Errata

This is a very draft list of “errata”. Some are confirmed; others are suspected but not confirmed one way or the other. Right now, the list is static and mostly unsorted. It probably covers only part of the full gamut of issues. It is not claimed to be a representative sample or to have complete coverage.

This list is editable! In the toolbar at the top of the page, click the edit button. You will be taken to the page source on GitHub. There, you can: make changes without commiting; preview those changes; commit changes; or submit a pull request. You can also file issues and request enhancements.

Edit button, in the toolbar at the top of every page (including this one).

Figure 10.1: Edit button, in the toolbar at the top of every page (including this one).

High Priority

Petroleum Refining Facilities

General Refinery Emissions

  • Category #10 Basic Processes. BY2011 NOx emissions are much lower than DataBank numbers would indicate. SOx and PM emissions are much more consistent.

  • Category #16 Other. CO2 emissions are large (~3 MMTCO2eq) and linearly increasing, which is inconsistent with profiles for other pollutants in this category. Given the large quantity (~ 3 MMTCO2eq), this may merit a reassessment.

  • Category #15 Flares and Blowdown Systems. CY1990-2000 estimates may merit a reassessment, or at least further explanation in the text. CY2010 CO and NOx estimates show an odd-looking peak.

  • Category #13 Cooling Towers. Roughly 60 ton/yr of PM10 appears suddenly at RY2005, then decreases by a factor of 10x over the next few years. This is traceable to just two cooling towers at the Phillips 66 refinery. The EF for one of them was immediately revised downward by a factor of 100x. (For more insights into PM emissions from cooling towers, try RY(2001:2018) %>% DB_point_source_PONSCO_emissions() %>% filter_categories("#13 Cooling Towers" = 13) %>% filter_pollutants("PM") %>% with_facility_name() %>% with_source_name() %>% unpack_source_codes() %>% pivot_table(rows = c("cat_id", "fac_name", "src_id", "src_name", "src_code")).)

Chemical Manufacturing Facilities

All Other Chemical Plants
  • Category #26 All Other Chem Mfg (Point). Almost 10,000 ton/yr of CO emissions stops abruptly at CY1997. Few other point-source categories even approach this magnitude of CO emissions. For perspective, all five refineries emit about 2,000 to 4,000 ton/yr CO from all of their sources combined (all equipment, all categories, everything). This is traceable to a single manufacturing plant, P#21 E I duPont de Nemours & Company. The facility has no recorded closure date in DataBank, but it was shut down in 1998.

Petroleum Products Evaporation and Leakage

GDF - Excess Emissions
  • Category #1197 GDF - Excess Emissions. Over 5,000 ton/yr ROG is missing for five years: CY1990-1994.1 This is a large omission. Among area-source categories, in CY1995, this category was the single largest overall emitter of ROG (excepting biogenics). In the growth profile assigned to this category #465 Projected Base On Excess Em, the growth factors for CY1990-1994 are exactly zero. The backcast flag is (currently) set to FALSE. In addition, the first applicable TOG emission factor for this category has an effective date of 1995-01-01. (For most categories, the first effective date will be 1975-01-01.)
Portable Fuel Container Spillage
  • Category #1434 Portable Fuel Container Spillage. The Bay Area total should not be equal to the statewide total. The text states that “CARB estimates 3.57 tons of uncontrolled ROG emissions per day statewide from portable fuel containers in 2011.” That 3.57 ton/day is equal to 1,304 ton/yr. In the BY2011 inventory, the estimated CY2011 total is 1,303 ton/yr ROG. (Perhaps the text is incorrect instead?)

