Printer Friendly

Ventilation rate investigations in Minnesota bars and restaurants.


A statistically representative sample of hospitality venues in the Minneapolis/St. Paul metropolitan area (19 drinking places, 9 limited-service restaurants, and 37 full-service restaurants) was chosen to assess patron and staff exposure to second-hand smoke (SHS). Each venue received three repeat visits on each of three different times of day and four different day types (Monday-Thursday, Friday, Saturday, and Sunday). Three to five of the venue visits were conducted for two hours, most often on Friday and Saturday nights. The remaining visits were conducted for at least 15 minutes. All visits included laser photometer fine particulate ([PM.sub.2.5]) concentration measurements, carbon dioxide ([CO.sub.2]) measurements, lit cigarette counts, and customer counts. The two-hour measurements also included measurements of ultraviolet absorbing particle and gas phase component concentrations. The rigorous design of this study provides improved confidence in the findings and reduces the likelihood of systematic bias.

Concentrations of SHS compounds in the air of restaurants and bars depend on both the strength of the source (i.e., the number of cigarettes smoked) and removal processes. While surface sorption and filtration affect removal, the dominant removal process is the venue's mechanical ventilation system(s). Field measurement of outdoor air ventilation rates in occupied buildings is difficult, particularly when the measurements are conducted without the occupants' knowledge. In this study, ventilation rate determinations were made using average [CO.sub.2] concentration measurements and occupancy counts. Those values were then corrected for changes in [CO.sub.2] observed at the beginning and end of the measurement interval.

Tracer gases have a long history of use in determining ventilation rates and buildings (Dick 1950, Dietz and Cote 1982, Sherman 1990, Persily 1997). The choice of an appropriate tracer gas is often difficult because of the need for human safety, ease of measurement, and reasonable cost. This paper describes the use of [CO.sub.2] as a tracer gas to estimate ventilation rates unobtrusively in a random selection of bars and restaurants before the implementation of a statewide smoking ban in the premises.


Ventilation was determined using measurements by field staff visiting the premises. Continuous real-time measurements of [CO.sub.2] concentration were made using a model 8762 IAQ-Calc TM manufactured by TSI, Inc. The device utilizes a dual-wavelength non-dispersive infrared sensor system with a resolution of 1 ppm and measurement range of 0 to 5,000 ppm. It has a temperature coefficient of 0.36% per [degrees]C and at 25[degrees]C is specified to be accurate to [+or-]3% of the reading or 50 ppm, whichever is larger. The IAQ-Calcs were set to log continuously with an averaging period of 1 minute. Prior to each weekend's monitoring, the two IAQ-Calcs were calibrated using the unit's built-in zero/span field calibration procedure. Nitrogen zero gas and a calibration gas with a [CO.sub.2] concentration of 2,516 ppm were used for the calibrations.

All visits were conducted unannounced without the knowledge of the venue staff or customers. This was expected to provide less biased results since the venue would not make any operational changes for the monitoring period and, since owner approval was not required, it allowed for a completely random sample of venues. A single person conducted the 15-minute or short-term visits, and the instruments were carried in either a briefcase or backpack. Two people conducted the two-hour visits to make it easier to bring in both the monitoring briefcase and catalogue case and remain in the venue for the two hours without drawing attention to themselves. The cases were typically kept on a chair near the monitor(s) with the inlet tube located at about table height. The "hum" from the cases was drowned out by the noise in the venues, and venue staff often did not ask about the equipment. The monitors were instructed to sit in the smoking section, in the area with the greatest concentration of people and well away from outside doors or doors to the kitchen. For only a few visits (typically nightclubs), the monitors were not allowed to keep the cases with them. When that occurred the visits were repeated at a later date.

In addition to the equipment measurements, each monitor used a Palm[R] handheld device with a customized database application to record observational data, including the time entering and leaving the venue; the start and end of the outdoor monitoring period; the number of staff, customers, and lit cigarettes; and the occupants' perception of smoke odor, cooking odor, wood odor, and stuffiness. For the two-hour visits the customer and lit cigarette counts were conducted every 15 to 20 minutes.

A number of factors affect the indoor [CO.sub.2] concentrations within an operating bar or restaurant (Waring and Siegel, 2007). The dominant source of [CO. sub.2] production is occupant respiration. In addition, [CO. sub.2] is produced by burning tobacco products and from unvented cooking combustion products. Finally, [CO. sub.2] is brought in and removed via the venue's ventilation system or air leakage through the building envelope. These factors can be related through the following mass balance equation:

V x dC / dt=Q [C.sub.0] - QC + G

Here, V is the volume of the venue, Q is the ventilation rate, [C.sub.o] is the outdoor concentration of [CO. sub.2], C is the indoor concentration of [CO. sub.2], and G is the net invenue [CO. sub.2] generation rate.

