Thermal Comfort Assessment in Comfort-Prone ...

Ann. Occup. Hyg., Vol. 51, No. 6, pp. 543–551, 2007 Ó The Author 2007. Published by Oxford University Press on behalf of the British Occupational Hygiene Society doi:10.1093/annhyg/mem032

Thermal Comfort Assessment in Comfort-Prone Workplaces PAOLO LENZUNI* and MICHELE DEL GAUDIO

Received 22 August 2006; in final form 31 May 2007 This paper’s main issue is a strong advocacy in favour of an a priori classification of thermal environments that can be really functional to comfort assessment: Class 1, environments where comfort conditions can be established (comfort-prone environments), and Class 2, environments where this is not practically feasible. The former, which are also identified here as ‘thermally unconstrained’ environments, because of the absence of elements preventing comfort from being pursued, are the subject of a novel classification scheme. In assembling such a scheme, the four standardized synthetic indexes (Predicted Mean Vote, Insulation REQuired, Predicted Heat Strain, Wet-Bulb Globe Temperature) have been carefully scrutinized, with special emphasis on the regions of overlap. Additional data from national technical documents and legislation have been used to help in assembling the discomfort assessment scheme. All available information has been reprocessed and cast in a form specific for use in comfort-prone environments. Classification takes place through placement in a four-level and in a six-level discomfort scale for cold and warm environments, respectively; for each area, a recommended descriptor as well as a time frame for intervention are specified. The new scheme also eliminates a few glitches and inconsistencies existing in the ISO 15265 scheme, mostly in the area of cold discomfort. Being solely concerned with comfort-prone environments and keeping an open mind with respect to all available information, the new classification scheme represents a simple and robust all-round tool, tackling issues related to both comfort assessment and to action planning for an optimized allocation of available resources. Keywords: comfort; risk assessment; thermal environment

The need for a reliable classification scheme and an associated intervention strategy remains undented. Accordingly, this paper’s first objective is to establish a new general framework for classification of thermal environments, specifically tailored to occupational risk assessment. Because of the extreme diversity of thermal environments, a single all-purpose study is very impractical to perform and definitely not user-friendly. Therefore, this paper’s second objective is to develop a classification scheme which focuses only on comfort-related issues. Stress-related issues, with emphasis on hyper- and hypo-thermal environments, will be dealt with in a future paper. The paper’s plan is the following: in Risk Assessment a set of definitions is introduced to single out comfort-prone environments, from those where health protection is the key issue, and provide a clear frame for the ensuing analysis of discomfort; in Cold Environments we discuss cold environments by crosschecking information obtained from the standardized

INTRODUCTION

Thermal risk assessment is usually perceived by occupational hygienists as a minor topic compared to its analogues in acoustics or toxicology. It is nevertheless a potentially very useful template, since it has to deal with both comfort- and stress-related issues at the same time. Risk assessment strategies for thermal environments have been recently tackled by international standard ISO 15265 (2004b), which links them to a single one-dimensional classification scheme (i.e. classification of thermal environments through one parameter, in a single sequence). Unfortunately, there are both conceptual and practical pitfalls in this document, which make it poorly suited as a tool for both worker’s care and technical action. *Author to whom correspondence should be addressed. Tel: þ39 055 289681; fax: þ39 055 210882; e-mail: [email protected] 543

Downloaded from http://annhyg.oxfordjournals.org/ at Medical College of Wisconsin Libraries-Serials on December 30, 2015

Department of Florence, Italian National Institute for Occupational Prevention and Safety, Via G. LaPira 17, 50121 Firenze, Italy

544

P. Lenzuni and M. Del Gaudio

indexes Predicted Mean Vote (PMV) and Insulation REQuired (IREQ), and minimum acceptable conditions set by a variety of national documents; eventually, results lead to the synthesis of a classification scheme laid out in the form of a set of limiting curves, specific for comfort-prone environments; this process is replicated in Warm Thermal Environments for warm and hot environments, with the appropriate standardized indexes.

