Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/6393029

Indoorairqualityinadentistryclinic

ArticleinScienceofTheTotalEnvironment·June2007

DOI:10.1016/j.scitotenv.2007.01.100·Source:PubMed

CITATIONS

102

READS

438

10authors,including:

C.G.Helmis

NationalandKapodistrianUniversityo…

166PUBLICATIONS2,178CITATIONS

SEEPROFILE

HelenaAFlocas

NationalandKapodistrianUniversityo…

136PUBLICATIONS1,736CITATIONS

SEEPROFILE

ChristosH.Halios

UniversityofReading

42PUBLICATIONS239CITATIONS

SEEPROFILE

GeorgeSgouros

RokasRenewables-Iberdrola

20PUBLICATIONS162CITATIONS

SEEPROFILE

Allin-textreferencesunderlinedinbluearelinkedtopublicationsonResearchGate,

lettingyouaccessandreadthemimmediately.

Availablefrom:HelenaAFlocas

Retrievedon:24November2016

ent 377 (2007) 349–365www.elsevier.com/locate/scitotenv

Science of the Total Environm

Indoor air quality in a dentistry clinic

C.G. Helmis a,⁎, J. Tzoutzas b, H.A. Flocas a, C.H. Halios a, O.I. Stathopoulou a,V.D. Assimakopoulos c, V. Panis b, M. Apostolatou a, G. Sgouros a, E. Adam b

a Division of Applied Physics, Department of Physics, University of Athens, University Campus, Build.Phys-5, Athens 157 84, Greeceb Department of Dentistry, University of Athens, Thivon 2, Goudi, 115 27, Athens, Greece

c Institute for Environmental Research and Suitable Development, National Observatory of Athens, 15236, P. Penteli, Greece

Received 24 July 2006; received in revised form 25 January 2007; accepted 27 January 2007Available online 16 April 2007

Abstract

The purpose of this work is to assess, both experimentally and theoretically the status of air quality in a dentistry clinic of theAthens University Dentistry Faculty with respect to chemical pollutants and identify the indoor sources associated with dentalactivities. Total VOCs, CO2, PM10, PM2.5, NOx and SO2 were measured over a period of approximately three months in a selecteddentistry clinic. High pollution levels during the operation hours regarding CO2, total VOCs and Particulate Matter were found,while in the non-working periods lower levels were recorded. On the contrary, NOx and SO2 remained at low levels for the wholeexperimental period. These conditions were associated with the number of occupants, the nature of the dental clinical procedures,the materials used and the ventilation schemes, which lead to high concentrations, far above the limits that are set by internationalorganizations and concern human exposure.

The indoor environmental conditions were investigated using the Computational Fluid Dynamics (CFD) model PHOENICS forinert gases simulation. The results revealed diagonal temperature stratification and low air velocities leading to pollutionstratification, accompanied by accumulation of inert gaseous species in certain areas of the room. Different schemes of naturalventilation were also applied in order to examine their effect on the indoor comfort conditions for the occupants, in terms of airrenewal and double cross ventilation was found to be most effective. The relative contribution of the indoor sources, which aremainly associated with indoor activities, was assessed by application of the Multi Chamber Indoor Air Quality Model (MIAQ) tothe experimental data. It was found that deposition onto indoor surfaces is an important removal mechanism while a great amountof particulate matter emitted in the Clinic burdened severely the indoor air quality. The natural ventilation of the room seemed toreduce the levels of the fine particles. The emission rates for the fine and coarse particulates were found to be almost equal, whilethe coarse particles were found susceptible to deposition onto indoor surfaces.© 2007 Elsevier B.V. All rights reserved.

Keywords: Dental clinics; Volatile organic compounds; Carbon dioxide; Particulate matter; Indoor air quality

1. Introduction

Recently the scientific community has becomeincreasingly interested in the air quality of indoor

⁎ Corresponding author. Tel.: +30 2107276927; fax: +30 2107295285.E-mail address: chelmis@phys.uoa.gr (C.G. Helmis).

0048-9697/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.scitotenv.2007.01.100

areas of hospitals and healthcare facilities. Operatingrooms, biochemical laboratories, infirmaries and privatepractices have been examined where the mixture ofpollutants, chemical compounds, microorganisms andbiological infectious agents in the air form indoor con-ditions which are dangerous to health for both patientsand health providers (Loizidou et al., 1992; San Jose-

350 C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

Alonso et al., 1999; Morawska, 2006). The comparisonof such an environment within hospitals with differentexposures to different risk factors, with or without air-conditioning, showed the positive effect of ventilationsystems on the improvement of indoor air quality, aslong as these systems are properly operated and wellmaintained (Holcatova et al., 2003) when outdoor con-centrations are low.

Indoor air quality in dental units has been alsoestimated and quantified with respect tomicrobial factors.Concentration measurements of microbial aerosols ingeneral dental practices have been performed in order tocarry out microbiological risk assessments (Bennett et al.,2000). The contamination levels have been analyzed byinvestigating the air, water and certain surfaces before,during and after dental treatments (Cellini et al., 2001;Monarca et al., 2002; Liguori et al., 2003). Otherresearchers have compared the levels of bacterial aerosolspollution between different dental environments aswell asdifferent positions within the same dental clinics (Grenier,1995; Kedjarune et al., 2000).

Regarding dental offices, the investigation of theenvironmental pollution has been focused on the con-tamination from infectious diseases (tuberculosis, hep-atitis, upper respiratory infections, and other viral orbacterial diseases) as produced by various dental pro-cedures (Micik et al., 1969; Mosley and White, 1975)and the methods for reducing airborne contamination(Miller and Micik, 1978; Littner et al., 1983; Kohn et al.,2003; Harrel and Molinari, 2004).

On the other hand, very few attempts have been madeto assess the air quality status of dental offices from thechemical point of view. Girdler and Sterling (1998) tested

Fig. 1. Ground plan of clinic, location of instruments for the pollutants m(●: location of DANTEC instruments, ⁎: dentistry chairs, |: bulkheads) (no

whether the exposure of dental staff to nitrous oxideduring inhalational sedation with nitrous oxide/oxygencomplied with specified occupational exposure standardswhile they assessed and determined the factors affectingthe levels of nitrous oxide pollution. More recently,Godwin et al. (2003) measured the concentration levels ofrespirable particulate matter, CO2 and VOCs in dentalclinics and estimated emission rates of indoor sources.

The objective of the present study was to evaluate theindoor environment of a selected clinic in the DentistryFaculty of Athens University with respect to CO2,TVOCs, PM10, PM2.5, SO2 and NOx and to identifypossible sources and relations between specific dentalactivities and pollution levels. Furthermore, the mechan-isms in the airflow and temperature fields associated withthe formation of the pollution levels were examined withthe application of the Computational Fluid Dynamics(CFD) model PHOENICS and the relative contribution ofthe indoor sources of particulate matter to the indoor airquality status was assessed with the aid of the MultiChamber Indoor Air Quality Model (MIAQ).

2. Experimental site, methodology andinstrumentation

The study took place in the Dentistry Faculty ofAthens University, which consists of two individual 5floor buildings (the Undergraduate Studies Building andthe Postgraduate Studies Building) connected by in-ternal corridors.

