3.7 Semiconductor Manufacture

    Category ID Description EIC
    2560 HFC-23 47047049990000
    2561 NF3 47047049990000
    2622 C2F6 (PFC-116) 47047049990000
    2623 C3F8 (PFC-218) 47047049990000
    2624 C4F8 (PFC-318) 47047049990000
    2625 CF4 (PFC-14) 47047049990000
    2626 SF6 47047049990000

    Introduction

    This document describes the methodology used to estimate greenhouse gas (GHG) emissions from high-global warming potential (high-GWP) gases, and more specifically those that are used in the semiconductor manufacturing industry.

    The use of these gases was adopted to replace and phase out ozone-depleting substances that were depleting the stratospheric ozone layer, under the 1987 Montreal Protocol and 1990 Clean Air Act Amendments. These replacement gases, including hydrofluorocarbons (HFCs) and perfluorocarbons (PFCs), though not harmful to the ozone layer contribute significantly to global warming due to their high heat trapping ability and long persistence in the environment. Owing to these characteristics, these gases are classified as GHGs with high global warming potential (GWPs). The San Francisco Bay Area (SFBA) GHG inventory for high-GWP emissions is derived from the California Air Resources Board’s (CARB) statewide GHG emissions inventory, and the SFBA inventory assigns these emissions by source categories (category numbers 2560-2561, 2622-2626), based on specific gases used in the Semiconductor Manufacture industry. In addition to HFCs and PFCs, high-GWP gases emitted in the semiconductor manufacturing subsector, for which emissions are reported herein, include NF3 and SF6. The predominant use of these compounds in the semiconductor manufacturing subsector is:

    • for cleaning chemical vapor deposition chambers, and,
    • for dry etching, which is a highly precise method of removing material from the surface of a semiconductor object by bombarding it with an ionized gas called plasma, that contains some of these high-GWP gases.

    Methodology

    The high-GWP source categories are considered area sources as they account for fugitive emissions from distributed devices that are not typically permitted by the Air District and hence are not systematically or annually catalogued. The inventory development for these categories uses an area source calculation approach, since the process-level fugitive emissions are widely distributed across thousands of emission release points and not necessarily measured directly.

    The emissions data for these high-GWP gases come directly from CARB’s Greenhouse Gas Inventory (CARB, 2023; latest data for year 2021) following the same source category classification used by CARB. The general methodology used by CARB to calculate emissions for the base year for these categories is as follows:

    Base Year Emissionscounty,pollutant =

    Emissionsstate;national,pollutant × Control Factorpollutant × Fractioncounty × Fractionin District × GWPpollutant

    Where:

    • Base Years: are years for which activity / throughput data are available in order to calculate emissions.
    • Emissionsstate;national,pollutant: is the amount of emissions from a larger area (e.g., state or national level) to be allocated to a smaller regional area based on a proportional measure, such as the ratio of county to state population.
    • Emission Factorpollutant: is a factor that allocates a mass amount of emissions of a particular pollutant per unit of activity.
    • Control Factorpollutant: is a fractional ratio (between 0 and 1) that estimates reductions in emissions from adopted rules and regulations.
    • Fractioncounty: is the fraction of total regional emissions (between 0 and 1) estimated to be allocated to a particular county.
    • Fractionin District: The Air District jurisdiction covers only a portion of Solano and Sonoma County and, therefore, an additional allocation is applied to these counties that proportions each county’s emissions that are within Air District’s boundary.
    • GWPpollutant: 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.

    This approach allows derivation of emissions data for the years 2000-2021. 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 Emissionscounty = Base Year(s) Emissioncounty x Growth Factor

    Where:

    • Growth Factor: is a scaling factor that is used to derive historical emissions estimates for years for which activity data and/or emissions are not available, and to forecast emissions for future years, using surrogates that are assumed to be representative of activity and/or emissions trends.

    More details on the county distribution of emissions, emission factors and controls are provided in the following subsections:

    County Fractions

    Both statewide employment data for the semiconductor and related device manufacturing industry (NAICS code 334413), and analogous county-specific data are available (GO-Biz, 2024) from the California Governor’s Office of Business and Economic Development. Since the CARB GHG emissions inventory dataset is statewide, the county fractions of emissions are determined using the ratios of the respective individual county employment data to the statewide data. The averages of the available employment data for 2019, 2020, and 2021, are used for the county fractions. Solano, Napa, and Marin Counties have no semiconductor manufacturing employment.

    ID Description

    BAAQMD Jurisdiction Fraction

    The Air District assumes that all semiconductor employment in Sonoma County is within the Air District jurisdiction. There is no employment for semiconductor manufacturing in Solano County.

