• Author / Uploaded
• Thavrin Than

• Categories
• Ventilación (Arquitectura)
• Escape de gas
• Naturaleza

Citation preview

Available online at www.sciencedirect.com

Building and Environment 39 (2004) 635 – 643

www.elsevier.com/locate/buildenv

Car park ventilation system: performance evaluation M.Y. Chan∗ , W.K. Chow Department of Building Services Engineering, The Hong Kong Polytechnic University, Hong Kong, China Received 22 June 2003; received in revised form 20 September 2003; accepted 14 October 2003

Abstract Unsatisfactory design of mechanical ventilation systems in car parks would give poor indoor environment. Site investigations on the ventilation systems in six car parks were carried out. Physical con5gurations of the car park and ventilation system installed were studied. Carbon monoxide levels were measured during the peak hours. From the results, parameters including the removal e7ectiveness and local air quality index were examined in conjunction with the measured carbon monoxide levels. Performance of di7erent types of ventilation systems was compared. It is found that the combined supply and exhaust system performs better than the exhaust-only system in controlling carbon monoxide levels at occupied zones (0.5 –1:5 m measured from =oor level) although more energy is consumed. ? 2004 Elsevier Ltd. All rights reserved. Keywords: Car park; Field measurement; Performance evaluation; Ventilation system

1. Introduction To cope with the increasing number of registered private cars, more multi-storey underground car parks were built in the Hong Kong Special Administrative Region (HKSAR) as the land is expensive [1]. A recent address by the Chief Executive of the HKSAR indicated the policy of striving for a better environment. Complaints of poor air quality within enclosed car parks have stirred up a hot issue in the community since 1993 [2]. Previous studies conducted in 1996 [3] revealed that the carbon monoxide (CO) levels in 25% of 103 car parks found in Hong Kong exceeded the World Health Organisation (WHO) standard 1-h time-weighted average (TWA) of 25 ppm CO level [4]. This means that it might bring adverse health impact to drivers, passengers and labour working in these premises. One of the aims of operating a ventilation system is to dilute the vehicular exhausts by introducing uncontaminated air and extracting the mixture away from the enclosure. In many projects, calculation of the amount of outdoor air required relied on the “well-mixed” condition. However, mixing depends on the air=ow patterns where the locations of supply points and extraction points of the ventilation system would be important. Note that supply and exhaust ∗

Corresponding author. E-mail address: [email protected] (M.Y. Chan).

0360-1323/$ - see front matter ? 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.buildenv.2003.10.009

positions are located preferably at opposite sides. The ratio of volumetric =ow rate of high- to low-level exhaust inlets was speci5ed in the Greater London Council [5] to be 1:2. The regulation does not specify the details of such inlets, with tailpipe level of vehicles taken as ‘low level’, and level above occupants’ height as ‘high level’ to avoid nuisance. The supply outlets are placed at high level or level close to the human breathing zone [6]. A possible explanation is fresh air can be delivered to occupants without mixing with vehicular exhaust [7]. The level of CO depends on air distribution, traNc density and ventilation rate. Piston =ow ventilation in many cases might be better than conventional mixing ventilation in terms of contaminant control. With ideal piston =ow, the air change eNciency is 100% better than conventional mixing [8,9]. However, piston =ow is not always practical due to congested traNc, space constraint resulting in supply and exhaust short-circuiting, and geometry (with sharp corners to give stagnant position) of the car park. The theory used in this paper assumes a mechanically ventilated, airtight enclosure where all the air enters or leaves via designated inlets and exhaust ducts except in5ltration compensation for the di7erence of supply and exhaust at the entrance. Six car parks were studied with performance of their ventilation systems evaluated. These car parks were selected because they are totally enclosed (roughly following the airtight assumption), underground and mechanically ventilated. The number of cars entering and leaving the car park