Classification according to ISO 15265 This standard establishes a single sequence based on the PMV [ISO 7730 (2005a)], IREQ [ISO/DIS 11079 (2005c)] and PHS [Predicted Heat Strain, ISO 7933 (2004a)] methods. Although physiologically correct, a single PHS-PMV-IREQ sequence is not well suited for risk assessment. Because it does not take into account the target that can be aimed at in a specific environment, it may easily provide correct but otherwise pointless indications. A non-climatized office can easily be rated as a stressing environment in a sweltering summer afternoon, but the resulting maximum exposure time Dlim is hardly of any use; at the other extreme, any outdoor workplace can have a PMV of order –1, or less, on a pleasant spring day, but with no feasible thermal manipulation, the associated discomfort index is again immaterial. Furthermore, overlap between different methods is often cumbersome and results both in no man’s lands that lie outside the scope of either method and in overlapping areas where two methods disagree (e.g. Holmer, 2004). Finally, classification of a thermal environment as either ‘moderate’ or ‘stressing’ is only possible after the fact, i.e. when measurements or other evaluation techniques have been put into action. A priori identification would instead be desirable, so that an appropriate measurement and analysis strategy can be selected and carried out. The following subsection outlines a framework in this respect. Definitions ‘Unconstrained’ or ‘comfort-prone’ thermal environments: those workplaces where no specific production needs exist that constrain the air temperature and possibly one or more of the other quantities (RH, va, tr, Icl; see Table 1 for symbols) relevant to the human thermal balance. Thermal comfort is a viable target in these environments. Any risk assessment strategy must accordingly be based on a discomfort index, whose numerical value determines both any restrictions on the worker’s exposure and the magnitude and timescale of technical interventions. These environments, hereafter identified as comfort prone, represent the focus of this paper.

‘Office-like’ thermal environments Because the focus of this paper is on thermally unconstrained (i.e. indoor, usually non-industrial) environments, discomfort on the cold side is largely determined by temperature and metabolic activity. Air velocity is less likely to be factor indoor. The two quantities ta and M are then left as the system’s only two free parameters while all other quantities have been locked at the following ‘typical’ values:

tr 5 ta va 5 0.1 m s1 RH 5 75% Icl 5 1 clo (1 clo 5 0.155 m2 K W1).

Constant-PMV and constant-IREQ hypersurfaces degenerate into curves that are drawn in the [ta–M] plane. On the warm side of comfort, the dominant roles are played by temperature, humidity and metabolic activity. Because of the complexity of a full threedimensional analysis, only the low-metabolism limit has been investigated, by restricting the investigation to the range M 1.4 met (1 met 5 58.15 W m2). Larger M values could be handled, but limit curves, particularly in high-discomfort environments, would require some adjustment. The two quantities ta and RH are then left as the system’s only two free parameters while all other quantities have been locked at the following typical values: tr 5 ta va 5 0.1 m s1 Icl 5 0.6 clo. Environments where all these assumptions are simultaneously met are identified as ‘office-like’ in


RISK ASSESSMENT

‘Constrained’ thermal environments: those workplaces where specific production needs exist that constrain the air temperature and one or more of the other relevant quantities (RH, va, tr, Icl). Because of the limitations so imposed on the thermal field, comfort cannot be established as a viable target. The most sensible objective is the elimination (minimization) of thermal stress. Accordingly, any risk assessment strategy must be based on a stress index. ‘Moderate’ thermal environments: those unconstrained thermal environments that display small (moderate) deviations from comfort, which can be handled by the human thermoregulatory system. ‘Vexing’ thermal environments: those unconstrained thermal environments that display large deviations from comfort. These are probably due to layout errors, severe mishandling of thermal issues or lack of maintenance. Extreme conditions, so far from comfort to even pose health risks, cannot be ruled out in principle. These may require extreme solutions, going as far as a work halt.

Thermal comfort assessment in comfort-prone workplaces

545

Table 1. Environmental and subjective quantities involved in PMV, PHS and IREQ and associated limits Quantity

Symbol

Effective range

Unit

PMV

PHS

IREQ °C

ta

10–30

10–50

þ10

tr

10–40

—

—

°C

Difference between mean radiant and air temperature

tr ta

—

0–60

—

°C

Water vapour partial pressure

pa

0–2700

0–4500

—

Pa

Air velocity Metabolic activity

va M

0–1 0.8–4

0–3 1–4.5

0.4–18 —

m s1 met

Clothing thermal resistance

Icl

0–2

0.1–1

.0.5

clo

this paper. While office work is definitely kept an eye on in this paper, this definition by no means excludes other activities that are performed without strenuous physical efforts and with similar clothing, in thermally unconstrained indoor environments. COLD ENVIRONMENTS