    Emission Factors, GWP Scaling and Local Controls

    The emissions data are sourced directly from CARB’s statewide GHG inventory which is based on the GWPs from IPCC’s Fourth Assessment Report (AR4; IPCC, 2007). These emissions were subsequently adjusted using GWPs from the more recent IPCC Fifth Assessment Report (AR5; IPCC, 2014). The adjustment was applied by multiplying each pollutant’s emissions by the ratio of its GWPs from 5th report by its AR5 GWP100-year to its AR4 GWP100-year.

    No controls have been added to these emissions as the Air District does not currently regulate emissions from these sources. However, state regulation, including those mandated by California Senate Bill 1206 (SB1206, 2022) requiring the phased reduction of these gases are incorporated in CARB’s forecast from 2022 to 2050. The forecast profile is further discussed below in the Future Projections subsection.

    Historical Emissions

    Historical emissions for years 1990-1999 are calculated by extrapolating a linear regression using 2000-2004 statewide CARB emissions as the basis. The backcast emissions are set to zero for any years in which the extrapolated line cross the zero axis (simulating the emergence of ozone depleting substance substitutes in the early 1990s).

    Future Projections

    Future projection forecasts are determined using a model in CARB’s 2022 Scoping Plan developed by Energy and Environmental Economics (E3) called California PATHWAYS Non-Energy GHGs Detailed HFCs Business as Usual (BAU) Reference Scenario (CARB, 2022). This modeled growth curve does not include aspirational targets but actual adopted regulations and developed policies that are being implemented. Since CARB’s modeling analysis stops at year 2045, the year 2045 emissions are held constant for 2045-2050 for all source categories.

    Assessment of Methodology

    The general methodology for determining emissions for these categories has not changed from the previous inventories, although all of the data inputs have been updated.

    Year

    Revision

    Reference

    2022

    1. Use base years of 2000-2021 from CARB’s GHG inventory released in 2023.
    2. Assume 2022 emissions to be the same as those in 2021, from the 2022 Scoping Plan modeling for the High-GWP sector.
    3. Update county proportions using fractions from GO-Biz as described above in the County Distribution/Fraction section
    4. Backcast is based on linear regression using CARB’s GHG emissions for the years 2000-2004. Any years for which the extrapolated emissions crossed the zero axis are set to zero.
    5. Forecast is based on CARB’s 2022 Scoping Plan Update’s BAU Reference Scenario from E3’s PATHWAYS modeling for Non-Energy GHGs Other: Industry.
    6. Update Sonoma/Solano County proportions as described in the BAAQMD Jurisdiction Fraction section
    7. No local controls are implemented. Statewide controls are included in the CARB BAU reference scenario
    8. Updated the GWP using IPCC Assessment Report 5
    1. CARB, 2023
    2. CARB, 2022
    3. GO-Biz, 2024
    4. Calculation using CARB, 2023 data
    5. CARB, 2022
    6. See text
    7. SB1206, 2022
    8. IPCC, 2014

    2015

    1. Previously, the Air District used the area source category 1891 and point source category 43 (point sources from the Air District’s internal permitting database) to report fugitive emissions from semiconductor manufacturing. Pollutant-specific emissions from Semiconductor Manufacturing were not explicitly identified. Also, Semiconductor Manufacturing emissions were only accounted for if they were reported as a part of point source category 43. It is quite likely that category 43 estimates are significantly underestimated, since CARB reporting requirements are not verified and validated, and may not cover all sources in the region.
    2. No controls were implemented
    1. CARB, 2019
    2. CARB, 2019

    2011

    1. District’s ODSS inventory was split into eight species-specific categories (for example, category 1752 represented HFC-125 emissions) that combined all end-uses and sectors, so specific contribution of high-GWP gas emissions to Semiconductor Manufacturing GHG inventory was not known.
    1. BAAQMD, 2011

    Sample Calculations

    An example of calculation for CF4 emissions in units of million metric tons of CO2 equivalents (MMTCO2eq) for Santa Clara County for year 2022 (category 2625) is shown below:

    Step 1

    Obtain 2021 statewide emissions from CARB’s GHG inventory for Category 2526 which is under CARB Sector activity code 30-20-13-29-24-000 for Industrial 🡪 Semiconductor Manufacture 🡪 CF4 pollutant (CARB, 2023)

    0.116 MMTCO2eq

    Step 2

    Estimate the adjustment factor for converting GWP for CF4 from AR4 to AR5 (GWP100(AR5)/GWP100(AR4));

    GWP100(AR5) = 6630

    GWP100(AR4) = 7390

    = 6630/7390

    = 0.8972

    Step 3

    Apply ratio to adjust the statewide GHG emissions to reflect the latest GWP

    0.116 MMTCO2eq

    × 0.8972

    = 0.104 MMTCO2eq

    Step 4

    Estimate the fraction of semiconductor manufacturing employment for the Bay Area relative to the State for 2022 (GO-Biz, 2024)