Before the main experiment, preliminary measure-ments of TVOCs, CO2 and particulate matter concentra-tions were performed in several areas of the Dentistry

easurements and measurement points of the airflow characteristicst in scale).

351C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

Faculty. From the results obtained, the Total TreatmentClinic, on the third floor of the Undergraduate StudiesBuilding, was selected according to its characteristics andhigh pollution levels. The Clinic has an area of 290m2 andoperates in two shifts (08:00–12:30 and 13:00–17:00)with 70–100 occupants in every shift. It is naturally (notmechanically) ventilated. Heating is achieved by centralheating radiators and air conditioners (A/Cs), which wererarely used. In Fig. 1, the ground plan of the Clinic ispresented, along with the position of the instruments usedfor the pollutant measurements.

In this room, the pollutants TVOCs, PM10, PM2.5,NO, NO2, SO2 and CO2 were monitored during theperiod from the 3rd December 2004 to the 3rd March2005, during both working days and weekends, usingthe following instrumentation:

• Portable instrumentation – two indoor air qualitymonitors (IAQRAE and ppbRAE of RAE systems)for TVOCs measurements (resolution: 10 ppb and1 ppb respectively, accuracy: 10%) and one monitor(IAQRAE of RAE systems) for CO2 measurements –were employed. The TVOCs and CO2 concentrationsrefer to 1-hour mean values, derived from 1-mincontinuous measurements. The measured values ofTVOCs are isobutylene equivalent and conversionfrom ppb to μg m−3 has been done by multiplyingthe measured value with the factor 2.3, according toAlevantis and Xenaki-Petreas (1996). The IAQRAEsystem provides also measurements of temperatureand relative humidity, as one hour mean values.

• Automated Horiba analyzers measuring NO, NO2,SO2, interfaced to a data logger giving 10-minaverage values. The NOx analyser uses a semicon-ductor sensor and the SO2 analyser an optical system.The principle of operation is the chemiluminescenceand UV fluorescence for the NOx and SO2 analyzersrespectively. The lowest detection limits (LDL) were0.98, 0.61 and 1.31 μg m−3

, respectively.• Particle samplers measuring PM10 and PM2.5, giv-ing mean concentrations calculated gravimetrically(weighing instrument KERN 770, accuracy 0.01 mg)from pre-set sampling periods (24 h, 10 h, 14 h)(Model 200 Personal Environmental Monitors (PEM)and SKC Universal DELUXE sampling air pumps of2 L min−1).

• Outdoor concentrations of PM10 and meteorologicaldata were collected from the air pollution monitoringstation operated by the Ministry of Environment(Goudi). The station was located at a distance of 50 mto the south of the Dentistry Faculty. From themeteorological data collected from the station, it was

evident that during 52% of the total experimentaltime the Dentistry Faculty was downwind of thestation, and 48% upwind.

During the whole experimental period a logbook waskept recording all the activities taking place in the clinic,including the number, the location and the duration of theopen windows, the number of students and personneloccupying the room and the nature of their work, thematerials used, as well as the cleaning processes andhours.

In order to quantify the ventilation prevailing in theclinic, the air change rates (ACH) were calculatedfollowing the methodology presented by Bartlett et al.(2004). ACH, measured in h−1, is the rate at whichoutside air replaces indoor air in a given space. Themethodology involves the solution of the mass-balanceequation for the CO2 concentrations, considering indoorhomogeneity and negligible deposition. Outdoor CO2

concentrations were frequently monitored during theexperiment and ranged on average at 1170 mg m−3.Indoor emission rate of CO2 was considered mainly dueto human respiration and was taken to be 589 mg min−1

CO2 per person (Godwin et al., 2003; Bartlett et al.,2004). The number of people in the clinic was estimatedaccording to the logbook records.

In order to further investigate the mechanismsassociated with the indoor environmental conditions ofthe clinic, a detailed examination both experimentallyand numerically of the air quality was conducted.Intensive TVOCs and PM measurements were per-formed, at two different locations of the room, in thecentral part (location K) and in the northern part(location B) simultaneously, during the period of 17–25 February 2005 (Fig. 1).

The indoor environmental conditions were examinedby applying the CFD model PHOENICS for the 19th,23rd, 24th and 25th February 2005 and the indoorproduction of Particulate Matter was assessed by employ-ing the indoor air quality model MIAQ for the 23rdFebruary. The necessary experimental data for theapplication of the previous models are spot mean airvelocity, temperature and turbulence intensity measure-ments at several indoor locations of the clinic, CO2

measurements at a fixed indoor position, all under differentventilation and occupational conditions, as well as surfacetemperature measurements of the indoor materials. Thedata were taken using the following instruments:

• DANTEC Flow Masters (type 54N60) for spot meanair velocity, temperature and turbulence intensitymeasurements of 1-min sets (accuracy 0.1 cm s−1,

Table 1Average concentration values of CO2, TVOCs, NOx and SO2, ranges of measured PM10 and PM2.5, temperature and relative humidity, as well asdiscomfort index, for the experimental period

Parameters measured Average backgroundvalue

Average working hoursvalue

Daily maximumrange

Daily valuesrange

Averagevalue

Limit values

CO2 (mg m−3) 1250 2200 1500–4600 1600 a1800TVOCs (μg m−3) 950 1900 2000–5500 1300 b1300, b2500NO (μg m−3) 50 65 10–650 55 –NO2 (μg m−3) 35 55 40–150 45 c250 (1 h)SO2 (μg m−3) 4 6.5 5–30 5 d350 (1 h)bTemperature (°C) 22.2 25.7 23.7–29.4 23.7 e20–24bRelative humidity (%) 26.7 25.6 24–41.7 26.1 f30–60%Discomfort indexDI Thom (°C)

19.1 21.1 19.8–22.7 19.9 gb21

PM10 indoor (μg m−3) 33–326 138 h50 (24 h)PM2.5 indoor (μg m−3) 23–229 75 i65 (24 h)PM10 outdoor (μg m−3) 4–40 14TVOCs outdoor (μg m−3) 140

The corresponding limit values are also shown.aASHRAE Standard 62-2001 rev. (2003).b(1) Molhave, 1995; Seifert, 1990; European Concerted Action, 1992. (2) Building Standard — State of Washington.cDirective 1999/30/EC.dDirective 1999/30/EC.eASHRAE Standard 55 (2004), for winter.fASHRAE Standard 55 (2004).gb21: normal conditions, 21–24: less than 50% of occupants feel discomfort, 24–27: more than 50% of occupants feel discomfort, 27–29: themajority of the occupants feel discomfort, 29–32: all the occupants feel discomfort, N32: need for medical treatment.hDirective 1999/30/EC.iUSEPA (US Environment Protection Agency, 1997).

352 C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

0.1 °C and 1%, for velocity, temperature and turbu-lence intensity, respectively).

• Infrared thermometer (Cole-Palmer, 08406) for surfacetemperature measurements of the indoor materials.

• IAQRAE for CO2 measurements.

