Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Evaluation of bioaerosols in elementary school classrooms in a coastal temperate zone Bartlett, Karen Hastings 2000

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2000-486028.pdf [ 8.54MB ]
Metadata
JSON: 831-1.0076831.json
JSON-LD: 831-1.0076831-ld.json
RDF/XML (Pretty): 831-1.0076831-rdf.xml
RDF/JSON: 831-1.0076831-rdf.json
Turtle: 831-1.0076831-turtle.txt
N-Triples: 831-1.0076831-rdf-ntriples.txt
Original Record: 831-1.0076831-source.json
Full Text
831-1.0076831-fulltext.txt
Citation
831-1.0076831.ris

Full Text

EVALUATION OF BIO AEROSOLS IN ELEMENTARY SCHOOL CLASSROOMS IN A COASTAL TEMPERATE ZONE. by Karen Hastings Bartlett B.A., The University of Victoria, 1973 M.Sc, The University of British Columbia, 1994 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Individual Interdisciplinary Studies Graduate Programme [Occupational Hygiene / Health Care and Epidemiology / Microbiology / Respiratory Medicine]) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December 1999 © Karen Hastings Bartlett, 1999 in presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of /AJft/\l(b1//H- IWr~r&>&CtpL/MA^ STiSbie<> 6fy&bU-ftlh t°H6^P^ The University of British Columbia 0CjU)9^V)^H^ f t ^ ' ^ ^ * Vancouver, Canada Date DE-6 (2/88) ABSTRACT Potential determinants of exposure to culturable airborne fungal and bacterial aerosols and carbon dioxide were examined as an aid to the interpretation and evaluation of indoor air quality assessments. Concentration measurments for culturable bioaerosols and CO2 were evaluated against published standards and guidelines. METHOD: All 39 schools from one British Columbia school district were enrolled in the study to ensure different building ages and construction materials, but the same maintenance protocols, were included. Schools were randomly assigned to winter, spring or fall sampling. Data collected included: number of occupants and patterns of occupancy, CO2 levels, temperature and relative humidity, total suspended particles, and air exchange rates using tracer gas (SF6) decay. Other characteristics of the classrooms included the presence or absence of forced air heat, carpets, live animals or aquaria, plants, and the siting of the school or portable classroom. Culturable indoor and outdoor aerosols of fungi and bacteria were collected. Determinants of exposure were modelled by constructing multiple linear regression equations for indoor fungi, indoor bacteria and indoor carbon dioxide. RESULTS: The multiple regression models were able to explain a considerable proportion of the variance for the outcomes of interest (total R2 = 0.59 for mesophilic fungi, 0.61 for bacteria, and 0.68 for CO2). Increased outdoor temperature and outdoor fungal counts were associated with higher concentrations for indoor fungi. Variables describing ventilation and conditions of occupancy were significant to all outcomes of interest but functioned differently in the models. For example, fungal concentration was higher in the presence of natural ventilation, but lower with increased mechanical ventilation. In contrast, CO2 was lower with both ventilation types, and lower with higher outdoor temperature. CONCLUSIONS: Using variables measured during an indoor air quality investigation, predictive models can be constructed which are useful in identifying determinants of bioaerosol and bioeffluent concentrations. Ranges of bioaerosol and bioeffluent concentrations for high occupancy buildings in a coastal temperate zone may differ from guidelines written for other indoor settings and climate zones. TABLE OF CONTENTS Page ABSTRACT ii LIST OF TABLES vii LIST OF FIGURES x ABBREVIATIONS, NOMENCLATURE AND SYMBOLS xi ACKNOWLEDGEMENTS xiii CHAPTER 1. GENERAL INSTROUCTION AND OBJECTIVES 1 1.0 General Introduction 1 1.1 Objectives 2 1.2 Overview of Thesis 3 CHAPTER 2. BACKGROUND AND LITERATURE REVIEW 4 2.1 Why Study Indoor Air Quality 4 2.1.1 Introduction to Building Related Illness (BRI) and Sick Building Syndrome (SBS) 4 2.1.2 The Controversy Surrounding SBS 6 2.1.3 Introduction to Indoor Air Quality Investigation 7 2.1.4 Introduction to Guidelines for Evaluation of Indoor Air Quality 8 2.2 Components of Indoor Air Quality Investigations 11 2.2.1 Introduction to Bioaerosols 11 2.2.1.1 The Role of Indoor Bioaerosols to Disease 11 2.2.1.2 Factors Contributing to Indoor Bioaerosol Concentrations 14 2.2.1.3 Factors Influencing the Evaluation of Indoor Bioaersol Cone. 17 2.2.2 Ventilation, CO2, Comfort Parameters and Particulate Matter 18 2.2.2.1 Indoor Air Quality and Mechanical Ventilation 18 2.2.2.2 Relationship Between CO2 and Ventilation 18 2.2.2.3 Investigation of the Ventilation System 21 2.2.2.4 Standards and Evaluation of Ventilation 21 2.2.2.5 Comfort Parameters (Temperature and Relative Humidity) 22 2.2.2.6 Particulate Matter 23 2.3 Introduction to Determinants of Exposure 26 CHAPTER 3. SITE SELECTION AND CHARACTERIZATION 28 3.0 Methods 28 3.1 Study Site Characterization 28 3.2 Classroom Sampling and Monitoring Schedule 29 3.3 Sampling Strategy 30 3.4 Power Estimates and Statistical Analyses 32 3.5 Results 36 iii 3.5.1 Descriptive Statistics of Sites 36 3.6 Summary of Site Characterization 39 CHAPTER 4. VENTILATION SYSTEMS, AIR EXCHANGE RATES, AND COMFORT PARAMETERS 41 4.1 Introduction to Methods for Measuring or Estimating Ventilation Rates 41 4.1.1 Air Exchange Rate by Tracer Gas Decay (SF^) 41 4.2 Methods for Measurement of Ventilation Parameters 42 4.2.1 Methods for Ventilation Assessment 42 4.2.2 Measurement of Air Exchange Rate by Tracer Gas Decay SF6 43 4.2.3 Measurement of Carbon Dioxide (CO2) Concentration 45 4.3 Methods for Measurement of Comfort Parameters 46 4.3.1 Temperature and Relative Humidity 46 4.3.2 Equilibrium Relative Humidity 46 4.3.3 Total Suspended Particles 47 4.4 Ventilation Variables and Data Transformation for Statistical Analyses 49 4.5 Results 51 4.5.1 Air Exchange Rates, Ventilation, Carbon Dioxide (CO2) and Comfort Parameters 51 4.5.2 Assessment of Mechanical Ventilation Systems 56 4.5.3 Relative Humidity and Equilibrium Relative Humidity 56 4.5.4 Temperature 57 4.5.5 Total Suspended Particles (TSP) 57 4.6 Summary of Comfort Parameters 57 CHAPTER 5. BIOAEROSOLS 59 5.1 Bioaerosols 59 5.1.1 Fungal Aerosols 59 5.1.1.1 Mesophilic Fungi 59 5.1.1.2 Thermotolerant Fungi 59 5.1.1.3 Xerophilic Fungi 59 5.1.2 Bacterial Aerosols 60 5.1.2.1 Mesophilic Bacteria 60 5.1.2.2 Thermophilic Bacteria 60 5.2 Methods 60 5.2.1 Fungal Aerosols 60 5.2.1.1 Mesophilic Fungi 60 5.2.1.2 Thermotolerant Fungi 62 5.2.1.3 Xerophilic Fungi 62 5.2.2 Bacterial Aerosols 63 5.2.2.1 Mesophilic Bacteria 63 5.2.2.2 Thermophilic Bacteria 64 5.2.3 Statistical analysis 65 5.3 Results 69 5.3.1 Fungal Aerosols 69 5.3.1.1 Mesophilic Fungal Aerosols 69 5.3.1.2 Thermotolerant Fungal Aerosols 72 IV 5.3.1.3 Xerophilic Fungal Aerosols 73 5.3.2 Bacterial Aerosols 75 5.3.2.1 Mesophilic Bacterial Aerosols 75 5.3.2.2 Thermophilic Bacterial Aerosols 76 5.4 Summary of Bioaerosols 76 CHAPTER 6. COMPARISON OF COMFORT PARAMETERS, VENTILATION AND BIOAEROSOL CONCENTRATIONS TO AVAILABLE GUIDELINES AND STANDARDS 77 6.1 Comfort Parameters 77 6.1.1 Relative Humidity 77 6.1.2 Temperature 77 6.2 Occupant Generated CO2 and Ventilation 78 6.3 Evaluation of Bioaerosols 81 6.3.1 Evaluation of Mesophilic Fungal Aerosols 81 6.3.2 Evaluation of Mesophilic Bacterial Aerosols 85 CHAPTER 7. RELATIONSHIPS BETWEEN INDOOR AIR QUALITY PARAMETERS 87 7.1 Methods 87 7.2 Results 88 7.2.1 Outcome Variable: Indoor Mesophilic Fungal Concentration (CFU/m3) 88 7.2.2 Xerophilic and Thermotolerant Fungi 93 7.2.3 Outcome Variable: Indoor Mesophilic Bacterial Concentration (CFU/m3) 94 7.2.4 Outcome Variable: CO2 Concentration (ppm) 98 7.3 Summary 103 CHAPTER 8. DISCUSSION 105 8.1 Non-compliant Rooms 106 8.1.1 C0 2 Concentration 106 8.1.2 Temperature and Relative Humidity 107 8.1.3 Fungal Concentration 107 8.1.4 Bacterial Concentration 110 8.2 Strengths of Study 111 8.2.1 Site Selection 111 8.2.2 Ventilation Systems, Air Exchange Rates and Comfort Parameters 111 8.2.3 Carbon Dioxide 112 8.2.4 Bioaerosols 112 8.3 Limitations of Study 113 8.3.1 Site Selection 113 8.3.2 Air Exchange Rates 114 8.3.3 Equilibrium Relative Humidity 115 8.3.4 Carbon Dioxide Measurement 115 8.3.5 Total Suspended Particles 116 8.3.6 Fungal Aerosols 116 8.3.7 Bacterial Aerosols 117 8.3.8 Generalization of the Data 118 8.3.8 Generalization of the Data 118 8.4 Contribution to the Literature 118 CHAPTER 9. CONCLUSIONS AND RECOMMENDATIONS 120 9.1 Conclusions from this Study 120 9.2 Recommendations for Further Investigation 121 9.3 Conclusions 123 LITERATURE CITED 124 APPENDIX I. FIELD DATA RECORDING FORMS 140 vi LIST OF TABLES Table Page 1. Definitions of size selective deposition of particulate matter 24 2. Summary of sampling schedule of schools over six sampling periods 29 3. Site specific, independent environmental variables 33 4. Variables describing fixed characteristics of buildings and sites 34 5. Room variables measured or observed on test day 35 6. Seasonal environmental characteristics of study periods 36 7. Description of fixed characteristics of classrooms by building age 37 8. Characteristics of room on test day by season (categorical variables) 38 9. Characteristics of room on test day by season (continuous variables) 39 10. Desirable characteristics of tracer gas for IAQ testing 41 11. Variables describing ventilation and room use 49 12. Carbon dioxide and comfort parameter variables 50 13. Seasonal distribution of ventilation, comfort parameters, and indoor CO2 55 14. Ventilation and comfort parameters by grouped building age or temporary building status. 55 15. Bacterial grouping by Gram stain and morphology 64 16. Fungal bioaerosol variables used for analysis 66 17. Variables used to describe fungal genera or groups for analysis 67 18. Bacterial bioaerosol variables used for analysis 68 19. Variables used to describe group morphology of bacterial counts 68 20. Ranking of mesophilic indoor and outdoor fungal groups 69 21. Indoor/Outdoor ratios of fungal counts (CFU/m3) grouped by season 72 22. Indoor and outdoor fungi isolated from elevated temperature incubation 73 23. Indoor and outdoor bacterial microflora 75 24. Number of rooms failing to meet RH guidelines by season 77 25. Number of rooms not conforming to temperature guidelines by season 78 26. CO2 concentrations exceeding recommended limits by season 79 27. Fungal counts found to be unaceptably high by selected criteria 82 28. Geometric means of mesophilic fungal counts and associated upper 95% confidence limits 84 29. Geometric confidence limits for indoor thermotolerant and xerophilic fungal means 85 30. 95% GUCL for bacterial concentration 86 31. Continuous site and ventilation characteristics related to indoor mesophilic fungal concentration 89 32. Categorical site and ventilation characteristics related to indoor mesophilic fungal concentration 89 33. Continuous environmental and occupancy variables significantly related to indoor mesophilic fungal concentration 90 34. Categorical environmental and occupancy variables significantly related to indoor mesophilic fungal concentration 90 35. Predictors of indoor mesophilic fungal concentration (In CFU/m ) including the independent variable, CO2 concentration 91 36. Predictors of indoor mesophilic fungal concentration (In CFU/m3) including the independent variable, mesophilic bacterial concentration 92 37. Predictors of indoor xerophilic fungal concentration (In CFU/m ) 94 38. Continuous site and ventilation characteristics related to indoor mesophilic bacterial concentration 95 39. Categorical site and ventilation characteristics related to indoor mesophilic bacterial concentration 95 40. Continuous environmental variables significantly related to indoor mesophilic bacterial concentration 96 41. Categorical environmental variables significantly related to indoor mesophilic bacterial concentration 96 42. Continuous occupancy variables significantly related to indoor mesophilic bacterial concentration 96 43. Predictors of indoor mesophilic bacterial concentration (In CFU/m ) 97 44. Continuous ventilation variables with statistically significant relationships to CO2 concentration 98 45. Categorical ventilation variables with statistically significant relationships to CO2 concentration 99 46. Continuous environmental variables with statistically significant relationships to CO2 concentration 99 viii 47. Categorical environmental variables with statistically significant relationships to CO2 concentration 100 48. Continuous occupancy variables with statistically significant relationships to CO2 concentration 100 49. Predictors of indoor CO2 concentration 102 ix LIST OF FIGURES Figure Page 1. Criteria for room selection within schools 31 2. Example of CO2 concentrations in room with radiant hot water heat, but no supply air 52 3. Example of CO2 concentrations in room with radiant hot water heat and mechanically supplied air 52 4. Example of CO2 concentrations in room equipped with forced supply air 53 5. Comparison of two measurements of air exchange rates in naturally vs. mechanically ventilated rooms. 54 6. Seasonal variation of selected indoor isolates of mesophilic fungal groups 71 7. Seasonal variation of selected outdoor mesophilic fungal groups 71 8. Seasonal variation of indoor xerophilic fungal groups 74 9. Percentage of classrooms meeting the ventilation rate criterion for rooms used for educational purposes 79 10. Air exchange rates determined by SF6 tracer gas decay measurements under static room conditions 80 ABBREVIATIONS, NOMENCLATURE AND SYMBOLS ACGIH ACH aerosol AHU AIHA ANSI ASHRAE atopy aw bioaerosols CEPA I2 cfm CFU C0 2 commensal bacteria geo. mean (GM) Gram stain GSD HVAC hygiene IAQ K N O 3 LOD LOQ American Conference of Governmental Industrial Hygienists Air changes per hour Airborne solid or liquid substance Air handling unit American Industrial Hygiene Association American National Standards Institute American Society of Heating, Refrigeration and Air-Conditioning Engineers exaggerated immunological reaction to antigens water activity biologically derived particulate matter Canadian Environmental Protection Act chi-square statistic cubic feet per minute Colony Forming Units Carbon dioxide saprophytic bacterial species associated with specific ecological niches (e.g. skin microflora) geometric mean differential stain dependent on bacterial cell wall structures geometric standard deviation Heating, ventilation and air-conditioning Study and promotion of health (from Hygeia, Greek goddess of health) Indoor Air Quality Potassium nitrate Limit of Detection Limit of Quantitation In natural log (base e) L/min litres per minute m cubic metre MEA Malt Extract Agar mesophilic bacteria bacteria with optimum growth at 35 - 37°C mesophilic fungi fungi with optimum growth at room temperature MgCh Magnesium chloride Mg(NC>3)2 Magnesium nitrate NaCI Sodium chloride NIOSH National Institute for Occupational Safety and Health olf emission rate of bioeffluents from a standard person Pa Pascal, a measurement of pressure = Newton/m2 ppm parts per million RH Relative Humidity SBS Sick Building Syndrome sd Standard Deviation SF6 Sulphur hexafluoride thermohygrometer temperature compensated relative humidity meter thermotolerant fungi fungi capable of growth at 37°C TSA Trypticase Soy Agar TSP Total Suspended Particles U.S. EPA United States Environmental Protection Agency VOC's Volatile Organic Compounds WCB Workers' Compensation Board (BC) w/v Weight of substance per volume solute xerophilic fungi fungi capable of growth on low water activity (aw) media YES Young Environmental System CO2, temperature and RH data logging monitor ACKNOWLEDGEMENTS I wish to sincerely thank the members of my committee for their enthusiastic and helpful support of this project, especially my supervisor, Dr. Susan Kennedy. I also wish to thank the British Columbia Health Research Foundation for the studentship awards (ST#10[98], ST#13[97], ST#6[96], and ST#23[95]) that made this study possible. I thank Victor Leung, lab manager for the Occupational Hygiene Programme, for his technical help with the equipment used for this study. Thanks are due to the teachers, students and staff of the participating school district for their interest and cooperation in this project. And finally, I wish to thank H.L. Hastings with all my heart for his moral support. khb Xlll CHAPTER 1 GENERAL INTRODUCTION AND OBJECTIVES 1.0 GENERAL INTRODUCTION There is a great deal of public attention focused on indoor air quality issues. The study of indoor air quality is complicated by the complexity of potential sources of indoor air pollution and limitations of measurement methodologies. Architects, building engineers and occupational hygienists have been challenged to address complaints by workers and members of the public who associate symptoms of ill health with the buildings where they work or live. The building itself becomes suspect as a contributing factor, in addition to ventilation, furnishings, and exogenous biological or chemical contaminants. Public pressure to provide answers to and relief from indoor air quality complaints has contributed to attention being focused on contaminants with which only a few researchers have extensive experience. In some cases there are unique challenges in developing standardized protocols for indoor air quality investigations due to limitations of measurement tools. As a result, adequate databases have not yet been developed to provide the information necessary to set exposure limits. Accurate measurement of exposure is requisite to being able to study putative health effects resulting from exposures to indoor air contaminants. The purpose of this study was to systematically examine a series of public buildings which were not previously identified as problem buildings. The intention was to determine building, site, environmental and use characteristics that could be used to establish expected ranges of selected indoor air quality data (fungal and bacterial aerosols, and CO2) as an aid in the interpretation of results from complaint buildings. Elementary schools were chosen to examine because parents and teachers have expressed concerns regarding possible links between indoor air and ill health. Although teachers are covered by provincial workers' compensation legislation, provision for evaluating indoor air quality as a specific complaint did not become part of the provincial health and safety regulations until 1998 (WCB 1998). The responsibility 1 for healthy schools is neither a public health mandate, which would be governed by the Ministry of Health, nor an education mandate which would be funded by the Ministry of Education. As a result, local school boards are bearing the responsibility for the investigation and resolution of indoor air quality complaints. At the present time there are no standardized protocols for school boards to use to guide indoor air quality investigations, nor guidance in the interpretation of the findings. This has resulted in differing standards for resolution of complaints. 1.1 OBJECTIVES The primary objectives of the study were as follows: • to describe the range of indoor environment characteristics encountered in public school buildings, not previously identified as problem buildings, in a coastal, temperate region; • to determine the relationships of • building (age, construction materials, furnishings, maintenance, cleanliness) • site (elevation, relationship to major arterial roads or semi-agricultural areas) • ventilation (type of ventilation and ventilation efficiency, presence of sound-absorbing or thermal insulation materials) • environmental (outdoor temperature, relative humidity, precipitation, wind speed, barometric pressure and indoor temperature, relative humidity, equilibrium relative humidity) • and use characteristics (occupancy numbers, demographics, patterns of use) to airborne fungal and bacterial concentrations found in these buildings; • and to compare these bioaerosol concentrations and ventilation efficiencies to published guidelines, where guidelines are available. A secondary objective of the study was to evaluate carbon dioxide (CO2) concentration as a marker of the contribution of the occupants' bioeffluent to the indoor environment by 2 examining the relationship between the building, ventilation, environmental and use characteristics and CO2. 1.2 OVERVIEW OF THESIS The outcomes of interest in the study were chosen to be culturable bioaerosols (fungi and bacteria) and a marker of bioeffluent, CO2. The background to indoor air quality issues and introduction to selected indoor air quality parameters follows in Chapter 2. The building, site, and environmental factors examined are described in detail in Chapter 3. Measurements and observations of ventilation systems and comfort parameters (relative humidity and temperature) are described in Chapter 4. Fungal and bacterial bioaerosol measurements are summarized in Chapter 5. The comparisons to available guidelines for fungal and bacterial aerosols and ventilation efficiency are presented in Chapter 6. The interrelationships between independent variables are explored and used to create multiple linear regression models in Chapter 7. Overall discussion of the results follows in Chapter 8. Conclusions and recommendations based on the observed measurements and models are summarized in Chapter 9. 3 CHAPTER 2 BACKGROUND AND LITERATURE REVIEW 2.1 WHY STUDY INDOOR AIR QUALITY? Indoor air quality issues have caught the attention of the public and are perceived to be an important health issue of the 1990's. In the United States, the National Institute for Occupational Safety and Health (NIOSH) carried out some 500 indoor air quality investigations between 1971 and 1988. In 1990 more than 2000 requests were received by NIOSH for health hazard investigations involving indoor air quality complaints in non-industrial sites (Nelson et al. 1995). 2.1.1 Introduction to Building Related Illness (BRD and Sick Building Syndrome (SBS) Some indoor air investigations are characterized by the discovery of a specific source or reason for building related illness such as the presence of microorganisms in the ventilation system associated with the development of hypersensitivity pneumonitis (Hood 1990; Thom et al. 1996). Cases where a specific disease or symptom is related to a building (indoor) exposure are referred to as building related illnesses or BRI's (Burge 19906, Burge 1995; Horvath 1997). An example of a bacterial BRI is respiratory disease associated with exposure to aerosolized Legionella where Legionella was also found in the water from which the aerosol was generated (Burge 1995). Other BRI's could include asthma, hypersensitivity pneumonitis and other specific respiratory illnesses (Horvath 1997). In 1983 the World Health Organization (WHO) defined SBS as "an increase in the frequency of building occupant reported complaints associated with acute non-specific symptoms ... in non-industrial environments that improve while away from the buildings" (WHO 1986). In a review of studies of indoor air quality complaints, Mendell (1993) found many which failed to identify a specific source of contaminant that could be the cause of the wide-ranging symptoms as reported below. 4 The generally reported symptoms related to SBS can be categorized by the system affected (Finnegan et al. 1984; Kreiss 1989; Hodgson 1992; Nelson et al. 1995). Some of these symptoms could also apply to BRI's (Burge 19906, Burge 1995; Horvath 1997). • upper respiratory or mucosal irritation (dry, itching, irritated or watering eyes; dry or sore throat; nasal symptoms, sinus congestion or sneezing; cough); • lower respiratory problems (tight chest; difficulty breathing; shortness of breath; wheezing); • central nervous system complaints (headache; depression; tiredness or lethargy; tension or irritability; memory deficits; dizziness or lightheadedness); • dermal problems (skin rash or dry skin); • sensory discomfort from odours; • gastrointestinal (nausea). The prevalence of SBS can be difficult to study, in part because outcomes may take the form of loss in productivity rather than absenteeism or compensation claims which are easier to track (Hockaday 1988). However, the prevalence of symptoms reported by building occupants on survey questionnaires can range from 1 - 50% of respondents (Finnegan et al. 1984; Kreiss 1989; Hill et al. 1992; Hodgson 1992; Chang et al. 1993; Mendell 1993; Jaakkola et al. 1994; Apter et al. 1994; Nelson et al. 1995; Willers and Andersson 1996). The public perception of indoor air quality problems seems to focus on modern building construction practices that rely on mechanical ventilation, and windows that can't be opened (Kreiss 1989; Hodgson 1992). The focus on mechanical ventilation comes in part from a study reported in the British Medical Journal by Finnegan et al. (1984), and a report from NIOSH which identified inadequate ventilation as the primary problem in 53% of indoor air quality investigations (Seitz 1990). However, in a study done of four non-problem buildings in Washington State, the two buildings that were associated with the most complaints were buildings with operable windows, one with a variable air volume ventilation system, the other with a constant air volume ventilation system (Nelson et al. 1995). 5 Tamblyn et al. (1992) identified two key components lacking in SBS research. There is no single feature or test considered diagnostic for SBS and no standardized questionnaires or methods for determining symptoms (Tamblyn et al. 1992). In an attempt to bring more rigour to the study of SBS, recent research has moved away from simple prevalence studies and has focused on systematically studying the relationship of components of the indoor environment to symptoms reporting. Double blind and cross-over trials of air handling systems have looked at ventilation (Tamblyn et al. 1992), relative humidity (Reinikainen et al. 1992) and recirculation of air (Jaakkola et al. 1994) in attempts to find solutions to indoor air quality complaints. The U.S. EPA currently has a large, ongoing study categorizing building factors and symptoms called the Building Assessment Survey and Evaluation (BASE) study (Apte and Daisey 1999). 2.1.2 The Controversy Surrounding SBS Although SBS is accepted by the public and indoor air quality researchers, the clinical relevance of SBS is still under considerable debate (Chang et al. 1993). The primary objection to the recognition of the entity of SBS is the lack of a reasonable hypothesis as to a source that could occasion such a diverse array of symptoms (Ryan and Morrow 1992; Chang et al. 1993; Rothman and Weintraub 1995; Horvath 1997). Detractors arguing against the plausibility of SBS symptoms being associated with indoor air contaminants point to the findings of studies where the primary correlate of SBS is gender or stress in the workplace (Hodgson 1992; Mendell 1993; Horvath 1997). In factories, exposure limits are routinely used to define a level of risk that is understood to be protective of the majority of workers (ACGIH 1999; WCB 1998). However, levels of contaminants in office air are much lower than in industrial settings, and monitoring data from non-problem buildings often does not differ significantly from problem buildings (Nelson et al. 1995). This makes the exposure data difficult to interpret, and a source or sources of the health complaints difficult to pinpoint. Exposure-response relationships are seldom identified (Ryan and Morrow 1992; Chang et al. 1993; Rothman and Weintraub 1995; Horvath 1997). There is little information available to guide in the interpretation of exposures that, while low, are confounded by a 6 multitude of sources and combinations of chemicals which may account for occupant complaints (Molhave 1982; Molhave et al. 1986; Fanger 19886). The background information on BRI and SBS illustrates the interest in associations between indoor air quality and complaints of ill health. However, the research undertaken for this thesis was not a health effects study. The approach taken for this work was to study factors which could clarify the role of environmental, building and occupant factors to levels of exposure to selected indoor air quality parameters which will be described in greater detail in Section 2.2. 2.1.3 Introduction to Indoor Air Quality Investigation In the United States, investigations into indoor air complaints have historically followed a protocol similar to the investigation of industrial sites. NIOSH has developed standardized protocols for teams of investigators to inspect buildings, administer symptom questionnaires and monitor for levels of suspected contaminants (Gormon and Wallingford 1989). Indoor air investigations typically include measurements of volatile organic compounds (VOC's), aldehydes, carbon dioxide, carbon monoxide, particulate matter, bioaerosols (fungi and bacteria), relative humidity and temperature, illumination, and noise (Finnegan et al. 1984; Kreiss 1989; Hodgson 1992; Chang et al. 1993; Levin 1995; Nelson ef al. 1995). • VOC's and aldehydes are most often associated with emissions from building materials and furnishings (Molhave 1992). VOC's may also be associated with bioeffluents from occupants (Fanger 1988a; Fanger 19886; Batterman and Peng 1995). • Indoor carbon monoxide is typically associated with incomplete combustion from burning of fuel for heating or other combustion processes (Engineering Interface 1989; Alberta Health 1993). • Illumination and noise levels are regulated by workplace standards as safety concerns (WCB 1998) or can function as sources of physical stressors (Levin 1995). 7 • Particulate matter may be generated indoors by occupant activity, combustion or natural sources (Hill et al. 1992; Etkin 1994). • Temperature and relative humidity play a large role in determining occupant comfort (ANSI/ASHRAE 1981; Levin 1995). • Carbon dioxide is produced by occupants and combustion sources, and is a surrogate measure of ventilation efficiency (Godish et al. 1986; Salisbury 1986; ASHRAE 1989; Engineering Interface 1989; Bearg 1993; Meckler 1993; Olcrest 19946; Batterman and Peng 1995; Jankovic et al. 1996). • The origin of airborne bacteria may be the bioeffluents from occupants (Morey 1985; Otten et al. 1986; Liu 1998) or from reservoirs within the building such as contaminated heating, ventilation and air-conditioning (HVAC) systems (Hood 1990; Teeuw et al. 1994; Burge 1995; Thorn et al. 1996; Weltermann et al. 1998). • The origin of airborne fungi may be from the free exchange of outdoor and indoor air, or may be from amplification sites within the building (Beaumont et al. 1984; Burge 1985; Burge et al. 1987; Hockaday 1988; Gravesen et al. 1994; Burge 1995; Dillon et al. 1996; Rand 1998; CMHC 1998). Section 2.2 will explore the components of an indoor air quality investigation in greater detail. 2.1.4 Introduction of Guidelines for Evaluation of Indoor Air Quality The study of indoor air quality is complex in nature due to the multiplicity of sources of contaminants and the difficulty in interpreting concentration levels which may be very close to ambient concentrations. Nonetheless, guidelines and standards are being developed by government and professional health and safety organizations based on research and epidemiologic studies. Examples of some of these guidelines follow: • Workplace exposures to chemical, physical and biological hazards: TLVs® and BEIs® Threshold Limit Values for Chemical Substances and Physical Agents; Biological Exposure Indices (ACGIH 1999). - Occupational Health and Safety Regulation: BC Regulation 296/97 (Workers' Compensation Board 1998) 8 • Comfort parameters: - Thermal Environmental Conditions for Human Occupancy (ANSI/ASHRAE 1981). • Ventilation: - Ventilation for Acceptable Indoor Air Quality (ASHRAE Standard 62-1989). • General Indoor Air Quality: - Indoor air quality assessment: a working manual (Alberta Health 1993). - Healthy Building Manual: Systems, parameters, problems and solutions (with Ontario IAQ parameters) (Engineering Interface 1989) Indoor Air Quality. Tools for Schools. IAQ Coordinator's Guide (EPA 1995). Indoor air quality in office buildings: A technical guide (Health Canada 1995a). • Bioaerosols (guidelines and sampling protocols): - Guidelines for assessment and sampling of saprophytic bioaerosols in the indoor environment (Burge et al. 1987). Field Guide for the Determination of Biological Contaminants in Environmental Samples (Dillon et al. 1996). - Bioaerosols: Assessment and Control (Macher et al. 1999). • Bioaerosols (guidelines for remediation and clean-up): Toxic mold [sic] clean up procedures: A guide for cleanup contractors (Canada Housing and Mortgage Corporation 1998). Guideline on assessment and remediation of Stachybotrys atra in indoor environments (New York City Department of Health 1998). Of the references listed above, the only one with the power of enforcement for provincial workers is the Workers' Compensation Act (WCB 1998). Two standards are referenced within the Workers' Compensation Act (ANSI/ASHRAE 1981; ASHRAE 1989). The other listed guidelines and standards are adopted by various jurisdictions and agencies. 