    = 0.739

    Step 5

    Estimate the CF4 emissions for the Bay Area by multiplying the employment fraction by the statewide GHG emissions

    = 0.104 MMTCO2eq

    × 0.739

    = 0.077 MMTCO2eq

    Step 6

    Apply growth factor from CARB Scoping Plan E3 forecasts to estimate 2022 emissions based on CARB’s 2021 GHG emissions inventory (CARB, 2022)

    = 0.077 MMTCO2eq

    × 1

    = 0.077 MMTCO2eq

    Step 7

    Estimate the fraction of semiconductor manufacturing employment for Santa Clara county relative to the Bay Area for 2022 (GO-Biz, 2024)

    = 0.911

    Step 8

    Calculate 2022 CF4 emissions for Santa Clara by multiplying the Bay Area CF4 emissions by fraction of the semiconductor manufacturing employment associated with Santa Clara relative to the Bay Area Apply

    = 0.077 MMTCO2eq

    × 0.911

    = 0.070 MMTCO2eq

    Emissions

    The table below summarizes greenhouse gas emissions for the base year 2022 in metric tons of CO2 equivalents (MTCO2eq).

    ID Description Total

    Summary of Base Year 2022 Emissions

    The High-GWP gases from the semiconductor manufacturing subsector contribute 0.3% of emissions (0.203 MMTCO2eq) to the 2022 base year San Francisco Bay Area regional GHG emissions inventory (65.341 MMTCO2eq). The fractions of the contribution to total High-GWP Gases’ and Industrial sector’s emissions are shown in the table below:

    Sector

    Emissions (MMTCO2eq)

    Emissions (% of High-GWP Gases)

    Emissions (% of Industrial Emissions)

    Industrial – semiconductor manufacturing

    0.203

    14.63

    1.13

    The largest fraction of the gas emissions reported here from the semiconductor manufacturing subsector come from CF4 (37.9%). The species breakdown is shown in the table below:

    High-GWP Species

    Emissions (MMTCO2eq)

    Emissions (% of all semiconductor manufacturing GHG emissions)

    CF4 (PFC-14)

    0.077

    37.9

    C2F6 (PFC-116)

    0.042

    20.7

    SF6

    0.033

    16.3

    HFC+PFC

    0.024

    11.8

    C3F8 (PFC-218)

    0.017

    8.4

    NF3

    0.008

    3.9

    C4F8 (PFC-318)

    0.002

    1.0

    Contribution of Semiconductor Manufacture Emissions by Sector
    Subsector Sector Subsector GHG Emissions (MMTCO2eq) Sector GHG Emissions (MMTCO2eq) % of Sector
    Semiconductor Manufacture Industrial 0.20 17.90 1.14%

    Contribution of Semiconductor Manufacture Emissions to Regional Total
    Subsector Subsector GHG Emissions (MMTCO2eq) Regional Total GHG Emissions (MMTCO2eq) % of Regional Total
    Semiconductor Manufacture 0.20 65.68 0.31%

    Trends

    The time series chart below shows the emission trends for all high-GWP gases by source category.

    Summary of Trends

    The historical semiconductor manufacturing high-GWP source emission trend shows a linear growth in emissions for the Bay Area from almost through the 1990s to a maximum in 2020, by design, as described above. The base year data indicate variations, but there is no consistent trend. There is a dip in emissions in 2009 that is due to C2F6 and CF4 reductions that year. Indeed, globally, there was a temporary reduction of approximately 15% in CF4 emissions in 2009, perhaps because of the impact of the financial crisis on aluminum and semiconductor production (Trudinger et al., 2016). The future trend shows small reductions in emissions, much different from the large future reductions seen for refrigerants (see High-GWP Gases methodology documentation). The assumed trend for the High-GWP Gases subsector results from the regulations limiting future maximum allowable GWPs for HFCs used in refrigerants, while the assumed growth profile for the Semiconductor Manufacturing subsector is the CARB Non-Energy Other: Industrial forecast (CARB, 2022). The use of CFCs is not regulated by SB 1206 and the one HFC in this set of categories, HFC-23, is not included in the Final Regulation Order on the Prohibitions on Use of Certain Hydrofluorocarbons in Stationary Refrigeration, Chillers, Aerosols-Propellants, and Foam End-Uses Regulation (SB 1206, 2022).

    Uncertainties

    The primary sources of uncertainties in this methodology stem from the accuracy of the emissions data collected statewide and assumed leakage rates used in developing CARB’s GHG inventory (CARB, 2022). Detailed discussions regarding the uncertainties for specific high-GWP gases can be found in CARB’s published study (Gallagher et al., 2014). Uncertainty levels for inventory-based emission estimates vary from 15% to 25% for the major high-GWP gases.