Regarding MIAQ, the indoor environment was consi-dered as a single zone occupying a volume of 820.7 m3,with a total of 885m2 surfaces (ceiling, floor and 4 walls).The ventilation of the roomwith open doors and windowswasmeasured to be 1.75m3 s−1, for the 23rd February, byuse of the instrument Dantec.

CO indoor concentrations were continuously moni-tored with the IAQRAE during the experiment, but thelevels were very low, below the limit of detection of theinstrument and thus are not presented. Ozone concen-trations could not be measured in this study due tomalfunction of the ozone analyzers.

3. General indoor air quality characteristics

In Table 1, the average hourly values of TVOCs,CO2, NOx and SO2 during the working hours, the non-working hours and the whole experimental period are

presented, along with the range of the maximum hourlyvalues. Table 1 also presents the average daily values ofPM10 and PM2.5 during the working days along with therange of the daily values, as well as the correspondingthreshold limit values. It should be noted that the term“background concentrations” will be used hereafter forthe values recorded during the weekends and non-working hours, when the clinic was not occupied.

According to Table 1, the relative humidity in the clinicranged at low levels during the experimental period. Thethermal comfort is studied with the aid of Thom dis-comfort Index DI (Thom, 1959), which reflects theproportionate contribution of air temperature (T) andrelative humidity (RH) to the human thermal comfort.FromTable 1 can be seen that the dailymaximum value ofthe index ranges between 19.8 and 22.7 °C, suggestingthat acceptable thermal comfort conditions prevail in theclinic, while there are periods that a small percentage (lessthan 50%) of occupants feel discomfort.

3.1. CO2

Indoor CO2 concentrations are associated withhuman presence, since CO2 is metabolic, as well as

Fig. 2. Average diurnal variation of occupancy and air exchange rates, for the whole experimental period in the Clinic, on working days.

353C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

with the ventilation of the indoor environment (Penman,1980; Bartlett et al., 2004). In Fig. 2 the average diurnalvariation of occupancy during workdays along with thecalculated ACH is presented. The occupancy bar isdivided in two sections, one for the medical staff (10persons) and one for the patients and students. Duringtime periods with no occupancy (0:00–7:00 LST and18:00–23:00 LST), with windows and doors closed,mean ACH values were quite low and they did notexceed 0.6 h−1 (average values below 0.3 h−1). On theother hand during working hours, mean ACH valuesranged between 1 h−1 and 10 h−1, while the average ACHvalue was 5 h−1. The highest values were achieved insome cases under cross ventilation.

During the experiment, CO2 concentrations in theTotal Treatment Clinic reached high levels due to the largenumber of occupants (students, patients and medicalstaff), even though ventilation conditions seemed accept-able. The maximum values ranged between 1500–4600 mg m−3 (Table 1), depending on whether theroom was full and the number of open windows. Undercross ventilation, the CO2 values presented a significantdecrease (30%) due to sufficient air renewal. The meanconcentration during working hours was 2200 mg m−3.The comparison of this value with the limit set by theAmerican Society of Heating, Refrigerating and AirConditioning Engineers for indoor spaces (satisfactoryair quality: b1000 ppm or 1800 mg m−3, ASHRAE,2003) indicated that the indoor conditions during thesehours fall into the category of unsatisfactory air quality.It should be noted that high CO2 levels do not affecthuman health but they indicate insufficient ventilation ofthe room, which offers discomfort to the occupants.

When levels exceed the value of 1000 ppm it is sug-gested that immediate action should be taken to enhanceair renewal. The background levels were significantlylower, around 1250 mg m−3. The measured concentra-tion levels were comparable with those found by Godwinet al. (2003) in the operating room of a dental office for a1-week period.

The analysis of the mean hourly variations of CO2

concentration during the experimental period, revealed adiurnal pattern (Fig. 3). During the first hours of the daythe values remained at low background levels and theyincreased when the students started to occupy the room.A primary peak appeared at around 11:00 LST. Then asmall decrease was recorded, being attributed to thebreak for the change of shifts. When the occupantsstarted to regather for the afternoon shift, CO2 valuesincreased again. A second peak of the day appearedaround 15:00 LST. Then the concentrations decreasedgradually as the number of occupants decreased at theend of the second shift and finally reached backgroundvalues. During the weekends the concentrations re-mained at similar background levels.

3.2. TVOCs

TVOCs concentrations were significantly high(Table 1) and they far exceeded the recommendedlimit of 300 μg m−3 for the indoor environment set bythe international bibliography (Molhave, 1995; Seifert,1990; European Concerted Action, 1992) as well as theBuilding Standard of 500 μg m−3, which is in force inthe State of Washington. In the former guidelines, it isindicated that TVOCs concentrations below 0.2 mg m−3

Fig. 3. Average diurnal variation of CO2 concentration, for the whole experimental period in the Clinic, on working days.

354 C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

do not cause any irritation or other symptoms, whilebetween 0.2–3 mg m−3 there is possible irritation anddiscomfort experienced by the occupants, depending onother parameters also. Above the level of 3 mg m−3 andup to 25 mg m−3, there are important symptoms ob-served, such as headache and heavy discomfort feeling,while over the level of 25 mg m−3, neurotoxical effectsmay also appear.

During the working hours, the high levels (averagevalue of 1900 μg m−3) were attributed to the nature ofthe materials and substances used for the dentalprocedures, which are mainly acrylic and act as strong(intense) TVOCs sources, but also due to the ventilationconditions and the large number of the occupants.

The maximum values observed, varying between2000–5500 μg m−3, were associated with the use ofdetergent products for cleaning and disinfecting thesurfaces of the working posts at the beginning and theend of each shift, which enhanced the TVOCs con-centrations. It is worth mentioning that each time thedental substance Kalocryl® (a PMMA material for postand core modeling) was used, the TVOCs reached ex-ceptionally high values (5500 μg m−3).

In their study, Godwin et al. (2003) found that themost abundant VOCs were methanol, methyl acrylate,methyl methacrylate and isobutyl methacrylate. In ourcase, there are three types of material that emit highVOCs: a) detergent products like Bacillol composed by1-propanol (45%), 2-propanol (25%), ethanol (5%)withglutaraldehyde (formaldehyde free), b) composite ma-terials for dental restoration, like Bis-GMA (Bisphenol

A-Glycidylmethacrylate) and HEMA (hydroxyethyl-methacrylate hydrogels dissolved in organic solvents,and c) Kalocryl® that is a PMMA (polymethylmetha-crylate) material.

It is important to note that even though the backgroundconcentration levels (950 μg m−3) were lower ascompared to those during the working hours, they werestill higher than the international limit and much higherthan the mean outdoor concentration of TVOCs (of115 μg m−3, as derived from spot outdoor measurementswith the portable instrumentation). This could be at-tributed to: a) the fact that all surfaces are cleaned at theend of the second shift and then the clinic is kept closedduring non-working hours and weekends, and b) to thepossibility that during the working hours the walls absorbTVOCs and re-emit them during the non-working hours,keeping the background values high.