The rate of change of the indoor concentration can be ignored when the number of occupants and ventilation rate are relatively steady, the ventilation rate is high, and/or the concentration is measured over a longer period of time.

Q = G / C- [C.sub.0] (2)

Instead of assuming a constant [CO.sub.2] concentration, the rate of change of indoor [CO.sub.2] concentration was approximated by the difference between the [CO.sub.2] concentrations at the beginning and end of each monitoring period over the length of each monitoring period:

dC / dt [congruent to] [C.sub.end] - [C.sub.beg] / [t.sub.visit] (3)

Combining Equations 1 and 3 gives an expression for the ventilation rate, Q, using a simplified version of the mass balance equation for the concentration of carbon dioxide that was measured on the premises during the visit.

Q [congruent to] G / C-[C.sub.0] - V / [t.sub.visit] [[C.sub.end] - [C.sub.beg] / C - [C.sub.0]] (4)


Q = ventilation rate ([m.sup.3]/s, cfm)

G = generation rate of [CO.sub.2] from people in the space ([m.sup.3]/s, cfm).

C = indoor [CO.sub.2] concentration ([m.sup.3]/ [m.sup.3], f[t.sup.3]/f[t.sup.3])

[C.sub.o] = outdoor [CO.sub.2] concentration ([m.sup.3]/ [m.sup.3], f[t.sup.3]/f[t.sup.3])

V = volume of the venue ([m.sup.3], f[t.sup.3])

[t.sub.visi] = time difference between [C.sub.beg] and [C.sub.end] of the visit (s)

[C.sub.end] = [CO.sub.2] concentration at the end of the visit ([m.sup.3]/ [m.sup.3], f[t.sup.3]/f[t.sup.3]), and

[C.sub.beg] = [CO.sub.2] concentration at the beginning of the visit ([m.sup.3]/ [m.sup.3], f[t.sup.3]/f[t.sup.3]).

Appendix C of ASHRAE 62-1 (ASHRAE 2007) specifies a [CO.sub.2] generation rate of 0.0043 L/s (0.0091 cfm) for a met level (M) of one where M is the met rating associated with that person's activity level. For each monitoring period, M was approximated using occupant activity data collected by the monitoring teams. The customers were divided into three categories of activity level, and the workers were divided into two categories (see Table 1).
Table 1.   Specified Met Level by Customer and Worker Type

Type                              Description           Met Level

Customer, seated          Seated at tables or bar             1.2
Customer, light activity  Playing video games or pool         2.0
Customer, heavy activity  Dancing or similar                  3.0
Worker, seated            Seated bouncer/lottery sales        1.2
Worker, typical           Server, bus tables, host            2.0

Note that in making the calculations of the ventilation rates the [CO.sub.2] contributions from lit cigarettes and unvented cooking combustion devices were deemed insignificant compared with the [CO.sub.2] contributed by the building occupants (Waring and Siegel, 2007).


Ventilation Rates

Ventilation rates were computed for 183 two-hour monitoring visits with valid indoor and outdoor [CO.sub.2] measurements (i.e., approximately three each two-hour visits per venue). While there was sufficient data to compute the ventilation rates for the short-term visits, only the rates from the two-hour visits were computed because the occupancy information was more reliable (i.e., repeated customer counts), and the longer duration two-hour data allowed for an adjustment due to the change in [CO.sub.2] concentration from start to the end of the monitoring period. The ventilation rates were reported as a volumetric rate and in air changes per hour (ach), which is the ventilation rate divided by the venue volume.


For the entire sample of 65 venues the minimum ventilation rate measured was 0.44 ach and the maximum 12.9 ach. The values at the 25th percentile, median, and 75th percentiles were 1.8, 2.9, and 4.6 ach respectively. The median ventilation rates were 2.9, 3.0, and 2.5 ach for the drinking places, full-service restaurants, and limited-service restaurants, respectively. The median for the entire sample was 2.9 ach.

The ventilation rates were compared to those recommended in ASHRAE Standard 62.1 in two different ways:

* the ASHRAE Standard 62.1 rates required for the site using assumptions about the occupancy that would be present at such a site (floor-area-based default occupancy) and

* the ventilation rate required by Standard 62.1 for the actual occupancy observed at the site.

For this normalization, we found that the 25th percentile, median, and 75th percentile of the distributions of the measured ventilation rates divided by the design rates based on default design occupancies for these venues were 0.34, 0.54, and 0.65. For the three types of venues, the comparison shows that for 75% or more of the visits the ventilation rates present were less than those recommended by ASHRAE using their default assumptions about occupancy at the site. On the other hand, the rates were typically larger than the rates recommended for the actual occupancies observed by the field measurement teams. For the entire sample, the distribution of measured ventilation rates normalized by the values that the standard requires for the occupancies observed was 1.07, 1.46, and 2.23 for the 25th, median, and 75th percentiles, respectively.



