AP03_6.vp 1 Introduction: the “indoor quality” concept The quality of indoor air is affected by all components of the environment, the so-called constituents of the microcli- mate (see Fig. 1.1, Table 1.1) [27]. Existence of the individual components is obvious from: (1) the differential equation of the environment (only agents creating exposing flows in the human organism can be taken into account), (b) the stress the- ory according to Selye (each constituent is created only by those agents or complexes of agents creating one type of stress in the human body, see [27]). Air quality is thus dependent on its temperature and relative humidity, the concentrations of odors and toxic materials, the number of aerosols and microbes in the air, contamination by radioactive gases, static electricity and the numbers of negative and positive ions in the air. The impact of each of these factors depends on the mag- nitude of the stimulus. Taking into account air quality only, i.e. the chemical components of indoor air, pleasant or un- pleasant odors dominate the perception of the environment by the occupant, if there are no significant indoor sources. That is-why odors have become the criteria for assessing over- all air quality [39], [41]. However, we always have to remember that it is solely the odor microclimate that we are evaluating and not the overall indoor air quality. This means that, if a significant source is present, e.g. water vapour or carbon monoxide, the hygro- thermal or the toxic constituent has to be evaluated according to its own criteria (e.g. acceptable relative humidity of the air or acceptable carbon monoxide concentration). Simultaneously, the increasing requirements for indoor air quality in buildings need more exact criteria in order to ascertain the real condition of the environment and to allow Acta Polytechnica Vol. 43 No. 6/2003 22 Indoor Air Quality Assessment Based on Human Physiology – Part 1. New Criteria Proposal M. V. Jokl Human physiology research makes evident that the Weber-Fechner law applies not only to noise perception but also to the perception of other environmental components. Based on this fact, new decibel units for odor component representing indoor air quality in majority locations have been proposed: decicarbdiox dCd (for carbon dioxide CO2) and decitvoc dTv (for total volatile organic compound TVOC). Equations of these new units have been proved by application of a) experimental relationships between odor intensity (representing odor perception by the human body) and odor concentrations of CO2 and TVOC, b) individually measured CO2 and TVOC levels (concentrations) – from these new decibel units can be calculated and their values compared with decibel units of noise measured in the same locations. The undoubted benefit of using the decibel scale is that it gives much better approximation to human perception of odor intensity compared to the CO2 and TVOC concentration scales. Keywords: indoor air quality, odors, air changes estimation. Fig. 1.1: Types of constituent AGENTS MICROCLIMATE TOXIC SOLIDS SOLID AEROSOLS AEROSOL MICRO-ORGANISMUS MICROBIAL TOXIC LIQUIDS MATERIAL LIQUID AEROSOLS AEROSOL TOXIC GASES TOXIC ODOURS ODOUR AIR (ITS MOVEMENT) SPACE (ITS COLOURNESS) MAN (AS AN OBJECT) WATER VAPOURS "P S Y C H IC M IC R O C L IM A T E " HEAT CONVECTION CONDUCTION EVAPORATION HYGRO-THERMAL RESPIRATION RADIATION LIGHT UV RADIATION LASER RADIATION ELECTRO-DYNAMIC ENERGETIC MICROWAVE RADIATION (RADAR) IONIZING RADIATION IONS IN SURROUNDING AIR ELECTRO-IONIC STATIC ELECTRICITY ELECTROSTATIC SOUND ACOUSTIC FORCE FIELD (GRAVITATION) VIBRATION Table 1.1: Common environmental agents and corresponding microclimate types better optimization of its level (see [29], [30]), to remove “sick building” symptoms, i.e. to get the real comfort within a building. Human physiology research makes evident that the Weber-Fechner law applies not only to noise perception but also to the perception of other environmental components. According to this law human body response (R) is proportion- ate to the logarithm of the stimulus (S; k is a constant): R S� k log . (1) Applied to acoustic component of the environment L SPL P P P � � 20 0 log [dB], (2) where L SPLP � � sound pressure level [dB); P � acoustic pressure (when measuring the RMS value, i.e. the square root of the arithmetic average of a set of squared instantaneous values); P0 � acoustic pressure at the threshold of hearing (for air P0 � 20 �Pa). Analogical relationship could be supposed for odor component of environment [27] which determines the neces- sary indoor air exchange (by which unpleasant odors should be removed, the toxic gases removal by