636

M.Y. Chan, W.K. Chow / Building and Environment 39 (2004) 635 – 643

Nomenclature A C(t)

area of the car park instantaneous carbon monoxide concentration of a point at time t (m3 m−3 ) average carbon monoxide concentration of the Ccp car park over a time period (m3 m−3 ) Ccp (∞) average carbon monoxide concentration of the car park at steady state (m3 m−3 ) Ce average carbon monoxide concentration of exhaust duct over a time period (m3 m−3 ) carbon monoxide concentration of exhaust duct Ce (t) at time t (m3 m−3 ) Ce (∞) carbon monoxide concentration of exhaust duct during steady state (m3 m−3 ) Cej carbon monoxide concentration of the jth exhaust inlet (m3 m−3 ) average carbon monoxide concentration of supCs ply air over a time period (m3 m−3 ) carbon monoxide concentration of supply duct Cs (t) at time t (m3 m−3 ) Cs (∞) carbon monoxide concentration of supply duct during steady state (m3 m−3 ) carbon monoxide concentration of the ith supply Csi outlet (m3 m−3 ) average carbon monoxide concentration over a CT time period T (m3 m−3 ) Cz average carbon monoxide concentration of the occupied zone over a time period (m3 m−3 ) G peak carbon monoxide generation rate (m3 s−1 ) m(t) volume of carbon monoxide that remained in the car park at time t (m3 )

was counted. CO level was monitored at various parts of the site including exhaust air duct, supply outlets, occupied areas (including lift lobby, staircase and pedestrian route), and some randomly selected areas. Results from measurement were used to assess the ventilation system performance. The indication of performance is e7ectiveness of pollutant removal and fresh air supply. The evaluation also focuses on the spatial variation of pollutant and fresh air supply. For car parks with bigger =oor area such as Site B, of volume 68; 000 m3 studied in this paper, the variation of pollutant levels and fresh air quantity might cause elevated exposure to vitiated vehicular exhausts. A series of parameters were used for assessment. The study eventually concludes on suggesting how car park ventilation systems can be assessed systematically. 2. General descriptions of sites Site A is a public housing estate with four parking levels. The total area of the car park is 2000 m2 and headroom 3 m,

m(∞) n Nop P Qf Q Qej Qinf Qsi T Qtj V  co co m−2   an cn   

volume of carbon monoxide that remained in the car park at steady state (m3 ) natural number number of operating cars per hour (h−1 ) total number of parking spaces outdoor air supply per =oor area (m3 s−1 m−2 ) total volumetric =ow rate of air (m3 s−1 ) volumetric =ow rate of the jth exhaust inlet (m3 s−1 ) volumetric =ow rate of outdoor air due to in5ltration at entrance (m3 s−1 ) volumetric =ow rate of the ith supply outlet (m3 s−1 ) time period (h) time step (h) volume of the car park (m3 ) ventilation eNciency (%) average carbon monoxide loading per hour (m3 h−1 ) average carbon monoxide loading per hour per =oor area (m3 h−1 m−2 ) activity level of car park (%) removal eNciency (%) nominal time constant of the ventilation system (min) turnover time of carbon monoxide (min) average carbon monoxide removal e7ectiveness (%) speci5c =ow (ach−1 ) local air quality index of occupied zone (%)

which provides 133 parking spaces for private cars. Levels 1 and 2 are naturally ventilated. Perforated bricks are laid along one side for natural ventilation. Levels 3 and 4 are mechanically ventilated and built underground, which are also the point of interest. The ventilation rate in terms of air changes per hour at levels 3 and 4 is 0.55. The ventilation system is an exhaust-only system. The exhaust inlets centre level is 2:5 m from the 5nished =oor. Site B is a two-level underground car park sited at a large commercial and shopping complex with an area of over 16; 000 m2 for each level. The headroom is 2 m, which is considered to be low. There are two types of exhaust inlet, 0.5 and 1:8 m centre level from the =oor. It provides a capacity of 914 parking spaces, with 395 on the 5rst level and 519 on the second level. Mechanical ventilation is provided in both levels by a combined system with eight exhaust fans and four supply fans serving each level. The supply rate is less than the exhaust rate. There is in5ltration at the main entrance and other possible air =ow paths. Outdoor air is directed to the supply outlets at the centre of the car park and distributed to the periphery by means of di7usion.

