Tobacco Documents Online

Progro.s Report Evaluation of Ventilation Systems February 15, 1993 APPENDIX D Evaluation of a Vertical Displacement Ventilation System Physioat Systems Tasks 4 N 0 N r CA 41 ~ ~o

EVALUATION OF A VERTICAL DISPLACEMENT VENTILATION SYSTEM Bjame W. Olesen', Makoto Koganei2, G. Thomas Holbrook', Julie Seelen' and James E. Woods' 'Indoor Environment Program, Virginia Polytechnic Institute and State University, USA 2Research and Development Center, Asahi Kogyosha Co., Japan ABSTRACT The effectiveness of a vertical displacement ventilation system was evaluated when contaminants including tobacco smoke were present The system tested supplied air through a perforated floor and carpet. The system performance was evaluated using ADPI, percentage dissatisfied due to draught, vertical temperature profile, air change effectiveness, and contaminant removal effectiveness. Gaseous contaminants were simulated with tracer gas. Also, particulates, CO, and CO2 were generated by cigarettes and occupants. Several combinations of supply air flow rate and thermal loads in the occupied zone were evaluated. Because of the method of air supply, very uniform air temperature and velocity conditions were obtained. The risk of draught was found to be negligible. The air change effectiveness of the system varied from 130 to 300%, increasing with increasing supply air flow rate and decreasing with increasing thermal load in the space. Contaminant removal effectiveness was less dependent on the supply air flow rate and thermal loads than the air change effectiveness and varied, depending on type of contaminant, from 80 to 580%. INTRODUCTION The cause for many cases of Sick Building Syndrome is poor ventilation due to either inadequate flow of outdoor air or poor distribution of the air in the system or in the occupied space. The distribution of the air influences both the thermal conditions and the indoor air quality in the space. The quality of the thermal environment is affected by the air temperature uniformity in the occupied space, i.e. vertical air temperature differences, and risk for draught. Draught sensation is the combined effect of air temperature, mean air velocity, and turbulence intensity. Indoor air quality is influenced by the quality of the supply air, the emission rates of sources in the space, the degree of contaminant dilution in the space, and the degree of contaminant removal from the occupied space. The requirements for the amount of outdoor air or the air change rates listed in existing standards and codes (1,2) assume complete mixing of the air in the ventilated space. Since complete mixing does not always occur, it is necessary to take into account how efficiently the air is distributed in the occupied zone. In the case where there are zones of stagnant air and/or short circuiting, it may be necessary to increase the amount of outdoor air required by code and the total supply air flow rate. In the case of displacement flow, the amount of outdoor air required by code may be reduced. To properly assess the ventilation in the breathing zone, one must determine: (1) The air renewal/air distribution

process: How quickly "old" air is replaced with "new" air in the occupied zone; and (2) The contaminant removal process: How effectively contaminants generated in the space are removed from the space and prevented from spreading in the occupied zone. These two processes are related but not identical and therefore need to be treated separately. For both air renewal and contaminant removal it is important to distinguish between results based on room average, occupied zone average or local conditions. Complete mixing here is referred to as having an air change effectiveness of 100%. The objective of this study is to evaluate the performance of a vertical displacement ventilation system with air supply through a perforated floor. Initial results of this evaluation and' a discussion of the measurement of air change effectiveness and contaminant removal effectiveness are presented. Data were obtained for several contaminants, two different supply air flow rates, and two thermal load conditions. METHODS Tests were performed in a test room with almost adiabatic walls and ceiling, served' by a vertical displacement ventilation system supplying air through a perforated floor/carpet and returning air through grilles in the ceiling (Figure 1). All tests used 100% outside air, i.e.,, no return air was recirculated. Before the start of the tests the supply and exhaust air flows were balanced. Thermal loads consisted of fluorescent lights (total of 240 W, wall mounted near the ceiling) and five persons who were simulated by electric blankets. These blankets were folded so the surface area was similar to that of a standard person. The output wattage, for each blanket (120 W), was determined from its electrical resistance at operating conditions. Supply air flow rates were 142 and 472 l/s and thermal load levels were 14 and 44 W/m2. Gaseous contaminants were simulated with tracer gas (SF6), dosed in the center of the room at 0.9 m. Particulates and CO were generated by 6 burning cigarettes also at 0.9 m. CO2 was generated by one person sitting in the center of the room (breathing level at 1.1 m). Total Volatile Organic Compounds (TVOC) were generated by both cigarettes and occupants. When the person was in the room, the thermal load output of the blanket in the center was reduced to 50 W. ® 0 O 7'TS 0 chair a F- 0 measurement location 0 cecurn a3r gtwes 2•sm--------------- 1,7m--------------- supply supply 1:1 m _ ,/' '~ '~ -+ 'r - -+ ~ -~ - - - - - - '~ 0; mLLLI0.6m :_. Z . ._.Z :_:~. pufonned aoorkaqm Chatr a chair chair Ch2ir a (a) Room plan with tobacco smoke dosing (b) Room section with measurement heights at "T" and tracer gas dosing at "S". annotated. Figure 1. Schematic of test room.

