Building and Environment, Vol. 29, No. 3, pp. 311-317, 1994 Copyright © 1994 Elsevier Science Ltd Printed in Great Britain. All rights rc~erved 0360-1323/94 $7.00+ 0.00
~ ) Pergamon 0360-1323(94)E0013-H
Contamination Distribution in Displacement Ventilation--Influence of Disturbances ELISABETH MUNDT* Tracer gas measurements have shown that in a room with displacement ventilation pollutants can be locked in at different levels. Sometimes well defined layers o f pollutants can be found. The distribution is very sensitive to disturbances; opening and closing o f a door can cause a great decrease in the local ventilation effectiveness. In spite o f this, or perhaps because o f this, a person can obtain a good air quality in the breathing zone, even if this zone is in a polluted layer. The convective plume around a body breaks through the polluted layers very rapidly.
the room. The spread of tracer gas released outside a convection plume is dependent on the type of gas used. The concentration profiles in the room indicate a very strong stratification. Low local air exchange indices are found outside the plume and a connection between the local ages (which can be higher than the age of the air in the exhaust duct) and the sensitivity to disturbances is presented.
INTRODUCTION D I S P L A C E M E N T ventilation provides a well defined air transport within the convection plumes from the lower part of the room to the upper part. The temperature gradient is stable and not very sensitive to disturbances by people or door openings [1]. However, the contamination distribution is not as stable and behaves in a different way to the temperature distribution [2-4]. The objective of this work is to study the behaviour of contamination distribution under different conditions (see Fig. 1). Different tracer gases are used to study the contamination distribution from different sources (heated and not heated) under laboratory conditions. The measurements are conducted under the same conditions that have been previously used to study convection flows and temperature gradients [1, 5, 6]. The ventilation effectiveness and the air exchange efficiency are measured as well as the ventilation indices and the local air exchange indices for some points. The influence of disturbances caused by opening and closing a door is studied. The change in the local air exchange indices for some points is measured when a person enters
MEASUREMENTS The measurements are conducted in a room with L = 3.6 m, W = 3.6 m and H = 2.4 m. The room is equipped with displacement ventilation and a heat source representing a person simulator (100 W) is placed in the centre of the room. A higher gradient is obtained by using a radiator (200 W) placed 1 m above the floor on one of the walls. The supply air temperature can be varied and the temperature difference between exhaust and supply is always kept so that the heat produced within the room is evacuated by the ventilation air.
l(c_'? Fig. 1. Temperature and concentration profiles in a room with displacement ventilation. The temperatures, measured by 56 thermocouples, and the concentrations, measured by two infrared analysers (Miran), are recorded at variable time steps and collected
*Department of Building Services Engineering, Royal Institute of Technology, Stockholm, Sweden. 311
312
E. Mundt
Tracer gas source
Exhaust air
--
-~
ulator
1 --
5,9 - i " 4,8_~_ 0.4 m
C
--~.4 m
3,7
~diator
Tracer gas source
,[0.4 2,6 --]
Xo
m
~ ~ lOm
Person simulator
B
f Supply air
10
Door
Elevation
Plan
Fig. 2. Elevation and plan of the room used for the measurements. Tracer gas concentrations (1-10) were measured at the levels shown in the left figure and the positions of the rods with the tubes in the room are shown in the right figure. for further analysis. Two tracer gases are used, N20 with a density of 1.8 kg/m 3 and a mixture of NzO and He with a density equal to air. The tracer gas is supplied either as a point source in the room, placed in the convective plume from the person simulator or outside it. F o r some measurements the tracer gas is introduced in the supply air. Figure 2 shows the r o o m and the placing of the different measuring points. Altogether 12 tubes are used to measure the concentrations, but only two at a time. Two portable rods with four tubes each (nos 2-5 and nos 6-9) at different heights are used and placed in different positions (A, B or C ; see Fig. 2). One tube (no. 1) is placed in the exhaust air and one tube (no. 10) is placed close to the floor (see Fig. 2). Two more tubes are used to monitor the exact time when the tracer gas enters the room. Two different ventilation air flows are used (150 and 93 m3/h) and two different heat loads (100 and 300 W). With the above mentioned variation for the tracer gas, this gives a total of 16 possible measuring situations; however, only some of these are measured and reported in this paper. The measurements are conducted in the following way. (1) Step up procedure to measure the mean ventilation effectiveness (see Fig. 3)
( e ) = z.. 100 r,
=
Ce(O0)
(c(oo))
°
100,
where V z, = -- = nominal time constant q z, = mean age of tracer gas leaving the r o o m
Ce (O0) = concentration in the exhaust duct < C ( ~ ) > = mean concentration in the room. The tracer gas is released as a point source by means of a perforated ball with a diameter of 35 ram. The measurements are started when the tracer gas reaches the room. (2) Measuring of the local ventilation effectiveness
c~(~)
• lOO,
(2)
where Cp(OO) = concentration at a point in the room. (3) Mixing of the air in the r o o m and changing the tracer gas supply to the supply air duct. (4) The mixing is stopped and the temperature gradient allowed to develop.
*a) Step up
C e(t)/ Ce(~)
(1)
C e(t)J
t Fig. 3. Illustration of the calculation of different ages of air. (a) Mean age of tracer gas leaving the room, z,, from a step up measurement with a point pollution source. (b) Mean age of air leaving the room, r,, from a step down measurement starting with an even concentration in the whole room. z, is equal to the nominal time constant. Calculation of the room mean age of air (z). If the measurement is made at a point in the room the area under the curve is equal to zp.
Contamination Distribution in Displacement Ventilation (5) Stopping the tracer gas supply and measuring the air
exchange efficiency (see Fig. 3)
<~o> =
T, n
~-w~-," 100,
(3)
z,.,'r)
where (z> = room mean age of air; and the local air exchange effectiveness ~'n
Cap= --" 100, "Cp
(4)
where Tp = local age of air at a point. During step 2 different influences of disturbances are studied. The temperature gradients, the volume flows in the convective plumes and the heights to which the convective plumes rise as well as the level at which they start to disintegrate can be calculated by methods presented earlier [1, 5, 6].
RESULTS The air exchange efficiency and the local air exchange effectiveness ~p for some points in the room are given in Table 1 for different situations. The values missing in Table 1 were not measured. The parameters were in one case measured without letting the gradient develop after the mixing of the room air. The gas supply and the fans were stopped at the same time and the measurements started. These values are also presented in Table 1. Once the temperature gradient was developed it was linear, stable and the same in the whole room except for the region close to the supply device. Table 2 presents the mean ventilation effectiveness (~> for the different cases. From Table 1 can be seen that the air exchange effÉciency <~> is only slightly influenced by the temperature gradient being formed before the measurements. This can be explained by the fact that the gradient is formed very quickly after the fans are turned off; measurements have shown that the gradient is rebuilt within 5 minutes (see Fig. 4). Table 1 also shows that the air exchange efficiency does not change much with the
Table 1. Air exchange efficiency and local air exchange effectiveness e,p for different situations with points 2-5 in the plume, position A, and points 6-9 in position B Ventilation flow rate
150 m3/h
Heat load (W) Gradient (°C/m) e~z e.:3 e~4
~a:
e~p6 e~7 e~8 cop9 eapl0
100
93 m3/h 300
100
300
0
0.6
1.5
0.6
2
80 242 207 181 128 219 140 116 127 589
77 242 168 162 125 182 113 117 121 407
72
74 264 182 150 132 120 88 83 122 372
72
150 103 104 115
148 103 109 115
313
ventilation air flow or the temperature gradient in the room. The local air exchange effectiveness cap, however, is influenced by both the formation of the gradient and the ventilation air flow as well as the resulting gradient in the room. Points 6 and 7 in particular have a lower air exchange effectiveness when the gradient is formed before the measurements starts. This can be explained by the fact that the temperatures below these points after the mixing are higher than after the gradient is formed so the air exchange at these points is very high in the beginning of the formation of the gradient. It can also be seen that points 7 and 8 have a very low air exchange effectiveness when the ventilation flow rate is decreased with the lower gradient ; the age of the air is then higher at these points than the age of the air in the exhaust. The mean ventilation effectiveness (Table 2) is naturally increasing when the tracer gas is released in the plume compared to when it is released outside the plume. A tracer gas with a density higher than air gives a lower mean ventilation effectiveness. The local ventilation effectivenesses at different levels are presented in Figs 5 and 6 for tracer gas N20-4-He. The local ventilation effectivenesses are shown as the band within which they varied during one experiment. The variations are caused by the fluctuations in Cp(oo). The tracer gas concentration in the exhaust air, Ce(~), was stable during the measurements. In these figures are also shown the levels at which the convection flows from the person simulator equal the ventilation flows. This is said to be the border between two zones in the room. The convection flow from the person simulator is equal to 26 l/s (93 m3/h) at 1.25 m above the floor and 41.6 1/s (150 m3/h) at 1.6 m above the floor [6]. The rising area of the plume is shown as well ; the top of this is the maximum height to which the plume rises for a certain gradient in the room and the lower level is the height at which the density difference in the plume has disappeared. If the gradient in the room is 2.0 °C/m the maximum height for the plume from the person simulator to rise is 2 m above the floor and the density difference disappears at 1.55 m above the floor. Figure 5 shows that if the pollutant is released in the plume the lower part of the room is free from tracer gas (shown as ~ oo in Fig. 5) even if the ventilation air flow is less than the convective air flow from the heat source. In case of a high temperature gradient the concentration has a maximum at the level at which the plume is spreading out horizontally. When the tracer gas is released outside the plume (Fig. 6) the concentration profile gives two maxima (minima of the local ventilation effectiveness) for a low air flow and a high temperature gradient. The concentrations fluctuate considerably, as can be seen from Figs 5 and 6. The results presented in Figs 5 and 6 showing minima in the local ventilation effectiveness at the levels where the plumes spread out horizontally in the cases with high temperature gradients might also explain some of the results in Table 1. The low local air exchange effectiveness cap at points 7 and 8 with the low gradient and the low flow rate implies a stagnation zone. When the gradient is increased the local air exchange effectiveness at these
314
E. Mundt Table 2. Mean ventilation effectiveness(e) with different tracer gases and gas release positions
Ventilation flow rate
150 m3/h
Heat load (W) Gradient (°C/m)
100 0.6
Tracer gas
N,O
Tracer gas outlet
N20 + He
(e>
..... . . . . . . . . . . . . .
N20 + He
N20 + He
room
plume
room
plume
room
183
89
I 192
V 116
1II --
VII 115
II 161
VI 92
i
~'Ft
1 mm
i
I Iil I lil
....
,
I |1i
I I
.
i / f : , ' /
13rain [ ['ff [ l " i l l rain, Fans stopl~! Ii.~ I" i ......... i
i i
.
i I
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
i I
/ /
J
plume room IV 164
.......
VIII 94
.....
20
,
(:i
0.5
0o0
i
"[-[
0.5 '
17rain
.
1.0
'f
! i1"/ ,'7,
.:_:..- ,!!:! .......
2.0 .......
1.0
19
N20 + He
plume
i Time=0, Fans started • I
.......
300 2
Height above fleer (in)
i
.....................
100 0.6
room
Height above floor (in)
2,0
300 1.5
plume
Case no. in Figs 5 and 6
'
93 m~/h
LU 21 Temperature (°C)
! /,","/ i't
0.0 19
fL 20
J %
"
t
21 Temperature (°C)
Fig. 4. Evolution of the temperature gradient in a room when the mixing fans are started and stopped. The ventilation flow rate is 150 m3/h.
points improves. This can be explained by the spread of the plume in this area for a higher gradient improving the air exchange in these points. Two types of experiments were made concerning the stability of the concentration in the room : first the influence of opening the door and closing it again, and second the influence on the local ventilation effectiveness in some points of a person entering the room. During the experiments the exhaust air concentration, C~(oo), was stable and the results are presented as the variation in the local ventilation effectiveness at different points as a function of time. The measurements were continued until steady state concentrations were obtained again after the disturbance. In case VI in Fig. 6 (low gradient, low air flow rate and tracer gas released outside the plume) the door was opened and closed and the concentration at point 7 (1.4 m above the floor at position C) was studied. Figure 7 shows the local ventilation effectiveness during the measurement. As can be seen the local ventilation effectiveness decreases greatly for about 5 minutes and then slowly returns to the initial value. From Table 1 the local air exchange effectiveness at this level is 88 and the nominal time constant for the low flow rate is 0.33 h; this gives the mean age of the air at the point to be 22
minutes. The time for the concentration to return to the initial value after the maximum is also about 20 minutes. In case IV in Fig. 5 (high gradient, low air flow rate and tracer gas released in the plume) the concentrations in point 4, position C, and point 8, position B (1.8 m above the floor), were studied when a person entered the room and stood at the positions B and C for 5 minutes at each place. Figure 8 shows the local ventilation effectiveness at these points. As can be seen the air is immediately cleaner at these points because of the convection plumes around the body. After the persons left the room the concentrations returned to the original ones, and the same time dependence as mentioned above was noticed. Other experiments showed the same effect ; each person creates its own convective plume taking the air from the lower parts of the room. The same experiments were done with case VIII in Fig, 3, when the concentration at the level 1.0 m is high. For this case the concentration at level 1.4 m and above increased and gave a decrease in the local ventilation effectiveness when a person entered the room. The temperature gradients were not influenced by the persons entering the room, only a slight move to a higher temperature was noticed because of the increased heat load in the room ; see Fig. 9.
Contamination Distribution in Displacement Ventilation
315
Height above floor (m) • 2.4
P l u m e r i s i n g area 1.5°C/m ~-~HI 2-0°C/m / SSl~"
-
2.0 lII
~
I
Convection flow = 4 1 . 6 1/s
o
o
i~,m~rv
w
Height of person simulator (100 W) Case Cue Case Case I SO
v I I I 250 300 350 L o c a l v e n t i l a t i o n effectivene~m Fig. 5. Local ventilation effectiveness at different levels at position B when the tracer gas source is in the plume,
0
I 150
I: low gradient, high flow rate II: low gradient, low flow rote HI: ~ gradient, high flow rate IV: high gradient, low flow rate
100
I 200
Height above floor (m) P l u n ~ rising area 1.5°C/m V 2"O°C/m
2.4
/ 2 . 0 -W~
"T ..... 1.6
.:....vl._. vn
) V
VI
= 4 1 . 6 Us
Vll
1.25 ~ Convection flow = 2 6 . 0 l/s
V
VII
1.0
• b
Vl
Height of person simulator (100 W) Case Case Case C~tse l 50
l 150
i l l 2,50 300 350 Local venti~tion eff~etiven~ Fig. 6. Local ventilation effectiveness at different levels at position C when the tracer gas source is outside the plume. 0
100
l 200
V: low 8rmlient, high flow rate Vl: low ~-miient, low flow rate VII: high gradienthigh flow rate 7111: h i g h g r a d i e n t , l o w f l o w r a t e v
316
E. Mundt Local 500
ventilation
Ep
effectiveness
400
/
300 200
I00 0 0 10 Door o)ened and closed
20
40
30 Time
(min)
Fig. 7. Local ventilation effectiveness at point 7 when the door is opened and closed.
Local 400
ventilation
effectiveness Ep
T1 One person enters the room and stands at Pos C
300
II
/I
fl
Point 4, P, )s C Point 8, Pq ps B
T2 The person moves to pos B
....
2OO
T3 Another person enters the room and stands at Pos C
100
0 T1
0 T2
20 T4
T3
30
40 Time (min)
T4 Both persons leave the room
Fig. 8. Influence of persons on the local ventilation effectiveness at the level 1.8 m.
Height above floor (m) m
m
:=5 '--10 t=15 = 20
i
m
i
l
i
m
t=25
j
r~
1.5'
1.0
0.5
,/
0o0
16
8
20
22 24 Temperature (°C)
Fig. 9. Variation of the temperature gradient, measured 0.8 m from position B, during the experiment shown in Fig. 8.
Contamination Distribution in Displacement Ventilation CONCLUSIONS In a room with displacement ventilation pollutants can be locked in at different levels depending on how they are produced, temperature gradients and ventilation air flows versus convective air flows. The upper zone is not a well mixed zone ; sometimes well defined layers of pollutants can be found. The concentrations can fluctuate greatly in the layer between the upper and lower zones. The distribution is very sensitive to disturbances; opening and closing of a door can cause a great decrease in the local ventilation effectiveness at a point. In spite of this, or perhaps because of this, a person can obtain a good air quality in the breathing zone, even if this zone is in a polluted layer. The convective plume around a
317
body breaks through the polluted layers and very rapidly increases the local ventilation effectiveness. The displacement ventilation acts as a demand controlled system for clean air from the lower parts of the room. If however the lower parts are polluted the reverse effect, a decrease in the local ventilation effectiveness, is obtained. This paper presents some results concerning the spread of contaminants in a room of moderate size with displacement ventilation. Further work is needed to investigate the air flow pattern in different situations in order to understand the influence of the different parameters on the air exchange and on the contamination distribution.
Acknowledgement--TheSwedish Council for Building Research supported this work.
REFERENCES 1. E. Mundt, Convection flows above common heat sources in rooms with displacement ventilation. Proc. Room-Vent 90, Oslo (1990). 2. R.B. Holmberg, L. Eliasson, K. Folkesson and O. Strindehag, Inhalation-zone air quality provided by displacement ventilation. Proc. Room-Vent 90, Oslo (1990). 3. J. Nickel, Air quality in a conferenceroom with tobacco smoking ventilated with mixed or displacement ventilation. Proc. Room-Vent 90, Oslo (1990). 4. H. Stymne, M. Sandberg and M. Mattson, Dispersion pattern of contaminants in a displacement ventilated room--implications for demand control. 12th AIVC Conference, Ottawa, Canada (1991). 5. E. Mundt, Temperturgradienter och konvektionsfl6den vid deplacerande ventilation. Meddelande nr 16, Installationsteknik, KTH, Stockholm (1991). 6. E. Mundt, Convection flows in rooms with temperature gradients--theory and measurements. Proc. Room- Vent 92, Aalborg (1992).