Similar to the CO2 diurnal pattern, TVOCs concentra-tions remained at lower levels early in the morning andconsiderably increased after 08:00 LST, while theyremained at high levels until the afternoon due to theuse of dental polymeric materials and the insufficientventilation of the room (Fig. 4). Three peaks were ob-served during the operation hours of the Clinic, associatedwith the use of detergent products for cleaning: the first at09:00 LST, the second at around 13:30 LST and the thirdat 17:00 LST. At the end of the last shift, the concentrationsdecreased gradually and reached background levels similarto those observed early in the morning.

Furthermore, during the weekends the concentrationsof TVOCs remained at background levels.

Fig. 4. Average diurnal variation of TVOCs concentration, for the whole experimental period in the Clinic, on working days.

355C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

3.3. Particulate matter

The measurements of PM10 and PM2.5 concentra-tions, as shown in Table 1, revealed very highconcentration levels due to the special materials usedand the dental procedures and treatments performed bythe students such as handpiece operation, trimming ofthe models, shaping of the temporaries, materialsagitation and materials mixing using hand mode.These sources combined with the large number ofpeople occupying the Clinic contributed to the increasedparticles levels, resulting in an average value of 138 μgm−3 for PM10 and 75 μg m−3 for PM2.5.

The highest levels were observed on the days when thenumber of the occupants was large and the windows anddoors were mostly closed. This situation can be seen on the20th December when concentrations reached 326 μg m−3

and 205 μg m−3 for PM10 and PM2.5 respectively. On thecontrary, on days with sufficient ventilation such as on the28th January (open doors and windows) the concentrationlevels were lower even though the number of people waslarge.

On most days, the daily concentration values of PM10

exceeded the limit of 50 μg m−3 suggested by theDirective 1999/30/EC. On the contrary, the internationallimit of 65 μg m−3 set by the USEPA (US EnvironmentProtectionAgency, 1997) for PM2.5 was exceeded only onhalf of the days.

The PM2.5 levels detected in this study are consistentwith the corresponding values reported by Godwin et al.(2003).

It is worth mentioning that the outdoor PM10

concentrations as measured at the air quality stationnearby the Dentistry Faculty (Table 1) were considerablylower than the indoor ones by an average factor of 12,suggesting the presence of indoor PM10 sources.

3.4. NOx and SO2

The measurements of NO, NO2 and SO2 concentra-tions showed that they remained at low levels for thewhole experimental period and within the recommendedoutdoor limits set by the Directive 1999/30/EC. Theaverage values obtained for these classic pollutants areshown in Table 1.

4. Detailed indoor air quality study

The TVOCs and Particle high concentrations ob-served in the experimental procedure implied the need toidentify the main mechanisms and sources contributingto these pollution levels. Thus, the indoor air quality wasfurther examined, including:

• Intensive measurements of TVOCs, PM10 and PM2.5,during the period of 17–25 February 2005, at twodifferent locations of the room, at the centre (locationK) and the northern part (location B) (Fig. 1).

• Application of the CFD model PHOENICS, for the19th, 23rd, 24th and 25th February 2005, in order toinvestigate the indoor environmental conditions ofthe clinic at peak working hours under different

Fig. 5. Variation of mean hourly TVOCs concentrations during the experimental period 17–25 February 2005, at two different locations, K and B, inthe Clinic.

356 C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

natural ventilation scenarios, by studying the airvelocity and temperature fields as well as the dis-persion of inert gaseous pollutants.

• Application of the numerical model MIAQ toestimate the relative contribution of the indoorsources of particulate matter to the indoor air qua-lity of the clinic during the 23rd February 2005, whenhigh values of PM2.5 and PM10 were observed. Thisday was selected because: a) it represents the mostcommon situation of the clinic on an everyday basisduring peak working hours, regarding occupancy,ventilation conditions and pollution levels, and b)data were adequate for the application of bothmodels.

Table 2Average TVOCs concentrations as well as daily PM10 and PM2.5 concentraFebruary 2005

Pollutants Indoor measurementlocations

Average backgroundvalue

Averavalue

TVOCs (μg m−3) K 1060 1880B 1500 3100

PM10 (μg m−3) KB

PM2.5 (μg m−3) KB

4.1. Intensive experimental measurements

The examination of the pollutant variations demon-strated that the concentrations at the two locations,followed the same diurnal pattern but their levels weredifferent (Fig. 5), with values at position B being twiceas high as those at position K.

More specifically, regarding TVOCs, the mean valueswere 1400μgm−3 and 2500μgm−3 at positions K andB,respectively, while the maximum values were 5500 and10500 μg m−3 (Table 2). Concerning PM10 and PM2.5,the maximum values (Table 2, Fig. 6) reached 439 and250 μg m−3 respectively at location B, while they were209 and 102 μg m−3 at location K.

tions at the measurement locations K and B for the period of 17–25

ge working hours Daily maximum valuesrange

Daily valuesrange

Averagevalue

2000–5500 14003000–10 500 2500

117–290 18975–439 24617–102 64193–250 223

Fig. 6. Variation of daily PM10 and PM2.5 concentrations during the experimental period 17–25 February 2005, at two different locations, K and B, inthe Clinic.

357C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

The difference between the concentration levels canbe attributed to the additional emissions of thesepollutants at the northern part of the room, wheremachines for grinding, smoothing and polishing acrylicmaterials are installed. The operation of these machinesacted as TVOCs and Particle sources, resulting inconsiderably higher concentrations at position B. Itshould be noted that the trimming and shaping facilitiesare performed in a separate area but unfortunately veryclose to the first four dental units of the clinic. Moreover,they are used only for very fine adjustments and im-provements to dentures, mouth guards, splints etc. Theproduction, trimming and shaping of the raw models ofdentures and mouth guards – that act as intense particlesources – takes place in a separate laboratory on thesecond floor of the building, totally separated from theexamined clinic facilities.

Furthermore, two sets of PM10 measurements wereperformed (08:00–18:00 LST and 18:00–08:00 LST)in order to record and compare the background andworking hours values (Fig. 6). During the evening andnight time (18:00–08:00) when there was no activity,the values of PM10 varied between 30–150 μg m−3.The concentrations during the working hours werehigher due to the dental treatments performed and thelarge number of occupants, reaching the value of439 μg m−3 when the ventilation conditions were poorand the value of 250 μg m−3 when air renewal wassufficient.

4.2. Numerical results

4.2.1. Indoor environmental conditionsThe general purpose CFD commercial software

package PHOENICS (Version 3.5.1, 2003, CHAM Ltd.)is able to solve the time averaged conservation equations ofmass, momentum, energy and chemical species in steadythree-dimensional flows, while turbulence and buoyancyeffects are taken into account (Spalding, 1981). The TotalTreatment Clinic was simulated in the PHOENICSenvironment with detailed geometry and real materialproperties (Fig. 7) and the model was validated against acertain experimental day (19/2, validation case), on whichthe clinic was empty, Door 1 and Windows 4, 8, 12 wereopen and heating radiators were off.