Generalities Both PMV (ISO 7730) and IREQ (ISO/DIS 11079) depend on the same set of six quantities; each quantity is listed in Table 1 along with its effective range for method validity. ISO 7730 also sets a second condition which must be fulfilled in order for the PMV method to apply, i.e. PMV must lie within the [2 to þ2] interval. No analogue exists for IREQ. It is very important to note that the standard itself specifies that it applies only to ‘indoor environments where thermal comfort is desirable, but where moderate deviations from thermal comfort occur’ (ISO 7730, Introduction), i.e. to what we define in this paper as moderate thermal environments (Definitions). IREQ is a stress index, and as such, it should not belong in a discussion on comfort-prone thermal environments. Its inclusion is, however, justified by the need to ensure a smooth transition between the low-discomfort (moderate) and the high-discomfort (vexing) thermal environments. Figure 1 shows the following curves in the [ta–M] plane: 1. ta 5 10°C: low end of the effective range for PMV and high end of the effective range for IREQ; 2. PMV 5 0.2: low end of the comfort range for Class A environments of PMV; 3. PMV 5 0.5: low end of the comfort range for Class B environments of PMV; 4. PMV 5 0.7: low end of the comfort range for Class C environments of PMV; 5. PMV 5 1: value associated to ‘slightly cold’ thermal sensation; 6. PMV 5 2: low end of the effective range for application of PMV method;

Fig. 1. Air temperatures as functions of metabolic activity, from PMV and IREQ.

7. IREQneutral 5 1 clo: low end of the durationlimited exposure (DLE) .8 h range, with reference to thermal neutrality; 8. IREQmin 5 1 clo: low end of the DLE .8 h range, with reference to minimum acceptable conditions; 9. DLE 5 120 min: high end of the range requiring immediate action (ISO 15265).

The low-discomfort regime National documents and regulations. Limits on minimum and maximum acceptable temperatures in workplaces have been set in many European and extra-European countries by such diverse institutions as technical associations, ministries or labour inspection units. Table A1 shows a list of relevant tasks in the context of office-like environments, implied metabolic activities, associated minimum acceptable temperatures and references to the original documents. Data are also presented in Fig. 2. Metabolic activities, which are not provided in the original documents (apart from the Belgian regulation, NR2,


Air temperature Mean radiant temperature

546


1993), have been associated to work tasks using Tables A1, B2 and B3 from ISO 8996 (2005b), and Table B1 from ISO 7730. Despite their different nature, cultural heritage and juridical relevance, these documents all share the use of air temperature and metabolic activity as the only two relevant quantities. The choice of the [ta–M] plane as the ideal ground to compare predictions from different sources and establishing a new classification scheme (see ‘Office-Like’ Thermal Environments) is accordingly further strengthened. Most documents include minimum temperatures at M 5 1.2 and M 5 1.7 met, which can be mutually compared to explore any sensitivity to economic, geographical or meteorological indicators of associated countries. Data show no any clear trend of this kind, possibly due to the strong homogeneity in the sample, in both respects. There is, however, a tendency of older documents [see, e.g. documents from Italy (NR7 1987) and Spain (NR8 1997), both more than 10 years old] to display lower values. This likely reflects the growing attention that is being paid to thermal issues at the workplace. Constant-discomfort limit curves. Office-like environments do not require special thermal needs. As such, they should be good candidates for Class C environments according to the recent classification scheme included in ISO 7730. This in turn suggests that PMV 5 0.7 might be the lowest discomfort limit curve. This choice, however, would conflict with the minimum allowable temperatures inside offices mandated by the national legislations and regulations discussed in National Documents and Regulations. Figure 3 displays the three curves with PMV 5 0.2, PMV 5 0.5 and PMV 5 0.7; the mean and standard uncertainty on the mean of data from Table A1; and the linear best fit to such data in the 1 M 2 met range

Fig. 3. Low-discomfort cold thermal environments.

tN 5 24:1 4:3 M:

ð1Þ

The air temperature associated to PMV 5 0.7 is always below legislation-mandated values (see Fig. 3), signalling the need for an additional limit curve (above PMV 5 0.7) on the low-discomfort side. However, moving up to PMV 5 0.5 does not provide substantial improvement: implied air temperatures still drop below legislation-mandated values if M . 1.3 met, which is rather common in office-like environments. It is PMV 5 0.2 that provides the most satisfactory answer in this sense because of its recognition in ISO 7730 as a comfort limit for high-quality environments (Class A). This choice may be criticized as it appears to pose an excessive thermal/economic burden to be enforced when M is very low. It must be kept in mind, however, that low-M activities are usually associated to cognitive tasks in offices (not so much in other office-like activities). These activities would certainly benefit from a high-quality thermal environment, which justifies the use of PMV 5 0.2. In the high-M limit (M 2 met), PMV 5 0.2 nicely matches values imposed by national regulations, reconciliating the two requirements as expected. The high-discomfort regime At first sight, PMV 5 2 would seem to be the most natural limit curve in this region. This value is also selected in ISO 15265 to represent the low end of the discomfort area (see its Table 7). Several elements must, however, be taken into account before coming to a conclusion. Whenever M 1.4 met, the PMV 5 2 curve lies below the ta 5 10°C boundary of the recommended range for application of the PMV method. Although this may be seen as just a minor formal violation, a lower reliability of PMV must be recognized anyway. A more worrisome element is that the PMV 5 2 curve lies 0.6 met to the left of the IREQmin 5 1 clo


Fig. 2. Minimum acceptable temperatures from national documents (see Table A1).


There are environments characterized by the simultaneous occurrence of PMV ,2 and IREQmin Icl IREQneutral, labelled as ‘constraint in the long term’; despite their claimed existence, they cannot really exist, as these two conditions are mutually exclusive. There are no environments characterized by Icl , IREQmin and DLE . 120 min; they are instead very much real (see Fig. 1). Environments with PMV . 2 do not require constraints on any timescale; on the opposite, they should be the target of substantial action, given the short exposures allowed by IREQ. A satisfactory classification scheme in cold environments can be established under the hypothesis that the concept of comfort itself becomes meaningless in areas where DLE , 480 min. In other words, the onset of DLE must signal the need for immediate action in comfort-prone environments. In practice, this leads to the selection IREQmin 5 1 clo for the last (most extreme) limit curve in a cold, comfort-prone thermal environment. This curve is well extrapolated to temperatures .10°C by the PMV 5 1 curve (see Fig. 1). So the limit curve is actually a combination of IREQmin 5 1 clo (at ta 10°C) and PMV 5 1 (at ta . 10°C). In practice, since transition takes place at M 2 met, PMV 5 1 will always be the relevant limit. Classification scheme of cold, comfort-prone environments Table 2 summarizes the classification scheme for cold thermal environments which has emerged in The Low-Discomfort Regime and The HighDiscomfort Regime under Cold Environments. Figure 4 provides a graphical synthesis of the same results. PMV is perfectly adequate to quantify discomfort in Areas A, B and C. However, IREQ is a more reasonable choice in Area D, where it provides a better approximation of the complex physiological reaction of human body to cold environments (Holmer, 2004), the more so as thermal conditions become colder. WARM THERMAL ENVIRONMENTS

Generalities ISO 7730, ISO 7933 and ISO 7243 (1989) are the relevant international standards on thermal comfort and stress in warm and hot environments. Methods advocated are PMV, PHS and WBGT (Wet-Bulb Globe Temperature), respectively. The first two such indexes depend on the same set of six quantities; each quantity is listed in Table 1 along with its effective range for method validity. WBGT is an empirical index which is still commonly used for risk assessment in hot thermal environments, mostly because


curve at ta 5 10°C (see Fig. 1). With a gradient of 10°C met1, that is equivalent to 6°C less, for the same metabolic activity. Because the two curves share a similar gradient, the 6°C distance roughly applies to any value of M in the 1–2 met range. Indeed, the PMV 5 2 curve is even located well to the left (i.e. at lower temperatures) of the more extreme curve DLE 5 120 min (see Fig. 1), which is identified in ISO 15265 as the upper limit of the area of immediate action. In more general terms, there appears to be a lingering paradox of non-negligible health risks (as signalled by DLE , 480 min), coexisting with medium discomfort (intermediate values of the predicted percentage of dissatisfied [PPD]). Example: the same environment (ta 5 tr 5 9.6°C, RH 5 75%, va 5 0.1 m s1, M 5 1.8 met, Icl 5 1 clo) which determines relatively small deviations from comfort (PMV 5 1.39 or PPD 5 45%) has DLE 5 240 min. Some inconsistencies between PMV and IREQ may be real as they can be traced to the different equations used in calculating thermal balance by the two methods. A recent in-depth investigation by Holmer (2004) has shown that for thermal neutral conditions, PMV predicts higher temperatures compared to IREQ, everything else being equal. Figure 1 shows that IREQneutral 5 1 clo lies very close to PMV 5 0.5 (the best match is actually with PMV 5 0.6). This in turn implies that PMV 5 0 lies well to the right of IREQneutral 5 1 clo, i.e. at warmer temperatures. Holmer’s (2004) Fig. 2 and our Fig. 1 agree that this shift is 3–4°C at M 5 2 met. The conflict in the regime far from comfort, however, does not lend itself to a simple explanation on energetic grounds. Existing differences in treatments of heat transfer and evaporative heat loss are too small. Indeed, it is not at all obvious that a mathematical or objective inconsistency exists in the first place. Coexistence of a subjective judgement of medium discomfort with a DLE , 480 min might well be real, as much as discomfort due to exposure to whole-body vibration coexists with risks of mechanical damage to the lumbar spine. That, however, does not at all mean that a conflict which is more of semantic than physiological in nature should be ignored. Occupational hygienists are accustomed to thinking that comfort and health risk areas do not overlap in thermal environments. This is indeed the case on the warm side of comfort, where a ‘grey’ or transition area is interposed between the two (see The High-Discomfort Regime under Warm Thermal Environments). On the contrary, this is not at all true on the cold side, where overlap is strong. This should be made clear. The current scheme included in ISO 15265 is inadequate for risk assessment and management of office-like environments, as it propagates three false beliefs.

547

548


Table 2. Proposed classification scheme of cold, comfort-prone environments Area Range

Discomfort

Action timescale

A

PMV , 0, ta tN

B

ta , tN, PMV –0.7 Low

Long term

C

1 PMV , –0.7

Intermediate

Short term

D

PMV , –1, IREQmin . 1 clo

High or very high Immediate

Very low

None

Figure 5 shows the following curves in the [ta–RH] plane, assuming M 5 1.2 met:

Fig. 4. Proposed classification scheme in cold, comfort-prone environments.

of its conceptual simplicity, along the lines laid out by ISO 7243. The calculation of WBGT relies on the assumption of light clothing (Icl 5 0.6); in addition, it follows different paths depending on the presence or absence of direct sunlight; limits also depend on the presence/absence of appreciable ventilation. Because we focus here on office-like thermal environments (see ‘Office-Like’ Thermal Environments), conditions appropriate to shaded, non-ventilated environments apply. WBGT’s strongest functional dependence is on the natural wet-bulb temperature tnw. This quantity is entirely specific to this index and has no unique direct relation to air temperature and humidity. In order to make predictions from PMV, PHS and WBGT comparable, the following empirical relations (del Gaudio and Lenzuni 2002) specific for use in lowventilation environments tNW 5 ta ð13 0:13 RHÞ

ð2aÞ

tNW 5 ta ð350=RHÞ

ð2bÞ

and

have been used for high (.50%) and low (,50%) values of RH, respectively. A smoothing function has also been used to match equations (2a) and (2b) in the transition region RH 5 35–65%.

1. ta 5 30°C: high end of the effective range for PMV; 2. pa 5 2700 Pa: high end of the effective range for PMV; 3. PMV 5 0.7: high end of the comfort range for Class C environments of PMV; 4. PMV 5 2: high end of the effective range for application of PMV method; 5. Dlim 5 480 min: high end of the range where no restriction on exposure is imposed according to PHS method; 6. WBGT 5 33°C: high end of the acceptable range for low-M activities. The low-discomfort regime National regulations. Although trailing its counterpart at cool temperatures, the concept of a maximum acceptable temperature in the workplace has also found its way in a few national documents, as shown in Table A2. Despite being widely recognized by the technical literature dealing with comfort in warm climates, humidity is not taken into account in these documents, which makes the quoted values hardly comparable with the outcomes of technical standards. There is some consideration of metabolic activity though, so that data appropriate for low metabolic activity (M 5 1.2 met) can be selected. The mean value of 28.1°C [from NR1 (2002), NR2, NR3 (2004), NR6 (2001a), NR8, NR9 (2006), NR10 (1983), NR11 (1986)] has been used in conjunction with RH 5 70% (commonly quoted as the maximum acceptable value for this latter quantity by itself). When plotted on top of the curves previously listed (Fig. 5), it shows up in close proximity of the PMV 5 1.3 curve, hinting at a possible identification of the latter as a transition from moderate to vexing thermal environments.