9 The guidelines pertaining to bioaerosols are still under consideration and discussion as to their applicability and validity. With the exception of the thermal comfort and ventilation standards, most of the indoor air quality guidelines and standards were written within the last five to ten years. Some specialized environments such as schools have only been able to make use of indoor air quality guidelines within the last year. This has not been due to lack of interest, however, and in some school districts, teachers have requested that indoor air quality become a focus of occupational health and safety departments (L. Sinclair, B.C. Teachers' Federation, personal communication). As of April, 1998, the Workers' Compensation Board of British Columbia (WCB) adopted regulations for indoor air quality within Occupational Health and Safety Regulation 296/97 (WCB 1998). For the first time teachers may invoke the health and safety regulations Sections 4.70 - 4.8 that specifically address indoor air quality, which require investigative procedures when complaints are reported (WCB 1998). Schools in Ontario, Newfoundland and Nova Scotia have been featured on national radio and television as having indoor air problems (Vancouver Sun 1998; Montreal Gazette 1999; Toronto Star 1999; Tanner 1999; Turner 1999), and in Nova Scotia, school boards are seeking guidance in drafting indoor air quality policy (Rand 1998). Interest in the air quality in British Columbia schools has been localized to individual school boards, and to date, neither the Ministry of Health nor the Ministry of Education has developed policy regarding investigative procedures. Central to the issue of investigating air quality complaints is the need for prospective, non-problem, regional data with which to compare complaint data, and the need to identify characteristics of building systems that may affect air quality. One of the emerging issues in indoor air quality research is that it is difficult to apply global standards to different climatic zones. An example of this comes from a 1998 provincial commission of inquiry which reported on the investigation of complaints of water damage in relatively new, residential buildings in British Columbia (Barrett 1998). This commission also served to focus attention on water damage to buildings in general. The commission found that changes in construction practices in B.C., and using building codes appropriate to mid or eastern Canadian climates resulted in housing at risk of water 10 damage and subsequent fungal colonization (Morris 1997; Barrett 1998). Regional differences in climate, building materials and construction techniques may have an impact on indoor air quality parameters. Microclimates and season are known to influence concentration and composition of bioaerosols (Beaumont et al. 1984; Wilken-Jensen and Gravesen 1984; Sneller 1984; Kozak et al. 1985; Reponen et al. 1992; Burge 1990c). The unanticipated effects of microclimates make global application of guidelines problematic and warrant the creation of local databases to assess their applicability. This present study was undertaken to develop a tool to aid in the interpretation of culturable fungal data for this geographic region, using relationships of the natural and built environment to fungal counts as a systematic approach to the interpretation of fungal measurement data. 2.2 COMPONENTS OF INDOOR AIR QUALITY INVESTIGATIONS 2.2.1 Introduction to Bioaerosols The ACGIH bioaerosols committee defines bioaerosols as "... those airborne particles that are living or originate from living organisms (Macher et al. 1999). Bioaerosols include microorganisms (i.e., culturable, non-culturable, and dead microorganisms) and fragments, toxins, and particulate waste products from all varieties of living things." The ACGIH bioaerosols committee (Macher et al. 1999) defines biologically derived airborne contaminants as "... bioaerosols, gases, and vapors [sic] that living organisms produce. Biologically derived materials are natural components of indoor and outdoor environments but, under some circumstances, biological agents may be considered contaminants when found indoors." The bioaerosols that will be considered in this thesis are culturable fungi and bacteria. 2.2.1.1 The Role of Indoor Bioaerosols to Disease Information available to the public from scientific (Wilson 1998; Schoen 1998), government (Health Canada 1995a; Health Canada 19956; New York Health Department 1998) and lay press (Montreal Gazette 1999; Toronto Star 1999; Vancouver Sun 1998; Tanner 1999; Turner 1999) has raised concern regarding the possible health 11 consequences of living or working in mouldy buildings. The concern becomes especially acute regarding school buildings and children as illustrated by newspaper titles such as "Toxic mould found in schools" (Vancouver Sun 1998), "Breaking out of the mould: indoor fungus is bad for your health, even worse for your children's" (Montreal Gazette 1999), "Toxic fungus found in school portables" (Wilson 1998) or "Air quality closes Nova Scotia school: fungi attacks other facilities" (Daily Commercial News 1997). Recently, attention has been focused on a few genera of fungi as being particularly dangerous. Stachybotrys, a strongly cellulolytic fungus most often found associated with dead plant material, has been implicated in cases of serious building related illness. Etzel and Dearborn (Etzel et al. 1998) were the first public health officials to suggest an association with Stachybotrys chartarum/atra in apartments with sometimes fatal cases of pulmonary hemorrhage in infants. Other studies have implicated Stachybotrys as the putative cause of a variety of deleterious health outcomes including neurologic (fatigue, headache), upper respiratory (sinus congestion, cough, shortness of breath), gastrointestinal (diarrhea) and dermatologic (rash) (Johanning et al. 1996; Jarvis et al. 1996; Hodgson et al. 1998; Jarvis et al. 1998; Sudakin 1999). Unfortunately, Stachybotrys is not the only toxigenic fungi that can be discovered in indoor air investigations. Memnoniella (Jarvis et al. 1996; Jarvis et al. 1998), Aspergillus versicolor (Hodgson et al. 1998), Fusarium (Health Canada 1995a) and certain strains of Penicillium (Gravesen et al. 1994) are also capable of producing toxins. In problem buildings these fungi can colonize damp building materials and correlations have been suggested between the amplification of toxigenic fungi and symptoms of sick building syndrome (Johanning et al. 1996; Sudakin 1998; Cooley et al. 1998). In addition to toxigenic fungi, at least 95 other fungal genera have been reported to be associated with allergic disease (Cruz et al. 1997). Occupants of buildings contaminated with microorganisms may have unique reactions to microbial populations and may not respond predictably to exposure gradients because individuals may be more sensitive, or allergic to specific aerosols (Beaumont et al. 1984; Brooks et al. 1992; Pope et al. 1993; Homer et al. 1999). Even non-atopic individuals vary in their reactions to fungal metabolites and mycotoxins (Burge 1990a). For example, a cell wall component of fungal mycelia and spores, (l-»3)-p-D glucan, has been shown to initiate an 12 inflammatory response in alveolar macrophages (Goto et al. 1994). Recent studies have found correlations between measurable (l-»3)-P-D glucan in office and home air and symptoms of cough, skin irritation or asthma (Rylander et al. 1992; Rylander et al. 1994). The amount of (l-»3)-p-D glucan found in fungal spores seems to be species specific and may help to explain why some, but not all, fungi are associated with health complaints (Fogelmark and Rylander 1997). Similarly, endotoxin, a lipopolysaccharide, cell membrane component of Gram-negative bacteria, causes an inflammatory response in airways and lung tissue (Folgelmark et al. 1994). In some investigations of building related illness, endotoxin or Gram-negative bacteria in humidification systems were strongly associated with the symptoms reported by building occupants (Hood 1990; Teeuw et al. 1994). It has been demonstrated in animal models that simultaneous exposure to endotoxin and (1—»3)-p-D glucan result in increased numbers of inflammatory cells and altered alveolar macrophage function (Fogelmark et al. 1994; Rylander and Fogelmark 1994; Pratt et al. 1994). The combination exposure of high levels of endotoxin and (l->3)-p-D glucan is thought to play an important role in the development of a transient, febrile disease affecting the lungs called organic dust toxic syndrome (Fogelmark et al. 1994). The ability of exposures to low concentrations of endotoxin and glucan to cause respiratory symptoms is not established, although Dutch researchers have shown relationships between these components in house dust and increased peak flow variability in asthmatic children (Douwes et al. 19986). Unless there are amplification sites to allow growth of Gram-negative bacteria (Hood 1990), occupants of buildings are the primary source of indoor bacteria (Meyer 1983; Spendlove and Fannin 1983) and bioeffluents (Fanger 1988a, 19886). Meyer (1983) reports that adults shed up to 3 grams of epithelial cells per day. These cells are colonized by commensal microflora which are continually being shed along with the desquamated cells. Other sources of bacteria may be from paper products and outdoor sources such as dirt and plants. These primarily Gram-positive bacteria are not usually considered to be pathogenic, but recent studies have identified cytotoxic toxins produced by common Bacillus species (Salkinoja-Salonen et al. 1998; Andersson et al. 1999). 13 Some measures of sick building syndrome were found to be associated with peptidoglycan, a cell wall component of bacteria (Liu 1998). 2.2.1.2 Factors Contributing to Indoor Bioaerosol Concentrations Fungi and bacteria are ubiquitous in the outdoor environment and are the normal flora of soil and plant surfaces (Burge 1990a). Not surprisingly, the airborne organisms outdoors, which range in size from 0.5-10 um, easily find a way into the interior space through fresh air intakes or simply through open windows and doors by entrainment in air currents (Burge 1995). One could expect the indoor concentrations of airborne bacteria and fungi to be similar to concentrations of outdoor flora (Burge et al. 1987; Flannigan 1995; Dillon et al. 1996; Burge et al. 1999) and to be qualitatively similar in composition (Beaumont et al. 1984; Burge et al. 1999). However, the presence and concentration of bioaerosols inside a building are thought to be influenced by the nature of the building itself, the materials and construction practice, and interior conditions of biologically available moisture and micronutrients (Spendlove and Fannin 1983; Burge 1985; Hockaday 1988; Adan et al. 1994; Becker 1994). The method of ventilating a building will influence the concentration of fungi, decreasing the concentration under some circumstances (Hake et al. 1999; Thomann and Tulis 1999) and increasing it in others (Burge 1985; Infante et al. 1994). The role of moisture and water activity (aw) in indoor bioaerosol concentrations Epidemiologic research has established a relationship between dampness in homes and buildings and respiratory symptoms experienced by occupants (Brunekreef et al. 1989; Dales et al. 1991; Kosinen et al. 1994; Wan and Li 1999). The association of the symptoms may be due in part to the presence of fungi in damp buildings, which is related to the need for elevated relative humidity (> 70% RH) for microbial growth (Burge 1985; Arundel et a/.1986). In order to understand the part that relative humidity plays in the promotion of microbial growth, the concept of biologically available water, or water activity (aw) must be understood. The quantity aw is the expression of the amount of available water in a substrate. The concept of aw is familiar to food scientists; it is commonly measured in food products as a predictor of microbial spoilage (Trailer and Christian 1978; Beuchat 1987; Trailer 14 and Scott 1992). A w is analogous to relative humidity in air and is expressed as the relative humidity (expressed as a decimal) in air in a closed container in which a substrate has come to equilibrium. This is a temperature-dependent measure (Troller 1978 and Christian 1978; Troller and Scott 1992). In food chemistry, the aw requirements of many fungi and bacteria have been determined and the information is used to incorporate solutes (e.g. salt, sugar) into food products as preservatives (Troller and Scott 1992). Mycologists are familiar with the aw requirements of fiingi, because the range is sufficiently wide that two types of media are needed to maximize the recovery of allergenically important species. Routine culture media have a water activity approaching 1.0. Xerophilic media (aw 0.65 - 0.9) address the requirements of fungal taxa or species that require a lower water activity by incorporating solutes that sequester biologically available water (Troller and Christian 1978; Beuchat 1987; Flannigan 1995). Building materials and indoor bioaerosol concentrations In addition to the simple requirements of biologically available water, building materials themselves contribute to the survival of microorganisms (Becker 1994; Pasanen et al. 1994; Vitanen 1995). The role building materials play is to furnish substrates to microflora that may enhance or inhibit microbial growth. For example, drywall, a commonly used interior finishing material, is essentially a sandwich of paper (cellulose), cornstarch, and an interior core of inorganic calcium sulfate. When exposed to water or high humidity, the interior core acts as an absorbent material to hold water, and the cellulose and cornstarch furnish sufficient nutrient to allow luxuriant fungal growth from settled spores (Pasanen et al. 1994; Adan et al. 1994; Becker 1994). Other cellulose-containing substrates may be wallpaper, acoustic ceiling tile or wood. Similarly, inert substrates such as fibrous glass or rock wool, which are used in insulation or as sound absorptive lining in air ducts, can support growth of fungal colonies due to the entrapment of dust and nutrient in the fibre mesh (Burge 1985; Ezeonu et al. 1994). These building materials are all susceptible to wetting, and their ability to retain water is a crucial difference among them. Once wet, they may furnish sufficient moisture to allow a bloom of fungi that requires high water activity (aw) such as Cladosporium or Stachybotrys (aw 0.94-0.98), or later favour fungi that require less water 15 activity, such as Penicillium ox Aspergillus (aw 0.75-0.8) (Pasanen et al. 1994; Flannigan 1995). Interest in the ability of building materials to furnish metabolically available water to promote microbial growth has taken on new life as architects and building engineers have been forced into the indoor air quality arena. The wood industry is very familiar with the need to treat green wood with fungicides to inhibit sap-staining fungal growth, but has only recently began to focus on the need to study the rewetting characteristics of wood (Foarde et al. 1994; Morris, 1997; Hazleden 1999). Anti sap-stain chemicals such as chlorophenols, which were in use until the 1980's, retain their fungal inhibitory capacity almost indefinitely (P. Morris, Forintek Canada Corp., personal communication), but are currently banned from use, due to their toxicity to humans. Not only have the anti sap-stain chemicals changed in the last ten years, but the kinds of woods being used in British Columbia have changed as well. Douglas fir was traditionally used to frame buildings, and Douglas fir has a high percentage of heartwood, which is relatively nutrient poor. Current lumber markets have replaced fir with spruce or pine, which have a higher percentage of sapwood which is nutrient rich for fungal growth (P. Morris, Forintek Canada Corp., personal communication). Because of these changes in wood used for framing, experimental methods are being developed to study the ability of construction wood to support the growth of bacteria and fungi (Vitanen 1995; Hazleden 1999). Other building materials are also being studied for their ability to sequester water and support microbial growth (Foarde et al. 1994; Adan et al. 1994; Pasanen et al. 1994). The role of other substrates in bioaerosol concentrations Kozak et al. (1985) found that housekeeping practices, and landscaping surrounding buildings, could also be associated with the concentration of indoor bioaerosol. Homes that lacked rigourous housekeeping, or that had excessive amounts of plant materials nearby, or shading of buildings by bushes and trees, tended to have higher indoor fungal counts than homes that were meticulously kept (Kozak et al. 1985). "House dust" is a combination of many organic substances, including human skin cells. If sufficient biologically available water is present, both fungi and bacteria are able 16 to use the organic components of dust as a nutrient source (Gyntelberg et al. 1994; Gravesen et al. 1994; Douwes et al. 1998a; Douwes et al. 19986). 2.2.1.3 Factors Influencing the Evaluation of Indoor Bioaerosol Concentrations As of the most recent publications by the American Conference of Governmental Hygienists (ACGIH) (Macher et al. 1999), there are no numeric guidelines against which to compare measurements of airborne bioaerosols in indoor air quality investigations. The ACGIH (Burge et al. 1999) suggests bioaerosols be evaluated on a case by case basis using criteria of pathogenicity, species or indoor/outdoor ratios not reflective of an outdoor source, or evidence of microbial growth on building materials. In contrast, Health Canada (1995a) has suggested using an interpretation based on studies done in office buildings which were used to establish baseline numeric values and unacceptability of certain fungal genera in an indoor environment. The challenge faced by school boards in determining whether there is a problem with bioaerosols in classrooms is in evaluating what guidelines to use to interpret bioaerosol measurements. The Health Canada (1995a) guidelines depend on the preexistence of well maintained, filtered, mechanical ventilation to achieve counts within range, even in non-complainant buildings. However, it becomes problematic to compare fungal counts in a naturally ventilated space that depends on open doors or windows for fresh air with those from a space supplied with fresh air through a mechanical system. Attention has recently focused on specific toxigenic fungi (Stachybotrys chartarum, Aspergillus versicolor, Fusarium spp.) (Health Canada 1995a, Burge and Otten 1999). Although Stachybotrys is easily identified by its distinctive spores using non-viable sampling methods, studies of exposure to Stachybotrys using culture - based methods have been hampered due to the relative difficulty in culturing the organism from air (Dillon et al. 1996). The lack of an agreed-upon exposure limit for Stachybotrys has challenged investigators to interpret the finding of low numbers of Stachybotrys spores in non-viable spore-capture samples. The decision to remediate a building may be based entirely upon the finding of potentially toxigenic or pathogenic fungal spores in any quantity from an indoor air sample. There are no reports in the literature to support an exposure-response relationship that, in the absence of visual colonization of building 17 materials, indicates a potential health hazard due to the presence of low concentrations of spores. 2.2.2 Ventilation, CO? Comfort Parameters and Particulate Matter 2.2.2.1 Indoor Air Quality and Mechanical Ventilation Interior spaces such as offices or school rooms have only two options for fresh air supply. Fresh air can be delivered mechanically, or by natural ventilation by opening windows or doors. Mechanical air delivery systems depend on fans to move air from the outside, and on ducting to deliver the air into the interior space. The air handling system can disseminate contaminants throughout the building in addition to delivering fresh air to the interior spaces. The comfort parameters, temperature and relative humidity, are largely controlled by the air handling system (Engineering Interface 1989; Bearg 1993). In cold climates, the air must be tempered, or warmed, before delivery to occupied spaces. Air handling units (AHU) may simply move a constant volume of air from the outside, while newer systems may vary the volume of outdoor air depending on outside temperatures, in order to save energy costs. Some mechanical systems only move air when heating is required, and do not supply air when room temperature is above the set-point temperature. Similarly, natural ventilation can be temperature dependent, with windows and doors opened during clement weather primarily to regulate comfort rather than to supply fresh air. 2.2.2.2 Relationship Between CO? and Ventilation Why measure CO?? CO2 concentration in buildings is often considered a surrogate measure of ventilation efficiency. Many studies have found evidence of occupant dissatisfaction when working in environments with slightly elevated CO2 (Salisbury 1986; Daneault et al. 1992; Hill et al. 1992; Kjaergaard and Pedersen 1992; Batterman and Peng 1995; Willers and Andersson 1996). However, under controlled conditions where volunteers were observed performing different tasks, there was little evidence to suggest measurable decrements of performance until short term levels of 2.5 - 8% CO2 (25,000 - 80,000 ppm) were reached (Yang et al. 1997; Maresh et al. 1997). Research in extreme 18 environments (Tansey et al. 1979) such as were found in submarines, reports that prolonged exposure to 10,000 ppm CO2 resulted in changes in calcium uptake into soft tissue from bone in animal model. The same researchers found epidemiologic evidence of increased incidence of urinary calculi in submarine crew members at similar, prolonged exposures (Tansey et al. 1979). However, volunteers exposed to 4,500 ppm CO2 for eight hours failed to show any measurable effects on respiration rate or on urine pH (Strieker et al. 1997). The CO2 concentration in buildings in cold climates, where air is recirculated for energy conservation, may approach 4,000 ppm (Lundquist et al. 1982) and have been reported with maxima as high as 5,900 ppm (Lee and Chang 1999). The 8-hour industrial exposure limit to CO2 is 5,000 ppm with a 15 minute exposure limit of 15,000 ppm (WCB 1998). Exposure limits for non-industrial workplaces are based on ASHRAE 62-1989 guidelines, which recommend ventilation rates that result in attainment of C0 2 levels below 1,000 ppm (WCB 1998 Section 4.70 -4.8). Guidelines for CO2 concentrations in non-industrial workplaces are lower than those regulated for industrial workplaces or that would be found in extreme environments. The toxicity of CO2 is not an issue at concentrations typically found in indoor air quality investigations (Strieker et al. 1997). CO2 is of interest because it is one substance of many produced by occupants as a bioeffluent and may be used as a surrogate measure of ventilation (Salisbury 1986; Meckler 1993; Olcerst 19946; Jankovic etal. 1996). Reduction of indoor air contaminants by fresh air Depending on diet and activity level, each person generates approximately 0.3 litres per minute CO2 (ASHRAE 1989). Any enclosed (e.g. indoor) space will accumulate an increasing concentration of CO2 over time if the volume of CO2 produced by the occupants is greater than the ability of the ventilation system to dilute the CO2 with fresh air. According to the principles of dilution ventilation, the introduction of sufficient uncontaminated air will dilute polluted air in an enclosed space (ACGIH 1992). It is assumed that ventilation which measurably reduces occupant-generated CO2 will 19 also reduce the concentration of other unidentified contaminants which may include infectious or allergenic particles (Wheeler 1993; Burge 19906). Dilution ventilation reduces the bioeffluents (volatile organic compounds (VOC's), commensal bacteria, and CO2) contributed by the occupants of the space. The "olf' is a unit of measurement of the olfactory contribution of VOC's produced by an average person, and is considered to be a major component of the perception of stale, stuffy or unacceptable air (Fanger, 1988a; Batterman and Peng 1995). The ASHRAE ventilation committee traditionally recognized the relationship between body odour in poorly ventilated offices and occupant complaints of stale air (Engineering Interface 1989). Olf factors cannot be measured directly, and are quantified by the perceptions of a trained panel of individuals. Batterman and Peng (1995) devised a method for measuring and differentiating occupant-generated VOC's from building VOC's and found a correlation between occupant generated VOC's and occupant generated CO2 (Pearson correlation coefficient r=0.75) (Batterman and Peng 1995). However, their method has not been widely used in routine indoor air quality investigations. The use of CO2 as a surrogate of ventilation efficiency is generally agreed upon (Godish etal. 1986; ASHRAE 1989; Meckler 1993; Persily 1993; Olcerst 1994a, 19946; Jankovic et al. 1996), and has been incorporated into regulatory guidelines (Health Canada 1995a; WCB 1998). Unfortunately, simple reduction of CO2 levels does not always result in the perception of improved air quality. Epidemiologic studies of sick buildings have reported mixed outcomes in symptom scores when ventilation is improved (Tamblyn et al. 1992; Mendell 1993; Apter et. al. 1994; Jaakkola et al. 1994; Nelson et al. 1995). Other sources of VOC's Volatile organic compounds in the indoor environment are defined by Molave (1982) as organic chemicals with boiling points between 50°C and 260°C. These chemicals can be associated with building materials, building furnishings and equipment, biological materials (e.g. fungal metabolic products) or with humans (Molhave 1982; Fanger 19886; Burge and Otten 1999). Concentrations of total VOC's in office air are typically orders of magnitude lower than permissible exposure limits for chemicals used 20 in industrial sites (Molhave 1982; Molhave et al. 1992; WCB 1998; ACGIH 1999), but still may be associated with complaints of poor air quality (Molhave 1982; Molhave et al. 1986; Molhave 1992). Research efforts have concentrated on subtle health effects that may be related to VOC's, such as changes in eye secretions to account for the often cited symptom of dry or irritated eyes (Molhave 1992; Kjaergaard 1992; Kjaergaard and Pedersen 1992), or a simple odorant effect to produce stress reactions (Molhave et al. 1986; Cone and Shusterman 1991; Silver 1992). VOC's in non-industrial settings are complex mixtures of compounds and are not often associated with a single source. VOC's can be sequestered in fleecy furnishings materials, such as the cloth used to cover dividers in open plan offices, to be released later under differing ventilation, humidity and temperature conditions. The ubiquity and assault of synthetic materials in all aspects of indoor environments have been suggested by some researchers and sufferers alike to have a relationship with symptoms of sick building syndrome (Molhave 1982; Molhave 1992). 2.2.2.3 Investigation of the Ventilation System The investigation of the air handling system typically includes a visual inspection of the ducts and filters for the presence of fibrous glass, general cleanliness, evidence of microbial contamination or conditions which could support microbial growth, examination of the supply and return air intakes for cross contamination, measurements of air flow and air flow balance within the building, and a characterization of the dilution capability of the supply air using measurement of carbon dioxide concentration (Bearg 1993; EPA 1995). The efficiencies of forced air systems vary. If air in a room is not adequately circulated, supply air may exit via the return air system without benefiting the occupants. The supply of air, and the mixing characteristics of the room, can be measured using a tracer gas such as sulfur hexafluoride (SF6) (Bearg 1993). 2.2.2.4 Standards and Evaluation of Ventilation The American Society of Heating, Refrigeration and Air-Conditioning Engineers, Inc. (ASHRAE), is responsible for researching and writing standards for acceptable 21 ventilation in indoor spaces. In British Columbia, ASHRAE standards are used for building codes and setting criteria for occupational indoor air quality. The standards currently in use that apply directly to indoor comfort and ventilation are ANSI/ASHRAE 55-1981 and ASHRAE 62-1989 (ANSI/ASHRAE 1981; ASHRAE 1989). The fresh air requirement varies with room use or design. For educational purposes (classrooms), the outdoor air requirement is 15 cfrn/person (equivalent to 0.008 m3/sec per person). This requirement is derived from a mass balance equation that takes into account the ratio of metabolic CO2 produced, to oxygen consumed, assuming a "normal" diet of fats, carbohydrates and protein, and light physical activity by occupants of an indoor space. If the provision of outdoor air is sufficient, the steady state CO2 level will be maintained below 1000 ppm. 2.2.2.5 Comfort Parameters (Temperature and Relative Humidity) Taken together, ventilation, temperature and relative humidity (RH) can be thought of as comfort parameters. When the temperature or RH are outside relatively narrow ranges in the indoor environment, occupant complaints may occur (Finnegan et al. 1984; Kreiss 1989; Engineering Interface 1989; Hodgson 1992; Hill et al. 1992; Reinikainen et al. 1992; Mendell 1993; Levin 1995; Batterman and Peng 1995; Willers and Andersson 1996). Temperature The contribution of controlled indoor temperature and relative humidity to comfort is recognized in ASHRAE Standard 55-1981 (ASHRAE 1981) and is referenced by other professional bodies (Engineering Interface 1989; Alberta Health 1993; Health Canada 1995a; EPA 1995). A comfortable temperature is subjective and depends on occupant activity levels and clothing. ASHRAE Standard 55-198 suggests ranges of temperatures appropriate for the season (e.g. 20 - 23.6 °C for winter) and which should result in appropriate comfort levels for at least 80% of occupants of the space. Relative humidity and equilibrium relative humidity Relative humidity is defined as the ratio of the amount of water vapour present in the air to the amount of water vapour present in saturated air at the same temperature and 22 barometric pressure (ANSI/ASHRAE 1981). Most building occupants are unaware of incremental changes in relative humidity. For example, at room temperature, a change in relative humidity from 30% to 70% would be perceived to be the same as a 1°C change in temperature (Engineering Interface 1989). However, excessively low RH can be associated with occupant complaints. Reinikainen et al. (1992) found statistically significant increases in symptom scores for skin, eye and nasal congestion when the relative humidity was lower than 20-30% in a cross-over trial conducted in Finland. The mass of water contained per unit of dry air is called the absolute humidity. The absolute humidity measured at the perimeter of a room can be different from that taken at the centre of a room (Engineering Interface 1989), because the water contained in the air may be deposited on room surfaces by condensing on cooler wall surfaces (windows or uninsulated walls). Furthermore, a water leak may wet building material from within. Water condenses out of air at the dew point temperature, and may provide sufficient water to support microbial growth (Vitanen 1995; Foarde et al. 1994; Adan et al. 1994; Pasanen et al. 1994). When the humidity at the building material surface is isolated from the room humidity, the resulting measurement is of the humidity in equilibrium with the material, or the equilibrium relative humidity. In biological terms, the water contained in building materials, that is available for support microbial growth, is expressed as water activity or aw (Troller and Scott 1992) and is measured as equilibrium relative humidity and is expressed as a decimal. Relative humidity recommendations for indoor space aim to limit the amount of available water to microorganisms while protecting furnishings and occupants from excessive dryness. ANSI/ASHRAE Standard 55-1989 (ANSI/ASHRAE 1989) recommends that optimal indoor relative humidity be maintained between 20 - 40% in winter. 2.2.2.6 Particulate Matter Particulate matter can be defined as any substance present in air as suspended, microscopic solids or droplets. Three size designations are defined by the ACGIH (1999) and are used by hygienists to define particulate matter in indoor environments and 23 workplaces. Particulate matter in the size range < 100 um is described as inhalable because the particles can enter and be deposited in the respiratory system. Atmospheric scientists use different nomenclature to describe three sizes of particulate matter, which are different, but which may deposit in roughly the same areas of the respiratory tract (CEPA 1998). These size definitions are listed in Table 1. Table 1. Definitions of size selective deposition of particulate matter ACGIH (1999) CEPA (1998) Area of deposition in respiratory tract designation designation (MMD)° (MMD)a Inhalable Total suspended Largest particles deposit in the nasal (100 urn) particles = TSP region; smallest, in the alveoli (gas-(< 40 um) exchange region). Thoracic PM 6 1 0 Deposits within the lung airways. (10 urn) (10 um) Respirable PM 2. 5 Deposits in the gas-exchange region. (4um) (2.5 um) " Mass median diameter of particles. b Particulate matter. The 50% mass median cut-off for open face cassette samplers is around 30 - 40 pm (Kenny et al. 1997). Prior to the introduction of size selective particle collection, particulate matter samples collected for occupational hygiene purposes were taken with open or closed face cassettes (ACGIH 1999). The particulate matter collected by this method would be in the size range of total suspended particles (TSP). In the outdoor environment, increasing levels of very small particulate matter (PM2.5) can be shown to be related to increased admissions to emergency rooms for respiratory complaints and increased mortality rates from respiratory complications and cardiovascular events (Vedal 1995). However, particulate matter concentrations in the indoor environment are typically lower than those outdoors. For example, the ratio of indoor to outdoor levels of PM10 is relatively stable (0.6-0.7) in the homes of non-smokers (Quackenboss et al. 1989). One longitudinal study conducted over a four year period in primary schools in Sweden showed an association with respirable particulate 24 matter with new cases of sick building syndrome, after chronic cases associated with total VOC's were accounted for (Norback et al. 1990). The source(s) of indoor particulate matter can be important in indoor air quality investigations (Quakenboss et al. 1989; Etkin 1994). In older studies, environmental tobacco smoke (ETS) was of interest (Quakenboss et al. 1989; Etkin 1994). However, in public buildings and schools, ETS has been eliminated by city by-laws (Vancouver City Health Department 1989; Vancouver City Health Department 1996) and health and safety legislation (WCB 1998). The composition of the majority of particulate matter found indoors is complex and the chemical composition is not often characterized. In a European study of 12 non-problem office buildings in Copenhagen, non-airbome dust samples were collected with a vacuum, and characterized for presence of Gram-negative bacteria, total volatile organic compounds and the ability to provoke histamine release. Relationships were found between: the prevalence of Gram-negative bacteria with general symptoms of fatigue, headache, concentration problems and upper respiratory tract, mucous membrane symptoms; between total volatile organic compounds and lack of concentration or heavy headedness; and between allergenic material and general malaise, dizziness and lack of concentration (Gyntelberg et al. 1994). Some fraction of the dust collected from carpets, etc., could potentially become airborne with sufficient agitation, such as foot traffic, or by the process of entrainment from normal vacuuming (Gyntelberg et al. 1994). Another potential source of indoor particulate matter is from the release of microscopic fibres from fibrous glass or rock wool, materials that have found their way into all forms of building materials, including insulation for sound and heat, and for light-weight strengthening of materials such as acoustical ceiling tiles. The movement of ventilation air may distribute respirable fibres (Infante et al. 1994; Horvath 1997) or microorganisms that have colonized the fibre batting (Burge 1985; Ezeonu et al. 1994). Respirable glass fibres may be responsible for some sick building syndrome symptoms such as itchy eyes or skin (Horvath 1997). IARC, the scientific body that evaluates and rates potentially carcinogenic substances, is currently evaluating fibrous glass as a carcinogen (Infante et al. 1994); however, there is no evidence to suggest exposures to concentrations typical of indoor environments would lead to cancer. 