Stationary Combustion of Fuels

Cogeneration
  • Category #292 Reciprocating Engines. Estimated CY1990-2007 TOG emissions are too low, compared to CH4. (The CH4/TOG ratio for those years is well above 1, generally in the range of 3x to 30x.) About 1,000 ton/yr TOG and 100 ton/yr ROG may be missing for those years.
Power Plants

  • Category #293 Gas Fired Boilers. The CO2/CO ratio for CY1990-2000 appears to be far too low, given what is observed for subsequent decades. This could be resulting in the absence of 5-10 MMTCO2eq from the CY1990 BAAQMD inventory. An absence of CO2 of that magnitude, in CY1990, could be substantially distorting the answer to the key question of whether the District is “on track” to reach its CY2030 target.

  • Category #294 Oil Fired Boilers. (Possibly related to above.) Emissions seem to be almost entirely absent from the BY2011 inventory. There are some very small emissions (~ 0.1 ton/yr of TOG and NOx) estimated for CY2008-2010, but that is it.

Oil Refineries - External Combustion

  • Category #301 Solid Fuel. Inventoried PM emissions increase suddenly from 3 ton/yr (until RY2008) to 300 ton/yr (RY2009 and later). This is entirely traceable to FCCU emissions at facility #10 Chevron Refinery.

  • Category #301 Solid Fuel. Inventoried SO2 emissions decrease suddenly from 1500 ton/yr (until RY2008) to 150 ton/yr (RY2009 and later). Again, this is entirely traceable to FCCU emissions at facility #10 Chevron Refinery. Some explanation is probably merited.

  • Category #299 Natural Gas & Other Gases. Inventoried TOG emissions increase by a factor of 2x beginning in RY2008. This is traceable to facilities #10 Chevron and #11 Shell, across a variety of src_codes, the largest being C1130235: CO-fueled boiler(s), 100-250 MMBTU/hr.

  • According to the text, “[p]ast base year historical emissions include interchangeable emission reduction credits (IERC) which were part of category 10. IERC banking emissions are now inventoried under categories 298-300.” It is unclear whether the BY emission estimates are of “what went into the air”, or include adjustments due to banking — and when/where/how much this distinction might matter.

Other External Combustion - Natural Gas
  • Categories #1590 and #1591 (NG Combustion, Commercial and Industrial). Throughput (cubic feet of natural gas consumed) should be double-checked against CEC data, especially for years CY1990-2000. See the example charts for Natural Gas Consumption in the Charting Annual Data cookbook. A large amount of GHG and (criteria-pollutant) emissions are at stake — several MMTCO2eq and hundreds or thousands of tons of PM, TOG, and NOx, respectively.

Mobile Combustion

Ships
  • Category #2158 (Ships in Transit, Container Ship, Main Engine, SFAB). All criteria-pollutant emissions are missing for years CY1990-2008. There are no emissions from either distillate fuel or residual fuel for these years.

Miscellaneous Other Sources

Road Dust
  • Categories #1759:1762 (Paved Roads). Emissions for the four Paved Roads categories are exactly flat. The paragraph under Growth states that “MTC’s CLM data show an increase of less than 2% for [1990 to 2011] (less than 0.1% per year), and this was used to create trends around the 2011 base year emissions.” However, the growth is not 2% per 20 years; it is exactly flat, because the relevant categories are all linked to profile #38 No Growth.

Medium (or maybe Low or High) Priority

Petroleum Refining Facilities

  • Refinery Process Vessel Depressurization: TOG emission rates for two isolated years (CY1999-2000) are above 3 ton/yr. There are no estimates before CY1999. Estimates after CY2000 are exactly zero. According to the text, “Emissions for category 20 have been relatively insignificant for last few years, at less than 5 pounds per day for the entire region.” 5 lb/day is approximately 1 ton/yr, and so should be visible on the chart. The growth profile is also inconsistent with the published emission numbers.

Chemical Manufacturing Facilities

Coatings and Inks
  • The paragraph under “Control” states that “the amended regulation has resulted in an 84% emission reduction since 1993,” but according to what’s in DataBank, the controlled fraction for category #23 seems to be 54%.
All Other Chemical Plants
  • Category #26 All Other Chem Mfg (Point). A blip of several hundred ton/yr CO occurs with a kiln at P#227 Criterion Catalysts & Technologies LP in RY2009 (possibly a unit error).

Other Industrial and Commercial Processes

Recycling and Shredding
  • PM emissions for #2223 Recycling and Shredding (Fugitive) are missing (not zero; actually missing) for years CY1990-1995. The growth profile assigned to this BY2011 category is zero for those years.
Concrete Batching
  • While the TSP estimates for the point-source component (category #39) align well with DataBank RY/PY data, the historical PM estimates exhibit differences on the order of 100 ton/yr for years CY1990-2000.
Cement Manufacturing & Combustion

  • Emissions of all pollutants for #1748 Cement Plant Combustion (Coal) are missing (not zero; actually missing) for CY2008 and CY2010-2030. However, emissions for CY2009 are present.

  • The growth profile assigned to #877 “Cement Plt. Comb. - Coal”, is also zero for the same years, and also has a spike at CY2009.

  • The text states “Projections of emissions to 2030 in all categories were based on ABAG’s 2009 Construction Employment growth profile.” But the charts show different rates of growth. None actually match ABAG2009 Construction Employment growth (+39.1%, CY2030/2011).

Rubber Manufacturing

  • Emissions of TOG and ROG classified as #46 Rubber Products Mfg (point) increase sharply at CY2011, from less than 1 ton/yr to 19 ton/yr.

  • Upon inspection with a pivot table, over 95% of this increase is attributable to the sudden appearance in the RY/PY inventory of a source S#32100 Fugitive Organic Emissions at facility P#110 Burke Industries. A constant 18.26 tons/yr TOG is attributed to this source. The facility has been in continuous operation since at least CY1990; it is implausible that no fugitive TOG or ORG was emitted from it prior to CY2011.

  • Interestingly, one can also see that RY1990 TOG and ROG emissions from P#110 S#32110 are also 18.26 ton/yr TOG, but between CY1991 and CY2010, TOG S#32100 disappears from the RY inventory.

  • The growth profile (#847) associated with this point-source category never dips below 100%, so all emissions before CY1990 should have been calculated to be 20 ton/yr or more, not zero. The backcast flag in DataBank is set to FALSE, but because this could have been changed between the generation of the CAP inventory and the BY2011 GHG inventory, it’s not solid evidence that the CAPs for this category were not backcast. Looking back in time, category #P46 can be found in the BY2008 inventory. In the BY2008 inventory, it comprises about 0.4-0.6 ton/yr TOG from CY1990 to 2030.

Asphaltic Concrete Plants
  • There is a peak in SO2 emissions at CY2010 that does not seem to match the profile of other pollutants nor the assigned growth profile (#817 Asphaltconcrplts (Modgc646)).
Oil Fields
  • For category #1575, the BY2011 emission factor for CH4, in Table 4.42, appears not to match the BY2011 Methodology PDF (page 4.21.1 - 2).
Gas Fields

  • For category #1576, the the BY2011 emission factor for CH4, in Table 4.45, appears not to match the BY2011 Methodology PDF (page 4.22.1 - 2).

  • The growth projection for #1576 (+32.7%, CY2030/2011) does not match the text (“Natural gas production in the Bay Area District is expected to grow approximately 1% per year”).

Sand Blasting
  • The paragraph under Control refers to Regulation 6 and [Rule 12-14], and states that “Controls can range up to 99% or higher.” However, no BY2011 control factors for the relevant categories (#42 or #1909) appear to be stored in DataBank. See DB_CTLFAC_CONCORDANCE and/or use BY(2011) %>% DB_control_factors() to pull the relevant CFs.
Other Miscellaneous Industrial/Commercial
  • Comparing across pollutants, emission values at CY2010 exhibit strange up/down variations. The general shapes of the rest of the years seems generally consistent both (a) across pollutants, and (b) with the assigned growth profiles.

Waste Management

POTWs

  • Projected growth for #50 POTWs (Point), at +17.8% (CY2030/2011), does not match the ABAG 2009 Total Population growth profile (+39.5%).

  • The county fraction for ALA in Table 5.1 is far too high (county population = 44%). In general, this table is inconsistent with county fractions computed directly from growth profile #657 Household Population (Table 10.45).

Land Farming

  • The projected growth (+15.8%, CY2030/2011) does not match the ABAG2009 Population growth profile.

  • The CY1990 emissions for TOG are sharply increased at CY1990. A probable cause: the emission factor for TOG changed abruptly, from CY1990 (TOG = 20 lb/ton) to CY1991 onwards (TOG = 2.24 lb/ton). Although this affects TOG, it is not reflected in published CY1990 emissions of CH4.

Composting

  • Like Other Miscellaneous Industrial/Commercial, when comparing across pollutants, emission values at CY2010 exhibit strange “peaks”.

  • Category #1936 Composting (Area):

    • County fractions for this area source are undocumented and do not match the fractions for category #1709.
    • Emission factors are time-varying (see chart) but only seem to change during a three year period, CY2010-2012. Why?

Petroleum Products Evaporation and Leakage

Industrial and Commercial Coatings: Marine
  • Categories #1930 and #1931 are missing emission data for CY2005-2008. These aren’t zeros; there are just no rows in the BY2011-annual-A.csv file, for these categories, for CY2005-2008.
Marine Loading

  • The very last paragraph under “Controls” refers to “Categories 795 - 802”, but #801 and #802 are not elsewhere mentioned, and no emissions for #801 or #802 appear in the published BY2011 inventory. (Are they unintentionally missing?).

  • Category #90 “Ballasting, Other Material” has time-varying emission factors. This is not inherently a problem, but the pattern looks remarkably like the assigned growth profile (#835 Ballasting-Other). This could be causing an unintentional “double-counting” (amplification) of the year-over-year variation. Related: the same thing happens with Farming Operations.

Non-Gasoline Terminals and Bulk Plants Storage

  • The text states that “There is no data available to project emissions for this category. Therefore, projected emissions are assumed to grow at the rate of 0.5% per year.” However, the long-term forecast for this category (#938) is -2.9% (CY2030/2011).

  • There is an apparent “cliff” in the TOG/ROG emissions where they drop from ~200 ton/yr (CY1990-2009) to ~20 ton/yr. Upon inspection of the RY/PY data, we can see that ~20 ton/yr is consistent with all the TOG values recorded in RY data for this category for many years, except for one: RY2008 (which is ~200 ton/yr). It seems plausible that either the 200 ton/yr datapoint was an artifact (which should not dictate all of CY1990-2009), or that the constant 20 ton/yr is a systematic underestimate. If true, then one or the other should be corrected.

Gasoline Truck Loading
  • There is a sudden “peak” in emissions circa CY2005. It is consistent with growth profile #589 Revised Bulk Plant Tputs but atypical compared to the rest of the profile. There is also a “seesaw” at CY2008-CY2008, dropping from ~400 ton/yr ROG to less than 1 ton/yr, then back up to ~400 ton/yr, then back down to a fairly constant ~20 ton/yr.
Sterilizer Venting
  • CY1994 emissions are exactly zero. Save for that year, estimated emissions are a constant 0.365 ton/yr ROG, despite (a) a positively sloped growth profile (CY1990-2030) and (b) steep changes in the historical control profile (CY1990-1995).
Solvent Cleaning
  • Projected long-term growth for #1245 Other Handwiping (Point), at +9.7% (CY2030/2011) is inconsistent with the text (“ABAG 2009 Manufacturing Employment”) and with #1935 Other Handwiping (Area), both of which are +28.4%.
Industrial/Commercial Coating
  • Projected long-term growth for Auto Refinishing categories, at +17.8% (CY2030/2011), is inconsistent with projected long-term growth for all other Industrial/Commercial Coating categories (+28.4%).
Vacuum Trucks
  • The paragraph under “Growth” refers to 1% growth, but the BY2011 growth profile for this category is flat.

