Indoor Air Quality
NATIONAL CENTER FOR HEALTH STATISTICS
RESEARCH TRIANGLE PARK, DURHAM COUNTY, NORTH CAROLINA
The Agency for Toxic Substances and Disease Registry (ATSDR) reviewed the results of air sampling collected at the offices of the National Center for Health Statistics (NCHS), located 4105 Hopson Park, Research Triangle Park, North Carolina. The Centers for Disease Control and Prevention's Office of Health and Safety requested this review to determine whether levels of volatile organic compounds (VOCs) represented a health hazard to building occupants. The National Center for Health Statistics is the sole tenant of a one-story, 55,000-square foot structure. Approximately 110 employees work there. In the approximately 15 months the National Center for Health Statistics has occupied this building, employees have complained about building odors, describing them as "sweet" and like "Play Doh," a children's modeling clay substance.
Sixteen roof-top package heating, ventilating, and air conditioning (HVAC) units and one centrally located energy recovery unit (heat exchanger) serve the building's HVAC requirements. Central Carolina Air Conditioning, Inc. indicates that the systems are designed to bring 4,000 cubic feet per minute (CFM) of outdoor make-up air into the building.
Several activities addressed indoor air quality concerns at this facility, including:
- CDC's Office of Health and Safety (OHS) conducted an indoor air quality investigation of the facility on April 11-13, 2001. OHS measured relative humidity, temperature, and carbon dioxide over a 48-hour period. Two air samples were collected for formaldehyde and for standard VOCs. Formaldehyde levels were measured at 0.1 parts per million (ppm) and 0.09 ppm. One VOC sample (collected in Room 1236) contained 0.079 ppm of acetone, 0.022 ppm of isopropyl alcohol, and 0.009 ppm of toluene. The other VOC sample (collected in Room 1082) contained 0.064 ppm of acetone. All other VOCs were less than 0.005 ppm (the detection limits of the method). These levels are also below ATSDR screening values and the Occupational Safety and Health Administration's (OSHA) Permissible Exposure Limits (PELs).
- Carbon Dioxide levels were less than 700 ppm and within guidance values established by the American Society of Heating, Refrigeration, and Air Conditioning Engineers (ASHRAE). Temperatures ranged from 71º to 76º F and humidity 49 to 57%. These values are within acceptable ranges as recommended by ASHRAE.
- In July 2002 the Office of Health and Safety collected eight samples for formaldehyde in the building. The levels of formaldehyde ranged from 0.06 to 0.1 parts per million (ppm).
- In August of 2001 the HVAC contractor installed 11 additional return air grilles in the hallways to improve air distribution within the return air plenum of the building.
- AGI Environmental Inc, the building owner's consultant, conducted an air quality survey on September 23, 2001. The consultant measured relative humidity, temperature, carbon dioxide, and carbon monoxide. The relative humidity ranged from 53 to 69%. Carbon monoxide was not detected; but the detection limit was not specified. Carbon dioxide levels were less than 500 ppm. Five samples were collected for formaldehyde. The formaldehyde levels ranged from 0.03 to 0.06 ppm.
- On December 18, 2001, and February 5, 2002, a NCHS consultant, EI Inc., collected air samples for volatile organic compounds using modified Environmental Protection Agency (EPA) Method 1P-1B. The samples were collected using multi-sorbent collection tubes. They were thermally desorbed and analyzed by gas chromatography and mass spectrometry. On January 14, 2002, a carpet sample was collected for a 24-hour VOC carpet emission test. The air sample was collected directly above the area where, on January 14, 2002, the carpet was removed.
The Office of Health and Safety concluded that the levels of parameters it tested were within acceptable ranges for an occupational environment. On April 24, 2001, the Office of Health and Safety recommended that the building owner operate the HVAC system 24 hours per day, 7 days per week for a 60-day period to accelerate offgassing of VOCs and to reduce odor complaints.
EI Inc. also measured flow rates of selected supply registers and noted some modest disparities between design and observed air flow rates within the occupied space.
Ten air samples were collected. The levels of total VOCs ranged from 2,654 microgram per cubic meter (µg/m3) to 5,148 µg/m3. The total VOC carpet emissions (from office 1240) were 3,149 micrograms per square meter per hour (µg/m2/hr).
In the December 2001 survey, EI. Inc. also collected samples for formaldehyde using EPA Method TO-15 VOCs-fungi and bacteria. The limited microbial sampling suggested that the building did not contribute to microorganism growth. Formaldehyde samples results ranged from 0.02 to 0.05 ppm. The TO-15 samples measured 31 and 28 ppb of isopropyl alcohol and 44 and 49 ppb of acetone.
Dr. Douglass Campbell, EI Inc. physician, interviewed several building occupants who were concerned about the building's air quality. Reported symptoms included headache, eye irritation, nose irritation, hives, itching skin, and scratchy dry throat.