Fig. 5. Air temperatures as functions of relative humidity, from PMV, PHS and WBGT.


549

Constant-discomfort limit curves. In analogy with cold environments, warm environments also suffer from a formal conflict between the two prerequisites for use of PMV. In detail, the PMV 5 2 curve is positioned at temperatures .30°C and as such it is entirely outside the effective range determining the validity of the PMV method (see Table 1). This in itself, however, is not a strong enough argument to delegitimate PMV 5 2 as a viable limit curve.

1. It has a trapezoidal shape: on the low-discomfort side PMV 5 2 is a very steep curve with a humidity gradient of 30% per °C. On the high-discomfort side, Dlim 5 480 min is a very shallow curve, with a mean humidity gradient of 6.75% per °C. Cooling of humans takes place more and more through sweating/evaporation as temperature increases; since the effectiveness of this mechanism decreases as humidity increases, this determines a clear trend towards shallower slopes as one proceeds to areas of more acute discomfort. 2. It is very wide, spanning 6.5°C at RH 5 50% and as much as 9°C at RH 5 30%. When projected against the background of human sensitivity to heat stress, this proves way too large to be accommodated within a single discomfort area and motivates the search for one or possibly more limit curves to be interposed between the two borders. To this aim, investigation of PMV and PHS inside the grey area has been performed. Extrapolation of PMV to warmer environments has been carried out by calculations of three curves with PMV 5 2.5, 3 and 3.5. Figure 6 shows no evidence of a smooth transition to PHS, since highPMV curves keep the same steep gradient as their low-PMV counterparts. As for constant-stress curves, the limit on water loss is more stringent than the limit on the core temperature in the low-stress regime. The latter has been ignored, and extrapolation of the PHS method to colder environments has been performed by synthesizing curves of constant water loss with fractions fw 5 2/3, 1/2 and 1/3 of the water-loss limit Dmax95 5 5% of body mass (ISO 7933). This is equivalent to allowing no more than 2500, 1875 and 1250 g day1, respectively, for an average 75-kg subject. Water loss is a credible descriptor of discomfort, as it quantifies the effort that the human body must produce to maintain an acceptable core temperature. Figure 6 shows that the curve with fw 5 1/3 repre-

Fig. 6. Constant-PMV and constant-water-loss curves in the transition region for low-Menvironments.

sents a good match to PMV 5 2. Therefore, the range fw 5 1/3–1 spans the entire grey area identified above, providing the desired smooth transition from steep in the low-discomfort limit to shallow in the high-discomfort limit. The role of WBGT can be ignored: while WBGT 5 33°C used to provide a more conservative limit compared to the old SWeat required (SWreq) index (Malchaire et al., 2001), this is no longer the case when compared to PHS. Actually, the WBGT limit curve overlaps quite well with the Dlim 5 480 min exposure limit. This eliminates the often quoted role of WBGT as a ground breaker to be used for preliminary risk assessment, possibly to be followed by an analytic assessment through PHS. Classification scheme of warm, comfort-prone thermal environments Table 3 summarizes the classification scheme for warm thermal environments which has emerged in The Low-Discomfort Regime and The HighDiscomfort Regime under Warm Thermal Environments, providing the mathematical expression of limit curves, the descriptor of election and the recommended timescale for action. Figure 7 provides a graphical synthesis of the same results for quick reference. A six-zone scheme has been deemed appropriate, implying the existence of five threshold curves: 1. The same arguments supporting the choice of PMV 5 0.7 in cold environments, lead naturally to the selection of PMV 5 0.7 as the upper border of the comfort area. No need for an additional curve at even lower discomfort values has emerged. 2. PMV 5 1.3 is the curve that comes closest to the few available data on the maximum allowable temperature in offices, indicated by national documents. This curve (which incidentally is associated to PPD 40%) has been selected to represent the transition from moderate to vexing thermal environments.


The high-discomfort regime The picture on the warm side of comfort is dominated by the presence of a huge ‘grey area’ shown in Fig. 5 between PMV 5 2 on the low-temperature side and Dlim 5 480 min on the high-temperature side. The grey area has two most relevant properties.