M.Y. Chan, W.K. Chow / Building and Environment 39 (2004) 635 – 643

637

Table 1 Summary of results of the six sites Parameters

Site A

Site B

Site C

Site D

Site E

Site F

A (m2 ) V (m3 ) P (number) Ventilation system Number of levels Headroom (m)  (ach−1 ) Q (m3 s−1 ) Qf (m3 s−1 m−2 ) co (m3 h−1 ) co m−2 (m3 h−1 m−2 ) Ce (m3 m−3 ) Ccp (m3 m−3 ) Cs (m3 m−3 ) Cz (m3 m−3 ) cn (min) an (min)  (dimensionless)  (dimensionless)  (dimensionless)  (dimensionless)  (dimensionless) CO at exhaust duct (ppm) Average CO (ppm) CO at occupied zone (ppm) CO at supply duct (ppm)

1454 4057 67 Exhaust-only 4 3 0.55 0.009 0.0004 13.3 0.009 6 8 0 14 145 109 75% 7% 75% 33% 43% 6 8 14 0

34,080 68,162 914 Combined 2 2 6.00 0.124 0.0033 42.3 1.240 100 129 2 40 13 10 77% 31% 77% 16% 250% 100 129 40 2

9896 22,102 413 Combined 3 2.2 3.80 0.057 0.0024 2.9 0.293 22 27 3 24 20 16 78% 17% 81% 26% 92% 22 27 24 3

4000 14,000 150 Exhaust-only 1 2.5 6.68 0.173 0.0065 3.4 0.860 37 55 5 30 13 9 70% 17% 66% 13% 122% 37 55 30 4

2000 9500 35 Combined 1 4.75 3.7 0.280 0.0049 0.3 0.140 7 11 3 9 29 16 56% 36% 69% 12% 80% 7.2 11 9 3

422 1358 42 Exhaust-only 1 3.2 4.46 0.040 0.0040 0.2 0.578 40 235 4 120 86 13 16% 24% 17% 2% 33% 40 235 120 4

The ventilation rate in terms of air changes per hour is 6 during normal operation and 9 during congested hours. In the course of CO monitoring, the ventilation system was operating at 6 air changes per hour. Site C is an institutional building with three levels. The total volume of the site is about 22; 000 m3 . The total capacity is 413 parking spaces, with 137 on the 5rst level, 144 on the second level and 132 on the third level. The ventilation system is a combined system with supply outlets and exhaust inlets at opposite sides. The exhaust inlets are located near =oor level (0:1 m) and at 1:6 m, while supply outlets are located at occupancy level (1:7 m). The ventilation rate in terms of air changes per hour is 3.8. Site D is a single =oor commercial underground car park with an area of 4000 m2 , which provides 150 parking spaces. The enclosure is ventilated by an exhaust-only system. The centre level of exhaust inlets is 2:2 m from the ground, which means that the vehicular exhausts would di7use through the breathing zone [6] before they are extracted away. The make-up air is drawn through the entrance by virtue of natural draughting. The ventilation rate in terms of air changes per hour is 6. Site E is a hotel basement car park, which serves exclusively for the tenants and sta7. The total area is 2000 m2 with high headroom of 4:75 m. The number of parking spaces is 35. The enclosure is ventilated by four exhaust fans and two supply fans. The supply outlets and exhaust inlets are 2.2 and 1:8 m from =oor level respectively. The supply and exhaust grilles are arranged alternately in an array.

The ventilation rate in terms of air changes per hour (ach) is 4, less than 6 ach as speci5ed in local codes. Site F is an underground car park with an area of 422 m2 and 42 parking spaces. The traNc is heavily congested on Sunday mornings. The car park is of an attendant-parking operation type. The sta7 of the car park is in charge of the operations in the reservoir space, directing the movement of incoming cars. The sta7 also operate the parking in the storage area. The storage of vehicles is closely packed together with no space left around. The ventilation rate in terms of air changes per hour is 4.46. The ventilation system is composed of two axial extraction fans at one end at 3 m above ground level. Fresh air is induced by means of natural draught at the entrance, implying that the assumption of airtight (without in5ltration) at building crack holds. The details of Sites A–F are shown in Table 1. The plans of Sites A–F are shown in Figs. 1–6.