Satisfactory agreement was obtained from the com-parison of the numerical results with the experimentaldata. Table 3 presents the experimentally and theoreticallyassessed values of velocity (Uexp andUth) and temperature(Texp and Tth) at several points of the domain (air outlets,20 cmabove the bulkheads and 2mover the patients). Themaximum difference that was observed between exper-imental and theoretical velocities was ±0.4 m s−1.Considering the low velocities prevailing in general inthe clinic, the previous disagreement was attributed to thegeometrical modeling simplifications that were made forcomputational economy reasons. Indoor temperature wasunderestimated by the model, with maximum differenceobserved of 3 °C. However, similar to the velocity, the

Table 3Experimental and theoretical velocities and temperatures, at severaldomain points, for the validation case

Measurement points of domain Uexp

(m/s)Uth

(m/s)Texp(°C)

Tth(°C)

Window4 (air outlet) 0.64 0.72 21.2 16.7Window8 (air outlet) 0.81 0.70 20.9 16.7Window12 (air outlet) 0.70 0.69 21.9 16.71 (dentistry chair) 0.05 0.02 22.7 20.72 (bulkhead) 0.16 0.02 22.5 20.73 (dentistry chair) 0.09 0.03 23.8 20.84 (bulkhead) 0.14 0.12 22.5 20.55 (dentistry chair) 0.07 0.05 22.4 20.76 (bulkhead) 0.16 0.04 22.6 20.77 (dentistry chair) 0.06 0.05 22.6 20.78 (bulkhead) 0.20 0.12 22.4 20.69 (dentistry chair) 0.10 0.07 22.5 20.710 (bulkhead) 0.22 0.13 21.6 20.711 (dentistry chair) 0.15 0.06 22.1 20.812 (bulkhead) 0.15 0.22 21.6 20.613 (dentistry chair) 0.16 0.19 22.0 20.814 (bulkhead) 0.24 0.22 22.2 20.715 (dentistry chair) 0.17 0.41 22.1 20.816 (bulkhead) 0.16 0.48 22.0 20.717 (dentistry chair) 0.14 0.52 21.9 21.118 (bulkhead) 0.05 0.21 22.6 20.519 (bulkhead) 0.08 0.02 22.0 20.620 (dentistry chair) 0.12 0.19 22.7 20.721 (bulkhead) 0.10 0.50 22.6 20.822 (dentistry chair) 0.12 0.14 22.3 20.823 (bulkhead) 0.23 0.23 21.8 20.724 (dentistry chair) 0.10 0.16 22.7 20.925 (bulkhead) 0.30 0.03 22.0 20.826 (dentistry chair) 0.14 0.10 23.3 20.827 (bulkhead) 0.21 0.07 22.1 20.728 (bulkhead) 0.17 0.07 22.2 20.729 (bulkhead) 0.14 0.05 22.4 20.730 (dentistry chair) 0.13 0.10 22.9 20.831 (bulkhead) 0.2 0.06 22.8 20.632 (dentistry chair) 0.16 0.09 23.6 20.733 (bulkhead) 0.16 0.05 22.9 20.634 (dentistry chair) 0.10 0.07 23.5 20.735 (bulkhead) 0.12 0.04 23.7 20.636 (dentistry chair) 0.10 0.06 24.1 20.737 (bulkhead) 0.11 0.05 24.0 20.638 (dentistry chair) 0.11 0.04 24.9 20.739 (bulkhead) 0.09 0.02 25.4 20.640 (dentistry chair) 0.10 b0.01 25.6 20.741 (bulkhead) 0.10 b0.01 24.9 20.6

358 C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

largest differences were observed only at few domainpoints (including the air outlets and the northwest areaclose to the wall), probably attributing to unidentified heatsources, which could not be taken into account in themodel. Further results of the validation case can be foundin Stathopoulou and Assimakopoulos (2006).

Having obtained satisfactory agreement betweenexperimental and numerical data, three cases werestudied, aiming at determining the best ventilation schemethat can offer better comfort conditions for the occupants,in terms of CO2 levels and indoor air renewal. The modelconfiguration settings of the three cases are summarizedin Table 4. The first case represents the most commonsituation in the clinic. The room is fully occupied at peakworking hours and ventilation conditions are the samewith those described in the validation case. Fresh aircomes in the clinic through door 1 (air inlet) and is takenout through windows 4, 8, 12 (air outlet) according to theexperimental data, forming a natural cross-ventilationscheme. One patient corresponds to one dentistry chairand one doctor standing above, while people are modeledas heat and scalar sources emitting metabolic CO2. Thereference CO2 concentration that was considered for theinterpretation of the model results corresponds to theaverage value measured at domain point 39 (Fig. 7),where the IAQRAE monitor was placed, based onmeasurements taken during different experimental days.

The second and third cases examine the effect ofdifferent natural cross-ventilation schemes applied in theclinic. These schemes consider differently oriented openwindows and doors, while they were selected according tothe openings that were mostly used by the working staff(according to the diary) and the fact that air exchanges(ACH) in dentistry clinics have to be low for reasons ofinfections control. In the second case, the middle door(numbered 2) and three south windows are open(numbered 14, 16, 18), while in the third case, door 2and four windows (4, 8, 16, 18) are open (two at the southside and two at the north side), thus forming a doublecross-ventilation scheme (see Fig. 7). It should also benoted that the heating radiators were off in all casesexamined.

The results of the three studied cases are summarizedbelow:

1st case (23rd February): The indoor airflow field ischaracterized by a single re-circulating vortex-like flowpattern, with relatively lower velocities in the centralarea, breaking into distinct vortices of various sizesamong the modeled objects (Fig. 8). The air velocitiesprevailing in general in the clinic are low, rangingfrom 0.1 to 0.3 m s−1, and they decrease with height.ASHRAE standards suggest that indoor air velocity

should not exceed the value of 0.8 m s−1 to keepcomfortable conditions. In this respect, it is interestingto note that such velocities are only observed close to theopen door, up to approximately 4 m away from it. Theindoor temperature field exhibits diagonal stratificationfrom the lower levels of the southern part of the room tothe higher ones of the northern part (Fig. 9). The averageplane values range between 20–24 °C, while values

Table 4Information and input model data (the asterisk (⁎) corresponds to experimental data)

Data 1st case 2nd case 3rd case

Domain dimensions (m) 6.78×44.69×2.65 6.78×44.69×2.65 6.78×44.69×2.65Grid cells 50×262×30 50×262×30 50×262×30Open door 1 2 2Open windows 4, 8, 12 14, 16, 18 4, 8, 16, 18⁎Ventilation rate (m3 s−1) 1.75 1.55 1.48⁎Inlet air velocity (m s−1), axial along (−x) 1.04 at z=0.7 m 0.8 at z=0.7 m 1.14 at z=0.7 m

1.025 at z=1.4 m 1.02 at z=1.4 m 0.9 at z=1.4 m1.055 at z=2.1 m 0.94 at z=2.1 m 0.6 at z=2.1 m

⁎Inlet air temperature (°C) 20.5 at z=0.7 m 22.8 at z=0.7 m 23 at z=0.7 m20.75 at z=1.4 m 23.1 at z=1.4 m 23.3 at z=1.4 m21.45 at z=2.1 m 23.3 at z=2.1 m 23.7 at z=2.1 m