The levels of formaldehyde were well below the Occupational Safety and Health Administration's (OSHA) Permissible Exposure Limit (PEL) of 0.75 ppm, based on an 8-hour time weighted average. Nevertheless, use of occupational exposure limits for comparison values in the indoor environments might not be protective of building occupants. Office workers may be more sensitive to environmental contaminants than industrial workers (AIHA 1993).
ATSDR has established a Minimal Risk Level for formaldehyde of 0.04 ppm for acute exposure (1 to 14 days of continuous exposure) and 0.03 ppm for intermediate exposure (up to 365 days of continuous exposure) (ATSDR 1999). MRLs are considered safe levels of daily human exposure to a hazardous substance over a specified duration of exposure. MRLs are set below levels that, based on current information, might cause adverse health effects in most people. But MRLs are based on noncancer health effects only. Also, because they are based on continuous exposure, MRLs were not developed as occupational exposure limits.
The acute MRL is based on a lowest known adverse effect level of (LOAEL) of 0.4 ppm. Study participants experienced eye and nasal irritation following exposure to 0.4 ppm of formaldehyde for 2 hours. This MRL includes an uncertainty factor of 9 to account for human variability and the use of a LOAEL.
The levels of formaldehyde appear to have decreased during the past 10 months from 0.1 ppm to less than 0.05 ppm. ATSDR does not expect that adverse health effects will occur when formaldehyde levels are less than 0.05 ppm in an office setting. The levels of formaldehyde, in combination with other VOCs, could have contributed to overall building odor and perceived poor air quality in the building.
VOCs are the primary source of most indoor odors. The levels (3,000 to 5,000 µg/m3) measured by using the multi-sorbent tubes suggest that VOCs could be a contributing factor to the odor and health complaints because these levels are 10-fold greater that levels typically observed in office buildings. The emission rate for the carpet sample is approximately 6-fold greater than the rate for so-called "low emission" carpets (Carpet and Rug Institute 2002). In addition to the carpet system (e.g., carpet, backing, adhesives), furniture, partitions, copiers and chemical cleaners can also contribute to the total VOC loading and odor.
In addition to VOCs, the following building-related factors are known to cause eye irritation: ozone, relative humidity less than 25% elevated carbon dioxide levels, tobacco smoke, dust, and fibers (Mølhave 1991). No smoking is allowed in the building. Also, ozone and particulate were not measured during the air quality surveys.
There are no health-based screening values for total VOCs or for majority of the individual VOCs detected in the multi-sorbent samples. Total VOCs are a complex mixture of specific compounds. Approximately 300 individual VOCs have been identified in indoor air (Mølhave 1991). Limited toxicological data are available on many of these compounds. No clear and consistent dose response relationship has been observed between levels of total VOCs and health outcomes. In a blind study, Mølhave exposed 150 persons who had a history of building discomfort to a mixture of 22 VOCs at levels of 0 µg/m3, 5,000 µg/m3, and 25,000 µg/m3. The participants' perception of poor air quality and odor intensity increased with total VOC levels. Subjects also reported perceived mucous membrane irritation at 5,000 µg/m3 and 25,000 µg/m3.
The relationship between odor and health is not fully understood. Schiffman and Shusterman describe three paradigms of how odors can produce health symptoms (Schiffman et al. 2000). The first is that symptoms are induced by exposure to odors at levels that also cause irritation or toxicological effects. Irritation rather that odor causes symptoms. In the second paradigm, exposure to odorous compounds produces health-related symptoms at levels of odor thresholds but below irritant levels. This typically occurs with exposure to certain types of sulfur-containing compounds or organic amines. These compounds have odor thresholds well below the levels that cause irritation. The mechanism by which health complaints result from these types of exposures is not well understood, but it could include psychological and genetic factors. The third paradigm associates an odor with a symptom where the odorant is part of a mixture containing a co-pollutant (e.g. dust or allergen) responsible for the health symptom. The odor acts as a marker for exposure.
Wolkoff (1999) suggests that odor thresholds for VOCs can be a sound endpoint for assessing comfort and health. Knudsen (1999) examined the relationship between odor intensity, i.e., perceived air quality, and VOC emissions of building products. Using an environmental chamber to measure VOC emission rates and sensory (odor) panels, researchers determined that some building products that are sensitive to oxidative degradation (e.g., carpets) result in secondary VOC emissions. Secondary VOC emissions result from chemical or physical processes (e.g., oxidation, thermal degradation, hydrolysis of building products). The odor index (the VOC level divided by its odor threshold) of primary VOCs decayed faster than the corresponding odor intensities. This suggests that secondary emissions rather than primary emissions are likely to affect perceived air quality over the long term. Some building products can affect perceived air quality despite selected VOC levels that are less than 10% of their odor threshold (odor index) (Knudsen et al. 1999). There is evidence that sensory irritation can occur from exposure to sub-threshold levels of individual VOCs combined in a mixture (Cometto-Muñiz et al. 1999). Wolkoff suggests that extremely low odor thresholds of unsaturated VOCslike unsaturated aldehydescould account for many odor concerns in buildings.
