Evaluation of the indoor air quality minimum ventilation rate procedure for use in California retail buildings



This research assesses benefits of adding to California Title-24 ventilation rate (VR) standards a performance-based option, similar to the American Society of Heating, Refrigerating, and Air Conditioning Engineers ‘Indoor Air Quality Procedure’ (IAQP) for retail spaces. Ventilation rates and concentrations of contaminants of concern (CoC) were measured in 13 stores. Mass balance models were used to estimate ‘IAQP-based’ VRs that would maintain concentrations of all CoCs below health- or odor-based reference concentration limits. An intervention study in a ‘big box’ store assessed how the current VR, the Title 24-prescribed VR, and the IAQP-based VR (0.24, 0.69, and 1.51 air changes per hour) influenced measured IAQ and perceived of IAQ. Neither current VRs nor Title 24-prescribed VRs would maintain all CoCs below reference limits in 12 of 13 stores. In the big box store, the IAQP-based VR kept all CoCs below limits. More than 80% of subjects reported acceptable air quality at all three VRs. In 11 of 13 buildings, saving energy through lower VRs while maintaining acceptable IAQ would require source reduction or gas-phase air cleaning for CoCs. In only one of the 13 retail stores surveyed, application of the IAQP would have allowed reduced VRs without additional contaminant-reduction strategies.

Practical implications

In most retail spaces, keeping concentrations of all volatile organic compounds (VOCs) within stringent health-based concentration limits would likely require substantially higher ventilation rates (VRs) than the current required minimum (this conclusion was driven by a small number of chemicals that exceeded limits). Saving energy in a retail space using an Indoor Air Quality Procedure (IAQP) with stringent concentration limits would generally also require source control or gas-phase air cleaning.


The goal of reducing energy use in commercial buildings while maintaining acceptable air quality is shared by building owners, operators, employers, and policy makers. Significant recent progress has been made in improving building energy performance through a combination of mandated building codes, technology advances, and recommendations by non-governmental organizations.

One issue that has received significant attention is the minimum requirement for ventilation (outdoor air supply) in commercial buildings because the heating and cooling needed to condition ventilation air makes up a small but significant portion of the total energy consumed by commercial buildings (Benne et al., 2009. However, ventilation is also essential for providing healthy, productive working environments for building occupants (Sundell et al., 2011; Seppänen and Fisk, 2004).

The retail industry employs a large number of workers and the stores are frequently visited by the general public. The 2007 Economic Census (Census, 2009) shows that in California alone, the ‘retail trade’ employed 1.7 million workers. The National Human Activity Pattern Survey (NHAPS) found that of the time people spent indoors outside of their home, 13% is spent in shopping mall and grocery stores (Klepeis et al., 2001). Retail buildings account for approximately 15% of commercial building energy use (Itron, 2006).

ANSI/ASHRAE Standard 62.1-2010 (ASHRAE, 2010) provides two alternate procedures for selecting minimum ventilation rates (VRs) for commercial buildings. In the ‘ventilation rate procedure’ (VRP), users adopt the minimum VRs listed in a table, regardless of the building's features. The prescribed minimum VRs were based historically on the rates needed to maintain occupant satisfaction with odors but recently have, in limited ways, considered indoor emissions from both occupants and the building itself and its furnishings. Current prescribed rates reflect a sum of two quantities: the minimum rate of outdoor air supply per unit floor area and the minimum rate of outdoor air supply per occupant. The VRP provides definitive guidance on minimum VRs, for a variety of building or space types, that can be used to specify heating, ventilation, and air conditioning (HVAC) system sizing during the design phase of a new building.

Standard 62.1 also includes an alternative (and rarely used) ‘Indoor Air Quality Procedure’ (IAQP) that contains both objective and subjective components and is intended to provide flexibility in the means of achieving IAQ; and it has been previously used with the objective of saving energy (Grimsrud et al., 1999; Bridges et al., 2013). In general for buildings with economizers lower minimum VRs reduce energy use (Benne et al., 2009). However, in some climates, and in some buildings without economizers, increased ventilation can save energy. The Bridges study used the IAQP to specify minimum VR in three stores that were below rates specified by the ASHRAE VRP. In contrast to the VRP, the IAQP is a performance-based approach that does not prescribe specific VRs by building use. The IAQP allows flexibility in the means used to achieve adequate IAQ levels, including outdoor air ventilation, source control, air cleaning, and other strategies. A comprehensive IAQP protocol (including both objective and subjective assessments of IAQ) is performed in stages, with the final stages occurring after the building is constructed and occupied. The first step in the IAQP is to specify a set of contaminants of concern (CoCs) and, based on guidelines from relevant, or ‘cognizant’ authorities, concentration limits not to be exceeded. Users of the IAQP are free to select which contaminants are considered and which guidelines should be used to determine maximum contaminant concentrations, subject to approval by the building code official.

To satisfy the objective component of the IAQP, indoor emission rates of all CoCs are calculated based on estimated building materials and contents as well as outdoor contaminant concentrations. An overall ventilation and design strategy must then be identified that will maintain indoor concentrations of all CoCs below the specified limits. VRs lower than those specified in the VR procedure are allowed as long as the IAQP users can demonstrate that CoC concentrations are below selected reference concentration limits (RCLs). Once the strategy is applied, and the building is constructed and occupied, a subjective test of the perceived air quality is performed to demonstrate that visitors and/or occupants are ‘satisfied’ with the air quality. The IAQP does not prescribe the procedure for assessing satisfaction with air quality or the level of satisfaction that must be achieved; however, it does describe how satisfaction may be determined. Subjective assessments of IAQ are normally based on survey responses, collected either from occupants after a period of time in the building (adapted responses) or from visitors immediately after they enter the building (unadapted responses).

