Vulcan

 Science  Methods  Documentation,  Version  2.0  

Kevin  Robert  Gurney,    
Arizona  State  University/Purdue  University  

Daniel  Mendoza,  Sarath  Geethakumar,  Yuyu  Zhou,  Vandhana  Chandrasekaran,  Chris  

Miller,  Advait  Godbole,  Nalin  Sahni,  Broc  Seib,  William  Ansley,  Sullivan  Peraino,  
Xueyao  Chen,  Utkarsh  Maloo,  Jonghun  Kam,  Jaymee  Binion    
Purdue  University    

Marc  Fischer,  Stephane  de  la  Rue  du  Can  

Lawrence  Berkeley  National  Laboratory  

1
  
Table  of  Contents  

1.0  Vulcan  data  source  introduction .............................................................................................. 3  

2.0  NEI  Point  CO2  Emissions.............................................................................................................. 7  
2.1  Data  reduction .......................................................................................................................... 8  
2.1.1  Material  and  pollutant  qualifiers ........................................................................... 8  
2.1.2  Time  period  consistency ........................................................................................... 9  
2.1.3  Missing  material  identification............................................................................... 10  
2.1.4  Idiosyncratic  adjustments ........................................................................................ 11  
2.2  Quantifying  CO2  Emissions .................................................................................................. 12  
2.2.1  CO  Emission  factor  retrieval .................................................................................... 13  
2.2.2  CO2  emissions  estimation.......................................................................................... 14  
2.3  Sources  of  uncertainty........................................................................................................... 14  
2.3.1  Pollutant  emission  factor   ......................................................................................... 15  
2.3.2  Heat  and  carbon  content   .......................................................................................... 15  
2.3.3  Utilizing  on  default  pollutant  Efs   .......................................................................... 16  
2.3.4  Summary  of  sensitivities   .......................................................................................... 16  
3.0  Cement  Production  CO2  Emissions ......................................................................................... 18  
4.0  Electricity  Production  CO2  Emissions .................................................................................... 21  
4.1  ETS/CEM  data   .......................................................................................................................... 21  
4.2  Cross-­‐matching  to  NEI  .......................................................................................................... 21  
4.3  Fuel  assignment   ...................................................................................................................... 22  
4.4  Sources  of  uncertainty   .......................................................................................................... 22  
5.0  NEI  Nonpoint  CO2  Emissions ..................................................................................................... 25  
5.1  Data  reduction .......................................................................................................................... 25  
5.1.1  Material  and  pollutant  qualifiers ........................................................................... 25  
5.1.2  Time  period  consistency ........................................................................................... 25  
5.2  Quantifying  CO2  emissions .................................................................................................. 27  
5.2.1  CO  emission  factor  retrieval .................................................................................... 27  
5.2.2  CO2  emissions  estimation.......................................................................................... 27  
5.2.3  Suspected  database  errors  and  corrections...................................................... 28  
5.3  Spatial  processing.................................................................................................................... 29  

2
 5.4  Sources  of  uncertainty   .......................................................................................................... 30  
5.4.1  Pollutant  emission  factor   ......................................................................................... 30  
5.4.2  Heat  and  carbon  content   .......................................................................................... 31  
5.4.3  Utilizing  only  default  pollutant  Efs ....................................................................... 31  
5.4.4  Summary  of  sensitivities   .......................................................................................... 32  
6.0  Transport  CO2  Emissions............................................................................................................. 33  
6.1  Onroad  sources......................................................................................................................... 33  
6.1.1  Vehicle  miles  traveled................................................................................................. 33  
6.1.2  CO2  emission  factors.................................................................................................... 37  
6.1.3  Time  structure   .............................................................................................................. 38  
6.1.3a  Traffic  data  records ........................................................................................ 38  
6.1.3b  Data  conditioning  and  gap  filling   ............................................................ 41  
6.1.3c  Application  of  ATR  data  ............................................................................... 43  
6.1.4  Spatial  rendering .......................................................................................................... 46  
6.1.4a  Roadway  rendering............................................................................................ 46  
6.1.4b  Rendering  to  regular  grid................................................................................ 46  
6.2  Nonroad  Mobile  Emissions.................................................................................................. 47  
6.3  Aircraft  emissions.................................................................................................................... 48  
6.4  Sources  of  uncertainty........................................................................................................... 50  
7.0  Sectoral  Assignment  and  Visualization ................................................................................. 51  
8.0  Temporal  Processing..................................................................................................................... 53  
8.1  Monthly  downscaling............................................................................................................. 53  
8.2  Sub-­‐monthly  downscaling ................................................................................................... 55  
8.3  Multiyear  time  structure ...................................................................................................... 57  
Reference ................................................................................................................................................... 61  
Appendix  A ................................................................................................................................................ 67  
Appendix  B   ............................................................................................................................................... 74

3
  
1.0  Vulcan  data  source  introduction  
The  Vulcan  United  States  fossil  fuel  CO2  emissions  inventory  is  constructed  from  five  
primary  datasets,  constituting  eight  data  types,  with  additional  data  used  to  shape  
the  space/time  distribution.  Figure  1.1  shows  a  schematic  of  the  data  sources  and  
how  they  are  processed  to  produce  CO2  emissions.    

 
Figure  1.1.  Vulcan  data  sources  and  processing  overview  

The  eight  data  types  can  be  succinctly  described  as  follows:  
o Point  sources:  non  electricity-­‐producing  sources  identified  as  a  specific  
geocoded  location  
o Non-­‐point  sources:  county-­‐level  aggregation  of  non-­‐geocoded  sources  
o Non-­‐road  sources:  mobile  surface  sources  that  do  not  travel  on  roadways  
such  as  boats,  trains,  snowmobiles,  etc.  
o Onroad  sources:  mobile  road-­‐based  sources  such  as  automobiles,  buses,  and  
motorcycles  
o Airport:  geolocated  sources  associated  with  taxi,  takeoff,  and  landing  cycles  
associated  with  air  travel  
o Aircraft:  gridded  sources  associated  with  the  airborne  component  of  air  
travel.  
o Electricity  Production:  geolocated  sources  associated  with  the  production  of  
electricity  
o Cement:  geolocated  sources  associated  with  cement  production  (non  fuel-­‐
based  emissions)  
The  point,  non-­‐point,  noroad,  and  airport  emission  data  files  come  from  the  
Environmental  Protection  Agency’s  (EPA)  National  Emissions  Inventory  (NEI)  for  

4
the  year  2002  which  is  a  comprehensive  inventory  of  all  criteria  air  pollutants  
(CAPs)  and  hazardous  air  pollutants  (HAPs)  across  the  United  States  [USEPA  
2005a].  
The  NEI  is  a  data  structure  with  which  the  EPA  can  meet  mandates  established  by  
the  Clean  Air  Act  (CAA)  pertaining  to  CAPs  and  HAPs.  The  CAPs  emissions,  the  
component  of  emissions  used  by  the  Vulcan  system,  are  collected  under  the  
Consolidated  Emissions  Reporting  Rule  (40  CFR  Part  51)  [USEPA  2002].  The  NEI  can  
be  used  to  track  progress,  drive  air  quality  modeling,  enable  emissions  trading,  and  
ensure  comprehensive  reporting  and  compliance.  
The  emissions  data  within  the  NEI  are  collected  from  state  and  local  agencies  and  
tribes  (S/L/T)  in  addition  to  other  data  sources  from  the  Department  of  Energy’s  
(DOE)  Energy  Information  Administration  (EIA)  and  EPAs  Clean  Air  Markets  
Division  (CAMD)  [DOE/EIA  2003;  ERG  and  EHP,  2004;  USEPA  2004a;  USEPA  2005b].  
All  of  this  data  is  inventoried  by  the  EPA  and  QA/QC  operations  are  performed  
before  releasing  the  data  as  the  NEI  [USEPA  2005c].  Currently,  the  Vulcan  system  
has  utilized  data  from  the  2002  NEI  and  this  forms  the  basis  of  much  of  the  2002  
CO2  Vulcan  inventory.  
The  NEI  database  is  composed  of  a  series  of  individual,  but  related,  data  files.  These  
data  files  share  common,  required  key  fields.  The  Vulcan  inventory  construction  
utilized  a  subset  of  these  database  fields  in  combination  with  other  data  streams  to  
produce  CO2  emissions.  
The  ETS/CEMs  data  is  collected  under  the  Acid  Rain  Program  (ARP),  which  was  
instituted  in  1990  under  Title  IV  of  the  Clean  Air  Act.  The  ARP  regulates  electrical  
generating  units  (EGUs)  that  burn  fossil  fuel  and  are  greater  than  25  MW  capacity  or  
are  less  than  25  MW  but  which  burn  coal  with  a  sulfur  content  of  greater  than  
0.05%  by  weight.  Covering  95%  of  CO2  emissions  from  the  electricity  production  
sector,  this  data  source  supplies  CO2  emissions  directly  and  is  either  directly  
measured  CO2  or  calculated  from  fuel  consumption  measurements  and  fuel  carbon  
content.  
The  Aero2k  dataset  supplies  the  other  component  of  aircraft  emissions,  that  
associated  with  airborne  emissions  (above  3000  ft).  The  Aero2K  database  quantifies  
CO2  emissions  (among  other  pollutants)  on  a  1°  x  1°  x  500  ft  grid  and  is  
incorporated  directly  into  the  Vulcan  inventory.  
The  National  Mobile  Inventory  Model  (NMIM)  County  Database  (NCD)  supplies  
vehicle  miles  traveled  (VMT)  data  for  each  combination  of  vehicle  type,  road  type,  
county,  and  month.  The  NMIM  NCD  is  part  of  the  NMIM  software  package  produced  
by  the  EPA.  This  is  combined  with  fleet  information,  vehicle  emission  factors,  and  a  
GIS  road  atlas  in  order  to  locate  emissions  as  roadway  line  sources  according  to  
vehicle,  road,  county,  and  month.  
Non-­‐fuel  combustion  cement  emissions  are  derived  from  individual  reported  
cement  facility  capacity  and  state  or  state-­‐aggregate  capacity  factors.  Geolocation  
was  accomplished  by  matching  postal  addresses  to  facility  locations  in  Google  Earth.  

5
The  Vulcan  effort  does  not  attempt  to  further  QA/QC  these  large  data  sources  and  
their  related  datasets  but  incorporates  this  data  at  “face  value”  with  exceptions  
noted  in  this  documentation.  Details  of  the  EPA  QA/QC  procedures  and  potential  
uncertainties  in  that  process  can  be  found  in  EPA  NEI  documentation  and  websites.    
Further  details  on  all  of  these  data  sources  and  their  incoporation  into  the  Vulcan  
inventory  is  provided  in  the  individual  document  chapters.  

6
2.0  NEI  Point  CO2  Emissions  
The  NEI  point  database  is  comprised  of  eight  related  files  described  in  Figure  2.1  
[USEPA  2006a;  ERG  2001a].  The  three  key  fields  that  define  a  “site”  in  the  point  
database  are  the  “state  and  county  FIPS”  code  (which  identifies  the  state  and  
county),  the  “state  facility  identifier”  (which  identifies  the  individual  emitting  
facility)  and  the  tribal  code  (used  in  place  of  a  state  and  county  FIPS  in  tribal  lands).    
 

 
Figure  2.1.  The  NEI  data  relationships1  
 
The  general  procedure  followed  to  generate  CO2  emissions  from  the  point  NEI  data  
is  to  utilize  the  existing  reporting  of  CO  emissions  at  the  facility  level.  As  depicted  in  
Figure  2.1  (with  the  correction  noted  in  the  figure  footnote),  each  site  or  facility  can  
have  multiple  emission  units  (different  buildings  or  portions  of  a  complex  facility  or  
site),  each  of  which  can  have  multiple  emission  processes  (eg.  energy  production,  
heaters,  kilns),  each  process  can  emit  more  than  one  pollutant  (toxics,  NOx,  CO,  etc),  
and  these  pollutants  can  be  emitted  by  more  than  one  stack  location.  
Where  CO  emissions  are  reported,  and  an  emission  factor  can  be  assigned,  CO  
emissions  are  relied  upon.  Where  data  on  CO  is  nonexistent  or  significantly  limited,  
NOx  emissions  are  used  –  though  this  occurs  in  a  very  limited  number  of  cases.    

1  This  figure,  reproduced  from  NEI  documentation  incorrectly  identifies  the  files  in  the  box  on  the  

lefthand  side.  The  database  labeled  “EP”  is  the  “Emissions  Process”,  the  database  labeled  “PE”  is  the  
“Emissions  Period”.    

7
The  NEI  point  source  data  files  are  primarily  comprised  of  processes  associated  
with  the  industrial  sector  (identifiers  are  supplied  in  the  NEI)  but  emissions  from  
residential,  commercial  and  mobile  sources  are  found  within  the  point  data2.  This  
sectoral  designation  is  important  when  representing  the  resulting  emissions  
spatially  and  categorically,  an  issue  that  is  discussed  in  section  7.0.  
Fossil  fuel  is  calculated  with  CO/NOx  emission  factors  and  CO2  emission  factors  are  
then  applied  to  these  throughput  values.  Details  of  this  process  are  as  follows:  
2.1  Data  reduction  
Because  the  NEI  contains  a  significant  amount  of  information  on  emission  processes  
that  do  not  consume  fossil  fuels  or  processes  that  contain  emissions  from  fossil  fuel  
combustion  other  than  NOx  and  CO,  the  first  step  in  utilizing  the  NEI  point  data  is  to  
reduce  the  data  to  the  subset  relevant  to  the  CO2  emissions  problem.  A  series  of  
reductions  are  made  to  the  original  NEI  point  dataset.  
2.1.1  Material  and  pollutant  qualifiers  
The  point  source  NEI  was  first  reduced  by  narrowing  the  database  through  
examination  of  the  emission  process  material/fuel  and  how  that  material/fuel  was  
utilized  in  the  emission  process  considered.  Only  records  that  had  the  following  
combination  were  considered  for  CO2  analysis:  
1)  the  pollutant  code  identified  either  CO  or  NOx    
      AND  
2)  the  material  code  (“Mat  code”)  could  be  matched  to  a  member  of  the  Vulcan  
fossil  fuel  list  (Table  2.1)  or  was  listed  as  “null”    
      AND    
3)  the  material  input/output  (“Mat  IO”)  identifier  was  set  to  “input”  (“I”)  or  
“null”  
The  goal  was  to  limit  the  processes  considered  to  those  producing  CO  or  NOx  (the  
cornerstone  to  generating  CO2  emissions  in  the  majority  of  the  Vulcan  inventory),  
burning  fossil  fuel  (as  opposed  to  processes  consuming  biotic  materials  or  
producing  fossil  fuels).  Consideration  of  the  “null”  entries  (which  were  ambiguous  
and  therefore  deemed  worthy  of  further  investigation)  is  made  later  on  in  the  data  
reduction.  
Though  throughput  information  (eg.  tons  of  coal  burned)  was  sometimes  included  
in  these  instances,  the  throughput  values  were  not  quality  controlled  by  the  EPA  and  
were  often  found  to  be  inconsistent  with  emissions.  

2
There are some records for which no sectoral assignment could be determined. However, these
occurences were isolated to the nonpoint data pipeline.

8
  

Table  2.1.  Material/fuel  and  phase  for  fossil  fuel  burning  processes  in  the  2002  NEI    
Material Phase Material Phase
Anthracite Culm Solid Jet A Fuel Liquid
Anthracite Solid Jet Fuel Liquid
Bituminous Coal Solid Jet Kerosene Liquid
Bituminous/Subbituminous Coal Solid Jet Naphtha Liquid
Butane Gas Kerosene Liquid
Coal Solid Lignite Solid
Coke Solid Liquified Petroleum Gas (LPG) Liquid
Coke Oven Gas Gas Lube Oil Liquid
Coke Oven or Blast Furnace Gas Gas Natural Gas Gas
Crude Oil Liquid Oil Liquid
Diesel Liquid Process Gas Gas
Diesel/Kerosene Liquid Propane Gas
Distillate Liquid Propane/Butane Gas
Distillate Oil Liquid Raw Coke Solid
Distillate Oil (Diesel) Liquid Refined Oil Liquid
Distillate Oil (No. 1 & 2) Liquid Refinery Gas Gas
Distillate Oil (No. 1) Liquid Residual Oil Liquid
Distillate Oil (No. 2) Liquid Residual Oil (No. 5) Liquid
Distillate Oil (No. 4) Liquid Residual Oil (No. 6) Liquid
Ethane Gas Residual/Crude Oil Liquid
Gas Gas Sour Gas Gas
Gasoline Liquid Subbituminous Coal Solid
1
Heat TBD Waste Oil Liquid
✝
 records  with  material  identified  as  heat  are  further  explored  for  physical  fuel  consumed  via  the  SCC  description.  

The  next  reduction  step  was  to  identify  only  those  processes  which  had  either  a  
non-­‐zero  NOx  or  CO  emissions  value  (or  both).  Fuel  throughput  and  CO2  emissions  
cannot  be  generated  without  one  or  the  other  of  these  two  pollutants  as  non-­‐zero  
values.  This  reduced  the  database  to  132,971  processes3.  65  processes  had  an  
unidentifiable  code  for  the  state  and  county  location  (the  “FIPS”  code),  further  
reducing  this  set  to  132,906  processes.  Of  the  132,971  processes,  XXXX  rely  on  NOx  
emissions  for  further  processing.  
2.1.2  Time  period  consistency  
Emissions  reporting  in  the  NEI  is  made  for  a  small  set  of  different  reporting  periods  
or  time  “types”  as  follows:  
o Type  27:  average  weekday  
o Type  28:  average  weekend  day  
o Type  29:  average  day  in  period  
o Type  30:  entire  period  total  
A  given  process  can  report  emissions  for  more  than  one  of  these  time  period  types.  
Only  processes  which  identify  time  type  30  are  retained  and  all  others  are  

3  If  an  emission  process  utilizes  emission  controls  and  those  controls  fully  eliminate  CO/NOx,  the  CO
2  
from  that  process  is  NOT  captured  in  the  Vulcan  inventory.  

9
removed.4    In  most  cases  the  time  type  30  is  a  complete  calendar  year  total  amount.  
These  annual  emissions  are  initially  divided  equally  amongst  the  total  number  of  
days  and  hours  in  the  year  (for  the  gridded  hourly  output).  Section  8.0  describes  
further  temporal  conditioning  of  the  point  emissions.  Most  facilities  with  emission  
time  type  30  estimate  the  emissions  for  a  period  of  365  days  or  8760  hours  per  year.  
However,  certain  facilities  report  timespans  for  a  specific  portion  of  the  year  making  
the  effective  operational  number  of  days  in  the  year  less  than  365.  In  such  cases,  the  
annual  emissions  reported  by  the  facility  are  equally  divided  amongst  the  reported  
number  of  days/hours  rather  than  365  days  (8760  hours).5    
Hence,  the  effective  calculation  is  as  follows:  

  (2.1)  

Where  E  is  emissions,  t  is  hourly  timestep,  p  is  the  reported  emissions  period,  and  
Δtp  is  the  number  of  days  in  the  reported  time  period  (most  commonly  365).    
There  are  also  cases  in  the  input  NEI  dataset  where  the  operational  start/end  date  
of  a  process  is  reported  as  a  year  other  than  2002.  These  are  a  mixture  of  typos  by  
the  reporting  agency  or  examples  where  a  previous  year  emissions  have  been  
“carried  over”  to  the  2002  database.  Such  records  are  modified  to  start  on  1/1/2002  
and  end  on  12/31/2002.  
After  removal  of  the  non-­‐30  time  types  (23,578  processes),  we  then  have  
109,328processes  remaining  in  the  database.  
2.1.3  Missing  material  identification  
In  order  to  explore  emission  processes  for  which  the  fuel  or  input/output  identifier  
was  listed  as  “null”,  the  NEI  input  format  (NIF)  source  classification  code  (SCC)  
lookup  table  was  used  to  fill  in  the  missing  information  and  confirm  the  material  
classifications  provided  by  the  NEI  material  code.6  This  exercise  further  identified  
how  the  material  was  used  in  the  emitting  process.  For  materials  listed  in  Table  2.1,  
only  actions  identified  as  “burned”  were  retained  in  the  Vulcan  point  inventory.  
Other  actions  such  as  “processed”,  “shipped”,  or  “produced”  were  not  considered  the  
purview  of  the  Vulcan  CO2  inventory  and  these  emitting  processes  were  removed.  
There  were  two  categories  of  emission  processes  that  did  not  meet  these  criteria  
and  the  most  common  were  as  follows:  
1) fugitive  emissions  (surface  oxidation)  from  fossil  fuel  throughput  (leakage  
from  pipelines,  spills,  etc);  

4
Version 2.0 of the Vulcan inventory will utilize the multiple time types to further structure emissions
during the emitting period.
5
However, as noted in Section 8.0, the emissions are forced to be constant for the year prior to performing
monthly and hourly downscaling.
6  Material  codes  are  actually  supplied  in  multiple  fields  in  the  NEI  which  are  often  contradictory.  The  

material  codes  are  associated  with  each  pollutant  field  in  addition  to  provided  as  an  independent  
field.  The  materials  identified  through  the  SCC  lookup  are  used  to  override  all  other  material  
classifications  and  form  the  basis  of  the  fuel  combusted.  

10
 2) emissions  based  on  the  production  of  a  material/fuel  other  than  those  
identified  in  Table  2.1.  For  example,  a  process  that  had  CO/NOx  emissions,  is  
using  natural  gas,  but  the  reported  NOx/CO  emissions  are  relative  to  the  
amount  of  ammonia  produced  rather  than  the  natural  gas  burned.  Without  
knowledge  regarding  how  much  fuel  is  burned  to  produce  ammonia  (in  this  
example),  a  reliable  estimate  of  throughput  cannot  be  calculated.  It  is  also  
unclear  whether  or  not  the  NOx/CO  emissions  are  indeed  related  to  the  fossil  
fuel  combustion  or  independently  related  to  the  production  of  the  non-­‐fossil  
material.  In  the  latter  case,  the  NOx/CO  emissions  related  to  the  fossil  fuel  
combustion  are  reported  elsewhere  and  hence,  included;  double-­‐counting  
would  be  the  result  of  including  emissions  for  the  non-­‐fossil  material.  In  the  
case  of  the  former  situation,  the  total  CO2  emissions  would  be  underreported  
via  these  instances  since  these  processes  are  removed  from  further  
consideration;  
15,996  processes  were  eliminated  at  this  step  as  they  had  no  information  by  which  a  
material  could  be  identified  or  were  not  burning  a  material  listed  in  Table  2.1.    
Elimination  of  these  processes  left  85,402  emission  processes.    
2.1.4  Idiosyncratic  adjustments  
A  series  of  individual  adjustments  were  made  to  the  NEI  point  data  due  to  
independent  data  or  instances  of  QA/QC  we  were  able  to  perform  on  the  NEI  
database.    The  following  lists  these  idiosyncratic  adjustments:  
1. Identification  of  a  typo  for  FIPS  13153,  state  facility  ID  15300003,  SCC  
39000201.  CO  emissions  were  listed  in  the  NEI  point  data  as  4128  tons.    
Emissions  should  be  28  tons  CO.  
2. Two  occurrences  of  FIPS  51019,  state  facility  ID  3,  SCC  39000189  and  CO  
emissions  of  3964.41  and  2098.06  tons.  The  NEI-­‐provided  emission  factor  
(221  lbs/ton  or  9.2  lbs/106BTU)  should  be  used  instead  of  the  FIRE-­‐supplied  
emission  factor.  
3. Three  occurrences  of  FIPS  13103,  state  facility  ID  10300007,  SCC  10200802  
and  CO  emissions  of  1018,  913.2,  and  8017  tons.  The  NEI-­‐provided  emisions  
factor  (18  lbs/ton  or  0.6  lbs/106  Btu)  should  be  used  instead  of  the  FIRE-­‐
supplied  emission  factor.  
4. One  occurrence  of  FIPS  5063,  state  facility  ID  506300036,  SCC  10200101  and  
CO  emissions  of  1683.7  tons.  This  should  utilize  an  emission  factor  of  90  
lbs/ton  (or  3.744  lbs/106  Btu).  
5. Two  occurrences  of  FIPS  40123,  state  facility  ID  826  and  SCC  39000201.  CO  
emissions  were  listed  in  the  NEI  point  data  as  381  and  373.8  tons.  Emissions  
should  be  81  and  73.8  tons,  respectively  (this  is  a  typo).  

11
 6. All  occurrences  of  SCC  102000704  and  39000701  are  assigned  a  material  
code  of  809  which  corresponds  to  “coke  oven  gas  or  blast  furnace  gas”  (see  
Table  2.1).7  
7. All  occurrences  of  SCC  102000707,  39000702,  and  39000789  are  assigned  a  
material  code  of  425  which  corresponds  to  “coke  oven  gas”  (see  Table  2.1).8  
8. Ten  SCCs  were  present  in  the  NEI  point  database  but  not  found  in  the  NIF  
SCC  lookup  table.  Four  of  these  SCCs  were  considered  viable  emission  
processes  via  the  SCC  description  text  supplied  in  the  NEI  point  database  (a  
fossil  fuel  was  burned  in  the  process).9  The  four  SCC  are:  
o 20100301:  Internal  Combustion  Engines;  Electric  Generation;  Gasified  
Coal;  Turbine  
o 10100818:  External  Combustion  Boilers;  Electric  Generation;  
Petroleum  Coke;  Circulating  Fluidized  Bed  Combustion  
o 30701415:  Industrial  Processes;  Pulp  and  Paper  and  Wood  Products;  
Hardboard  (HB)  Manufacture;  "Tube  dryer,  direct  NG-­‐fired,  blowline  
blend,  PF  resin,  hardwood  
o 10102018:  External  Combustion  Boilers;  Electric  Generation;  Waste  
Coal;  Circulating  Fluidized  Bed  Combustion  
9)  The  emission  factors  for  the  Hansen  Permanente  Plant  (facility  id:  43130317)  
in  Santa  Clara  county,    CA  (FIPS:  6085)  had  two  processes  (SCCs:  39000899,  
39000201)  for  which  we  will  not  reject  the  supplied  emission  factors  even  
though  they  are  outside  the  stated  bounds.  They  do  not  supply  units  but  we  
are  confident  that  they  are  lbs  CO/ton.  
10)  SCC:  39000899  (coke  combustion)  will  utilize  a  CO  emission  factor  of  0.220  
lbs  CO/106  Btu  instead  of  the  default  value  of  0.021  lbs  CO/106  BTU.  This  
emission  factor  was  found  as  an  NEI  provided  EF  in  a  few  cases  and  appears  
more  consistent  with  anticipated  results.  
11)  for  plant  id:  1191680  and  SCC:  10300603  in  Middlesex,  MA  (FIPS:  25017),  
the  CO  emissions  were  incorrectly  reported  as  tons  (as  4900  tons)  and  
should  have  been  reported  as  lbs  (which  results  in  2.45  tons  CO/year).  
12)  All  cases  of  SCC  39000201  will  utilize  the  CO  EF  identified  in  point  9):  1.427  
lbs  CO/106  Btu.    
2.2  Quantifying  CO2  emissions  
With  the  data  reduction  complete,  each  process  is  examined  in  order  to  retrieve  
information  by  which  an  amount  of  emitted  CO2  can  be  produced.  The  CO2  emission  
quantity  is  determined  from  the  provided  CO  and/or  NOx  emissions  amount  in  
combination  with  an  emission  factor  (EF)  for  one  or  both  of  these  pollutants  and  an  

7  These  processes  are  common  in  steel  production  and  were  assigned  a  material  type  “process  gas”.  

Personal  communication  with  Indiana  State  Environmental  officials  provided  the  more  specific  fuel  
type  (and  a  more  accurate  emission  factor).  In  addition  to  Indiana,  Pennsylvania  and  Illinois  report  
these  SCCs.    
8  See  previous  footnote.  
9
The material type was identfied through examination of the CO and NOx material codes.

12
emission  factor  for  CO2.  The  CO/NOx  EF  used  is  chosen  from  three  different  
alternatives:  1)  the  EF  provided  in  the  NEI  data  itself  for  the  particular  process  in  
question  and  for  the  particular  pollutant  (CO  or  NOx),  2)  the  EF  retrieved  from  the  
FIRE  database,  a  collection  of  standard  EFs  applied  to  specific  SCC/control  
combinations  [USEPA  1997;  USEPA  2006b;  WebFIRE  2005],  and  3)  a  default  EF  
value  (provided  in  Appendix  A,  Tables  A.1  and  A.2).  
The  basic  process  by  which  CO2  emissions  are  created  is  as  follows:  
 

  (2.2)  
where  C,  is  the  emitted  amount  of  carbon,  PE  is  the  equivalent  amount  of  
uncontrolled  criteria  pollutant  emissions  (CO  or  NOx  emissions),  p  is  the  
combustion  process  (e.g.  industrial  10  MMBTU  boiler,  industrial  gasoline  
reciprocating  turbine),  f  is  the  fuel  type  (e.g.  natural  gas  or  bituminous  coal),  PF  is  
the  emission  factor  associated  with  the  criteria  pollutant,  and  CF  is  the  emission  
factor  associated  with  CO2  (provided  in  Appendix  A,  Table  A.3).    
When  CO  emissions  are  available,  these  are  used  to  generate  the  fuel  consumed  (and  
hence,  CO2  emissions)  because  the  question  of  emission  control  is  of  a  lesser  
concern  with  CO  as  it  is  with  NOx  emissions.    
2.2.1  CO  emission  factor  retrieval  
The  following  series  of  logical  steps  trace  the  procedure  for  retrieving  the  most  
reliable  CO  and  NOx  emission  factors  (PF)  for  each  process  retained  in  the  Vulcan  
system.  In  each  case,  the  retrieval  of  an  emission  factor  is  based  on  the  process  
under  consideration  and  the  material  processed.  The  procedure  is  determined  by  
the  SCC  provided  in  the  NEI  point  database  and  the  material  as  determined  in  
previous  steps  (see  section  2.1.3).  
Where  emission  factors  are  supplied  in  physical  units  (emitted  amount  per  volume  
or  mass  of  fuel),  they  are  converted  to  thermal  units  (emitted  amount  per  106BTU)  
for  use  in  the  Vulcan  emission  calculations.  Appendix  A,  Table  A.3  provides  fuel  heat  
contents  used  in  this  process.  
Retrieval  options:  
************************************************************************  
1.  There  is  a  PF  provided  within  the  NEI  and  there  is  a  FIRE  PF  (or  multiple).  Is  
the  provided  NEI  PF  within  the  tolerance  thresholds10  of  the  FIRE  PF  (or  any,  if  
multiple)?    
-­‐  If  so,  retrieve  the  NEI  provided  PF    
-­‐  If  not,  retrieve  the  FIRE  PF  (the  largest,  if  multiple)  
 

10  The  factor  must  be  within  a  factor  of  three  larger  than  that  supplied  or  within  75%  lower.    

13
 2.  There  is  a  PF  provided  within  the  NEI,  but  no  available  PF  in  the  FIRE  
database.  Is  the  NEI  provided  PF  within  the  tolerance  thresholds  of  the  default  
PF?  
-­‐  If  so,  retrieve  the  NEI  provided  PF    
-­‐  If  not,  retrieve  the  default  PF  
 
3.  There  is  no  PF  provided  within  the  NEI,  but  there  is  a  FIRE  PF  (or  multiple)  
-­‐  Retrieve  the  FIRE  PF  (use  largest,  if  multiple)    
 
4.  There  is  no  PF  provided  within  the  NEI  and  there  is  no  FIRE  PF  
-­‐  Retrieve  the  default  PF  
************************************************************************  
The  next  step  in  the  CO2  emissions  calculation  is  the  estimation  of  the  fuel  
throughput  for  the  considered  process.  This  is  computed  as  the  ratio  of  the  mass  of  
emitted  pollutant  divided  by  the  PF  (with  appropriate  units  ascertained).11    
2.2.2  CO2  emissions  estimation  
Once  the  material/fuel  throughput  has  been  produced,  a  CO2  EF  is  applied  (provided  
in  Appendix  A,  Table  A.3).  The  CO2  EF  is  variously  referred  to  as  “carbon  coefficient”  
or  “carbon  factor”  in  the  literature.  For  this  study,  it  represents  the  mass  of  carbon  
or  CO2  emitted  per  unit  energy  of  fuel  consumed  (since  all  fuel  is  previously  
converted  to  energy  units,  all  CO2  EFs  are  thus  standardized).  Emission  factors  for  
CO2  are  based  on  the  fuel  carbon  content  and  assume  a  gross  calorific  value  or  high  
heating  value,  as  this  is  the  convention  most  commonly  used  in  the  US  and  Canada  
[URS,  2003].  Emission  factors  are  reported  as  units  of  carbon  dioxide  as  opposed  to  
units  of  carbon  and  assume  100%  oxidation  of  fuel  carbon  to  CO2  for  natual  gas,  
99%  for  coal  and  oil  [IPCC  1996;  DOE/EIA  2007b].  
2.3  Sources  of  Uncertainty  
The  computation  of  CO2  emissions  in  the  point  data  source  includes  a  number  of  
self-­‐reporting  uncertainty  sources  which  we  designate  here  as  “categorical”  and  
“numerical”  uncertainties.  Categorical  uncertainties  include  the  following:  
1. Time  period  designation  
2. Fuel  designation  
3. SCC  designation  
Errors  in  these  information  sources  imply  that  the  facility  operator  or  office  tasked  
with  estimating  pollutant  emissions  mis-­‐categorized  the  time  period  for  which  
emissions  were  estimated,  the  fuel  being  consumed  or  the  SCC  code  for  which  the  
pollutant  emissions  were  estimated.  Estimating  the  liklihood  that  categorical  errors  
were  made  is  difficult.  Quantifying  how  that  categorical  error  would  impact  the  final  
CO2  emission  estimate  is  also  difficult.  Given  the  nature  of  the  reporting  
(professionals  tasked  with  complying  with  air  quality  regulatory  law)  and  the  

11  This  fuel  throughput  calculation  assumes  that  the  fuel  estimated  is  the  amount  of  fuel  “burned”  in  

the  combustion  process.    

14
difficulty  associated  with  estimating  the  potential  errors,  this  study  considers  these  
sources  of  uncertainty  to  be  low.  Nonetheless,  uncertainty  associated  with  category  
3.  is  attempted  below.  
The  numerical  sources  of  uncertainty  in  the  CO2  calculation  include  the  following:  
1. Pollutant  emission  quantity  reported  
2. Provided  pollutant  emission  factor  
3. Default  pollutant  emission  factor  
4. CO2  emission  factor  (carbon  content  of  fuel)  
5. Heat  content  conversions  
Among  these  uncertainty  sources,  3  through  5:  the  CO2  emission  factor,  the  heat  
content,  and  the  default  pollutant  emission  factor,  can  be  quantified  with  available  
data.  The  first  two  uncertainty  sources  are  difficult  to  quantify.  Unlike  the  
categorical  uncertainties,  however,  these  are  both  much  more  likely  to  contain  
errors  and  those  errors  would  have  a  direct  and  potentially  large  impact  on  the  CO2  
emissions  estimation.  
In  order  to  provide  a  first  order  sense  of  the  impact  of  the  quantifiable  components  
of  the  last  three  sources  of  numerical  uncertainty,  we  take  a  sensitivity  approach.    
The  sensitivity  approach  asks  the  question:  how  wrong  could  the  CO2  emissions  
estimate  be,  given  typical  variations  in  the  underlying  sources  of  uncertainty?  These  
variations  are  conservatively  interpreted  as  a  one-­‐sigma  spread  on  the  central  
estimate  of  the  CO2  emissions  (though  the  variations  described  below  are  likely  
higher  than  a  true  one-­‐sigma  spread  of  an  actual  sample  of  underlying  factors).  
2.3.1  Pollutant  emission  factor  
For  the  default  pollutant  emission  factors,  a  range  of  values  is  used  as  a  form  of  
sensitivity.  The  range  reflects  values  in  the  WebFire  database  as  well  as  a  range  of  
values  that  are  self-­‐reported  in  the  NEI  point  database  itself.  For  example,  for  
industrial  pulverized  bituminous  coal  combustion,  values  ranging  from  0.5  lbs  
CO/ton  to  22.86  lbs  CO/ton  are  included  in  the  sensitivity  test.  These  represent  the  
highest  and  lowest  possible  values  for  CO  emissions/unit  fuel  available  in  the  
WebFire/NEI  combined  datasets.  The  lower  CO  emission  factor  will  lead  to  a  greater  
amount  of  fuel  consumed  and  a  greater  CO2  emission.  Whereas  the  high  CO  emission  
factor  will  do  the  opposite  (result  in  lower  CO2  emissions).  This  range  also  
incorporates  the  categorical  error  3.  In  the  first  list  above  as  this  spread  of  CO  
emission  factors  generally  reaches  across  SCC  values  within  a  specific  fuel  
designation.  These  extreme  ranges  are  considered  2-­‐sigma  errors  and  hence,  the  
distance  between  the  central  EF  and  the  hi  and  lo  extremes  are  halved  to  arrive  at  a  
one-­‐sigma  value.  
2.3.2  Heat  and  carbon  content  
As  described  in  section  2.2.1,  pollutant  emissions  that  are  reported  in  mass  or  
volume  units  are  first  converted  to  emissions  per  unit  thermal  content  (per  106  btu).  
This  requires  the  use  of  a  heat  content  conversion  which  is  dependent  upon  the  fuel  
considered  as  provided  in  Appendix  A,  Table  A.3.  This  alters  equation  2.2  as  follows,  

15
   (2.3)  

where  HCf  is  the  heat  content  which  is  only  dependent  upon  the  fuel  consumed  in  
the  combustion  process.    
Fuel  heat  contents  can  vary  substantially  and  are  generally  associated  with  the  
parent  fuel  formation/location  (coal  mine,  oil  well,  etc).  Variation  (one  standard  
deviation)  in  heat  content  is  derived  from  fuel  samples  and  is  described  and  
quantified  in  DOE/EIA  [2007b].  The  largest  variation  in  heat  content  is  found  for  
coal  and  is  derived  from  sampling  coal  from  each  producing  state  destined  for  
power  plants  in  the  United  States.  Depending  upon  coal  rank,  variation  (standard  
deviation  about  the  mean  value)  in  heat  content  ranges  from  4  to  12%.  Additional  
analysis  was  performed  here  by  quantifying  the  variation  in  coal  delivered  to  power  
plants  using  the  DOE/EIA  form  423  database  and  consistent  results  were  found  
[DOE/EIA,  2002a;  2002b].    
Variation  in  heat  content  for  the  remaining  fuel  categories  are  partly  derived  from  
the  DOE/EIA  form  423  database  or  quantitatively  identical  to  the  variation  assigned  
for  the  carbon  coefficient  (CO2  emission  factor).  Variation  in  the  CO2  emission  factor  
is  similarly  derived  from  DOE/EIA  [2007b].  The  largest  variation  is  for  refinery  
gases  (33%).  Variation  in  the  heat  content  and  carbon  content  of  fuel  are  generally  
correlated.  We  treat  them  as  uncorrelated  and  additive.  This  is  likely  a  conservative  
approach.    These  stated  variations  are  considered  a  one-­‐sigma  spread.  
2.3.3  Utilizing  only  default  pollutant  EFs  
Finally,  the  provided  pollutant  emission  factor  can  be  tested  somewhat  by  
substituting  all  provided  pollutant  emission  factors  with  default  factors  in  all  
instances.  This  bypasses  both  the  acceptance  of  provided  emission  factors  and  the  
SCC-­‐specific  WebFire  emission  factor  lookup  and  defaults  to  the  values  in  Appendix  
Table  A.1  and  Table  A.2.  
2.3.4  Summary  of  sensitivities  
Hence,  we  have  four  sensitivity  tests:  
1. vary  the  default  pollutant  emission  factors  (hi  and  lo  cases)  
2. vary  the  fuel  heat  content  (hi  and  lo  cases)  
3. vary  the  fuel  carbon  content  (hi  and  lo  cases)  
4. utilize  only  default  emission  factor  
The  first  three  can  be  quantified  in  a  directional  sense  to  arrive  at  a  “hi”  and  “lo”  CO2  
emissions  estimate  whereas  test  4  will  cause  results  to  vary  in  both  numerical  
directions.  Results  are  produced  which  isolates  the  impact  of  each  of  these  tests  and  
the  combination  of  all  four  sensitivity  tests.  The  combination  sensitivity  test  is  as  
follows:  
Low-­‐end  pollutant  emission  factors  +  hi-­‐end  heat  content  +  hi-­‐end  CO2  EF.  

16
This  combination  senstivity  test  is  run  with  and  without  utilization  utilization  of  
default  emission  factors.  
 

17
3.0  Cement  
CO2  is  emitted  from  cement  manufacturing  as  a  result  of  fuel  combustion  and  as  
process-­‐derived  emissions  [van  Oss  2005]  .  The  emissions  from  fuel  combustion  are  
captured  in  the  fossil  fuel  combustion  emission  processes.  The  process-­‐derived  CO2  
emissions  result  from  the  chemical  process  that  converts  limestone  to  calcium  oxide  
and  CO2.  This  occurs  during  “clinker”  production  (clinker  is  the  raw  material  for  
cement  which  is  producing  by  grinding  the  clinker  material).  
3.1  Emissions  estimation  
Estimation  of  CO2  emissions  from  clinker  production  utilizes  two  datasets.  The  first  
is  the  data  provided  by  the  Portland  Cement  Association  [PCA  2006].  The  PCA  
document  provides  the  annual  clinker  capacity  at  individual  facilities,  postal  
addresses,  facility  name,  zip  code  and  contact  phone  numbers.  The  capacity  data  
reflects  conditions  for  the  calendar  year  2006.  
The  other  dataset  utilized  is  the  Minerals  Yearbook  produced  by  the  United  States  
Geological  Survey  [USGS  2003].  The  USGS  Yearbook  provides  the  capacity  factor  (or  
percent  utilization  of  capacity)  for  2002  on  a  statewide  or  multi-­‐state  basis  (some  
states  are  quantified  individually,  others  are  part  of  an  aggregate).    
Clinker  production  for  2002  is  estimated  by  multiplying  the  USGS-­‐suppled  capacity  
factor,  defined  at  the  state  or  state-­‐aggregate  level,  by  the  individual  facility  capacity  
(appropriate  to  the  state  or  state-­‐aggregate  capacity  factor)  provided  by  the  PCA  
document.  The  sum  of  the  individual  PCA-­‐reported  capacities  for  all  facilities  in  a  
state  or  multi-­‐state  aggregate  can  be  compared  to  the  USGS-­‐reported  equivalent.  
This  is  presented  in  Figure  3.1a.  

   
Figure  3.1  Comparison  of  PCA-­reported  [PCA,  2006]  statewide  or  multi-­state  
aggregate  a)  clinker  capacity  and  b)  clinker  production  to  that  reported  by  the  USGS  
[USGS  2003].  The  1:1  line  is  also  shown.  Units:  kilotonnes/year.  
The  USGS  reported  capacity  (94,241  kt/year)  is  consistently  higher  (25%)  than  that  
provided  by  the  PCA  reference  document  (75,239  kt/year).  The  large  outlier  value  is  
the  datum  for  the  sum  of  Michigan  and  Wisconsin.  

18
The  same  relationship  can  be  constructed  for  production  and  this  is  shown  in  Figure  
3.1b.  
The  USGS  reported  production  is  larger  (27%)  at  the  state  or  state-­‐aggreate  level  
compared  to  the  data  reported  in  the  PCA  document.  
3.2  Geolocation  
The  geolocation  for  each  of  the  individual  facilities  was  achieved  by  entering  the  
PCA  document’s  facility  address  into  Google  Earth  and  visually  inspecting  the  scene  
for  the  primary  emitting  stack  of  the  cement  facility.  This  approach  succeeded  in  
locating  all  105  facilities  present  in  the  PCA  document.  
These  geolocation  points  are  checked  against  the  cement  facilities  reported  through  
the  NEI  point  database  (see  section  2.0).  82  of  the  105  facilities  present  in  the  PCA  
database  are  found  (with  geolocation)  in  the  NEI  point  data.  The  average  percent  
difference  between  the  82  Google  Map  identiied  locations  and  those  entered  in  the  
NEI  point  datebase  is  -­‐0.01%  and  0.01%  for  the  latitude  and  longitude,  respectively.    
3.3  CO2  emissions  factor  
The  CO2  emission  factor  used  in  the  Vulcan  Project  is  0.59  metric  tonnes  CO2/short  
ton  of  clinker  produced12.  This  emission  factor  is  the  result  of  a  calculation  that  
reflects  IPCC  recommendation  on  the  incorporation  of  cement  kiln  dust.  The  
calculation  is  as  follows:  
  (3.1)  
Where  Ei  is  the  CO2  emissions  in  tonnes  of  CO2  from  facility  i  and  Pi  is  the  clinker  
produced  by  facility  i  in  units  of  metric  tonnes.  The  factor,  0.525  metric  tonnes  
CO2/metric  tonne  of  clinker,  is  an  emission  factor  recommended  by  the  World  
Business  Council  for  Sustainable  Development  and  consistent  with  the  
Intergovernmental  Panel  on  Climate  Change  emission  factors  when  corrected  for  
typical  MgO  contents  in  clinker  [WBCSD  2005].  As  this  emissions  factor  does  not  
account  for  the  fact  that  a  percentage  of  the  clinker  precursor  materials  remain  in  
the  kiln  in  the  form  of  cement  kiln  dust  (CKD),  the  IPCC  recommends  that  emissions  
from  CKD  be  included  as  equal  to  2%  of  total  process-­‐related  CO2  emissions.  
The  EIA  estimates  cement  manufacturing  in  2002  to  account  for  43  MtCO2/year  out  
of  a  total  69.4  MtCO2/year  for  their  entire  industrial  process-­‐derived  CO2  emissions  
[DOE/EIA  2007a].  The  latter  value  includes  both  limestone  and  soda  ash  
manufacturing  which  are  currently  not  included  in  the  Vulcan  inventory.13  These  
estimates,  in  turn,  are  based  upon  throughput  estimates  from  the  U.S.  Geological  
Survey.  Vulcan  estimates  a  total  of  44.22  MtCO2/year  which  compares  well  with  the  
cement  manufacturing  estimate  from  the  EIA.  

12
This is equivalent to 0.536 metric tonnes of CO2/metric tonne of clinker produced.
13  These  categories  will  be  included  in  Vulcan  2.0.  

19
3.4  Sources  of  Uncertainty  
The  primary  sources  of  uncertainty  in  the  calculation  of  CO2  emissions  from  cement  
manufacturing  are  as  follows:  
1. Uncertainty  in  facility  clinker  capacity  
2. Uncertainty  in  state  or  state-­‐aggregate  capacity  factor  
3. Unaccounted  for  sub-­‐state  variation  in  capacity  factors  
4. Unaccounted  for  variation  in  CO2  emission  factor  (temperature,  MgO,  FeO  
contents,  etc)  
Numbers  1  through  3  are  external  data  sources  with  no  uncertainty  estimate  
included.  Hence,  construction  of  a  probability  density  function  associated  with  the  
incoming  data  is  difficult.  For  the  uncertainty  sensitivity  analysis  performed  in  the  
Vulcan  Project,  an  attempt  is  made  to  reflect  a  range  of  possible  values  for  1,  2,  and  
4.  A  high-­‐end  estimate  is  generated  by  assuming  an  increase  of  10%  in  these  three  
sources  of  uncertainty.  A  low-­‐end  estimate  is  generated  by  assuming  a  decrease  of  
10%  in  all  three  of  these  sources  of  uncertainty.  These  are  considered  one-­‐sigma  
errors.  

20
4.0  Electricity  Production  CO2  Emissions  
4.1  ETS/CEM  data  
The  emissions  from  electricity  production  is  primarily  supplied  by  data  obtained  
from  the  DOE/EIA  and,  most  importantly,  the  EPA’s  CAMD  Emission  Tracking  
System/Continuous  Emissions  Monitoring  system  (ETS/CEMs)  data  for  Electrical  
Generating  Units  (EGUs)  [ERG  and  EHP,  2004;  USEPA  2004a;  USEPA  2005b;  
Ackerman  &  Sundquist,  2008;  Petron  et  al.,  2008].  The  ETS/CEMs  data  is  collected  
under  the  Acid  Rain  Program  (ARP),  which  was  instituted  in  1990  under  Title  IV  of  
the  Clean  Air  Act.  The  ARP  regulates  EGUs  that  burn  fossil  fuel  and  are  greater  than  
25  MW  capacity  or  are  less  than  25  MW  but  which  burn  coal  with  a  sulfur  content  of  
greater  than  0.05%  by  weight.  In  addition  to  heat  input,  these  facilities  are  required  
to  engage  in  continuous  monitoring  and  reporting  of  sulfur  oxides  (SOx),  CO2,  and  
nitrogen  oxides  (NOx)  emissions.  These  data  are  reported  directly  as  hourly  CO2  
emissions  monitored  from  an  emitting  stack  or  through  a  calculation  based  on  
records  of  fuel  use.  All  emitting  locations  are  geocoded  to  latitude,  longitude  and  
postal  address.    
Because  the  ETS/CEMs  data  within  the  NEI  are  reduced  to  the  annual  total  
emissions,  the  original  hourly  ETS/CEMs  reporting  is  utilized  in  the  Vulcan  
inventory.  No  attempt  is  made  to  gap-­‐fill  missing  data  or  adjust  emissions  in  any  
way  (time  gaps  may  be  due  to  peaking  units  or  shutdowns,  etc).  There  are  1241  
facilities  in  the  hourly  data,  consistent  with  the  annual  files  available  from  the  EPA.  
Furthermore,  the  total  CO2  emissions  for  all  of  the  ETS/CEMs  data  as  calculated  
from  the  hourly  emissions  is  0.60  GtC/year,  consistent  with  the  annual  files.  
4.2  Cross-­‐matching  to  NEI  
Removal  of  the  ETS/CEMs  facilities  from  the  NEI  must  be  accomplished  to  avoid  
double-­‐counting  of  CO2  emissions.  There  were  1241  ETS/CEMs  individual  facilities  
in  2002  (which  constitute  a  much  larger  number  of  “processes”)  and  the  identifying  
and  emissions  data  associated  with  these  facilities  can  be  downloaded  from  the  
CAMD  website  [USEPA  2008a].  
Cross-­‐matching  the  ETS/CEMs  and  NEI  processes  was  accomplished  by  attaining  the  
Registry  ID  associated  with  the  ETS/CEMs  facilities  from  the  EPA  Envirofacts  data  
warehouse  [USEPA  2008b].  The  Registry  ID  is  a  common  identifier  for  the  two  
reporting  systems.  This  procedure  led  to  911  facility  cross-­‐matches.  An  additional  
129  matches  were  identified  from  the  common  ORISPL  code,  an  identifying  code  
utilized  by  the  DOE  and  often  found  in  the  NEI.  
The  remaining  201  facilities  were  approached  through  a  combination  of  proximity  
and  address/facility  name  matching.  All  facilities  within  0.05  degrees  in  latitude  and  
longitude  were  retrieved  from  the  NEI  point  database  and  these  were  individually  
inspected  to  determine  which,  if  any,  were  referencing  the  same  emitting  facility.  
Alternative  facility  names  were  determined  that  these  were  searched  within  the  
NEI.  This  effort  achieved  an  additional  118  matched  facilities.  All  of  the  matched  
facilities  were  then  removed  and  the  separate  hourly  CO2  emissions  ETS/CEMs  data  
was  used  in  the  Vulcan  inventory.  