25 2.3 INTRODUCTION TO DETERMINANTS OF EXPOSURE There is great value in being able to create predictive models by using field data to identify variables that explain variance in the data set. The use of regression statistics allows the correlation between variables to be used to develop equations that can predict future values of one variable, given a value of the other variable (Munro and Page 1993). The discipline of epidemiology has made extensive use of linear regression modelling to identify variables related to disease outcomes. Similarly, occupational hygienists have been able to use measurements, and information collected in workplaces, to identify variables related to outcomes of exposure; to determine which elements of tasks, controls, work organization or environment may act to increase or decrease the exposure of interest (Burstyn and Teschke 1999). Measured exposures may be correlated with other measurements contributing to the exposure. In multiple linear regression, more than one variable can be fit to the model, and will result in a better prediction of the outcome variable by accounting for more of the overall variance in the outcome. However, when multiple variables are entered into a model they may affect one another, and the selection of appropriate variables requires careful attention to correlations among the array of potential explanatory variables. Determinants of exposure studies may be either experimental or observational. In experimental designs the researcher has control of the variables of interest and can vary conditions - for example, testing different ventilation designs. In observational studies the researcher is faced with a multiplicity of factors at the study sites that are not controlled. The challenge of observational studies is to adequately capture the relevant data. The researcher must make a priori decisions about what factors will be measured or recorded and to plan sufficient samples to have the power to incorporate the chosen factors into the model. Strategies to increase the representativeness of the data and decrease systematic biases include random selection of the sampling sites, days, and areas sampled (Burstyn and Teschke 1999). The study reported here was an observational study. Measurements and observations were made on the sampling day, and incorporated the daily routines of the 26 classroom and school. Teachers were not asked to vary their routines to accommodate the study. 27 CHAPTER 3 SITE SELECTION AND CHARACTERIZATION 3.0 METHODS 3.1. Study Site Characterization The study was carried out in all elementary schools in one school district in the Greater Vancouver Regional District of British Columbia, an urban area in a coastal temperate zone, with a population of about 1.8 million people. The urban and suburban core spreads over the Fraser River delta, with the Strait of Georgia to the west, the Fraser valley to the south and east and the Coast Mountain range to the north. Invitations were extended to two local school districts to participate in the study, one of which accepted the invitation. The health and safety coordinator of the participating school district became the central contact for communication to the schools and the administration, and arrangements for access, and permission to monitor study rooms, were made through individual school principals. There were 39 elementary schools in the participating school district, representing a range of ages, construction styles, and placements within commercial, residential and semi-agricultural environments. The school district had a centralized purchasing department for furnishings and floor coverings, centralized maintenance coordination for building repairs, and control over custodial cleaning protocols. Three rooms were systematically selected for study at each school. None of the rooms were identified a priori as being "problem" rooms by the occupants, the school district management or the investigator. Only one room was excluded from the analysis because the school was closed due to snow conditions on the sampling day, giving a total of 116 rooms. The three rooms were chosen to be representative of the varied environmental conditions within the school. Indoor environments were not considered to be homogeneous because the majority of schools had been subjected to continued expansion, including the construction of additional wings, annexes or temporary 28 buildings called portables. These expansions incorporated a variety of heating systems which, within each school, could include any or all of the available systems; for example, no forced air (hot water radiators), individual or unit ventilators, or centralized and multiple air handling systems, including constant or variable air volume units. These differing heating systems influenced other forms of natural ventilation, such as the possibility of or need for opening doors and windows. Schools were randomly assigned to a fall, winter or spring sampling period. The seasonal periods were defined by historic temperature ranges using records from Environment Canada, and coincided with horticultural changes in vegetation. Schools were not typically used for classroom instruction from the end of June to the first Monday in September. There were additional breaks from normal classroom schedules for a two-week period which included the Christmas and New Year's holidays, and for a one week period in March for spring break. During the breaks in instructional time, custodial staff thoroughly cleaned and waxed the floors, or shampooed carpets. Maintenance projects, such as installing new flooring or painting, were typically done during non-instructional time. The fall sampling period, as determined by the school year and seasonal changes, commenced two weeks after the start of the school year and ended with the first hard frost, which normally occurs in this climate within the first two weeks of November. The winter sampling period commenced two weeks after the resumption of school in January to the end of February. The spring sampling period began in April and ended the first week of June. Table 2 summarizes the seasonal sampling periods. Table 2. Summary of sampling schedule of schools over six sampling periods Fall Winter Spring n=6 (Sept. 17/96-Oct. 31/96) n=6 (Jan. 16/96-Feb. 22/96) n=7 (Apr. 16/96-June 6/96) n=7 (Sept. 16/97-Nov. 6/97) n=6 (Jan. 14/97-Feb 19/97) n=7 (Apr. 8/97-May 29/97) 3.2 Classroom Sampling and Monitoring Schedule Each classroom included in the study was visited twice, the first time to measure the air exchange rate, and the second to complete all other measurements and observations. Because concerns were raised by school administrators regarding the use 29 of sulphur hexafluoride (SF6) gas to measure the air exchange rate (see page 40 for description and protocol), the tracer gas decay procedure was conducted during periods of time when students and teachers were not present in the classroom. The tracer gas studies were performed during the unoccupied periods immediately preceding the sample collection period, when seasonal heating and/or ventilation were in use. For the fall sampling period, the classrooms were evaluated in the last two weeks of August; for winter, the last two weeks of December; for spring, the week of spring break in March. During the second visit to the classroom, observations were made of building age, siting, construction, indoor and outdoor environmental conditions, furnishings, general maintenance, landscaping, number of occupants and occupancy patterns. Measurements were taken of relative humidity, equilibrium relative humidity, temperature, CO2 concentration, bacterial and fungal bioaerosols, total suspended particulate matter and ventilation air flows. Care was taken to minimize differences between rooms that may have been due to transient conditions, such as the reactivation of air handling systems after a weekend. Sampling days were standardized to fall on a Tuesday, Wednesday or Thursday. The sampling schedule was examined, and excluded any week that began with a Monday holiday, or that had a holiday or professional day within the sampling week. Individual class schedules were examined to determine, within the Tuesday to Thursday schedule, which days represented normal occupancy patterns for the room, to avoid sampling on a day when the class was on a field trip. There were a few unavoidable schedule clashes that coincided with classes being recessed unexpectedly for general assemblies to observe, for example, Hallowe'en or Valentine's Day. The differences in hours of occupancy were accounted for by using the actual hours of occupancy as an independent variable for statistical analysis. 3.3. Sampling Strategy Three classrooms were chosen for study in each school. The rooms were chosen using the flowchart in Figure 1. 30 gure 1. Criteria for room selection within schools •<^Yes Does school have portables? Select one portable No Does school have more than one air handling system? > < Yes Select one classroom from each system No Does school have more than one floor? • X : Yes Select one classroom from each floor No Does school have more than one wing? ^ " ^ ^ Yes Select one classroom from each wing. No Does school have an annexe? - K Yes Select one classroom from the annexe. No Do the rooms face different compass directions? " • < Yes Select one classroom from each direction. Is the room occupied for the majority of the day? Select classroom that is normally occupied. Does teacher give permission to be in room? No • X Yes Select room where permission is granted. 31 3.4. Power Estimates and Statistical Analyses Bioaerosol data were available from previous studies undertaken during summer and winter at different sites within the province (Bartlett 1993; Bartlett and Nakahara 1994; Bartlett 1995) suggesting the value of testing for seasonal variation in bioaerosol concentrations. The power of the study, as designed, was estimated for a comparison of one determining factor (season), given the fixed number of schools. The school district that volunteered for the study had 39 elementary schools that were available for enrollment in the study. A power calculation (NCSS/PASS®, Version 1.0-8/91, Kaysville, UT) was performed in order to determine the power that the study would have to reject the null hypothesis of no difference in fungal concentrations taken during different seasons. For the power calculation, colony counts from the previous studies were log transformed to the natural log and the resulting geometric standard deviations were used for the estimate of variance. It was estimated that, given a geometric standard deviation of 1.06 from the available data, there would be a power of approximately 0.65-0.72 (a = 0.05) to detect a mean difference of 50% in colony counts for 40 sites. This would translate into a statistically significant difference in the following example geometric means: 700 CFU/m (summer) and 300 CFU/m (winter) with standard deviations in the range of 600 and 300 respectively. As the results later indicated, rooms (n=l 16) rather than sites (n=39) were the primary unit of analysis, suggesting that the actual power of the study was higher. Daily observations and measurements were recorded using a worksheet developed for this project. A sample worksheet is appended (Appendix I). A data input form was developed from the daily worksheets using Epi Info version 6.04b (Centers for Disease Control, Atlanta, GA). Data were analysed using SPSS version 7.5 for Windows (SPSS, Inc., Chicago). Data were cleaned by examining frequency tables and ranges for all variables and randomly rechecking entries for accuracy of transcription. Descriptive statistics were calculated for all variables (means, ranges, standard deviations, and counts for categorical data). Data were examined for normal distribution by observation of histograms and statistically by computing Lillefor's probabilities for Kolmogorov-Smirnov one sample 32 tests against a normal distribution. Data that approximated log normality were log transformed to the natural log (base e) to allow the use of parametric statistics. Comparison of means (null hypothesis of equal means) was performed by Student t or of multiple means by one way analysis of variance (ANOVA) using Scheffe's test for post hoc comparisons. Potential explanatory variables were grouped as shown in Tables 3-5 and included site specific outdoor environment variables (Table 3), fixed building characteristics (Table 4) , and room (as occupied on the day of testing) variables (Table 5). Table 3. Site specific, independent environmental variables Variable name Type of variable Description Season Categorical fall; winter; spring (see Table 2 for definitions) Mean outdoor temperature Continuous Source: Environment Canada Measurement: temperature averaged over 24-hour period (°C) Mean outdoor relative humidity Continuous Source: Environment Canada Measurement: average of maximum and minimum RH measured on sampling day (%)• Barometric pressure Continuous Source: Environment Canada Measurement: barometric pressure at noon on sampling day (kPa) Rain Continuous Source: Environment Canada Measurement: precipitation for 24-hour period (mm) Snow (any) Categorical no; yes Wind speed Continuous Source: Environment Canada Measurement: average wind speed for 24-hour period (km/hr) Wind direction Categorical Source: Environment Canada Measurement: 360° compass direction divided into 16 sectors over 24 hour period 33 Table 4. Variables describing fixed characteristics of buildings and sites Variable name Type of variable Description Location of room Elevation above sea level Building age Building construction Categorical Continuous Continuous Categorical main building; annexe; portable classroom Source: Geographical survey map Measurement: metres Source: School records Measurement: year of construction masonry; concrete; wood frame; cinder block; portable (aluminum siding) Central forced air system Categorical no; yes Stand alone or unit ventilator Categorical no; yes Presence of carpet in classrooms Categorical Ceiling treatment Categorical none; < 49% floor surface carpeted; > 50% floor surface carpeted plaster or wood; cellulose or acoustical tile Room volume Continuous Source: room measurement in m School siting Categorical on major roadway; quiet residential street; adjacent to trees or agriculture 34 Table 5. Room variables measured or observed on test day Variable name Type Description Number of occupants Continuous Source: average head count over the course of the school day Occupant age Continuous Source: average age of students based on grade level of class % of day room empty Continuous Source: observation of time room empty during school day Activity level Continuous Source: observation of time occupants actively moving around Quiet sitting Continuous Source: observation of time occupants quietly sitting in room Room use pattern Categorical room never empty; empty <24% of day; empty 25-49% of day; empty > 50% of day Cleanliness of room Categorical very dusty; dusty; "lived-in"; cleaner than norm; excellent cleaning; meticulous cleaning (Kozak et al. 1985) Live animal resident Categorical no; yes Aquarium Categorical no; yes Plants/plant material Categorical no; yes Clutter Categorical normal clutter; unusually cluttered Moisture Categorical no signs of moisture; old water stains on room surfaces; current signs of moisture 35 3.5 RESULTS 3.5.1 Descriptive Statistics of Sites All 39 schools were included in the study. The schools varied in size and student population. The total number of students in the district was 13,851. The smallest school had 127 students; the largest, 664. The district mean was 355 with a median of 371. The average number of students per classroom was 23.9 (sd 5.1) and represented kindergarten to grade 7 (5 - 13 year olds). Information and measurements were collected on environmental, building and room characteristics that were thought to contribute to the outcomes of interest - fungi, bacteria, and CO2. Table 6 summarizes the environmental characteristics of the sites, by season. Table 6. Seasonal environmental characteristics of sampling periods Variable Winter" n = 35 Spring" n = 42 Fal l c n = 39 *p-value Mean (sd) [range] Mean (sd) [range] Mean (sd) [range] Temperature (°C) 3.6 (3.58)t [-6.7-9.8] 11.9 (3.20) [7.2-19.2] 11.4(2.96) [4.3-17.1] < 0.001 Relative humidity (%) 80.5 (9.33) [59 - 97] 71.4 (7.82)t [59 - 89] 79.1 (9.36) [66 - 95] < 0.001 Barometric pressure (kPa) 101.6(1.07) [99.1 - 103.2] 101.6 (0.80) [99.1 - 102.8] 101.3 (0.87) [99.5 - 102.5] NS Rain (mm) 5.4 (7.82) [0-31.8] 3.8 (6.56) [0 - 29.6] 5.4 (8.66) [0 - 42.4] NS Wind speed (krn/hr) 11.6 (6.02) [0.6-24.9] 12.0 (4.07) [5.0-20.1] 11.4 (6.24) [2.7-29.0] NS *ANOVA. t Group significantly different from other groups by Scheffe post hoc procedure. " Mid January to end of February. b Mid April to beginning of June. c Mid September to beginning of November. Mean spring and fall temperatures were very similar, with significantly lower winter temperatures. The relative humidity in spring was significantly lower than in fall or winter. 36 Wind direction was a categorical variable. The prevailing wind direction was from the east on 49% of study days and did not vary with season. Fixed characteristics of schools are shown in Table 7. The core school buildings were built between 1915 and 1995, with 51% of schools being built between 1950 and 1970. Table 7. Description of fixed characteristics of classrooms by building age Building variables Original Original Original construction construction construction <1949 1950-1970 1971-1995 Count (%) Count (%) Count (%) Total number of study rooms 30°/l 16 (26%) 59*7116 (51%) 277116(23%) Construction materials Masonry 8 (27%) 0 2 (7%) Concrete/cinderblock 4 (13% 6 (10%) 5 (19%) Wood frame 18 (60%) 53 (90%) 3 (11%) Prefabricated 0 0 17 (63%) Urban placement On major arterial road 19 (63%) 23 (39%) 9 (33%) On residential street 9 (30%) 28 (47%) 5 (19%) Heavily treed/agricultural 2 (7%) 8 (14%) 13 (48%) Elevation (metres) 100-299 m 4(13%) 15 (25%) 6 (22%) 300-499 m 18 (60%) 43 (73%) 18(67%) 500-699 m 8 (27%) 1 (2%) 3 (11%) Location of classroom tested First floor 14 (47%) 54 (92%) 22 (81%) Second floor 14 (47%) 5 (8%) 5 (19%) Third floor 2 (7%) 0 0 Availability of ventilation Central forced air 10 (33%) 21 (36%) 6 (22%) Unit ventilators 5 (17%) 7 (12%) 5 (19%) Axial fan 10 (33%) 9 (15%) 16 (59%) None 5 (17%) 22 (37%) 0 Ceiling treatment Cellulose acoustical tile 22 (73%) 55 (93%) 26 (96%) Other 8 (27%) 4 (7%) 1 (4%) Amount of carpet in rooms None 6 (20%) 14 (24%) 1 (4%) < 49% of floor 12 (40%) 24 (41%) 2 (7%) > 50% of floor 12 (40%) 21 (36%) 24 (89%) " Includes 1 room in an annexe. b Includes 2 rooms in annexes. c Includes 17 portables. 37 Classrooms within a single school could be built of different construction materials, particularly when rooms were added as need or finances allowed. Seventeen schools used prefabricated, portable classrooms to accommodate the student population. The presence of centralized ventilation, carpet and cellulose ceiling tiles varied within each school. A grading system designed by Kozak et al. (1985) was used to describe the amount of plant material near the building. The landscaping at all of the schools was judged to be minimal and without excessive shade or shrubbery close to the buildings, although 23 (20%) of the rooms were near treed parks or semi-agricultural areas. Room characteristics that varied on the day of testing included age and number of occupants, room use patterns, presence of plant material, live animals or aquaria, and the use of passive forms of ventilation such as opening windows and doors. Transient ventilation characteristics will be discussed further in Chapter 4. The indoor environmental characteristics of interest are listed in Tables 8 and 9. Table 8. Characteristics of room on test day by season (categorical variables) Room variable Winter Spring Fall n=35 n=42 n=39 Count (%) Count (%) Count (%) Presence of aquaria Yes 5 (14%) 5 (12%) 3 (8%) No 30 (86%) 37 (88%) 36 (92%) Presence of plants Yes 10 (29%) 18(43%) 17 (44%) No 25 (71%) 24 (57%) 22 (56%) Presence of live animals Yes 3 (9%) 4 (10%) 2 (5%) No 32 (91%) 38 (90%) 37 (95%) Unusually cluttered Yes 9 (26%) 11(26%) 7(18%) No 26 (74%) 31 (74%) 32 (82%) Materials other than school furniture Yes 19 (54%) 25 (60%) 20 (51%) No 16(46%) 17 (40%) 19 (49%) Signs of moisture None 20 (57%) 18(43%) 18(46%) Stains, currently dry 12 (34%) 22 (52%) 17 (44%) Signs of current moisture 3 (9%) 2 (5%) 4 (10%) 38 Table 9. Characteristics of room on test day by season (continuous variables) Variable Winter n=35 Mean (sd) [range] Spring n=42 Mean (sd) [range] Fall n=39 Mean (sd) [range] *p-value Occupants per 100 m3 room volume 9.0 (2.27) [3 - 14] 8.7 (2.34) [3-13] 9.3 (2.36) [5-16] NS Age of occupants 9.3(1.88) [6-12.5] 9.71 (2.34) [5.5-12.5] 9.27(1.64) [6-12.5] NS Room empty (%ofday) 28.1 (17.6) [0 - 84] 26.2 (12.38) [0 - 63] 20.8 (13.38) [2 - 73] NS % of day spent sitting by occupants 45.7 (24.49)t [0 - 84] 59.6 (16.86) [0 - 90] 66.4(15.92) [15-91] O.001 *ANOVA. t Group significantly different from other groups by Scheffe post hoc procedure. The measurements and observations of the variables describing room use characteristics were fairly evenly distributed over the study period. However, students were observed to spend more time sitting quietly at their desks in the spring and fall than in the winter sampling period (p < 0.001). A grading system designed by Kozak et al. (1985) was used to describe the quality of the housekeeping in the rooms. In general, the rooms looked "lived in" without heroic attempts to keep surfaces clean of dust. Rooms were vacuumed and biodegradable garbage was removed daily. Only 27 (23.3%) of the rooms were described as "excessively cluttered" with paper products or furniture, while half of the rooms (55%) contained extraneous materials other than school furniture and books. Other materials included science fair projects, art projects, seedlings, upholstered furniture and stuffed toys. 3.6 SUMMARY OF SITE CHARACTERIZATION Classrooms tested represented a range of building styles which included masonry, concrete, wood frame, cinder block and pre-fabricated portables. The rooms varied with respect to furnishings, with 33% of rooms having less than one half of the room carpeted and 49% being entirely carpeted. Cellulose or acoustical tile predominated as a ceiling 39 treatment in most (89%) rooms. Only a minority of rooms had plants, live animals or aquaria as part of their furnishings. Thus, potential substrates for micro-organism growth and sources of other indoor contaminants varied widely among classrooms. The building, site, environmental and room use characteristics were evenly distributed during the sampling periods. The activity level of the students appeared to vary seasonally (percent of day spent sitting was greater in the fall and less in the winter p < 0.001) which may be related to the need for more diverse activities during inclement weather. Outdoor environment characteristics varied with season as expected. 40 CHAPTER 4 VENTILATION SYSTEMS, AIR EXCHANGE RATES, COMFORT PARAMETERS 4.1 INTRODUCTION TO METHODS FOR MEASURING OR ESTIMATING VENTILATION RATES 4.1.1 Air Exchange Rate by Tracer Gas Decay (SFk) The air exchange rate is the volumetric flow rate divided by the room volume. This can be conceptualized as the rate at which dilution air "flushes" the room. Assuming that supply air mixescompletely with the room air, this rate can be directly calculated by measuring the supply air flow entering a room by forced air. However, it is difficult to measure a flow for a naturally ventilated room which has no mechanism to exchange the room air. The ability of fresh air to dilute contaminants in a room can be assessed by measuring the decreasing concentration of an introduced contaminant - for example, a tracer gas - over a period of time. An exponential function is used to describe the dilution of the tracer gas by uncontaminated air and allows the calculation of the air exchange rate. The characteristics of the ideal gas for use as a tracer gas have been described by Bearg (1993) and are listed in Table 10. Table 10. Desirable characteristics of tracer gas for IAQ testing (Bearg 1993) Measurable at very low concentrations Inert, nonpolar, and not absorbed Nontoxic, nonallergenic Nonflammable and nonexplosive Not a normal constituent of air Measurable by portable equipment Measurable by a technique that is free of interference by substances normally in air Relatively inexpensive 41 Sulphur hexafluoride (SFe) meets all of these characteristics and was chosen for this study. However, although SF6 is inert and non-toxic at the parts per million (ppm) concentrations used for testing, it was felt that there would be resistance from teachers and parents to unnecessary chemical exposures to room occupants. In rooms without mechanically supplied air, dilution of the tracer gas sometimes required more than 90 minutes. It would have been difficult to maintain constant room conditions for prolonged periods of time during the course of a normal school day. Therefore, tracer gas decay measurements were taken during unoccupied periods when the ventilation system was operating. This resulted in measurement of room ventilation characteristics during a period of time unaffected by the multiplicity of modifications to the ventilation caused by occupancy (windows or doors open, movement into and out of room, etc.). 4.2 METHODS FOR MEASUREMENT OF VENTILATION PARAMETERS 4.2.1 Methods for Ventilation Assessment For each room, the presence or absence of ventilation was assessed visually by locating supply or exhaust vents. Vents were further investigated to determine if they were connected to an air-handling unit (AHU). The vent and ducting were checked for overall cleanliness by visual inspection of the grid covering the vent and the duct as far as a xenon-lamp (Eveready ® Industrial, St. Louis, MO) flashlight beam penetrated. If present, the central AHU was inspected and the conditions of the air filters and air intake vents were noted. The condition of the mechanical room was rated on a categorical scale for presence of standing water or storage of chemicals. The velocity of supply and exhaust air was measured using a TSI model 8360 VelociCalc Plus (TSI, Inc. St. Paul, MN) thermal anemometer using a modified grid traverse across the face of the vent (ACGIH, 1992). The modification of the grid traverse was to divide the face of the vent into equal sections and measure the velocities at the centre of each section. The number of measurements varied with the size of the vent, with a minimum of four readings for the smallest vents (8.9 cm x 24 cm) and up to 12 readings for the largest vents (40 cm x 75 cm). Corresponding flows (cubic metres per 42 second) were calculated from velocities (metres per second) by multiplying the average velocity over the face of the vent by the area of the vent. The VelociCalc Plus TSI model 8360 was factory calibrated. At the beginning of the project the velocity measurement of the VelociCalc Plus was checked against the velocity pressure measured using a Pitot tube and magnehelic gauge in an experimental fan plus duct construction made for this purpose in the UBC Occupational Hygiene Programme laboratory. Three fan speeds were used to create velocity pressures ranging from 0.09 - 6.1 mm water, corresponding to velocities from 4-10 m/s. The relationship of velocity pressure to velocity is given in equation 1 (ACGIH 1992). V = 4.043 JVP L i . Where: V= velocity in m/s VP = velocity pressure in mm of water The VelociCalc velocity measurements were 13% lower than the calculated velocities. The approximate error of the Pitot tube at 4 m/s (the limit of detection of the Pitot tube) is given to be 15% (ACGIH 1992). Accordingly, the VelociCalc velocity measurements were used without further correction. The limit of detection for the VelociCalc velocity measurements was taken to be 0.2 m/s, as per the manufacturer's specifications. For field measurements, the direction of the flow of air through the vents was determined visually by noting the direction in which indicator smoke moved in relation to the vent. The source of indicator smoke was a lit incense stick. 4.2.2 Measurement of Air Exchange Rate by Tracer Gas Decay (SFt) The tracer gas decay method used for this study was a modification of ASTM Standard E741-83 described by Bearg (1993). The tracer gas decay protocol was as follows. The doors and windows were closed in the room to be tested. A Miran 1 A, portable infrared spectrophotometer 43 (Foxboro, Nortex, ON) was set to 10.7 urn wavelength, 0.75 m pathlength, with Teflon® tubing attached to the air intake to standardize the sampling height to 82 cm above floor level. The full-scale range of the analogue display allowed concentrations of SF6 in the range of 0.5 - 5 ppm to be measured. The Miran analogue scale was calibrated prior to the tracer gas sampling periods (section 3.1.2) by introducing incrementally increasing micro litre volumes of 100% SF 6 into a closed loop through the detector (total volume 5.6 L). The precision of the test method was determined by recharging the room with SF6 and reanalysing the decay curve in 10% of the sample rooms. The mean coefficient of variation of the 12 duplicated tests was 13% (range 1 - 30%). For the test, a 43 cm diameter portable fan (AirKing CML 20P-3) was placed on the floor, with the exhaust side facing into the room, and at a 45° angle to the wall. The fan was run at full speed for at least 15 minutes to ensure that the room air was reasonably well mixed. SF6 was introduced at the intake side of the fan by bolus injection of gas taken from a Tedlar® bag (SKC Inc., Eighty four, PA) using two 60 mL disposable, plastic syringes (Becton Dickinson, Mississauga, ON). Enough SF6 was introduced into the room to achieve a final concentration in the room air of 4 - 5 ppm (120 - 240 mL depending on room size). The SF6 was considered well mixed with the room air when the Miran detector reading had stabilized. The fan through which the bolus injection was made was turned to the lowest setting for the remainder of the readings. The analogue scale deflection was recorded over a 90-minute period at one-minute intervals during the mixing stage and at approximately five-minute intervals when the readings had stabilized. The data points were plotted on semilog graph paper to visualize the shape of the decay curve. The resulting straight-line relationship between SF6 concentration and elapsed time was an indication that the assumption of well-mixed room air was met, an assumption required by the mass balance equation used to calculate the air changes per hour. The number of air changes per hour for the room was calculated using the following simplified equation (equation 2) describing the decrease in tracer gas concentration over time (Bearg 1993). 44 1 I = 2 V AT 1-2 J Where: I = air changes per hour T = time in hours In = natural log C\ - tracer gas concentration at time 1 Ci = tracer gas concentration at time 2 The concentrations C\ and C? were taken from the straight line portion of the semilog-plotted graph. This section of the graph best represented well-mixed room air conditions. 4.2.3 Measurement of Carbon Dioxide (CO?") Concentration CO2 concentrations in ppm were measured using a YES 203 (Young Environmental Systems, Inc., Richmond, BC), a non-dispersive, infra-red monitor. CO2 concentrations were averaged over five minute intervals and stored in memory. The monitor was placed in the room prior to the start of the school day, and was removed approximately two hours after the room occupants had left at the end of the school day. The monitor was placed in an area of the room that was assumed to have well mixed air, but was not in a draft from ventilation sources. These placement requirements were generally met by placing the monitor on top of a partition dividing the classroom from the cloakroom. The monitor was returned to the laboratory at the end of each day and downloaded using a Pentium 75 personal computer (PC) and TrendReader™ software (Version 1.0. ACR Systems, Inc.). The ASCII data files were then read into Microsoft Excel 97© spreadsheets for further analysis. For analyses in which a single daily average C0 2 concentration was required, the average concentration over the six hours per day the room was occupied for teaching purposes was used. The YES 203 was calibrated at the beginning of each sampling period (fall, winter spring) and at the end of the last sampling period using a two point calibration curve using certified clean carrier gas (0 ppm) and CO2 calibration gas (900 ppm) (Praxair, 45 Mississauga, ON). The manufacturer's specifications for the YES 203 stated an accuracy of ± 50 ppm for the instrument. 4.3 METHODS FOR MEASUREMENT OF COMFORT PARAMETERS 4.3.1 Temperature and Relative Humidity Room temperature and percent relative humidity (RH) were measured using two external channels of the YES 203. The temperature probe used a NTC thermistor and the RH probe used a sulfonated polystyrene wafer (Young Environmental Systems, Inc., Richmond, BC). Data were continuously averaged (5 minute intervals) and stored in memory until downloaded using the same procedure as for the CO2 data. Data for the outdoor temperature, RH and meteorological conditions for the corresponding sampling day were taken from the records of the nearest Environment Canada meteorological station (maximum 10 km away from sampling site). In order to assess the accuracy of the YES RH and temperature probe, a second measurement of room temperature using a sling psychrometer was taken at 12:30 p.m. each sampling day. The sling psychrometer uses two thermometers, one wetted with a cotton wick, the other dry. The relationship of wet to dry bulb readings is used to determine the RH using a psychrometric chart (Pita 1989: 176). The average of 106 sling psychometric readings was 45.3% compared to an average of 41.8% for the corresponding YES RH measurements. The Pearson correlation coefficient for comparison of the two measurement tools was 0.90. The average of 102 dry bulb temperature measurements was 21.7 °C compared to an average of 22.4 °C for corresponding YES temperature measurements. The Pearson correlation coefficient for comparison of the two measurement tools was r=0.88. As the YES RH and temperature measurements were not consistently biased lower or higher than the sling psychrometer measurements, no additional correction factors were applied to the YES data. 4.3.2 Equilibrium Relative Humidity The equilibrium relative humidity (ERH) was defined for this study as the RH, isolated from the room RH, at a wall surface, expressed as a decimal. The ERH was measured using portable, battery operated thermohygrometers (Cole Parmer Instruments, 46 G37101-00, Niles, IL). The thermohygrometers were calibrated at the beginning of the study by suspending the units over Petri dishes containing saturated salt solutions in sealed, air-tight containers. The four salt solutions used were MgCb^EbO (aw at 25°C 0.33), Mg(N03)2-6H20 (aw 0.53), NaCl (aw 0.75) and KN0 3 (aw 0.94) (Trailer and Christian 1978). The time required to equilibrate was experimentally determined by recording the change over time of the RH display on the thermohygrometer enclosed with the calibration saturated salt solutions and from laboratory mock-ups of the thermohygrometer affixed to a wall as described below. The equilibrium relative humidity at two points in the room was measured by sealing a thermohygrometer against the wall surface using an air-tight plastic box with a flange that allowed it to be fixed to the wall with vinyl tape. The standardized time to equilibration chosen for the study was four hours. The two points in the room were (1) on or near to the windowsill and (2) on the wall nearest the corridor door or remote from the windows, whichever applied. 4.3.3 Total Suspended Particles Glass fibre filters (Type AE, Gelman Sciences, Ann Arbour, MI) were preconditioned for 48 hr in a humidity (50 ± 5%) and temperature (22 ± 2 °C) controlled room. Filters were weighed three times each on a Sartorius micro-balance (Sartorius, Southfield, MI) after eliminating surface static charge by neutralization with an alpha particle emitter. The accuracy of the micro balance as given by the manufacturer's specifications was + 0.005 mg. The filters were mounted into polystyrene, three piece cassettes (SKC, Eighty-four, PA) with cellulose backing pads and the cassettes sealed with vinyl tape. Battery operated SKC sampling pumps (44XR) were calibrated at the beginning and checked at the end of each sampling period using a calibrated rotameter (Matheson, Montgomeryville, PA) to a flow of 2 L/min. The rotameter was calibrated against a primary standard using a biuret and soap solution. Two pumps were used inside the classroom and two pumps were used outside the classroom nearest the source of fresh air for the room. Cassettes were attached to the pumps by vinyl tubing and hung from tripods 81 cm above floor or ground level. The sample time was recorded to calculate 47 flow rate and was approximately 8 hr (range 6 - 8.8 hr). A total of 59 (13% of samples) field blank filters mounted in cassettes were taken to the sample site, opened for 8 hr and treated identically to samples with the exception that no air was drawn through the filter. All filter cassettes were returned to the laboratory at the end of each sampling day and stored at 4°C prior to reconditioning for 48 hr in the humidity and temperature controlled room, prior to post-weighing as described for the pre weight. The storage conditions were in consideration of future chemical speciation of the collected dust. All of the filters from a sampling season were batched for reconditioning and weighing. Three filters from the same lot as the sample filters were retained in the weighing room as a laboratory control and were reweighed on each weighing day. The field blank filters were weighed to determine the limit of detection of the method. The LOD was taken as three times the standard deviation of 59 pre and post weights. The mean (0.007 mg) plus three times the standard deviation (0.04) of the net weights was used to calculate the LOD. For each sample the LOD was determined by dividing the LOD by the volume of air collected for that sample. The particulate matter data was normally distributed. Filter weights that were below the limit of detection were assigned a value of the LOD divided by the V2 (Hornung and Reed 1990). 48 4.4 VENTILATION VARIABLES AND DATA TRANSFORMATION FOR STATISTICAL ANALYSIS The distributions of the average CO2 concentrations, air exchange rates and air flows were positively skewed. These data were tested for normality by the Lilliefors test and the null hypothesis of normality was rejected (p < 0.001). These variables were log-transformed to the natural log prior to analysis to allow the use of parametric statistics. Ventilation and room use variables analyzed are listed in Table 11. Table 11. Variables describing ventilation and room use Variable name Type of variable Description Ventilation operating on test day Categorical no; yes Supply air flow Continuous Source: VelociCalc Plus Measurement: m3/s Exhaust air flow Continuous Source: VelociCalc Plus Measurement: m3/s Type of mechanical ventilation Cateogrical None; central air supply; unit ventilator; supply or exhaust fan. Mechanical room Categorical dry; wet; excessive storage of chemicals; no centralized supply air Air changes per hour (In) (SF6 tracer gas decay) Continuous Source: Calculated from SF6 tracer gas decay protocol Abbreviation: ACH (unoccupied) Outside door Corridor door Categorical Categorical no outside door; outside door open intermittently; outside door open <24% of day; outside door open >25% of day open < 49% of day; open > 50% of day; no corridor door (portable classroom) Windows Categorical closed all day; open < 49% of day; open> 50% ofday Room use Categorical room never empty; empty <24% of day; empty 26-49% of day; empty > 50% of day 49 Table 12 lists the variables used to describe CO2 and comfort parameters in the room on test days. Table 12. Carbon dioxide and comfort parameter variables Variable name Type of variable Description Outdoor C 0 2 Continuous Source: YES data logger Measurement: outside measurement in ppm averaged over 30 minutes Indoor C 0 2 (In) Continuous Source: YES data logger Indoor measurement in ppm averaged over 6-hour period room occupied with students & teachers Mean indoor temperature Continuous Source: YES data logger Measurement: temperature averaged over 6- hour occupied day (°C) Equilibrium relative humidity at Continuous Source: Thermohygrometer windowsill Measurement: RH at the end of 4 hours. Equilibrium relative humidity at Continuous Source: Thermohygrometer corridor door Measurement: RH at the end of 4 hours. Mean indoor relative humidity Continuous Source: YES data logger Measurement: RH averaged over 6-hour occupied day (%). Total Suspended Particles Continuous Source: Air drawn through glass fibre filter. Measurement: Gravimetric analysis (mg/m3). 50 4.5 RESULTS: 4.5.1 Air Exchange Rates, Ventilation, Carbon Dioxide (CO?) and Comfort Parameters It became apparent in the school rooms examined that in addition to the fixed characteristics of the ventilation system, it was necessary to evaluate the additional modifications to the ventilation on a daily basis. Some changes were voluntary (opening windows or doors for ventilation) and some were systematic (air handling unit was not supplying ventilation) on the test day. In this study, 37 (32%) of rooms were potentially served by central forced air; on the test day only 29 systems were supplying ventilation air. Unit ventilators were present in 17 (15%) rooms and were operating in 15 rooms on the test day. Fans for supply or extraction of air were available in 35 (30%) rooms but only 18 were in use on the test day. The 54 (47%) rooms that were not mechanically supplied with ventilation air on the test day were considered to be naturally ventilated. Windows were opened for all or part of the day less often in the winter (23%) than in either spring (43%) or fall (59%) (x 2 pO.Ol). Examples of the relationship between the type of ventilation and the concentration of CO2 throughout the day are shown below. Arrows mark the periods when students entered or left the room. Figure 2 illustrates a room with radiant hot water heat and no mechanically supplied ventilation air. Figure 3 illustrates a room with radiant hot water heat, with supply air ventilation system operating on the test day. Figure 4 illustrates a room with a centrally supplied, forced air system that provides both heat and ventilation air. 51 Record # 69 C02 2500 2000 E a. a. CM o o 1500 1000 500 open windows students leave leave for gym -C02 \ enter 21 students + teacher radiant heat, supply fan broken # .«? .»?> ft* .<?• .<N .4 .<£ .<#> .«§> .<£ # • ^ .<> v> <?- •O' <& \ N Time Figure 2. Example of CO2 concentrations in room with radiant hot water heat, but no supply air. Record #57 C02 1600 1400 1000 E a. a. CN o o 800 600 200 additional 22 ..buddies leav ^ leave for day leave tor music — C02 enter 22 students + 1 teacher radiant heat + separate supply fan %v <by <b- Q v o>- q>- ,J>? » v v » v " <VV <VU <V? »>v v " ^ V V Is V ^ y -b- fcv fc-Time Figure 3. Example of CO2 concentrations in room with radiant hot water heat and mechanically supplied air. 52 It is readily apparent that the decay rates, and the maximum CO2 peaks, vary with the availability of ventilation. In Figure 2, the decay curve is flattened compared to either of Figures 3 or 4. The maximum CO2 peak in the fully ventilated room (Figure 4) was about 850 ppm compared to peaks above 2000 ppm in the non-ventilated room (Figure 2). Figure 5 illustrates the mean difference in air exchange rates between naturally (no mechanical ventilation) and mechanically ventilated rooms measured during using SF6 tracer gas decay (significant difference in means, p < 0.001). 53 3 o CD Q. (/) CD D ) C (0 _£Z o < 3 2 1 0 -1 -2 -3 4 -4 -5 -6 — ^ ^ ^ ^ ^ ^ ^ ^ ^ CP N = 53 Natural 62 Mechan ica l Natural vs. Mechanical Ventilation Figure 5. Comparison of air exchange rates in naturally vs. mechanically ventilated rooms. Although fixed building characteristics were randomly sampled in different seasons, there were strong seasonal associations with CO2 concentration, and comfort parameters as is shown in Table 13. There were no significant seasonal differences between air exchange rates. Mean temperature was significantly lower in winter, and relative humidities were all significantly different from each other for each season. 54 Table 13. Seasonal distribution of ventilation, comfort parameters and indoor CO2 Season n Ventilation rate (unoccupied)" ACH* GMC (GSD)d c o 2 ppm GM (GSD) Indoor Relative Humidity % Mean (sd) Indoor temperature °C Mean (sd) Fall 38 0.26 (6.20) 908 (1.39) 47.7 (6.19)t 22.9(1.61) Winter 35 0.34 (3.01) 1185 (1.35)t 36.9 (6.98)t 21.6(1.22)1 Spring 42 0.46 (2.86) 816(1.46) 41.6 (7.44)f 22.8(1.92) *p-value NS O.001 <0.001 <0.01 *ANOVA. f Group significantly different from other groups by Scheffe post hoc procedure. a Measured by SF6 decay. b Air changes per hour. c Geometric mean. d Geometric standard deviation. Table 14. Ventilation and comfort parameters by grouped building age or temporary building status. Year of n Ventilation c o 2 Indoor Indoor construction rate Relative temperature of room (unoccupied)" Humidity ACH* ppm % °c GMc(GSD)rf GM (GSD) Mean (sd) Mean (sd) Before 1950 30 0.42 (2.64) 905.8 (1.47) 44.5 (7.81) 22.6(1.68) 1951-1974 59 0.31 (3.87) 939.8 (1.38) 41.4 (8.57) 22.4(1.82) After 1975 10 1.20 (4.1 l)t 804.1 (1.51) 43.2 (4.32) 22.5(1.01) Portable 17 0.18(4.94) 1155 (1.54) 40.8 (8.47) 22.4(1.80) classroom *p-value <0.005 NS NS NS fGroup significantly different from other groups by Scheffe post hoc procedure. a Measured by SF6 decay. b Air changes per hour. c Geometric mean. d Geometric standard deviation. There was a statistically significant relationship between the year of building construction and the measurement of air exchange rate, with portable classrooms (primarily constructed in the late 1980's and early 1990's) having the lowest rates and buildings built after 1975 having the highest. There were no statistically significant differences in temperature or relative humidity in buildings of different ages as shown in Table 14. 4.5.2 Assessment of Mechanical Ventilation Systems The 37 mechanical rooms housing forced central air heating units were examined. The majority (78%) were found to be relatively clean and dry; 8% had wet floors or leaking water pipes in the mechanical room, and 14% used the room for the storage of whiting (white powder used to mark playing fields). However, none of the air handling units (AHU) was seen to have signs of moisture within the filters or fans. Filters in AHU's varied with respect to the amount of accumulated dust, but no significant pressure drops across the filter were found. There were no statistically significant relationships between the condition of the mechanical rooms and any of the outcomes of interest (fungal, bacterial or CO2 concentration). Visual inspection of the ventilation ducting confirmed that none was lined with fibrous glass. 4.5.3 Relative Humidity and Equilibrium Relative Humidity The equilibrium relative humidity measurements did not indicate systemic differences between the humidity at the window (40.9%, sd 8.49) vs. the opposite wall (40.7%, sd 8.56), nor between the equilibrium humidity at either wall and the room humidity (42.2%, sd 8.12). The mean relative humidity and both measurements of equilibrium relative humidity were significantly lower in the winter (p<0.001). In winter, the season when rooms are heated, the difference between indoor relative humidity and outdoor humidity is greatest (outdoor less indoor = 43%, vs mean difference of 30% for spring and fall). This illustrates the effect of heating outdoor air and the resultant lowering of the indoor relative humidity. For example, if outdoor air has a relative humidity of 60% at 10 °C, it contains 4.6 grams of water per kilogram of 56 relative hiunidity (the ratio of the partial pressure to the saturation pressure) falls to 28% (Pita 1989). 4.5.4 Temperature The temperature was well controlled in the majority of rooms tested, with only 8/116 (7%) having mean temperatures higher than 24 °C and only 1/116 (0.8%) with a mean temperature less than 20°C. None of the schools had air conditioning. 4.5.5 Total Suspended Particles (TSP) The concentrations of TSP were low in both indoor and outdoor samples. There were 43 (37%) indoor and 100 (86%) outdoor samples that were below the LOD. The high number of samples below the LOD resulted in a lack of sensitivity for this variable. The concentration of indoor TSP was higher than outdoor TSP (mean 0.1688 vs. 0.1089 mg/mJp< 0.001). There were no significant relationships found between indoor particulate matter concentration and variables hypothesized a priori to be of interest -namely, presence of carpet, decreased standards of cleanliness, presence of live animals and plants, increased clutter or unusually plentiful stuffed furniture or toys. Due to the large number of samples below the LOD, and the lack of apparent association with potential contributing factors, no further analyses were conducted on these data. 4.6 SUMMARY OF COMFORT PARAMETERS The schools in this district were not homogeneous as to availability or use of ventilation. Differences in relative humidity, temperature and CO2 concentrations were influenced by the availability and use of ventilation and the outside climate conditions, or season. Buildings constructed during different periods and using different building codes and architectural materials also accounted for some differences in ventilation availability. Although many schools built after 1975 were designed to be energy efficient buildings, often with mechanical ventilation, older buildings may have been retrofitted with mechanical ventilation, and in newer buildings the ventilation system may not have been adequate for the purposes for which it was being used. 57 With few exceptions the temperature and relative humidity were within ASHRAE 55-1981 guidelines. However, the 47% of classrooms exceeded the ASHRAE 62-1989 guidelines for occupant-generated CO2 concentration with mean levels above 1000 ppm. A more complete discussion of the guidelines pertaining to ventilation and comfort parameter guidelines can be found in Chapter 6. 58 CHAPTER 5 BIOAEROSOLS 5.1 BIOAEROSOLS 5.1.1 Fungal Aerosols 5.1.1.1 Mesophilic Fungi Mesophilic fungi were defined for the purpose of this study as those with optimal growth at room temperature on semi-synthetic, non selective culture media (Burge et al. 1985; Flannigan 1995). This group of fungi can potentially encompass thousands of species, the vast majority of which are saprophytic. 5.1.1.2 Thermotolerant Fungi Thermotolerant fungi were defined as those capable of growth at body temperature, or 37°C (Burge et al. 1987; Flannigan 1995). Differential incubation temperature allows for selective culture of some organisms that may otherwise be overgrown by more numerous mesophilic fungi. Some fungi, capable of growth at body temperature, may become opportunistic pathogens, particularly for immunocompromised hosts, as is the case for Aspergillus fumigatus, Aspergillus niger, Aspergillus flavus, Paecilomyces, and Scopulariopsis (Gravesen et al. 1994). 5.1.1.3 Xerophilic Fungi Xerophilic fungi were defined as organisms capable of growth on media with low water activity (Flannigan 1995). Xerophilic fungi become important in indoor air quality investigations due to their unique ability to survive and colonize building surfaces that would normally inhibit fungi that require water activities above 0.9. Xerophilic fungi can grow at water activities of 0.65-0.9. The ability to grow on relatively dry substrate allows building materials and house dust to be suitable substrates for fungal growth. By using a differential, low water activity media, fungi such as Aspergillus versicolor, Eurotium, or 59 xerophilic Cladosporium and Penicillium can be cultured without being overgrown by more numerous competing fungi (Flannigan 1994). 5.1.2. Bacterial Aerosols 5.1.2.1. Mesophilic Bacteria Mesophilic bacteria were defined for this study as bacteria whose optimum condition for growth was at 35 - 37°C on non-enriched, non-selective media (Burge 1995). 5.1.2.2 Thermophilic B acteria Thermotolerant bacteria are defined as bacteria capable of growth at 55°C (Burge et al. 1987). Differential incubation temperature allows non-competitive growth of thermo-tolerent organisms. Thermoactinomyces or Saccharopolyspora are among the biological agents associated with hypersensitivity pneumonitis (Duchaine et al. 1999: Salvaggio 1997). Wet building materials as well as HVAC drip pans have been implicated as reservoirs for these organisms (Weltermann et al. 1998) and have been found in water damaged materials in school buildings (Thom et al. 1996). The ACGIH bioaerosols committee recommends culturing for thermophilic bacteria as part of routine indoor air investigations, because normal background levels are very low and the presence of thermophilic actinomycetes are indicative of unusual conditions in the building that would allow growth of these organisms (Otten et al. 1986). 5.2 METHODS 5.2.1 Fungal aerosols 5.2.1.1 Mesophilic Fungi Fungal aerosol samples were taken using the following protocol. A 100 x 15 mm disposable Petri dish (Phoenix Biomedical, Mississauga, ON) filled with 45-46 mL malt extract agar (MEA: w/v maltose, 1.275%; dextrin, 0.275%; glycerol, 0.235%; pancreatic digest of gelatin, 0.078%; agar 1.5%) (BBL Becton Dickinson and Company, Cockeysville, MD) was placed in each of two Andersen N-6 sampling heads (Graseby Andersen, Atlanta, GA) (Burge et al. 1987; Dillon et al. 1996). The N-6 sampling head had 400, 0.25 mm diameter holes through which air was drawn. Particles larger than 60 0.65 urn were impacted onto the agar surface. The two N-6 sampling heads were placed on tripods placed approximately 152 cm apart, with the sampling head 81 cm above floor or ground level. Air was drawn through the sampling head at 0.0283 m3 per minute by a high volume pump (AirCon-2, Gilian Instrument Corp., W. Caldwell, NJ or Andersen vacuum pump, Andersen, Atlanta, GA). Air flow was controlled by a critical orifice. The two sampling heads were run simultaneously for five minutes and constituted duplicate samples. The time the samples were taken was standardized to be at the end of the school day (approximately 3:00 p.m.). Fungal aerosol samples were taken in the centre of the classroom for the indoor sample, and outdoors near the source of fresh air for the room. The end of the school day was chosen as the time during the day when the bioaerosols would be optimally entrained in the indoor air. A MEA media control plate was included each sampling day. The control plate was taken into the field and treated identically to the sample plates, but was not opened. The media control was returned with the sample plates and incubated as described. Samples were returned to the laboratory for incubation for seven days at room temperature and exposed to a seasonally variable light/dark cycle. At the end of seven days the colonies growing on the agar at the impaction sites were counted with the assistance of a stereoscopic microscope (Nikon SMZ 1, Tokyo, Japan). Colony counts were adjusted for the probability of more than one spore entering one of the 400 sieve holes using the positive hole correction table (Andersen 1958). The samples were then stored at 4°C until colony identification was performed. Fungal colonies were first characterized by examination under a stereoscopic microscope (5 - 30 x magnification). Colonies were further identified to genus by microscopic examination using phase contrast micoscopy at 400 x magnification (Nikon labophot 2-pol, Tokyo, Japan) using 2 cm wide Scotch ™ tape (No. 600, 3M Company, St. Paul, MN) to lift mycelia, conidial heads and spores from the colony surface onto a microscope slide (Koneman and Roberts 1985). Conidial structures were used for morphologic identification (Raper et al. 1949; Raper and Fennell 1973; Pitt 1979; Malloch 1981; Rippon 1982; Kern 1985; Koneman and Roberts 1985; Smith 1990; Kwon-Chung and Bennett 1992; Gravesen et al. 1994; Samson et al. 1994). Colonies that did not have conidial structures or spores were grouped together as "sterile mycelia." 61 In order to validate the identification protocol, a series of 10 tape-mounted slides were sent to a reference laboratory (University of Alberta Microfungus Collection & Herbarium, Lynne Sigler, curator) for confirmation of the identification. The cultures were chosen to represent mesophilic, thermotolerant and xerophilic fungi which are differentiated only by primary culture characteristics, but do not represent differing microscopic morphology by the protocols used. There was 100% agreement for the commonly encountered genera in reported studies - namely, Aspergillus, Alternaria, 2 different species of Cladosporium, Mucor, Paecilomyces, 2 different species of Penicillium, Trichoderma and sterile mycelia. The reference laboratory was asked to identify the slide mount to the generic level, which was the level to which cultures were identified for this study. The limit of detection of the method was based on the presence of one countable colony on the sample media. Thus, the limit of detection (LOD) for the sampling time of 5 mm at 0.0283 m /min was 7 CFU/m . The limit of quantitation (LOQ) was based on the ACGIH bioaerosols committee recommendation that a minimum of 30 colonies is required to reduce the coefficient of variation of the method to 20% (Dillon et al. 1996). This results in a LOQ of 210 CFU/m3. For the purposes of this study, values below the limit of detection were assigned a nominal value of one colony for data analysis purposes only. 5.2.1.2 Thermotolerant Fungi The sampling protocol for thermotolerant fungal aerosols was identical to that for mesophilic fungi, with the following exceptions. Samples were returned to the laboratory and incubated at 37°C (NuAire model NU-2700, ESBE Scientific) for 48 h. At the end of 48 h, colonies were counted. Plate counting and identification protocols were identical to those for mesophilic fungi. 5.2.1.3 Xerophilic Fungi The sampling protocol for xerophilic fungal aerosols was identical to that for mesophilic fungi, with the following exceptions. The sampling media was 10% NaCl-MEA (Flannigan, 1995) and contained, w/v sodium chloride (NaCl), 10%; maltose, 62 2%. Incubation times and temperatures were as for mesophilic fungi and identical protocols were used for plate counting and identification of cultures. 5.2.2 Bacterial aerosols 5.2.2.1 Mesophilic Bacteria Bacterial aerosol samples were taken using the following protocol. A 100 x 15 mm disposable Petri dish (Phoenix Biomedical, Mississauga, ON) filled with 46-47 mL trypticase soy agar (TSA: w/v pancreatic digest of casein, 1.5%; papaic digest of soybean meal, 0.5%; sodium chloride, 0.5%; agar 1.5%) (BBL Becton Dickinson and Company, Cockeysville, MD) was placed in each of two Andersen N-6 sampling heads (Graseby Andersen, Atlanta, GA). Bacterial aerosol samples were taken in the centre of the classroom for the indoor sample, and outdoors near the source of fresh air for the room. The end of the school day was chosen as the time during the day when the bioaerosols would be optimally entrained in the indoor air by occupant activities. A TSA media control was included each sampling day. The control plate was taken into the field and treated identically to the sample plates, but was not opened. The media control was returned with the sample plates and incubated as described. Samples were returned to the laboratory for incubation for 48 h at 37°C. At the end of the incubation the colonies growing on the agar surface were counted with the assistance of a stereoscopic microscope (Nikon SMZ 1, Tokyo, Japan). The samples were then stored at 4°C until Gram stains were performed. Bacterial colonies were first characterized by examination under a stereoscopic microscope (5 - 30 x magnification). Colonies were further grouped by morphologic features by examining heat fixed smears stained by the 4-step Gram procedure (Kruczak-Filipov 1992) by using light micoscopy at 1000 x magnification (Nikon labophot 2-pol, Tokyo, Japan). The reagents for the Gram staining procedure were commercially prepared (Becton Dickinson Microbiology Systems, Sparks, MD). Bacterial cultures were grouped according to the scheme presented in Table 15, adapted from determinative bacteriology (Holt et al. 1994). Only rudimentary biochemical characterization was used for grouping purposes as the mesophilic bacteria found in this study were primarily Gram-positive cocci and rods. Additional characterization to genus level was not 63 attempted for this study. Gram-negative bacteria were found in such low numbers as to make species characterization unimportant. Table 15. Bacterial grouping by Gram stain and morphology Description of group Designation Gram positive rods spore forming Bacilli group non-spore forming Coryneform group filamentous, branching Actinomycetes group Gram negative rods fermentative, oxidase negative Fermentative rods group non fermentative, oxidase positive Non-fermentative rods group Gram positive cocci catalase positive, fermentative, coagulase negative Staphylococci group catalase positive, fermentative, coagulase positive Staphylococcus aureus catalase positive, non-fermentative Micrococci group catalase negative, cell division on one plane Streptococci group The limit of detection of the method was based on the presence of one countable colony on the sample media. Thus, the limit of detection (LOD) for the sampling time of 5 min at 0.0283 m3/min was 7 CFU/m3. The limit of quantitation (LOQ) was based on the ACGIH bioaerosols committee recommendation that a minimum of 30 colonies is required to reduce the coefficient of variation of the method to 20% (Dillon et al. 1996). This results in a LOQ of 210 CFU/m3. For the purpose of this study a value of one colony was assigned for the LOD. 5.2.2.2 Thermophilic Bacteria The sampling protocol for thermophilic bacterial aerosols was identical to that for mesophilic bacteria with the following exceptions. Samples were returned to the laboratory and incubated at 55°C (Precision® Mechanical Convection Incubator Model 4EM, Chicago, IL) for 72 h. At the end of 72 h, colonies were counted. Grouping protocols were identical to those for mesophilic bacteria. 64 5.2 3 Statistical Analyses Data were analysed using SPSS version 7.5 for Windows (SPSS, Inc., Chicago). The distributions of the bioaerosol concentrations were positively skewed. These data were tested for normality by the Lilliefors test and the null hypothesis of normality was rejected (p < 0.001). The data were log-transformed to the natural log. Examination of the histograms and testing for normality by the Lilliefors test of the log transformed data suggested that parametric statistics could be used for analysis. Calculation of the 95% confidence intervals of the geometric mean counts The 95% confidence intervals of the geometric means were calculated using log-normal statistics (Hawkins et al. 1991). In exposure monitoring applications the arithmetic mean is more meaningful than the median (Hawkins et al. 1991). Using the approach of Hawkins et al. (1991), the geometric mean was used to calculate an arithmetic mean, using equation 3, as shown here. ln(M) = \n(GM) + (0.5)(\n(GSD)Y Where: In = natural log M= calculated arithmetic mean from geometric mean GSD = geometric standard deviation The upper and lower geometric confidence limits are then calculated around the In arithematic mean using the geometric standard deviation. Equation 4 (Hawkins et al. 1991) was used for these calculations. ln(GLCL ,GUCL ) = ln(M) ± (r 0 9 7 5 d f ( l n ( ^ D ) Where: In = natural log GLCL = Geometric lower confidence limit GUCL = Geometric upper confidence limit M= calculated arithmetic mean from geometric mean t = statistic from t distribution #=n- l GSD = geometric standard deviation 65 Bioaerosol variables used for statistical analysis Fungal bioaerosol variables are listed in Table 16. Table 16. Fungal bioaerosol variables used for analysis Variable name Type of variable Description Indoor mesophilic fungal count Continuous Source: Room temperature MEA culture Measurement: arithmetic mean CFU/m 3 of duplicate samples Outdoor mesophilic fungal count Continuous Source: Room temperature MEA culture Measurement: arithmetic mean CFU/m 3 of duplicate samples Indoor thermotolerant fungal count Continuous Source: 37°C incubated MEA culture Measurement: arithmetic mean CFU/m 3 of duplicate samples. Outdoor thermotolerant fungal count Continuous Source: 37°C incubated MEA culture Measurement: arithmetic mean CFU/m 3 of duplicate samples. Indoor xerophilic fungal count Continuous Source: Room temp. 10% NaCl-MEA culture Measurement: arithmetic mean CFU/m 3 of duplicate samples. Outdoor xerophilic fungal count Continuous Source: Room temp. 10% NaCl-MEA culture Measurement: arithmetic mean CFU/m 3 of duplicate samples. Additional variables were created for descriptive purposes. Twenty-six variables were used to describe the genera of mesophilic fungi identified as a percent of the total sample count. From these 26, additional variables were created by multiplying each of these percentages by the arithmetic mean of the paired samples to obtain a genus count. This count was log transformed to the base e. Twelve variables were used to describe the genera of identified thermotolerant fungi as a percent of the total thermotolerant sample. Ten variables were used to describe the genera of identified xerophilic fungi as a percent of the total xerophilic sample. Additional variables to describe the log transformed genera count were created as described for the mesophilic fungi. The descriptions of these variables are listed in Table 17. 66 Table 17. Variables used to describe fungal genera or groups for analysis Description of Mesophilic genera Thermotolerant Xerophilic genera variables for all morphology genera morphology morphology genera morphologies — Indoor % of sample Acremonium Aspergillus Aspergillus Aspergillus - Outdoor % of sample Aureobasidium Beauveria Aureobasidium - (In) Indoor count Botrytis CFU/m3 Chrysonilia Chrysosporium Chrysonilia - (In) Outdoor count Cladosporium Cladosporium CFU/m3 Epicoccum Eurotium Geotrichum Mucor Nigrospora Oidiodendron Geotrichum Eurotium Paecilomyces Paecilomyces Paecilomyces Penicillium Penicillium Penicillium Rhizopus Rhizopus Scopulariopsis Scopulariopsis Sporotrichum Trichoderma Trichoderma Ulocladium Wallemia Wallemia Yeast Yeast Sterile mycelia Sterile mycelia Sterile mycelia Unidentifiable Unidentifiable Unidentifiable Other Other Other \ 67 Bacterial bioaerosol variables used for analysis are listed in Table 18. Table 18. Bacterial bioaerosol variables used for analysis Variable name Type of variable Description Indoor mesophilic bacterial count Continuous Source: TSA cultured at 37°C Measurement: arithmetic mean CFU/m 3 of duplicate samples Outdoor bacterial count Continuous Source: TSA cultured at 37°C Measurement: arithmetic mean CFU/m 3 duplicate samples of Indoor thermophilic bacterial count Continuous Source: TSA cultured at 55°C Measurement: arithmetic mean CFU/m 3 duplicate samples. of Outdoor thermophilic bacterial count Continuous Source: TSA cultured at 55°C Measurement: arithmetic mean CFU/m 3 duplicate samples. of Additional variables were created for descriptive purposes. Eight variables were used to describe group morphology of mesophilic bacteria as a percent of the total sample count. From these eight, additional variables were created by multiplying each sample percentage by the arithmetic mean to obtain a group count. This count was In transformed. Two variables were used to describe group morphology of thermophilic bacteria as a percent of the total thermophilic sample. Additional variables to describe the In transformed genera count were created as described for the mesophilic bacteria. The descriptions of these variables are listed in Table 19. Table 19. Variables used to describe group morphology of bacterial counts Description of variables for all group morphologies Mesophilic group morphology Thermophilic group morphology - Indoor % of sample Actinomycetes group Actinomycetes group - Outdoor % of sample Bacilli group Bacilli group - (In) Indoor count CFU/m3 Corneform group - (In) Outdoor count CFU/m3 Micrococci group Staphylococci group Streptococci group Gram negative rods (non-fermenting) 68 5.3 RESULTS 5.3.1 Fungal Aerosols 5.3.1.1 Mesophilic Fungal Aerosols The overall geometric mean indoor fungal count was 323.8 CFU/m3 (GSD 2.80; range 14 - 18,583 CFU/m3). The 95% confidence interval of the mean was 454.6 - 664.8 CFU/m . The overall geometric mean outdoor fungal count was significantly higher at 445.86 CFU/m3 (GSD 2.10; range 89 - 2392) (p<0.01). The variability of mesophilic fungal counts was assessed by taking side by side samples. Of 116 indoor samples, 113 were duplicate samples. The mean overall coefficient of variation of the indoor duplicates was 15.3% (range 0 - 85%; median 12.3%). Of 116 outdoor samples, 112 were duplicate samples. The mean overall coefficient of variation was 13.3 % (range 0 - 78%; median 8.85%). 100% of the mean indoor counts were above the LOD of 7 CFU/m3; 63.8% were above the LOQ of 210 CFU/m3, similarly, 100% of the mean outdoor counts were above the LOD and 84.5% were above the LOQ. The fungal groups present in 10% or more of the samples are shown in Table 20. The indoor and outdoor concentrations were compared by Student t test. Table 20. Ranking of mesophilic indoor and outdoor fungal groups Fungal group Indoor Positive Outdoor Positive Indoor CFU/m3fl samples CFU/m3 samples vs. GMb (GSD)C % GM (GSD) % outdoor p-value Total recovery 323 (2.8) 100 446 (2.1) 100 0.001 from MEA Cladosporium 50 (3.7) 98 77 (2.7) 100 <0.005 Penicillium 45 (4.6) 96 27 (4.7) 88 O.05 Sterile mycelia 42 (5.6) 90 108(4.1) 97 O.001 Yeast 20 (6.0) 79 28 (5.3) 85 NS Aureobasidium 7.5 (5.2) 67 15 (4.6) 85 <0.001 Botrytis 3.6 (6.4) 35 12(7.1) 66 O.001 Aspergillus 2.1 (3.7) 30 1.6(2.6) 19 NS Paecilomyces 1.3(2.0) 11 1.5 (2.6) 15 NS a Colony forming units per cubic metre. b Geometric mean. c Geometric standard deviation. 69 The ranking of the most commonly found mesophilic fungi in the indoor samples was: Cladosporium > Penicillium > sterile mycelia > yeast > Aureobasidium > Botrytis > Aspergillus > Paecilomyces. The ranking of the most common outdoor mesophilic fungi was: sterile mycelia > Cladosporium > yeast > Penicillium > Aureobasidium > Botrytis > Aspergillus > Paecilomyces. Penicillium spp. were found more often indoors than outdoors (p < 0.05 %2) and in higher concentration (p < 0.05). In contrast, Cladosporium spp., was found equally in both indoor and outdoor samples, but in lower concentration indoors (p < 0.005). Botrytis and Aureobasidium were found in significantly more outdoor samples (p < 0.001 %2) and in higher concentrations (p < 0.001). Aspergillus was isolated more often in indoor than outdoor samples (p < 0.05 %2), but in similar concentration. The grouping together of fungal colonies that did not produce spores, identified as "sterile mycelia," included basidiomycetes, but these were not enumerated separately from the other non-sporulating colonies. This group was commonly found in both indoor and outdoor samples, with outdoor samples having higher geometric mean counts than indoor samples (41.8 vs. 108.4 CFU/m3 respectively) (p < 0.001). The fungal groups varied seasonally, as shown in Figure 6. Seasonal concentrations of indoor isolates were significantly lower in winter for the predominant fungal groups with the exception of Aspergillus spp. and the grouped yeast. With the exception of the grouped yeast, isolates of fungi from outdoors also showed significant seasonal patterns; Aspergillus was below the limit of detection during the winter sampling season. Figure 7 illustrates the concentration of outdoor fungal groups by season. 70 Seasonal variation of indoor fungal groups Cladosporium Penidllium Sterile mycelia Yeast Aureobasidium Botrytis Aspergillus +Scheffe post hoc procedure Fungal groups Figure 6. Seasonal variation of selected indoor mesophilic fungal groups. Seasonal variation of outdoor fungal groups *p < 0.05 *p < 0.001 *p < 0.001 *p < 0.001 + + •p < 0.001 • Winter • Spring • Fall *p < 0.05 Sterile mycelia Cladosporium Yeast Penicillium Aureobasidium Botrytis Aspergillus "Scheffe post hoc procedure Fungal groups Figure 7. Seasonal variation of selected outdoor mesophilic fungal groups. 71 Table 21. Indoor/outdoor ratios of fungal counts (CFU/m3) grouped by season. Fungal group Parameter Winter Spring Fall Mesophilic n 35 42 39 mean (sd) 0.90 (0.150) 0.99 (0.229) 0.96 (0.104) range 0.54-1.27 0.63-1.91 0.71-1.21 Thermotolerant n 22 36 35 mean (sd) 1.16(0.838) 0.97 (0.588) 1.14(0.568) range 0-2.83 0-2.19 0-2.63 Xerophilic n 32 42 39 mean (sd) 1.06 (0.601) 1.15 (0.495) 1.04 (0.325) range 0-2.08 0.40 - 2.76 0-2.26 The concentration of Botrytis sp. was dramatically higher in the fall season both indoors and outdoors. This fungus is particularly associated with the colonization of high carbohydrate plant materials such as fruit. The study region is richly supplied with wild blackberry bushes whose berries mature in late August. The relative ranking of the indoor and outdoor fungal groups remained the same over the two sampling cycles for each season (data not shown). Indoor/outdoor ratios for samples were computed. Table 21 illustrates the indoor/outdoor ratios for mesophilic, thermotolerant, and xerophilic fungi grouped by season. There were no significant seasonal differences for any of the fungal groups. 5.3.1.2 Thermotolerant Fungal Aerosols Fungi capable of growth at elevated temperatures (37°C) are a subset of the greater representation of environmentally isolated fungi. The overall geometric mean indoor thermotolerant fungal count was 8.67 CFU/m3 (GSD 3.90; range 0-1525 CFU/m3). The 95% confidence interval of the mean was 17 - 28 CFU/m3. The overall geometric mean outdoor fungal count was 6.89 CFU/m3 (GSD 3.63; range 0 - 148 CFU/m ) which did not differ significantly from indoor counts. The mean coefficient of variation of thermotolerant fungal side by side samples was 36.6% (range 0 - 106%; median 28.3%) for 56 pairs (48%) with counts above the LOD. 2.6 % mean indoor counts were above the LOQ. Of 50 outdoor samples above the LOD, the mean overall coefficient of variation was 27.4% (range 0 - 98%; median 72 21.1%). No mean outdoor counts were above the LOQ. The variation between duplicated samples was greater for thermotolerant fungal samples than for the duplicated mesophilic fungal samples due to the low concentrations of organisms. Incubating plates at 37 °C allowed the recovery of Aspergillus spp. in 58 (50%) indoor samples vs. 35 (30%) of samples incubated at room temperature (p < 0.01 %2). Similarly, elevated incubation of outdoor samples recovered Aspergillus spp. in 60 (52%) vs. 22 (19%) (p < 0.001 %2). Unlike the mesophilic fungi, the recovery of thermotolerant fungal groups did not vary seasonally for indoor samples. The thermotolerant fungal groups isolated from 10% or more of samples are listed in Table 22. Table 22. Indoor and outdoor fungi isolated from elevated temperature incubation Fungal group Indoor CFU/m 3 f l GM* (GSD)C Positive samples % Outdoor CFU/m 3 G M (GSD) Positive samples % Indoor vs. outdoor p-value Total recovery from MEA 9(3.91) 84 7 (3.64) 81 NS Aspergillus 3.2 (4.37) 50 3.3 (3.96) 51 NS Yeast 2.3 (3.08) 39 1.5 (2.31) 23 <0.005 Sterile mycelia 1.4 (2.03) 21 1.7(2.50) 30 O.05 Paecilomyces a r~\ i r 1.3 (1.98) 13 1.1 (1.57) 7 NS  Colony forming units per cubic metre. b Geometric mean. c Geometric standard deviation. 5.3.1.3 Xerophilic Fungal Aerosols The xerophilic fungi are a subset of the larger group of mesophilic and hydrophilic fungi. The overall geometric mean indoor xerophilic fungal count was 51.9 CFU/m3 (GSD 4.67; range 0 - 2143 CFU/m3). The 95% confidence interval of the mean was 128.5 - 226.7 CFU/m3. The overall geometric mean outdoor fungal count was 42.5 CFU/m3 (GSD 3.63, range 0 - 796 CFU/m3) and did not differ significantly from indoor counts. The mean coefficient of variation of xerophilic fungal side by side samples was 25.1% (range 0 - 98%; median 19.5%) for 100 pairs of indoor samples with counts above 73 the L O D . O f 100 outdoor duplicates above the L O D , the mean overall coefficient of variation was 29.5% (range 0 - 122%; median 23.3%). While 92% o f mean indoor xerophilic fungal samples were above the L O D ; 15.5% were above the L O Q . Similarly, 95.7% of mean outdoor samples were above the L O D and 12.3% were above the L O Q . Cladosporium spp., Penicillium spp. and Aspergillus spp. were isolated from xerophilic media at lower concentrations than M E A media, however, Eurotium, a xerophile, was only isolated on xerophilic media, (11% of indoor and 6% of outdoor samples). Figure 8 illustrates the seasonal variation of indoor xerophilic fungi. Seasonal variation of indoor xerophilic fungi 4 .5 3.5 *p< 0.001 • Winter • Spring • Fall Cladosporium Penicillium ^ Scheffe post hoc procedure Sterile mycelia Fungal groups Aspergillus Eurotium Figure 8. Seasonal variation of indoor xerophilic fungal groups. Xerophilic Cladosporium spp. were significantly reduced in the winter; however, there were no significant differences in the recoveries by season for the other xerophilic genera. The seasonal variation o f the outdoor xerophilic fungi were not different from the seasonal variation of the indoor xerophilic fungi. 74 5.3.2 Bacterial Aerosols 5.3.2.1 Mesophilic Bacterial Aerosols The overall geometric mean indoor mesophilic bacterial count was 225.9 CFU/m3 (GSD 2.75; range 4 - 1686 CFU/m3). The 95% confidence interval of the mean was 312.6 - 453.8 CFU/m . The overall geometric mean outdoor bacterial count was significantly lower at 26.1 CFU/m3 (GSD 2.56; range 0 - 179 CFU/m3) (pO.001). The overall coefficient of variation of the indoor duplicate samples was 15.0% (range 0 - 71%; median 11.68%). The overall coefficient of variation of the outdoor samples was 35.1 % (range 0 - 93%; median 28.3%). 99% of the mean indoor bacterial samples were above the LOD and 64% were above the LOQ. While 96% of the mean outdoor bacterial counts were above the LOD, none were above the LOQ of the method. The bacterial population recovered in this study was almost entirely Gram-positive, but varied in its composition indoors compared to outdoors. The bacterial groups recognized on >_10% of samples are listed in Table 23. Table 23. Indoor and outdoor bacterial population Bacterial group Indoor C F U / m 3 f l GM* (GSD)C Positive samples % Outdoor C F U / m 3 GM (GSD) Positive samples % Indoor vs. outdoor p-value Total recovery from TSA 226 (2.75) 100 26 (2.56) 99.2 O.001 Micrococci 126 (3.23) 100 3.1 (3.59) • 54 O.001 Bacilli 21 (4.31) 87 6.4 (3.65) 79 O.001 Staphylococci 8.9 (6.88) 61 1.2(1.75) 11 O.001 Coryneform 8.9 (5.28) 70 3.5 (3.36). 61 O.001 Actinomycetes 1.5 (2.58) 17 1.6 (2.34) 26 NS a Colony forming units per cubic metre. b Geometric mean. c Geometric standard deviation. Micrococci and staphylococci were both significantly higher in number of positive samples (p < 0.001 %2), and concentrations from indoor samples than outdoor samples. In contrast, the Gram-positive rods, the sporing and non-sporing groups were 75 significantly higher in concentration in outdoor samples, but were isolated from similar numbers of samples. No coagulase positive staphylococci were cultured in this study. 5.3.2.2 Thermophilic Bacteria Although 76 (66%) of indoor and 74 (64%) of outdoor samples showed growth on TSA plates incubated at 55°C, the vast majority were thermotolerent Bacillus spp. Of the indoor samples, only 4 (3%) and 3 (2.6%) outdoor samples were positive for bacteria belonging to the actinomycetes group. The counts were very low for these samples, ranging from 2 -5 CFU/m3. Only the thermophilic actinomycetes were a priori of interest for this study; there is no evidence that low counts of thermophilic bacilli contribute to indoor air quality concerns. With the exception of one sample, the samples that were positive for indoor thermophilic actinomycetes also recovered outdoor thermophilic actinomycetes. No further analyses were conducted for the thermophilic bacteria due to low numbers and correlation between indoor and outdoor cultures. 5.4 SUMMARY OF BIOAEROSOLS Bioaerosols in the indoor and outdoor environment were ubiquitous. The mean indoor levels of mesophilic fungal aerosols were lower than outdoor levels, and mean indoor levels of bacterial aerosols were higher than outdoor levels. The source of indoor fungi is normally from outdoor air unless fungi are colonizing interior building materials. The source of indoor bacteria is normally from the bioeffluents of occupants (human or otherwise) unless bacteria are also colonizing building materials or ventilation system features. The larger group of saprophytic fungi were differentially cultured to enhance the growth of thermotolerant and xerophilic fungi. Both of these sub groups were present in lower numbers than was found for side by side samples of mesophilic fungi. Bacteria were also differentially cultured to enhance the growth of genera capable of growth at 55 °C. Species of Bacillus growing at elevated temperatures were not specifically of interest, and the presence of thermophilic actinomycetes was extremely rare. 76 CHAPTER 6 COMPARISON OF COMFORT P A R A M E T E R S , VENTILATION A N D BIOAERSOL CONCENTRATIONS TO A V A I L A B L E GUIDELINES A N D STANDARDS 6.1 COMFORT P A R A M E T E R S The classrooms were evaluated using existing guidelines and regulations. 6.1.1 Relative Humidity A S H R A E standard 62-1989 specifies that relative humidity should be maintained between 30% and 60%. However other researchers (Arundel et al. 1986) have proposed a range from 40% to 60% to accommodate contact lens wearers, and as the zone minimizing adverse health effects from bacteria, viruses, fungi, dust mites, respiratory infections and asthma. Although there is some disagreement with the observations of Arundel et al., other Canadian publications have used data from this publication to suggest appropriate ranges of relative humidity (Engineering Interface 1989). Health Canada (1995a) states that ideal indoor relative humidity levels are 35% in the winter and 50% in the summer and states a range between 25% and 60%. Table 24 summarizes the number of rooms considered in this study that would fail to meet these guidelines. Table 24. Number of test rooms failing to meet R H guidelines by season Guideline and reference Winter n=35 Spring n=42 Fall n=39 Count (%) Count (%) Count (%) < 30% (ASHRAE 62-1989) 5 (14%) 4 (9.5%) 0 < 40% (Arundel et al. 1986) 22 (63%) 17 (41%) 5 (13%) > 60% (ASHRAE & Arundel 0 0 2(5%) op. cit.) 6.1.2 Temperature The comfort zone temperatures depend to some extent on the clothing worn by the occupants and their activity level. The ranges of temperatures suggested by A S H R A E standard 55-1981 are also cited by Health Canada (1995a). For winter the range is 20.3 - 24.2 °C and for summer the range is 23.1 - 26.4 °C (ASHRAE 1981) for 77 mechanically ventilated rooms. Table 25 summarizes the number of rooms falling outside the suggested ranges by season. Table 25. Number of rooms not conforming to temperature guidelines by season Guideline endpoints Winter n=35 Spring n=42 Fall n=39 Count (%) Count (%) Count (%) < 20.3 °C (winter range) 5 (14%) NA 5 NA < 23.1 °C (summer range) NA 24 (57%) 24 (62%) > 24.2 °C (winter range) 1(3%) NA NA > 26.4 °C (summer range) NA 1 (2%) 2 (5%) "NA = not applicable. 6.2. OCCUPANT GENERATED C0 2 AND VENTILATION The current standard (ASHRAE 62-1989) for schools is for the provision of ventilation equivalent to 8 L/sec per person (0.008 m3/second per person). Based on mass balance equations provided in Appendix D (ASHRAE 1989), if this ventilation rate were met, it would prevent CO2 levels from rising above 1000 ppm for the stated occupancy limits. The ASHRAE ventilation committee is currently reviewing Standard 62-1989. A proposed change will be to describe adequate ventilation in terms of occupant-generated CO2 above the ambient level, in light of the possibility that the ambient level may be above 350 ppm in areas such as cities with heavy traffic or other sources contributing to outdoor CO2. The resulting context is that ventilation is adequate if the occupant-generated CO2 is no more than 650 ppm above ambient levels (E. Chessor, ASHRAE ventilation committee member, personal communication). The Workers' Compensation Board of British Columbia recently adopted indoor air quality guidelines in Part 4: General Conditions (WCB 1998) and referenced ASHRAE 62-1989 as the objective for CO2 levels, including the new wording as described in the previous paragraph. Health Canada (1995a) set more conservative objectives for office buildings at 850 ppm which could be achieved with a ventilation rate of 10 L/s per person. Table 26 summarizes the CO2 measurements from classrooms in this study. 78 Table 26. CO2 concentrations exceeding recommended limits by season CO2 ppm Winter n=35 Count (%) Spring n=42 Count (%) Fall n=39 Count (%) *p-value < 850 ppm 6(17%) 21 (50%) 17(44%) > 851 ppm 29 (83%) 21 (50%) 22 (56%) <0.005 < 1000 ppm 10(29%) 29 (69%) 23 (59%) > 1001 ppm 25 (71%) 13 (31%) 16(41%) 0.001 *X2 test. Figure 9 illustrates the percentage of classrooms meeting the ventilation rate criterion of 8 L/s per person for rooms used for educational purposes as specified in ASHRAE 62-1989. Room ventilation Over 8 Us 8% Under 8 L/s Figure 9. Percentage of classrooms meeting the ventilation rate criterion for rooms used for educational purposes (ASHRAE 1989). As can be seen in Figure 9, the majority of classrooms in the study did not provide the ASHRAE guideline ventilation rate of 0.008 m3/s per person. As seen in Table 26, about half of the rooms were below 1000 ppm. The difference between these two illustrations is that the ventilation rate (0.008 m /s per person) is based on full day occupancy. None of the rooms were continuously occupied for the entire day, and the CO2 levels decreased during the periods the rooms were empty, therefore more rooms met the ASHRAE 62-1989 criteria for average CO2 concentration. 79 Another way of expressing ventilation is the air exchange rate. Although this is not a regulated value, the acceptable level for living areas of a residence is 0.35 ACH of outdoor air (ASHRAE 1989; Fernandez-Caldas et al. 1994; Turner et al. 1995). Offices and other mechanically ventilation spaces typically have ventilation rates of 1.0 ACH (Turner et al. 1995). Figure 10 summarizes the air exchange rates for the classrooms, during unoccupied periods. 50 Air changes per hour Figure 10. Air exchange rates determined by SF6 tracer gas decay measurement under static room conditions. The vertical line indicates 0.35 ACH. Approximately half (60/115) of the classrooms studied had air exchange rates below the suggested level of 0.35 ACH as measured by SF6 tracer gas decay, which agrees well with the proportions of classrooms with and without access to mechanical ventilation. 6.3 BIOAEROSOLS 6.3.1 Mesophilic Fungal Aerosols The most current publication of the ACGIH bioaerosols committee (Macher et al. 1999) differs in opinion from an often referenced, earlier publication which suggested numeric standards for the evaluation of fungal aerosols (Burge et al. 1987). Rather, the general principles include a critical examination of the presence of indicator organisms (e.g. Stachybotrys, Fusarium), and recommednations that indoor levels be lower than outdoor levels realizing that outdoor levels under snow cover are very low. In this coastal temperate zone, only 5 sampling days had snow cover. Health Canada Health Canada (1995a p 50) proposes numeric guidelines based on data collected in office buildings. These guidelines are as follows: "Significant numbers of certain pathogenic fungi should not be present in indoor air (e.g., Aspergillus fumigatus, Histoplasma and Cryptococcus). Bird or bat droppings near air intakes, in ducts or buildings should be assumed to contain these pathogens. Action should be taken accordingly. Some of these species cannot be measured by air sampling techniques." "The persistent presence of significant numbers of toxigenic fungi (e.g., Stachybotrys atra, toxigenic Aspergillus, Penicillium and Fusarium species) indicates that further investigation and action should be taken accordingly." "The confirmed presence of one or more fungal species occurring as a significant percentage of a sample in indoor air samples and not similarly present in concurrent outdoor samples is evidence of a fungal amplifier. Appropriate action should be taken." - "The 'normal' air mycoflora is qualitatively similar and quantitatively lower than that of outdoor air. In federal government buildings, the 3-year average has been approximately 40 CFU/m3 for Cladosporium, Alternaria, and non-sporulating basidiomycetes." "More than 50 CFU/m of a single species (other than Cladosporium or Alternaria) may be reason for concern present [sic]. Further investigation is necessary." 81 "Up to 150 CFU/m is acceptable if there is a mixture of species reflective of the outdoor air spores. Higher counts suggest dirty or low efficiency air filters or other problems." "Up to 500 CFU/m is acceptable in summer if the species present are primarily Cladosporium or other tree and leaf fungi. Values higher than this may indicate failure of the filters or contamination in the building." "The visible presence of fungi in humidifiers and on ducts, mouldy ceiling tiles and other surfaces requires investigation and remedial action regardless of the airborne spore load." No Stachybotrys or Fusarium species were cultured from air samples in this study. There were 15 samples with greater than 50 CFU/m3 of a single species that was not reflective of outdoor flora. Morphologically similar Penicillium sp. were isolated in 14/116 (12%) and Aspergillus niger was isolated in 1/116 (0.8%) samples as the predominant species representing over 50 CFU/m . The indoor/outdoor ratios and other criteria for fungal evaluation are summarized in Table 27. Table 27. Fungal counts found to be unacceptably high by selected criteria Variable Winter n=35 Spring n=42 Fall n=39 Count (%) Count (%) Count (%) Indoor > outdoor 6 (17%) 12 (29%) 12 (41%) (Macher et al. 1999) Samples > 150 fungi 21 (60%) 38 (91%) 35 (90%) CFU/m3 (Health Canada 1995a) Samples > 500 5(14%) 16 (38%) 20 (51%) CFU/m3 (Health Canada 1995a) Aspergillus 12 (34%) 19 (45%) 26 (68%) fumigatus ° (37°C) (Health Canada 1995a) "Range of concentrations ( 7 - 1492 CFU/mJ). The Health Canada criteria for evaluation of bioaerosol concentrations were based on data collected primarily from office buildings. Many office buildings have HVAC 82 systems that function to reduce the fungal bioaerosol concentration indoors except in the cases when the HVAC system is contaminated and is the source of fungal aerosols. Using criteria based on a data base from office environments makes the interpretation somewhat problematic for buildings that are used differently and have less sophisticated air handling systems, if they have forced air at all. Depending on the criteria cited, 30 to 94 rooms would have failed, based on a simplistic interpretation of culturable fungal counts or the presence of potentially pathogenic fungi (Aspergillus fumigatus). ACGIH bioaerosols committee guideline criteria If a room were completely open to the outdoor environment by windows and doors open to the prevailing wind direction, the indoor air concentration of fungal spores should be quantitatively and qualitatively the same as the outdoor air concentration. The ratio in this case would be 1.0. However, if a room is closed to the outside environment and only filtered, forced air is supplied, then the ratio should be considerably less than unity. In a 1987 protocol by the ACGIH bioaerosols committee (Burge et al. 1987) the suggested cut off for indoor levels of fungal counts was approximately 1/3 that of outdoor fungal counts, with a similar qualitative representation of fungal species. In this study, no indoor counts were less than 1/3 of outdoor counts; 32 (27.6%) samples had indoor counts equal to or greater than outside counts, and 94 (81%) samples had indoor counts greater than 150 CFU/m3. In climates with prolonged snow cover in winter, the use of indoor/outdoor ratios can be problematic if the outdoor count is very low (Burge et al. 1999); however, in this study there were no statistically significant differences in indoor/outdoor ratios due to season (mean indoor/outdoor ratio 0.95). 95% confidence intervals of means A goal of this research study was to develop an evaluation tool that could be used locally. Using the approach of Hawkins et al. (1991), the calculated arithmetic mean mesophilic fungal count for all indoor samples was 549.7 CFU/m with a geometric standard deviation of 2.81. The 95% geometric confidence interval of the mean was 454.6 - 664.8 CFU/m3. In the study classrooms, 23 rooms (20%) recorded fungal counts higher than the geometric upper confidence limit. 83 This approach could be further developed to establish other data ranges - for example, for seasons or for mechanical vs. natural ventilation. Geometric confidence intervals calculated for the mean exposures in these groups could be used for local conditions that share exposure conditions. An example using seasonal and ventilation data is shown in Table 28. In this example, seasons were chosen because there are temperature and ecological conditions that can be grouped together. The presence or absence of ventilation was chosen because of the strong relationship ventilation has to indoor air quality variables. Only the geometric upper confidence limit (GUCL) is calculated because concentrations lower than the geometric lower confidence limit would not be considered out of compliance. Table 28. Geometric means of mesophilic fungal counts and associated upper 95% geometric confidence limits. Variable n Arithmetic mean " (GSD)C CFU/m 3 G U C L * Counts > GUCL Winter 35 293.8 (2.81) 419 8 (7%) Spring 42 764.3 (2.97) 1072 6 (5%) Fall 39 536.7(1.99) 671 11 (9%) Natural ventilation 54 635.9 (2.48) 816 10(8.6%) Mechanical ventilation 62 460.6 (2.96) 607 14(12%) " Mean calculated from geometric mean (Hawkins et al. 1991). b Geometric upper confidence limit. c Geometric standard deviation. In the examples given in Table 28, the fungal counts in 25 rooms sampled seasonally were greater than the 95% GUCL. In contrast, 24 rooms had counts above the 95% GUCL for the groups defined by use of natural or mechanical ventilation. The geometric mean concentrations of the selectively cultured fungi, the thermophiles and xerophiles, tended to be higher for indoor than outdoor samples, but not significantly so. Although there are no guidelines that discuss numeric limits for these subpopulations of fungi, a similar approach could be taken using broad exposure grouping to develop exposure groups. Table 29 summarizes the 95% geometric confidence intervals of the means for these fungal populations. 84 Table 29. Geometric confidence limits for indoor thermotolerant and xerophilic fungal means. Variable Thermotolerant 95% GUCL" CFU/m3 Counts > GUCL Xerophilic 95% GUCL CFU/m3 Counts > GUCL Winter 25 5 (4%) 149 4 (3%) Spring 42 3 (3%) 169 7 (6%) Fall 33 8 (7%) 449 5(4%) Natural ventilation 38 7 (6%) 383 9 (8%) Mechanical ventilation 27 9 (8%) 163 10 (9%) " Geometric upper confidence limit. The seasonal evaluation of thermotolerant fungi identified 16 rooms with counts greater than the 95% GUCL. Six of the 16 rooms also would have been identified based on mesophilic fungal concentrations. Ten rooms were unique in having only thermotolerant fungi elevated above the 95% GUCL. The differential incubation allowed unrestricted growth of the thermotolerant microflora without competition from other mesophilic flora on the plates incubated at room temperature. Of the 16 room counts identified in the xerophilic seasonal groupings, 13 room counts were also identified in the mesophilic data set, 4 were identified in the thermotolerant data set and 4 rooms were uniquely identified. 6.3.2 Evaluation of mesophilic bacterial aerosols There are no Canadian guidelines per se regarding appropriate levels of bacterial aerosols. Infection control efforts in hospital environments have suggested levels of <175 CFU/m3 bacteria in sterile operating theatres (Morey 1985) but these criteria are clearly inappropriate for this setting. The presence of high concentrations of Gram-negative flora is known to contribute to endotoxin exposure (Hood 1990; Teeuw et al. 1994). Recent work from Finland suggests that some Bacillus spp. isolated from indoor environments may produce cytotoxins (Salkinoja-Salonen et al. 1998; Andersson et al. 1999) but no exposure limits have yet been proposed. The Hong Kong Interim Indoor Air Quality Guidelines suggest an upper limit of 1000 CFU bacteria/m3 (1-hour average) (Lee and Chang 1999). This is in general agreement with levels referenced by an AIHA 85 publication (Dillon et al. 1996). There were two classrooms with bacterial concentrations above 1000 CFU/m3 in this study. The relationship of ventilation efficiency to spread of contagious particles (Burge 19906; Wheeler 1993) would suggest that increased levels of commensal bacteria may be a surrogate for the potential spread of infectious disease in the same environment if infectious particles were present. Table 30 illustrates the relationship of ventilation and bacterial concentrations. Table 30. 95% GUCL for bacterial concentrations Variable n Arithmetic mean " (GSD) CFU/m3 9 5 % GUCL 6 Counts > GUCL Natural ventilation 53 427.1 (2.09) 524 16 (14%) Mechanical ventilation 62 306.6 (3.02) 406 12 (10%) a Calculated from the geometric mean (Hawkins et al. 1991). b Geometric upper confidence limit. Using these data, 16 naturally and 12 mechanically ventilated rooms identified counts which exceeded the upper confidence limit of the mean. 86 CHAPTER 7 RELATIONSHIPS BETWEEN INDOOR AIR QUALITY PARAMETERS 7.1 METHODS The outcome (dependent) variables of interest were indoor fungal, indoor bacterial, and indoor CO2 concentrations. Indoor CO2 was chosen as a marker for bioeffluent, or "olf factor, suggested as a potential contribution to indoor air quality complaints (Fanger 1988a; Fanger 19986). Multiple linear regression equations were created for each outcome variable. Independent variables describing site and environmental conditions for the study are listed in Tables 3 - 5 in Chapter 3. Independent variables describing ventilation and comfort parameter measurements and the dependent variable of C0 2 concentrations are listed in Tables 11 and 12 in Chapter 4. Dependent variables describing measurements of mesophilic, thermotolerant and xerophilic fungal, and mesophilic bacterial concentrations are listed in Tables 16 and 18 in Chapter 5. All continuous variables were examined to determine if the data followed a normal distribution. The outcome variables describing bioaerosol and CO2 concentrations were positively skewed and approximately log normally distributed (Lilliefors test). The values for these measurements were log transformed to the natural log to allow the use of parametric statistics. All independent, continuous variables were tested for linear correlation to each other (Pearson r). Independent variables that were highly correlated were examined to determine the logic of the relationship and to choose only those variables that were appropriate to include in multiple linear regression models for the outcomes of interest based on prior knowledge or theory. Similarly, all categorical variables were tested for statistically significant relationships with other independent variables and choices made as for continuous variables. Categorical variables with more than two descriptors were coded into vectors using dummy coding (Munro and Page 1993). 87 Statistical power for this analysis was estimated by a "rule of thumb" that suggests having 30 samples per independent variable (Nunnally 1978). There were 116 samples which would suggest adequate power for models containing no more than four significant independent variables. Multiple linear regressions were constructed using a standard procedure, entering those independent variables most highly related to the outcome variable in univariate analysis and which, a priori, were predicted to contribute to the outcome. Independent variables were retained in the multiple regression models if they were significantly related to the outcome (p < 0.05) and they contributed to an increase in the adjusted R2. In cases where independent variables may have described similar conditions, the most appropriate variable was chosen by individually entering each into the model to assess which contributed the most predictive power. Because of the large number of highly correlated variables, many of which are surrogates for the same general construct, and because of the relatively complex nature of the relationships among potential predictor variables, interaction terms were not considered for inclusion in the models. It was felt that this would increase the difficulty in interpreting the results, especially given the already low power of the models using greater than four independent variables. 7.2 RESULTS 7.2.1 Outcome Variable: Indoor Mesophilic Fungal Concentration (CFU/m3) Independent univariate variables were examined to determine which would be entered into the regression model for indoor mesophilic fungal concentration. The independent variables could be grouped into three broad categories, ventilation, environment or occupant related. The statistically significant ventilation variables are summarized in Tables 31 and 32, environmental and occupancy variables in Tables 33 and 34. 88 Table 31. Continuous site and ventilation characteristics related to indoor mesophilic fungal concentration Continuous variables n Correlation coefficient p-value Pearson r Building age 105 - 0.248 <0.05 Supply air (m3/sec) 116 - 0.340 < 0.001 Air changes per hour " 115 - 0.283 < 0.005 a Measured by SF6 tracer gas decay. Table 32. Categorical site and ventilation characteristics related to indoor mesophilic fungal concentration Variable n Mesophilic fungi CFU/m* *p-value GM" (GSD)6 Ventilation used on test day Yes 62 255.8 (2.96) <0.005 No 54 420.7 (2.48) Supply air ventilation used Yes 56 227.9 (2.79) < 0.001 No 60 445.9 (2.55) Location of classroom Main building 96 286.3 (2.57) Portable 17 469.0 (2.75) <0.005 Annexe 3 1748.8 (8.31)t School placement arterial 51 332.0 (3.03) residential 42 409.1 (2.08) O.05 treed/agriculture 23 195.8 (3.27)t Building age < 1949 30 385.8 (1.89) 1950- 1974 59 324.1 (3.10) 1975 - 1995 10 96.6(1.67)t 0.001 portable classroom 17 469.0 (2.75) | Group significantly different from other groups by Scheffe post hoc procedure. " Geometric mean. b Geometric standard deviation. 89 Table 33. Continuous environmental and occupancy variables significantly related to indoor mesophilic fungal concentration Continuous variables n Correlation coefficient p-value Pearson r Mean indoor relative humidity (%) 116 + 0.396 < 0.001 Mean indoor temperature (°C) 116 + 0.194 <0.05 Mean outdoor temperature (°C) 116 + 0.458 O.001 Equilibrium RH (at window) 116 + 0.306 0.001 Equilibrium RH (at corridor) 116 + 0.347 O.001 Outdoor meso.fungi (In CFU/m3) 116 + 0.418 < 0.001 Indoor bacteria cone. (In CFU/m3) 116 + 0.364 < 0.001 Table 34. Categorical environmental and occupancy variables significantly related to indoor mesophilic fungal concentration Variable n Mesophilic fungi CFU/m3 p-value GM f l (GSD) Season Winter 35 172.3 (2.81)t Spring 42 422.8 (2.97) < 0.001 Fall 39 422.7 (2.00) Snow cover Yes 5 88.6 (2.26) < 0.005 No 111 341.8(2.73) Signs of moisture in room None 56 244.9 (3.15) Old stains 51 430.7 (2.34) <0.05 Current signs 9 346.9 (2.32) t Group significantly different from other groups by Scheffe post hoc procedure. a Geometric mean. b Geometric standard deviation. Selection of variables for the regression equation (mesophilic fungal concentration). Independent variables were selected for ventilation, site specific and environmental factors. The most highly associated ventilation measure was supply air flow, and the voluntary use of windows to supplement ventilation. Building age was the most significant site specific variable. The environmental factors in the model were the outdoor temperature and fungal concentration. The occupancy factors that were 90 significant were CO2 and mesophilic bacterial concentrations (presumably as markers of bioeffluents). Only one occupancy factor could be entered because these factors were highly correlated with each other. The regression equation including CO2 is summarized in Table 35; the model including mesophilic bacterial concentration is summarized in Table 36. Table 35. Predictors of indoor mesophilic fungal concentration (In CFU/m3) including the independent variable "CO2 concentration" Variable Unstandardized coefficients P Std. error Standardized t p-coefficients value Supply air flow (m3/sec) - 1.047 0.320 Windows open <'/2day 0.549 0.194 > 1 / 2 d a y 0.029 0.155 default = not open Indoor C0 2 (In ppm) 0.625 0.212 Building age < 1949 -0.534 0.224 1950-1974 -0.325 0.199 1975- 1995 - 1.609 0.300 default = portable Mean outdoor temperature (°C) 0.113 0.017 Outdoor fungi (In 0.305 0.099 CFU/m3) Intercept (In) fungi (CFU/m3) = - 0.898 R 2 = 0.581 Std error of the estimate = 0.6960 -0.220 -3.270 0.001 0.193 2.825 0.006 0.014 0.190 NS 0.224 2.943 0.004 - 0.228 -2.391 0.019 -0.158 -1.633 NS - 0.439 -5.362 O.001 0.540 6.653 O.001 0.220 3.075 0.003 -0.552 0.582 91 Table 36. Predictors of indoor mesophilic fungal concentration (In CFU/m3) including the independent variable "mesophilic bacterial concentration" Variable Unstandardized coefficients (3 Std. error Standardized t p-coefficients value Supply air flow (m7sec) -0.868 Windows open < y2 day 0.552 > Vi day - 0.066 default = not open Indoor bacteria 0.260 (In CFU/m3) Building age < 1949 - 0.589 1950-1974 -0.461 1975- 1995 - 1.476 default = portable Mean outdoor temperature (°C) 0.099 Outdoor fungi (In 0.276 CFU/m3) Intercept (In) fungi (CFU/m3) = 2.355 R 2 = 0.593 Std error of the estimate = 0.681 0.322 0.191 0.148 0.072 0.219 0.193 0.300 0.015 0.099 -0.183 0.196 -0.031 0.255 0.253 0.225 0.406 0.478 0.198 -2.692 0.008 2.892 0.005 -0.448 NS 3.633 O.001 -2.692 -2.389 0.008 0.019 -4.918 O.001 6.484 O.001 2.782 0.006 3.457 0.001 The presence of supply air was associated with a reduced mesophilic fungal concentration, while the use of open windows to supplement ventilation was associated with higher indoor fungi counts. The best variable to describe "season" was outdoor temperature, with lower counts associated with colder outdoor temperatures. Lowest indoor fungal counts were found in the newest buildings except for "portables" in which the highest counts were seen. As expected, there was a strong relationship with outdoor concentration of fungi. There was a positive correlation with markers of occupancy. Both independent variables describing bioeffluent, CO2 and bacterial concentration, contributed independently to the model, but could not be included in the same model. When both bioeffluent variables were added to the model, CO2 ceased to be significant (p = 0.264) 92 and the significance of bacterial concentration was reduced (p = 0.033 vs. p = O.001) with concomitant reductions in the standardized coefficient values (P) of both variables. This example illustrates the problems of putting highly correlated variables into a multiple linear regression model. The positive correlation between indoor fungal counts and bacteria may be due to the ability of fungi to use shed skin cells as a nutrient source (Homer et al. 1999), but the presence of skin cells, mesophilic bacteria and CO2 are all inextricably linked as occupant generated bioeffluent. Alternate measures of ventilation were tried in the model (air exchange rate, and exhaust flow) but none functioned as well as supply air flow. Although ventilation systems are often held suspect as disseminators of bioaerosols, this apparently contradictory finding could be explained if the filters are reasonably well maintained, and the supply air contributed cleaner air to the room which reduced the concentration of indoor spores. Other variables describing occupancy were tried in the model (number of occupants, room use patterns) but none was as effective in the model as indoor CO2 or mesophilic bacteria. Variables describing furnishings, presence of carpets, animals, plants, or aquaria were not significant in the model. Measurements of indoor relative humidity were strongly correlated with outdoor temperature, and only one of the two variables could be used in the model. Outdoor temperature was chosen as the stronger predictor. 7.2.2 Xerophilic and Thermotolerant Fungi Xerophilic and thermotolerant fungi are simply subsets of the larger group of saprophytic fungi whose source is the outdoor environment. Selective media or incubation conditions were used to enhance the growth of these groups of fungi for the purpose of enumerating those organisms that may be potentially pathogenic or may have a specific niche in building materials. The predictive power of the multiple linear regression equation was less for the outcome of indoor xerophilic and again very much less for the outcome of thermotolerant fungi. The numbers of samples above the LOD and the concentrations for these outcomes were less than for the mesophilic fungi with a reduction of significant relationships to independent variables. All variables found significant for mesophilic 93 fungi were offered into the model for either xerophilic or thermotolerant fungi. The predictors of indoor xerophilic fungal concentration are summarized in Table 37. Table 37. Predictors of indoor xerophilic fungal concentration (In CFU/m3) Variable Unstandardized coefficients P Std. error Standardized t coefficients P-value Ventilation available & operating test day - 0.651 Indoor C0 2 (In ppm) 0.737 Building age < 1949 1950- 1974 1975 - 1995 default = portable Mean outdoor temperature (°C) Outdoor xerophilic fungi (CFU/m3) Intercept (In) fungi (CFU/m3) = - 3.558 R2 = 0.508 Std error of the estimate = 1.1172 -0.164 - 0.081 -1.015 0.124 0.480 0.222 0.332 0.352 0.317 0.476 0.028 0.098 -0.212 0.177 - 0.046 - 0.026 -0.186 0.397 0.401 -2.924 2.220 -0.466 -0.256 -2.134 4.372 4.893 0.004 0.029 NS NS 0.035 O.001 O.001 -1.430 0.156 Thermotolerant fungi The only independent variable significantly related to indoor thermotolerant fungal concentrations was outdoor thermotolerant fungal concentration (R2 = 0.19). Higher numbers of samples above the LOD would be needed to increase the predictive power of the independent variables for this outcome. 7.2.3 Outcome Variable: Indoor Mesophilic Bacterial Concentration (CFU/m3) Independent variables were examined for their relationships with the outcome variable of bacterial concentration in univariate analyses. The variables were broadly grouped into three areas, namely, site, environment and occupant related measurements. Those site variables that were significantly related to indoor mesophilic bacteria are 94 summarized in Tables 38 and 39, environmental measurement in Tables 40 and 41 and occupancy characteristics in Table 42. Table 38. Continuous site and ventilation characteristics related to indoor mesophilic bacterial concentration Continuous variables n Correlation coefficient p-value Pearson r Exhaust ventilation (mVsec) 115 -0.189 <0.05 Supply air (m3/sec) 115 . -0.322 < 0.001 Air changes per hour (hi) a 115 - 0.456 < 0.001 Room volume (m3) 116 + 0.214 <0.05 Measured from SF6 tracer gas decay. Table 39. Categorical site and ventilation characteristics related to indoor mesophilic bacterial concentration Variable n Mesophilic bacteria CFU/m3 p-value GM» (GSD)* Ventilation available Yes 85 188.6 (2.83) < 0.001 No 30 381.3 (2.01) Ventilation used on test day Yes 62 166.3 (3.02) < 0.001 No 53 325.5 (2.09) Central forced air available Yes 36 117.8(3.51) < 0.001 No 79 305.4 (2.01) Elevation above sea level m 100-199 6 195.1 (1.68) 200 - 299 18 228.4 (2.37) 300-399 37 257.6 (2.53) < 0.05 400 - 499 42 267.7 (2.52) 500 - 599 12 90.9 (4.34)f Building age < 1949 30 235.3 (2.63) 1950- 1974 58 268.2 (2.24) 1975- 1996 10 65.8 (4.71)t < 0.001 portable 17 247.6 (2.45) f Group significantly different from other groups by Scheffe post hoc procedure. 0 Geometric mean. Geometric standard deviation. 95 Table 40. Continuous environmental variables significantly related to indoor mesophilic bacterial concentration Continuous variables n Correlation coefficient Pearson r p-value Mean indoor relative humidity 115 + 0.272 < 0.005 Mean outdoor relative humidity 115 + 0.333 < 0.001 Indoor mesophilic fungi 115 + 0.364 < 0.001 (In CFU/m3) Indoor thermotolerant fungi 115 + 0.305 < 0.001 (In CFU/m3) Table 41. Categorical environmental variables significantly related to indoor mesophilic bacterial concentration Variable n Mesophilic bacteria CFU/nr* *p-value GM f l (GSD)6 Season Winter 35 266.2 (2.45) Spring 42 165.5 (3.16) <0.05 Fall 38 276.7 (2.37) Precipitation Yes 71 265.9 (2.46) <0.05 No 44 175.1 (3.08) Signs of moisture in room None 55 168.0 (2.96)t Old stains 51 280.1 (2.41) < 0.005 Current signs 9 425.5 (1.80) * ANOVA. | Group significantly different from other groups by Scheffe post hoc procedure. a Geometric mean. b Geometric standard deviation. Table 42. Continuous occupancy variables significantly related to indoor mesophilic bacterial concentration Continuous variables n Correlation coefficient p-value Pearson r Indoor C0 2 cone. (In ppm) Til + 0.584 % of day occupants spent moderately active 115 +0.235 % of day occupants spent sitting 115 -0.195 quietly < 0.001 <0.05 <0.05 96 Selection of variables for the regression equation for indoor bacterial concentration. The significant variables were examined. Choices had to be made between continuous and categorical variables that described the same relationship to the outcome variable to choose the variable that would contribute the most to the equation. Also, variables were scrutinized to determine the logic of the relationship to the outcome variable. Finally, dummy coding was used to enable the inclusion of nominal categorical variables. The final multiple linear regression model is presented in Table 43. Table 43. Predictors of indoor mesophilic bacterial concentration (In CFU/m3) Variable Unstandardized coefficients p Std. error Standardized t coefficients P-value Ventilation rate (unoccupied) In ACH -0.146 0.056 Indoor C0 2 (In ppm) 1.509 0.218 Building age < 1949 0.354 0.222 1950-1974 0.414 0.198 1975 - 1995 - 0.388 0.295 default = portable Signs of moisture Old stains 0.298 0.134 Current moisture 0.610 0.250 default = no signs % of day spent in -0.0088 0.003 quietly sitting Room volume (m3) 0.00424 0.001 Mean outdoor 0.03781 0.016 temperature (°C) Intercept (In) bacteria (CFU/m3) = - 6.505 R2 = 0.606 Std error of the estimate = 0.6652 -0.194 0.549 0.153 0.205 -0.109 0.147 0.163 -0.182 0.211 0.184 -2.624 0.010 6.924 O.001 1.592 NS 2.092 0.039 -1.317 NS 2.216 0.029 2.445 0.016 -2.683 0.009 3.217 0.002 2.381 0.019 -4.172 O.001 The term describing ventilation is the air exchange rate in the unoccupied room, which was negatively correlated with the outcome. Terms describing the physical characteristics of the room are the age of the building and room volume. Buildings in the 97 group built between 1950 and 1974 had higher bacterial counts than either the older or younger buildings. The room volume was positively correlated with the bacterial counts, which was independent of the number of occupants. Environmental characteristics on the test day were entered as signs of moisture and outdoor temperature. Lower outdoor temperature may favour reduced ventilation, or more indoor activities for occupants. Increased activity was positively correlated to bacterial concentration in univariate analysis, but was not retained in the final regression model. Terms describing occupancy patterns are the indoor CO2 concentration and the percent of day the occupants spent quietly sitting. The indoor concentration of CO2 is predicted both by number of occupants and efficiency of ventilation, which is logically related to the generation of bioeffluent which includes bacteria. The CO2 concentration was a more powerful predictor of bacterial concentration than numbers of occupants or any ventilation term that described "in use" characteristics on the test day because the CO2 concentration term combines occupant density and ventilation as an indicator of bioeffluent. Variables tested in the model but not significantly associated with bacterial counts were measurements of exhaust or supply ventilation, type of ventilator; building construction type or placement, presence of carpet, plants, animals or aquaria; measures of cleanliness; indoor and outdoor relative humidity; or age of occupants. 7.2.4 Outcome Variable: CO? Concentration (In ppm) Independent univariate variables were examined to determine which would be entered into the regression model for CO2 concentration. The independent variables could be grouped into three broad categories, ventilation, environment or occupant related. The statistically relevant ventilation variables are summarized in Tables 44 and 45, environmental variables in Tables 46 and 47, and occupant variables in Table 48. Table 44. Continuous ventilation variables with statistically significant relationships to CO2 concentration Continuous variables n Correlation coefficient p-value Pearson r Air changes per hour " (In) 115 -0.314 0.001 Exhaust flow mVsec n « • 1 1 1 1 . 116 - 0.223 <0.05 a Air changes per hour calculated from SF6 tracer gas decay. 98 Table 45. Categorical ventilation variables with statistically significant relationships to CO2 concentration Independent variable n C 0 2 GM (GSD) ppm *p-value Mechanical ventilation possible Yes 86 893.3 (1.45) < 0.005 No 30 1117.8(1.38) Mechanical ventilation in used on test day Yes 62 844.6(1.43) < 0.001 No 54 1079.1 (1.41) Central forced air Yes 37 763.9(1.38) < 0.001 No 79 1046.7(1.42) Vane axial fan (supply or exhaust) Yes, in use on test day 18 839.2(1.26) Present, not on 17 1256.0(1.40)1 < 0.005 None 81 916.4(1.46) Windows Not open on test day 53 1058.0(1.41) Open < Yz day 18 1007.0(1.46) < 0.001 Open > l/ 2 day 45 810.1 (1.42)f Corridor door No corridor door (portable) 17 1154.7(1.54) Open < Vi day 9 1012.8(1.53) O.05 Open > V2 day 90 905.6(1.41)1 Outside door No outside door/never open 35 939.1 (1.43) Open intermittently 73 986.5 (1.44) Open < % day 6 744.7(1.39) <0.05 Open > % < V2 day 2 454.7(1.13)1 * ANOVA. f Group significantly different from other groups by Scheffe post hoc procedure. Table 46. Continuous environmental variables with statistically significant relationships to CO2 concentration Continuous variables n Correlation coefficient Pearson r p-value Mean indoor temperature °C Mean outdoor temperature °C 116 116 - 0.390 - 0.460 < 0.001 < 0.001 99 Table 47. Categorical environmental variables with statistically significant relationships to CO2 concentration Independent variable n C 0 2 GM" (GSD)° *p-value ppm Season Winter 35 1184.8 (1.35) Spring 42 816.2(1.46) < 0.001 Fall 39 907.8 (1.39) Building construction Masonry or concrete 20 1115.7(1.27) Wood frame 74 875.6(1.41) Cinder block 5 792.3 (1.82) < 0.005 Portable 17 1154.7 (1.54) *ANOVA. t Group significantly different from other groups by Scheffe post hoc procedure. " Geometric mean. b Geometric standard deviation. Table 48. Continuous occupancy variables with statistically significant relationships to CO2 concentration Continuous variables n Correlation coefficient p-value Pearson r Number of occupants in room 116 + 0.192 <0.05 Occupants per 100 m3 of room 116 + 0.184 <0.05 volume % of day spent moderately active 116 + 0.244 <0.01 Concentration of indoor bacteria 116 + 0.584 < 0.001 Selection of variables for the regression equation (CO? concentration) The significant variables were examined. Choices were made between continuous and categorical variables that described the same relationship to the outcome variable to choose the variable that would contribute the most to the equation. Also, variables were scrutinized to determine the logic of the relationship to the outcome variable. Finally, dummy coding was used to enable the inclusion of nominal categorical variables. Variables describing ventilation conditions, environmental conditions and occupancy conditions were entered into a regression equation. In the first model created, one of the variables (In indoor bacterial concentration) which had a very high correlation 100 with In CO2 concentration was included. Both CO2 and indoor bacteria are products of human occupancy and level of activity, and comprise part of the contribution of bioeffluent to the perception of indoor air quality. However, it would not be reasonable to use bacterial counts to predict C0 2 concentration. Although the source of much of the mesophilic bacterial population was the occupants, simply substituting the term describing the number of occupants for the term describing the concentration of bacteria reduced the predictive power of the equation. Various combinations of variables describing occupants and occupation patterns were tested as substitutes for the single variable describing bacterial concentration. The inclusion of two terms, one describing the number of occupants per 100 m3 of indoor space and the other describing the percentage of the day the room was empty, resulted in a final model that was stable and had a similar predictive value and standard error of the estimate as did the original equation in which bacteria was included. The variable describing the percentage of day the room was empty was not, by itself, significantly correlated with C0 2 concentration. Variables describing ventilation were examined for the best predictive power in the equation. It was found that although there was a reasonably good (p < 0.05) negative correlation between measured exhaust ventilation and CO2 concentration, the exhaust variable was highly skewed due to the small number of rooms equipped with exhaust capabilities. Variables describing environmental conditions were also closely examined. The categorical variable describing season and the continuous variables describing indoor and outdoor temperatures were independently tested in the model, and the best variable, mean outdoor temperature, was chosen for the final equation. The final regression equation is given in Table 49. 101 Table 49. Predictors of indoor CO2 concentration Variable Unstandardized coefficients P Std. error Standardized coefficients P t p-value ACH (tracer gas decay) - 0.088 0.023 - 0.241 -3.825 O.001 Windows open < Vi day - 0.009 0.062 -0.010 -0.160 NS > l/ 2 day -0.168 0.048 - 0.222 -3.475 0.001 default = not open Outside door open intermittently - 0.066 0.047 - 0.087 -1.419 NS < Vi day - 0.357 0.101 -0.215 -3.533 0.001 > V2 day - 0.605 0.163 -0.214 -3.719 0 . 0 0 1 default = never open Mechanical ventilation operating on test day Mean outdoor temperature °C Mean indoor relative humidity % Number of occupants per m3 room volume % of day room empty Intercept (In) C0 2 ppm = 6.599 R2 = 0.682 Std error of the estimate = 0.2200 -0.138 - 0.048 0.021 0.027 - 0.005 0.047 0.006 0.003 0.009 0.002 -0.186 -2.956 0.004 - 0.632 -8.501 O.001 0.450 6.046 O.001 0.167 2.871 0.005 -0.216 -3.497 0.001 38.59 O.001 Also tried in the model but not retained were variables describing the construction, age and placement of the buildings; the specific types of ventilators; and the presence of animals, plants or aquaria in the rooms. Measurements of both natural and mechanically supplied ventilation were important in the model. All forms of ventilation (windows, doors and whether a mechanical system was operating) served to reduce the concentration of C0 2 . The number of air changes per hour, an expression of the efficiency of ventilation as designed for the room, was also highly significant in the model. Lower outdoor temperatures predicted higher C0 2 presumably due to the necessity of closing windows or reducing fresh air into the forced air heating systems. Rooms that were never or seldom empty 102 during the day tended to accumulate higher CO2 concentrations. Greater numbers of occupants per 100 m3 of room volume were associated with higher CO2 levels. The relationship of relative humidity may be due to the trend for RH to be lower in ventilated rooms than in non-ventilated rooms. 7.3 SUMMARY The study was conducted during the defined seasons of fall, winter and spring. Season, or the climate (temperature, relative humidity) defined by the season, had significant effects on measurements of bioaerosols and ventilation efficiency. Fungal aerosols were significantly higher indoors and outdoors during spring and fall (p < 0.001). Indoor mesophilic bacterial aerosols were significantly lower in spring (p< 0.05). Carbon dioxide concentrations were highest in winter (p < 0.001) when the ventilation efficiencies were lowest (p < 0.005). Each outcome parameter was affected differently by the climate. Outdoor fungal concentrations vary with season (as defined by characteristic temperature and relative humidity), and to a lesser extent by the amount or proximity of foliage or decaying plant material (Kozak et al. 1985). In this study the indoor fungal concentration was strongly related to the outdoor concentration which would suggest the necessity of including outdoor measurements in order to evaluate indoor concentration. Fungal concentrations were correlated with indoor relative humidity and equilibrium relative humidity (r = 0.396 and 0.347 respectively, p < 0.001). Indoor humidities were significantly lower in winter, and were well below 60% RH, an indirect measure of equilibrium relative humidity of the building materials (Pasanen et al. 1994). Building construction materials, presence of carpet, plants, or live animals, age of occupants, housekeeping or placement of schools were tested in the regression equations but were not significantly related to outcome bioaerosol concentrations and were not retained in the models. CO2 and mesophilic bacterial concentrations, the bioeffluents released by occupants, were lowest in spring which coincided with high ventilation efficiencies. Opening windows and doors to air out rooms significantly decreased occupant bioeffluents, while at the same time allowing entry of outdoor fungal spores. The simple entry of fungal spores cannot be regarded as a detriment for the majority of people who 103 would be exposed to the same spores by going outside, but may present a problem for the atopic individual who may not be able to find relief from exposure to allergens. Because seasonal variation contributes to worst case or best case sampling scenarios, the seasonal contribution to indoor air quality parameters should be acknowledged in future study designs. The use or availability of ventilation also had significant effects on the outcomes. Use of mechanical ventilation significantly decreased bioeffluent concentrations (p <0.001) and the use of supply air reduced fungal concentrations (p O.OOl). These data would suggest that air quality could be improved by furnishing clean, uncontaminated air to rooms. The observation of mechanically supplied air reducing fungal concentration would seem to be at variance with investigations of sick building syndrome, which often identify the HVAC system as the source of biologic contaminants. However, the ability of properly maintained air supply systems to dilute fungal contaminants has recently been confirmed in studies being conducted at Duke University (Thomann and Tulis 1999) where the air stream was isolated from the room air prior to mixing, and was found to have lower spore concentrations. 104 CHAPTER 8 DISCUSSION This study was initiated to investigate and evaluate the range of selected indoor characteristics encountered in school rooms and to determine the relationships of environment, building, ventilation and occupancy to the outcome measures of bioaerosols and CO2. Rooms were randomly chosen to represent the range of building ages, construction styles and ventilation sources commonly found in this geographic area. Although the relationship between indoor bioaerosols and symptoms of sick building syndrome were discussed in order to motivate this work, it should be re-emphasized that this was not a study of health effects. This study was undertaken to systematically develop a database of exposure measures using field measurements to describe the "normal" range of values found under differing conditions of season and ventilation. In the school district studied, ventilation systems were sometimes available but not used, or not used to their potential. Many, but not all of the portable classrooms had reversible vane axial fans mounted on the side of the unit. Teachers were reluctant to use these fans due to the noise created in a confined space. In the winter the fans could not be used for supply air because the air would be untempered. In some cases teachers were unaware that the units existed. Another type of ventilator in operation was an independent, gas-fired unit mounted in the side wall of the room. Some of these unit ventilators were found to have their intake louvres closed, restricting the unit to recirculating air rather than providing fresh air. A few schools were situated near industrial areas whose malodorous effluents were sufficiently objectionable that the teachers couldn't use windows or doors for supplemental ventilation. Ideally the ventilation provided would be filtered to remove allergen bearing particles and would deliver sufficient fresh air to dilute contaminants within the room. 105 The occupancy patterns in the classrooms were different from the stereotypical office environment where workers tend to be at their desks during most of the morning and afternoon resulting in bimodal CO2 peaks. There were only a few cases where, for security reasons or for outdoor odour problems, rooms were kept closed and locked. These rooms also tended to have high bacterial concentrations by the end of the day. The concurrent finding of high CO2 and bacterial levels could lead to the perception of "stale" air and hence to complaints about the indoor air quality as suggested by Fanger (19886), Molhave (1982) and Batterman and Peng (1995). 8.1 NON-COMPLIANT ROOMS 8.1.1 CO? Concentrations Local workplace regulations have recently been enacted that address indoor air quality as it applies to C0 2 (WCB 1998). Using the criteria of 1000 ppm (650 ppm from occupants) as a surrogate of acceptable ventilation, 47% of the classrooms studied here would fail. The relationship of CO2 concentration to indoor air quality is not well understood, and elevated C0 2 is often interpreted as a health hazard, which has not been shown to be the case. However, inadequate ventilation in rooms occupied by numerous students or by environmentally sensitive students may lead to complaints of stuffiness or stale air due to the olf factors produced by the occupants (Fanger 19886) or other indoor sources of pollutants. The ASHRAE committee currently reviewing ventilation guidelines will be making new recommendations in the near future that may decrease the target ventilation rate for public buildings (E. Chessor, WCB, personal communication). Although this may lessen the need for immediate remediation on the grounds of compliance, lowering the ventilation requirement will not likely improve the perception of the indoor environment (Fanger 19886; Mendell 1993; Batterman and Peng 1995). In infection control models, reducing the ventilation rate serves to increase the probability of contagion spread (Burge 19906; Wheeler 1993). 106 8.1.2 Temperature and Relative Humidity The temperature was generally well regulated in these study schools. Exceptions to this were nine rooms (8%) that were outside target ranges for season and for these rooms the primary reason seemed to be due to thermostat controls that were not responsive to conditions in the room being tested. The relative humidity, with the exception of two rooms, was well below the target of 60% RH. The rooms examined could not be described as damp, nor were large areas of water damaged building materials seen. Nine rooms were below 30% RH (ASHRAE 1989) and 44 (38%) rooms were below 40% RH (Arundel et al. 1986). These low humidities may contribute to complaints of dry eyes for wearers of contact lenses (Arundel et al. 1986; Engineering Interface 1989). It would be inadvisable, however, to humidify these rooms unless there were pressing and consistent complaints, as the potential problems posed by increased humidity far outweigh those imposed by low humidity in this temperate climate. 8.1.3 Fungal Concentrations The wide range of counts, spatially and temporally, encountered in bioaerosol studies has been regarded as the greatest challenge to interpretation of such data. In this study, temporal variation was illustrated by the variation in seasonal counts. For individual samples, however, the mean coefficient of variation of duplicate samples for mesophilic fungi was 15.3% (range 0 - 85%). The coefficient of variation of duplicates was below 20% for 89 (79%) of samples, the CV level considered acceptable for biologic systems (Dillon et al. 1996). The overall geometric mean indoor mesophilic fungal count was 323.8 CFU/m3 (range 14-18,583). The highest indoor count value in the study was of a nearly pure culture of Penicillium. This was clearly an abnormal result and subsequent investigation located a faulty drain running beneath the room. Other elevated, non representative counts of Penicillium (n = 14 rooms) or Aspergillus (n = 1 room) were similarly investigated and discussed with district maintenance staff. In some of these cases, water damaged carpet had recently been removed, or in one case a salmon enhancement aquarium had overflowed. It was possible with the aid of the plant operations staff to 107 identify rooms where remediation had already begun due to previous leaks. This would seem to support the case for an educated maintenance and support staff. It becomes more problematic when other criteria are used to evaluate fungal counts. If guidelines more appropriate to office buildings are used for school buildings (Health Canada 1995a), 94 (81%) of rooms would be reevaluated for having counts over 150 CFU/m3. Forty-one (35%) classrooms had counts greater than 500 CFU/m3 (Health Canada 1995a). However, by referring to the current ACGIH criteria (Burge et al. 1999), 30 (26%) classrooms had indoor counts higher than outdoor counts. Of these, 13 (11%) had a different ranked flora indoors than outdoors. The other 17 (15%) had either the same ranked flora as outdoors or the predominant fungal group found was sterile mycelia which could not be further characterized. Examination of the distribution of data from this study enabled 95% confidence intervals to be calculated around the geometric means for indoor fungi and bacteria. Twenty-three mesophilic fungal concentrations exceeded the upper confidence limit for the data set. The practice of using 95% confidence intervals around a mean to assess single point measurements comes from exposure assessments in workplace settings where the exposure being measured is created from an industrial process (NIOSH method, Leidel et al. 1977). The data from exposure monitoring studies is often log-normal in distribution, as was the bioaerosol data. The geometric standard deviation of industrial processes is usually in the range of 1.5 to 3.5 (Hawkins et al. 1991). The geometric standard deviation for mesophilic fungi was 2.8; for mesophilic bacteria, 2.7. There were no measurements of either parameter that were below the limit of detection. The 95%o confidence intervals for mesophilic fungi and mesophilic bacteria could be applied to this geographic/climatic location, but could not be broadly applied to other climatic zones. Another approach would be to narrow the spread of the data by defining more stringent localized criteria, for example, grouping together data from a season or based on a room characteristic such as ventilation use, an approach similar to creating homogeneous exposure groups in workplaces (Hawkins et al. 1991). However, such groupings did not greatly narrow the geometric standard deviations in this data set. In the case of the thermotolerant and xerophilic fungi, (GSD of 3.9 and 4.7 respectively) the 108 interpretation of 95% confidence intervals was more problematic because of the greater spread of the data and the number of samples falling below the limit of detection. If it were possible to continue this study, it would be of value to investigate each room that failed to meet either the Health Canada guidelines or the ACGIH recommendations to further identify building or room features that may contribute to these counts. For example, the interior of the wall cavity may have had unseen water damage or colonization of fungi. This may be further explored by drilling holes in the wall and visually exploring the wall cavity by fibre-optic methods, or by measuring the conductance of insulation or building material to determine wetness within the wall. Additional information that could potentially be of use would be surfacial geologic maps that identify soil types associated with former stream beds under the foundations. Measurements of pressure differentials between rooms and, for example, crawl spaces may also contribute to future investigations of fungal sources. Possible future approaches to the estimate of fungal exposure would be to use surrogate measures of spore mass, particularly of the allergen or toxin assays. In vitro tests are currently under development. At Wageningen Agricultural University, The Netherlands, enzyme linked assays are being developed to detect Penicillium and Aspergillus extracellular polysaccharide (EPS) and (l-»3)-(3-D glucan (Boleij et al. 1995; Douwes et al. 1998a). In Sweden a Limulus derived assay is used to detect (1—»3)-|3-D glucan (Goto et al. 1994) and a polymerase chain reaction (PCR) tests specific to Stachybotrys chartarum has been developed (Land and Must 1999). In Canada and Finland, a fungal membrane component, egosterol, is measured as a surrogate of fungal mass (Miller and Young 1997; Pasanen et al. 1999). Improving techniques of measurement or surrogate measurement of fungal mass will contribute to the research that is needed to determine potential links between fungal concentration and ill health. For some individuals the mechanism may be through allergy, for other individuals other fungal components may be responsible (Burge 1995; Dillon et al. 1996; Cruz et al. 1997; Sudakin 1998). For example, non specific inflammatory responses to glucans may play an important role (Goto et al. 1994; Douwes etal. 1998a). 109 8.1.4 Bacterial Concentrations There are no Canadian guidelines for acceptable bacterial concentrations in public buildings. In this study it can be seen that the bioeffluents (CO2 and mesophilic bacteria) are correlated (Pearson r = 0.584, p < 0.001). The source of indoor CO2 and bacterial concentrations higher than outdoor levels is primarily from occupants of the space. One jurisdictional example of incorporation of abacterial concentration into indoor air quality criteria is the Hong Kong Interim Indoor Air Quality Guidelines which lists 1,000 CFU/m (1 hour average) as an upper limit for indoor bacteria (Lee and Chang 1999). The sampling protocol to determine a one hour average specifies the use of a Portable Air Sampler (PAS) and a flow rate of 20 mL/min. In this study only two rooms (1.7%) exceeded 1,000 CFU/m3, however, the differing sampling methodologies used make direct comparisons problematic to interpret. No rooms in this study were found to be contaminated with predominately Gram-negative bacteria, nor were elevated levels of thermophilic actinomycetes found. High concentrations of Gram-negative bacteria are a source of exposure to endotoxin (Burge 1985; Hood 1990; Burge 1995). High concentrations of thermophilic actinomycetes have been associated with cases of hypersensitivity pneumonitis (Burge et al. 1987; Salvaggio 1997; Weltermann et al. 1998). The low concentrations of these groups of organisms are consistent with the assumption that the schools in this study were not identifiable as "sick schools" based on bacterial characterization. 110 8.2 STRENGTHS OF STUDY 8.2.1 Site Selection The schools were randomly assigned to a sampling season which allowed seasonal effects to be measured. The number of schools to be tested, and the finite length of the seasons, required two seasonal cycles to complete the study. There are few long term studies of schools in the literature (Norback et al. 1990; Cooley et al. 1998), and the bulk of published reports are from clusters of sampling periods occurring over, at most, one season (Daneault et al. 1992; Gyntelberg et al. 1994; Koskinen et al. 1994; Willers and Andersson 1996; Thorn et al. 1996; Liu 1998; Lee and Chang 1999). The schools within the study district were built over an eighty-year period, exemplifying the use of different building materials. The use and availability of ventilation varied within the district. However, there was common maintenance supervision and use of cleaning protocols throughout the district. This allowed both naturally and mechanically ventilated rooms to be sampled within the same area and season, and sometimes within the same school. Minimal differences between schools were seen in general cleanliness or landscaping. 8.2.2 Ventilation Systems, Air Exchange Rates, and Comfort Parameters The use of full day CO2 data allowed the graphic visualization of the dynamic quality of school room use. The ranges of CO2 minima and maxima during the day showed a different pattern than that normally seen in office buildings, where a fewer number of occupants have more predictable work habits. The value of the CO2 data was greatly enhanced by having both the observation of room use patterns and the resultant CO2 concentrations. These data could potentially be used to further define C0 2 generation rates which could result in suggesting maximum student populations in rooms otherwise unable to control occupant bioeffluents (Persily 1993; Olcerst 1994a; Olcerst 19946; Jankovic et al. 1996). The measurements and observations made of the ventilation system and air exchange rates may be used to prioritize strategies to increase comfort for occupants of the classrooms by identifying those rooms that lack sufficient fresh air to dilute indoor contaminants. i l l 8.2.3 Carbon Dioxide Use of a data logging instrument allowed full day CO2 concentrations to be measured. The dynamic quality of the CO2 concentrations would have been missed if only grab samples had been taken. Care was taken in the placement of the instrument to represent an area of the room with well mixed air. The monitor also had to be placed in an area not readily reached by students whose natural curiosity was to want to take the instrument apart or to see the numbers change by blowing on the detector. In most rooms these requirements were met by placing the monitor on the top of the partition dividing the classroom from the cloak-room. However, the potential bias of this placement was not tested. 8.2.4 Bioaerosols The choice of using culturable bioaerosol counts as the measurement of exposure remains controversial in the field of bioaerosol research. Alternatives could have included spore counts or chemical surrogates of fungal mass such as egosterol measurement (Miller and Young 1997), but for the reasons described below, cultural counts were chosen. It was necessary to be able to identify the components of the aerosol population. Although not all fungal or bacterial species would be able to grow on the culture media chosen, there is no evidence that fungi cultured simultaneously indoors and outdoors would be preferentially biased except in cases where overwhelming competition from another organism was encountered. However, it was equally possible that the competing organism would be encountered indoors or outdoors, depending on the circumstances. Although the microscopic examination of spores allows limited identification, spore counts cannot differentiate between Penicillium and Aspergillus. Cell counts cannot differentiate between Micrococcus and Staphylococcus, nor between saprophytic Staphylococcus and Staphylococcus aureus Identification of the airborne fungi to genus level or to bacterial group allowed samples to be processed in a timely fashion. The time and resources required to identify each colony to the species level would have been prohibitive for large numbers of 112 samples. In order to use statistical tests, large numbers of samples were required to have sufficient power to avoid type II errors (inability to detect a true difference between means). As was seen in the data set, differential colony counts made inferences about the source or origin of the bioaersol possible. Identification to species would be absolutely necessary in the investigation of cases of purported building related illness. The addition of selective media and incubation conditions allowed the recovery of fungi that would not otherwise have been enumerated. Eurotium, a xerophilic fungi, was recovered from only 10% NaCl-MEA, although it will grow on MEA. Exposure to high concentrations of this organism has been associated with a reported case of hypersensitivity pneumonitis (Yoshida, et al. 1990). Aspergillus fumigatus, a thermotolerant fungus, was recovered on MEA incubated at 37°C more often than on plates incubated at room temperature. This organism is a potential pathogen for the immunocompromised population, is potentially allergenic, produces toxins, and has been associated with cases of hypersensitivity pneumonitis (Smith and Moss, 1985; Day 1996; Cruz, et al. 1997). In this study the geometric mean counts of the xerophilic and thermotolerant organisms were generally low. However, it has been suggested by Gravesen et al. (1994) that xerophilic organisms may find unique ecological niche in substrates such as dust or building materials that are intermittently wet. This information is necessary for proactive building maintenance. 8.3 LIMITATIONS OF STUDY 8.3.1 Site Selection The school district selected for study was not randomly chosen. Due to financial and logistical constraints the schools were required to be in the lower mainland British Columbia. Two school districts that fulfilled the expectations of the study design were invited to participate. The first district approached turned down the opportunity. The health and safety coordinator of the second district may have realized the value to the district of the data that would be generated, and accepted. However, the increased awareness of indoor air quality issues shown by this coordinator may have influenced the management and maintenance of the schools within his control and biased the schools towards better than average care. 113 No systematic information was collected from teachers or students regarding health effects or complaints about the air quality. However, anecdotal evidence supported the assumption that the schools were not pristine and not without complaints. Anecdotally there were individual complaints of perceived increased incidence of respiratory symptoms and asthma. However, no systematic health surveys have been conducted in this school district. At the same time there were commendations of the district for the progressive approach taken to health and safety concerns. The most important feature biasing the results was the attention paid to repairs. The general maintenance of the district seemed to be above average at the time of the study. However, with decreased funding levels this may be in the process of change and revisiting the same district in future may yield very different results. Classrooms within the schools were systematically chosen using a decision tree to represent different conditions of ventilation or placement within the building. This may have introduced a bias since a truly random sample of classrooms was not attempted, and air quality measurements unique to randomly chosen rooms may have been missed. Specialized rooms such as computer laboratories or gymnasiums were specifically excluded as were all non-instructional rooms or offices. These choices may have biased the results towards less problematic environments. On only two occasions of the room selection was the first choice vetoed by the principal on the grounds that the teacher would not accommodate the disruption of the testing equipment, etc. Alternate rooms were chosen with approximately the same features and it was not felt this unduly biased the results. 8.3.2 Air Exchange Rates A limitation of the study was that the tracer gas decay trials were done when the rooms were not occupied. The protocol of testing during school breaks was developed when principals voiced resistance to the testing because it was too disruptive to the classroom procedures. The natural curiosity of the students regarding instrumentation, and the length of the test period would have made the tracer gas procedure difficult to do during normal school hours. In addition, classroom use was very different from typical 114 office use in that students were in and out at irregular intervals which would have made even closing doors or windows for the length of the trial problematic. 8.3.3 Equilibrium Relative Humidity The measurement of equilibrium relative humidity was attempted by sealing a thermohygrometer into an airtight box affixed to the wall. In all classrooms tested the equilibrium RH was close to the RH of the room. This indicated that the walls were in equilibrium with the room and condensation moisture was not present. The time chosen for the test (four hours) was based on laboratory mock-ups conducted in the Occupational Hygiene Laboratory. Moisture problems inside the wall could not be ruled out using this method. The vapour pressure of the water within the wall would have to be high enough to penetrate the painted or wallpapered surface in order to selectively increase the humidity at the wall surface. Water damage inside walls would not be expected to be evenly distributed throughout the wall, and the standardized placement of the thermohygrometer may have missed localized moisture accumulation. However, in a subset of rooms, a Delmhorst BD-9 moisture meter was used to penetrate the surface of the wall and to determine wall moisture directly. All such measurements were within the "green" zone (0 - 80% moisture for plaster or 10 - 15% for wood). It was accepted that these data indicated walls were not currently wet. It is possible a measurement device that could access the interior cavity of the wall or insulation may have revealed sources of moisture not observed using the techniques described. 8.3.4 Carbon Dioxide Measurement Outdoor CO2 measurements were only available for the first seasonal cycle due to the failure of a heater within the instrument which allowed measurements to be taken when the temperature was below room temperature. This failure did not affect indoor readings as all rooms were maintained at temperatures reasonably close to 20 °C. The monitor required AC power to function. There were three rooms with temporary gaps in the data when children interfered with the power cord. However, these 115 gaps did not represent a significant part of the day, and the missing data was assigned an average value based on the activity level in the room during the interval in question. 8.3.5 Total Suspended Particles A limitation of this study was the low sensitivity of the particulate matter measurements. This was due in part to the choice of filter media, and in part to the choice of size selectivity of the sampling equipment. The filter media and size selectivity would have been appropriate if subsequent speciation of the dust was warranted. The high limit of detection of the method resulted in unacceptably high numbers of samples below the limit of detection. Current guidelines for indoor air dust levels are size selective for PMio or PM2.5 particulate matter. This made a direct comparison with TSP measurements problematic. Formulae for the conversion of TSP to P M i 0 are available, however, given the lack of sensitivity of the measurement these were not applied to the data. 8.3.6 Fungal Aerosols A methodological limitation of the study was the choice of using culturable bioaerosol counts as the measurement of exposure. It is well established that the culturable fraction of the bioaerosol represents only a fraction of the total cell or spore content of the air (Burge et al. 1999). There is no direct conversion method to estimate total spores given culturable data, nor culturable fraction given spore count data. The media chosen for the study were general purpose media recommended by the ACGIH bioaerosols committee for culturable samples (Dillon et al. 1996). However, other media choices were possible and could have returned different results. Media more nutritionally enriched may have allowed the growth of fastidious organisms, and more inhibitory media may have restricted the growth of fast growing organisms that may have obscured the presence of slower growing colonies. The counts reported in this study were entirely dependent on the media and incubation conditions used which remained constant throughout the study. The media chosen for the culture of xerophilic organisms (10% NaCl-MEA) is a more restrictive media than DG-18, a media designed for use in food science that has been adapted for aerobiology (Beuchat 1987). However, this choice was considered 116 acceptable even though the counts were lower than if DG-18 had been used. The aw of 10% NaCl-MEA was within in the range of the xerophiles of interest, and allowed the recovery of Eurotium, a xerophilic fungus. The sampling protocol for culturable bioaerosols restricted the samples to point samples (grab samples) and the variation throughout the day was not assessed. Recent work by H. Neumeister-Kemp at Technische Hygiene, Freie Universitat Berlin, Germany, confirms diurnal fluctuations of individual fungal species' concentrations (H. Neumeister-Kemp, personal communication). The samples could not be taken during the course of the school day due to the excessive noise made by the high volume pumps and the excessive curiosity shown by the children. The limitation of reliance on airborne culture may have influenced the recovery of Stachybotrys. Stachybotrys spores are large, and when actively growing, wet and sticky, preventing them from easily becoming airborne. Although Stachybotrys grows readily on MEA, it is unlikely that it would be recovered due to the competitive advantage of faster growing organisms such as Cladosporium and Penicillium (Dillon et al. 1996). The viability of Stachybotrys spores are known to be low even under optimum circumstances (Muilenberg and Burge 1994). The absence of Stachybotrys from the identified cultures cannot be taken as an absence of spores; however, there were no visible signs of fungal colonization in the classrooms. Fungal colonies that did not sporulate could not be identified by the methods used in this study. The grouping together of these all non-sporing organisms as "sterile mycelia" obscured the representation of Basidiomycetes, and may have resulted in the underrepresentation of Alternaria species, an allergenic fungi commonly found in indoor environments (Macher et al. 1999). Sparsely sporulating Alternaria may have been counted in the "sterile mycelia" group. 8.3.7 Bacterial Aerosols Many of the same methodological limitations apply to the viability of bacterial cells from air samples. The predominantly Gram-positive bacterial population is indicative of organisms that are able to survive desiccation. The methodology chosen for this study may have restricted the recovery of Gram-negative aero flora even though the 117 media chosen supports luxuriant growth of water derived cultures of Pseudomonas sp. based on previous testing. 8.3.8 Generalization of the Data The field data indicating a relationship between ventilation and bioeffluents (bacteria and CO2) should be generalizable to the principles of dilution ventilation, number of occupants and room usage. Other independent variables that may be unique to this study, for example, building age, may be measures of other building related factors such as availability of ventilation or construction practices. It would require additional study in a different school system to determine alternate variables to use to construct more generalizable predictive models. The fungal data for this study was collected entirely from an urban, coastal, temperate climate zone. The concentrations of indoor and outdoor fungi may differ significantly in rural, agricultural areas or in climates with dry cold winters or extremes of relative humidity. In this study, none of the schools tested had air-conditioning available. Some types of air conditioning systems dehumidify the air as it is cooled, creating potential problems with the resulting condensate. The schools studied may have been above average in respect to having daily janitorial attention, a centralized maintenance staff and were without significant overgrowth of landscaping plants or bushes. Although bioaerosol concentrations would vary with climate and proximity to composting or agricultural sources, relationships to ventilation and outdoor temperature should remain valid. This data set may not bear any resemblance to either arctic or tropical climates. It is not expected that buildings known to be colonized with fungi would exhibit similar relationships to buildings only transiently populated with outdoor fungal spores. 8.4 CONTRIBUTION TO THE LITERATURE This is the first study in British Columbia undertaken to establish a database of bioaerosol measurements in non-complaint elementary schools. The study design incorporated several unique features not previously reported in the literature. 118 The sampling periods were strictly defined by season. In British Columbia, spring and fall day time temperatures are similar, however, there are significant differences in the growth or decay of vegetation. Winter temperatures were significantly colder, although there were only a few sampling days with snow cover on the ground. Indoor fungal concentrations were found to be related to methods of ventilation which have not been previously reported for this climate, nor have any studies evaluated fungal concentrations found in schools against criteria developed for use in office environments. Selective media and incubation conditions allowed fungal genera to be enumerated that would normally be missed or underrepresented by using only one media suited for the culture of mesophilic, hydophilic fungi and these relationships have not been reported for non-problem buildings. The use of field data to construct predictive equations to clarify the relationships between building systems, ventilation, and bioeffluent factors on indoor concentrations of culturable bioaerosols has not previously been reported. 119 CHAPTER 9 CONCLUSIONS AND RECOMMENDATIONS 9.1 CONCLUSIONS FROM THIS STUDY The primary objectivies of the study were met. Specifically, ranges of culturable fungal and bacterial aerosol, and occupant generated CO2 concentration in elementary school classrooms were described. The relationships between specific building, environment and occupant factors and the outcomes of interest were explored. The concentration ranges of fungal aerosols were compared to currently available guidelines. There are few guidelines with which to compare concentrations of bacterial aerosols, however, the relationship between bacterial concentrations and characteristics of ventilation were discussed. The ranges of values measured in this study for fungal aerosols and CO2 concentrations exceeded commonly used guidelines for the interpretation of indoor air quality, suggesting that guidelines promulgated for use in office buildings are insufficient for the interpretation of public buildings such as school classrooms. More data are needed to develop CO2 concentration guidelines for high occupancy rooms. More data are also needed to develop bioaerosol concentration guidelines for the ranges of climatic zones found in Canada. This study had sufficient power to show significant relationships among commonly measured indoor air quality parameters. The season and environmental conditions in which the sampling was conducted as well as building characteristics such as availability and use of ventilation were significantly related to the outcomes of indoor carbon dioxide, mesophilic bacterial and mesophilic and xerophilic fungal concentrations. Specifically, in these non-problem rooms, the factors that were associated with higher mesophilic fungal concentration were the use of natural ventilation (open windows), increased outdoor temperature, increased outdoor fungal counts, and a factor presumed to be a surrogate for "occupants" (bacteria or CO2 concentration). The factors that were associated with decreased fungal concentrations were increased 120 mechanical ventilation and all ages of buildings except a group of portable classrooms built primarily in the late 1980's and early 1990's. A similar constellation of factors were associated with xerophilic fungal concentrations. In contrast, increased natural ventilation was associated with lower CO2 concentration, while both CO2 and bacteria were reduced with increased mechanical ventilation. Increased outdoor temperature, however, was associated with increased bacterial and decreased C0 2 concentration. A secondary objective, to evaluate carbon dioxide concentration as a marker of the contribution of the occupants' bioeffluents to the indoor environment, was also explored. Although this study was not able to measure all components contributed by occupants to the indoor environment which may influence the perception of indoor air quality, some intriguing relationships were investigated between bacterial aerosol and CO2 concentrations and measures of ventilation. Clear differences in concentration of fungal, bacterial and occupant-generated CO2 were shown in relation to the presence and use of mechanical ventilation compared to natural ventilation. However, the relationships are complex, and methods of natural ventilation which may be useful for decreasing the concentration of bacterial aerosols and CO2 may serve to increase the concentration of saprophytic fungal aersols. No single indoor air quality parameter can be used to fully characterize indoor air quality. The value of this study was to suggest relationships that may be of value in determining systems that may be out of balance with each other, and which may indicate problems that should addressed by building engineers or maintenance personnel. Additional data from health effects or epidemiologic studies may be necessary to determine health-based limits of exposure to contaminants of indoor air. 9.2 RECOMMENDATIONS FOR FURTHER INVESTIGATION: • Indoor air quality guidelines formulated primarily for use in office buildings are conservative regarding C0 2 and bioaerosol concentrations when used to evaluate typical elementary school exposures. Additional research is necessary to formulate appropriate concentration guidelines for indoor air quality in naturally ventilated and 121 • Classrooms without access to mechanical ventilation were not able to meet current B.C. Workers' Compensation Board guidelines for indoor air quality based on CO2 limits. Classrooms that do not currently have mechanical ventilation may benefit from provision of tempered fresh air to dilute occupant-generated CO2 and bioeffluents. Pre and post intervention studies could be conducted if the rooms without current ventilation were provided with ventilation. • Indoor RH was below 60% for 98% of rooms tested. Humidity measured at the wall surface was in equilibrium with the room humidity. The presence of moisture in wall cavities could not be ruled out. Future investigations should include the examination of wall cavities using a horoscope or other fibre-optic tool to test the relationship of hidden moisture with indoor fungal concentrations. • Rooms with indoor fungal counts which exceeded concurrent outdoor counts (when outdoor counts are above the LOQ), or which were not representative of outdoor flora should be examined for sources of bioamplification. • Rooms with indoor concentrations of fungi higher than outdoor concentrations should be re-examined to determine if these data are consistent. Records of room repairs or interventions that may contribute to fungal concentrations should be obtained from maintenance personnel. • There were insufficient numbers of rooms with plants, live animals or aquaria to demonstrate additional sources for exposure to culturable fungal and bacterial aerosols. Additional data would be required to fully evaluate potential contributions of these furnishings to indoor microbial exposures. • This study failed to show additional risk of exposure to airborne fungi and bacteria in rooms that were carpeted. The carpets in the rooms examined were dry at the time of testing, and may have passivly trapped settled particles. Additional studies could be planned to investigate the potential of carpet materials to support microbial growth under differing moisture or humidity conditions. • Occupant activity and the way rooms were used influenced the concentration of bacterial aerosols and CO2. These contributions to the indoor envirorment by the occupants themselves may have specific implications for high occupancy rooms such as classrooms and may require guidelines specific for these rooms. 122 9.3 CONCLUSIONS For the present it would seem prudent to evaluate fungal concentrations in B.C. using the following criteria. • Fungal concentrations can be expected to be higher in the fall and spring and lower in winter. • Indoor concentrations should be lower than outdoor concentrations, when outdoor concentrations exceed the LOQ. • The predominating indoor fungal species should be reflective of the outdoor flora sampled on the same day near the source of fresh air for the room. • Sample concentrations exceeding the 95% confidence interval of the fungal mean for the season should be further investigated. 123 LITERATURE CITED 1999. Breaking out of the mould: indoor fungus is bad for your health, even worse for your children's [Stachybotrys chartarum]. Montreal Gazette, March 4, p D4. 1999. Black, slimy mould a danger to health: experts agree more research needed on links to illness [Stachybotrys]. Toronto Star January 29, pp. F1-F2. 1998. Toxic mould found in schools. Vancouver Sun May 19, p. A8. 1997. Air quality closes Nova Scotia school: fungus attacks other facilities. Daily Commercial News 70(238): A5. ACGIH. 1999. TLVs® and BEIs® Threshold Limit Values for Chemical Substances and Physical Agents; Biological Exposure Indices. American Conference of Governmental Industrial Hygienists, Cincinnati. 1992. Industrial Ventilation. 21st ed. A Manual of Recommended Practice. American Conference of Governmental Industrial Hygienists, Cincinnati. Adan, O.C.G., R.A. Samson, and J.Th.M. Wijnen. 1994. Fungal resistance tests; a proposed method for testing resistance of interior finishes. In R.A. Samson, B. Flannigan, M.E. Flannigan, A.P. Verhoeff, O.C.G. Adan, E.S. Hoekstra, (editors). Health Implications of Fungi in Indoor Environments. Air Quality Monographs Volume 2. Elsevier, Amsterdam Alberta Health. 1993. Indoor air quality assessment: a working manual. Alberta Health, Environmental Services. Edmonton. Andersen, A.A. 1958. New Sampler for the Collection, Sizing and Enumeration of Viable Airborne Particles. Journal of Bacteriology 76: 471-484. Andersson, M.A., R. Mikkola, J. Peltola, J. Helin, C. Apetroaie, and M. Salkinoja-Salonen. 1999. Toxin producing Bacillus strains isolated from indoor environment. In Proceedings of Indoor Air 99, The 8th International Conference on Indoor Air Quality and Climate. Edinburgh, Scotland. Construction Research Communications Ltd., London, UK. Vol. 2, pp 529-530. ANSI/ASHRAE. 1981. ASHRAE Standard 55-1981. Thermal Environmental Conditions for Human Occupancy. American Society of Heating, Refrigerating and Air-conditioning Engineers, Inc. Atlanta. Apte, M.G. and J.M. Daisey. 1999. VOCs and "sick building syndrome": application of a new statistical approach for SBS research to U.S. EPA base study data. In Proceedings of Indoor Air 99, The 8th International Conference on Indoor Air Quality and Climate. 124 Edinburgh, Scotland. Construction Research Communications Ltd., London, UK. Vol. l,pp 117-122. Apter, A., A. Bracker, M. Hodgson, J. Sidman, and W-Y. Leung. 1994. Epidemiology of the sick building syndrome. Journal of Allergy and Clinical Immunology 94:277-288. Arundel, A.V., E.M. Sterling, J.H. Biggin and T.D. Sterling. 1986. Indirect health effects of relative humidity in indoor environments. Environmental Health Perspectives 65: 351-361. ASHRAE. 1989. ASHRAE Standard 62-1989. Ventilation for Acceptable Indoor Air Quality. American Society of Heating, Refrigerating and Air-conditioning Engineers, Inc. Atlanta. Bartlett, K.H. 1993. Ambient and indoor particulate monitoring in selected B.C. communities. Report written for the B.C. Ministry of Health, Environmental Health Assessment, Protection and Safety. Bartlett, K.H. 1995. Burnaby aerosol project. Report written for the B.C. Ministry of Health, Environmental Health Assessment, Protection and Safety. Bartlett, K.H.and B. Nakahara. 1994. Winter monitoring for ambient and indoor particulate in the communities of Creston, Lumby, Quesnel, Duncan, Parksville and Victoria. Report written for the B.C. Ministry of Health, Environmental Health Assessment, Protection and Safety. Barrett, D. 1998. The renewal of trust in residential construction: Commission of inquiry into the quality of condominium construction in British Columbia. Submitted to the Lieutenant-Governor in Council, Government of British Columbia. Batterman, S. and C. Peng. 1995. TVOC and CO2 concentrations as indicators in indoor air quality studies. American Industrial Hygiene Association Journal 56(1): 55-65. Bearg, D.W. 1993. Indoor Air Quality and HVAC Systems. Lewis Publishers, Boca Raton. Beaumont, F., H.F. Kauffman, H.J. Sluiter, and K. deVries. 1984. A volumetric-aerobiologic study of seasonal fungus prevalence inside and outside dwellings of asthmatic patients living in Northeast Netherlands. Annals of Allergy 53(6): 486-492. Becker, R. 1994. Fungal disfigurement of constructions - analysis of the effects of various factors. In RA. Sampson, B. Flannigan, M.E. Flannigan, S.P. Verhoeff, O.C.G. Adan, and E.S. Hoekstra (editors). Health Implications of Fungi in Indoor Environments. Air Quality Monograph Volume 2. Elsevier Press, Amsterdam. 125 Beuchat, L.R. 1987. Influence of water activity on sporulation, germination, outgrowth, and toxin production. In L.B. Rockland and L.R. Beuchat (editors). Water Activity: Theory and Applications to Food. Marcel Dekker, Inc. New York. Boleij, J., E. Buringh, D. Heederik, and H. Kromhout. 1995. Measurement methods for biologic agents. In Occupational Hygiene of Chemical and Biological Agents. Elsevier Science, Amsterdam, The Netherlands. Brooks, B.O., F.D. Aldrich, G.M. Utter, J.A. DeBroy and R.D. Schimke. 1992. Immune responses to pollutant mixtures from indoor sources. Annals of the New York Academy of Sciences 641: 199-214. Brunekreef, B., D.W. Dockery, F.E. Speizer, J.H. Ware, J.D. Spengler, and B.G.Ferris. 1989. Home dampness and respiratory morbidity in children. American Review of Respiratory Disease 140: 1363-1367. Burge, H.A. 1985. Indoor sources for airborne microbes. In R.B. Grammage, S.V. Kaye, and V.A. Jacobs (editors). Indoor Air and Human Health. Lewis Publishers, Inc. 1990a. The fungi. In P.R. Morey, J.C. Feeley, Sr., and J.A. Otten (editors). Biological Contaminants in Indoor Environments, ASTM STP 1071. American Society for Testing and Materials, Philadelphia. 19906. Risks associated with indoor infectious aerosols. Toxicology and Industrial Health 6(2): 263-274. 1990c. Bioaerosols: prevalence and health effects in the indoor environment. Journal of Allergy and Clinical Immunology 86(5): 687-701. 1995. Aerobiology of the indoor environment. Occupational Medicine: State of the Art Reviews 10(1): 27-40. Burge, H.A., M. Chatigny, J. Feeley, K. Kreiss, P. Morey, J. Otten, and K. Peterson. 1987. Guidelines for assessment and sampling of saprophytic bioaerosols in the indoor environment. Applied Industrial Hygiene 5(2): R10-R16. Burge, H.A., J.M. Macher, D.K. Milton, and H.M. Ammann. 1999. Data Evaluation. In Bioaerosols, Assessment and Control. American Conference of Governmental Hygienists. Cincinnati, OH. Burge, H.A. and J.A. Otten. 1999. Fungi. In Bioaerosols, Assessment and Control. American Conference of Governmental Hygienists. Cincinnati, OH. Burstyn, I. and K. Teschke. 1999. Studying the determinants of exposure: a review of methods. American Industrial Hygiene Association Journal 60: 57-72. 126 CEP A. 1998. National Ambient Air Quality Objectives for Particulate Matter. Parti. Science Assessment Document. Report by the CEPA/FPAC Working Group on Air Quality Objectives and Guidelines. Health Canada, Ottawa, ON. CMHC. 1998. Toxic mold cleanup procedures: A guide for cleanup contractors. Canada Mortgage and Housing Corporation, draft 5b. Ottawa, ON. Chang, C C , R.A. Ruhl, G.M. Halpern, and M.E. Gershwin, 1993. The sick building syndrome. I. Definition and epidemiological considerations. Journal of Asthma 30(4): 285-295. Cooley, J.D., W.C Wong, C A . Jumper, and D.C. Straus. 1998. Correlation between the prevalence of certain fungi and sick building syndrome. Occupational and Environmental Medicine 55(9): 579-584. Cone, J.E. and Shusterman, D. 1991. Health effects of indoor odorants. Environmental Health Perspectives 95: 53-59. Cruz, A., M. Saenz de Santamaria, J. Martinez, A. Martinez, J. Guisantes, and R. Palacios. 1997. Fungal allergens from important allergenic fungi imperfecti. Allergologia et Immunopathologia 25(3): 153-158. Dales, R.E., H. Zwanenburg, R. Burnett, and C A . Franklin. 1991. Respiratory health effects of home dampness and molds among Canadian children. American Journal of Epidemiology 134(2): 196-203. Daneault, S., M. Beausoleil, and K. Messing. 1992. Air quality during the winter in Quebec day-care centers. American Journal of Public Health 82(3): 432-434. Day, J.H. 1996. Allergenic respiratory responses to fungi. In The Mycota: A Comprehensive Treatise on Fungi as Experimental Systems for Basic and Applied Research. Volume VI. Human and Animal Relationships. K. Esser and PA. Lemke (editors), D.H. Howard and J.D. Miller (volume editors). Springer-Verlag, Berlin. Dillon, H.K., PA. Heinsohn, and J.D. Miller. 1996. Field Guide for the Determination of Biological Contaminants in Environmental Samples. American Industrial Hygiene Association, Fairfax, VA. Douwes, J., G. Doekes, J. Heinrich, A. Koch, W. Bischof, and B. Brunekreef. 1998a. Endotoxin and P(l—»3)-D-glucan in house dust and the relation with home characteristics: a pilot study in 25 German houses. Accepted for publication in Indoor Air. In J. Douwes (Ph.D. thesis) Respiratory Health Effects of Indoor Microbial Exposure: A Contribution to the Development of Exposure Assessment Methods, Agricultural University of Wageningen, The Netherlands. 127 Douwes, J, A. Zuidhof, G. Doekes, S. van der Zee, M. Boezen, and B. Brunekreef. 19986. P(l->3)-D-glucan and endotoxin in house dust and peak flow variability in asthmatic children. Submitted for publication. In J. Douwes (Ph.D. thesis) Respiratory Health Effects of Indoor Microbial Exposure: A Contribution to the Development of Exposure Assessment Methods, Agricultural University of Wageningen, The Netherlands. Duchaine, C , A. Meriaux, G. Brochu, K. Bernard, and Y. Cormier. 1999. Saccharopolyspora rectivirgula from Quebec dairy barns: amplification of simplified criteria for the identification of an agent responsible for farmer's lung disease. Journal of Medical Microbiology 48(2): 173-180. Engineering Interface Ltd. 1989. Healthy Building Manual: Systems, parameters, problems & solutions. Prepared for Coordinating Council of Commercial, Institutional and Agricultural Energy Management Task Forces. Minister of Supply and Services Canada, Ottawa, ON. EPA. 1995. Indoor Air Quality. Tools for Schools. IAQ Coordinator's Guide. U.S. Environmental Protection Agency, Washington, D.C. Etkin, D.S. 1994. Particulates in Indoor Environments: Characterization and Health Effects. Cutter Information Corp., Arlington, MA. Etzel, R.A., E. Montana, W.G. Sorenson, G.J. Kullman, T.M. Allan, and D.G. Dearborn. 1998. Acute pulmonary hemorrhage in infants associated with exposure to Stachybotrys atra and other fungi. Archives of Pediatrics and Adolescent Medicine 152(8): 757-762. Ezeonu, I.M., J.A. Noble, R.B. Simmons, D.L. Price, S.A. Crow, and D.G. Ahearn. 1994. Effect of relative humidity on fungal colonization of fibreglass insulations. American Society of Microbiologists Annual Meeting, Q233. Fanger, P.O. 1988a. The olf and decipol. ASHRAE Journal 30(10): 35-38. 19886. Hidden olfs in sick buildings. ASHRAE Journal 30(11): 40-43. Fernandez-Caldas, E., W.L. Trudeau, and D.K. Ledford. 1994. Environmental control of indoor biologic agents. Journal of Allergy and Clinical Immunology 94(2): 404-411. Finnegan, M.J., C.A.C. Pickering, and P.S. Burge. 1984. The sick building syndrome: prevalence studies. British Medical Journal 289: 1573-1575. Flannigan, B. 1995. Guidelines for evaluation of airborne microbial contamination of buildings. In Proceedings of the International Conference Fungi and Bacteria in Indoor Air Environments, Health Effects, Detection and Remediation. E. Johanning and C.S. Yang (editors). Eastern New York Occupational Health Program, Latham, NY. 128 Foarde, K.K., E.C. Cole, D.W. VanOsdell, DJ. Bush, D.L. Franke, and J.C.S. Chang. 1994. Characterization of environmental chambers for evaluating microbial growth on building materials. In R.A. Sampson, B. Flannigan, M.E. Flannigan, S.P. Verhoeff, O.C.G. Adan, and E.S. Hoekstra (editors). Health Implications of Fungi in Indoor Environments. Air Quality Monograph Volume 2. Elsevier Press, Amsterdam. Fogelmark, B. and R. Rylander. 1997. (l-»3)-P-D glucan in some indoor air fungi. Indoor and Built Environment 6:291-294. Fogelmark, B., M. Sjostrand, and R. Rylander. 1994. Pulmonary inflammation induced by repeated inhalations of P(l—»3)-D-glucan and endotoxin. International Journal of Experimental Pathology 75:85-90. Godish, T., J. Rouch, D. McClure, L. Elro, and C. Seaver. 1986. Ventilation system performance in a new classroom building assessed by measurements of carbon dioxide levels. Managing Indoor Air for Health and Energy Conservation. Proceedings of the ASHRAE Conference IAQ'86. American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. pp. 603-610. Gormon, R.W. and K.M. Wallingford. 1989. The NIOSH approach to conducting indoor air quality investigations in office buildings. In N.L. Nagda and J.P. Harper (editors). Design and Protocol for Monitoring Indoor Air Quality, ASTM STP 1002. American Society for Testing and Materials, Philadelphia. Goto, H , K. Yuasa, and R. Rylander. 1994. (1—>3)-P-D glucan in indoor air, its measurement and in vitro activity. American Journal of Industrial Medicine 25:81-83. Gravesen, S., J.C. Frisvad, and R.A. Samson. 1994. Microfungi. Munksgaard, Copenhagen. Gyntelberg, F., P. Suadicani, J.W. Nielsen, P. Skov, O. Valbjom, PA. Nielsen, T. Schneider, O. Jorgensen, P. Wolkoff, C.K. Wilkens, S. Gravensen, and S. Norn. 1994. Dust and the sick building syndrome. Indoor Air 4:223-238. Hake, W., H.G. Neumeister-Kemp, P.C. Kemp and H. Martiny. 1999. Reduction of micro-organisms by HVAC system filters and analysis of the micro-organisms passing through the filters. In Proceedings of Indoor Air 99, The 8th International Conference on Indoor Air Quality and Climate. Edinburgh, Scotland. Construction Research Communications, Ltd, London, UK. Vol. 4, p. 43. Hawkins, N.C., S.K. Norwood, and J.C. Rock (editors). 1991. A strategy for occupational exposure assessment. Appendix III. American Industrial Hygiene Association, Akron, OH. 129 Hazleden, D. 1999. Envelope drying rates experiment start-up report. Report prepared for Canada Mortgage and Housing Corporation by Forintek Canada Corporation, contract 98-2221, Vancouver, BC. Health Canada. 1995a. Indoor air quality in office buildings: A technical guide. A report of the Federal-Provincial Advisory Committee on Environmental and Occupational Health. Ottawa, ON. 19956. Fungal contamination in public buildings: A guide to recognition and management. Health Canada Federal-Provincial Committee on Environmental and Occupational Health. Ottawa, ON. Hill, B.A., B.F. Craft, and J.A. Burkart. 1992. Carbon dioxide, particulates, and subjective human responses in office buildings without histories of indoor air quality problems. Applied Occupational and Environmental Hygiene 7(2): 101-111. Hockaday, P.S. 1988. The architect's concern about indoor pollution. In R.B. Kundsin (editor). Architectural Design and Indoor Microbial Pollution. Oxford University Press. Hodgson, M. 1992. Field studies on the sick building syndrome. Annals of the New York Academy of Sciences 641: 21-36. Hodgson, M.J., P. Morey, W.Y. Leung, L. Morrow, D. Miller, B.B. Jarvis, H. Robbins, J.F. Halsey, and E. Storey. 1998. Building-associated pulmonary disease from exposure to Stachybotrys chartarum and Aspergillus versicolor. Journal of Occupational and Environmental Medicine 40(3): 241-249. Holt, J.G., N.R. Krieg, P.H.A. Sneath, and J.T. Staley, S.T. Williams (editors) 1994. Bergey's Manual of Determinative Bacteriology, 9th Edition. Williams & Wilkins, Baltimore. Hood, M. A. 1990. Gram negative bacteria as bioaerosols. In Biological Contaminants in Indoor Environmental, ASTM STP 1071. P.R. Morey, J.C. Feeley, Sr., J.A. Otten (editors). American Society for Testing and Materials, Philadelphia. Horner, W.E., S.B. Lehrer, and J.E. Salvaggio. 1999. Aerobiology. In Allergens and Allergen Immunotherapy, 2n d Edition. R.F. Lockey and S.C. Bukantz (editors). Marcel Dekker, Inc., New York. Homung, R.W. and L.D. Reed. 1990. Estimation of average concentration in the presence of nondetectable values. Applied Occupational and Environmental Hygiene 5(1): 46-51. Horvath, E.P. 1997. Building Related Illness and Sick Building Syndrome: from the specific to the vague. Cleveland Clinic Journal of Medicine 64(6): 303-309. 130 Infante, P.F., L.D. Schuman, J. Dement, and J. Huff. 1994. Fibrous Glass and Cancer. American Journal of Industrial Medicine 26: 559-584. Jaakkola, J.J.K., P. Tuomaala, and O. Seppanen. 1994. Air recirculation and sick building syndrome: a blinded crossover trial. American Journal of Public Health 84:422-428. Jankovic, J.T., R. Ihle, and D.O. Vick. 1996. Occupant generated carbon dioxide as a measure of dilution ventilation efficiency. American Industrial Hygiene Association Journal 57(8): 756-759. Jarvis, B.B., W.G. Sorenson, E.L. Hintikka, M. Nikulin, Y. Zhou, J. Jiang, S. Wang, S. Hinkley, R.A. Etzel, and D. Dearborn. 1998. Study of toxin production by isolates of Stachybotrys chartarum and Memnoniella echinata isolated during a study of pulmonary hemosiderosis in infants. Applied and Environmental Microbiology 64(10): 3620-3625. Jarvis, B.B., Y. Zhou, J. Jiang, S. Wang, W.G. Sorenson, E.L. Hintikka, M. Nikulin, P. Parikka, R.A. Etzel, and D.G. Dearborn. 1996. Toxigenic molds in water-damaged buildings - dechlorogriseofulvins from Memnoniella echinata. Journal of Natural Products 59(6): 553-554. Johanning, E., R. Biagini, D. Hull, P. Morey, B. Jarvis, and P. Landsbergis. 1996. Health and immunology study following exposure to toxigenic fungi (Stachybotrys chartarum) in a water-damaged office environment. International Archives of Occupational and Environmental Health 68(4): 207-218. Kenny, L.C., R. Aitken, C. Chalmers, J.F. Fabries, E. Gonzales-Fernandez, H. Kromhout, G. Liden, D. Mark, G. Riediger, and V. Prodi. 1997. A collaborative European study of personal inhalable aerosol sampler performance. Annals of Occupational Hygiene 41(2): 135-153. Kem, M.E. 1985. Medical Mycology. FA . Davis Company, Philadelphia. Kjaergaard, S. 1992. Assessment of eye irritation in humans. Annals of the New York Academy of Sciences 641: 187-198. Kjaergaard, S. and O.F. Pedersen. 1992. Sensitivity of the eyes to airborne irritant stimuli: influence of individual characteristics. Archives of Environmental Health 47(1): 45-50. Koneman, E.W. and G.D. Roberts. 1985. Practical Laboratory Mycology, 3 rd ed. Williams & Wilkins, Baltimore. Koskinen, O., T. Husman, A. Hyvarinen, T. Reponen, and A. Nevalainen. 1994. Respiratory symptoms and infections in mouldy dwellings and day-care centres. In R.A. Sampson, B. Flannigan, M.E. Flannigan, S.P. Verhoeff, O.C.G. Adan, and E.S. Hoekstra 131 (editors). Health Implications of Fungi in Indoor Environments. Air Quality Monograph Volume 2. Elsevier Press, Amsterdam. Kozak, P.P., J. Gallup, L.H. Cummins, and S.A. Gillman. 1985. Endogenous mold exposure: environmental risk to atopic and nonatopic patients. In R.B. Gammage, S.V. Kaye, and V.A. Jacobs (editors). Indoor Air and Human Health. Lewis Publishers, Inc., New York. Kreiss, K. 1989. The epidemiology of building-related complaints and illness. Occupational Medicine: State of the Art Reviews 4(4): 575-592. Kruczak-Filipov, P. and R.G. Shively. 1992. Gram stain procedure. In H.D. Isenberg (editor), Clinical Microbiology Procedures Handbook, Volume 1. American Society for Microbiology, Washington, DC. Kwon-Chung, K.J. and J.E. Bennett. 1992. Medical Mycology. Lea&Febiger, Philadelphia. Land, C.J. and A. Must. 1999. Detection and Identification of Stachybotrys chartarum from damp gypsum boards by the polymerase chain reaction (PCR. In Proceedings of Indoor Air 99, The 8th International Conference on Indoor Air Quality and Climate. Edinburgh, Scotland. Construction Research Communications, Ltd, London, UK. Vol. 4, pp. 939-941. Lee, S.C. and M. Chang. 1999. Indoor air quality investigations at five classrooms. Indoor Air 9(2): 134-138. Leidel, N.A., K.A. Busch, and J.J. Lynch. 1977. Occupational exposure sampling strategy. National Institute for Occupational Health and Safety, DHEW (NIOSH) publication number 77-173. Cincinnati, OH. Levin, H. 1995. Physical factors in the indoor environment. Occupational Medicine: State of the Art Reviews 10(1): 59-94. Liu, L.-J. 1998. Bacterial exposures and health effects among school children in South Carolina. Epidemiology 9(4): Abstract 452. Lundquist G.R., M. Iversen and J. Dorsgaard. 1982. Indoor climate in low-ventilated day-care institutions. Environment International 8: 139-142. Macher, J., H.A. Ammann, H.A. Burge, D.K. Milton, P.R. Morey (editors). 1999. Bioaerosols: Assessment and Control. American Conference of Governmental Hygienists. Cincinnati, OH. Malloch, D. 1981. Moulds. Their Isolation, Cultivation and Identification. University of Toronto Press, Toronto. 132 Maresh, CM. , L.E. Armstrong, S.A. Kavouras, G.J. Allen, D.J. Casa, M. Whittlesey, and K.E. LaGasse. 1997. Physiological and psychological effects associated with high carbon dioxide levels in healthy men. Aviation, Space, and Environmental Medicine 68(1): 41-45. Meckler, M. 1993. Evaluating IAQ compliance in buildings utilizing carbon dioxide level monitoring. Environments for People, IAQ '92. American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. pp. 355-364. Mendell, M.J. 1993. Non-specific symptoms in office workers: a review and summary of the epidemiologic literature. Indoor Air 3: 227-236. Meyer, B. 1983. Indoor air quality. Addison-Wesley Publishing Company, Inc. Reading, MA, pp. 105-109. Miller, J.D. and J.D. Young. 1997. The use of egosterol to measure exposure to fungal propagules in indoor air. American Industrial Hygiene Association Journal 58:39-43. Molhave, L. 1982. Indoor air pollution due to organic gases and vapours of solvents in building materials. Environment International 8: 117-127. . 1992. Human reactions to controlled exposures to VOCs and the "total VOC" concept. In H. Knoppel and P. Wolkoff (editors). Chemical, Microbiological, Health and Comfort Aspects of Indoor Air Quality - State of the Art in SBS. ECSC, EEC, EAEC. Brussels and Luxenbourg. Printed in the Netherlands. Molhave, L., B. Bach, and O.F. Pedersen. 1986. Human reactions to low concentrations of volatile organic compounds. Environment International 12: 167-175. Morey, P.R. 1985. Overview. In R.B. Gammage, S.V. Kaye, and V.A. Jacobs (editors) Indoor Air and Human Health. Lewis Publisher Inc., New York. pp. 133-137. Morris, P.I. 1997. Moisture problems in Vancouver condominiums. Paper presented to The Structural Board Association Annual Meeting. Muilenberg, M.L. and H.A. Burge. 1994. Filter cassette sampling for bacterial and fungal aerosols. In R.A. Samson, B. Flannigan, M.E. Flannigan, A.P. Verhoeff, O.C.G. Adan, and E.S. Hoekstra (editors). Health Implications of Fungi in Indoor Environments. Air Quality Monographs Volume 2. Elsevier, Amsterdam. Munro, B.H. and E.B. Page. 1993. Statistical Methods for Health Care Research. 2n d ed. J.B. Lippincott Company. Philadelphia, PA. 133 Nelson, N.A., J.D. Kaufman, J. Burt, and C. Karr. 1995. Health symptoms and the work environment in four non-problem United States office buildings. Scandinavian Journal of Work and Environmental Health 21:51-59. New York City Department of Health. 1998. Guideline on assessment and remediation of Stachybotrys atra in indoor environments. NY City Department of Health, Bureau of Environmental and Occupational Disease Prevention. 1999. Facts about Mold. NY City Department of Health, Bureau of Environmental Investigations. Norback, D., M. Torgen, and C. Edling. 1990. Volatile organic compounds, respirable dust, and personal factors related to prevalence and incidence of sick building syndrome in primary schools. British Journal of Industrial Medicine 47:733-741. Nunnally, J.C. 1978. Psychometric theory. In Statistical Methods for Health Care Research. B.H. Munro and E.B. Page. 1993. J.B Lippincott Company. Philadelphia, PA. Olcerst, R. 1994a. A technique to use data loggers to measure effective ventilation and air exchange rates by carbon dioxide tracer. American Industrial Hygiene Association Journal 55(9): 833-835. 19946. Quantification of an indoor population by means of air exchange rates and carbon dioxide gas concentrations. American Industrial Hygiene Association Journal 55(11): 1080-1083. Otten, J.A., P.R. Morley, HA. Burge, M A Chatigny, J.C. Feeley, and K. Peterson. 1986. Airborne viable microorganisms in office environments: Sampling protocol and analytical procedures. Applied Industrial Hygiene 1(1): 36-44. Pasanen, A-L., P. Kalliokoski, and M. Jantunen. 1994. Recent studies on fungal growth on building materials. In R.A. Samson, B. Flannigan, M.E. Flannigan, A.P. Verhoeff, O.C.G. Adan, and E.S. Hoekstra (editors). Health Implications of Fungi in Indoor Environments. Air Quality Monographs Volume 2. Elsevier, Amsterdam. Pasanen, A.-L., K. Yli-Pietila, P. Pasanen, P. Kalliokoski, and J. Tarhanen. 1999. Egosterol content in various fungal species and biocontaminated building materials. Applied and Environmental Microbiology 65(1): 138-142. Persily, A.K. 1993. Ventilation, carbon dioxide and ASHRAE Standard 62-1989. ASHRAE Journal 35: 40-44. Pita, E.G. 1989. Air conditioning principles and systems: an energy approach. Wiley, New York. pp. 171-206. 134 Pitt, J.I. 1979. The Genus Penicillium and its Teleomorphic States Eupenicillium and Talaromyces. Academic Press, Toronto. Pope, A . M . , R. Patterson, and H. Burge. 1993. Indoor Allergens: Assessing and Controlling Adverse Health Effects. National Academy Press, Washington, D.C. Pratt, D., M.F. Carvalheiro, J. Dutkiewicz, H. Goto, R. Jacobs, J. Liesivuori, E. Melbostad, A . Andersen-Rask, and D. Sekimpi. 1994. Endotoxin and ( l -»3)-P-D glucan. American Journal of Industrial Medicine 25: 139-140. Quackenboss, J.J., M.D. Lebowitz, and C D . Crutchfield. 1989. Indoor-outdoor relationships for particulate matter: exposure classification and health effects. Environment International 15: 353-360. Rand, T. 1998. Mold contamination in Atlantic Canada schools: burdens, amplifying sites and school inspection policy benefits. Conference Proceedings of the Third International Conference on Bioaerosols, Fungi and Mycotoxins: Health Effects, Assessment, Prevention and Control held September 23-25, 1998, Saratoga Spring, N Y . Raper, K . B . and D.I. Fennell. 1973. The Genus Aspergillus. Robert E. Krieger Publishing Company, Huntington, N Y . Raper, K . B . , C. Thorn, and D.I. Fennel. 1949. A Manual of the Penicillia. The Williams & Wilkins Company, Baltimore. Reinikainen, L . M . , J.J.K. Jaakkola, and O. Seppanen. 1992. The effect of air humidification on symptoms and perception of indoor air quality in office workers: a six-period cross-over trial. Archives of Environmental Health 47(1): 8-15. Reponen, T., A . Nevalainen, M . Jantunen, M . Pellikka, and P. Kalliokoski. 1992. Normal range criteria for indoor air bacteria and fungal spores in a subarctic climate. Indoor Air 2: 2 6 - 3 1 . Rippon, J.W. 1982. Medical Mycology. The Pathogenic Fungi and the Pathogenic Actinomycetes, 2 n d Edition. W.B. Saunders Co., Philadelphia. Rothman, A . L . and M.I. Weintraub. 1995. The sick building syndrome and mass hysteria. Neurologic Clinics 13(2): 405-412. Ryan, C M . and L A . Morrow. 1992. Dysfunctional buildings or dysfunctional people: an examination of the sick building syndrome and allied disorders. Journal of Consulting and Clinical Psychology 60(2): 220-224. Rylander, R. and B. Fogelmark. 1994. Inflammatory responses by inhalation of endotoxin and (l-»3)-P-D glucan. American Journal of Industrial Medicine 25: 101-102. 135 Rylander, R., V. Hsieh, and C. Courteheuse. 1994. The first case of sick building syndrome in Switzerland. Indoor Environment 3: 159-162. Rylander, R., H. Persson, H. Goto, K. Yuasa, and S. Tanaka. 1992. Airborne beta -l-»3-glucans may be related to symptoms of sick buildings. Indoor Environment 1: 263-267. Salisbury, SA. 1986. Measuring carbon dioxide levels as an indicator of poor building ventilation: a case study. Managing Indoor Air for Health and Energy Conservation. Proceedings of the American Society of Heating, Refrigeration and Air-Conditioning Engineers, Inc. Conference, pp. 78-82. Salkinoja-Salonen, M.S., MA. Andersson, R. Mikkola, A. Paananen, J. Peltola, H. Mussalo-Rauhamaa, R. Vuorio, N.-E. Saris, P. Grigorjev, J. Helin, and T. Timonen. 1998. Toxigenic microbes in indoor environment: Identification, structure and biological effects of the aerosolized toxins. Conference Proceedings of the Third International Conference on Bioaerosols, Fungi and Mycotoxins: Health Effects, Assessment, Prevention and Control held September 23-25, 1998, Saratoga Spring, NY. Salvaggio, J.E. 1997. Extrinsic allergic alveolitis (hypersensitivity pneumonitis): past, present and future. Clinical and Experimental Allergy 27(Suppl 1): 18-25. Samson, R.A., B. Flannigan, M.E. Flannigan, A.P. Verhoeff, O.C.G. Adan, and E.S. Hoekstra. 1994. Health Implications of Fungi in Indoor Environments. Elsevier, Amsterdam. Schoen, D. 1998. Dangers of the damp and dark. Canadian Medical Association Journal 159(2): 133. Seitz, T A . 1990. NIOSH indoor air quality investigations 1971 through 1988. In D.M. Weekes and R.B. Gammage (editors). Proceedings of the Indoor Air Quality International Symposium. American Industrial Hygiene Association, Akron OH. ppl63-171. Silver, W.L. 1992. Neural and pharmacological basis for nasal irritation. Annals of the New York Academy of Sciences 641: 152-163. Smith, E.G. 1990. Sampling and Identifying Allergenic Pollens and Molds. Blewstone Press, San Antonio. Smith, J.E. and M.O. Moss. 1985. Mycotoxins: Formulation, Analysis and Significance. Chapter 2. The Toxigenic Fungi. John Wiley & Sons, Toronto. Sneller, M.R. 1984. Mould allergy and climate conditions. In Al-Doory, Y. and J. F. Domson (editors). Mould Allergy. Lea & Febiger, Philadelphia. 136 Spendlove, J.C. and K.F. Fannin. 1983. Source, significance and control of indoor microbial aerosols: human health aspects. Public Health Reports 98(3): 229-244. Strieker, S., M. Bourgeau, E. Fonberg, and D. Parent. 1997. Physiological responses to elevated carbon dioxide levels in buildings. Indoor + Built Environment 6:301-308. Sudakin, D.L. 1998. Toxigenic fungi in a water-damaged building: An intervention study. American Journal of Industrial Medicine 34:183-190. Sudakin, D.L. 1999. Mycotoxins and pulmonary hemorrhage. Archives of Pediatrics and Adolescent Medicine 153(2): 205-206. Tamblyn, R.M., R.I. Menzies, R.T. Tamblyn, J.P. Farant, and J. Hanley. 1992. The feasibility of using a double blind experimental cross-over design to study interventions for sick building syndrome. Journal of Clinical Epidemiology 45(6): 603-612. Tanner, M. 1999. Study links basement moulds to health problems. The Vancouver Sun, Saturday, March 27, p. 115. Tansey, W.A., J.M. Wilson, and K.E. Schaefer. 1979. Analysis of health data from 10 years of Polaris submarine patrols. Undersea Biomedical Research, Submarine Supplement S217-S246. Teeuw, K.B., C.M.J.E. Vandenbroucke-Grauls, and J. Verhoef. 1994. Airborne Gram-negative bacteria and endotoxin in sick building syndrome. Archives of Internal Medicine 154: 2339-2345. Thomann, W.R. and J.J. Tulis. 1999. The relative contribution of the HVAC system to the total glass fiber and fungal spore levels in indoor spaces. In Proceedings of Indoor Air 99, The 8th International Conference on Indoor Air Quality and Climate. Edinburgh, Scotland. Construction Research Communications Ltd., London, UK. Vol. 4 pp 1029-1031. Thorn, A., M. Lewne, and L. Belin. 1996. Allergic alveolitis in a school environment. Scandinavian Journal of Work, Environment and Health 22(4): 311-314. Troller, J.A. and J.H.B. Christian. 1978. Water Activity and Food. Academic Press, New York. Troller, J.A. and V.N. Scott. 1992. Measurement of water activity (aw) and acidity. In C. Vanderzant and D.F. Spittstoesser (editors). Compendium of Methods for the Microbiological Examination of Foods, Third Edition. American Public Health Association. Turner, J. 1999. Black, slimy mould a danger to health. Canadian Press Newswire February 3. 137 Turner, W . A . , D . W . Bearg, and T. Brennan. 1995. Ventilation. Occupational medicine: State o f the Art Reviews 10(1): 41-57. Vancouver City Health Department. 1989 Bylaw 6580 section 6, smoking in public conveyances. 1996. By law 7560 section 6, smoking in restaurants. Vedal , S. 1995. Health effects of inhalable particles: implications for British Columbia. Report prepared for the A i r Resources Branch, British Columbia Ministry o f Environment, Lands and Parks. Vitanen, H A . 1995. Models of the critical time of humidity and temperature conditions for the development of mould fungi in pine and spruce sapwood. Paper presented to The International Research Group on Wood Preservation, 26 t h Annual Meeting. Wan, G . - H . and C.-S. L i . 1999. Dampness and airway inflammation and systemic symptoms in office building workers. Archives of Environmental Health 54(1): 58-63. Weltermann, B . M . , M . Hodgson, E . Storey, A . C . DeGraff Jr., A . Bracher, S. Groseclose, S.R. Cole, M . Cartter, and D . Phillips. 1998. Hypersensitivity pneumonitis: a sentinal event investigation in a wet building. American Journal o f Industrial Medicine 34(5): 499-505. Wheeler, A . 1993. A perspective of ventilation for health care and related facilities. In Proceedings of the Workshop on Engineering Controls for Preventing Airborne Infections in Workers in Health Care and Related Facilities. P .J. Bierbaum and M . Lippmann (editors). U . S . Department of Health and Human Services, Cincinnati, O H . Wilken-Jensen, K . and S. Gravesen (editors). 1984. Atlas of Moulds in Europe Causing Respiratory Allergy. Foundation for Allergy Research in Europe. A S K Publishing. Willers, S.and S. Andersson. 1996. Sick building syndrome symptoms among staff in schools and kindergartens: are the levels of volatile organic compounds and carbon dioxide responsible? Indoor + Built Environment 5(4): 232-235. Wilson, P. 1998. Toxic fungus found in school portables (Stachybotrys atra fungus). Medical Post 34(20): 31. Workers' Compensation Board ( W C B ) . 1998. Occupational Health and Safety Regulation: B C Regulation 296/97. British Columbia. World Health Organization (WHO). 1986. Indoor A i r Quality Research. E U R O Reports and Studies 1031-1064. 138 Yang, Y., C. Sun, and M. Sun. 1997. The effect of moderately increased COv concentration on perception of coherent motion. Aviation, Space and Environmental Medicine 68(3): 187-191. Yoshida, K., M. Ando, K. Ito, T. Sakata, K. Arima, S. Araki, and K. Uchida. 1990. Hypersensitivity pneumonitis of a mushroom worker due to Aspergillus glaucus. Archives of Environmental Health 45(4): 245-247. APPENDIX I. FIELD DATA RECORDING FORM Date: Log Sheet Weather: Temperature indoors: Barometric presure: _ Site identification: Principal: Temperature outdoors: %RH Absolute humidity Teacher: Site characterization: Age of building Construction: AHU/heating system: Location of classroom: Number of occupants in classroom Room volume: Windows: open closed avaerage age of children: open Orientation of windows: _ Floor covering: carpet Other openable but closed Drapes? age of carpet Ceiling: acouttical tile Interior cleanliness: 5 6 Plants/organic debris in classroom: Comments: other 10* Signs of moisture: Activity in room: sitting: moderate: active: empty: Activity as % of day: sitting: empty: Landscaping: CAV Shade level: .Minimal moderate: active: Outdoor organic debris: Low CLV Moderate Moderate _. UN* Marked* _ High* 141 Equilibrium relative humidity: Location #1: description: Time on: time of reading: %RH temp. Location #2: description: Time on: time of reading: %RH temp. • Bioaerosols: Bacteria: sample location: Indoor TSA thermophiles (ID): Outdoor TSA thermophiles (ID) _ CFU: CFU/m3: CFU: CFU/m3: Indoor TSA thermophiles (ID): Outdoor TSA thermophiles (ID) _ CFU: CFU/m3: CFU: CFU/m3: Indoor TSA mesophiles (ID): Outdoor TSA mesophiles (ID) CFU: CFU/m3: CFU: CFU/m3: Indoor TSA mesophiles (ID): Outdoor TSA mesophiles (ID) CFU: CFU/m3: CFU: CFU/m3: • Bioaerosols: Fungi: sample location: Indoor MEA thermophiles (ID): Outdoor MEA thermophiles (ID) CFU: CFU/m3: CFU: CFU/m3: Indoor MEA thermophiles (ID): Outdoor MEA thermophiles (ID) CFU: CFU/m3: CFU: CFU/m3: Indoor MEA mesophiles (ID): Outdoor MEA mesophiles (ID) _ CFU: CFU/m3: CFU: CFU/m3: Indoor MEA mesophiles (ID): Outdoor MEA mesophiles (ID) _ CFU: CFU/m3: CFU: CFU/m3: • Bioaerosols: Xerophilic Fungi: sample location: Indoor XMEA mesophiles (ID): Outdoor XMEA mesophiles (ID) CFU: CFU/m3: CFU: CFU/m3: Indoor XMEA mesophiles (ID): Outdoor XMEA mesophiles (ID) • CFU: CFU/m3: CFU: CFU/m3: 142 Open face cassette: Pump # Pump # Indoor filter number: Outdoor filter number: Flow on: Flow off: Flow on: Flow off: Time on: Time off: Time on: Time off: _ Elapsed time: Average flow: Elasped time: Average flow: Pre weight: ^ Post weight: Pre weight: Post weight: _ Net weight: Cone. Net weight: Cone. Duplicate: Open face cassette: Pump # ; Pump # Indoor filter number: Outdoor filter number: Flow on: Flow off: Flow on: Flow off: Time on: Time off: Time on: Time off: _ Elapsed time: Average flow: Elasped time: Average flow: Pre weight: Post weight: Pre weight: Post weight: _ Net weight: Cone. Net weight: Cone. CO2: Meter on: Minimum: Meter off: Maximum: Vent i la t ion: Supply Velocity dimension (area) Vent # Vent # Vent# flow Sketch of layout: Outdoor check: Occupied average: Return Velocity dimensions (area) flow Vent # Vent # Vent# 143 Ventilation: SF 6: quantity Notes: Use back of page to record Miran readings: Condition AHU/penthouse: Condition of ducting: Sketch layout of room: "Kozak, P.P. et al. (1985) Endogenous mold exposure: environmental risk to atopic and non-atopic patients. In Gammage et al. Indoor Air and Human Health, Lewis Publishers, NY. General cleanliness/maintenance: Level 5: very cluttered with minimal or no attempt at dust control Level 6: cluttered with obvious areas of increased dust, attempt at dust control infrequent and at irregular intercals, tolerance of some degree of debris accumulation Level 7: fair attempt at dust control but premises looks "lived in." Dust control is only partially implemented Level 8: cleaner premises with increased attempt at dust control but still having a "lived in" appearance Level 9: excellent cleaning but some areas of relaxed standards Level 10: extremely meticulous dust control Landscaping: CAV: cared for/average vegetation - few to average number of shrubs/trees, all of which are well pruned CLV: cared for/lush vegetation - somewhat more than average number of shrubs/trees with healthy appearance. UN: uncared for/natural - generally uncared for appearance; premises located in rural areas; only natural vegetation or previously landscaped but allowed to grow wild Shade levels: Minimal: trees and shrubs too small to produce shade Moderate: average amount of shade around building Marked: building rarely has direct light Level of organicdebris: Low: no significant dead vegetation within 100 ft. of building Moderate: some dropped leaves in the area or where ivy or some other material is used for ground cover High: large number of dropped leaves: grass not removed after cutting, presence of hay, compost or stands of dried weed or grass 144 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0076831/manifest

Comment

Related Items