Stationary Combustion of Fuels

Domestic Fuel, LPG & Liquid Fuel

  • For category #286 Domestic LPG, the growth is described as “follow[ing] household population data”. But the assigned growth profile is #658 Persons per Household, which is the number of persons in an average household (fairly constant), not the total population.

  • Minor: for category #287 Distillate Oil, the forecast is described as “2% reduction per year”; it seems to be about 1.7% (with rounding, OK).

Domestic Solid Fuel (Wood)

  • The CY1990-2000 growth profiles for Sonoma County are odd, compared to the profiles for other counties. This artifact is apparent in two growth profiles: #612 Modified Households - Firepl; and #613 Modified Households Woodstov.

  • The text states that “projections of emissions are based on household growth,” but the actual forecast is exactly flat (zero growth).

Power Plants

  • For category #297 Oil Fired Turbines, the CY1990-2010 decline in SO2 emissions does not appear to be explained by either the growth profile or the control profile (for this category, there are no control factors for SO2 in DataBank, only NOx).

  • For category #1595 Gas Fired Turbines, a sharp rise and fall in CO, CY2002-2009, does not match other pollutants.

  • For category #1597 Power Imports, county fractions for CC and SOL are zero.

Oil Refineries - External Combustion
  • Category #299 Natural Gas & Other Gases:
    • NOx emissions are higher than PY data would suggest, CY1990-2000.
    • Both CO and SO2 emissions are lower than PY data, CY1995-2000.
Turbines

  • Category #305 Gas Fired Turbines shows a sharp up/down tick at CY2010. If you look at an RY pivot table, you can see that this is entirely due to one source. If you look at a PY pivot table, the effect disappears — that same source is very steady, year-over-year. It’s possible that an error existed at the time that the RY snapshot was made, but was fixed before the PY data were frozen.

  • Category #305 Gas Fired Turbines shows variations in SO2 emissions, CY1995-2000, that are not explained by an SO2 control profile (there isn’t one) and are not consistent with variations in other uncontrolled pollutants (e.g. TOG). They are, however, fairly consistent with RY and PY data. This may be OK.

Other External Combustion - Natural Gas
  • Category #307 NG Combustion (Point, Misc): Growth profile #631 Gh, Thrpt Based does not match the values found in RY2006 and RY2007, which results in an artificial “peak”. This is noticeable for all pollutants, but especially uncontrolled pollutants. PM and TOG emissions for CY1993-2000 are also under-estimated by this growth profile. Time-varying emission factors would be a way to solve this.
Other External Combustion - All Other Fuels
  • For category #312 Other Fuels, the ratio (CH4 + ROG) / TOG is about 1.5 for CY2000-2030. This ratio should never be more than 1/1. The inventoried estimates are a few hundred tons per year of each of these pollutants (CH4, ROG, and TOG).

Waste Burning

Incineration

  • Category #1579 Incineration (Area) has a large and linear decline from CY1990-1995 that is unexplained. This may be attributable to BAAQMD regulations.

  • This category has a CO/CO2 ratio of about 0.003, comparable to a well-maintained boiler. Typical CO/CO2 ratios for garbage and vegetation burning (“open burning”) may be an order of magnitude higher. See, for example, Akagi (2011). (NOTE: in the BY2011 inventory, Category #1579 is associated with CO2_bio, rather than CO2.)

Planned Agricultural Burning
  • Looking at category #316 Field Crops, CY2001-2008, there is a substantial “surge” in inventoried emissions. This apparent surge is on the order of hundreds of tons of most criteria pollutants. If there is reason to believe that this really happened in the world, it’s probably worth an explanation. It appears to be reflected in — or attributable to — DataBank growth profiles #726 and #727.