Fig. 1. Site A plan view.

638

M.Y. Chan, W.K. Chow / Building and Environment 39 (2004) 635 – 643

mixing gives the CO volume m (m3 ) and concentration at exhaust duct Ce (m3 m−3 ) at time t without natural in5ltration: dm(t) G − QCe (t) = : (1) dt At non-peak hours, Ce (t) tends to be a steady-state value Ce (∞), for zero rate of change of m(t): G Ce (∞) = : (2) Q Putting in Eq. (2), the volume of CO remained in the car park at steady-state, m(∞) can be estimated as
∞ [Ce (∞) − Ce (t)] dt: (3) m(∞) = Q 0

Fig. 2. Site B plan view.

Steady state condition: The spatial average of CO level at steady state in the car park Ccp (∞) is the theoretical concentration under perfect mixing condition at steady state m(∞) Ccp (∞) = (4) V or
Q ∞ Ccp (∞) = [Ce (∞) − Ce (t)] dt: (5) V 0 Note that a mixing factor might be added for incomplete mixing. The average concentration Ccp (∞) will be equal to the exhaust concentration Ce (∞) if the ventilation is of conventional mixing type, under perfect mixing condition, leading to G Ccp (∞) = : (6) Q

Fig. 3. Site C plan view.

4. Key parameters The following parameters were measured in this 5eld study:

Fig. 4. Site D plan view.

3. Carbon monoxide balancing In a car park with ventilation rate Q (m3 s−1 ) and CO generation rate of G (m3 s−1 ), mass balancing for perfect

• Exposure to CO: The time prolonged of the occupants inside the car park and concentrations prevalent are the two major elements in assessing the exposure to pollutants. It is common to describe the two components together as time-weighted average [10]. The levels of CO prevalent in the car parks may =uctuate with time and so averaging is necessary for a better description of the concentration. The resulting exposure level to CO for a time period T; CT , i.e. the T hour time-weighted average, can be expressed in terms of the average CO concentration Ci at the ith hour [11]: n 1 CT = Ci : T

(7)

i=1

• Exhaust and supply concentration: For an enclosed car park with several supply outlets and exhaust inlets

M.Y. Chan, W.K. Chow / Building and Environment 39 (2004) 635 – 643

639

Fig. 5. Site E plan view.

the ith supply [12]: n i=1 Qsi Csi : Cs =  n i=1 Qsi

(8)

Similarly, Ce is given in terms of the CO concentration Cej at the jth exhaust inlet: n j=1 Qej Cej Ce = n : (9) j=1 Qej Neglecting the change of air density, conservation of volumetric =ow rate in terms of the volumetric =ow rate Qinf of outdoor air due to in5ltration at the entrance gives n 

Qsi + Qinf =

i=1

n 

Qej :

(10)

j=1

Note that i stands for the ith supply and j for the jth exhaust. Usually, the exhaust volume is greater than the supply volume. The di7erence is compensated by in5ltration at the entrance. For simplicity, Fig. 6. Site F plan view.

Q=

n  i=1

of volumetric =ow rates Qsi and Qej at the ith outlet and jth inlet, respectively, the =ow is steady and change of air density is negligible. The average supply (Cs ) and exhaust (Ce ) concentration over a time period of n supply grilles are de5ned in terms of the CO concentration Cs at

Qsi + Qinf =

n 

Qej :

(11)

j=1

• Speci5c 6ow: Ventilation rate has always been expressed in terms of ‘air changes per hour’, giving the erroneous impression that air is completely replaced at the given air change rate. But for the design practice in Hong Kong, while fresh air can be supplied at the given