The IAQP is designed to allow the minimum VR that, in combination with other measures, will verifiably maintain acceptable IAQ Although the IAQP allows VRs to be reduced relative to those required by the VR procedure, application of the IAQP in some circumstances might require higher VRs. The IAQP has been used to a limited extent; at least one large retail organization uses the IAQP to specify the minimum ventilation requirements for its stores throughout the United States (Grimsrud et al., 1999). In this case, applied IAQP-based VRs are significantly lower than the alternative prescribed rates. It is unclear whether building owners or designers would ever voluntarily design or operate a building with alternate IAQP-based VRs that exceeded the minimum prescribed VRs, thus achieving improved IAQ at a greater energy cost.

In California, there is ongoing consideration of the merit of incorporating an IAQP-like procedure into the state's Title-24 building efficiency standards (California Energy Commission, 2008). This issue is also of interest outside of California due to the broad-based goal of balancing building energy and indoor air quality concerns. A recent report details some limitations of the current IAQP specified in ASHRAE 62.1-2010 (Apte et al., 2013). The report also describes some limitations associated with standards that assume that prescribing fixed minimum VRs will ensure adequate IAQ even though the standards do not consider building features such as the use of air cleaning equipment or the strength of indoor pollutant sources.

This study has three main goals: (i) assessing, in a set of California retail stores, the adequacy of the VRs currently prescribed by Title-24 in providing the IAQ level specified by IAQP process; (ii) determination of the VRs needed to satisfy the objective component of the IAQP; and (iii) evaluating whether several VRs implemented experimentally in a ‘big box’ store, including the current VR as measured before the intervention, the Title 24-prescribed VR, and a calculated IAQP-based VR, would achieve adequate IAQ, assessed both objectively and subjectively. Two types of data were collected to evaluate the IAQP in California retail buildings; observational data from stores functioning as usual, and data from an intervention study in one big box store. Through the process of applying the IAQP, the research team developed specific recommendations for potential future California ventilation standards based on IAQ. The data from this study can inform discussions about the possibility of adding an IAQP-like option to Title-24.

Study methods

For the first two goals described above, CoCs and RCLs were selected for use in evaluating IAQ. VRs and indoor and outdoor CoC concentrations were measured in 13 stores, including three grocery, five furniture, and four apparel one big box store. Mass balance models were employed to calculate indoor emission rates for CoCs in each store. Mass balance models were then used, with these emission rates and typical outdoor air contaminant concentrations, to calculate, for each store, the ‘IAQP’ VR that would maintain indoor CoC concentrations below selected RCLs. If ventilation was not able to maintain concentrations of a CoC below the concentration limit, for example, because of high outdoor air concentrations, that CoC was neglected in the calculation of IAQP VRs. These IAQP VRs were compared with observed VRs and to the Title 24-prescribed VRs.

For the second goal described above, a VR intervention study was performed in the sole big box store among the group of stores assessed, to determine how objectively assessed indoor contaminant levels (from measured air concentrations) and subjectively assessed IAQ (from subject surveys) varied with VR. The three VRs in the intervention study were an approximation of the store's current VR (0.24 h−1), the Title 24-prescribed VR (0.69 h−1), and the calculated IAQP-based VR (1.51 h−1).

First, we determined the CoCs to be used to evaluate IAQ, based on research that has identified indoor air contaminants known to impact health. The CoCs were based on VOCs and aldehydes previously identified in commercial-building indoor environments by Parthasarathy et al. (2011). A more extensive analysis of the data in this work was performed in the LBNL report (Dutton et al., 2013), that included particles, ozone, and carbon monoxide. In this report, we assessed the effect of changes in VRs, increased particle filtration, and source control measures for each of the 13 stores. Indoor particle concentrations in the three grocery stores originated substantially from indoor sources, and in the other ten stores, outdoor air was the dominant source of particles. The analysis found that even in grocery stores that had higher indoor than outdoor particle concentrations, increased VRs were ineffective for controlling indoor particles. For buildings with significant indoor sources of particles in locations where outdoor particle concentrations are consistently low relative to our reference locations, it may be appropriate to consider using ventilation to control indoor particles IAQP calculations. However, in the buildings surveyed in this study the preferred methods of controlling indoor particle concentrations were enhanced particle filtration or indoor particle source controls. Additionally, the study identified only outdoor sources of ozone, and carbon monoxide in most commercial buildings, so these contaminants were not considered to be drivers of the IAQP-based minimum VRs.

Reference concentration limits were identified for each CoC, where available, from lists of RELs from California's Office of Environmental Health and Hazard Assessment (OEHHA, 2008). Where no OEHHA RELs existed, concentration limits from alternative relevant authorities were referenced, including the U.S. Environmental Protection Agency (U.S. EPA, 2014), the National Institute for Occupational Safety and Health (NIOSH, 2014), and the Agency for Toxic Substances and Disease Registry (ATSDR, 2013). In addition to health effect thresholds, odor thresholds were considered in determining concentration limits because these are relevant to perceived IAQ (Fanger, 1988). Thresholds for odor and pungency of VOCs were obtained from Cain and Schmidt (2009), Hodgson and Levin (2003), and Nagata (1993). The lowest thresholds found among these studies, from Parthasarathy et al. (2012), were used to screen our CoCs.

