gas analysis system incorporated a sampling box, gas analyser and a computer with analysis and control software. The computer communicates with the measuring system and can be programmed to perform the required gas sampling measurements. Twelve gas sampling tubes, velocity and temperature sensors were positioned at dierent points in the room. Table 1 lists the sampling points and sensor locations in the tests and Fig. 1 shows the location of measuring stands Test procedure Initially 16 pilot tests were carried out with 2 simple set-ups. One with just the heated box in the room and the second with the box and the desk. These were carried out to validate the measurement techniques proposed for this work and to observe any trends in the data for a simple model before starting the complicated task of interpreting the data for a room with many components. After the pilot tests were completed a number of test congurations were investigated with the emphasis on the air quality at the breathing zone of the mannequin. Table 2 lists the 12 congurations and the room ventilation a 3, 4, 5, 7, 8, 10, 0.14, 0.55, 0.92, 11 and ,1.32,1.55 and 1.8 Breathing (Seated zone mannequin) 1.63 (Standing mannequin) Plume of (Seated mannequin mannequin) 1.8 (Standing mannequin) Inlet 6 Inlet 0.4 Exhaust 2 Exhaust 2.3 a This is for tests using a constant emission of SF 6 gas above the heated box to measure the neutral height and the concentration of gas. loads that were used with the three dierent DV units. Full details of the test condition for each DV unit can be found in Ref. [4] Measurements The rst priority in this work was to establish whether there is a dierence between the perceived air quality and the air quality in the rest of the occupied zone. Because the local mean age of air at a point represents the time the supply air takes to reach that point, this term has been assumed throughout this paper to represent the quality of the air at that point. This is considered a plausible measure of the air

4 812 H. Xing et al. / Building and Environment 36 (2001) Table 2 Congurations tested in the Reading chamber Cong. Mannequin Air Ventilation load Inlet no. posture change (W=m 2 ) temp. rate ( C) (h 1 ) DV1 DV2 DV3 1 Seated Seated Seated Seated Seated n=a 18 6 Standing Standing Standing Standing Standing n=a Seated n=a Standing n=a 18 quality since the residence time should be an indicator of the degree of contamination of the air at a point. The tracer decay gas technique was used to determine the age of air at a number of points in the room using SF 6. The local age of air at a point and the mean age of air in the room can be calculated using the following expressions: C 0 p (t)dt p = ; (1) = C(0) tc 0 e (t)dt C 0 e (t)dt : (2) The local mean age of air was calculated for all the sample points within the working compartment of the chamber (points 2 12) for each test condition, and these were averaged to obtain the room mean age. The data from these tests have been analysed and are presented in the result section. The local air exchange index (the rate of exchange of air at a particular point within the room) and the local ventilation eectiveness were also calculated using E p = n 2 p ; (3) p = C e(t) C p (t) : (4) Other tracer gas measurements were carried out using the steady state concentration method to measure the neutral height (height of the fresh air layer), and the ventilation eectiveness (VE) for buoyant pollutant removal. Here, the sampling tubes were all attached to stand 10 (see Table 1) at dierent heights to measure the concentration pro- le. The neutral height was determined by measuring the height at which the measured concentration increases significantly with a small increase in height. In this case, the local ventilation eectiveness was calculated from the measured concentration data using Eq. (5). = C e( ) C s ( ) (5) C p ( ) C s ( ) Other measurements carried out during the tests were air temperatures and velocities. 3. CFD Simulations Computational uid dynamics is a useful tool for predicting the diusion of airborne contaminants in a room. ACFD program called VORTEX [5], developed at the University of Reading, was used to provide a microscopic prediction of the diusion of the tracer gas in the whole space as well as air velocity and temperature data. Three-dimensional simulations were carried out using a nominal grid of The CFD simulations were found to be particularly useful for understanding, in greater detail, the air diusion process in the room. The CFD code was initially validated for all test conditions listed in Table 2. The age of air, temperature and velocity measurements were compared with the predictions. The code was also used to simulate the eects of the air change rate and DV diuser positions on the indoor air quality. 4. Results and discussion 4.1. Experimental results For each conguration listed in Table 2, the age of air was measured for 11 sampling points. 9 of those sampling points situated in the occupied zone were volume weighted to obtain an average. The local mean age of air at the breathing zone and the occupied zone are plotted against the room load in Figs. 4 and 5 for DV units 1 3 with seated and standing mannequin postures. Figs. 4(a) (c) show that the local mean age of air at the breathing zone of a seated mannequin is between 35 and 50% lower than that in the occupied zone for the three types DV units tested. The results would suggest that the best air quality at the breathing zone should be achieved with the at wall mounted DV unit (DV 1). However smoke tests revealed that the ow from this DV unit was not uniform and most of the supply air diverted towards the mannequin. The spread from DV unit 2 (semi-circular unit) was more uniform and thus the age of air at the breathing zone was found to be higher. For the oor DV units (DV 3), the air does not ow over the oor as expected but actually spreads from the unit at an angle to the oor. Therefore, the air entrained by the mannequin is a mixture of the supply air and room air, hence the age of the air in this case is higher. Generally, the mean age of air for the three DV units was found to decrease with an increase in ventilation load for both the breathing zone and the occupied zone up to a load of about 30 W=m 2. The results from all the DV units would suggest that, for the same air change rate the higher the inlet temperature (low ventilation load congurations) the

5 H. Xing et al. / Building and Environment 36 (2001) Fig. 4. Local mean age of air for a seated mannequin (5 ac=h): (a) DV unit 1, (b) DV unit 2, (c) DV unit 3. higher the local mean age of air is for both the breathing and occupied zones. Figs. 5(a) (c) show the local mean age of air at the breathing zone and the occupied zone for a standing mannequin in the chamber. The dierence between the data for the occupied zone and that for the breathing zone is lower than that for a seated mannequin. The average of the mean Fig. 5. Local mean age of air for a standing mannequin (5 ac=h): (a) DV unit 1, (b) DV unit 2, (c) DV unit 3. age of air for the breathing zone is only about 20% lower than that for the occupied zone in the case of DV unit 1 and about 10% lower for DV units 2 and 3. This dierence appears to get smaller as the ventilation load increases which indicates that more mixing was present at higher loads. Fig. 6 shows the eect of ventilation load on the local mean age of air for a number of points in the room. At

6 814 H. Xing et al. / Building and Environment 36 (2001) Fig. 6. Eect of ventilation load on the local mean age of air for a seated mannequin for DVI and 5 ac=h. Fig. 7. Eect of ventilation load on the local mean age of air for a standing mannequin for DVI and 5 ac=h. point 5 (stand 2 in Fig. 1), which is a point 0:5 m from DV unit 1 and 0:6 m above the oor, one can see that the local age of air increases signicantly for conguration 4 which represents a high ventilation load 43 (W=m 2 ) and a low air supply temperature (18 C). That is, the cool air from the DV unit drops to the oor quicker than in the other test congurations and miss point 5. However, changes in the ventilation load seem to have little eect on the local age of air at point 3. Point 3 is 0:5 m from the DV unit and 0:15 m above the oor (see Fig. 1 and Table 1) and lies within the supply air stream from the DV unit. Hence, changes in ventilation load do not aect the air quality in the fresh air stream close to the DV unit. Similar results were also observed for the other DV units. When considering the local mean age of air results for all the DV units, it was found that the local mean age in a plume (points 4, 8 and 9) generally increases with an increase in load for a constant inlet temperature. In Fig. 7, for DV unit 1 with a standing mannequin, one can see that changes in the ventilation load aect the local mean age of air at some points in the room. Again, at point 3 changes in the load have little eect on the local mean age of air. Furthermore, an increase in the air change rate decreased the local mean age of air at most points. The breathing zone results for a seated and a standing mannequin are compared in Figs. 8(a) (c). It can be seen that the age of air for a seated mannequin is lower than that for a standing mannequin for all the DV units, except for low ventilation loads. It can also be seen in Fig. 8(c) that the mean age of air for the seated mannequin in the case of DV unit 3 (oor DV units) is generally higher than that for the other two DV units (Figs. 8(a) and (b)). This highlights the fact pointed out earlier that the supply air from the oor DV unit does not ow over the oor and thus the air entrained by the mannequin is not all fresh air.

7 H. Xing et al. / Building and Environment 36 (2001) Fig. 8. Local mean age of air at the breathing zone (5 ac=h): (a) DV unit 1, (b) DV unit 2, (c) DV unit 3. Figs. 9(a) (c) show that there is little variation in the mean age of air with load in the occupied zone whether there is a seated or a standing mannequin in the chamber for the three DV units. The local air exchange index for a seated mannequin in a room tted with DV unit 1 is plotted in Fig. 10. It is clear that the local air quality at the breathing zone is better than that at all the other points except those close to the DV units. The ventilation eectiveness prole in the room has been calculated using the data from the constant emission measurements for obtaining the neutral height [6]. It can be seen Fig. 9. Local mean age of air in the occupied zone (5 ac=h): (a) DV unit 1, (b) DV unit 2, (c) DV unit 3. from Fig. 11 for DV unit 1 that the ventilation eectiveness at the breathing zone is twice that at another point of the same height in the room. This conrms that the mannequin entrains air from the lower fresh air displacement ow zone. Although the prole with DV unit 2 was similar to that with DV unit 1, the eectiveness at a height of 0:55 m was much lower than that for the case of DV unit 1. This could be due to the fact that DV unit 2 produced a greater spread of

8 816 H. Xing et al. / Building and Environment 36 (2001) Fig. 10. Eect of ventilation load on the local air exchange index for a seated mannequin for DVI and 5 ac=h. Fig. 12. Comparison between measured and simulated age of air for test conguration 4, DV1. Fig. 11. Ventilation eectiveness for DVI conguration 2 with a seated mannequin. Fig. 13. Comparison between measured and simulated velocity for con- guration 4, DV1. air over the oor. The ventilation eectiveness in the lower part of the chamber was much lower for DV unit 3 than for the other two DV units. However, in the upper recirculating zone the eectiveness was almost the same as that for the other DV units. The ventilation eectiveness at the breathing zone was about 5.5 compared to about 8 for both of the other two DV units. The ventilation eectiveness at the breathing zone of a standing mannequin was generally lower than that for a seated mannequin CFD results CFD simulations were initially carried out to compare the results with measurements. Generally, good agreement was found between the CFD prediction and experimental results. Figs compare the age of air, velocity and temperature between CFD and experiment results for DV1 (conguration 4). The CFD predicted the age of air (Fig. 12) well at points 8 and 9 which are in the plume of the mannequin. There are big dierences between points 11 and 12 which are located in the corner of the chamber. Fig. 14 shows that the CFD predicted the temperature very well at the measured points. Fig. 14. Comparison between measured and simulated temperature for conguration 4, DV1. Good agreement can also be found for the velocity (Fig. 13) data except for point 16, which is close to the inlet and 0:35 m above the oor. Compared to point 9, which is on the same stand as point 16 but at 0:14 m above the oor, one can see that there is a large drop in the velocity with height near the diuser. This is mainly because of the diculty in simulating the inlet diuser. Abig opening instead of a mesh inlet surface obviously reduces the inlet air ow momentum with height above the oor. Other comparisons can be found in Ref. [4].

9 H. Xing et al. / Building and Environment 36 (2001) Fig. 15. Comparison of age of air for DV1 conguration 2 at dierent air change rates CFD results. Fig. 17. Age of air for dierent diuser positions CFD results. Fig. 18. Velocity vectors obtained from CFD for conguration 2, DV1. The inlet is on wall 1, the section is in the middle of the mannequin. Fig. 16. Age of air in breathing zone and near the DV unit CFD results. The eects of the air ow rate and air inlet position on the indoor air quality were investigated by the CFD simulations for DV unit 1 conguration 2. The air ow rates were 3, 5 and 7 air changes per hour and the air inlet position were changed from the partition wall (wall 1 in Fig. 1) to the side wall and the wall behind the mannequin. Fig. 15 shows the variation of the age of air at the sampling points for air change rates of 3, 5 and 7. The age of air increases at all sampling points when the air change rate decreases. Fig. 16 shows the eect of air change rate on the age of air near the inlet (point 3, 5), in the mannequin plume (point 8) and in the breathing zone (point 9). Fig. 17 shows a comparison of the age of air at the sampling points for dierent diuser positions. It is shown that the age of air in the breathing zone is similar. This means that the position of the DV unit has no major inuence on the age of air in breathing zone. However, the age of air is slightly higher when the diuser is on wall 1 than that when it is on walls 2 and 6. This is because, in the latter two cases, the diuser is nearer to the mannequin than in the former case. There is little dierence in the age of air at point 4, which is above the computer box. Other points in the room, such as 3, 5, 10 and 11 are inuenced more by the location of the diuser. Figs are velocity vector plots in the middle plane through the mannequin with the diusers on wall 1, wall 2 and wall 6, respectively. These show clearly the air ow patterns from the diuser and the air entrained around the mannequin. Although the diuser is located in three dierent positions, the air entrained around the mannequin remains similar. Figs show plots of the temperature contours for the above three cases. Again, the plume above the mannequin is largely unaected by the diuser position. From the above results, one can also deduce that only the ow pattern at low level has changed by changing the diuser position. In summary, the eect of the inlet position on the air quality is more noticeable at certain locations such as near the inlet and room corners. The air ow pattern at low level (near oor) will be aected by the ow com-