ventilation usually highly exceeds ordinary requirements and must be solved separately): � �Lodor odork� log � � 0 [dB], (3) where Lodor represents the odor concentration level in dB (the value representing the human response, i.e. odor perception [dB]), � represents the odor concentration in a building inte- rior [g � m�3], [ppm], �0 represents the odor concentration threshold value [g � m�3], [ppm]. The psycho-physical scale by Jaglou (the scale of perceived odors) could be applied to odor concentration levels [18]. In relationship to the percentage of dissatisfied (PD) its experi- mentally formed course is presented in Fig. 1.2. It is a logarithmic function (see equation (1)) that proves the logarithmic form of equation (3). The odor indoor air quality, decisive for air exchange, is dominantly determined by two odor substances: carbon diox- ide CO2, which cannot be neglected in rooms occupied by a higher number of people (it is produced by respiration) and the complex of volatile organic compounds (VOC) produced by the majority of upholstery and building materials. The following equations could be written for CO2 and TVOC: Lodor CO odor CO CO CO k 2 2 22 0 � � � � � log � � [dB CO2], (4) Lodor TVOC odor TVOC TVOC TVOC k� � � � � log � � 0 [dB TVOC], (5) where � CO2 and �TVOC represent the concentrations of CO2 and TVOC. Even though these equations (4, 5) look very promising from the physiological point of view, they must be experimen- tally proved. Furthermore, at least two points are necessary for each equation: (1) minimum threshold value, i.e. the weakest odor that can be detected (odor tresholds are statisti- cally determined points usually defined as the point at which 50 % of a given population will perceive on odor” [24]), (2) any second point can be chosen. We prefer the maximum threshold (limit) value, i.e. the beginning of the toxic range. The weakest odor that the smell organ of a healthy human can register has an intensity of l, according to the Yaglou psycho-physical scale [18], and corresponds approximately to a percentage dissatisfaction (PD) of 5.8 % (see Fig. 1.2). If we respect the similarity theory (see [21]), according to which analogous phenomena are governed by the same laws (e.g., concerning the perception of noise, odor etc. with intensity as a logarithmic function), then the corresponding minimal value for thermal comfort, as defined by Fanger [16], is 5 which is not too dissimilar from 5.8 for Yaglou’s odor value, taking into account the demanding nature of the experimen- tal procedure. There is a good collection of experimental values in litera- ture, therefore it has not been necessary to rely on our own measurements; even complete curves (see for example Fig. 1.2) are available. 2 Carbon dioxide For a long time, the odor constituent has been evalu- ated on the basis of CO2 concentration and its limit value of 1000 ppm, introduced by von Pettenkoffer (see [42]), was used to determine the minimum amount of fresh air (25 m3� h�1 per person). CO2 is the most important biologically active agent whose production is proportional to human met- abolic ratio [24]. In practice, monitoring CO2 levels for the purpose of controlling fresh air supply has proved satisfactory for lecture theatres, halls, cinemas, theatres and similar spaces where the load imposed by occupants can vary rapidly. In order to prove the equation (4), another experimental relationship besides the experimental relationship presented in Fig. 1.2 must be available: namely, the relationship between PD and CO2 concentration. This is presented in Fig. 1.3. Now we are able to shape the equation (4). The first point, the minimum threshold value for CO2 can be taken as 5.8 % dissatisfaction (Yaglou psycho-physical scale: 1), which is 485 ppm, i.e. 875 mg � m�3 Figs. 1.2 and 1.3 [15]. The second point used was the short-term exposure limit, which is the begin- ning of the toxic range, i.e. 15000 ppm. This is based on [22] from the Health and Safety Executive (HSE) Occupational Exposure Limits of Great Britain. © Czech Technical University Publishing House http://ctn.cvut.cz/ap/ 23 Acta Polytechnica Vol. 43 No. 6/2003 ODOR INTENSITY D IS S A T IS F IE D [% ] Fig. 1.2: The relationship between odor intensity (Yaglou’s psycho-physical scale) and the percentage of dissatisfied sedentary subjects during light activity (smoking not permitted) [18] So we are able to formulate the equation for odor level: L i odor CO CO 2 ppm decicarbdiox, dCd� 90 485 2log [ ] [ ] � (6) � �� �k kodor CO odor CO2 2log 15000 485 135 90� � � or L i odor CO CO -3 2 mg m decicarbdiox, dCd� � 90 875 2log [ ] [ ] � (6a) where decicarbdiox, dCd, is a new decibel unit for