Observational field study methods

Ventilation rates and indoor and outdoor CoC concentrations were measured simultaneously in 13 California retail stores, including one big box store. Full details of methods and data collected are described in Chan et al. (2012), with estimates of measurement uncertainty given in our results section. For both the intervention and observational studies, particle counts and inorganic contaminants such as ozone and carbon monoxide were measured continuously using real-time instruments. Time-integrated concentrations of VOCs, aldehydes, and particle mass were measured. Contaminant concentrations were measured in one to four indoor locations. Outdoor concentrations were measured using a set of identical instruments placed either on the building rooftop or at ground level outside the store. The outside air VR was measured using the SF6 decay method.


Multisorbent tubes containing Carbopack B and Carbopack X were used to capture VOCs with a range of different vapor pressures. Volatile carbonyl samples were collected using dinitrophenyl hydrazine (DNPH)-coated cartridges (Waters Sep-Pak®, Milford, MA, USA). Ambient ozone was removed with potassium iodide scrubbers prior to each DNPH sample. One-hour samples were collected at each of the 1–4 indoor and single outdoor locations at 1 Lpm, using a novel low-cost sampling system developed specifically for the study. Multiple samples were collected immediately following a change in the study VR and then periodically throughout the remainder of each study. Four sorbent tube samples (two DNPH and two Carbopak) were collected simultaneously to provide duplicates of each sample type. Thermal desorption-gas chromatography/mass spectrometry was used to quantitatively analyze the VOC samples by following U.S. EPA Methods TO-1 and TO-17 (U.S. EPA, 1984, 1999). Multipoint internal standard calibrations were performed using pure compounds and 1-bromo-3-fluorobenzene as the reference compound. A duplicate set of samples was collected at all times. DNPH cartridges were extracted with 2-ml aliquots of acetonitrile, and the extracts were analyzed by high-performance liquid chromatography (HPLC) with ultraviolet (UV) detection at λmax = 360 nanometers (nm) (Agilent 1200).

Analysis method

As noted above, the CoCs used for our assessment included VOCs and aldehydes. For each CoC in each store, a simple steady-state mass balance model was used to calculate indoor emission rates based on measured indoor and outdoor contaminant concentrations and measured VRs. Where there were multiple sampling periods, emission rates were calculated for each period. Appendix S1 describes the mass balance model used to calculate VOC emission rates.

We used calculated indoor CoC emission rates and typical outdoor air contaminant concentrations for each store in a mass balance model to calculate an ‘IAQP’ VR that would maintain indoor concentrations of particle and VOC CoCs below selected concentration limits.

Average measured concentrations of indoor VOCs and aldehydes were compared with the relevant concentration limits. For each store, we identified a subset of these contaminants whose indoor concentrations approached or exceeded the limits. The minimum VRs necessary to maintain concentrations below limits were then calculated for these particular store-relevant VOC CoCs. IAQP-based VRs were compared with the observed VRs and the Title-24 prescribed VRs. A detailed IAQP calculation method for VOCs can be found in Appendix S1. As an alternative to increased ventilation, we calculated the percentage reduction in indoor contaminant emission rates required to reduce indoor concentrations below the concentration limits, assuming prescribed Title-24 VRs. The source control calculation method is detailed in Appendix S1.

Intervention field study methods

A VR intervention study was performed in the big box retail store that was part of this research. This single-story big box store is located in northern California and has a sales floor area of 10 000 m2, consisting of one large open room. The ceiling height is 4.2 m, above which is a 3-m-high plenum. A very broad range of products is available in the store, including clothing, house-wares, groceries, toys, and sporting goods. Within the main retail area are two potential sources of combustion-generated pollutants: a small fast food outlet and a chain retail coffee outlet. Thermal conditioning is provided by five conventional variable air volume rooftop units with economizers and an additional fan-powered exhaust system. An experimental evaporative cooling unit also provides additional ventilation air.

Three VRs were studied: a low rate approximating the store's current VR (0.24 h−1), a medium rate based on the Title-24 prescribed VR (0.69 h−1), and a high rate based on the calculated IAQP-based VR (1.51 h−1). The store's mechanical VRs were fixed at these rates for 1-week study periods each allowing for small variations due to infiltration. During each study period, IAQ was assessed objectively by measuring indoor contaminant concentrations and subjectively by surveys regarding perceived air quality and acute health symptoms.

Pollutant measurements in the intervention study

For each 1-week study period, concentrations of a range of indoor contaminants, including VOCs, particles, and inorganic gases, were monitored at up to four locations within the store and one location on the roof. Indoor sampling stations were located approximately in the center of each of the store's four quadrants, at heights between 1.4 and 1.8 m, subject to availability of shelf space and electrical power.

Ventilation rate measurements

Prior to the intervention study in the big box store, we established an empirical relationship between mechanical damper setting in the rooftop units and measured VRs. This relationship enabled us to quickly select the damper settings and fan speeds to produce the desired three VRs at the beginning of each week of the subjective response study. The infiltration rate was not assessed, however, care was taken to ensure that the configuration settings resulted in a slight positive pressurization of the store, reducing unintentional infiltration. The electricity use from all additional fans including kitchen and bathroom exhaust was monitored by the store's building manager and these data were shared with us. To achieve our objective of attaining our desired whole-building VRs, it was not necessary or desirable to disable or alter the flow rates of these smaller constant rate fans.