The uncertainties associated with the computed ventilation rate were calculated using the partial derivative method with the flow rate uncertainty determined by the square root of the sum of the squared terms (Beckwith et al. 1982).


The manufacturer specified accuracy of the TSI IAQ-Calc used for [CO.sub.2] measurements is the greater of [+ or -]50 ppm or 3% of the reading. Since all measurements at the venue are made with the same instrument and instrument drift is slow compared to the sampling period, the uncertainty in the concentration difference (e.g., indoors--outdoors and end--beginning) is smaller than if the two measurements were made with different instruments. The differences between calibration gas measurements shortly after completion of a "field calibration" and the actual calibration gas concentration were used to develop the concentration difference uncertainty specified in Table 2.
Table 2. Uncertainty Estimates

Occupant CO2 Production                              [+ or -] 15%

Venue Volume                                         [+ or -] 10%

CO2 Concentration Difference  Larger of [+ or -] 4% of reading or
                                                           15 ppm

Time of Visit                                      [+ or -] 1 min

Air Mixing                                            [+ or -] 6%

In addition to instrument [CO.sub.2] measurement error, there is uncertainty in the concentration difference due to incomplete mixing of [CO.sub.2] within the venue and the use of a single location measurement to represent the average concentration in the venue. Sources of [CO.sub.2] (primarily the occupants in the room) are typically distributed throughout the room. Air is a low-viscosity fluid and mixes well, particularly if forced-air heating and cooling systems are used to distribute the air. However, this lack of distributed sampling points is estimated to add approximately 6% to the uncertainty of the concentration difference measurement. This 6% uncertainty is added in quadrature to the instrument measurement error and was determined from a comparison of simultaneous smoking and nonsmoking section [CO.sub.2] measurements for two-hour monitoring visits to 11 venues.

Uncertainties in the computed ventilation rates are calculated from measurement uncertainties collected in Table 2.

Combining terms in the computation for ventilation rates gives uncertainties in the computed ventilation rates ranging between [+ or -]15% and [+ or -]30% for 90 percent of the measurements.

The quality of the ventilation estimate using the steady-state assumption (Equation 2) is improved by correcting for the change in [CO.sub.2] concentrations at the beginning and end of the monitoring (Equation 4). However, the change for using the non-equilibrium correction term is small. Figure 3 presents a comparison of the steady-state calculation of the ventilation rates and those calculated including the non-equilibrium equation. For ventilation rates less than approximately 3 ach the non-equilibrium correction tends to lower the calculated value, and for those larger than 3 ach the correction is about evenly positive and negative.


Thus at higher ventilation rates the estimate associated with the steady-state assumption of Equation 2 is adequate. However, in situations with ventilation rates less than about 3 ach the simple steady-state expression of Equation 2 will over-estimate the actual ventilation rate.

The relative uncertainties of the calculated ventilation rates are presented in Figure 4. The diamonds in the figure show the uncertainty for the steady-state calculation, and the squares represent the uncertainty for the non-equilibrium calculation (Equation 4). For most of the calculated values the uncertainty is dominated by the assumed 15% uncertainty of the [CO.sub.2] respiration generation rate. The relative uncertainty for the non-equilibrium calculations is generally higher, with about 73% of the values having an uncertainty less than 20% compared to 90% for the steady-state method. However, the improved accuracy of the calculated value for the non-equilibrium method is expected to offset the higher uncertainty.



A summary of the ventilation results and a comparison of the results with the values required for the occupancies specified by Standard 62.1 are important for those considering building changes to improve energy efficiency and indoor air quality in these building types. Field measurement of outdoor air ventilation rates in occupied buildings is difficult, particularly when the measurements are conducted without the occupants' knowledge. In this study, ventilation rate measurements were made from [CO.sub.2] concentration measurements and occupancy counts.

Ventilation systems are designed to provide specific amounts of ventilation air to building occupants to remove pollutants generated by occupant activities and other sources within the building. These values are set by building codes and standards (cf. for example ASHRAE Standard 62.1) and are typically verified before initial occupancy of the building. Unless systems are well maintained and adjusted regularly, the amount of ventilation provided may decrease over time. In addition, the level of ventilation may be reduced by owners attempting to reduce energy costs if they perceive that code-required levels of ventilation do not provide a recognizable benefit to their business. Similarly, maintenance and rebalancing systems are expenses that owners may forego.