The thermal environment in the occupied zone was evaluated measuring air velocity (3 min. mean and standard deviation), and air temperatures at four heights (0.1 m, 0.6 m, 1.1 m, 1.7 m) and at four locations (Figure 1). From these measurements ADPI, vertical air temperature differences, and percentage dissatisfied due to draft (PD) were calculated (3,4,5). ADPI was calculated as the percentage of points meeting the criteria: -1.7 °CSA:5 1.1 °C,where 0=t,-t„-8 (v,-0.15) (1) where: 0= effective draught temperature (°C); t,,, = average test zone air temperature (°C); t, = corrected air temperature (°C); v, = mean air velocity (m/s). The percentage of dissatisfied due to draught (PD) was calculated with: PD=(34-t,,,)(v,-0.05)°'0 ( 0.37 SD, + 3.14 ) 2) where: SD, = Standard deviation of air velocity. The air change effectiveness of the ventilation system was calculated by dividing the local age-of-air in. the return by the local age-of-air at 1.1 m height in the center of the room. The age-of-air in a point in the room is a measure of the time it takes the supply air to reach that location in the room (6,7). The local age-of-air in the center of the room at 1.1 m(breathing level of a sedentary person) was selected to represent the breathing level of the entire room. The age-of-air was measured using the step-up method dosing tracer gas (SF6) in the supply duct, as described by Sutciiffe (8). Contaminant removal effectiveness, e, is a measure of how effectively contaminants generated in the space are removed from the space and prevented from spreading in the occupied zone, calculated as in (9): E`=(Ce-C.)L(Cr-C,) 100% where: C. = contaminant concentration in the extract (return) air; C, = contaminant concentration in the supply air; C, = average contaminant concentration in the space. Gaseous contaminants were measured with the photo-acoustic detection method, and particulates with the piezo-balance method. RESULTS AND DISCUSSION (3) The thermal environment was very uniform (Table'1). Vertical temperature differences were less than 1.0°C between ankle (0.1 m) and head (1.1 m) for a sedentary person and less than 1.5°C for a standing person (0.1 m and 1.7 m). Mean air velocities were always lower than 0.10 m/s in the occupied zone and for most locations about 0.05 m/s. The ASHRAE-113 recommends that ADPI should be more than 80%. When the system was operated at the higher air flow, the index dropped to 65%. Since the measured air velocities in most cases were about 0.05 m/s, the corrected air temperature due to air velocity, calculated as -8 (v, - 0.15 m/s), was +0.8°C. Since the limiting effective draught temperature is 1.1 °C, the maximum acceptable difference between the test point temperature and the average zone temperature is then only 0.3°C. ADPI calculation

implies that the optimal air velocity is 0.15 m/s, but it has been demonstrated that at 0.15 m/s people may experience draft (4,10). The data for the percentage dissatisfied due to draught show that all points in the occupied zone were lower than the 15% recommended in ISO/DIS 7730 and ASHRAE 55-92. This indicates that, in this case, the criteria used for ADPI do not give a correct evaluation of the impact that the air velocity distribution may have on comfort. Thus, it is recommended that these criteria be modified for consistency with the ISO/DIS 7730 and ASHRAE 55-92 standards. Table 1. ADPI, Percentage Dissatisfied due to draught (percentage of points where PD<15% and average PD), and Air Change Effectiveness (ACE) for two different supply rates and two thermal load conditions. Load W/m2 13 44 Supply rate 142 Us (300 cfrn) I Supply rate 472 Us (1000 cfm) ADPI PD<15% 100% 94% 100% 100% PD avg 0.7% 0.3% ACE n ADPI 146%1 1 65% e test was interrupt , ere ore this v ue ne to be PD<15% 100% PD avg 1.6% ACE 300% 180% v i wi ' er testing. The air change effectiveness was always higher than 100~'o and increased with higher supply air flow rate (Table 1). At the higher supply rate and low thermal load, the air change effectiveness was about 300%. Theoretically, the maximum air change effectiveness for piston flow is 200%, based on the average age-of-air of the space measured at half the room height. Since for this study, the effectiveness was based on the local age-of-air measured at one third the room height (1.1 m level), the resulting air change effectiveness was 300%, indicating piston flow. Increasing the thermal load to 44 W/m2 decreased the air change effectiveness to around 180% probably caused by the fact that the thermal loads disturbed the piston flow pattern and increased mixing. The contaminant removal effectiveness is influenced by the supply flow rate, by the thermal loads, and by the type of source. For the higher air flow rate, the contaminant removal effectiveness was higher for all contaminants except COZ, ranging from 200 to 500% - significantly higher than for complete mixing, i.e., 100%. C02, generated by the person sitting in the room, was emitted with a lateral impulse as opposed to the other contaminants which had a negligible initial velocity. Therefore the contaminant removal effectiveness for COZ was not significantly influenced by the supply flow rate or thermal loads but mainly dominated by the fact that it had an initial velocity. The removal' effectiveness was below 100% only for the gaseous contaminants (SF6) in the case of low supply rate and high thermal load. This was because the SF6 is a "cold" source, its density is greater than air causing the SF6 to mix with the air below the dosing point, and also because in this study it was dosed above an electric blanket. For the low thermal condition, the blanket was turned off, no thermal flow was provided, and SF6 spread in all directions resulting in low local concentrations. For the high thermal condition, the blanket underneath the dosing point was turned on, providing a thermal flow upward, taking the SF6 upward in its plume, resulting in higher concentrations. The cigarettes are "warm" sources, although they do not significantly influence the heat load in the room, they influence the local thermal conditions because of their small heat plumes. These heat plumes take the generated contaminants upwards. See concentration profiles for CO, TVOC, and particulates in Figure 2.