⁎Inlet air turbulence intensity (%) 17.5 at z=0.7 m 11 at z=0.7 m 42 at z=0.7 m32 at z=1.4 m 10 at z=1.4 m 22 at z=1.4 m17 at z=2.1 m 21 at z=2.1 m 48 at z=2.1 m

⁎Surface temperatures of objects (°C) ⁎13–22.5 ⁎13–22.5 ⁎13–22.5Total heat emissions per patient (W) 83.8 [ASHRAE (2004)] 83.8 [ASHRAE (2004)] 83.8 [ASHRAE (2004)]Total heat emissions per doctor (W) 126 [ASHRAE (2004)] 126 [ASHRAE (2004)] 126 [ASHRAE (2004)]Number of patients 38 38 38Number of doctors 38 38 38⁎Reference CO2 concentration (mg m−3) 2530 2479 1913

359C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

vertically increase by 1 °C in the southern area and 2 °Cin the northern one. Variations elsewhere in the domainare less than 1 °C. It is also interesting to note that theaverage temperature and air velocity at the breathingheight of the patients and doctors are 23 °C and 0.15 ms−1, respectively.

To examine the indoor air renewal and the dispersionof inert gases in the clinic, the spatial distribution of themetabolic CO2 concentration during peak workinghours is illustrated in Fig. 10. Pollutants accumulate inthe northern area, where higher temperatures occur,relating well with the temperature stratification observed(Fig. 9). CO2 concentrations significantly increase from319 mg m−3 at the south side to 2522 mg m−3 at thenorth side, indicating that this cross-ventilation schemeeffectively refreshes only the air of the south part of theclinic. The northern part suffers from uncomfortableconditions since CO2 levels exceed the maximumacceptable value of 1800 mg m− 3, according toASHRAE. The average plane values across the rowsof the dentistry chairs range at high levels, as well(2024 mg m−3 at the east row and 2100 mg m−3 at thewest row). It could be concluded that only about onethird of the indoor space of the clinic (the south area)maintains comfortable conditions regarding CO2 levelsand air renewal.

2nd case (24th February): The airflow field does notdepict significant changes compared with the previouscase. The re-circulating flow pattern is maintained withsmall differences owing to the different location of theopen door. Small differences are also observed in the

vortices of the x–z and y–z planes, mainly attributing tothe inlet profile that favours formation of large vortices(not shown). The temperature field retains the diagonalstratification pattern seen in the previous case, extendingfrom the middle door, which is now open, to the northwall and small vertical stratification is formed in thesouth part of the room. Also, higher temperatures occur(by approx. 2.5 °C) mainly due to the higher measuredair inlet temperature profile (not shown).

The CO2 concentration field follows the temperaturestratification pattern observed in the previous case (notshown). This is attributed to the low air velocities thatprevail in the clinic, which can not offer adequate airmixing throughout the indoor space and thus thermallydriven air motion dominates. As a result, warm areas areformed, joint with spot concentration areas due topollutants trapping. The average CO2 concentration ofx–z planes decrease along the north–south axis from2429 mg m−3 at the north wall to 818 mg m−3 in front ofthe open door and increases again to 2256 mg m−3 at thesouth wall. The reference CO2 concentration that wasgiven for the interpretation of the model results is2479 mg m−3 and corresponds to the mean value mea-sured during the experiment on that day. Similarly, withthe 1st case, this cross-ventilation scheme does not helpair renewal in the north area, while it offers relativelybetter conditions in the southern part of the room.However, at both sides CO2 levels exceed the limit value(ASHRAE) of 1800mgm−3 while acceptable conditionsare only obtained in the middle area close to the opendoor.

Fig. 7. Perspective view of geometrical domain.

Fig. 8. Airflow field (m s−1) across rows of bulkheads.

360 C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

Fig. 9. Air temperature field (°C) across rows of bulkheads.

Fig. 10. CO2 concentration field (×2530 mg m−3) across rows of bulkheads.

361C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

362 C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

3rd case (25th February): The airflow and temper-ature fields in this case are not considerably alteredcompared to those of the 2nd case (not shown), mainlydue to the similar characteristics of the air inlet profilesthat were measured on both days. The small differencesobserved do not have an important effect on the indoorconditions. Furthermore, the opening of one extrawindow compared to the other two cases and the doublecross ventilation do not seem to be able to reduce theindoor temperature levels, since heat sources due to thelarge number of people are significant.

Regarding indoor air renewal, CO2 levels are consid-erably reduced in the double cross-ventilation scheme,although the CO2 concentration patterns do not showimportant spatial changes as compared to the previouscase (not shown). CO2 levels are kept well below the limitof 1800 mg m−3 in the major part of the room and theyexceed it only in the northern area. The average CO2

concentration decreases along the north–south axis from1856 mg m−3 at the north wall to 325 mg m−3 in frontof the open door and increases again to 1722 mg m−3 atthe south wall. The reference CO2 concentration thatwas given for the interpretation of the model results is1913 mg m−3. All the above results indicate that thedouble cross-ventilation scheme is the best among thethree scenarios examined, since it offers relative betterconditions for the occupants in terms of air renewal andcomfort.

At this point it is important to note that the pollutionstratification pattern that was found to prevail in theclinic, independently of the ventilation scheme applied,was also confirmed by the intensive measurements(presented in Section 4.1), recording increased pollutionlevels at the northern location of the room compared tothe central one.

Besides ventilation, there are other room features thatare expected to contribute to the indoor CO2 concen-tration patterns, such as the arrangement of the dentistrychairs in the clinic, since these actually act as local heatand CO2 sources. If the spatial distribution of the chairswas different, the concentration patterns would also bedifferent.

Furthermore, the outdoor meteorological conditionsalso affect the prevailing indoor patterns, since theyinfluence the inflow–outflow relationship. The indoorair velocity and temperature fields are expected to beinfluenced and thus the dispersion of inert gases in theroom. Air drafts are developed and the more windowsare open, the more drafts are expected to appear, thuscausing an error to the simulated fields.

As a result of the above, outdoor CO2 levels may alsohave an impact on the indoor concentrations, but not on

the distribution patterns. This was also confirmed by theresults of the 3rd ventilation case. However, it should benoted that experimental measurements of outdoorbackground CO2 concentrations were in general low,compared to the metabolic CO2 produced indoors by theoccupants.

Since pollution sources in the clinic are significant dueto the nature of the dental operations and emissions arecontinuous during working hours, it is highlighted thatadequate ventilation must take place either through theopening ofmorewindows or by applying other ventilationtechniques. The latter is preferable since high air changesare not allowed in dentistry clinics for reasons of infectioncontrol and orientation of open windows was not found toimprove significantly the indoor air quality. However, it isessential that good ventilation should be applied beforethe changing of shifts, especially for the staff that is moreexposed.