640

M.Y. Chan, W.K. Chow / Building and Environment 39 (2004) 635 – 643

‘air change rate’, the ‘replaced air’ is normally a mixture of ‘stale’ and ‘fresh’ air. That is why there is a degree of mixing describing the mixing proportion of stale and fresh air. For this reason, ‘speci5c =ow’ () is useful for a car park of volume V : =

3600 × Q : V

(12)

• Car park CO loading per hour: The quantity of CO discharged from vehicles is a key factor in determining the ventilation rate. The ideal case is to count the total number of cars operating and measure the individual exhaust quantity from each vehicle. However, it is almost impossible to carry out such measurement and so monitoring the average concentration of CO in the exhaust air stream and volumetric =ow rate of all exhaust inlets was proposed (with in5ltration and ex5ltration neglected). Averaging over time will give the average CO loading per hour. The car park CO loading (co ) per hour can be given in terms of the time step (Qtj ) of each measurement as n j=1 Qej Cej Qtj co = : (13) T The total sampling time T is given by T=

n 

Qtj :

(14)

The monitoring is normally expected to cover the congested period and at least for a working shift of 8 h. • Activity level: Activity level () is the ratio of total number of operating cars observed to total capacity of the car park in an hour. There are automatic counters at the entrance and exit in these car parks and so the number of operating cars Nop (includes those entering or leaving the car park) can be recorded. The total capacity of the car park is the total number of parking spaces (P) provided: Nop ; P

(15)

• Nominal time constant of ventilation system: Time constant is the time interval from the time the air is present at a particular location to the time the air leaves the car park. A long nominal time constant is normally associated with poor indoor air quality. The overall average time constant in a car park is called the ‘nominal time constant’ (an ) of the whole car park. It is also the inverse of the speci5c =ow and hence is derivable from the ventilation rate and the volume of the car park [12]. It is independent of the internal =ow pattern or pollutant properties. The unit of nominal time constant interpreted here is minutes: an =

V : Q × 60

cn =

m(∞) : G × 60

(16)

• Turnover time of CO: It is similar to that of nominal time constant of air, but it represents the turnover time (cn ) of CO on an average [12]. The unit of turnover time

(17)

• CO ventilation e9ciency: The ventilation eNciency () re=ects the performance of a system with regard to the average ability to remove pollutants. It is based on the average car park concentration, supply concentration and average exhaust concentration over a time period [9]. It is given by the equation =

Ce − Cs : Ccp − Cs

(18)

If the supply concentration is near zero, or the exhaust concentration and average concentration are much higher than the supply concentration, then Eq. (18) can be approximated to =

Ce : Ccp

(19)

The ventilation eNciency at steady state is (∞) =

j=1

=

interpreted here is minutes. The mathematical expression is given by

Ce (∞) : Ccp (∞)

(20)

From Eqs. (2) and (4) (∞) =

an : cn

(21)

• CO removal e:ectiveness: CO removal e7ectiveness () is the description of movement and dilution of CO within a car park. Indices of CO removal e7ectiveness are dependent on both the characteristics of air=ow and the characteristics of the CO emission [9]. It is given by the ratio of average concentration of CO at the exhaust and the zone average over a time period. For conventional mixing ventilation, perfect mixing will give 100% e7ectiveness. Short-circuiting will give zone average concentration greater than the exhaust concentration. The CO removal e7ectiveness will range between zero and unity. For piston =ow ventilation, the CO removal e7ectiveness will be greater than unity. The mathematical expression is given by =

Ce : Ccp

(22)

This is the average e7ectiveness over a time period. • CO removal e9ciency: The CO removal eNciency () is a normalised value of the CO removal e7ectiveness. If the ventilation system is of a conventional mixing type, complete mixing of pollutants within the car park is 0.5. The de5nition of mixing ventilation is always explained in terms of the age of air. Complete mixing is a condition where the local mean age of air throughout the enclosure is the same and equal to the nominal time constant.

M.Y. Chan, W.K. Chow / Building and Environment 39 (2004) 635 – 643

It can also be explained in terms of the local =ow rate. If complete mixing occurs, the local =ow rate throughout the enclosure is the same and equal to the aggregate =ow rate [12]. The supply air and enclosure air are mixed by the actions of the supply momentum and buoyancy. Under this phenomenon, the supply air is used to dilute the concentration of CO and other pollutants in the car park. The mixture of air and CO or other pollutants is then extracted away via the exhaust duct. The above operation occurs as numerous =ow 5elds in the enclosure, but the mixing ratio of air and pollutant varies from one location to another. Imperfect mixing is due to short-circuiting of supply and exhaust. Short-circuiting gives a value of  between 0 and 0.5. Piston =ow ventilation gives a value between 0.5 and unity. Piston =ow ventilation is a unidirectional =ow of air in which supply air propels the pollutant ahead of it like a front. The action is analogous to a piston inside a cylinder, ‘sweeping’ the air to the outlet. That is why it is called “piston =ow”. The two-air side partitioned, with no or very few mixing is the basic requirement. In order to achieve this criterion, the air turbulence must be reduced to a minimum so that the dispersion of pollutant is minimised: =

Ce − Cmin ; Cmax − Cmin

(23)

where Cmax and Cmin are the maximum and minimum values of CO levels inside the car park, respectively. • Local air quality index at occupied zone: The local air quality index () at occupied zone is the ratio between the average concentration of CO at the exhaust (Ce ) and at the occupied zone (Cz ) over a time period, e.g. shro7 oNce, lifts lobby, pedestrian walkway, etc., =

Ce : Cz

(24)

5. Results The results on the measured key parameters are summarised in Table 1. The following are observed. Site A was not heavily utilised, so the surveyed activity level () was only 7.4%. The 5gure was averaged from an 8-h survey. The average CO loading (co ) was 13:3 m3 h−1 . The normal rate is expected to be 15 –20% as speci5ed in the literature such as ASHRAE Handbook [13]. Both the maximum concentration and 8-h time-weighted average did not exceed the values speci5ed by WHO [4]. The CO removal e7ectiveness () was 75%, which means that the average concentration was higher than the exhaust concentration (Ce ). Imperfect mixing was the reason for this phenomenon. The local air quality index () at the occupied zone was 43%, which was worse than the average. The removal eNciency () was 0.33, which implies conventional mixing.

641

Site B was heavily utilised throughout the survey. The surveyed activity level () was 31%, which means there were over 280 cars operating in the car park per hour. The CO loading (co ) was 42:29 m3 h−1 . The average loading of CO per =oor area (co m−2 ) was also the highest among the six sites. The fresh air supplied (Qf ) to the site per =oor area was ranked fourth in descending order. It is the main reason why CO level (Ccp ) was the highest, with 78 ppm for 8 h. Referring to the WHO standard [4], 50 ppm is the maximum permissible exposure level for the 8-h time-weighted average. Site C was heavily congested in the mornings and evenings. The usage other than these periods was relatively low. The 8-h time-weighted average was suppressed to 21:65 ppm due to the above reason. The maximum level in the morning reached a peak of around 200 ppm. The original design concept was a piston =ow system, but the derivation of removal eNciency () had shown that it is a conventional mixing system. Removal eNciency is between 0% and 100% (Eq. (23)). For Site D, peak hours were found in the evenings and during holidays. The 8-h time-weighted average was 55 ppm, which slightly exceeded the recommended level [4]. The ventilation system is an exhaust-only system. At some particular locations, the CO levels (e.g. non-pedestrian route) were extremely high because of insuNcient outdoor air supply. The average CO removal e7ectiveness () was 66.7%. The e7ectiveness was said to be fairly acceptable. The surveyed activity level () of Site E was 36.2%, which was the highest. However, the 8-h time-weighted average was 24 ppm. It was due to suNcient outdoor supply, good supply air distribution and high headroom. The outdoor air supplied per unit =oor area (Qf ) was 0:0049 m3 s−1 , which was the second highest. The supply outlets are evenly distributed along the walkway for users and the lift lobby. Outdoor air should be delivered to the occupied areas directly. Site F is of the attendant operated type and it was heavily congested. Note that attendant operated type is more eNcient for smaller car parks in terms of parking and de-parking time required. The density of parking was so high that the area per parking space was less than the recommended value of the Australian Standard, i.e. 23 m2 per space [14]. The area per parking space is only 10 m2 . There is no local standard governing the parking space requirement in Hong Kong. Momentarily, high levels of CO (say up to 235 ppm) were recorded at the areas near the exhaust outlet. The surveyed activity level () was 24.4% averaged over 8 h, but during some occasional periods, the activity level () was over 100% on Sunday morning. The activity level is de5ned as the percentage of turnover per hour. It is possible to have 5gures greater than 100%, implying that some cars stay in the car park for less than an hour. The local air quality index () was 35%, being the worst one of the six car parks. It is because the exit staircase is close to the parking area and exhaust outlet without isolation. Fresh air