10 818 H. Xing et al. / Building and Environment 36 (2001) ing directly from the diuser. However, the analysis shows that there is no major inuence on the air quality in the upper part of the occupied zone, especially in the thermal plumes. 5. Conclusions Fig. 19. Velocity vectors for conguration 2, DVI. The inlet is on wall 2, The section is in the middle of the mannequin. Fig. 20. Velocity vector obtained from CFD for conguration 2 DVI, the inlet is on wall 6, the section is in the middle of the mannequin. The perceived (breathing zone) air quality (represented by the mean age of air) for a seated mannequin in a room ventilated using displacement systems was between 35 and 50% better than the average air quality in the occupied zone. This dierence depends on the type of air terminal device (DV unit) which is used for supplying the air. The perceived mean age of air for a standing mannequin was between 10 and 20% better than the mean age of air in the occupied zone and this dierence also depended on the type of DV unit used. The average perceived air quality for a seated mannequin was between 15 and 33% better than that for a standing mannequin again depending on the type of DV unit used. The air quality in the occupied zone was found to be better for a semi-circular wall DV unit than for a at wall DV unit or oor DV units. However, the air quality was found to be better for wall DV units than for oor DV units. The ventilation eectiveness at the breathing zone for both the seated and standing mannequin was greater than that for a point at the same height in the chamber for the tests with all the DV units, because the Fig. 21. Temperature contours obtained from CFD for conguration 2, DV1. The inlet is on wall 1, the section is in the middle of the mannequin.

11 H. Xing et al. / Building and Environment 36 (2001) Fig. 22. Temperature contours obtained from CFD for conguration 2, DV1. The inlet is on wall 2, the section is in the middle of the mannequin. Fig. 23. Temperature contours obtained from CFD for conguration 2, DV1. The inlet is on wall 6, the section is in the middle of the mannequin. mannequin entrains fresh air from the fresh air layer on the oor into the breathing zone. The local air exchange index was found to be highest close to the DV unit and at the breathing zone the index was greater than at most other points in the room. Acomparison between the CFD simulation results and the measured results showed a good agreement at most positions except those points near the DV unit. The inlet position will have an eect on the air quality at certain locations such as near the inlet and room corners. The airow pattern at low level (near the oor) will be aected by the ow coming from the diuser. No major inuence was found on the air quality in the upper part of the occupied zone, especially in the thermal plumes, for the three diuses tested. Acknowledgements The Engineering and Physical Sciences Research Council, UK has supported this work under Grant Reference GR=K This work was also in collaboration with Halton Products Ltd and BSRIA.

12 820 H. Xing et al. / Building and Environment 36 (2001) References [1] Stymne H, Sandberg M, Mattsson M. Dispersion pattern of contaminants in a displacement ventilated room implications for demand control. Proceedings of the 12th AIVC Conference, Ottawa, p [2] Brohus H, Nielsen PV. Contaminant distribution around persons in rooms ventilated by displacement ventilation. Proceedings of the ROOMVENT 94, Cracow, Poland, Vol. 1, p [3] Olesen BW. Thermal comfort. Bruel and Kjaer Technical Review, 2, [4] Hatton A, Xing H, Awbi HB. A study of the air quality in room ventilated with displacement systems, Research Report, Department of Construction Management and Engineering, The University of Reading, Reading, UK, [5] Gan G, Awbi HB. Numerical simulation of the indoor environment. Build Environ 1994;29: [6] Xing H, Awbi HB. The neutral height in a room with displacement ventilation. In: Awbi HB, editor, Air distribution in rooms: ventilation for health and sustainable environment. Vol. 2. Oxford, UK: Elsevier, p