Whole-building VRs that include both mechanical ventilation and limited infiltration were measured using a tracer gas decay procedure. We did not measure mechanical air intake flow rates. Small volumes of sulfur hexafluoride (SF6) tracer gas were injected directly into the outside air inlets of all six of the store's main air-handling units, producing indoor concentrations of approximately 1.5 parts per million (ppm). Simultaneous injections took place over a period of several minutes through delivery tubes connected to valves calibrated to produce approximately homogeneous concentrations throughout the store. Through normal operation of the air-handling units, fresh outside air (that did not include tracer gas) gradually replaced the store air that contained the tracer gas, causing the concentration of tracer gas to decay. Tracer gas concentrations were measured for approximately 1 h to coincide with the VOC sampling. Tracer concentrations were measured using Miran SapphIRe® Model 250B (Thermo Fisher Scientific Inc., Waltham, MA, USA) infrared gas analyzers calibrated with primary standard calibration gases at four indoor locations. We performed a calibration study of the Miran SapphIRe®, by comparing measured decay rates under controlled conditions with decay rate measurements based on bag samples collected at 5-min intervals. Bag samples were analyzed by gas chromatograph equipped with an electron capture detector which was calibrated using tracer gas standards with an estimated measurement uncertainty of approximately 2%. We found that the Miran instruments responded linearly over a range of SF6 concentrations of 100–1500 parts per billion (ppb) with an average error of 5%, resulting in an overall estimate of uncertainty in the tracer gas measurement of approximately 5%. VRs were calculated using the age-of-air method (ASHRAE, 1997) with tracer concentration decay data from within the 300–1500 ppb range. The average VR for each sample collection period was given by the reciprocal of the average of the age of air at the four stations. The age-of-air method was considered the most appropriate because it does not require homogeneous distribution of tracer throughout the sampling period. However, in addition, as a secondary verification of the results, the air change rate was also computed using curve fitting to the exponential decay of SF6 concentrations using standard method ASTM E 741 (2006).

Subjective measurements of air quality

Subjectively perceived IAQ was assessed in the big box store using three independent groups of untrained human subjects. For these assessments, store VRs were held constant to the degree possible during each of 3-study weeks. VRs were held constant by deactivating outdoor air economizer controls, maintaining constant fan speeds, and maintaining constant damper settings for outdoor, recirculated, and exhaust airflows for a several-day period.

Once the VR had been held constant for 3–4 days prior to the day of the subjective study to achieve approximately steady-state indoor contaminant concentrations, subjects took a series of surveys of perceived IAQ and symptoms. Prior work by Dalton (2000) found that ‘repeated or prolonged exposure to an odorant typically leads to stimulus-specific decreases in olfactory sensitivity to that odorant, but sensitivity recovers over time in the absence of further exposure.’ Therefore, subjects took surveys both immediately upon entering the building and at 12-min intervals at five additional marked locations within the store. Figure B1 in Appendix S1 shows the map provided to subjects with an example of an assigned route connecting the survey station locations (identified by the label #station number). Questions on all six paper surveys were identical and assessed olfactory responses and short-term health-related symptoms. All subjects took the initial survey (#1) at the same location immediately after entering the store, to assess perceived air quality and symptoms before their olfactory responses became adapted to the store environment. Subjects then performed a pre-determined sequence of four surveys, one in each of the store's four quadrants (#2–5). Each of these four surveys was preceded by a 12-min period of adaptation to the local indoor air, achieved by walking around within the pre-defined quadrant. The sequence of quadrants in which subjects took these four surveys was ordered to reduce any bias associated with local variations in the store environment. Subjects were randomly assigned to proceed through the quadrants following one of six possible orders. The last survey inside the store (#5) was taken by all subjects near the main entrance, and the final survey (#6) was performed at the survey station near the main entrance where the first survey (#1) had been completed. The final survey was given after subjects exited the store for 3 min to refresh their olfactory systems so that they were no longer in a state of adaptation to IAQ conditions. A total of 129 subjects returned completed surveys: 44 for the first week, 41 for the second, and 44 for the final study week.

Appendix S1, Figure B2 shows an example page of the survey that subjects completed at station one; the subsequent five surveys that subjects completed were identical. In addition to questions about air quality, subjects answered questions about their age, gender, employment status, income range, and marital status. Health factors and shopping habits were also assessed: whether the subject smoked, had ever been diagnosed with asthma, or had any of a list of common allergies. A detailed breakdown of age, demographic, and health factors can be found in Appendix S1. A human subjects protocol for this study was developed, reviewed, and approved by Lawrence Berkeley National Laboratory's Committee for the Protection of Human Subjects.

In response to the survey questions on olfactory responses and subjective symptoms, subjects were free to use a pen to make a mark corresponding to their perception of air quality anywhere along a (horizontal or vertical) linear scale. Figure 1 shows an example. Survey responses were translated manually to electronic data by measuring the distance of each response from the left side of the scale (or the bottom of the scale for vertically oriented scales). To check the quality of survey grading, 20% of the surveys were measured and digitized a second time by a different person and the results compared; no significant differences in scale readings were found.

Figure 1.

Survey response example

In response to the question ‘How do you rate the air quality?’ subjects were free to mark anywhere along either (but not both) of two scales that ranged from ‘Clearly unacceptable’ to ‘Just unacceptable’ or from ‘Just acceptable’ to ‘Clearly acceptable.’ Subjects’ responses were translated into a numerical scale ranging from −1 to 0 representing unacceptable responses and 0–1 for acceptable responses. (Note: there were no responses at 0 on either the acceptable or unacceptable scale). In addition to this continuous scale, responses were also simplified into a binary scale of ‘unacceptable’ or ‘acceptable,’ depending on which of the two scales subjects had marked. For the purpose of determining the relationship between VR's and IAQ, we will consider the primary results to be the continuous scale of IAQ acceptability, because it uses the full detail in the responses, rather than the binary scale.

As noted above, surveys one and six were administered to subjects immediately after entering the store from outside, to obtain unadapted responses. Surveys two, three, four, and five, taken after subjects had been in the store for at least 12 min, provided adapted responses. Analyses were performed separately for adapted and unadapted responses.