4.2.2. Relative contribution of indoor PM sources to airquality

TheMulti chamber Indoor Air Quality model (MIAQ)is a general mathematical model for both indoor aerosoldynamics and the concentrations of chemically reactivecompounds in indoor air. MIAQ links a flexibledescription of building and ventilation system structureto a mechanistically sound analysis of particle dynamicsand indoor chemistry (Nazaroff and Cass, 1986, 1989). Inthe following, an attempt is made to assess the indoorparticulate sources in the Clinic during the 23rd February.For this purpose, consecutively numerical experimentswere performed. During these numerical experiments, themeasured outdoor particulate concentrations, indoortemperature and relative humidity along with thetemperature of the surfaces and the geometric character-istics of the indoor surfaces were set as inputs to themodel. Then, indoor sources with varying strengths weresimulated during the consecutive numerical experiments,until the average indoor concentrations calculated by themodel were equal to the indoor measured ones.

More specifically, the simulated room was consideredto be a single zone occupying a volume of 820.7 m3, witha total of 885 m2 surfaces (ceiling, floor and 4 walls).Ventilation of the simulated room was calculated with themethod presented in Section 2. Two aerosol size rangeswere considered: one is 0.1–2.5 μm and accounts forPM2.5 and the second 2.5–10 μm stands for PM10–PM2.5.Outdoor values of PM10 which were used as an input forthe model were measured at a fixed ambient stationoperated by the Ministry of Environment (see Section 3).The corresponding outdoor concentration values forPM2.5 were considered to be at 60% of PM10 values

363C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

since outdoor particulates were mainly associated withautomobile combustion (Chaloulakou et al., 2003, 2004).

Indoor sources were considered to vary with time.Thus, the number of the dentistry chairs that wereoccupied during each hour (according to the logbookkept during the experimental period) was considered asan index of the relative source strength during each hour.During the numerical experiments it was considered thatindoor particles were controlled by the interplay ofventilation, deposition, coagulation and indoor sources.The iteration step was set to 1 min and there were 30iteration steps in order to have the concentration results.

The time evolution of the simulated indoor PM10 andPM2.5 concentrations (not shown) revealed that indoorPM10 and PM2.5 concentrations reach extremely highvalues compared to the corresponding outdoor values,indicating the importance of the indoor emittedparticulates during the working hours. The highestindoor concentrations are observed during the timeperiod when the windows and doors are closed, reachingvalues as high as 456 μg m−3 and 367 μg m−3 for PM10

and PM2.5 concentrations respectively. The particulatelevels were reduced substantially (below 50 μg m−3)when the room was evacuated and the ventilationpermitted fresh outdoor air to enter indoors. It is in-teresting to note that according to the numerical exper-iments described, an average of about 2400 μg min−1 ofPM2.5 and a total of about 5000 μg min−1 of PM10 areemitted in the indoor environment. It is worth men-tioning that the estimated emission rates account notonly for the direct particulate emissions due to the variousclinical operations conducted within the experimentalroom, but also include particulate sources such as re-suspension. The PM2.5 emission rate found here is greatercompared to the respective values found in Godwin et al.(2003), where emission rates in the operatory room wereestimated to be 1217 μg min−1. It should be noted thoughthat in their case 7 people were present within theexperimental room throughout the working day, while inour case an average of 23 patients per hour were examined.

Finally, it is worth noting that even though the emis-sion rates for the fine (PM2.5) and coarse (PM10–2.5)particulates were found to be almost equal in ourexperiment (2400 μg min−1 and 2600 μg min− 1

respectively), the corresponding mass concentrations aremuch lower for the latter (average PM2.5 and PM10–2.5

values during the working hours were 102 μg m−3 and74 μg m−3 respectively). This is most probably due to thehigher deposition rates in accordance to the gravitationalsettling (Riley et al., 2002), indicating the importance ofthe deposition as a removal mechanism of the indoorcoarse particulates.

It should be noticed that the model does not explicitlytake into consideration all the dynamic processes (e.g.nucleation, condensation, re-suspension) affecting theindoor aerosol. However, in the simulations describedabove, such processes were included in the source term(Nazaroff and Cass, 1989).

5. Conclusions

The indoor air quality of a dentistry clinic was studiedboth experimentally and theoretically. It was found thatthe indoor air quality with respect to TVOCs, CO2 andPM was critical in the dentistry clinic due to the use ofspecific substances for dental operations, cleaningprocesses and also the high number of occupants in theroom. The commonly used natural ventilation schemesare not able to offer sufficient air renewal throughout theindoor space of the clinic. This causes accumulation andtrapping of air pollutants in certain areas of the room,especially in the northern part and, thus, in the formationof localized high pollution spots, isolated from the generalflow. The CO2 concentrations were high during theworking hours, compared to the international standards.Depending on the occupancy and the ventilation condi-tions, they reached the highest values when the maximumnumber of people and insufficient air renewal occurred.Acceptable CO2 levels for comfort conditions were onlyobserved at a relatively small indoor area close to the opendoor. Differently oriented open windows were not foundto considerably affect the indoor conditions, while thedouble cross-ventilation scheme was the most effective.

However, it should be taken into account that inhospitals and clinics, although the opening of windowsaids in air cleaning, high ACH are not allowed due toinfection control reasons and therefore the naturalventilation should be limited. Air removing fans at theceiling and air suction tubes in close vicinity to the dentalchairs would help, as well as the operation of an HVACsystem.

The significantly high levels of TVOCs concentra-tions recorded in the clinic were attributed to the use ofacrylic substances and dental materials. Exceptionallyhigh values were associated with the use of the dentalsubstance Kalocryl®. The detergents products used forthe decontamination of the working posts also contrib-uted to the enhancement of TVOCs, leading to peaks atthe beginning and the end of the shifts and increasedbackground values well above the limits set for theindoor environment. These high background valuescould be attributed to the cleaning after the end of thesecond shift and the closing of the clinic until the nextworking day, and the re-emission of the walls. Thus, it is

364 C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

recommended that the indoor air of the clinic should berenewed either by natural or mechanical ventilationduring non-working hours. For the spots of high TVOCsemissions, the operation of an automated system issuggested, which could consist of a sensor for TVOCsthat will activate a fan to remove the air above that area,until the TVOCs levels decrease.

The PM concentration levels were significantly highduring the working days, far exceeding the EU limit forthe outdoor environment. The emission and diffusion ofthe particles in the clinic can be attributed to handpieceoperation, the trimming of the models, the shaping of thetemporaries, the material agitation and the materialsmixing using hand mode. In specific, indoor values weremuch higher than the outdoors by a factor of 12.

The estimation of the relative contribution of theindoor particles sources, by using the MIAQ model,gave an average of about 2400 μg min−1 for PM2.5 and atotal of about 5000 μg min−1 for PM10, directly emittedin the indoor environment, severely affecting indoor airquality. Natural ventilation through the opening of doorsor windows reduced the indoor particulates, while thecoarse particles were substantially susceptible todeposition onto indoor surfaces. Even though theemission rates for the fine and coarse particulates werefound to be almost equal, the corresponding massconcentrations are much lower for the latter, mostprobably due to the higher deposition rates. RegardingNO, NO2 and SO2, no indoor sources were present, thusthe concentrations were found to be low and followedthe outdoor low values and variations.