    Additional uncertainty is introduced when scaling state-level inventory to produce Bay Area county specific emissions, primarily due to using scaling factors based on surrogate parameters representing each of the four sectors. The uncertainty for the semiconductor manufacturing subsector is considered less than that for the other high-GWP subsectors, since the relevant employment scaling is quite specific.

    CARB updates its GHG inventory annually, incorporating new data and occasionally revising past estimates. Therefore, the base year inventory data may change year to year. In future inventories, the current assumption that 2022 emissions are identical to those in 2021 will be replaced with actual 2022 data when CARB publishes the next inventory update.

    Contact

    Author: Sally Newman

    Reviewer: Abhinav Guha

    Last Update: 08/19/2025

    References

    BAAQMD. 2011. BAAQMD internal documentation for 2011 GHG inventory.

    BAAQMD. 2015. Bay Area Emissions Inventory Summary Report: Greenhouse Gases. Base Year 2011, Bay Area Air Quality Management District. http://www.baaqmd.gov/~/media/Files/Planning%20and%20Research/Emission%20In ventory/BY2011_GHGSummary.ashx?la=en

    CARB. 2019. Greenhouse Gas Inventory 2019 Edition: Years 2000-2017 (AR4 GWPs), California Air Resources Board. https://ww2.arb.ca.gov/ghg-inventory-archive

    CARB. 2022. 2022 Scoping Plan. 2022 Scoping Plan for Achieving Carbon Neutrality, California Air Resources Board. https://ww2.arb.ca.gov/our-work/programs/ab-32-climate-change-scoping-plan/2022-scoping-plan-documents

    CARB. 2023. Greenhouse Gas Inventory 2023 Edition: Years 2000-2021 (AR4 GWPs)

    https://ww2.arb.ca.gov/ghg-inventory-archive

    Gallagher, G.; Zhan, T.; Hsu, Y-K.; Gupta, P.; Pederson, J.; Croes, B.; Blake, D. R.; Barletta, B.; Meinardi, S.; Ashford, P.; Vetter, A.; Saba, S.; Slim, R.; Palandre, L.; Clodic, D.; Mathis, P.; Wagner, M.; Forgie, J.; Dwyer, H.; Wolf, K. 2014: High-global Warming Potential F-gas Emissions in California: Comparison of ambient-based versus inventory-based emission estimates, and implications of refined estimates. Environ Sci. Technol. 48, 1084-1093. (See Supporting Information). http://pubs.acs.org/doi/suppl/10.1021/es403447v

    Go-BIZ. 2024. California Business and Economic Development. url for Semiconductors and Microelectronics: https://business.ca.gov/industries/semiconductors-microelectronics/. url for spreadsheet data file: https://business.ca.gov/wp-content/uploads/2023/01/CHIPS-Data-Request-Industry-Data-11-2-22.xlsx

    IPCC. 2007. Forster, P., V. Ramaswamy, P. Artaxo, T. Berntsen, R. Betts, D.W. Fahey, J. Haywood, J. Lean, D.C. Lowe, G. Myhre, J. Nganga, R. Prinn, G. Raga, M. Schulz and R. Van Dorland: Changes in Atmospheric Constituents and in Radiative Forcing. In: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M.Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. 106 pp. Available here: https://www.ipcc.ch/site/assets/uploads/2018/02/ar4-wg1-chapter2-1.pdf

    IPCC. 2014. Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II, and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, R.K. Pachauri and L.A. Meyers (eds.)]. IPCC, Geneva, Switzerland, 151 pp. Available here: https://www.ipcc.ch/site/assets/uploads/2018/02/SYR_AR5_FINAL_full.pdf

    SB 1206. 2022. Senate Bill No. 1206. Skinner. Hydrofluorocarbon gases: sale or distribution. https://leginfo.legislature.ca.gov/faces/billNavClient.xhtml?bill_id=202120220SB1206; Final Regulation Order: https://ww2.arb.ca.gov/sites/default/files/barcu/regact/2020/hfc2020/frorevised.pdf

    Trudinger, C.M., P.J. Fraser, D.M. Etheridge, W.T. Sturges, M.K. Vollmer, M. Rigby, P. Martinerie, J. Muhle, D.R. Worton, P.B. Krummel, L.P. Steele, B.J. Miller, J. Laube, F.S. Mani, P.J. Rayner, C.M Harth, E. Witrant, T. Blunier, J. Schwander, S. O’Doherty, and M. Batte. 2016. Atmospheric abundance and global emissions of perfluorocarbons CF4, C2F6 and C3F8 since 1800 inferred from ice core, firn, air archive and in situ measurements. Atmos. Chem. Phys. 16, 11733-11754. doi:10.5194/acp-16-11733-2016.