5.6 Commercial Cooking

Category ID	Description	EIC
29	Cooking - Charbroiling	69068060000000
935	Bakeries - Small Bakeries	42041260120000
1710	Cooking - Deep Fat Frying	69068260000000
1711	Cooking - Unspecified (Griddles)	69068460000000

Introduction

This document describes the Area Source Methodology used to estimate emissions of carbon dioxide (CO₂) from commercial cooking operations in the Bay Area. Commercial cooking refers to the cooking of meat products such as steak, hamburger, poultry, pork, and seafood and french fries at food service establishment, including full-scale restaurants and limited-service restaurants, as well as cafeterias, buffets, and catering facilities. The emissions inventory focuses exclusively on process-related emissions from cooking activities, excluding any emissions from fuel combustion. Emissions are attributed to the specific cooking equipment or devices used in these establishments.

Based on numerous studies, the following types of equipment are the primary contributors to commercial cooking emissions and are the focus of this inventory:

Charbroilers – both underfired and chain-driven units
Deep-fat fryers
Other common commercial cooking appliances – such as clamshell grills and flat griddles

The CO₂ emissions calculated under this category only accounts for emissions generated from cooking meats and their marinades. Emissions from the combustion of fuels used by the cooking device are covered in other source categories (e.g. category # 1591, Commercial Natural Gas Combustion – Other, EIC 060-995-0110-0000) that are covered in the Commercial Combustion - Natural Gas methodology documentation.

Methodology

This source category is classified as an area source because individual cooking operations are small, widespread, and not feasible to track as point, mobile, or biogenic sources. However, these sources can be aggregated geographically—typically at the county level—and their emissions are estimated collectively using the methodology described below:

Base Year(s) Emissions _{county,pollutant}=

Activity Data × Emission Factor_pollutant × Control Factor_pollutant × Fraction_county× Fraction_{in District}× GWP_pollutant

Where:

Activity Data: is the total regional throughput or activity data for applicable base year(s).
Emission Factor_pollutant: is a factor that allocates an amount of emissions, in mass, of a particular pollutant by unit of activity data.
Control Factor_pollutant : is a fractional ratio (between 0 and 1) that captures the estimated reduction in emissions as a result of Air District rules and regulations.
Fraction_county: is the fraction of total regional emissions (between 0 and 1) estimated to be allocated to a particular county.
Fraction_{in District} : The BAAQMD jurisdiction covers only a portion of Solano and Sonoma County. For this reason, additional allocation must be done for these counties to determine the proportion of the county’s emissions occurring within BAAQMD’s jurisdiction.
GWP_pollutant is the Global Warming Potential of a particular GHG pollutant. The current version of the GHG emissions inventory incorporates the global warming potential (GWP) reported in the Fifth Assessment report of the Intergovernmental Panel for Climate Change (IPCC, 2014). The GWPs for the three principal GHGs are 1 for carbon dioxide (CO₂), 34 for methane (CH₄), and 298 for nitrous oxide (N₂O), when calculated on a 100-year basis with climate-carbon feedback included.

Once base year emissions are determined, historical backcasting and forecasting of emissions relative to the base year emissions are estimated using growth profiles as follows:

Current Year Emissions _county = Base Year(s) Emission _county × Growth Factor

Estimating emissions from commercial cooking is a complex process due to the need to account for multiple variables, including the types of meat cooked, cuisine types, and the gross revenue of restaurants. These factors are incorporated into the activity data used to estimate emissions by cooking equipment type. The methodology begins by uniformly estimating the total quantity of meat and french fries cooked across various equipment types at the county level. This involves determining the number of restaurants in each county, categorized by cuisine type and gross revenue, to reflect differences in cooking practices and output. Data on the amount of meat cooked per equipment type is based on a 2001 survey conducted by the Pacific Research Institute (PRI) under sponsorship from the California Air Resources Board (CARB). Estimates of french fry throughput come from a 2020 industry report prepared for Potatoes USA (Technomic, 2020), which characterizes cooking volume within the food service sector. Emissions are calculated by multiplying the total quantity of meat or french fries cooked on each type of equipment by pollutant-specific emission factors for CO₂. Additional details on activity data, county-level distribution, emission factors, and applicable control measures are provided in the following subsections.

Activity Data / Throughput

The activity data for this source category includes the amount of meat and potatoes cooked on each type of cooking device by county. These amounts are estimated based on the number of restaurants in a county, the type of cuisine cooked at each restaurant, and the average amount of meat or potatoes cooked per restaurant that operate at least one of the cooking devices based on their cuisine type and gross revenue.

Number of Restaurants

To estimate emissions from commercial cooking, accurate restaurant counts by cuisine type are essential. In 2024, CARB purchased a proprietary restaurant dataset for California from Dun & Bradstreet (D&B, 2024), which includes information on cuisine type for each establishment. CARB processed the dataset and provided the Air District with aggregated restaurant counts by cuisine type for each Bay Area County (see Table 1). Cuisine categories include ethnic (e.g., Chinese, Japanese, Thai), family-style, fast food, seafood, and steakhouses/barbecue. As part of the data cleaning process, CARB removed entries located on residential parcels, vacant lots, wilderness areas, and parks to improve accuracy.

In previous base-year inventories, the District relied on detailed industry reports from the California Employment Development Department (EDD), which provided county-level counts of privately owned food service businesses. These included full-service and limited-service restaurants (e.g., fast-food franchises), as well as caterers, buffets, and cafeterias. At the time, the EDD dataset was considered the most complete source, as efforts to obtain accurate counts from third-party vendors and county health departments had proven unsuccessful.

Earlier attempts by the District to use proprietary datasets from InfoUSA (2017) and D&B (2023) revealed major limitations, including:

Duplicate listings at the same address
Closed or ownership-changed businesses
Establishments without on-site cooking
Listings for plant-based restaurants
Entries unrelated to food service

Similarly, restaurant counts from county health departments—which conduct annual inspections—were found to be significantly lower than those provided by EDD or third-party vendors. While many of the earlier data quality issues have been addressed in the D&B dataset processed by CARB, some inaccuracies remain. Notably, duplicate listings at the same address persist and are estimated to inflate restaurant counts by approximately 8–13%, leading to a corresponding overestimation of emissions (see Uncertainties section for further discussion).

Table 1. Number of restaurants by cuisine-type from CARB (2024).

County	Ethnic	Family	Fast Food	Seafood	Steak & BBQ	Total
Alameda	1179	192	297	39	77	1784
Contra Costa	634	107	176	16	55	988
Marin	208	28	28	5	15	284
Napa	93	23	29	3	11	159
San Francisco	1119	222	117	43	56	1557
San Mateo	583	117	114	31	38	883
Santa Clara	1240	214	293	50	107	1904
Solano	146	22	71	9	16	264
Sonoma	298	33	76	5	13	425

To better reflect differences in cooking activity among restaurants, the number of establishments within each cuisine type was further distributed by gross revenue. This step accounts for the general relationship between a restaurant's size and the quantity of meat cooked—larger, higher-grossing restaurants are expected to serve more patrons and therefore cook more meat. While exceptions exist (e.g., high-end restaurants with limited menus and high prices), the overall assumption that higher revenue correlates with higher meat throughput is considered reasonable for most establishments.

Gross revenue data was sourced from third-party vendor datasets purchased from InfoUSA (2017) and D&B (2023). These datasets were previously used to develop detailed restaurant inventories for community-level assessments conducted under California Assembly Bill 617 (AB 617).

As of 2024, restaurant inventories were completed for Richmond-North Richmond-San Pablo, East Oakland, and Bayview Hunters Point under the AB617 program (A detail restaurant inventory was not developed for West Oakland, when it was adopted in 2019). Within each defined community boundary, restaurants were classified into five gross revenue categories:

Class A: Less than $500,000,
Class B: $500,000 to $1 million,
Class C: $1 million to $2 million,
Class D: $2 million and $3 million, and
Class E: Over $3 million.

The majority of restaurants across all cuisine types—approximately 80%—fall into Class A, indicating they are relatively small businesses. A notable exception is the fast food category, where most establishments fall into Class C. Each county was assigned a distribution of restaurants by revenue class using data from the AB 617 community inventory located within its boundaries. For example:

The revenue distribution from East Oakland was applied to all of Alameda County
Richmond–North Richmond–San Pablo data was applied to Contra Costa County
Bayview–Hunters Point data was applied to San Francisco County

For counties without AB 617 community data, average percentages across all three communities were used to assign revenue distributions by cuisine type (see Table 2).

Table 2: Percentage of Cuisine Type by Revenue by County (AB617 Community)

Cuisine Type	Revenue	Contra Costa (%)	Alameda (%)	San Francisco (%)	All other counties (average %)
BBQ	A	81.0	81.0	81.0%	81.0
	B	9.5	9.5	9.5	9.5
	C	4.8	4.8	4.8	4.8
	E	4.8	4.8	4.8	4.8
Ethnic	A	88.7	88.5	88.5	88.6
	B	4.3	6.1	6.1	5.7
	C	7.1	4.8	4.8	5.2
	D	0	0.6	0.6	0.5
Family	A	84.6	68	80	75.5
	B	5.1	10.7	6.7	8.2
	C	5.1	13.3	6.7	9.4
	D	5.1	6.7	6.7	6.3
	E	0	1.3	0	0.6
Fast Food	A	28	35.3	28	31.7
	B	0	7.8	0	4.0
	C	68	52.9	68	60.4
	D	4	3.9	4	4.0
Seafood	A	100	100	100	100

Finally, the number of restaurants by cuisine type and gross revenue was determined for each county by multiplying the number of restaurants by cuisine (Table 1) by the percentage of restaurants in each revenue classification (Table 2).

Cooking Equipment

The next step involves estimating the percentage of restaurants, by cuisine type, that use one or more types of commercial cooking equipment. This information is drawn from the Pacific Research Institute (PRI) study (see Table 3), which surveyed restaurants and recorded the number of each equipment type in use. Initially, the methodology included multiplying these percentages by the average number of equipment units per restaurant, based on cuisine type. However, this step is no longer required. The updated emission estimation methodology uses process rates that quantify the amount of meat cooked per week per restaurant, assuming the restaurant operates at least one unit of the relevant equipment. As a result, the specific count of equipment units is no longer factored into the emissions calculation.

Table 3. Type of cooking equipment used per restaurant category

Equipment Type	Ethnic	Family	Fast Food	Seafood	Steak and BBQ
Percent of restaurants with equipment (PRI Table 4)
Chain-Driven charbroilers	3.5	10.1	18.6	0.0	6.9
Underfired charbroilers	47.5	60.9	30.8	52.6	55.2
Deep-fat fryer	81.9	91.4	96.8	100.0	82.8
Flat griddles	62.7	82.9	51.9	36.8	89.7
Clamshell griddles	4.0	1.4	14.7	10.5	0.0

The percentage of restaurants with equipment by cuisine was then multiplied by the number of restaurants by cuisine and gross revenue estimated in the Number of Restaurants section to derive the number of restaurants by cuisine, gross revenue, and county that use at least one of each equipment type.

Process Rates

Emissions are estimated based on the quantity of food cooked on each type of equipment, multiplied by pollutant-specific emission factors. While restaurant-level equipment estimates were originally used to scale emissions, the updated methodology applies cooking rates directly on a per-restaurant basis for those known to operate a particular equipment type.

Meat Cooking Rates
Meat products—including hamburgers, steak, chicken (skinless and skin on), pork, and seafood—represent the majority of food prepared on charbroilers, flat griddles, and deep-fat fryers. Cooking volumes by meat type and equipment were derived from the PRI study. Weekly meat throughput (in pounds) was assigned to each restaurant revenue class using the following approach:

Class A: Assigned the lower bound of the 95% confidence interval
Classes B/C: Assigned the geometric mean
Classes D/E: Assigned the upper bound of the 95% confidence interval

All meat types reported in the PRI study were included in the inventory, except for the "Other" category, which was excluded due to the absence of specific emission factors.

Potato Cooking Rates
A separate analysis was conducted for potatoes, based on the U.S. EPA’s 2020 National Emissions Inventory (NEI) Technical Support Document: Commercial Cooking (USEPA, 2023). Most potatoes cooked in deep-fat fryers at limited-service and fast casual restaurants are frozen (primarily as french fries), with frozen varieties being used at a rate approximately ten times greater than fresh potatoes, which can be roasted, baked, or fried. National data on frozen potato purchases by restaurant type were sourced from a 2020 Technomic report prepared for Potatoes USA. To allocate this national quantity to Bay Area counties, the number of limited-service and full-service restaurants per county was divided by the national total. Because the D&B dataset only includes California restaurants, national restaurant counts could not be derived from it. Instead, the District used data from the California EDD (CAEDD, 2023), which provides workforce and business establishment statistics. The number of full-service and limited-service restaurants in each Bay Area County and nationwide (as of Q4 2023) from EDD was used to calculate county-level allocation fractions (see Table 4).

Since the latest available potato sales data are from 2020—a year likely impacted by the COVID-19 pandemic—the District adjusted the national purchase volume to reflect 2023 conditions. Growth multipliers were developed by dividing each Bay Area county’s 2023 population by its 2020 population, using data from the California Department of Finance (CADOF, 2024). These multipliers were then applied to scale 2020 potato sales to more accurately represent 2023 consumption (see Table 4). The quantity of french fries cooked in limited and full service restaurants are estimated by multiplying the national sales volume in 2020 of 5,593 million lbs of frozen potatoes by the fraction of restaurants in each county compared nationally and the growth multiplier to project 2020 potato sales to 2023. The amount of french fries cooked in limited and full service restaurants are then summed by county.

Table 4. Fraction of Restaurants in County and Growth Projections by County for 2023

County	Fraction of Limited and Full Service Restaurant compared to National Total	Growth Multiplier for 2023 to be applied to 2020 frozen potatoes Sales	Year 2023 Estimated Frozen Potatoes Cooked (lbs)
Alameda	0.0167	0.978	5.47 x 10⁷
Contra Costa	0.00732	0.982	2.30 x 10⁷
Marin	0.00172	0.964	4.19 x 10⁶
Napa	0.000861	0.972	2.15 x 10⁶
San Francisco	0.0142	0.954	4.20 x 10⁷
San Mateo	0.00875	0.964	2.89 x 10⁷
Santa Clara	0.0194	0.975	6.44 x 10⁷
Solano	0.00325	0.981	1.09 x 10⁷
Sonoma	0.00443	0.982	1.35 x 10⁷

County Distribution / Fractions

Only restaurants located within the District’s jurisdiction including parts of Solano and Sonoma Counties were incorporated into this inventory. Unlike other area source categories, the county distribution or fractions are not used for this inventory as commercial cooking inventory was developed following a bottom-up approach by estimated activity and operations on an individual restaurant basis.

ID	Description	ALA	CC	MAR	NAP	SF	SM	SNC	SOL	SON
1711	Cooking - Unspecified (Griddles)	0.21	0.12	0.03	0.02	0.19	0.11	0.23	0.03	0.05
29	Cooking - Charbroiling	0.22	0.12	0.03	0.02	0.19	0.11	0.23	0.03	0.05
935	Bakeries - Small Bakeries	0.27	0.12	0.04	0.01	0.13	0.17	0.14	0.02	0.10

Emission Factors

CO₂ emissions are emitted from the combustion of grease drippings on the cooking surface and are estimated from measured emission factors by cooking equipment type and meat cooked (Kuehn et al., 1999). The study estimates CO₂ emission factors from grease drippings only by subtracting the CO₂ emitted from operation of the cooking equipment from the total CO₂ emitted from the entire cooking process.

Table 5. Emission Factors by Equipment and Meat Cooked

Equipment Type	Meat/Food	CO₂ Emission Factors (lb pollutant/1000 lbs meat cooked)
Underfired Charbroilers	Hamburger	350
	Steak	184
	Poultry (with and without skin)	75
	Pork	75
	Seafood	35
Deep Fat Fryers	Poultry (with and without skin)	-
	Pork	-
	Seafood	-
	Potatoes	-
Flat Griddles	Steak	10.3
	Hamburger	19.6
	Poultry (with and without skin)	-
	Pork	-
	Seafood	-
Clamshell Griddles	Steak	-
	Hamburger	-
	Poultry (with and without skin)	-
	Pork	-
	Seafood	-
Chain-Driven Charbroilers	Hamburger (controlled)	350
	Hamburger (uncontrolled)	350
	Steak	184
	Poultry (with and without skin)	75
	Pork	75
	Seafood	35

Control Factors / Emission Controls

On December 5, 2007, the District adopted Regulation 6-2: Commercial Cooking Equipment (BAAQMD, 2007), to reduce PM₁₀ and volatile organic compound (VOC) emissions from commercial cooking equipment through a requirement to install control devices including certified catalytic oxidizer on chain-driven charbroilers that are mostly operated in franchise fast food restaurants. Certified controls are required for under-fired charbroilers in later years starting in 2010. However, to date, no control devices for under-fired charbroilers have completed the District’s certification process.

As both controls for chain-driven and under-fired charbroilers do not abate CO₂ produced from the cooking process, as such, there is no expectation of an impact on CO₂ process emissions from the full implementation and enforcement of this Rule.

Historical Emissions / Backcast

Emissions from commercial cooking between 1996 and 2007 were estimated using historical counts of eating establishments in the Bay Area, as reported in the Taxable Sales in California (Sales and Use Tax) publications. For years prior to 1996, emissions were estimated using the 2009 total population growth profile published by the Association of Bay Area Governments (ABAG) (ABAG, 2009).

Non-CO₂Emissions for charbroilers are lower starting in 2009 due to the implementation of controls under Regulation 6-2 for chain-driven charbroilers. CO₂ emissions are mostly in sync with gross revenue and meat/food consumption throughputs. A notable revision to the emissions inventory methodology was introduced with the 2023 inventory update, which incorporated gross revenue by cuisine type to more accurately estimate meat throughput. This refinement led to a significant adjustment to past inventory estimates. It is important to note that the apparent reduction in emissions in recent years is not reflective of actual changes in commercial cooking activity, but rather the result of improved methodological approaches.

Future Projections / Growth Factor

Future emissions for all categories are based on linear projection of the meat cooked data for 2023 scaled by the total population growth projections in future years from the California Department of Finance (CADOF, 2024).

Sample Calculations

The table below shows an example calculation for estimating base year 2023 PM_2.5emissions from family restaurants (with gross revenue of less than $500K – classified as A) using under-fired charbroilers to cook hamburgers in Alameda County.

Step 1	Estimate the number of family restaurants in Alameda County provided by CARB D&B, 2024	192
Step 2	Estimate the percentage of family restaurants that are classified as A based on gross revenue from AB617 data from East Oakland	68% or 0.68
Step 3	Estimate the number of family restaurants in Alameda County with gross revenue < $500K by multiplying the number of restaurants by the percentage of restaurants classified as A	192 × 0.68 = 130.6
Step 4	Obtain the percentage of family restaurants with under-fired charbroilers from the PRI Study (PRI, 2001)	60.9% or 0.609
Step 5	Estimate the total number of family restaurants (gross revenue less than $500K) in Alameda County that operate at least one under-fired charbroiler by multiplying the number of restaurants by the percentage of restaurants with the cooking equipment	130.6 × 0.609 = 79.54
Step 6	Obtain the CO₂emission factor for hamburgers cooked on under-fired charbroilers (lbs/1000 lbs meat; Kuehn et al., 1999)	350
Step 7	Obtain the 95% upper confidence limit lower bound quantity of hamburger cooked on under-fired charbroilers per week (lbs/1000 lbs of meat cooked) per restaurant from the PRI Study PRI, 2001	69.21
Step 8	Convert the weekly hamburger cooked to annual quantity on under-fired charbroilers (lbs/lbs of meat)	69.21 x 52 weeks/yr × (1/1000) = 3.60
Step 9	Estimate the annual CO₂emissions of hamburgers cooked on under-fired charbroilers per family restaurant (with gross revenue less than $500K) in Alameda County (lbs/yr) by multiplying the annual amount of hamburger cooked by the emission factor	3.60 × 350 = 1260
Step 10	Estimate the total CO₂emissions of hamburgers cooked on under-fired charbroilers from all family restaurants (gross revenue < $500K) in Alameda County by multiplying the annual PM2.5 emissions per restaurant by the number of restaurants in Alameda that operate at least one under fired charbroiler. Convert the emissions into units of million metric tons per CO2 equivalent (MMTCO₂eq/yr)	1260 lbs/yr x 79.54 x × 1/2000 tons/lbs × 0.907 MT/tons = 45.44 MT/yr = 4.54×10^-5 MMTCO₂eq/yr

Assessment of Methodology

Significant changes were incorporated into this base year inventory as compared to the previous base years. Details regarding the updates are listed in the table below.

Year	Revision	Reference
2023	Restaurant counts by cuisine type in each county based on D&B from CARB Used the percentage of restaurants that operate at least one of cooking equipment by cuisine type Assigned the percentage of restaurants in each cuisine type by gross revenue to quantity the amount of meat cooked by high, medium, and low amounts Emission factors by type of meat cooked on specific cooking appliance taken from SCAQMD and USEPA. Ratioing method used to fill-in emissions factors where factors are not available. PM_2.5is assumed to be 97% of PM₁₀ for chain-driven charbroilers CO₂ emission factors from grease combustion only derived from ASHRAE report NO_x, SO_x, and CO assumed to be products from fuel combustion and excluded from inventory Projected population data by county used for forecast Inclusion of throughput and emissions from frozen potatoes cooked in deep fat fryers	D&B, 2024 PRI, 2001 AB617 community assessments SCAQMD, 1998 and USEPA, 2023. District methodology of ratioing existing emission factors to fill in missing factors. EPA Wagon Wheel, 2023 Kuehn et al., 1999 Updated methodology to be consistent with other Air Districts CADOF, 2024 Technomics, 2020
2015	Number of restaurants based on California EDD for full-service and limited service restaurants, cafetarias, grills buffets, and buffets Continued to use the number of each type of cooking equipment based on report prepared for South Coast Air Quality Management District report (1997) Emission factors from study performed for SCAQMD CO and CO₂ emissions estimated from combustion Total organic gas (TOG) emitted is considered to be all reactive (TOG = ROG) PM_2.5 is assumed to be 97% of total PM and PM₁₀ is 100% of total PM Projected population data from ABAG used for forecasts	CA EDD, 2015 Welch, 1997 Welch and Norbeck, 1998. Please note that emission factors by meat were applied across other equipment if factors were not available. Some emission factors were used in error and have been corrected in 2023 update. Emission factors based on ASHRAE report (Kuehn et al., 1999) CARB, 2022 organic specification profile Not referenced, but erroneously listed as 77% in methodology documentation ABAG, 2013
2011	Number of restaurants based on data provided by Bay Area County Health Department. The District eliminated restaurants not open to the public, gone out of business, and were less likely to cook. Emissions factors from SCAQMD and the University of Minnesota Number of each type of cooking equipment based on South Coast Air Quality Management District report (1997) – 33% of restaurants have under-fired charbroilers, 4% have chain-driven charbroilers, 62% have deep fat fryers, and 52% have griddles. Total organic gas (TOG) emitted is considered to be all reactive (TOG = ROG) Growth profile from ABAG 2009 were used to forecast. Seasonal variations in sales were account assuming higher sale during summer months due to longer daylight hours and warmer weather	Not referenced Welch and Norbeck, 1998 and University of Minnesota (not referenced) Welch, 1997 CARB organic speciation profile (not referenced) ABAG, 2009 Not referenced
Past Inventories to 1996	Historical number of eating establishments in the Bay Area were based on “Taxable Sales in California” (Sales and Tax Use Tax) publications. Prior to 1996, ABAG’s 2009 total population growth profiles were used.	Not referenced

Emissions

The table below shows the total emissions by criteria air pollutant in tons per year for commercial cooking categories.

ID	Description	CO2	CO2_bio	Total
29	Cooking - Charbroiling	8834.5	0.0	8834.5
1711	Cooking - Unspecified (Griddles)	526.6	0.0	526.6
935	Bakeries - Small Bakeries	0.0	226.7	226.7

Summary of Base Year 2022 Emissions

The tables below show the contribution of commercial cooking GHG emissions to the overall regional total and to the Commercial & Residential sector.

Contribution of Commercial Cooking Emissions by Sector

Subsector	Sector	Subsector GHG Emissions (MMTCO2eq)	Sector GHG Emissions (MMTCO2eq)	% of Sector
Commercial Cooking	Commercial + Residential	0.009	12.85	0.07%

Contribution of Commercial Cooking Emissions to Regional Total

Subsector	Subsector GHG Emissions (MMTCO2eq)	Regional Total GHG Emissions (MMTCO2eq)	% of Regional Total
Commercial Cooking	0.009	65.68	0.01%

Trends

The time series chart below shows the emission trends for commercial cooking.

Summary of Trends

Long-term projections for the commercial cooking sector are inherently uncertain due to the industry's high turnover rate. Nationwide, an estimated 30% of restaurants fail, with approximately 17% closing within their first year of operation (Mohamed, A., 2024). While failure rates vary by city, common reasons include inexperience, poor customer service, operational inefficiencies, weak leadership, and resistance to evolving industry trends. However, the dynamic nature of the industry means that new restaurants frequently open to replace those that close, resulting in a relatively stable overall number of establishments in many areas. These fluctuations are difficult to capture in the emissions inventory.

To account for this variability, the inventory assumes that restaurant growth is proportional to population growth. Accordingly, future emissions for all commercial cooking categories are projected using a linear population growth forecast from the California Department of Finance (CADOF, 2024).

Uncertainties

Numerous uncertainties affect the accuracy and representativeness of the commercial cooking emissions inventory. These uncertainties stem primarily from the use of dated data sources, assumptions regarding restaurant characteristics, and limitations in available emission factors and activity data. Key sources of uncertainty include:

Incomplete Source Coverage: This inventory does not account for all commercial cooking emissions. Excluded sources include institutional kitchens (e.g., schools, prisons), public events (e.g., fairs, sporting events), caterers, and non-permanent operations such as food trucks.
Restaurant Turnover and Representativeness: The restaurant industry is highly volatile, with approximately 30% of restaurants closing annually and 17% of independently owned full-service restaurants failing within the first year (Mohamed, 2024). As a result, restaurant counts represent only a snapshot in time. Consequently, associated attributes such as cuisine type and gross revenue may not accurately reflect current operations, particularly when a closed restaurant is replaced by a non-restaurant business or a conceptually different establishment.
Outdated Activity and Throughput Data: Much of the activity data, including meat throughput, cuisine classifications, and cooking equipment profiles, is based on a 2001 survey of California restaurants conducted by the Pacific Research Institute (PRI). Given the age of this data, it may no longer reflect current cooking practices, equipment use, or consumer preferences. For example, the rise of fusion cuisine and specialized, single-dish restaurants may not be well represented in the original survey framework. CARB is currently soliciting contractors to complete a three year study to update the activity data reported in the PRI study which will be available in 2029.
Limitations in Equipment Usage Assumptions: The PRI survey focused on restaurants operating charbroilers but did not account for differences in charbroiler size or usage patterns. For example, small one-foot charbroilers are often used only to sear grill marks on food without actually cooking the meat. These differences in operational practices could significantly affect emission estimates.
Use of Gross Revenue as a Proxy for Throughput: Gross revenue was used as a proxy to estimate meat throughput, based on the assumption that higher-grossing restaurants serve more food. However, this approach may overestimate emissions for high-end fine dining establishments, which may generate high revenue from limited, premium offerings but cook smaller quantities of meat compared to similarly grossing chain restaurants. The inventory does not differentiate between restaurant types such as fine dining and fast casual franchises.
County-Level Assumptions from Localized Assessments: The distribution of restaurants by cuisine type and gross revenue was derived from third-party datasets (D&B and InfoUSA) used in AB 617 community assessments. These local distributions were extrapolated to represent entire counties, although the District has no independent verification that these assumptions are accurate or representative of the broader regional restaurant population.
Unaccounted Emission Controls: The inventory does not include reductions from local control measures or voluntary best practices implemented by restaurant operators. Examples include electrostatic precipitators (ESPs) used in areas such as San Francisco’s Embarcadero and San Jose’s Santana Row, as well as wet scrubbers, taller exhaust stacks, and routine cleaning of exhaust ducts—all of which can significantly reduce emissions but are not captured in this inventory.
Duplicate Restaurant Listings: The D&B and InfoUSA datasets are estimated to contain 8% to 13% duplicate entries, varying by county. These duplicates can lead to an overestimation of emissions by up to 13%. While efforts were made to identify and flag duplicate records, manual validation and removal were not feasible due to resource constraints.
Outdated Emission Factors: Emission factors from Welch et al. (1998) were developed using an EPA test method that is no longer valid. More recent studies or updated factors using current EPA-approved test methods may yield different results. Additional research is needed to determine whether more accurate and representative emission factors are available, especially across different meat types and equipment.

Contact

Author: Virginia Lau

Reviewer: Nicholas Tang

Last Update: June 20, 2025

References

ABAG. 2009. Forecasts and Projections, Association of Bay Area Governments. https://abag.ca.gov/our-work/land-use/forecasts-projections

BAAQMD. 2007. Regulation 6 Rule 2: Commercial Cooking Equipment, Bay Area Air Quality Management District. https://www.baaqmd.gov/~/media/dotgov/files/rules/reg-6-rule-2-commercial-cooking-equipment/documents/rg0602.pdf?rev=42fc0966398c43f9b585572708a5ea70&sc_lang=en

California AB 617. (C. Garcia, Chapter 136, Statutes of 2017)

https://leginfo.legislature.ca.gov/faces/billNavClient.xhtml?bill_id=201720180AB617

CADOF. 2024. E-5 Population and Housing Estimates, 2024, California Department of Finance. https://dof.ca.gov/forecasting/demographics/estimates/e-5-population-and-housing-estimates-for-cities-counties-and-the-state-2020-2024/

CAEDD. 2023. California Employment Development Department, 2015 and 2023.

https://labormarketinfo.edd.ca.gov/cgi/dataanalysis/areaselection.asp?tablename=ces

CARB, 2022. ORGPROF.

https://ww.2arb.ca.gov/speciation-profiles-used-carb-modeling

D&B. 2024. Restaurant Listing for California, purchased by CARB and aggregated listing provided to BAAQMD.

D&B. 2023. Restaurant Listing for San Francisco County only, purchased by Air District in support of AB617 emissions inventory and analysis.

InfoUSA. 2017. Restaurant list for regional modeling.

Kuehn, et al., 1999. Comparison of emissions from selected commercial kitchen appliances and food products. American Society of Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE) Transactions, 105, PART 2/-. https://store.accuristech.com/standards/4285-rp-745-comparison-of-emissions-from-selected-commercial-kitchen-appliances-and-food-products?product_id=1719327&srsltid=AfmBOorJZ3WUhEf3qfWkdDiQXEUw8E7fJ5B85z4w8ye4Br9eCkCYqJ9E

McDonald, et al. , 2003. Emissions from Charbroiling and Grilling of Chicken and Beef, Table 2, Journal of Air and Waste Management Assoc, 53: 185-194

Mohamed, A., 2024. Restaurant Failure Rate Statistics and Management Insights. November 27, 2024. https://blogs.oregonstate.edu/nexus/2024/11/27/restaurant-failure-rate-statistics-and-management-insights/#:~:text=According%20to%20the%20National%20Restaurant,can%20be%20challenging%20and%20demanding.

PRI. 2001. Charbroiling Activity Estimation - Prepared for the CARB and California Environmental Protection Agency, Public Research Institute, San Francisco State University, San Francisco. June 30, 2001.

https://ww2.arb.ca.gov/sites/default/files/classic/research/apr/reports/l943.pdf

Technomic, 2020. Domestic Sales and U.S. Potato Utilization Report - 2019/2020 Marketing Year, prepared for Potatoes USA for Potato Operator Meeting Presentation.

https://potatoesusa.com/wp-content/uploads/2020/12/MY-20-Sales-Utilization-Final-Report.pdf

USEPA. 2023. 2020 National Emissions Inventory Technical Support Document: Commercial Cooking, Office of Air Quality Planning and Standards, Air Quality Assessment Division, United States Environmental Protection Agency. EPA-454/R-23-001s. March 2023. https://www.epa.gov/system/files/documents/2023-03/NEI2020_TSD_Section19_CommercialCooking.pdf

US EPA, 2023. Wagon Wheel Emission Factors.

Welch, W.A., 1997. Further Development of Emission Test Methods and Development of Emission Factors for various Commercial Cooking Operations. Prepared for South Coast Air Quality Management District.

Welch, W.A. and Norbeck, J.M. 1998. Development of Emission Test Methods and Emission Factors for Various Commercial Cooking Operations. Prepared for South Coast Air Quality Management District. Other Documents and Presentations, Symposium Paper

for American Society of Heating, Refrigerating, and Air-Conditioning Engineers Annual Meeting.

Toronto, June. TO-98-14-3, 1998. https://store.accuristech.com/standards/to-98-14-3-development-of-emission-test-methods-and-emission-factors-for-various-commercial-cooking-operations?product_id=1711344&srsltid=AfmBOor9YWrXceW_CMjLV8e-WRbMBt31d7C0xgkdQxV-VHNY8v0AJfwN