550


Table 3. Proposed classification scheme of warm, comfort-prone environments Area

Range

Discomfort

Action timescale

A

0 , PMV 0.7

Very low

None

B

0.7 , PMV 1.3

Low

Long term

C

1.3 , PMV 2

Intermediate

Intermediate

D

PMV . 2, fw 1/2

High

Short term

E

1/2 , fw 1 [ Dlim 480 min

Very high

Very short term þ work halt

Dlim , 480 min

Extreme

Immediate þ work halt

F

CONCLUSIONS

A priori discrimination of thermal environments where comfort is a reasonable target (here termed ‘comfort-prone’ environments) from those where it is not is a fundamental prerequisite for thermal risk assessment. Classification according to the viability of comfort allows the most cost-effective use of the results of the comfort assessment procedure. In the context of comfort-prone environments, a classifier has been developed which mostly relies on information derived from standardized indexes (PMV, IREQ and PHS) with some support from national technical documents/legislation, particularly in cold environments. The proposed scheme eliminates a few mistakes of ISO 15265, which includes one class that does not exist (at least for low values of metabolic activity) while ignoring one which should instead be included. More in general, this document is poorly suited for comfort assessment because of its one-sequence approach extending from extremely hot to extremely cold environments. As a consequence of an in-depth investigation of the high-discomfort regime, the proposed classification scheme identifies a larger number of comfort areas than ISO 15265 (six instead of four) in warm and hot environments. This allows a better identification of high-priority cases. Constant-water-loss curves have been identified as a reliable tracer of constant discomfort inside the grey area that lies between moderate (PMV , 2) and health threatening (Dlim , 480 min) environments. Each comfort area comes with limit curves, a recommended descriptor and an action timescale, for an optimization of resource allocation.

Table A1. Minimum workplace acceptable temperatures Definition of activity

Austria

Intermediate work

1.7

18

Light work

1.2

19

Belgium Light work Very light work Canada Intermediate (Quebec) work—standing

Fig. 7. Proposed classification scheme in warm, low-M, comfort-prone environments.

Reference M ta (met) (°C)

Country

1.7a 18 1.0a 20 20

16

Light work standing

1.7

17

Light work sitting

1.4

19

Light work sitting, especially mental work or work involving reading or writing

1.2

20

NR1 NR2 NR3


3. PMV 5 2 provides the last curve within the range of applicability of the PMV method. 4. Moving further up the discomfort scale, water loss becomes the best tracer of discomfort. Because it is located roughly at the centre of the grey area, the curve with fw 5 1/2, i.e. a water loss equal to 2.5% of the body mass (1875 g day1 for a standard body, ISO 7933), has been picked as the limit curve for very strong discomfort. Should this scheme be applied to higher values of M, a higher water loss should be selected. Example: for M 5 1.6 met, the most useful choice is fw 5 2/3 (2500 g day1). 5. Similarly to what happens in cold environments, the onset of DLE must signal the need for immediate action in comfort-prone environments, and as such, it represents the natural choice for the boundary of the last (most extreme) discomfort curve. This implies the selection of the curve with constant water loss equal to 5% of the body mass (3750 g day1). This curve should be used, in the context of classification of comfort-prone environments, as a mere mathematically convenient expression, deprived of its original meaning of an exposure limit for day-long shifts.

Thermal comfort assessment in comfort-prone workplaces Table A1. Continued Country

Reference M ta (met) (°C)

Definition of activity

1.7

15

Sedentary or static work with slight physical exertion

1.2

18

1.7

18

2.0

17

Light work standing and/or moving

1.7

19

Intermediate work sitting

1.7

19

France

Light physical exertion Germany Intermediate work standing and/or moving

Italy

Spain

UK

Light work sitting

1.2

20

Intermediate work without frequent movements

2.0

13

Light work with frequent movements

1.8

14

Light work without frequent movements

1.7

15

Precision work Conversation, reading and studying

1.4 1.2

17 18

Light work

1.2

17

Sedentary work typical of offices

1.7

14

Light work in factories

1.7

16

Classrooms (teachers)

1.6

18

Offices and laboratories

1.2

20

NR4 (2005)

NR5 (2000) NR6

NR7

NR8

NR9

a

Provided by the original document.

Table A2. Maximum acceptable temperatures for activities with M51.2 met Country

ta (°C)

Reference

Australia Austria

30 25

NR11 NR1

Belgium

30

NR2

Canada

26.8a

NR12 (2001b)

France

30

NR10

Germany

26

NR6

Spain

27

NR8

UK

30

NR9

a

Mean value from four Canadian provinces.