The total VOC level in most offices is less than 200 µg/m3, and individual VOC compounds do not typically exceed 50 µg/m3 (Morey et al. 2000). Building occupant complaints are commonly encountered when total VOC levels exceeds 3,000 µg/m3 (Morey et al. 2000). Still, exposure to individual VOCs with extremely low odor thresholds can elicit complaints well below the 3,000 µg/m3 level (Morey et al. 2000). Mixtures can show additive or hyper-additive effects at levels below individual odor thresholds. Therefore the use of guidance values based solely on total VOCs should be viewed with caution.
In summary, insufficient scientific information exists with which to assess the health impact of total VOCs on building occupants.
The National Institute for Occupational Safety and Health (NIOSH) has observed that HVAC systems are associated with up to 60% of indoor air quality concerns (Morey et al. 2000). ASHRAE recommends 20 cubic feet per minute (CFM) of outdoor (fresh air) per person. This standard was intended to achieve acceptable air control on the basis of controlling bioeffluents from building occupants rather than controlling building material-related odors. Because the source of carbon dioxide (C02) is human respiration, the low C02 readings observed during these surveys might not necessarily indicate adequate supply of outdoor air or good mixing, especially if the offices were sparsely occupied at the time of the measurements.
The levels of formaldehyde in the NCHS office are well below occupational exposure limits. The levels of formaldehyde have decreased to 0.05 ppm or less over the past 10 months. ATSDR does not anticipate that these current levels of formaldehyde will result in adverse health effects.
Levels of total VOCs are elevated above those typically found in office buildings. Sufficient scientific information does not exist for ATSDR to make a health determination about the levels of total VOCs measured at the NCHS office. Limited emission testing and air sampling suggests that the carpet system could be the principal source of the VOCs and odor.
While ATSDR cannot make a health determination, prudent public health practice calls for implementing odor control and reduction efforts, and maintaining an adequate supply of outdoor air to the occupied work spaces.
Specific recommendations include:
- Conduct additional emission testing and VOCs sampling, if necessary, to better characterize the source of the VOCs,
- Identify, evaluate and control any potential contributors to secondary VOC emissions, such as excessive moisture or ozone from office equipment,
- Conduct periodic air sampling for formaldehyde to ensure that the levels remain below 0.05 ppm, and
- Perform periodic inspections and balancing of the HVAC system to ensure adequate distribution of outdoor air to the occupied spaces.
Following the receipt of additional data, the Office of Health and Safety should reevaluate these recommendations. These actions should have the beneficial effects of improving perceived air quality, employee productivity, and decreasing the amount of employee sick time.
CDC Building Management should consider adding so-called "low emissions" specifications to its building specification requirements for new office space.
Peter J. Kowalski, MPH, CIH, CSP
Environmental Health Scientist
Agency for Toxic Substances and Disease Registry
Division of Health Assessment and Consultation
Exposure Investigation and Consultation Branch
Chief, Consultations Section
Agency for Toxic Substances and Disease Registry
Division of Health Assessment and Consultation
Exposure Investigation and Consultation Branch
- Agency for Toxic Substances and Disease Registry. 1999. Toxicological profile for formaldehyde. Atlanta: US Department of Health and Human Services.
- American Industrial Hygiene Association. 1993. The industrial hygienist's guide to indoor air quality investigations. Fairfax VA: American Industrial Hygiene Association.
- Carpet and Rug Institute. 2002. Available at: URL: http://www.carpet-rug.com/drill_down_2.cfm?page=8&sub=6
- Cometto-Muñiz JE, Cain WS, Abraham MH, Gola JMR. 1999. Chemosensory detectability of 1-butanol and 2-heptanone singly and in binary mixtures. Physiol Behav 67(2):269-76.
- Knudsen. HN, Kjaer UD, Nielson PA, Wolkoff P. 1999. Sensory and chemical characterization of VOC emissions from building products: impact of concentration and air velocity. Atmos Environ 33:1217-30.
- Mølhave L. 1991. Volatile organic compounds, indoor air quality and health. Indoor Air 4:357-376.
- Morey PR, Horner E, Epstein BL, Worthan AG, Black MS. 2000. Indoor air quality in non-industrial occupational environments. In Harris RL, editor. Patty's industrial hygiene. Volume 4, 5th edition. New York: John Wiley and Sons. p. 3149-3221.
- Schiffman SS, Walker JM, Dalton P, Lorig TS, Raymer JH, Shusterman D. 2000. Potential health effects of odor from animal operations, waste water treatment, and recycling of byproducts. J Agromed 7(1):7-81.
- Wolkoff P. 1999. How to measure and evaluate volatile organic compound emissions from building products-a perspective. Sci Total Environ 227 (2-3):197-213.