- Email: [email protected]
Experimental study on air change effectiveness: Improving air distribution with all-air heating systems
Building and Environment 125 (2017) 515e527
Contents lists available at ScienceDirect
Building and Environment journal homepage: www.elsevier.com/locate/buildenv
Experimental study on air change effectiveness: Improving air distribution with all-air heating systems Hideyuki Amai a, b, *, Shichao Liu b, c, Atila Novoselac b a
Obayashi Corporation, USA Department of Civil, Architectural and Environmental Engineering, The University of Texas at Austin, Austin, TX, USA c Department of Civil and Environmental Engineering, Worcester Polytechnic Institute, USA b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 10 August 2017 Received in revised form 12 September 2017 Accepted 12 September 2017 Available online 18 September 2017
A major challenge of all-air heating applications is poor air distribution, which is often associated with a high temperature stratification. The two metrics that are commonly used for design and assessment of supply air distribution in the space are: air distribution performance index (ADPI) and the Air Change Effectiveness (E), respectively. All-air heating systems often produce stagnant air in the occupied part of the room. In this case, E may be very low while relatively uniform temperature in this occupied zone results in acceptable ADPI. Since ventilation design is based on ADPI, many all-air heating systems often produce very low E. This experiment based study identifies situation with very low E and provides simple strategies to improve it. The study provides additional design criteria to the ADPI diffuser selection guide that helps with optimal diffuser selection and adjustments. The results show that additional design criteria significantly improve E as well as temperature distribution, measured by temperature effectiveness (ƐT), with all-air heating systems. Appropriate adjustment of the diffuser may improve E and ƐT up to 30%, while the lower supply-room air temperature difference may increase E and ƐT in average 75% and 45%, respectively. Also, proper return air inlet location significantly improves E and ƐT for all-air heating. However, there are certain trades off as: the diffuser adjustment also may require seasonal adjustment for cooling and heating operation, lower supply air temperature difference requires more fan power, and floor exhaust placement may need more space for duct work. © 2017 Elsevier Ltd. All rights reserved.
Keywords: Ventilation effectiveness Mixing ventilation ADPI All-air heating Diffuser adjustments
1. Introduction Mixing ventilation is the most common type of air distribution, used much more than alternative air distribution strategies such as piston, displacement, stratum, and personalized ventilation [1]. Achieving the ideal mixing in the room depends on an air distribution design; it can apply various types of supply air diffusers and return air inlets (exhaust) to create a mixing ventilation in different space types [2]. The impact of supply air diffusers on room air temperature and velocity under cooling applications are well studied [3], while there is much less information on the impact on the fresh air dispersion in heating applications. Also, there is very little information on the impacts that location of exhaust has on the air distribution. A major challenge of mixing ventilation is in heating application (all-air heating) as it causes poor air
* Corresponding author. 55W 5TH AVE APT 01A, San Mateo, CA, USA. E-mail address: [email protected] (H. Amai). https://doi.org/10.1016/j.buildenv.2017.09.017 0360-1323/© 2017 Elsevier Ltd. All rights reserved.
distribution due to a high temperature stratification [4e8]. Fisk et al. [4] conducted experiments that used overhead all-air-heating system that supplied minimum air supply flow rate of typical VAV systems. The air change effectiveness was significantly lower than 1.0 in each experiment. The measured air change effectiveness was in the range of 0.69e0.91 with mean value of 0.81. Offermann et al. [5] measured ventilation effectiveness and ADPI under heating conditions. For the ceiling supply/return configuration, ventilation effectiveness was 0.73 when temperature difference of supply air temperature and room average temperature was 8 C. Krajcik et al. [6,7], measured air change efficiency and temperature effectiveness in a test chamber with various combinations of radiant floor heating and mixing ventilation. Theses all-air heating systems often produce stagnant air in the occupied space of the room with relatively uniform low temperature in this stagnant zone. However, with stagnant cold air in occupied zone, the fresh hot supply air short circuit in the upper part of the room causes very poor ventilation effectiveness in the occupied space. In any application, it is important to consider both thermal
516
H. Amai et al. / Building and Environment 125 (2017) 515e527
comfort of occupants and effective distribution of supply air to the occupants not only for cooling but also for heating. An effective air distribution improves indoor air quality by providing fresh air in all areas of the occupied zone and thermal comfort by providing uniform temperature distribution and acceptable velocities [9]. The widely used design method for indoor air distribution is Air Diffusion Performance Index (ADPI) [11]. The index, percentage of the whole occupied zone with acceptable air velocity and temperature ranges, quantifies the performance of supply air diffusers to create a uniform thermal environment in the occupied zone. ADPI provides engineers and practitioners a simple tool to design air distribution that is usually associated with complex heat transfer and fluid dynamics [12]. The concept of ADPI first introduced and developed the correlations between diffuser properties and indoor air distribution performance for five types of terminal diffusers at cooling conditions. Recent studies have updated the ADPI metric to include 15 diffuser types [3,13e15]. Also, these studies removed the deficiency of ADPI method related to applicability to only cooling systems; the new ADPI guideline is valid for both heating and cooling conditions. Furthermore, the studies identified the conditions where poor ventilation effectiveness appears in all air heating. Specifically, when warm supply air is delivered in the room with a weak jet momentum, the supply air short circuit in the ceiling level generating stagnant air in the occupied zone. Locations of supply air diffuser and exhaust impact air distribution and their ventilation effectiveness in the space have been investigated by many researchers [16e18]. Sinha et al. [16] compared impact of different inlet and outlet locations using models and computational fluid dynamics. The study found that the most effective combination of inlet and outlet positioning is with inlet near the floor and exhaust near the ceiling. When considering position of air supplies in the upper part of the room, Lee et al. [17] experimentally compared high wall jet from grill diffusers with typical ceiling diffusers. Their results show that the air inlet position and type are important determinants in the distribution of airborne contaminant concentrations. Overall, the ceiling diffuser produced more efficient ventilation than the wall jet air inlet. In Khan's study [18], the arrangements of wall inlet and outlet greatly influence contaminant concentration. However, the influence of the outlet location is minimal with ceiling diffuser inlet. Since the air near exhaust is not driven by jet momentum but by negative pressure in the air, the velocities near exhausts are relatively small. Therefore, the effects of the exhaust location on room airflow pattern are relatively small in most applications. However, the exhaust location influences Air Change Effectiveness and Contaminant Removal Effectiveness [2]. Location of exhaust is even more important when mixing ventilation provides heating. ASHRAE standard 62.1 [19] assumes Air Change Effectiveness (E) of 0.8 when a 0.8 m/s jet does not reach to the lower part of the room or when the supply-room temperature difference is larger than 8 C. In the literature, Ventilation Effectiveness is defined as description of air distribution system's ability to remove internally generated pollutants from a building zone or space, while the E is an air distribution system's ability to deliver ventilation air to occupied zone or space [10]. In the following section of the paper, we will use E as the index that describes quality of the air distribution and the ventilation effectiveness. Our previous study [20] conducted extensive experiments on E in mixing ventilation. These data combined with the results from the latest ADPI study [3] and provided comprehensive data set on diffuser performance considering both uniformity of the temperature field and range of E in the both cooling and heating applications. E and temperature effectiveness (ƐT) were slightly higher than 1.0 in cooling applications. However in heating application, E
was significantly decreased at small T0.25/L even though ADPI was in acceptable range (ADPI higher than 80%); this poor E due to the short circuit of the supply air. E within the acceptable ADPI range of T0.25/L (recommended range) was 0.56e0.87. This short circuiting of hot air puts performance of many diffusers in the range that is lower than specified value in ASHRAE standard 62.1, E > 0.8. To overcome the challenges of poor ventilation under heating conditions, this study examines simple strategies that may improve E and ƐT under heating conditions, while maintaining acceptable ADPI; specifically, it evaluates the impacts of diffuser deflector adjustment with linear slot diffusers and adjustable blades grills, room-supply air temperature difference with vertical flow of linear slot diffusers, and exhaust locations with adjustable blade grill. In addition, it provides new diffuser selection data in the form of the ADPI for diffusers with a vertical jet projection (vertical flow) categorized as Group E in ASHRAE Handbook [21]. The following section briefly summarizes indices utilized to evaluate the performance of the mixing ventilation systems used in this study. It is followed by the methodology section that describes experimental setup and matrix of experiments. The results section compares different indices to evaluate impacts of analysed ventilation performance strategies, while the conclusion section synthesizes major findings. 2. Theoretical background and description of used indices This section explains three indices: Air Diffusion Performance Index (ADPI), Air change effectiveness (E) and Temperature effectiveness (ƐT), that are most commonly utilized to evaluate air distribution performance. It provides short overview of variables and physical phenomena and explains the basics for interpretation of the tables and graphs in the results section. 2.1. ADPI ADPI is defined as the percentage of the occupied zone that maintains acceptable velocity and temperature. The region of acceptable velocity and temperature is determined by local Effective Draft Temperature (EDT) that combines air temperature difference and air speed [22e24]. EDT for cooling condition is defined as,
EDTðqÞ ¼ Ti Ta 8:0ðVi 0:15Þ½ C
(1)
where Ti is temperature at the test point, i; Ta is spacious average temperature ( C) and Vi is local air speed (m/s). EDT for heating condition is defined as [3,13].
EDTðqÞ ¼ Ti Ta 9:1ðVi 0:15Þ½ C
(2)
ADPI incorporates the throw and the characteristic length, and it provides design variables for selecting diffusers. The dominant diffuser property for air distribution is the supply jet throw length at which the jet velocity decreases to a selected terminal value of 0.25 m/s (T0.25) [25]. The characteristic length (L) describes the room geometry in the form of the distance at which jet travels. The ratio T0.25/L is a dimensionless number that characterizes a supply diffuser momentum (including capacity of a diffuser to mix/entrain surrounding air) for given flow rate. T0.25/L has the largest impact on ADPI, and together T0.25/L and ADPI are used in the diffuser selection guideline provides by ASHRAE Handbook [11]. The guideline lists the relationships between ADPI and a dimensionless ratio of T0.25/L for various diffuser types at different thermal loads. One is able to design an HVAC terminal system, such as diffuser selection and layout, by ensuring T0.25/L of the system renders ADPI greater
H. Amai et al. / Building and Environment 125 (2017) 515e527
than 80% which is the threshold of air distribution design. Since T0.25/L is a critical parameter determining air distribution [20], we will utilize it to assess the performance of air distribution throughout this study. T0.25 is defined by different air flow rate in isothermal conditions from manufacture's catalog data for several diffuser types including adjustable blade diffusers [11]. However, T0.25/L for a vertical flow has not defined in the literature. Throw of warm vertical jet may significantly travel less than isothermal jet by strongly influence of buoyancy force. In this study, manufacture's catalog T0.25 for vertical flow (isothermal flow) is modified by measured supply air temperature and room air temperature, utilizing empirical chart in ASHRAE Handbook [21]. From this chart, the throw correction factor for vertical flow regarding room supply air temperature differences is determined. In addition, ceiling height is defined L for vertical flow. 2.2. Air change effectiveness (E) The Air Change Effectiveness (E) as defined in ASHRAE standard 129 [26] is often utilized as ventilation effectiveness index. It defines how well the supply air is distributed through the occupied space of the room. There are numerous ways to experimentally determine ventilation effectiveness [27e29]. However, this research utilizes the age of air approach to calculate E. The age of air is determined from the tracer gas decay test. Age of air at a location i in the space is defined as
1 Ai ¼ C0
Z∞ Ci ðtÞdt
(3)
0
where C0 is the initial concentration of tracer gas at the time t ¼ 0 and C(t) is the decayed concentration measured at time t. The local air change effectiveness, Ei, is defined as
Ei ¼
tn Ai
½
(4)
In the equation above tn is the nominal time constant, which is reciprocal value of the air change rate in the room (tn ¼ Room Volume/Volume flow rate). The arithmetic mean of Ei in the occupied zone of the room is defined as E. For perfect mixing, E is 1 and is smaller than 1 in a case where the occupied zone has a stagnant air. It can be larger than 1 where the supply air reach occupied zone more effectively than with perfect mixing, i.e. with a displacement ventilation [8]. 2.3. Temperature effectiveness (ƐT) Temperature effectiveness [30], known as ventilation effectiveness for heat removal [31], evaluates the temperature gradient in the test space. The temperature effectiveness (ƐT) is defined as
εT ¼
TSA TEA 100½% TSA 〈T〉O
517
(5)
where TSA is the supply air temperature, TEA is the exhaust air temperature and0 is the average temperature in occupied space. 3. Methodology The first part of the methodology section describes the diffusers tested in this study. The second part explains the two series of experiments: (1) experiments related to ADPI and (2) experiments related to E and ƐT. These two series of experiments were conducted in the same test room at the University of Texas at Austin and the testing methodologies followed previous studies procedures described greater details in Refs. [3,13,14,20]. 3.1. Tested diffusers Fig. 1a shows the linear slot diffusers (Model SDS75, frame size 190 mm 1200 mm, Price Industries, Inc.) with vertical flow used in the experiments. The diffusers provide various airflow patterns by adjustabling deflectors. The performance of linear slot diffusers with horizontal projection can be found in the previous study related to air change effectiveness [20]. The vertical flow allows to supply primary air directly to the occupied space and may increase E and ƐT. However, a higher air velocity in occupied space may compromise ADPI. Fig. 1b shows adjustable blade diffusers (Model 51DV, frame size: 150 mm 600 mm, Nailor HVAC, Inc.) used as high side wall diffusers. This specific model is a good representative of all high side wall adjustable blade diffusers [3], [15]. The diffusers also allow different airflow directions by adjustment of blade angles. To evaluate the impact of adjustments on E and ƐT, blades angle were set three patterns: 0 horizontal, 45 Upward and 45 Downward. The ADPI with each adjustments was determined in our previous studies [3], [15]. In previous studies, 45 Upward projection allowed supply air to easily attch to the ceiling and slide along the ceiling due to Coanda effect, resulting in better ADPI under cooling conditions. However, it may cause higher thermal stratifications and lower E under heating conditions. 45 Downward projection directly supplies air to the occupied zone which may increase E. However, it may also cause lower ADPI because of higher velocity in occupied space. 0 Horizontal blade position was considered as the nominal setting, and different exhaust locations were tested with nominal setting. 3.2. ADPI measurements ADPI measurements were conducted in a test chamber located
Fig. 1. Diffusers used in this study: a) Linear slot diffuser with vertical flow, b) Adjustable blade grill with 0 Horizontal, 45 Upward, and 45 Downward adjustment.
518
H. Amai et al. / Building and Environment 125 (2017) 515e527
Fig. 2. Experimental setup of a test room for ADPI measurements. (Chamber geometry, temperature and velocity measurement locations).
at the Center for Energy and Environmental Resources at the University of Texas at Austin, with the size of 5.5 m 4.5 m 2.7 m. The chamber is equipped with a precise environmental control system. Detailed technical specification for this state-of-the-art indoor environment related research facility can be found in our previous publication [32]. Fig. 2 illustrates experimental setups of the test room for ADPI measurements. ADPI measurements were conducted under the same chamber and setups as our previous studies [3] [13], [14]. Heating load was simulated by adjusting the temperature of the cooled wall connected to a dedicated chiller, which together mimicked exterior wall or window in winter condition. Although particle image velocimetry (PIV) measurement [33] can be used to better visualizing the airflow fields, velocity and temperature at sixty locations in the occupied zone at four different heights above the floor (0.1 m, 0.6 m, 1.1 m, and 1.7 m) were measured with twelve hot-sphere anemometers (HT-400, SENSOR, Poland, accuracy: ±0.03 m/s þ -3%, temperature: ±0.2 C). The measurements were repeated five times for each experiment to obtain sixty locations
with the twelve available sensors. At the same time, the vertical temperature (0.1 m, 0.6 m, 1.1 m, 1.4 m, 1.8 m, 2.2 m) was measured at five different locations by thermistors (Model 44033, OMEGA, Accuracy: ±0.1 C). Furthermore, supply and exhaust air temperatures were monitored during the experiments to ensure the stability of the chamber conditions. The uncertainty of ADPI measurement depends on the accuracy of the instruments used and several factors relate to experimental setup; the detailed procedure for calculation of this uncertainty is described in our previous study [13]. Due to the high accuracy of velocity and temperature sensors in the studies, the uncertainty is ±2.7% of absolute value of ADPI. 3.2.1. Experimental matrix for ADPI measurements Table 1 shows the experimental conditions for ADPI measurements. 2 slots and 4 slots linear slot diffusers with vertical flow were tested. Each setup was consisted of several cases with various T0.25/L (related to air change rate). For instance, Cases (A.1-6) had the same experimental setup but different T0.25/L such as 0.5 in Case
Table 1 Experimental conditions for ADPI measurements. Case#
Supply Diffusers Type
A.1-6
Linear slots) A.7-12 Linear slots) A.13- Linear 18 slots) A.19- Linear 24 slots)
Adjustment Slot Diffuser (2 Slot Diffuser (4 Slot Diffuser (2 Slot Diffuser (4
Vertical
T0.25/L [-] (Air change rate [h1])
Temperature difference between supply and exhaust DT ¼ TEA-TSA [ C]
Exhaust Location
0.5 (2.1), (8.6) 0.4 (2.1), (8.6) 0.6 (2.1), (8.6) 0.5 (2.1), (8.6)
0.8 (3.3), 1.1 (4.5), 1.4 (5.8), 1.7 (7.2), 2.0
5
EX Cn
0.7 (3.3), 0.9 (4.5), 1.2 (5.8), 1.4 (7.2), 1.8
5
1.0 (3.3), 1.3 (4.5), 1.6 (5.8), 1.9 (7.2), 2.2
2
0.8(3.3), 1.1 (4.5), 1.4 (5.8), 1.7 (7.2), 1.9
2
H. Amai et al. / Building and Environment 125 (2017) 515e527
A.1 and 0.8 in Case A.2. The T0.25 derived as described in theoretical background and description of used indices section. All experiments were conducted with heating conditions. 3.3. E and ƐT measurements Fig. 3 illustrates experimental setup for E and ƐT measurements. The experimental setup allowed adjustable blade diffusers with high side wall position, and 2 slots and 4 slots linear slot diffusers with vertical flow. Dimensions of the plenum box for adjustable blade diffuser is also described in Fig. 3. Furthermore, Fig. 3 shows the five specific exhaust locations to evaluate the impact of the exhaust locations. There were three locations for ceiling mounted positions: EX Cd, EX Cn, and EX Cw, and the two locations near floor: EX Fd and EX Fw. Tracer gas decay test using CO2 as tracer gas was conducted to measure Ei and calculate E as described in previous section. The decay of CO2 concentration was measured by in-situ sensors
519
(TELAIR model 7001, Accuracy: ± 50 ppm or 5%) at 18 locations simultaneously (Fig. 3 provide the exact sensor locations). Tracer gas injection was ensured near perfect mixing throughout the space by 3 powerful mixing fans (Lasko Box Fan, Model #B20200, 55 cm 11 cm 57 cm) before the tracer gas decay started. In addition, vertical temperature distributions, supply and exhaust temperature were measured. Average temperature at 0.1 m, 0.6 m, 1.1 m, 1.4 m, and 1.8 m (T0.1, T0.6, T1.1, T1.4 and T1.8) above the floor utilized to determine ƐT. The uncertainty of measurements of air change effectiveness has been already discussed in our previous study [20]. From repetitive experiments, the uncertainty of local Ei was 6% on average and maximum 14% and E was 6% on average and maximum 11%. 3.3.1. Experimental matrix for E and ƐT measurements Table 2 shows the experimental conditions for E and ƐT measurements. The experiments were conducted with 2 slots and 4 slots linear slot diffusers with vertical flow, and adjustable blade
Fig. 3. Experimental setup of a test room for E and ƐT measurements (Chamber, adjustable blade grille box and exhaust box geometry, temperature and CO2 concentration measurement locations).
520
H. Amai et al. / Building and Environment 125 (2017) 515e527
Table 2 Experimental conditions for E and ƐT measurements. Case# Supply Diffusers Type B.1-8 Linear Slot Diffuser (2 slots) B.9- Linear Slot Diffuser 17 (4 slots) B.18- Linear Slot Diffuser 23 (2 slots) B.24- Linear Slot Diffuser 29 (4 slots) B.30- Adjustable Blade 36 Grill B.37- Adjustable Blade 43 Grill B.44- Adjustable Blade 50 Grill B.51- Adjustable Blade 57 Grill B.58- Adjustable Blade 63 Grill B.64- Adjustable Blade 71 Grill B.72- Adjustable Blade 78 Grill
Adjustment Vertical
T0.25/L [-] (Air change rate [h1])
EX Cn
0.7 (3.3), 0.9 (4.4), 0.9 (4.5), 1.2 (5.8), 1.4 5 (8.6) 1.3 (4.5), 1.6 (5.8), 1.9 (7.2), 2.2 (8.6) 2
EX Cn EX Cn
2
EX Cn
0.9(2.1), 1.2 (3.3), 1.4 (4.5), 1.6 (5.8), 1.9 (7.2), 2.1 5
EX Cn
0.4(2.1), 0.5 (3.3), 0.7 (4.5), 0.8 (5.8), 0.8 (7.2), 0.9 5
EX Cn
0.4(2.1), 0.5 (3.3), 0.7 (4.5), 0.8 (5.8), 0.8 (7.2), 0.9 5
EX Cn
0.9(2.1), 1.2 (3.3), 1.4 (4.5), 1.6 (5.8), 1.9 (7.2), 2.1 5
EX Cw
0.9(2.1), 1.2 (3.3), 1.4 (4.5), 1.6 (5.8), 1.9 (7.2), 2.1 5
EX Cd
0.9(2.1), 1.2 (3.3), 1.4 (4.5), 1.6 (5.8), 1.9 (7.2), 2.1 5
EX Fw
0.9(2.1), 1.2 (3.3), 1.4 (4.5), 1.6 (5.8), 1.9 (7.2), 2.1 5
EX Fd
Vertical Vertical
0.5 (2.1), 0.8(3.3), 1.1 (4.5), 1.4 (5.8), 1.7 (7.2), 2.0 (8.6)
0 Horizontal 0.8 (1.6), (8.6) 45 Upward 0.3 (1.6), (8.6) 0.3 (1.6), 45 Downward (8.6) 0 Horizontal 0.8 (1.6), (8.6) 0 Horizontal 0.8 (1.6), (8.6) 0 Horizontal 0.8 (1.6), (8.6) 0 Horizontal 0.8 (1.6), (8.6)
Exhaust Location
0.8 (3.3), 1.1 (4.4), 1.5 (5.8), 1.7 (6.9), 1.7 5
0.5 (2.1), (7.2), 2.0 0.4 (2.1), (6.9), 1.5 0.6 (2.1),
Vertical
Temperature difference between supply and exhaust (DT ¼ TEA-TSA) [ C]
0.7 (3.0), (8.6) 0.6 (3.0), (7.2), 1.8 1.0 (3.3),
Fig. 4. Linear slot diffusers with vertical flow under DT ¼ -5 C (Case A.1-6 Case A.7-12, Case B.1-8 and Case B.9-17), a) and c): T0.25/L vs. E (left side y-axis) and ADPI (right side yaxis), b) and d): T0.25/L vs. ƐT. *1; Solid mark: TSA < 8 C above0, Mark without fill: TSA > 8 C above 0, *2; T0.25/L is based on corrected T0.25 from 0 and TSA.
H. Amai et al. / Building and Environment 125 (2017) 515e527
diffusers with three different adjustments under heating conditions. In addition, five different exhaust locations with nominal settings of adjustable blade diffusers were tested. 4. Results This section is divided into two parts, presenting results from two series of experiments: 1) experiments with linear slot diffusers with vertical flow, and 2) experiments with adjustable blade diffusers with different deflector angles and exhaust locations. 4.1. Vertical flow with linear slot diffusers The first part examines the range of T0.25/L that can achieve ADPI higher than 80% and according E and ƐT. Then we discusse the results of ADPI, E and ƐT with lower temperature difference between supply and exhaust (DT). Finally, the section examines some results of velocity and temperature measurements from ADPI experiments. 4.1.1. ADPI, E and ƐT Fig. 4 shows the results of ADPI, E and ƐT with DT ¼ -5 C. Fig. 4 displays the results of 2 slots diffusers (Fig. 4a and b) slots diffusers (Fig. 4c and d). The solid marks indicate that supply air
521
temperatures (TSA) is less than 8 C above average occupied space temperature (0). Marks without fill indicate that TSA is 8 C or more than 0. With 2 slots linear slot diffusers (Fig. 4a), higher ADPI found in lower and higher T0.25/L. Although ADPI is high in lower T0.25/L, E and ƐT are low. The range of T0.25/L that can achieve ADPI higher than 80% with TSA less than 8 C above 0 (herein after recommended range) is 1.6e2.1. E and ƐT are approximately 0.9 within recommended range. The small momentum of supply air at low T0.25/L cannot provide enough mixing in the space and causes high thermal stratification and low ADPI. Conversely, a strong vertical momentum increases overall mixing performance of the space, resulting in higher ADPI, E and ƐT; however the excessive air speed may resuts in discomfort asociated with a draft in the area below the diffuser. Different from the results of 2 slots, ADPI is decreased as T0.25/L is increased with 4 slots diffusers (Fig. 4c). ADPI higher than 80% is found only in lower T0.25/L at which high thermal stratification causes lower E and ƐT. However, higher T0.25/L improves both E and ƐT due to an increased mixing effect. 4 slot diffusers performs quite different from 2 slot diffusers regarding ADPI. This will be further discuss in the later results section with descriptions of room air velocity and temperature fileds. Fig. 5 shows the results of ADPI, E and ƐT measurements with
Fig. 5. Linear slot diffusers with vertical flow under DT ¼ -2±2 C (Case A.13-18, Case A.24-29, Case B.18-23 and Case B.24-29), a) and c): T0.25/L vs. E (left side y-axis) and ADPI (right side y-axis), b) and d): T0.25/L vs. ƐT. *1; Solid marks: TSA < 8 C above0, Marks without fill: TSA > 8 C above 0, *2; T0.25/L is based on corrected T0.25 from 0 and TSA.
522
H. Amai et al. / Building and Environment 125 (2017) 515e527
DT ¼ -2. ADPI is higher than 80% for both 2 slots and 4 shots diffusers within the tested ranges of T0.25/L, implying that small DT enhances ADPI significantly. The highest E is greater with 2 slots diffusers than with 4 slots diffusers: E is 1.1 with 2 slots diffusers, 0.8 with 4 slots diffusers respectively. The highest ƐT is also greater with 2 slots diffusers than with 4 slots diffusers: 1.05 with 2 slots diffusers, 0.9 with 4 slots diffusers respectively.
4.1.2. Room air velocity and temperature Fig. 6 shows the results of velocity and temperature fields with ADPI measurements (Case A.1, A.3, A.5, and A.12). Temperature fields show with temperature differences between supply air and point of measurements (DTSA-i ¼ TSA-Ti). For Case A.1 (2 slots, T0.25/L: 0.5 (2.1 h1)), air velocity below 1.7 m is less than 0.25 m/s. Temperature stratification (maximum temperature difference within
Fig. 6. Velocity and temperature fields in ADPI measurements (Case A.1, A.3, A.5, and A.12). *Refer to Fig. 2 for location of measuring points (L,C,R and D1-D5).
H. Amai et al. / Building and Environment 125 (2017) 515e527
occupied space) is less than 2 C. For Case A.3 (2 slots, T0.25/L: 1.1 (4.5 h1)), jet from the diffuser (velocity higher than 0.25 m/s) reaches to 1.7 m. Temperature stratification is 4 C between 0.1 m and 1.7 m. Because of the bouyance effect and weak jet from diffuser, jet from the diffuser could not reach to the bottom ended in higher thermal stratification. It is considered that thermal stratification in occupied zone that results in low ADPI. In Case A.5 (2 slots, T/L0.25: 1.7 (7.2 h1)), jet from diffuser reaches to 0.6 m and temperature stratification it the lowest among shown results. Finally, with Case A.12 (4 slots, T/L0.25: 1.8 (8.6 h1)), jet from diffuser reached to 1.1 m. Temperature stratification is high and difference is about 3.5 C. Although Case 12 has almost same T0.25/L as Case A.5 calculated by the method in theoretical background section, the results of velocity and temperature fields are different. It is considered that correction factor calculated does not accurately normalize throw of 2 slots and 4 slots diffusers since the emprical table in ASHRAE handbook [21] is with supply opening of aspect
523
ratio 1.0 and air flow along with the perimeter wall. It is also considered the different diffusers' width affect temperature fields. 4.2. Adjustable blade diffusers with high side wall supply This section examines results of adjustable blade diffusers. The first paragraph examines different blade adjustment. The results combined with ADPI measurements from previous study [3]. The next paragraph discusses the results of different exhaust locations with nominal adjustment. 4.2.1. ADPI, E and ƐT Fig. 7 shows the results of ADPI, E and ƐT measurements with different blade angles: 0 Horizontal, 45 Downward, and 45 . ADPI shown in Fig. 7a, c and 7e with right side of y-axis extracted from a previous study [3]. The solid marks indicate that TSA, is within 8 C above0. Marks without fill indicate that TSA is 8 C or higher
Fig. 7. Adjustable blade grill with different blade angles (Case B.30-36, Case B.37-43 and Case B.44-50, and ADPI results from a previous study [3]), a), c) and e): T0.25/L vs. E (left side y-axis) and ADPI (right side y-axis), b), d) and f): T0.25/L vs. ƐT. *1; Solid marks: TSA < 8 C above0, Marks without fill: TSA > 8 C above 0, *2; T0.25/L is based on isothermal T0.25 from manufactures catalog.
524
H. Amai et al. / Building and Environment 125 (2017) 515e527
than0. With 0 Horizontal adjustments in Fig. 7a and b, the ranges of E and ƐT within recommended ranges are 0.65e0.98, and 0.71e0.95 respectively. E is significantly decreased when T0.25/L is small. Even within recommended range, E and ƐT are approximately 0.6 and 0.7 respectively. The thermal stratification for such conditions is quite high since TSA is 8 C or higher than 0. With 45 Downward adjustments in Fig. 7c and d, the ranges of E and ƐT within recommended ranges are 0.92e0.97 and 0.81e0.90 respectively. ADPI changes reversely with E and ƐT when T0.25/L is increased, since downward jets from diffuser with higher speed increases draft discomfort in the occupied zone while the jets improves mixing effect that improves ventilation effectiveness. With 45 Upward adjustments in Fig. 7e and f, the ranges of E and ƐT within recommended ranges are similar to 0 Horizontal adjustment: 0.68e0.97 and 0.70e0.90, respectively. E is significantly decreased when T0.25/ L is smaller than recommended range since the low momentum of diffuser jet has weak capacity to well mix the occupied air.
4.3. Impact of exhaust locations with nominal adjustment Fig. 8 shows the results of E and ƐT with different exhaust locations with the 0 Horizontal adjustments. Fig. 8a shows E, and Fig. 8b shows ƐT. The vertical dash lines display the recommended range with EX Cn. Results shows that similar to EX Cn, E and ƐT are almost equivalent to those for EX Cd. The location of EX Cw yield lower E and ƐT than EX Cn. The maximum E and ƐT are 0.7 because short circuit occurs when the exhaust is located at the opposite side of the diffuser. With near floor exhaust (EX Fw and EX Fd), E and ƐT are significantly higher than the ceiling mounted exhaust especially with low T0.25/L. With EX Fw, E and ƐT are greater than 1.0 when T0.25/L is lower than 1.7. E and ƐT slightly decrease as T0.25/L increase. As exhaust located opposite side of the diffuser, short circuit flow may increase once jet from the diffuser reaches to the opposite side of the wall. With Ex Fd, E and ƐT are higher than 1.0 and they are not sensitive to T0.25/L. The results suggest that near floor exhaust significantly improves ventilation effectiveness and air distribution performance compared to ceiling mounted exhaust. A higher supply airflow rate (higher T/L0.25) may not be helpful for effective air
distribution for the floor mounted exhaust. 5. Discussion This section discusses the results of experiments and the improvements of E and ƐT. The fist section summarizes the ranges of E and ƐT within recommended ranges for tested diffusers. The next section discusses the possible measures to improve E and ƐT while maintaining acceptable ADPI. 5.1. Range of E and ƐT within recommended range Table 3 provides the summary of E and ƐT within the recommended range. This table intends to update the design guideline in a previous study [20]. The data shown in the table are from the experiments with DT ¼ -5 C. The ranges of heating loads are calculated from DT and air flow rate of the exhaust air stream. 4 slots linear slot diffusers with a vertical flow is not covered by the table as the recommended range could be found from this study. E and ƐT are higher than 0.8 for 2 slots linear slot diffusers with vertical flow. With adjustable blade diffusers, both E and ƐT increase with the increase of T0.25/L. Results also show that E can be less than 0.8 with 0 Horizontal and 45 Upward conditions. E and ƐT are higher than 0.8 with 45 Downward, and recommended range is narrower with 45 Downward than with 45 Upward. 5.2. Improvement of E and ƐT This paragraph discusses improvements of E and ƐT regarding different supply diffuser adjustments, supply and exhaust air temperature differences, and exhaust locations. 5.2.1. Diffuser adjustments Fig. 9 compares vertical and horizontal flow adjustments of linear slot diffusers under DT ¼ -5 C. Results of E and ƐT with the horizontal flow adjustments are extracted from a previous study [20]. Fig. 9 also shows the recommended ranges with dash lines. The ranges indicate with air change rate to compare different T0.25/ L. With 2-slot diffusers in Fig. 2a and b, the recommended range is
Fig. 8. Adjustable blade grill (0 Horizontal adjustment) with different exhaust locations (Case B.30-36, Case B.51-57, Case B.58-63, Case B.64-71 and Case B.72-78), a): T0.25/L vs. E, b): T0.25/L vs. ƐT. *1; Solid marks: TSA < 8 C above0, Marks without fill: TSA > 8 C above 0, *2; T0.25/L is based on isothermal T0.25 from manufactures catalog.
H. Amai et al. / Building and Environment 125 (2017) 515e527
525
Table 3 The ranges of E and ƐT within recommended range of T0.25/L in terms of ADPI. Diffuser Type
Linear Slot Diffusers Adjustable Blade Grill
a
Vertical 2 slots 0 Horizontal 45 Downward 45 Upward
Recommended range of T0.25/L regarding ADPI
Range of E and ƐT within recommended T0.25/L regarding ADPI
T0.25/L
E
ƐT
Loads
Loads 2
[-]
[W/m ]
[-]
[-]
[W/m2]
1.4e2.1
12e31
0.87e0.95
0.84e0.95
10e27
1.1e2.2a 0.6e0.8a 0.6e1.0a
35-40a 35-40a 35-40a
0.65e0.98 0.92e0.97 0.68e0.97
0.71e0.95 0.81e0.90 0.70e0.90
9e36 9e33 9e35
Results extracted from previous studies [3].
narrower with vertical flow than with horizontal flow. E and ƐT increase about 25% and 30% at the maximum with the vertical flow, respectively. Adjusting air flow directions under heating mode may increase both E and ƐT. However, careful consideration must be given as direct air flow to occupied space may also decrease ADPI. Proper adjustment of diffusers and control of air flow rate under heating condition are necessary to improve E and ƐT. Fig. 10 compares E and ƐT of different angles adjustments for adjustable blade diffusers under DT ¼ -5 C. The dash lines display the recommended ranges with air flow rate. The recommended range is narrower with 45 Downward than with 0 Horizontal and 45 Upward. 45 Downward is able to increase E and ƐT about 30%
and 15% at the maximum, compared to the 0 Horizontal condition, respectively. Similar to the adjustment of the linear slot diffuser, downward blades can direct air flow to the occupied space to improve E and ƐT under heating conditions. However, air flow rate needs to be properly controlled to avoid significant decrease in ADPI. 5.2.2. Room supply air temperature difference (DT) Fig. 11 compares DT ¼ -2 C and DT ¼ -5 C under the same heating load for linear slot diffusers with vertical flow. The figure examines how DT effects ventilation performance. Lower DT may significantly increase E and ƐT. Compared to the Cases with DT ¼ -
Fig. 9. Linear slot diffusers with vertical and horizontal flow (Case B.1-8, Case B.9-17 for vertical flow and results from a previous study [20] for horizontal flow), a) Air change rate vs. E, b): Air change rate vs. ƐT. *1; Solid marks: TSA < 8 C above0, Marks without fill: TSA > 8 C above 0.
526
H. Amai et al. / Building and Environment 125 (2017) 515e527
Fig. 10. Adjustable blade grills with different deflector adjustments (Case B.30-36, Case B.37-43 and Case B.44-50), a): Air change rate vs. E, b) Air change rate vs. ƐT. *1; Solid marks: TSA < 8 C above0, Marks without fill: TSA > 8 C above 0.
5 C, Cases with DT ¼ -2 C increase E about 75% on average (100% at the maximum) and also increase ƐT about 45% on average (65% at the maximum), respectively. This analysis provides supportive data for HVAC designers to determine optimal design DT. A lower DT may require less fresh air (ventilation rate) to satisfy required ventilation rate in occupied zone and less heating energy as it improves both E and ƐT. On the other hand, lower DT requires higher air flow rate to remove the same heating load which requires more fan energy. For example, 2.5 times more air is required with DT ¼ -5 C to remove same heating load than with DT ¼ -2 C if distributed air is perfectly mixed in the space. The improvement of E and ƐT with lower DT need to consider this trade-off before designers decide room supply air difference. 5.2.3. Return air inlet locations As shown in Fig. 8 proper exhaust location may significantly increase E and ƐT. Locating exhaust near floor significantly increases E and ƐT: about 70% with both EX Fw and EX Fd at the maximum, compared to EX Cn. However, locating exhaust near floor may
require more duct work or room space for HVAC system as typical HVAC system installed in ceiling plenum space. In addition, the results show that improper exhaust location may decrease E and ƐT. With Ex Cw, E and ƐT decrease about 30% at the maximum compared to EX Cn location. A designer also needs to consider proper exhaust locations, especially when the high side wall supply strategy is utilized. 6. Conclusions This study conducted experimental measurements of ADPI, E and ƐT in a full-scale test room. Combined with previous studies [3] [13], [15] [20], the results provide supportive data for optimal diffuser selections in mixing ventilation, with an emphasis on the improvement of E and ƐT in heating applications. Proper applications of each tested strategy: diffuser adjustment, lower DT and exhaust location may significantly improve E and ƐT. This study shows that the proper adjustment of diffuser, lower DT and different exhaust location may improve E about 25e30% at the
Fig. 11. Linear slot diffusers with vertical flow under same heating load (DT ¼ -5 C and DT ¼ -2 C). a): E, b): ƐT.
H. Amai et al. / Building and Environment 125 (2017) 515e527
maximum, 75% in average and 70% at the maximum, respectively. However, designer also have to consider other aspects of those strategies such as narrower range of recommended T0.25/L with vertical flow and trade off of fan power with lower DT.
[10]
Nomenclature
[11]
ADPI EDT T0.25 L E Ei Ai C0 C(t)
tn
ƐT TSA Ti0 DT
Air Diffusion Performance Index Effective Draft Temperature Terminal Velocity Value of 0.25 m/s Characteristic Length Air Change Effectiveness, Arithmetic Mean of Ei in Occupied Zone Local Air Change Effectiveness Age of Air at a Location i Initial Concentration of Tracer Gas at the time t ¼ 0 Decayed Concentration Measured at Time t Nominal Time Constant Temperature Effectiveness Supply Air Temperature Temperature at a Location i Average Temperature in Occupied Space Exhaust Air and Supply Air Difference, TEA-TSA
Acknowledgement The authors wish to thank the American Society of Heating, Ventilation, and Air Conditioning Engineers (ASHRAE RP-1546) and Nailor and Price for their generous donation of diffusers. The author also wishes to thank Obayashi Corporation to financially support author's study at the University of Texas at Austin. References [1] G. Cao, H. Awbi, R. Yao, Y. Fan, K. Siren, R. Kosonen, J. Zhang, A review of the performance of different ventilation and airflow distribution systems in buildings, Build. Environ. 73 (2013) 171e186. [2] D. Muller, C. Kandzia, R. Kosonen, A.K. Melikov, P.V. Nielsen, Mixing Ventilation - Guidebook on Mixing Air Distribution Design 19, REHVA guidebook, 2013. [3] S. Liu, A. Novoselac, Expansion and Updating of the Air Diffusion Performance Index Method (RP 1546), ASHRAE Research Project Report, TC 5.3, Indoor Air Distribution, 2016. [4] W.J. Fisk, D. Faulkner, D. Sullivan, F. Bauman, Air change effectiveness and pollutant removal efficiency during adverse mixing conditions, Indoor Air 7 (1997) 55e63. [5] F.J. Offermann, D. Int-Hout, Ventilation effectiveness measurements of three supply/return air configurations, Environ. Int. 15 (6) (1989) 585e592. [6] M. Krajcik, A. Simone, B.W. Olesen, Air distribution and ventilation effectiveness in an occupied room heated by warm air, Energy Build. 55 (2012) 94e101. [7] R. Tomasi, M. Krajcik, A. Simone, B.W. Olesen, Experimental evaluation of air distribution in mechanically ventilated residential rooms: thermal comfort and ventilation effectiveness, Energy Build. 60 (2013) 28e37. [8] A. Novoselac, J. Srebric, Comparison of air exchange efficiency and
[9]
[12] [13] [14]
[15]
[16] [17] [18]
[19]
[20] [21] [22] [23]
[24]
[25]
[26]
[27] [28]
[29] [30] [31] [32]
[33]
527
contaminant removal effectiveness as IAQ indices, ASHRAE Trans. 109 (2) (2003) 339e349. Chung, K. C and Hsu, S. P, Effect of ventilation pattern on room air and contaminant distribution, Build. Environ., 36(9), pp. 989e998. ASHRAE, Handbook of Fundamentals Chapter 16. Ventilation and Infiltration, American Society of Heating, Refrigerating and Air- Conditioning Engineers, 2009. ASHRAE, Handbook of Fundamentals Chapter 57. Room Air Distribution. American Society of Heating, Refrigerating and Air- Conditioning Engineers, 2009. P.L. Miller, R.T. Nash, A further analysis of room air distribution performance, ASHRAE Trans. 77 (2) (1971) 205. S. Liu, A. Novoselac, Air diffusion performance index (ADPI) of diffusers for heating mode, Build. Environ. 87 (2015) 215e223. S. Liu, J. Clark, A. Novoselac, Air diffusion performance index (ADPI) of overhead-air-distribution at low cooling loads, Energy Build. 134 (2017) 271e284. S. Liu, A. Novoselac, The effect of deflectors on air diffusion performance index of adjustable diffusers: cooling condition (RP1546), Sci. Technol. Built Environ. 22 (2016). S.L. Sinha, R.C. Arora, S. Roy, Numerical simulation of two dimensional room air flow with and without buoyancy, Energy Build. 32 (2000) 121e129. E. Lee, J.A. Khan, C.E. Feigley, M.R. Ahmed, J.R. Hussey, An investigation of air inlet types in mixing ventilation, Build. Environ. 42 (3) (2007) 1089e1098. J.A. Khan, C.E. Feigley, E. Lee, M.R. Ahmed, S. Tamanna, Effects of inlet and exhaust locations and emitted gas density on indoor air contaminant concentrations, Build. Environ. 41 (7) (2006) 851e863. ASHRAE Standard 62.1-2010, Ventilation for Acceptable Indoor Air Quality. American Society of Heating, Refrigerating and Air- Conditioning Engineers, Inc, 2010. H. Amai, A. Novoselac, Experimental study on air change effectiveness in mixing ventilation, Build. Environ. 109 (2016) 101e111. ASHRAE, Handbook of Fundamentals Chapter 20. Space Air Diffusion, American Society of Heating, Refrigerating and Air- Conditioning Engineers, 2013. J. Rydberg, P. Norback, Air distribution and draft, ASHRAE Trans. 55 255 (1949). H.E. Straub, M.M. Chen, Distribution of Air within a Room for Year-round Air ConditioningdPart II. University of Illinois Engineering Experiment Station Bulletin, 442. H.E. Straub, F.S. Gilman, S. Konzo, Distribution of Air within a Room for Yearround Air Conditioning -Part I 435, University of Illinois Engineering Experiment Station Bulletin, 1956. ANSI/ASHRAE Standard 70e2006 Method of testing the performance of air outlets and air inlets American Society of Heating, Refrigerating and air conditioning engineers. ANSI/ASHRAE Standard 129-1997, Measuring Air-change Effectiveness. American Society of Heating, Refrigerating and Air- Conditioning Engineers, 2002. I. Mundt, M.H. Mathisen, V.P. Nielsen, A. Moser, REHVA Guidebook No. 2. Ventilation Effectiveness, REHVA, 2004. SHASE Standard 115-2010, Field Measurement Methods for Ventilation Effectiveness in Rooms, Society of Heating, Air-conditioning and Sanitary Engineers of Japan, 2010. S. Cui, M. Cohen, P. Stabat, D. Marchio, CO2 tracer gas concentration decay method for measuring air change rate, Build. Environ. 84 (2015) 162e169. D. Etheridge, M. Sandberg, Building Ventilation, Theory and Measurement, Wiley, New York, 1996, pp. 471e473. H.B. Awbi, G. Gan, Evaluation of the overall performance of room air distribution, Proc. indoor air 3 (1993) 238. S. Liu, A. Novoselac, Lagrangian particle modeling in the indoor environment: a comparison of RANS and LES turbulence methods, HVAC&R Res. 20 (4) (2014) 480e495. X. Cao, J. Liu, N. Jiang, Q. Chen, Particle image velocimetry measurement of indoor airflow field: a review of the technologies and applications, Energy Build. 69 (2014) 367e380.
Contents lists available at ScienceDirect
Building and Environment journal homepage: www.elsevier.com/locate/buildenv
Experimental study on air change effectiveness: Improving air distribution with all-air heating systems Hideyuki Amai a, b, *, Shichao Liu b, c, Atila Novoselac b a
Obayashi Corporation, USA Department of Civil, Architectural and Environmental Engineering, The University of Texas at Austin, Austin, TX, USA c Department of Civil and Environmental Engineering, Worcester Polytechnic Institute, USA b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 10 August 2017 Received in revised form 12 September 2017 Accepted 12 September 2017 Available online 18 September 2017
A major challenge of all-air heating applications is poor air distribution, which is often associated with a high temperature stratification. The two metrics that are commonly used for design and assessment of supply air distribution in the space are: air distribution performance index (ADPI) and the Air Change Effectiveness (E), respectively. All-air heating systems often produce stagnant air in the occupied part of the room. In this case, E may be very low while relatively uniform temperature in this occupied zone results in acceptable ADPI. Since ventilation design is based on ADPI, many all-air heating systems often produce very low E. This experiment based study identifies situation with very low E and provides simple strategies to improve it. The study provides additional design criteria to the ADPI diffuser selection guide that helps with optimal diffuser selection and adjustments. The results show that additional design criteria significantly improve E as well as temperature distribution, measured by temperature effectiveness (ƐT), with all-air heating systems. Appropriate adjustment of the diffuser may improve E and ƐT up to 30%, while the lower supply-room air temperature difference may increase E and ƐT in average 75% and 45%, respectively. Also, proper return air inlet location significantly improves E and ƐT for all-air heating. However, there are certain trades off as: the diffuser adjustment also may require seasonal adjustment for cooling and heating operation, lower supply air temperature difference requires more fan power, and floor exhaust placement may need more space for duct work. © 2017 Elsevier Ltd. All rights reserved.
Keywords: Ventilation effectiveness Mixing ventilation ADPI All-air heating Diffuser adjustments
1. Introduction Mixing ventilation is the most common type of air distribution, used much more than alternative air distribution strategies such as piston, displacement, stratum, and personalized ventilation [1]. Achieving the ideal mixing in the room depends on an air distribution design; it can apply various types of supply air diffusers and return air inlets (exhaust) to create a mixing ventilation in different space types [2]. The impact of supply air diffusers on room air temperature and velocity under cooling applications are well studied [3], while there is much less information on the impact on the fresh air dispersion in heating applications. Also, there is very little information on the impacts that location of exhaust has on the air distribution. A major challenge of mixing ventilation is in heating application (all-air heating) as it causes poor air
* Corresponding author. 55W 5TH AVE APT 01A, San Mateo, CA, USA. E-mail address: [email protected] (H. Amai). https://doi.org/10.1016/j.buildenv.2017.09.017 0360-1323/© 2017 Elsevier Ltd. All rights reserved.
distribution due to a high temperature stratification [4e8]. Fisk et al. [4] conducted experiments that used overhead all-air-heating system that supplied minimum air supply flow rate of typical VAV systems. The air change effectiveness was significantly lower than 1.0 in each experiment. The measured air change effectiveness was in the range of 0.69e0.91 with mean value of 0.81. Offermann et al. [5] measured ventilation effectiveness and ADPI under heating conditions. For the ceiling supply/return configuration, ventilation effectiveness was 0.73 when temperature difference of supply air temperature and room average temperature was 8 C. Krajcik et al. [6,7], measured air change efficiency and temperature effectiveness in a test chamber with various combinations of radiant floor heating and mixing ventilation. Theses all-air heating systems often produce stagnant air in the occupied space of the room with relatively uniform low temperature in this stagnant zone. However, with stagnant cold air in occupied zone, the fresh hot supply air short circuit in the upper part of the room causes very poor ventilation effectiveness in the occupied space. In any application, it is important to consider both thermal
516
H. Amai et al. / Building and Environment 125 (2017) 515e527
comfort of occupants and effective distribution of supply air to the occupants not only for cooling but also for heating. An effective air distribution improves indoor air quality by providing fresh air in all areas of the occupied zone and thermal comfort by providing uniform temperature distribution and acceptable velocities [9]. The widely used design method for indoor air distribution is Air Diffusion Performance Index (ADPI) [11]. The index, percentage of the whole occupied zone with acceptable air velocity and temperature ranges, quantifies the performance of supply air diffusers to create a uniform thermal environment in the occupied zone. ADPI provides engineers and practitioners a simple tool to design air distribution that is usually associated with complex heat transfer and fluid dynamics [12]. The concept of ADPI first introduced and developed the correlations between diffuser properties and indoor air distribution performance for five types of terminal diffusers at cooling conditions. Recent studies have updated the ADPI metric to include 15 diffuser types [3,13e15]. Also, these studies removed the deficiency of ADPI method related to applicability to only cooling systems; the new ADPI guideline is valid for both heating and cooling conditions. Furthermore, the studies identified the conditions where poor ventilation effectiveness appears in all air heating. Specifically, when warm supply air is delivered in the room with a weak jet momentum, the supply air short circuit in the ceiling level generating stagnant air in the occupied zone. Locations of supply air diffuser and exhaust impact air distribution and their ventilation effectiveness in the space have been investigated by many researchers [16e18]. Sinha et al. [16] compared impact of different inlet and outlet locations using models and computational fluid dynamics. The study found that the most effective combination of inlet and outlet positioning is with inlet near the floor and exhaust near the ceiling. When considering position of air supplies in the upper part of the room, Lee et al. [17] experimentally compared high wall jet from grill diffusers with typical ceiling diffusers. Their results show that the air inlet position and type are important determinants in the distribution of airborne contaminant concentrations. Overall, the ceiling diffuser produced more efficient ventilation than the wall jet air inlet. In Khan's study [18], the arrangements of wall inlet and outlet greatly influence contaminant concentration. However, the influence of the outlet location is minimal with ceiling diffuser inlet. Since the air near exhaust is not driven by jet momentum but by negative pressure in the air, the velocities near exhausts are relatively small. Therefore, the effects of the exhaust location on room airflow pattern are relatively small in most applications. However, the exhaust location influences Air Change Effectiveness and Contaminant Removal Effectiveness [2]. Location of exhaust is even more important when mixing ventilation provides heating. ASHRAE standard 62.1 [19] assumes Air Change Effectiveness (E) of 0.8 when a 0.8 m/s jet does not reach to the lower part of the room or when the supply-room temperature difference is larger than 8 C. In the literature, Ventilation Effectiveness is defined as description of air distribution system's ability to remove internally generated pollutants from a building zone or space, while the E is an air distribution system's ability to deliver ventilation air to occupied zone or space [10]. In the following section of the paper, we will use E as the index that describes quality of the air distribution and the ventilation effectiveness. Our previous study [20] conducted extensive experiments on E in mixing ventilation. These data combined with the results from the latest ADPI study [3] and provided comprehensive data set on diffuser performance considering both uniformity of the temperature field and range of E in the both cooling and heating applications. E and temperature effectiveness (ƐT) were slightly higher than 1.0 in cooling applications. However in heating application, E
was significantly decreased at small T0.25/L even though ADPI was in acceptable range (ADPI higher than 80%); this poor E due to the short circuit of the supply air. E within the acceptable ADPI range of T0.25/L (recommended range) was 0.56e0.87. This short circuiting of hot air puts performance of many diffusers in the range that is lower than specified value in ASHRAE standard 62.1, E > 0.8. To overcome the challenges of poor ventilation under heating conditions, this study examines simple strategies that may improve E and ƐT under heating conditions, while maintaining acceptable ADPI; specifically, it evaluates the impacts of diffuser deflector adjustment with linear slot diffusers and adjustable blades grills, room-supply air temperature difference with vertical flow of linear slot diffusers, and exhaust locations with adjustable blade grill. In addition, it provides new diffuser selection data in the form of the ADPI for diffusers with a vertical jet projection (vertical flow) categorized as Group E in ASHRAE Handbook [21]. The following section briefly summarizes indices utilized to evaluate the performance of the mixing ventilation systems used in this study. It is followed by the methodology section that describes experimental setup and matrix of experiments. The results section compares different indices to evaluate impacts of analysed ventilation performance strategies, while the conclusion section synthesizes major findings. 2. Theoretical background and description of used indices This section explains three indices: Air Diffusion Performance Index (ADPI), Air change effectiveness (E) and Temperature effectiveness (ƐT), that are most commonly utilized to evaluate air distribution performance. It provides short overview of variables and physical phenomena and explains the basics for interpretation of the tables and graphs in the results section. 2.1. ADPI ADPI is defined as the percentage of the occupied zone that maintains acceptable velocity and temperature. The region of acceptable velocity and temperature is determined by local Effective Draft Temperature (EDT) that combines air temperature difference and air speed [22e24]. EDT for cooling condition is defined as,
EDTðqÞ ¼ Ti Ta 8:0ðVi 0:15Þ½ C
(1)
where Ti is temperature at the test point, i; Ta is spacious average temperature ( C) and Vi is local air speed (m/s). EDT for heating condition is defined as [3,13].
EDTðqÞ ¼ Ti Ta 9:1ðVi 0:15Þ½ C
(2)
ADPI incorporates the throw and the characteristic length, and it provides design variables for selecting diffusers. The dominant diffuser property for air distribution is the supply jet throw length at which the jet velocity decreases to a selected terminal value of 0.25 m/s (T0.25) [25]. The characteristic length (L) describes the room geometry in the form of the distance at which jet travels. The ratio T0.25/L is a dimensionless number that characterizes a supply diffuser momentum (including capacity of a diffuser to mix/entrain surrounding air) for given flow rate. T0.25/L has the largest impact on ADPI, and together T0.25/L and ADPI are used in the diffuser selection guideline provides by ASHRAE Handbook [11]. The guideline lists the relationships between ADPI and a dimensionless ratio of T0.25/L for various diffuser types at different thermal loads. One is able to design an HVAC terminal system, such as diffuser selection and layout, by ensuring T0.25/L of the system renders ADPI greater
H. Amai et al. / Building and Environment 125 (2017) 515e527
than 80% which is the threshold of air distribution design. Since T0.25/L is a critical parameter determining air distribution [20], we will utilize it to assess the performance of air distribution throughout this study. T0.25 is defined by different air flow rate in isothermal conditions from manufacture's catalog data for several diffuser types including adjustable blade diffusers [11]. However, T0.25/L for a vertical flow has not defined in the literature. Throw of warm vertical jet may significantly travel less than isothermal jet by strongly influence of buoyancy force. In this study, manufacture's catalog T0.25 for vertical flow (isothermal flow) is modified by measured supply air temperature and room air temperature, utilizing empirical chart in ASHRAE Handbook [21]. From this chart, the throw correction factor for vertical flow regarding room supply air temperature differences is determined. In addition, ceiling height is defined L for vertical flow. 2.2. Air change effectiveness (E) The Air Change Effectiveness (E) as defined in ASHRAE standard 129 [26] is often utilized as ventilation effectiveness index. It defines how well the supply air is distributed through the occupied space of the room. There are numerous ways to experimentally determine ventilation effectiveness [27e29]. However, this research utilizes the age of air approach to calculate E. The age of air is determined from the tracer gas decay test. Age of air at a location i in the space is defined as
1 Ai ¼ C0
Z∞ Ci ðtÞdt
(3)
0
where C0 is the initial concentration of tracer gas at the time t ¼ 0 and C(t) is the decayed concentration measured at time t. The local air change effectiveness, Ei, is defined as
Ei ¼
tn Ai
½
(4)
In the equation above tn is the nominal time constant, which is reciprocal value of the air change rate in the room (tn ¼ Room Volume/Volume flow rate). The arithmetic mean of Ei in the occupied zone of the room is defined as E. For perfect mixing, E is 1 and is smaller than 1 in a case where the occupied zone has a stagnant air. It can be larger than 1 where the supply air reach occupied zone more effectively than with perfect mixing, i.e. with a displacement ventilation [8]. 2.3. Temperature effectiveness (ƐT) Temperature effectiveness [30], known as ventilation effectiveness for heat removal [31], evaluates the temperature gradient in the test space. The temperature effectiveness (ƐT) is defined as
εT ¼
TSA TEA 100½% TSA 〈T〉O
517
(5)
where TSA is the supply air temperature, TEA is the exhaust air temperature and
Fig. 1. Diffusers used in this study: a) Linear slot diffuser with vertical flow, b) Adjustable blade grill with 0 Horizontal, 45 Upward, and 45 Downward adjustment.
518
H. Amai et al. / Building and Environment 125 (2017) 515e527
Fig. 2. Experimental setup of a test room for ADPI measurements. (Chamber geometry, temperature and velocity measurement locations).
at the Center for Energy and Environmental Resources at the University of Texas at Austin, with the size of 5.5 m 4.5 m 2.7 m. The chamber is equipped with a precise environmental control system. Detailed technical specification for this state-of-the-art indoor environment related research facility can be found in our previous publication [32]. Fig. 2 illustrates experimental setups of the test room for ADPI measurements. ADPI measurements were conducted under the same chamber and setups as our previous studies [3] [13], [14]. Heating load was simulated by adjusting the temperature of the cooled wall connected to a dedicated chiller, which together mimicked exterior wall or window in winter condition. Although particle image velocimetry (PIV) measurement [33] can be used to better visualizing the airflow fields, velocity and temperature at sixty locations in the occupied zone at four different heights above the floor (0.1 m, 0.6 m, 1.1 m, and 1.7 m) were measured with twelve hot-sphere anemometers (HT-400, SENSOR, Poland, accuracy: ±0.03 m/s þ -3%, temperature: ±0.2 C). The measurements were repeated five times for each experiment to obtain sixty locations
with the twelve available sensors. At the same time, the vertical temperature (0.1 m, 0.6 m, 1.1 m, 1.4 m, 1.8 m, 2.2 m) was measured at five different locations by thermistors (Model 44033, OMEGA, Accuracy: ±0.1 C). Furthermore, supply and exhaust air temperatures were monitored during the experiments to ensure the stability of the chamber conditions. The uncertainty of ADPI measurement depends on the accuracy of the instruments used and several factors relate to experimental setup; the detailed procedure for calculation of this uncertainty is described in our previous study [13]. Due to the high accuracy of velocity and temperature sensors in the studies, the uncertainty is ±2.7% of absolute value of ADPI. 3.2.1. Experimental matrix for ADPI measurements Table 1 shows the experimental conditions for ADPI measurements. 2 slots and 4 slots linear slot diffusers with vertical flow were tested. Each setup was consisted of several cases with various T0.25/L (related to air change rate). For instance, Cases (A.1-6) had the same experimental setup but different T0.25/L such as 0.5 in Case
Table 1 Experimental conditions for ADPI measurements. Case#
Supply Diffusers Type
A.1-6
Linear slots) A.7-12 Linear slots) A.13- Linear 18 slots) A.19- Linear 24 slots)
Adjustment Slot Diffuser (2 Slot Diffuser (4 Slot Diffuser (2 Slot Diffuser (4
Vertical
T0.25/L [-] (Air change rate [h1])
Temperature difference between supply and exhaust DT ¼ TEA-TSA [ C]
Exhaust Location
0.5 (2.1), (8.6) 0.4 (2.1), (8.6) 0.6 (2.1), (8.6) 0.5 (2.1), (8.6)
0.8 (3.3), 1.1 (4.5), 1.4 (5.8), 1.7 (7.2), 2.0
5
EX Cn
0.7 (3.3), 0.9 (4.5), 1.2 (5.8), 1.4 (7.2), 1.8
5
1.0 (3.3), 1.3 (4.5), 1.6 (5.8), 1.9 (7.2), 2.2
2
0.8(3.3), 1.1 (4.5), 1.4 (5.8), 1.7 (7.2), 1.9
2
H. Amai et al. / Building and Environment 125 (2017) 515e527
A.1 and 0.8 in Case A.2. The T0.25 derived as described in theoretical background and description of used indices section. All experiments were conducted with heating conditions. 3.3. E and ƐT measurements Fig. 3 illustrates experimental setup for E and ƐT measurements. The experimental setup allowed adjustable blade diffusers with high side wall position, and 2 slots and 4 slots linear slot diffusers with vertical flow. Dimensions of the plenum box for adjustable blade diffuser is also described in Fig. 3. Furthermore, Fig. 3 shows the five specific exhaust locations to evaluate the impact of the exhaust locations. There were three locations for ceiling mounted positions: EX Cd, EX Cn, and EX Cw, and the two locations near floor: EX Fd and EX Fw. Tracer gas decay test using CO2 as tracer gas was conducted to measure Ei and calculate E as described in previous section. The decay of CO2 concentration was measured by in-situ sensors
519
(TELAIR model 7001, Accuracy: ± 50 ppm or 5%) at 18 locations simultaneously (Fig. 3 provide the exact sensor locations). Tracer gas injection was ensured near perfect mixing throughout the space by 3 powerful mixing fans (Lasko Box Fan, Model #B20200, 55 cm 11 cm 57 cm) before the tracer gas decay started. In addition, vertical temperature distributions, supply and exhaust temperature were measured. Average temperature at 0.1 m, 0.6 m, 1.1 m, 1.4 m, and 1.8 m (T0.1, T0.6, T1.1, T1.4 and T1.8) above the floor utilized to determine ƐT. The uncertainty of measurements of air change effectiveness has been already discussed in our previous study [20]. From repetitive experiments, the uncertainty of local Ei was 6% on average and maximum 14% and E was 6% on average and maximum 11%. 3.3.1. Experimental matrix for E and ƐT measurements Table 2 shows the experimental conditions for E and ƐT measurements. The experiments were conducted with 2 slots and 4 slots linear slot diffusers with vertical flow, and adjustable blade
Fig. 3. Experimental setup of a test room for E and ƐT measurements (Chamber, adjustable blade grille box and exhaust box geometry, temperature and CO2 concentration measurement locations).
520
H. Amai et al. / Building and Environment 125 (2017) 515e527
Table 2 Experimental conditions for E and ƐT measurements. Case# Supply Diffusers Type B.1-8 Linear Slot Diffuser (2 slots) B.9- Linear Slot Diffuser 17 (4 slots) B.18- Linear Slot Diffuser 23 (2 slots) B.24- Linear Slot Diffuser 29 (4 slots) B.30- Adjustable Blade 36 Grill B.37- Adjustable Blade 43 Grill B.44- Adjustable Blade 50 Grill B.51- Adjustable Blade 57 Grill B.58- Adjustable Blade 63 Grill B.64- Adjustable Blade 71 Grill B.72- Adjustable Blade 78 Grill
Adjustment Vertical
T0.25/L [-] (Air change rate [h1])
EX Cn
0.7 (3.3), 0.9 (4.4), 0.9 (4.5), 1.2 (5.8), 1.4 5 (8.6) 1.3 (4.5), 1.6 (5.8), 1.9 (7.2), 2.2 (8.6) 2
EX Cn EX Cn
2
EX Cn
0.9(2.1), 1.2 (3.3), 1.4 (4.5), 1.6 (5.8), 1.9 (7.2), 2.1 5
EX Cn
0.4(2.1), 0.5 (3.3), 0.7 (4.5), 0.8 (5.8), 0.8 (7.2), 0.9 5
EX Cn
0.4(2.1), 0.5 (3.3), 0.7 (4.5), 0.8 (5.8), 0.8 (7.2), 0.9 5
EX Cn
0.9(2.1), 1.2 (3.3), 1.4 (4.5), 1.6 (5.8), 1.9 (7.2), 2.1 5
EX Cw
0.9(2.1), 1.2 (3.3), 1.4 (4.5), 1.6 (5.8), 1.9 (7.2), 2.1 5
EX Cd
0.9(2.1), 1.2 (3.3), 1.4 (4.5), 1.6 (5.8), 1.9 (7.2), 2.1 5
EX Fw
0.9(2.1), 1.2 (3.3), 1.4 (4.5), 1.6 (5.8), 1.9 (7.2), 2.1 5
EX Fd
Vertical Vertical
0.5 (2.1), 0.8(3.3), 1.1 (4.5), 1.4 (5.8), 1.7 (7.2), 2.0 (8.6)
0 Horizontal 0.8 (1.6), (8.6) 45 Upward 0.3 (1.6), (8.6) 0.3 (1.6), 45 Downward (8.6) 0 Horizontal 0.8 (1.6), (8.6) 0 Horizontal 0.8 (1.6), (8.6) 0 Horizontal 0.8 (1.6), (8.6) 0 Horizontal 0.8 (1.6), (8.6)
Exhaust Location
0.8 (3.3), 1.1 (4.4), 1.5 (5.8), 1.7 (6.9), 1.7 5
0.5 (2.1), (7.2), 2.0 0.4 (2.1), (6.9), 1.5 0.6 (2.1),
Vertical
Temperature difference between supply and exhaust (DT ¼ TEA-TSA) [ C]
0.7 (3.0), (8.6) 0.6 (3.0), (7.2), 1.8 1.0 (3.3),
Fig. 4. Linear slot diffusers with vertical flow under DT ¼ -5 C (Case A.1-6 Case A.7-12, Case B.1-8 and Case B.9-17), a) and c): T0.25/L vs. E (left side y-axis) and ADPI (right side yaxis), b) and d): T0.25/L vs. ƐT. *1; Solid mark: TSA < 8 C above
H. Amai et al. / Building and Environment 125 (2017) 515e527
diffusers with three different adjustments under heating conditions. In addition, five different exhaust locations with nominal settings of adjustable blade diffusers were tested. 4. Results This section is divided into two parts, presenting results from two series of experiments: 1) experiments with linear slot diffusers with vertical flow, and 2) experiments with adjustable blade diffusers with different deflector angles and exhaust locations. 4.1. Vertical flow with linear slot diffusers The first part examines the range of T0.25/L that can achieve ADPI higher than 80% and according E and ƐT. Then we discusse the results of ADPI, E and ƐT with lower temperature difference between supply and exhaust (DT). Finally, the section examines some results of velocity and temperature measurements from ADPI experiments. 4.1.1. ADPI, E and ƐT Fig. 4 shows the results of ADPI, E and ƐT with DT ¼ -5 C. Fig. 4 displays the results of 2 slots diffusers (Fig. 4a and b) slots diffusers (Fig. 4c and d). The solid marks indicate that supply air
521
temperatures (TSA) is less than 8 C above average occupied space temperature (
Fig. 5. Linear slot diffusers with vertical flow under DT ¼ -2±2 C (Case A.13-18, Case A.24-29, Case B.18-23 and Case B.24-29), a) and c): T0.25/L vs. E (left side y-axis) and ADPI (right side y-axis), b) and d): T0.25/L vs. ƐT. *1; Solid marks: TSA < 8 C above
522
H. Amai et al. / Building and Environment 125 (2017) 515e527
DT ¼ -2. ADPI is higher than 80% for both 2 slots and 4 shots diffusers within the tested ranges of T0.25/L, implying that small DT enhances ADPI significantly. The highest E is greater with 2 slots diffusers than with 4 slots diffusers: E is 1.1 with 2 slots diffusers, 0.8 with 4 slots diffusers respectively. The highest ƐT is also greater with 2 slots diffusers than with 4 slots diffusers: 1.05 with 2 slots diffusers, 0.9 with 4 slots diffusers respectively.
4.1.2. Room air velocity and temperature Fig. 6 shows the results of velocity and temperature fields with ADPI measurements (Case A.1, A.3, A.5, and A.12). Temperature fields show with temperature differences between supply air and point of measurements (DTSA-i ¼ TSA-Ti). For Case A.1 (2 slots, T0.25/L: 0.5 (2.1 h1)), air velocity below 1.7 m is less than 0.25 m/s. Temperature stratification (maximum temperature difference within
Fig. 6. Velocity and temperature fields in ADPI measurements (Case A.1, A.3, A.5, and A.12). *Refer to Fig. 2 for location of measuring points (L,C,R and D1-D5).
H. Amai et al. / Building and Environment 125 (2017) 515e527
occupied space) is less than 2 C. For Case A.3 (2 slots, T0.25/L: 1.1 (4.5 h1)), jet from the diffuser (velocity higher than 0.25 m/s) reaches to 1.7 m. Temperature stratification is 4 C between 0.1 m and 1.7 m. Because of the bouyance effect and weak jet from diffuser, jet from the diffuser could not reach to the bottom ended in higher thermal stratification. It is considered that thermal stratification in occupied zone that results in low ADPI. In Case A.5 (2 slots, T/L0.25: 1.7 (7.2 h1)), jet from diffuser reaches to 0.6 m and temperature stratification it the lowest among shown results. Finally, with Case A.12 (4 slots, T/L0.25: 1.8 (8.6 h1)), jet from diffuser reached to 1.1 m. Temperature stratification is high and difference is about 3.5 C. Although Case 12 has almost same T0.25/L as Case A.5 calculated by the method in theoretical background section, the results of velocity and temperature fields are different. It is considered that correction factor calculated does not accurately normalize throw of 2 slots and 4 slots diffusers since the emprical table in ASHRAE handbook [21] is with supply opening of aspect
523
ratio 1.0 and air flow along with the perimeter wall. It is also considered the different diffusers' width affect temperature fields. 4.2. Adjustable blade diffusers with high side wall supply This section examines results of adjustable blade diffusers. The first paragraph examines different blade adjustment. The results combined with ADPI measurements from previous study [3]. The next paragraph discusses the results of different exhaust locations with nominal adjustment. 4.2.1. ADPI, E and ƐT Fig. 7 shows the results of ADPI, E and ƐT measurements with different blade angles: 0 Horizontal, 45 Downward, and 45 . ADPI shown in Fig. 7a, c and 7e with right side of y-axis extracted from a previous study [3]. The solid marks indicate that TSA, is within 8 C above
Fig. 7. Adjustable blade grill with different blade angles (Case B.30-36, Case B.37-43 and Case B.44-50, and ADPI results from a previous study [3]), a), c) and e): T0.25/L vs. E (left side y-axis) and ADPI (right side y-axis), b), d) and f): T0.25/L vs. ƐT. *1; Solid marks: TSA < 8 C above
524
H. Amai et al. / Building and Environment 125 (2017) 515e527
than
4.3. Impact of exhaust locations with nominal adjustment Fig. 8 shows the results of E and ƐT with different exhaust locations with the 0 Horizontal adjustments. Fig. 8a shows E, and Fig. 8b shows ƐT. The vertical dash lines display the recommended range with EX Cn. Results shows that similar to EX Cn, E and ƐT are almost equivalent to those for EX Cd. The location of EX Cw yield lower E and ƐT than EX Cn. The maximum E and ƐT are 0.7 because short circuit occurs when the exhaust is located at the opposite side of the diffuser. With near floor exhaust (EX Fw and EX Fd), E and ƐT are significantly higher than the ceiling mounted exhaust especially with low T0.25/L. With EX Fw, E and ƐT are greater than 1.0 when T0.25/L is lower than 1.7. E and ƐT slightly decrease as T0.25/L increase. As exhaust located opposite side of the diffuser, short circuit flow may increase once jet from the diffuser reaches to the opposite side of the wall. With Ex Fd, E and ƐT are higher than 1.0 and they are not sensitive to T0.25/L. The results suggest that near floor exhaust significantly improves ventilation effectiveness and air distribution performance compared to ceiling mounted exhaust. A higher supply airflow rate (higher T/L0.25) may not be helpful for effective air
distribution for the floor mounted exhaust. 5. Discussion This section discusses the results of experiments and the improvements of E and ƐT. The fist section summarizes the ranges of E and ƐT within recommended ranges for tested diffusers. The next section discusses the possible measures to improve E and ƐT while maintaining acceptable ADPI. 5.1. Range of E and ƐT within recommended range Table 3 provides the summary of E and ƐT within the recommended range. This table intends to update the design guideline in a previous study [20]. The data shown in the table are from the experiments with DT ¼ -5 C. The ranges of heating loads are calculated from DT and air flow rate of the exhaust air stream. 4 slots linear slot diffusers with a vertical flow is not covered by the table as the recommended range could be found from this study. E and ƐT are higher than 0.8 for 2 slots linear slot diffusers with vertical flow. With adjustable blade diffusers, both E and ƐT increase with the increase of T0.25/L. Results also show that E can be less than 0.8 with 0 Horizontal and 45 Upward conditions. E and ƐT are higher than 0.8 with 45 Downward, and recommended range is narrower with 45 Downward than with 45 Upward. 5.2. Improvement of E and ƐT This paragraph discusses improvements of E and ƐT regarding different supply diffuser adjustments, supply and exhaust air temperature differences, and exhaust locations. 5.2.1. Diffuser adjustments Fig. 9 compares vertical and horizontal flow adjustments of linear slot diffusers under DT ¼ -5 C. Results of E and ƐT with the horizontal flow adjustments are extracted from a previous study [20]. Fig. 9 also shows the recommended ranges with dash lines. The ranges indicate with air change rate to compare different T0.25/ L. With 2-slot diffusers in Fig. 2a and b, the recommended range is
Fig. 8. Adjustable blade grill (0 Horizontal adjustment) with different exhaust locations (Case B.30-36, Case B.51-57, Case B.58-63, Case B.64-71 and Case B.72-78), a): T0.25/L vs. E, b): T0.25/L vs. ƐT. *1; Solid marks: TSA < 8 C above
H. Amai et al. / Building and Environment 125 (2017) 515e527
525
Table 3 The ranges of E and ƐT within recommended range of T0.25/L in terms of ADPI. Diffuser Type
Linear Slot Diffusers Adjustable Blade Grill
a
Vertical 2 slots 0 Horizontal 45 Downward 45 Upward
Recommended range of T0.25/L regarding ADPI
Range of E and ƐT within recommended T0.25/L regarding ADPI
T0.25/L
E
ƐT
Loads
Loads 2
[-]
[W/m ]
[-]
[-]
[W/m2]
1.4e2.1
12e31
0.87e0.95
0.84e0.95
10e27
1.1e2.2a 0.6e0.8a 0.6e1.0a
35-40a 35-40a 35-40a
0.65e0.98 0.92e0.97 0.68e0.97
0.71e0.95 0.81e0.90 0.70e0.90
9e36 9e33 9e35
Results extracted from previous studies [3].
narrower with vertical flow than with horizontal flow. E and ƐT increase about 25% and 30% at the maximum with the vertical flow, respectively. Adjusting air flow directions under heating mode may increase both E and ƐT. However, careful consideration must be given as direct air flow to occupied space may also decrease ADPI. Proper adjustment of diffusers and control of air flow rate under heating condition are necessary to improve E and ƐT. Fig. 10 compares E and ƐT of different angles adjustments for adjustable blade diffusers under DT ¼ -5 C. The dash lines display the recommended ranges with air flow rate. The recommended range is narrower with 45 Downward than with 0 Horizontal and 45 Upward. 45 Downward is able to increase E and ƐT about 30%
and 15% at the maximum, compared to the 0 Horizontal condition, respectively. Similar to the adjustment of the linear slot diffuser, downward blades can direct air flow to the occupied space to improve E and ƐT under heating conditions. However, air flow rate needs to be properly controlled to avoid significant decrease in ADPI. 5.2.2. Room supply air temperature difference (DT) Fig. 11 compares DT ¼ -2 C and DT ¼ -5 C under the same heating load for linear slot diffusers with vertical flow. The figure examines how DT effects ventilation performance. Lower DT may significantly increase E and ƐT. Compared to the Cases with DT ¼ -
Fig. 9. Linear slot diffusers with vertical and horizontal flow (Case B.1-8, Case B.9-17 for vertical flow and results from a previous study [20] for horizontal flow), a) Air change rate vs. E, b): Air change rate vs. ƐT. *1; Solid marks: TSA < 8 C above
526
H. Amai et al. / Building and Environment 125 (2017) 515e527
Fig. 10. Adjustable blade grills with different deflector adjustments (Case B.30-36, Case B.37-43 and Case B.44-50), a): Air change rate vs. E, b) Air change rate vs. ƐT. *1; Solid marks: TSA < 8 C above
5 C, Cases with DT ¼ -2 C increase E about 75% on average (100% at the maximum) and also increase ƐT about 45% on average (65% at the maximum), respectively. This analysis provides supportive data for HVAC designers to determine optimal design DT. A lower DT may require less fresh air (ventilation rate) to satisfy required ventilation rate in occupied zone and less heating energy as it improves both E and ƐT. On the other hand, lower DT requires higher air flow rate to remove the same heating load which requires more fan energy. For example, 2.5 times more air is required with DT ¼ -5 C to remove same heating load than with DT ¼ -2 C if distributed air is perfectly mixed in the space. The improvement of E and ƐT with lower DT need to consider this trade-off before designers decide room supply air difference. 5.2.3. Return air inlet locations As shown in Fig. 8 proper exhaust location may significantly increase E and ƐT. Locating exhaust near floor significantly increases E and ƐT: about 70% with both EX Fw and EX Fd at the maximum, compared to EX Cn. However, locating exhaust near floor may
require more duct work or room space for HVAC system as typical HVAC system installed in ceiling plenum space. In addition, the results show that improper exhaust location may decrease E and ƐT. With Ex Cw, E and ƐT decrease about 30% at the maximum compared to EX Cn location. A designer also needs to consider proper exhaust locations, especially when the high side wall supply strategy is utilized. 6. Conclusions This study conducted experimental measurements of ADPI, E and ƐT in a full-scale test room. Combined with previous studies [3] [13], [15] [20], the results provide supportive data for optimal diffuser selections in mixing ventilation, with an emphasis on the improvement of E and ƐT in heating applications. Proper applications of each tested strategy: diffuser adjustment, lower DT and exhaust location may significantly improve E and ƐT. This study shows that the proper adjustment of diffuser, lower DT and different exhaust location may improve E about 25e30% at the
Fig. 11. Linear slot diffusers with vertical flow under same heating load (DT ¼ -5 C and DT ¼ -2 C). a): E, b): ƐT.
H. Amai et al. / Building and Environment 125 (2017) 515e527
maximum, 75% in average and 70% at the maximum, respectively. However, designer also have to consider other aspects of those strategies such as narrower range of recommended T0.25/L with vertical flow and trade off of fan power with lower DT.
[10]
Nomenclature
[11]
ADPI EDT T0.25 L E Ei Ai C0 C(t)
tn
ƐT TSA Ti
Air Diffusion Performance Index Effective Draft Temperature Terminal Velocity Value of 0.25 m/s Characteristic Length Air Change Effectiveness, Arithmetic Mean of Ei in Occupied Zone Local Air Change Effectiveness Age of Air at a Location i Initial Concentration of Tracer Gas at the time t ¼ 0 Decayed Concentration Measured at Time t Nominal Time Constant Temperature Effectiveness Supply Air Temperature Temperature at a Location i Average Temperature in Occupied Space Exhaust Air and Supply Air Difference, TEA-TSA
Acknowledgement The authors wish to thank the American Society of Heating, Ventilation, and Air Conditioning Engineers (ASHRAE RP-1546) and Nailor and Price for their generous donation of diffusers. The author also wishes to thank Obayashi Corporation to financially support author's study at the University of Texas at Austin. References [1] G. Cao, H. Awbi, R. Yao, Y. Fan, K. Siren, R. Kosonen, J. Zhang, A review of the performance of different ventilation and airflow distribution systems in buildings, Build. Environ. 73 (2013) 171e186. [2] D. Muller, C. Kandzia, R. Kosonen, A.K. Melikov, P.V. Nielsen, Mixing Ventilation - Guidebook on Mixing Air Distribution Design 19, REHVA guidebook, 2013. [3] S. Liu, A. Novoselac, Expansion and Updating of the Air Diffusion Performance Index Method (RP 1546), ASHRAE Research Project Report, TC 5.3, Indoor Air Distribution, 2016. [4] W.J. Fisk, D. Faulkner, D. Sullivan, F. Bauman, Air change effectiveness and pollutant removal efficiency during adverse mixing conditions, Indoor Air 7 (1997) 55e63. [5] F.J. Offermann, D. Int-Hout, Ventilation effectiveness measurements of three supply/return air configurations, Environ. Int. 15 (6) (1989) 585e592. [6] M. Krajcik, A. Simone, B.W. Olesen, Air distribution and ventilation effectiveness in an occupied room heated by warm air, Energy Build. 55 (2012) 94e101. [7] R. Tomasi, M. Krajcik, A. Simone, B.W. Olesen, Experimental evaluation of air distribution in mechanically ventilated residential rooms: thermal comfort and ventilation effectiveness, Energy Build. 60 (2013) 28e37. [8] A. Novoselac, J. Srebric, Comparison of air exchange efficiency and
[9]
[12] [13] [14]
[15]
[16] [17] [18]
[19]
[20] [21] [22] [23]
[24]
[25]
[26]
[27] [28]
[29] [30] [31] [32]
[33]
527
contaminant removal effectiveness as IAQ indices, ASHRAE Trans. 109 (2) (2003) 339e349. Chung, K. C and Hsu, S. P, Effect of ventilation pattern on room air and contaminant distribution, Build. Environ., 36(9), pp. 989e998. ASHRAE, Handbook of Fundamentals Chapter 16. Ventilation and Infiltration, American Society of Heating, Refrigerating and Air- Conditioning Engineers, 2009. ASHRAE, Handbook of Fundamentals Chapter 57. Room Air Distribution. American Society of Heating, Refrigerating and Air- Conditioning Engineers, 2009. P.L. Miller, R.T. Nash, A further analysis of room air distribution performance, ASHRAE Trans. 77 (2) (1971) 205. S. Liu, A. Novoselac, Air diffusion performance index (ADPI) of diffusers for heating mode, Build. Environ. 87 (2015) 215e223. S. Liu, J. Clark, A. Novoselac, Air diffusion performance index (ADPI) of overhead-air-distribution at low cooling loads, Energy Build. 134 (2017) 271e284. S. Liu, A. Novoselac, The effect of deflectors on air diffusion performance index of adjustable diffusers: cooling condition (RP1546), Sci. Technol. Built Environ. 22 (2016). S.L. Sinha, R.C. Arora, S. Roy, Numerical simulation of two dimensional room air flow with and without buoyancy, Energy Build. 32 (2000) 121e129. E. Lee, J.A. Khan, C.E. Feigley, M.R. Ahmed, J.R. Hussey, An investigation of air inlet types in mixing ventilation, Build. Environ. 42 (3) (2007) 1089e1098. J.A. Khan, C.E. Feigley, E. Lee, M.R. Ahmed, S. Tamanna, Effects of inlet and exhaust locations and emitted gas density on indoor air contaminant concentrations, Build. Environ. 41 (7) (2006) 851e863. ASHRAE Standard 62.1-2010, Ventilation for Acceptable Indoor Air Quality. American Society of Heating, Refrigerating and Air- Conditioning Engineers, Inc, 2010. H. Amai, A. Novoselac, Experimental study on air change effectiveness in mixing ventilation, Build. Environ. 109 (2016) 101e111. ASHRAE, Handbook of Fundamentals Chapter 20. Space Air Diffusion, American Society of Heating, Refrigerating and Air- Conditioning Engineers, 2013. J. Rydberg, P. Norback, Air distribution and draft, ASHRAE Trans. 55 255 (1949). H.E. Straub, M.M. Chen, Distribution of Air within a Room for Year-round Air ConditioningdPart II. University of Illinois Engineering Experiment Station Bulletin, 442. H.E. Straub, F.S. Gilman, S. Konzo, Distribution of Air within a Room for Yearround Air Conditioning -Part I 435, University of Illinois Engineering Experiment Station Bulletin, 1956. ANSI/ASHRAE Standard 70e2006 Method of testing the performance of air outlets and air inlets American Society of Heating, Refrigerating and air conditioning engineers. ANSI/ASHRAE Standard 129-1997, Measuring Air-change Effectiveness. American Society of Heating, Refrigerating and Air- Conditioning Engineers, 2002. I. Mundt, M.H. Mathisen, V.P. Nielsen, A. Moser, REHVA Guidebook No. 2. Ventilation Effectiveness, REHVA, 2004. SHASE Standard 115-2010, Field Measurement Methods for Ventilation Effectiveness in Rooms, Society of Heating, Air-conditioning and Sanitary Engineers of Japan, 2010. S. Cui, M. Cohen, P. Stabat, D. Marchio, CO2 tracer gas concentration decay method for measuring air change rate, Build. Environ. 84 (2015) 162e169. D. Etheridge, M. Sandberg, Building Ventilation, Theory and Measurement, Wiley, New York, 1996, pp. 471e473. H.B. Awbi, G. Gan, Evaluation of the overall performance of room air distribution, Proc. indoor air 3 (1993) 238. S. Liu, A. Novoselac, Lagrangian particle modeling in the indoor environment: a comparison of RANS and LES turbulence methods, HVAC&R Res. 20 (4) (2014) 480e495. X. Cao, J. Liu, N. Jiang, Q. Chen, Particle image velocimetry measurement of indoor airflow field: a review of the technologies and applications, Energy Build. 69 (2014) 367e380.