odor level (decibel carbon dioxide) caused by CO2 production by humans, � iCO2 is indoor air concentration, kodorCO2 is constant. Besides experimental functions already applied to equa- tion (4), a lot of individually measured CO2 levels from vari- ous locations (Tables 1.2 and 1.3, Fig. 1.4) are available for the verification of the equation (4). New decibel units dCd can be calculated from the mea- sured CO2 concentrations and their values can be compared Acta Polytechnica Vol. 43 No. 6/2003 24 Fig. 1.3: The percentage of dissatisfied sedentary subjects as a function of the carbon dioxide concentration above outdoors Sample Mean, dCd / ppm Range, dCd / ppm Study S NS S NS Public places 39 schools (NS) – – 28 dCd 990 ppm – – �5 to 68 425–2800 Smedje et al 1994 10 schools (NS) – – 38 1300 – – – – Thorstensen et al 1990 14 town halls (S) 20 800 – – 1 to 38 500–1300 – – Skov, Valbjorn 1987 5 office buildings (S) – – – – <26 <950 – – Palonen, Sepanen, 1990 4 office buildings (S) 20 800 – – �8 to 38 400–1300 – – Loewenstein 1989 26 office buildings (S) 11 639.4 – – �13 to 52 350–1850 – – Reynolds et al 1990 10 kindergarten – – 27 962.7 – – 25 to 66 915–2590 Piade et al 1988 1 office building (S) �6 420 – – – – – – Jaakola et al 1990 10 offices (5 S, 5 NS assumed) 8 590 4 533 – – – – Proctor et al 1989 1 office – – – – 8 to 13 600–675 – – Berglund et al 1982 1 library – – 16 731.5 – – – – Berglund et al 1988 9 office buildings (S) 15 710.6 – – – – – – Jones 1980 Homesa) 10 homes and apartments 14 692 �1 to 40 470–1360 van der Wal et al 1990 Table 1.2: Measured CO2 levels in various locations [23] © Czech Technical University Publishing House http://ctn.cvut.cz/ap/ 25 Acta Polytechnica Vol. 43 No. 6/2003 Sample Mean, dCd / ppm Range, dCd / ppm Study S NS S NS 18 homes 6 570 �13 to 24 350–900 Keskinen, Graeffe 1989 57 homes 16 730.5 – – van Dangen, van der Wal 1990 living room b.r. 29 1024 – – Op’t Veld, Slijpen 1993 living room a.r. 23 882 – – Op't Veld, Slijpen 1993 kitchen b.r. 29 1013 – – Op't Veld, Slijpen 1993 kitchen a.r. 21 836 – – Op't Veld, Slijpen 1993 bedroom b.r. 25 927 – – Op't Veld, Slijpen 1993 bedroom a.r. 13 672 – – Op't Veld, Slijpen 1993 S – smoking, NS – nonsmoking, a.r. – after renovation, b.r. – before renovation, negative values of dCd – measuring methods allow estimating of values below the detection threshold, a)no distiction made between smoking/non-smoking. Table 1.2: Measured CO2 levels in various locations [23] (continue) Before renovation After renovation WHO 1987 guideline values living room kitchen bedrooms living room kitchen bedrooms (dCd) mean 29 29 25 23 21 13 35 CO2 mean 1024 1013 927 882 836 672 1200 (ppm) std 184 277 193 160 122 148 CO mean 3.9 4.3 4.2 2.2 2.0 1.4 10 (8 h) (mg � m�3) std 0.4 0.2 1.6 1.0 0.5 0.5 CH2O* mean 665 577 530 405 357 231 120 (0.5 h) (�g � m�3) std 214 51 234 167 153 188 TVOC (ref CH4) mean – – – 4.5 4.1 2.9 – (mg � m�3) std – – – 1.0 1.0 1.0 NO2 mean 84 160 30 30 34 16 150 (�g � m�3) std 40 127 9 16 14 3 (24 h) Resp.dust mean 30 – – 30 – – 70 (�g � m�3) std 16 – – 15 – – (PM10 24 h) RH(%) mean 42 41 57 45 44 45 (30–70) % std 4 6 9 4 3 4 * including a.o. other aldehydes, C5H12, C6H14 Table 1.3: Measured indoor quality parameters before and after renovation (mean values and standard deviations, n � 16) [38] with decibel units for noise measured in the same locations (Fig. 1.5). A perfect agreement is evident. 26 © Czech Technical University Publishing House http://ctn.cvut.cz/ap/ Acta Polytechnica Vol. 43 No. 6/2003 Fig. 1.4: The measured CO2 levels in various buildings (see Table 1.2); the psychophysical scale slightly modified by Fanger (1988), S – smoking, NS – nonsmoking, a.r. – after renovation, b.r. – before renovation 3 Volatile organic compounds Although CO2 is a good indicator of the perceived air quality by sedentary persons, it is frequently also an un- suitable indicator: it does not represent further perceived sources of air contamination, such as building materials and fittings, especially carpets and other floor covering materials, producing VOC. This was why Fanger [17] proposed a new system based on the units of the “olf ” and “decipol”. It was presented in a number of international periodicals and publications and in 1992 became the EU recommended method of evaluating indoor air quality (EUR 14449 EN, 1992). It was not, how- ever, accepted for the BSR/ASHRAE Standard 62-1989 R (1989). There are certain obvious problems with this system, as were analysed by Oseland [39], Jokl et al [28] and especially by Parine [41]. Fanger’s system, as used in the EC standard, is based, rather than on CO2, on a new criterion: the total of all vola- tile organic