The responses to the air quality surveys were plotted against several demographic, health, and environmental variables. Temperature and humidity were considered potential confounding variables in the relationship of VR to air quality and symptoms; however, the influences of temperature and humidity could not be separated from the influence of ventilation because only a single VR measurement was obtained for each of the three study VRs.

Survey responses were used to determine whether the changes in VRs caused statistically significant changes in perceived air quality and short-term health symptoms. Paired t-tests, assuming equal variance of response, were used to explore variables that might confound the relationship between VR and acceptability of air quality. These variables included gender, asthma status, temperature, humidity, and allergy status. We also employed a non-parametric Wilcoxon signed-rank test that does not require normality, equal variance for comparison. A key objective of this analysis was to determine whether our air quality satisfaction target was achieved (as required by the IAQP) at each VR level. Based on prior precedent, an 80% acceptability rate was used as the minimum requirement for perceived air quality. The 80% rate of acceptability has historically been used to establish minimum VR requirements in standards.


Measurements from 13 stores

Table 1 lists the store dimensions and measured air change rates in the 13 buildings. The table includes the average, range of air change rates measured within each store and the percentage spacial variation in tracer starting concentration. We used subjective judgment to determine when the tracer had become sufficiently well mixed after being injected, and we calculated the percentage variation in tracer concentration by starting with the maximum variation (maximum – minimum concentration), and dividing by the average initial store concentration.

Table 1. Description of 13 retail study buildings (includes three ventilation conditions for store BB)
Retail storeDescription/California LocationSales floor area (m2)Indoor store height (m)Measured air change rates (1/h), decayMeasured air change rates (1/h), age of airTracer gas% rangeApprox. distance to nearest highway (km)
  1. aStore was surveyed while mechanical ventilation was in operation; F01-F03, F05, A01, A02 had natural ventilation only.

  2. Stand alone stores are denoted by b, remaining stores were attached to an adjacent store.

G01a,bGrocery, Berkeley327080.68 (0.61–0.79)0.75 (0.71–0.79)70.31
G02a,bGrocery, Walnut Creek183971.88 (1.70–2.06)1.45 (1.30–1.59)160.19
G03a,bGrocery, Tarzana3307100.70 (0.66–0.73)0.77 (0.72–0.82)70.06
F01Furniture, San Francisco64141.14 (0.98–1.32)1.12 (0.94–1.35)390.31
F02bFurniture, Oakland153362.84 (2.63–3.03)2.38 (2.17–2.78)470.31
F03bFurniture, Berkeley67840.35 (0.20–0.61)0.39 (0.33–0.51)230.94
F04a,bHardware, Fremont511080.64 (0.59–0.68)0.68 (0.63–0.75)140.63
F05Hardware, Long Beach79042.28 (2.21–2.35)1.16 (0.86–1.25)120.63
A01Apparel, Oakland12033.86 (3.42–4.29)2.33 (2.27–2.38)30.06
A02Apparel, Oakland8033.04 (NA)2.22 (NA)NA0.06
A03aApparel, San Mateo107090.41 (0.39–0.44)0.52 (0.51–0.53)210.31
A04aApparel, Long Beach99550.46 (0.45–0.47)0.54 (0.51–0.57)60.06
BB Baselinea,bBB retail, Davis10 21940.44 (0.36, 0.53)0.43 (0.35, 0.52)130.06
BB Lowa,b0.25 (0.22, 0.26)0.24 (0.21, 0.26)44
BB Med.a0.67 (0.42, 1.26)0.69 (0.41, 1.20)27
BB Higha1.50 (1.28, 1.79)1.51 (1.23, 1.78)16

Excluding the big box store, variation in initial tracer concentration was on average 11% for the mechanically ventilated and 27% for the naturally ventilated stores. In the big box store, baseline, medium and high VR scenarios, the store air appeared to be initially well mixed. During the medium VR scenario, significant spatial variations in VR were observed. Post hoc analysis of the fan electricity data revealed that one of the six remote terminal unit exhaust fans that had reactivated in error. However, the overall VR for the space met our target rate based on the Title 24-prescribed minimum VR.

There are multiple sources of uncertainly in the measurement of air change rates. We estimate the uncertainties in the instrument measurement of tracer concentrations to be 5%. The calculation of air change rates from the tracer concentration data is a further source of uncertainty, factors that contribute to this overall uncertainty include, temporal variations in airflow rates, observed differences in initial tracer concentrations within the buildings, and inherent uncertainty in the VR calculation methods. These factors varied significantly between the stores, in general, we expect that the largest source of uncertainty comes from the spatial variation in initial tracer gas concentration and the potential that measurement locations do not lead to an overall average air change rate representative of the indoor volume. The range of local air change rates, each based on the reciprocal of a local age of air, is an indicator of measurement uncertainty resulting from imperfect mixing, and these ranges are given in Table 1. The true averaged air change rate is expected to fall within the range of measurements of local air change rates.

Comparing predicted air change rates using the age-of-air method and decay methods found an average difference of 16% for the 13 stores. Prior work by Sandberg (1985) indicates that in cases when these two methods provide very similar results uncertainty in the overall measurement is likely to be lower.

Volatile organic compounds and aldehydes – observed indoor concentrations vs. reference concentration limits

Table 2 lists the measured concentration ranges of several CoCs measured in the 13 study buildings, compared with appropriate concentration limits. Note that two different RCLs were employed for formaldehyde – the 9 μg m3 chronic REL from OEHHA that is exceeded in a substantial fraction of buildings and sometimes also exceeded in outdoor air, and the higher 19.6 μg m3 chronic RCL from NIOSH. Where available, reference outdoor concentrations were based on the 90th percentile of the last 4 years of outdoor measured state annual average (by county) concentration data (2007–2011) (CARB, 2012). Alternatively, where no statewide reference data were identified, reported outdoor concentration data were used from a recent sampling study conducted in California (Bennett et al., 2011). For the IAQP calculations, reference outdoor data were considered preferable to using the measured outdoor data collected coincidently with the measured indoor data, because in most cases the reference outdoor concentrations exceeded our coincident-measured outdoor concentrations, resulting in higher (more conservative) IAQP-based VRs.

Table 2. Summary of indoor concentrations of key contaminants of concern and reference levels
 Chronic Guidelines (μg/m3)Indoor concentrations (μg/m3) (average [Min–Max])Outdoor ref. concentration (μg/m3)
  1. aCARB, 2012; bBennett et al., 2011, Table 45.

Formaldehyde9 (OEHHA), 19.6 (NIOSH)19.4 [4.7–58.1]5.8a
Acetaldehyde9 (EPA)13.9 [2.7–34.8]3.3a
Octanal2.1 (Odor)2.3 [0.2–11.9]0.81b
Acrolein0.02 (EPA)8.4 [0.5–43.9]2.9a
Hexanal32.4 (Odor)14.7 [0.7–58.4]1.11b

Table D1 through Table D3 in Section D of the Appendix S1 show the IAQP-based minimum VRs for the top four (VOC or aldehyde) contaminants in each store (excluding the big box store) that had concentrations exceeding, or closest to, the applicable RCLs. The tables also show, for the top four CoCs, the percentage reduction in indoor sources of contaminants that would be necessary to limit indoor concentrations to the RCLs if Title-24 VRs were applied in each store. Table D4 shows the same information for the big box store for the top three contaminants exceeding or closest to the applicable RCLs. In the big box store, only the top three most significant contaminants were considered because for the majority of contaminants’ indoor concentrations were significantly below any available RCL. For the intervention study, we calculated the IAQP-based VR using outdoor contaminant concentration data measured during the baseline characterization. Table D5 gives IAQP calculation results using measured outdoor concentrations. Figure 2 summarizes the VRs measured, the VR as calculated using Title-24 prescribed minimum rates, and the IAQP-based VRs for each of the 13 stores. The IAQP-based VR was the largest of the minimum VR based on the calculations for each CoC.

Figure 2.

Ventilation rates for 13 retail stores

For the three grocery stores, acetaldehyde and octanal defined the IAQP-based VRs. IAQP VRs for the three grocery stores, G01, G02, and G03, were 2.9, 6, and 3 ACH, respectively. For G01 and G02, these rates are significantly higher than the current VRs. In G02, the IAQP-based rates are also significantly higher than the Title-24 prescribed minimum rates. At all three grocery stores high concentrations of acrolein, above RCL, were observed both in and outdoors. Within the context of the IAQP, where outdoor concentrations exceeded RCL's, ventilation could not be used to define the IAQP-based minimum VR.

In all five furniture stores, formaldehyde was the dominant CoC in the objective IAQP calculation. When the OEHHA REL was used, the minimum IAQP-based VRs all exceeded both measured VRs and Title-24 rates. If the less stringent NIOSH chronic RCL standards were applied for formaldehyde, IAQP-based rates would still exceed Title-24 rates in the majority of stores.

Formaldehyde was again the dominant driver of the IAQP-based VRs in the apparel stores, with IAQP VRs exceeding current VRs or Title-24 VRs in stores A01 and A02. Referring to Table D3 in the Appendix S1, the formaldehyde emission rates for A02 of 407 μg/h m2, were approximately 3 times larger than the next highest emission rate, as a result of this high emission rate the IAQP-based VR for A02 was significantly higher than any of the other stores. By contrast, in store A03, IAQP-based VRs were significantly lower than rates prescribed by Title-24, indicating that the store may be overventilated.

The IAQP calculation for the big box retail store identified formaldehyde as the principal driver of the IAQP VR; all other measured contaminants were below reference levels at the surveyed baseline VR. The IAQP VR exceeded the existing VR but was slightly below the Title-24 VR.

Intervention study

For the intervention study, indoor concentrations of CoCs were measured throughout each of the three periods when the VR was altered. Indoor concentrations of both PM2.5 and PM10 at all VR levels (Dutton et al., 2013) were found to be significantly lower than the maximum California annual standards. As reported in Dutton et al. (2013) increased ventilation was found to have no significant impact on indoor particle concentrations. Results from the VOC IAQP VR calculations, which were performed using the big box store observational field study data, showed that the concentrations of formaldehyde, acetaldehyde, and octanal were closest to chronic RCLs. Formaldehyde was the only COC that exceeded the chronic RCL. Results of the IAQP calculation in Table D5 predicted that formaldehyde would be the most significant driver of an IAQP-based VR and that a VR of 1.3 h−1 would be sufficient to lower indoor concentrations of formaldehyde to meet the most stringent reference guideline, OEHHA's REL. The actual measured VR during the study week with the ‘high’ IAQP-based VR was found to be 1.5 h−1 (see Table 1). In the hours following the changes in VRs usually on the Monday of the study week, indoor concentrations of several of the contaminants including formaldehyde differed from concentrations predicted using models that assume an emission rate that is independent of VR. We theorized that this effect was the result of absorption and de-absorption of certain VOCs on indoor surfaces significantly. By the Thursday of each week when the subjective study was performed, steady-state contaminant concentrations were consistent with our simple mass balance models. The results of steady-state indoor concentrations of formaldehyde, acetaldehyde, and octanal at the three intervention VRs are given in Figures 3, 4 and 5, along with corresponding RCLs. These results demonstrate that the application of the IAQP-based VR successfully achieved its objective of controlling indoor concentrations of all CoCs including formaldehyde, to at or below RCLs.

Figure 3.

Formaldehyde concentrations in ppb

Figure 4.

Acetaldehyde concentrations in ppb

Figure 5.

Octanal concentrations in ppb

Survey results

Survey participant demographics

The majority (62%) of survey respondents were between the ages of 18 and 22, 98% identified themselves as non-smokers, 34% had previously been diagnosed with allergies, and 16% had previously been diagnosed with asthma. The male-to-female ratio was significantly different from our preferred 50–50 balance. A full breakdown of age, demographic, and health factors (including gender, employment status, asthma, and allergy status) is in Appendix S1, Figure C1 and gives the gender ratios of the survey respondents for each study VR.

Subject response analysis

Figure B1 in Appendix S1 shows a map of locations. Surveys completed at stations 2, 3, 4, and 5 were all considered adapted responses; surveys collected at stations 1 and 6 were unadapted responses. Supplemental Table C4 summarizes the binary response regarding air quality acceptability for each study VR, weighted by gender ratio, and presented separately for initial unadapted, average unadapted, and average adapted responses. The average binary responses for unadapted subjects were 90.3%, 95.5%, and 98.9% acceptability at low, medium, and high VRs, respectively, suggesting increasing acceptability of air quality correlated with increasing VRs. In all cases, the targeted minimum acceptability rating of 80% was exceeded; that is, more than 80% of both males and females rated the IAQ as acceptable. Supplemental Table C5 shows mean reported values of acceptability for continuous response on a scale from −1.0 to 1.0. Comparing the standard deviation results across the study VRs indicates that the responses had approximately equal variance. Unadapted subjects’ continuous responses do not show as consistent a positive trend toward acceptability with increasing VR as is seen in the binary responses. Adapted subjects showed no evident trend using either the binary or continuous scales of IAQ satisfaction. For unadapted subjects, a t-test from a linear regression that accounted for repeated measures indicated a positive association between acceptability and VR, with a slope coefficient of 0.074 (95% confidence (−0.019, 0.167), P = 0.119). However, based on the criteria of P < 0.05, this association was statistically non-significant. During the three periods of survey collection, average temperature ranged between 20 and 23°C, and average relative humidity ranged between 22 and 41. Figure 14 of the report by Dutton et al. (2013) shows the approximately normal distribution of the subjects continuous scale responses, at the three study VRs.

Based on the results of our paired t-test and non-parametric Wilcoxon signed-rank tests of the continuous responses, there was no statistically significant difference between the mean air quality reported for adapted responses of men and women (P = 0.24). There was no statistically significant difference between mean air quality reported by asthmatics and non-asthmatics, either for adapted responses (P = 0.45) or for unadapted responses (P = 0.71). There was no statistically significant difference between mean air quality reported by allergic and non-allergic respondents, either for adapted responses (P = 0.065) or for unadapted responses (P = 0.41). No significant correlation was found between the subjects’ responses with either temperature or relative humidity, for either adapted, or unadapted subjects.


Key findings

Existing volatile organic compounds

Measured indoor VOC and aldehyde concentrations exceeded RCLs for at least one contaminant in all stores. Based on modeling, at Title-24 VRs, only two of the stores (A03 and the BB store) would have maintained all indoor CoC concentrations below specified limits. Thus, for the majority of buildings, neither the surveyed VRs nor prescribed Title-24 VRs were sufficient to protect occupants from CoC exposures exceeding concentration limits.

Indoor Air Quality Procedure calculations

Using the OEHHA REL, formaldehyde was the dominant driver of the IAQP-based VR in all stores with the exception of the three grocery stores. However, there are significant differences in published formaldehyde RCLs, ranging from the stringent 9 μg/m3 (OEHHA) to 98 μg/m3 (World Health Organization, 2010). When we used RCLs based on NIOSH exposure limits, formaldehyde was still the most significant driver in the majority of stores; however, when we applied the World Health Organization reference exposure limit for formaldehyde, then acetaldehyde and octanal become the dominant drivers of IAQP VRs.

Indoor Air Quality Procedure-based VRs exceeded Title 24-prescribed VRs in all stores except A03 and the big box store. In 62% of the stores, the IAQP-based VRs exceeded 3 ACH, and in 54% the rates exceeded 5 ACH. In these cases, using ventilation alone to manage indoor contaminants would likely be prohibitively expensive because of both the resulting increase in ongoing energy use and the possible increase in ventilation system size required to provide these higher VRs. Under these circumstances, the alternative strategies to consider are source control or application of air cleaning systems for VOCs. However, source control is complex because of the large number and changing nature of sources, and it is not clear that air cleaning technologies for VOCs are sufficiently effective and affordable for widespread use in buildings (Fisk, 2007).

Intervention study

The intervention study in the big box store found that, for the CoC that were closest to chronic RCLs (formaldehyde, acetaldehyde, and octanal), increased ventilation was effective at lowering steady-state indoor concentrations. At the IAQP-based VR, indoor concentrations of all CoCs were maintained below concentration limits. At all VRs in the big box store, subjective satisfaction with air quality exceeded the 80% target for both men and women. There was a significant increase in perceived air quality with increasing VRs as subjects first entered the store. For adapted responses, there was no clear relationship between perceived air quality and VRs. Comparing mean responses from unadapted (0.49) and adapted responses (0.45) showed that perceived air quality decreased marginally as subjects remained within the store for more than a few minutes and also as subjects moved deeper into the store. Subjective responses to perceived indoor air quality are affected by both odor and irritation (Cometto-Mu and Cain, 1995), and irritation caused by exposure to airborne contaminants has been shown to increase with increased exposure (Shusterman et al., 2006). Odor is also thought to play a larger role in subjective IAQ satisfaction than mucosal irritation (Cometto-Mu and Cain, 1995). Therefore, one explanation for our results is that the subjects’ odor perception fatigued over time, but exposure-related irritation increased, with the overall effect being a marginal decrease in perceived IAQ. An alternative interpretation is that local variations in odor counter a general adaption to the store environment. Either way, this result has potentially broader implications for ventilation standards that rely on perceived IAQ and use adapted subjects.

Study limitations

This study alone is insufficient as a basis for general recommendations regarding minimum commercial VRs in retail stores. Our sample of 13 retail stores included a range of common retail stores; our results are likely less applicable to stores with very different outdoor environments or indoor sources. However, the results of this study, together with results from Chan et al. (2012) and a related study evaluating chronic health effects as a function of VRs (Parthasarathy et al. 2012), provide a picture of the minimum VRs required to limit contaminant concentrations to acceptable levels in a majority of retail stores.

Our models of objective IAQ exclude several factors including gas-particle phase changes or gas-phase reactions. Inclusion of these factors was considered unlikely to significantly affect the results of our analysis. The mass balance models used to estimate objective IAQ included several significant assumptions that include a well mixed space and negligible, deposition, filtration, or penetration losses more details are available in Appendix S1. Although these uncertainties are thought to be significant, they are not expected to systemically bias the emission rate estimates or the IAQP-based VRs that are based on those emission rates.

This study provides limited data on the subjective portion of the IAQP for setting minimum VRs because, as a result of cost constraints, we performed perceived IAQ surveys for our IAQP assessment in only a single big box retail store.

For several reasons outlined in Dutton et al., (2013), the survey of perceived air quality did not include potentially sensitive populations such as children, the elderly, and people self-identifying as particularly sensitive to airborne contaminants. Our study plan called for each week's subjects to include 50% women and 50% men. However, the limited number of available subjects meant we were unable to balance gender during each week.

Implications of applying an Indoor Air Quality Procedure

The key change in approach when using the IAQP rather than the VRP is to consider outdoor air ventilation as just one of several possible tools for achieving adequate IAQ. Considering a wider range of tools would be an important step toward reducing energy use in buildings while maintaining or improving IAQ. In theory, this is a win–win strategy. In practice, however, this study shows that applying the IAQP with stringent concentration limits can increase energy use unless effective alternatives to increased ventilation are also considered. In buildings with already weak indoor contaminant sources, application of the IAQP to lower VRs has immediate potential to save energy without requiring additional control methods.

Improved specifications for an Indoor Air Quality Procedure

Currently, users of the ASHRAE IAQP have complete flexibility to select ‘critical contaminants’ and RCLs even though many users will not have the expertise necessary to select the contaminants most relevant to occupants’ health. Moreover, in the current framework, IAQP users can make selections to provide the answer they desire rather than to most accurately reflect the conditions in their buildings. It is therefore recommended that future versions of an IAQP, including any version developed for inclusion in Title-24, include lists of critical contaminants to be considered and appropriate contaminant limits to be applied.

Risks of Indoor Air Quality Procedures and ventilation rate procedures

There are inherent risks in broad adoption of the IAQP. If IAQP users select source control measures or air cleaning procedures that do not work as well as expected, at first and over time, IAQ may be degraded. In addition, if new sources of contaminants are introduced into a building, previously established source control or air cleaning technologies may no longer be sufficient. However, the current VR procedure does not assure that good IAQ is maintained either because the current procedure places no constraints on indoor pollutant emission rates and does not require use of high-efficiency particle filters. Development of additional forms of VR standards that combine elements of current VR and IAQ procedures in practical ways might allow improvements over both and also allow for energy to be saved without degrading IAQ.


  • In a sample of 13 retail stores, measured VRs generally exceeded the minimum requirements specified in California's Title-24 Standards; however, in a majority of stores, concentrations of selected VOCs exceeded the most stringent contaminant limits.
  • When ASHRAE's IAQP was applied in 13 stores, the minimum VRs needed in 11 stores to maintain concentrations of CoCs below stringent limits were higher, often substantially, than the minimum VRs specified in Title-24. Thus, application of the IAQP would only enable reduced VRs and associated energy savings when source control or gas-phase air cleaning of indoor contaminants was implemented.
  • When IAQP was applied in 10 of 13 stores, formaldehyde control drove the IAQP VR, based on California's stringent RCL of 9 μg m3. In the remaining three stores, the IAQP VRs were dictated by acetaldehyde or octanal. Even when using the higher NIOSH formaldehyde RCL, 19.6 μg m3, formaldehyde remained the driver for the IAQP VR in many buildings.
  • In our intervention study of a big box store, more than 80% of simulated shoppers were satisfied with IAQ at all three VRs studied, including one VR below the minimum VR specified in Title-24.


The research reported here was supported by the California Energy Commission Public Interest Energy Research Program, Energy-Related Environmental Research Program, award number 500-09-049 under contract DE-AC03-05CH11231 between the U.S. Department of Energy and the University of California. The authors thank: Marla Mueller for program management; Sebastian Cohn for assistance with measurements and sample analysis; Marion Russell for assistance with sample analysis; staff at the big box building for their assistance with the intervention study; and members of the project advisory committee for their reviews of a draft of this document.