Results from the measurements in this study support these concerns. The median ventilation rate for all venues was just 50% of the design ventilation that should be provided to these sites (assuming a default occupancy value). This result was consistent across venue types observed in the study. To be fair, however, venue ventilation rates determined from indoor [CO.sub.2], outdoor [CO.sub.2], and occupancy show that 80% of venues were ventilated at values above ASHRAE recommendations for the observed occupancy levels (i.e., the observed occupancies were less than the default values typically used in the buildings' designs). Therefore, elevated concentrations of measured SHS constituents are unlikely to be caused by low ventilation rates and should be attributed instead to SHS sources present in the venues. For the simpler control systems most often used in hospitality venues, it is unlikely that building operators employed systems that actually adjusted ventilation rates for the occupancy present during these measurements.


Measurements showed that actual ventilation rates were less than 50% of the design ventilation rates for the venues based on default occupant density; however, 80% were above the ventilation rate required for the actual occupancies observed in the spaces. These results support the observation that pollutant sources must be controlled to reduce exposure in these venues--smoking exposure is a source problem, not a ventilation problem.


This research project was funded in part by ClearWay Minnesota through Grant Number RC 2006-0050. The contents of this report are solely the responsibility of the authors and do not necessarily represent the official views of ClearWay MinnesotaSM. The authors also wish to thank Ryan Strandjord, who prepared the monitoring equipment, performed field monitoring when necessary, and extracted the monitored data. The high quality of the data is largely due to his diligent work on this project. We are also indebted to the field monitors who tirelessly carried out the thousands of monitoring visits: Trent Byers, David Farrar, Matt Hruby, Melanie Larson, Aaron Norman, Carrie Quinlan, Angie Thomas, Jeff Thomas, Angela Vreeland, McKinzie Woelfel, and Nate Woelfel.


ASHRAE. 2007. ANSI/ASHRAE Standard 62.1-2007, Ventilation for Acceptable Indoor Air Quality, Appendix C.

Atlanta: American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.

Beckwith T.G., N.L. Buck, and R.D. Marangoni. 1982. Mechanical Measurements. Addison Wesley. Reading, Massachusetts. pp. 269-270.

Dick, J.B. 1950. Measurement of ventilation using tracer gas technique. Heating, Piping, Air Conditioning Journal, 5: 131-137.

Dietz, R.N., and E.A. Cote. 1982. Air infiltration measurements in a home using a convenient per fluorocarbon tracer technique. Environment International 8: 419-33.

Persily, A.K. 1997. Evaluating building IAQ and ventilation with indoor carbon dioxide. ASHRAE Transactions 103(2).

Sherman, M.H. 1990. Tracergas techniques for measuring ventilation in a single zone. Building and Environment 25: 365-74.

Waring, M.S., and J.A. Siegel. 2007. An evaluation of the indoor air quality in bars before and after a smoking ban in Austin, Texas. Journal of Exposure Science and Environmental Epidemiology 17: 260-268.


Sriram Somasundaram, Manager, Building Energy Systems & Technologies Group, Pacific Northwest National Laboratory: How many times was each venue visited so that you could get different occupancy patterns in each of them?

Martha Hewett: Each venue was visited approximately 36 times, but these measurements, from the longer two hour visits, were made primarily on Friday and Saturday nights and so were at high occupancy levels.

Carlos Lisboa, Engineer, BLC Navitas, LDA: To validate the assumptions made in the calculation of the calculated ventilation rates, namely the [CO.sub.2] generation rate by internal sources, did the team measure directly the ventilation airflow in ducts or diffusers/grilles in any of the places analyzed?

David Grimsrud: No, direct measurements of flows in ducts were not made in this set of measurements. These measurements were made without direct involvement of the owners or managers of the restaurants visited. However, the technique has been used in other settings where duct flow measurements and tracer decay measurements were also used as comparisons. See, for example, Grimsrud et al., 2011. In this study of three large retail stores, the ratio of ventilation rates measured by the TAB contractor and the mass-balance model using [CO.sub.2] concentrations was 1.09 +/-0.23; the ratio of the tracer gas decay measurements of ventilation rates to the mass-balance model results was 0.94 +/-0.13.

David L. Bohac, PE


Martha J. Hewett


Kristopher I. Kapphahn

David T. Grimsrud, PhD


David L. Bohac is the Director of IAQ, Martha J. Hewett is the Director of Research, and Kristopher I. Kapphahn is a research technician at the Center for Energy and Environment, Minneapolis, MN. David T. Grimsrud is principal of Grimsrud & Associates, Minneapolis, MN.
COPYRIGHT 2012 American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Bohac, David L.; Hewett, Martha J.; Kapphahn, Kristopher I.; Grimsrud, David T.
Publication:ASHRAE Transactions
Article Type:Report
Geographic Code:1USA
Date:Jan 1, 2012
Previous Article:Effect of hybrid ventilation system on indoor environment and annual cooling load in a high-rise building.
Next Article:Parametric Laboratory Evaluation of Residential Heat Pump Water Heaters.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters