Kitchen hood performance in food service operations.

Introduction

Cooking processes in commercial kitchens create heat, gases, vapors, fumes, and particles. Ventilation that controls these emissions is needed to reduce a variety of risks. Risks include fire from grease fumes deposited on surfaces, contamination of foods by grease, and occupational exposure to chemicals and heat.

Public health sanitarians regularly inspect food service establishments to reduce public health risks. While the emphasis of these inspections is on consumer protection against foodborne disease, food service codes also address general environmental conditions. Sanitarians have an opportunity to improve public health and safety by assessing the effectiveness of a facility's ventilation systems.

Although quantitative guidelines for hood performance exist, kitchen hood assessments are typically qualitative evaluations of cleanliness. Two questions are addressed in this study: 1) Are hoods meeting quantitative operating guidelines? and 2) Do qualitative inspections identify hoods that are not meeting the guidelines?

The Risks

The risks from inadequately operated and maintained kitchen hoods are linked to the release of grease fumes. Fumes are formed when cooking oils and oils in food are heated. During heating the vapor pressure of oil increases, and the oils are volatilized into vapors. As these vapors move away from the heat source and cool, the molecules condense into very small liquid and solid particles. The chemical composition of these fumes varies with the cooking oils and food products heated and the chemical changes that may occur during the cooking process. The aerosol size of these fumes is very small, generally less than 1.0 microns ([micro]m) in aerodynamic diameter. This small size enables them to stay aloft and, if not appropriately controlled by ventilation, to deposit on surfaces throughout a facility. This surface deposition can create potential fire and food contamination hazards. The small size of these fumes also allows penetration deep into the lungs of exposed workers.

Inhalation Exposure

Cooking fumes pose an inhalation hazard to exposed workers. Carcinogenic polycyclic aromatic hydrocarbons (PAHs) have been found in cooking fumes (Siegmann & Sattler, 1996, 1999). These fumes also have been found to be genotoxic and associated with increased rates of lung cancer (Wu, Chiang, Ko, & Lee, 1999). The PAHs identified in cooking fumes do not have specific occupational exposure limits. Some of them are covered as a mix in the coal tar pitch volatile standard. This standard is not particularly relevant to the exposures of restaurant workers. Nonetheless, personnel working in food service establishments are exposed to substances known to be hazardous, even if occupational exposure limits have not been published for them. Another study conducted in Finland confirmed that food industry workers may be exposed to relatively high levels of airborne impurities (Vainiotalo & Matveinen, 1993). Occupational exposure to these fumes can be controlled with proper local exhaust ventilation.

Fire

Over 40 percent of the fires in eating and drinking establishments are attributed to cooking (U.S. Fire Administration, 2001). Uncontrolled aerosolized grease can coat surfaces with flammable material. Properly operating kitchen hoods can capture fumes and prevent widespread grease contamination, reducing fire risks (National Fire Protection Association [NFPA], 1997). Grease accumulation in hood systems can still present localized risk of fire, and proper maintenance and cleaning of these systems is needed to keep fire risks low. If a fire does occur, a properly operating ventilation system can slow the spread of flames outside of the hood (NFPA, 1997). Qualitative visual inspections are the typical assessment hoods receive during food service inspections. This aspect of the inspection is usually of lower priority than the food safety aspects and is often done cursorily or overlooked.

Local Exhaust Hoods

Local exhaust ventilation hoods control emissions from cooking. Fumes are drawn into hoods, where the grease can be collected. If the system is properly operating and regularly cleaned, fire and food sanitation risks are reduced. The use of kitchen hoods also has been shown to reduce airborne exposures to fumes (Chaing, Wu, & Ko, 1998; Siegmann & Sattler, 1999). The relevant National Fire Protection Association standard, NFPA 96, states that adequate airflows should be maintained and adequate hood size should be used (NFPA, 2001). This standard does not, however, specify performance criteria for hoods. The American Conference of Governmental Industrial Hygienists (ACGIH) and the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) provide quantitative performance guidelines for the operation of kitchen hoods.

ACGIH design guidelines for kitchen range hoods are based on the geometry of the hood (ACGIH, 2001). Hoods are classified according to the following categories: hood against wall (HAW), low side-wall hood (LSW), and island-type hood (ITH). Low side-wall hoods are similar to HAWs, but the canopy width can be up to 1 foot shorter than the width of the cooking surface. For hoods of this type, the recommended volumetric flow rate ([Q.sub.rec]) is based on the dimensions of the hood. The [Q.sub.rec] assumes that hood dimensions (distance from cooking surface to hood, hood overhang, etc.) also meet recommendations.

ASHRAE flow rate design guidelines are based on the type of cooking process as well as the hood geometry (ASHRAE, 1995). Hoods that control heat or water vapor when grease is not present are classified as Type II hoods. Examples of Type II hoods are those used with pizza ovens and other operations solely involving baking. Type I hoods are used for processes that generate fumes and grease. They are further categorized as light, medium, heavy, and extra-heavy duty. Light-duty hoods are used for processes involving ovens, steamers, and small kettles. Medium-duty hoods control emissions from large kettles, ranges, and griddles. The heavy-duty classification is reserved for upright broilers, charbroilers, and woks. The extra-heavy-duty classification refers to hoods used for cooking with solid fuel. Extra-heavy-duty operations were not observed in this study.

The ASHRAE hood geometry classifications are similar to the ACGIH scheme. Hoods are categorized according to the following categories: wall-mounted canopy, back-shelf canopy, island canopy, pass-over style, and eyebrow type. The wall-mounted, back-shelf, and island canopy types are analogous to ACGIH's HAW, LSW, and ITH classifications and were the types evaluated in this study. The ASHRAE guidelines provide a range of acceptable flow rates. For this study, the midpoint of the recommended range was considered the performance target, [Q.sub.rec].

Restaurant Inspection

Public health sanitarians regularly inspect restaurants to reduce the risk of foodborne illness. As public and environmental health professionals, these inspections are also an opportunity to ensure worker safety and health, even though that may not be their primary purpose. Food service inspections rarely, if ever, assess airflow through range hoods, however.

The Ohio Uniform Food Service Code, in addition to identifying food protection requirements, addresses ventilation in food service establishments (Ohio Department of Health, 2001). The code states: "If necessary to keep rooms free of excessive heat, steam, condensation, vapors, obnoxious odors, smoke, and fumes, mechanical ventilation of sufficient capacity shall be provided." Thus the code does address worker protection from inhalation hazards and requires adequate ventilation. Other state codes have similar requirements. While the performance standard for these hoods is not specified in the code, the ACGIH and ASHRAE guidelines are appropriate benchmarks for assessing the adequacy of kitchen ventilation.

Methods

Public health sanitarians were accompanied on 60 restaurant inspections over a period of two months. A total of 89 hoods were evaluated. Each hood was classified as a hood against wall (HAW), island-type hood (ITH), or low side-wall hood (LSW). The hood was sketched onto graph paper, and the dimensions were measured. Measurements included length, width, slot dimensions (if present), face area, and distance from cooking surface.

The authors determined the average velocity with a multipoint velocity traverse of the face or the slots using a TSI Velocicalc Plus thermo-anemometer. At each point the velocity was recorded when the anemometer reading had stabilized. If the hood was a slotted, slot velocity ([V.sub.slot]) was measured at multiple locations for every other slot. Velocity at every third slot was measured for some hoods with more than 25 slots. If the hood had a duct takeoff that was not slotted, the face area was visually divided into rectangles of equal area. Face velocity ([V.sub.face]) was measured in the center of each rectangle. For 90 percent of the hoods, a measurement was taken for at least every 1 square foot ([ft.sup.2]) of face area. For a few of the larger hoods, there was a maximum of 2.5 [ft.sup.2] of face area for each measurement. Using the total slot area or face area and the average slot or face velocity, the authors calculated the volumetric flow rate through the hood ([Q.sub.hood]): [Q.sub.hood] = [V.sub.avg]A.

[FIGURE 2 OMITTED]

The dimensions and information on the hood classification were used to calculate the recommended volumetric flow rates ([Q.sub.rec]) for both ACGIH and ASHRAE guidelines.

In addition to the quantitative measurements, the field researcher rated the cleanliness of each hood on a scale of 1 (very clean) to 5 (very dirty). These values were later converted to ratings of either "clean" for scores of 1 and 2, or "dirty" for scores of 3, 4, and 5.

Results and Discussion

The most common type of hood in the restaurants inspected was the hood-against-wall configuration. The number and percentage of each type of hood evaluated in the study are shown in Figure 1. For the purpose of determining the ASHRAE target flow rate, the process also was classified as calling for a light-, medium- or heavy-duty Type I hood or as a Type II hood. Most processes used Type I medium duty (61 percent), with 26 percent using Type I heavy duty and 9 percent using Type II. Only 3 percent of the hoods were Type I light duty.

The required hood airflow is a function of the size and shape of the hood; the dimensions of the cooking surface; and, for ASHRAE guidelines, the type of cooking taking place. Large hoods and heavier-duty hoods require more flow than smaller or lighter-duty hoods. In comparing the airflow performance of hoods, the absolute flow rate is not an adequate metric. The actual flow rate as a proportion of the required flow recommendation is the key criterion. For the study reported here, the authors normalized the performance of each hood to allow systems of different types and sizes to be compared. The metric for assessing hood flow rates is the actual flow as a percentage of the recommended flow rate. This value was calculated as follows: Percentage of recommended rate = [Q.sub.hood]/[Q.sub.rec].

In analysis of any data, the first--and often ignored--step is to examine the distribution of the data. The distribution indicates which descriptive statistics, such as measure of central tendency (mean, median, or geometric mean) and inferential statistics, are appropriate. For example, t-tests used to compare the mean values between groups require that the data be normally distributed. Figure 2 shows the frequency distribution for percentages of ACGIH-recommended values. The distribution clearly is not symmetrical. The percentages of ASHRAE-recommended values had a similar distribution. Visually, the data appear to be log normally distributed. Figure 3 is a log-probability plot of the percentage of the recommended value for each data point versus the percentiles for each point. The approximately straight line of this plot with a logarithmic x-axis suggests that the sample is log normally distributed. The data pass formal tests of log normality in the Kolmogorov-Smirnov goodness-of-fit test (p < .05). Since the data are log normally distributed, the appropriate measures of central tendency and variability are the geometric mean (GM) and the geometric standard deviation. Also, because the data are log normally distributed, inferential statistics, such as t-tests, need to be done on the log-transformed values. Examination of the distribution of a data set is not a trivial exercise but is required for appropriate interpretation of data. Since so many environmental data are not normally distributed, all environmental health professionals should be aware of the importance of this initial step of data analysis.

Table 1 summarizes the descriptive statistics for hood performance in terms of percentage of the recommended ACGIH and ASHRAE guideline values. The GM performance of all hoods was less than 85 percent of the recommended flow rates. Some hoods were operating at less that 20 percent of the recommended flow. Table 2 breaks down the same information by geometry and use type. The GM values appear different among hood types, but because of variability within each hood type, the difference is not significant.

[FIGURE 3 OMITTED]

Figure 4 illustrates the proportion of the hoods meeting the airflow performance guideline. Clearly most hoods were not operating at the recommended flow rates. Hood performance was worse when the ASHRAE guidelines were used as a benchmark. Overall, 39 percent and 24 percent of the hoods, respectively, met the ACGIH and ASHRAE guidelines. As mentioned earlier, the ASHRAE recommendations account for the type of cooking being done. This approach resulted in higher target flow rates for some operations. When hoods were classified by both geometry and use, the number in each category became smaller, so it was more difficult to generalize results. Heavyduty hood-against-wall hoods, however, had a pass rate of only 18 percent (n = 17), compared with a pass rate of 36 percent (n = 39) for medium-duty HAWs.

While adequate ventilation is part of food service regulations, it is not a specific point on most food service inspection sheets. On the Ohio reports, ventilation is covered under the "Physical Facilities" section of the report, although none of the five checkpoints of that section specifically direct the inspector to any sort of evaluation of the ventilation system. Sanitarians noted inadequate ventilation flow rates by qualitative visual inspections only twice during this study, which included the 89 hoods. Both hoods were in the same establishment and were identified as operating at 73 percent and 24 percent of the recommended ACGIH flow rate and 51 percent and 34 percent of the recommended ASHRAE flow rate. The geometric mean for all hoods was 83.6 percent of the ACGIH target flow and 75.9 percent of the ASHRAE target flow. So the two hoods that were identified as problems were performing worse than most, with the second hood performing far worse. Overall, however, sanitarians failed to identify hoods operating at inadequate flow rates.

Sanitarians noted hood cleanliness during nine inspections for a total of 15 hoods. Of the 15 dirty hoods they noted, five actually were rated "clean" by the project researcher. The average cleanliness score of the noted hoods was not statistically different from that of the hoods not noted as dirty by sanitarians. Of the 18 hoods judged "very dirty" by the researcher, only three were noted by the inspecting sanitarians. The number of identifications of dirty hoods may even have been inflated because of the fact that a researcher specifically evaluating the ventilation systems accompanied the inspectors.

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

The hood flow performance was compared with the cleanliness rating to see if there was a relationship. Figure 5 illustrates the comparisons. The geometric mean of the ACGIH target flow for hoods rated as clean was 82 percent. For hoods rated dirty, the geometric mean of the ACGIH target was 84.2 percent. An independent t-test on the percentage of recommended flow and the relative-cleanliness ranking showed no statistical difference between the groups. Thus, the cleanliness of the hood is not a predictor of ACGIH airflow performance. When the ASHRAE criteria were used, however, the extent to which the flow rate approached the target does seem to be related to cleanliness rating. For "clean" hoods, the geometric mean of the percentage of recommended flow was 96 percent; it was 65 percent for "dirty" hoods. An independent-samples t-test for a difference between the geometric mean of the "clean"- and "dirty"-rated hoods indicated that the "clean" hoods in general were closer to the target flow rates. The reason is not clear. It may be that restaurants that are generally more conscientious about the cleanliness of their equipment also are more concerned with the correct operation of that equipment.

Conclusions and Recommendations

Most of the kitchen hoods in the restaurants inspected did not meet the recommended flow rates specified by either of the guideline-publishing organizations. Inadequate hood flow can create food safety and fire hazards by allowing grease fumes to spread throughout the facility and deposit on surfaces. In addition, these fumes pose a health risk to workers exposed at elevated concentrations.

Restaurant operators need to be more diligent about the operation of their kitchen ventilation units. Manufacturers of kitchen hoods provide performance criteria for their products that are in line with the guidelines from ACGIH and ASHRAE. Restaurants should be sure that contractors installing or working on ventilation systems are familiar with the published guidelines. Following installation, units should be evaluated in place to ensure adequate flow. They should be periodically re-evaluated to detect any change in performance.

Restaurant inspections could be an important part of reducing risks associated with poorly performing kitchen hoods. The current practice of qualitatively assessing hoods, however, rarely identifies hoods that are not achieving adequate airflow rates. Also, during the study reported here, sanitarians rarely noted hoods that needed cleaning. Having ventilation as a separate evaluation point would serve as a reminder to the sanitarian to at least check the cleanliness of the hoods. Determination of the flow through a hood is time consuming and could not be practically expected to be a part of every inspection. Periodic quantitative evaluation in some proportion of the inspections, however, would be a useful complement to more regular qualitative evaluations of hood performance. The heavy-use hood-against-wall type had the lowest performance rate. Perhaps targeting this category of hood with periodic quantitative assessment would be a resource-effective way of reducing risks.

Future investigations of restaurant hood performance should include the assessment of workers' exposures. The previous exposure studies were done in Switzerland, Taiwan, and China. Exposure information based on U.S. practices would help to further delineate the hazards associated with inadequate kitchen ventilation. The exposure data should be linked to ventilation performance to better assess risks.

FIGURE 1 Number of Each Hood Type Evaluated

Island; n = 10        11%
Against wall; n = 66  74%
Low side; n = 13      15%

Note: Table made from pie chart.

TABLE 1 Flow Rates as a Percentage of the Recommended Flow Rate (n = 89)

                              Percentage of ACGIH  Percentage of ASHRAE
                              Recommendation       Recommendation

Geometric mean (GM)                   83.6                75.9
Geometric standard deviation           1.87                2.07
Low                                   13.1                10.0
High                                 373                 791

TABLE 2 Hood Type Flow Rates as a Percentage of the Recommended Flow
Rate

                   Percentage of ACGIH   Percentage of ASHRAE
                   Recommendation        Recommendation
Geometry           Geometric Mean  GSD   Geometric Mean  GSD

  HAW                    83.0      1.92      72.4        2.00
  ITH                    76.5      1.99      99.3        1.71
  LSW                    90.3      1.56      76.1        2.66

ASHRAE usage type

  Type I light           --        --        93.4        1.76
  Type I medium          --        --        82.2        1.95
  Type I heavy           --        --        66.7        2.06
  Type II                --        --        55.0        2.69

Acknowledgements: This research was supported by a grant from the Bowling Green State University Faculty Research Committee. The authors would like to thank the participating sanitarians and health department officials for their cooperation and assistance in this research.

REFERENCES

American Conference of Governmental Industrial Hygienists. (2001). Industrial ventilation: A manual of recommended practice (24th ed.). Cincinnati, OH: Author.

American Society of Heating, Refrigerating and Air-Conditioning Engineers. (1995). ASHRAE handbook: HVAC applications. Atlanta, GA: Author.

Chaing, T.A., Wu, P.F., & Ko, Y.C. (1998). Prevention of exposure to mutagenic fumes produced by hot cooking oils. Environmental and Molecular Mutagenesis, 31, 92-96.

National Fire Protection Association. (1997). Fire protection handbook (18th ed.). Quincy, MA: Author.

National Fire Protection Association. (2001). NFPA 96 standard for ventilation control and fire protection of commercial cooking operations. Quincy, MA: Author.

Ohio Department of Health. (2001). Uniform food safety code, Ohio administrative.code. http://www.odh.state.oh.us/Rules/Final/Chap 3717_1/Fr3717_1_1st.htm (3 Dec. 2004).

Siegmann, R., & Sattler, K. (1996) Aerosols from hot cooking oils, a possible health hazard. Journal of Aerosol Science, 27(Suppl. 1), S493-S494.

Siegmann, R., & Sattler, K. (1999). Identification of carcinogens in cooking oil fumes. Environmental Research, 80(1), 18-22.

Vainiotalo, S., & Matveinen, K. (1993). Cooking fumes as a hygienic problem in the food and catering industries. American Industrial Hygiene Association Journal, 54(7), 376-381.

U.S. Fire Administration. (2001). Fire in the United States, 1989-1998 (12th ed.). Emmitsburg, MD: Federal Emergency Management Agency, U.S. Fire Administration, National Fire Data Center.

Wu, P.F., Chiang, T.A., Ko, Y.C., & Lee, H. (1999). Genotoxicity of fumes from heated cooking oils in Taiwan. Environmental Research, 80(2), 122-126.

Charles B. Keil, Ph.D., C.I.H.

Hailu Kassa, Ph.D., R.S.

Kenny Fent

Corresponding Author: Charles B. Keil, Associate Professor, Environmental Health Program, Bowling Green State University, 223 Health Center, Bowling Green, OH 43403-0280. E-mail: [e-mail].