642

M.Y. Chan, W.K. Chow / Building and Environment 39 (2004) 635 – 643

supply at the main entrance was heavily degraded by car exhaust at the entrance queue.

6. Discussion The measured 8-h time-weighted average of CO level at Site B was 78 ppm, the highest among the six sites. It was due to the high loading of CO emission from cars. The emissions from vehicles are close to the occupied areas. During the monitoring of CO at Site B, the ventilation system was operated at normal ventilation rate. The amount of fresh air supplied to unit =oor area was nearly the least. The dilution of CO was not as good as the others. For CO removal e7ectiveness (), the values lie between 17.2% and 80.5%. It is a ratio of exhaust concentration to average CO levels, which compares the level of CO prevalent at average condition. All of them are less than unity. It demonstrates the imperfect mixing of air and CO at all sites (see Table 1). However, some of them are very close to unity. The removal e7ectiveness is satisfactory at these sites. Normalised removal eNciency () examines the type of ventilation system. For conventional mixing ventilation, the values lie between 0 and 0.5. For piston =ow ventilation system, the values lie between 0.5 and unity. The normalised removal eNciency () of the six sites falls between 0 and 0.5. The type of ventilation is conventional mixing. The nominal time constant (an ) of the six sites ranged from 9 to 110 min. This parameter shows the average duration of stay of outdoor air from entering to leaving the car park. The ideal case would be as short as possible. As speci5ed in ASHRAE Application Handbook [13], the recommended ventilation rate is 6 ach, corresponding to a time constant of 10 min. However, it does not absolutely guarantee a good air quality even if the time constant is very short. Another parameter, turnover time (cn ) of CO will determine the average duration of stay of CO inside the car park. If the ventilation eNciency of a system is invariable, a short time constant will result in a short turnover time. A better indoor air quality can be expected. If the supply concentration (Cs ) is zero, or the exhaust concentration (Ce ) and car park average concentration are much higher than the supply concentration, ventilation eNciency () can be reduced to CO removal e7ectiveness () at steady-state condition. Removal eNciency () shows the ability of the system to remove pollutants. CO removal effectiveness () shows the control of average CO level in comparison to the exhaust level. If the mixing is perfect, and the system is conventional mixing,  will also be equal to  whatever the supply concentration or the di7erence is. The resulting value is 100%. It is almost impossible to measure the emission and CO left in the car park at steady state accurately. The only way to obtain the turnover time of CO is to take the ratio as

shown in Eq. (17) under stipulated conditions. It is a derived quantity and not an independent parameter. 7. Conclusion The removal e7ectiveness () had shown that the six sites employed conventional mixing system. It is very diNcult to obtain piston =ow ventilation even when the supply and exhaust ends are separate at opposite sides with careful arrangement. The undesired mixing e7ect was due to car movement, momentum force of supply air and buoyancy force of car exhausts. Although the removal eNciency () of piston =ow is twice that of conventional mixing, elevation of concentration may cause high levels of pollutant near the exhaust inlet. If the occupied areas (lift lobby, pedestrian route) are close to the exhaust inlets, piston =ow ventilation will cause adverse health e7ects to the occupants. This condition will end up with a local air quality index () of CO equal to unity or less than unity. A good distribution of supply air will normally have an even local air quality index greater than unity at all areas. Surplus supply of outdoor air at occupied zone is a good solution to improve the local air quality index. Drivers and car park attendants spend most of their time in these areas, normally taking a few minutes to park the cars. If the concentration of pollutants in these areas can be suppressed down to a relatively low level, the exposure to vitiated exhaust can be greatly reduced. The local air quality indices () for the combined systems at Sites B, C and E were 250%, 91.7% and 80%, respectively. These show that a relatively low level of CO was maintained at the occupied zone, indicating that air quality was improved. Various literature [13,15] had speci5ed the outdoor air supply for each operating car in underground car parks. The most often seen 5gure is 0:236 m3 s−1 per operating car with an average simultaneous usage of 3.5% and 5% peak of the total capacity of the car park. This is equivalent to a 5gure of 0.00826 and 0:0118 m3 s−1 per parking space for average and peak, respectively. All sites satisfy the requirement except Site A. The removal eNciency () indicates that the concentration gradient within the car park is quite large. It might be due to the stagnant zones created and the degree of mixing, which varied a lot with di7erent locations. Relatively high level of CO was prevalent at occupied areas. At Sites A and F, the local air quality indices () were 43% and 33%, respectively. Piston =ow ventilation had been proved to be more superior to conventional mixing provided that the occupied areas are not close to the exhaust inlets. It is concluded that it is diNcult to achieve piston =ow ventilation practically under various constraints. Mixing ventilation becomes the only alternative. The combination of mixing ventilation and local exhaust strategy at tailpipe level near parking stall would be the preference. The local exhaust inlets can minimise the

M.Y. Chan, W.K. Chow / Building and Environment 39 (2004) 635 – 643

emission strength so that the burden of dilution ventilation system can be reduced. Acknowledgements The authors wish to thank the cooperation of the car park management teams from various organisations. References [1] Transport Department, Hong Kong Special Administrative Region. Monthly TraNc & Transport Digest, December 1999. [2] Hong Kong Consumer Council. Choice, February 1993. p. 14 –20 (in Chinese). [3] Chan MY, Burnett J, Chow WK. Personal exposure to carbon monoxide in underground car parks in Hong Kong. Indoor Built Environment 1997;6:350–7. [4] World Health Organization. Guidelines for air quality. Geneva, Switzerland: WHO; 2000. http://www.who.org. [5] Greater London Council. London Building Acts (Amendment) Act 1939, Section 20, Code of Practice. UK: Greater London Council; 1939.

643

[6] American Society of Heating Refrigerating and Air-conditioning Engineers. Thermal environmental conditions for human occupancy. ASHRAE Standard ANSI/ASHRAE 55, UK, 1995. [7] Chan MY. Critical evaluation by 5eld studies on the performance of ventilation systems in enclosed car park in Hong Kong. PhD thesis, Department of Building Services Engineering, The Hong Kong Polytechnic University, Hong Kong, China, 1999. [8] Koskela HK, Rolin IE, Norell LO. Comparison between forced– displacement and mixing ventilation in a garage. ASHRAE Transaction 1991;97:1119–26. [9] Air In5ltration and Ventilation Centre. A guide to contaminant removal e7ectiveness. Technical Note AIVC 28.2, UK, December 1991. [10] Health and Safety Executive. EH40/96 occupational exposure limits, UK, 1996. [11] Chow WK. On ventilation design for underground car parks. Tunnelling and Underground Space Technology 1993;10:225–45. [12] Air In5ltration and Ventilation Centre. A guide to air change eNciency. Technical Note AIVC 28, UK, February 1990. [13] ASHRAE Handbook: Heating, ventilating, and air-conditioning systems and applications. Atlanta, GA, USA: American Society of Heating, Refrigerating and Air-Conditioning Engineers; 2003. [14] Council of Australian Standards. Mechanical ventilation for acceptable indoor-air quality. Australian Standard 1668, Part 2, 1991. [15] American Conference of Government Industrial Hygienists. Industrial ventilation, a manual of recommended practice, 21st ed. USA, 1998.