Mobile Combustion

Recreational Boats
  • Emissions of many pollutants drop to exactly zero at five-year intervals (CY2020, 2025, 2030, …, etc.), and then return to the previous trend.
Lawn & Garden Equipment
  • N2O emissions for a five-year interval (CY2026-2030) are exactly zero.
Transportation Refrigeration Units (TRUs)
Agricultural Equipment
  • Like TRUs, N2O emissions for a five-year interval (CY2026-2030) are exactly zero. Also, #1646 Ag Equip (Diesel) is linked to growth profile #464 Commercial Boats. (?)
Construction Equipment
  • SO2 emissions appear spiky, particularly between CY1998-2010.
Industrial Equipment

  • TOG emissions for #1667 (Natural Gas) exhibit the same zero-every-5-years artifact that is present in Recreational Boats, only this time it is isolated to one pollutant (TOG).

  • SO2 from #1666 (Diesel) grows much more rapidly than any other pollutant.

Light Commerical Equipment

  • CO2 and CH4 for #1672 (Diesel) drop to exactly zero at CY2010, then resume the previous trend.

  • N2O for #1668 (Gasoline, 4-stroke) is exactly zero for CY2026-2030 (same as rec boats, TRUs, ag equip, etc.).

Off Road Motorcycles & All Terrain Vehicles
  • Emissions for several sub-categories exhibit odd behavior at 5-year intervals, similar to the above.

Miscellaneous Other Sources

Farming Operations
  • Emission factors for both of these two categories (#749 and #1435) vary by time in a way that strongly resembles the growth profiles. If this is accidental, it could be causing a “double-counting” of the magnitude of the decrease from CY1990-2010. If it is intentional, it probably warrants some extra documentation.
Animal Waste, Livestock
Animal Waste, Other
  • County fractions (Table 10.26 are inconsistent with the text.
Sanitary Sewers
  • County fractions in Table 10.28 are inconsistent with those computed directly from growth profile #657 Household Population (10.45). Perhaps this is due to the use of a subpopulation or another data source? Or a problem with profile #657?
Accidental Structural Fires
  • The BY2011 emission factors in Table 10.7 appear not to match the relevant table in the the BY2011 Methodology PDF (page 10.3.1 - 2).
Accidental Auto Fires
Accidental Vegetation Fires
  • The BY2011 throughputs (burned acreage), in Table 10.10, appear not to match the relevant table in the BY2011 Methodology PDF (page 10.4.1 - 2).
Pesticides
  • Emission factors for categories #1190 and #1191 vary by time. This could be OK, but it is unusual to see in the BY2011 inventory.
Consumer Products
  • The CY2011 TOG emissions listed in Table 10.38 appear slightly different than the values on page 10.14.1-1 of the BY2011 Methodology PDF.

Low Priority

Petroleum Products Evaporation and Leakage

Asphalt Paving
  • The text states “organic emission factors used for hot-mix, emulsified, and road oils were 0.04, 17.9, and 2.19 lb/ton of asphalt applied, respectively.” There are control factors in DataBank for #95 Road Oil (97.5% controlled, CY1990-2030) and #97 Emulsified (55.3% controlled, CY1990-2030), but nothing for #96 Hot Mix. If we apply those (DataBank) CFs to the (DataBank) EFs listed in Table 6.38, then we do in fact arrive at the EFs mentioned in the text (#95 = 2.2 lb/ton and #97 Emulsified). So the math works out, but it is odd to have a mixture of controlled (for #96) and uncontrolled (for #95 and #97) EFs stored in DataBank.
Industrial/Commercial Coating

Stationary Combustion of Fuels

Back-up Diesel Generator
  • Category #1598 should be (re)assessed.

Mobile Combustion

Ships
  • BY2011 categories are actually #2041:2220, not #1787:1887.

  1. No relevant rows are present in the file BY2011-annual-A.csv; hence the NA(s).↩