REFERENCES del Gaudio M, Lenzuni P. (2002) Esposizione a stress da alte temperature dei lavoratori delle cave di marmo di Massa e Carrara [in Italian]. In Nicolini O, Nataletti P, Peretti A et al., editors. Proceedings of dBA 2002, Artestampa Modena. pp. 91–7. Holmer I. (2004) Cold but comfortable? Application of comfort criteria to cold environments. Indoor Air; 14: 27–31. ISO. (1989) ISO 7243: 1989: Hot environments—estimation of the heat stress on working man, based on the WBGT-index

(wet bulb globe temperature). Geneva: International Organization for Standardization. ISO. (2004a) ISO 7933: 2004: Ergonomics of the thermal environment—analytical determination and interpretation of heat stress using calculation of the predicted heat strain. Geneva: International Organization for Standardization. ISO. (2004b) ISO 15265: 2004: Ergonomics of the thermal environment—risk assessment strategy for the prevention of stress or discomfort in thermal working conditions. Geneva: International Organization for Standardization. ISO. (2005a) ISO 7730: 2005: Ergonomics of the thermal environment—analytical determination and interpretation of thermal comfort using calculation of the PMV and PPD indices and local thermal comfort criteria. Geneva: International Organization for Standardization. ISO. (2005b) ISO 8996: 2005: Ergonomics of the thermal environment—determination of metabolic rate. Geneva: International Organization for Standardization. ISO. (2005c) ISO/DIS 11079: 2005: Ergonomics of the thermal environment—determination and interpretation of cold stress when using required clothing insulation (IREQ) and local cooling effects. Geneva: International Organization for Standardization. Malchaire J, Piette A, Kampmann B et al. (2001) Development and validation of the predicted heat strain model. Ann Occup Hyg; 45: 123–35. National Regulations 1 (NR1). (2002) Bundes-Arbeitssta¨ttenverordnung—B-AStV [in German]. Bundesgesetzblatt f}ur die } Republik Osterreich, Ausgegeben am 27, September 2002. National Regulations 2 (NR2). (1993) Algemeen Reglement voor de Arbeidsbescherming (ARAB), Titel II, Hoofdstuk II, Afdeling I: Arbeidsklimaat [in Flamish]. Federale Overheidsdienst Werkgelegenheid, Arbeid en Sociaal Overleg, Brussels. National Regulations 3 (NR3). (2004) Confort thermique a l’interieur d’un etablissement [in French]. Commission de la sante` et de la securite` du travail du Quebec. National Regulations 4 (NR4). (2005) Temperatur i arbejdsrum pa˚ faste arbejdssteder, At-vejledning A.1.12 [in Danish]. Arbeidstilsynet, Copenhagen. National Regulations 5 (NR5). (2000) Conception de lieux du travail [in French]. Institute Nationale de Recherche et de Securite` (INRS), Paris. National Regulations 6 (NR6). (2001a) Arbeitssta¨ttenRichtlinie—Arbsta¨tt 5.006 Raumtemperaturen [in German]. Bundesarbeitsblatt 6-7/2001 S. 94, Bundesministeriums fu¨r Arbeit und Soziales. National Regulations 7 (NR7) UNI 8852. (1987) Impianti per la climatizzazione invernale per gli edifici adibiti ad attivita` industriale ed artigianale. Norme per l’ordinazione, l’offerta ed il collaudo [in Italian]. Ente Nazionale Italiano di Unificazione (UNI), Milano. National Regulations 8 (NR8). (1997) Disposiciones mı´nimas de seguridad y salud en los lugares de trabajo-Anexo III: Condiciones ambientales en los lugares de trabajo [in Spanish]. National Regulations 9 (NR9). (2006) CIBSE guide A: environmental design. Chartered Institution of Building Services Engineers (CIBSE), London. National Regulations 10 (NR10). (1983) Arret Prolonge´ des Installations de Conditionnement d’Air, Recommandation R 226 [in French]. Institute Nationale de Recheche et de Securite` (INRS), Paris. National Regulations 11 (NR11). (1986) Occupational Health, Safety and Welfare Act. South Australian Consolidated Acts, Adelaide, SA. National Regulations 12 (NR12). (2001b) Canadian health and safety regulations with respect to thermal conditions in the workplace. Canadian Centre for Occupational Health and Safety (CCOHS), Hamilton, Ontario. Available from http://www.ccohs.ca/oshanswers/phys_agents/hot_cold.html.


Denmark Work with limited physical exertion

551