In conclusion, it can be stated that the TotalTreatment Clinic of Dentistry Faculty is characterizedby significant pollutant emissions due to the nature ofthe dental procedures. Therefore, it is essential to im-prove the ventilation in order to prevent the occupantsfrom inhaling high pollutant concentrations, especiallythe doctors who spend most of their everyday time inthese environments.

Acknowledgment

The present study is co-funded by the EuropeanSocial Fund and National Resources in the frameworkof the project PYTHAGORAS (EPEAEK II).

References

Alevantis LE, Xenaki-Petreas M. Indoor air quality in practice. In:Santamouris M, Asimakopoulos DN, editors. Series: EnergyConservation in Buildings, CIENE, EC; 1996. Appendix A.

ASHRAE Standard 62-2001. rev. Ventilation for Acceptable Air Quality.Atlanta: ASHRAE; 2003.

ASHRAE Standard 55. Thermal Environment Conditions for HumanOccupancy. Atlanta: ASHRAE; 2004.

Bartlett KH, Martinez M, Bert J. Modeling of occupant-generated CO2

dynamics in naturally ventilated classrooms. J Occup Environ Hyg2004;1:139–48.

Bennett AM, Fulford MR, Walker JT, Bradshaw DJ, Martin MV,Marsh RD. Microbial aerosols in general dental practice. Br Dent J2000;189(12):664–7.

Cellini L,DiCampli E,DiCandiaM,Chiavaroli G.Quantitativemicrobialmonitoring in a dental office. Public Health 2001;115(4): 301–5.

Chaloulakou A, Kassomenos P, Spyrellis N, Demokritou P, KoutrakisP. Measurements of PM10 and PM2.5 particle concentrations inAthens, Greece. Atmos Environ 2003;37:649–60.

Directive 1999/30/EC of the European Council on 22 April 1999.European Concerted Action. Indoor Air Quality and its impact on man,

Guidelines for ventilation requirements in buildings. Commissionof the European communities, Joint Research Centre, 1992 ReportNo. 11, EUR 14449 EN, 1992.

Girdler NM, Sterling PA. Investigation of nitrous oxide pollutionarising from inhalational sedation for the extraction of teeth inchild patients. Int J Paediatr Dent 1998;8:93–102.

Godwin CC, Batterman SA, Sahni SP, Peng CY. Indoor environmentquality in dental clinics: potential concerns from particulate matter.Am J Dent 2003;16(4):260–6.

Grenier D. Quantitative analysis of bacterial aerosols in two differentdental clinic environments. Appl Environ Microbiol 1995;61(8):3165–8.

Harrel S, Molinari J. Aerosols and splatter in dentistry: a brief review ofthe literature and infection control implications. J Am Dent Assoc2004;135(4):429–37.

Holcatova I, Beneova V, Hartlova D. Comparison of the environmentin operating theatres in two hospitals. Indoor Built Environ2003;12(1–2):121–4.

Kedjarune U, Kukiattrakoon B, Yapong B, Chowanadisai S, Leggat P.Bacterial Aerosols in the dental clinic: effect of time, position andtype of treatment. Int Dent J 2000;50:103–7.

Kohn WG, Collins AS, Cleveland JL, Harte JA, Eklund KJ, MalvitzDM. Centers for disease control and prevention (CDC). Guidelinesfor infection control in dental healthcare settings. MMWR.Recomm Rep-Adm Conf U S 2003;52(RR-17):16–7.

Liguori G, Gombos F, Marinelli A, Lucariello A, Galle F, Marinelli P.Microbial environmental monitoring in the dental surgery room.Ann Ig Med Prev Comunita 2003;15(2):123–33.

Littner MM, Kaffe I, Tamse A. Occupational hazards in the dentaloffice and their control. I. Measures for controlling infections.Quintessence Interna 1983;14(1):67–9.

Loizidou M, Asimakopoulos DN, Lagoudi A, Petrakis M. Anassessment on the indoor air quality of houses and hospital inAthens. In: Lester JN, Perry R, Reynolds GL, editors. Quality ofthe indoor environment; 1992. p. 597–608.

Micik RE, Miller RL, Mazzarella MA, Ryge G. Studies on dentalaerobiology. I. Bacterial aerosols generated during dental proce-dures. J Dent Res 1969;48(1):49–56.

Miller RL, Micik RE. Air pollution and its control in the dental office.Dent Clin North Am 1978;22(3):453–76.

Molhave L. Evaluations of VOC emissions from materials andproducts: solid flooring materials. In: Maroni M, editor. Proceed-ings of Healthy Buildings '95 Conference, September 10–14 1995,1. Italy: Milan; 1995. p. 145–62.

Monarca S, Grottolo M, Feretti D, Gigola P, Zerbini I, Alberti A, et al.Environmental monitoring of infective risks in Italian dental offices.Minerva Stomatol 2002;51(7–8):319–26.

365C.G. Helmis et al. / Science of the Total Environment 377 (2007) 349–365

Morawska L. Droplet fate in indoor environments or can we preventthe spread of infection? Indoor Air 2006;16:335–47.

Mosley JW, White E. Viral hepatitis as an occupational hazard ofdentists. J Am Dent Assoc 1975;90(5):992–7.

Nazaroff WW, Cass GR. Mathematical modeling of chemically reactivepollutants in indoor air. Environ Sci Technol 1986;20:924–34.

Nazaroff WW, Cass GR. Mathematical modeling of indoor aerosoldynamics. Environ Sci Technol 1989;23:157–66.

Penman JM. An experimental determination of ventilation rate inoccupied rooms using atmospheric carbon dioxide concentration.Build Environ 1980;15:45–7.

Querol X, Alastuey A, Ruiz CR, Artinano B, Hansson HC, HarrisonRM, et al. Speciation and origin of PM10 and PM2.5 in selectedEuropean cities. Atmos Environ 2004;38:6547–55.

Riley WJ, McCone TE, Lai ACK, Nazaroff WW. Indoor particulatematter of outdoor origin: importance of size-dependent removalmechanisms. Environ Sci Technol 2002;36(2):200–7.

San Jose-Alonso JF, Velasco-Gomez E, Rey-Martinez FJ, Alvarez-Guerra M, Gallego-Pelaez C. Study on environmental quality of asurgical block. Energy Build 1999;29:179–87.

Seifert B. Regulating indoor air. In: Walkinshaw, editor. Indoor Air '90,Proceedings of the 5th International Conference on IndoorAirQualityandClimate, 29 July–August 3 1990, Toronto, Canada, vol. 5. ; 1990.p. 35–49.

Spalding DB. A general purpose computer program for multi-di-mensional one- and two-phase flow. Math Comput Simul 1981;23:267–76.

Stathopoulou OI, Assimakopoulos VD. Investigating environmentalconditions in dentistry clinics using computational fluid dynamics.Proceedings of the International Conference on Protection andrestoration of the environment VIII, July 3–7 2006 Chania,Greece; 2006. p. 575–6.

Thom EC. The discomfort index. Weatherwise 1959;12:57–60.U.S. Environmental Protection Agency 1997. Latest Findings on

National Air Quality: 1997 Status and Trends. www.epa.gov.