compounds (TVOC) produced by humans and especially by building materials, furniture and other fit- tings. TVOC is also used for outside air quality as used for ventilation purposes, especially in areas with sources of con- tamination (chemical and other factories). TVOC is defined by the World Health Organisation (WHO) as a set of agents (toluene, xylene, pinene, 2-(2et- oxyetoxy), ethanol etc.) with a melting point below room temperature and a boiling point in the range 50–260 °C. More detailed definitions also exist. Humans sense TVOC by means of olfactory (smell) sen- sors (see [27]). Adaptation during the course of exposure is small. The response of the human organism to indoor TVOC has been classified as acute sensing of the environment, acute © Czech Technical University Publishing House http://ctn.cvut.cz/ap/ 27 Acta Polytechnica Vol. 43 No. 6/2003 or sub-acute inflammation of the skin or mucous membranes, or a sub-acute and mild stress reaction [15]. In order to prove the equation (5), besides the ex- perimental relationship presented in Fig. 1.2, another experimental relationship must be available: namely, the relationship between PD and TVOC concentrations. It is presented in Fig. 1.6. Now we are able to shape the equation (5). The first point, the minimum threshold value for TVOC, can be taken as 5.8 % dissatisfaction (Yaglou psychophysical scale: 1) which is 50 �g � m�3, see Fig. 1.6 (adapted from Fig. 1.2 in EUR 14449 EN, see [30]). The second point used was the short-term exposure limit, which is the beginning of the toxic range, i.e. 25000 �g m�3, which has been estimated by Molhave [36]. So we are able to formulate the equation for odor level: L iodor TVOC TVOC� 50 50 log � [decitvox, dTV] (7) ( � �k kodor TVOC odor TVOClog 25000 50 135 50� � � ) where decitvoc, dTv, is a new decibel unit for odor level caused by TVOC release from building materials and other sources (decibel TVOC), kodor TVOC is a constant. Besides experimental functions already applied to equa- tion (5), a lot of individually measured TVOC levels from vari- ous locations (Tables 1.3 and 1.4, Fig. 1.7) are available for the verification of the equation (5). New decibel units dTv can be calculated from the measured TVOC concentrations and their values can be com- pared with decibel units for noise measured in the same locations (Fig. 1.5). A perfect agreement is evident. 4 Conclusions 1. The undoubted benefit of using the decibel scale is that it gives a much better approximation to human preception of odor intensity compared to the CO2 and TVOC con- centration scales. This is because the human olfactory organ (see [27]) reacts to a logarithmic change in level 28 © Czech Technical University Publishing House http://ctn.cvut.cz/ap/ Acta Polytechnica Vol. 43 No. 6/2003 Fig. 1.6: The percentage of dissatisfied sedentary subjects as a function of the total volatile organic compound concen- tration (TVOC) above outdoors Sample TVOC [�g � m�3] Odor level [dTv] Study min mean max min mean max Living rooms and bedrooms old houses 130 200 240 21 30 34 Brown, Crump 1993 new houses 330 460 580 41 48 53 non-asthmatics 72 320 1600 8 40 75 Nordbäck et al 1993 asthmatics 67 540 8300 6 52 111 Office buildings a) without SBS during working hours 43 51 61 �3* 0 4 Ekberg 1993 during night-time 37 44 49 �6* �3* 0 during weekend 37 – – �6* – – b) with SBS (liquid process photo copiers) – 5000 90000 – 100 163 Broder et al 1993 SBS � Sick Building Syndrome * Measuring methods allow to estimate values under the detection threshold Table 1.4: Measured TVOC values in various locations which corresponds to the decibel scale, where a change of 1 dB is approximately the same relative change every- where on the scale. 2. The new decicarbdiox and decitvoc values also fit very well with the dB values for sound, e.g., the optimal odor value of 30 dB corresponds to the ISO Noise Rating Acceptable Value NR 30 for libraries and private offices. They can therefore be compared to each other. 3. It is possible, by comparing dCd and dTv values, to esti- mate, which component – CO2 or TVOC – plays a more important role and hence which sources of contamina- tion are more serious. © Czech Technical University Publishing House http://ctn.cvut.cz/ap/ 29 Acta Polytechnica Vol. 43 No. 6/2003 30 © Czech Technical University Publishing House http://ctn.cvut.cz/ap/ Acta Polytechnica Vol. 43 No. 6/2003 References See presented at the end of Part 3 (p. 44). Prof. Ing. Miloslav V. Jokl, DrSc. phone: +420 224 354 432 email: miloslav.jokl@fsv.cvut.cz Department of Engineering Equipment of Buildings Czech Technical University in Prague Faculty of Civil Engineering 166 29 Prague 6, Czech Republic