Table 2. Contaminant Removal Effectiveness (CRE) for different contaminants and for two different supply rates and two thermal load conditions. Conta- Load CRE (%); supply rate 1421/s CRE (%); supply rate 4721/s minant W/m2 1.1 m 1.7 m average 1.1 m 1.7 m average 13 216 188 201 - - - COZ 44 217 112 204 176 106 213 13 259 237 241 - - - CO 44 382 241r 315 392 451 582 13 363 374 364 - - - TVOC 44 340 238 297 372 404 534 13 219 219 241 - - - SF6 44 92 81 133 403 220 411 parti- 13 323 - - - - - culate 44 298 - - 461 - - (142 U.) (4n u.) Contaminant Symbol (Ce Cs) (Ce' Cs) C02 0 106ppm 32.9ppm CO 13 8.85ppm 2.98ppm TVOC ~ 4.S7PPm 1.74ppm Open symbols for 1421/s. SF6 O 5.30ppm 1.84ppm solid symbols for 472 I/s. Particulate ~ 0.68mg/m3 026mg/m3 std. dev. velocity 3.2 2.4 0.0 0.0 0.5 1.0 1.5 (Ce Cl)/(Cr'Cs) 0.0 0.5 1.0 1.5 (Ce Cd/(Cr Cg) 3.2 0.8 0.0 0.00 0.03 0.06 0.09 Yelocity, m/s Figure 2. Profiles of data averaged over four locations in test room at 44W/m2 and at two supply flow rates.

The problem with obtaining meaningful contaminant removal data is determining the proper sampling locations to arrive at a representative data set. The contaminant concentration can be measured at one height, e.g., breathing level, or averaged over a volume, e.g., the entire occupied zone. Also the distance from the source to sample locations should be taken into account as well as the source temperature and type of contaminant. Until a standard methodology is developed, the sample locations utilized to characterize the room or zone of interest, and the source and its contaminants should be included when presenting contaminant removal effectiveness data. CONCLUSION Even though the data reported here are limited and the tests must be repeated to obtain statistically significant results, it is clear that the performance of a displacement ventilation system does not depend solely on the system itself. Other factors such as the thermal load condition, the total supply rate, and the type and location of the contaminant generated in the room to play important roles in determining the overall system performance. Criteria used in determining the ADPI must be reviewed to reflect results on peoples' perception of draught. The definition and measurement method for contaminant removal effectiveness needs to be further detailed. ACKNOWLEDGEMENT We are pleased to acknowledge that this study is being funded by Philip Morris USA. We also gratefully acknowledge the support from Asahi Kogyosha Co., Ltd. for Dr. Koganei as a Visiting Scholar at the Virginia Tech Indoor Environment Program. REFERENCES 1. ASHRAE. Std. 62-89: Ventilation for Acceptable Indoor Air Quality. Atlanta, 1989 2. EEC Report No. 11. Guidelines for Ventilation Requirements in Buildings. European Concerted Action (COST-613). Commission of the European Communities, Luxembourg, 1992. 3. ASHRAE. Std. 113-90: Method of Testing for Room Air Diffusion. Atlanta, 1990. 4. ASHRAE. Std. 55-92: Thermal Environmental Conditions for Human Occupancy. Atlanta, 1992. 5. ISO 7730. Moderate thermal environments - Determination of the PMV and PPD indices and specification of the conditions for thermal comfort. International Standards Organization, Geneva, 1984. Revised: 1993. 6. Sandberg M, Sjoberg M. The use of movements for assessing air quality in ventilation rooms. Buildings and Env., 1983;18:181-197. 7. Skaaret E, Mathisen HM. Ventilation Efficiency. Env. Int. 1982;8 8. Sutcliffe H. Technical Note No. 28: A Guide to Air Change Efficiency. AIVC, UK, 1990. 9. Brouns C, Waters B. Technical Note No. 28-2: A Guide to Contaminant Removal Effectiveness. AIVC, UK, 1991. 10. Fanger PO, Melikow AK, Hanzawa H, Ring J. Air turbulence and sensation of draught. Energy and Buildings 1988;12: #1.