21
The  83  unmatched  ETS/CEM  facilities  accounted  for  2.1  Mtonnes  of  CO2  or  0.34%  of  
the  total  ETS/CEMs  2002  CO2  emissions.  No  further  attempt  was  made  to  remove  
these  facilities  from  the  NEI  and  it  remains  unclear  whether  or  not  these  facilities  
are  included  in  the  NEI.  
4.3  Fuel  assignment  
In  order  to  maintain  the  ability  to  report  CO2  emissions  according  to  fuel  at  each  
emitting  process  or  record,  the  exact  fuel  or  fuel  mix  at  each  of  the  ETS/CEMs  
facilities  was  identified  through  matching  with  the  EIA  form  906  data  which  
provides  a  detailed  summary  of  key  characteristics  at  all  power  production  facilities  
[DOE/EIA  2008].  The  EIA  form  906  data  provides  a  listing,  for  the  year  2002,  of  the  
fuel  share  at  reporting  power  production  facilities  in  the  US.  Using  this  data,  1167  
matches  were  made  through  direct  ORSPL  code  matching.  Five  additional  facilities  
were  matched  through  a  combination  of  state  location  and  facility  name.  The  fuel  
mix  at  the  remaining  89  facilities  were  identified  through  a  combination  of  online  
searching  of  utility  websites  and  direct  contact  with  facility  operators.    
After  eliminination  of  the  ETS/CEMs  facilities  within  the  point  NEI  database,  we  
have  101,758  processes  in  the  NEI  since  a  single  facility  can  have  multiple  processes  
associated  with  it.  
Some  electric  generation  is  further  captured  in  the  NEI  point  file  (with  no  obvious  
match  to  ETS/CEMs  facilities)  and  these  emissions  are  assumed  to  be  associated  
with  facilities  that  are  too  small  to  be  included  in  the  ETS/CEMs  system.  They  add  a  
small  amount  of  CO2  emissions  to  the  final  value  (~0.014  GtC/year)  and  are  added  
to  the  utility  sector  in  the  final  Vulcan  sectoral  output.  
Purdue  University  utilizes  a  power  plant  (the  Wade  facility)  for  generating  onsite  
electricity.  In  2002,  this  facility  was  not  required  to  report  emissions  under  the  Acid  
Rain  legislation  and  reporting  of  local  air  pollutants  was  not  located  in  the  NEI  (the  
reason  for  this  is  still  under  investigation).  Hence,  this  facility  was  individually  
added  to  the  electricity  generation  sector  of  the  Vulcan  data  product  from  locally  
provided  data  (Robin  Ridgway,  personal  communication).  
4.4  Sources  of  Uncertainty  
Recent  research  has  attempted  to  estimate  uncertainties  associated  with  power  
plant  CO2  emissions  in  the  U.S.  through  the  comparison  of  two  power  plant  CO2  
emissions  data  sources  [Ackerman  and  Sundquist,  2008].  The  first  is  calculated  CO2  
emissions  accomplished  by  the  DOE/EIA.  This  calculation  includes  data  collected  
from  each  power  plant  on  the  physical  consumption  of  fuel  and  the  heat  content  of  
that  consumed  fuel  [DOE/EIA  2010].  Hence,  the  amount  of  thermal  energy  
consumed  at  each  power  plant  is  calculated  (see  
www.eia.doe.gov/cneaf/electricity/page/eia906_920.html  for  a  legacy  of  the  forms  
used  to  collect  this  information).  The  consumed  thermal  energy  is  combined  with  a  
fuel-­‐specific  CO2  emission  factor  (the  quantity  of  CO2  emissions  per  unit  energy)  
provided  by  the  DOE/EIA  (see  Appendix  A  of  DOE/EIA  2010).  The  second  source  in  
Ackerman  and  Sundquist  [2008]  is  the  EPA’s  eGRID  database  which  combines  the  
ETS/CEMs  data  described  previously  in  this  document  with  a  calculation  of  CO2  

22
emissions  based  on  fuel  consumption  data  supplied  by  the  DOE/EIA.  This  latter  data  
is  attributed  in  eGRID  documentation  to  the  same  sources  as  used  by  the  DOE/EIA  
in  its  CO2  emissions  estimation.  Differences  between  eGRID  and  the  DOE/EIA,  for  
that  subset  of  facilities  for  which  there  is  only  a  fuel  calculation-­‐based  method  
available,  are  presumably  due  to  emission  factors  and  related  assumptions.    
The  Ackerman  and  Sundquist  [2008]  study  generated  subsets  of  the  total  facility  list  
based  on  the  type  of  facility  (combined  heat  &  power  vs  non-­‐CHP,  for  example)  and  
the  method  employed  to  report  CO2  emissions.  For  all  fossil  fuel  facilities,  the  study  
found  5.1%  and  11%  for  average  signed  and  average  absolute  differences  between  
the  two  datasets.  They  found  that  the  largest  percentage  discrepancies  were  in  cases  
where  the  eGRID  database  reported  an  ETS/CEMs  value  and  the  DOE/EIA  reported  
a  fuel-­‐based  calculation  value.  In  these  cases,  they  found  5.4%  and  16.6%  
differences  in  the  average  signed  and  averaged  absolute  comparisons,  respectively.  
Where  the  facility  had  a  mixture  of  fuel-­‐based  calculation  and  stack  monitoring  (as  
with  multiple  boilers),  the  values  were  21.7%  and  24.4%  respectively.  
Of  course,  all  of  these  percent  differences  do  not  take  into  account  the  size  of  the  
emissions  themselves  but  treats  all  facilities,  regardless  of  size,  as  equal  when  
generating  the  percent  difference  statistics  (we  call  these  “unweighted”  statistics).  
When  the  mass  of  CO2  emissions  are  considered  the  Ackerman  and  Sundquist  
[2008]  study  concludes  that  all  fossil  fuel  facilities  result  in  a  3.4%  difference  
(signed  difference)  in  CO2  emissions  for  the  U.S..  Unfortunately,  an  absolute  
difference  is  not  calculated  with  the  CO2  emissions  magnitude  included  in  the  
analysis.  
Assigning  an  uncertainty  to  the  power  plant  emissions  in  the  Vulcan  data  product  
remains  a  challenge  even  with  the  analysis  performed  by  Ackerman  and  Sundquist  
[2008].  Some  of  the  differences  found  are  due  to  differing  methodological  treatment  
between  the  eGRID  and  DOE/EIA  studies  and,  as  such,  is  not  necessarily  a  reflection  
of  uncertainty  of  the  ETS/CEMs  data  per  se.  However,  that  component  of  the  study  
comparing  facilities  utilizing  ETS/CEMs  devices  vs  fuel  calculations  may  be  
considered  a  proxy  for  the  uncertainty  associated  with  these  monitoring  devices.  
This  is  an  imperfect  metric  because  the  differences  in  the  two  datasets  reflect  not  
only  the  potential  uncertainty  in  ETS/CEMs  monitoring  and/or  fuel  consumption  
amounts,  but  in  the  methodological  application  of  emission  factors  and  fuel  heat  
contents.  Most  importantly,  the  differences  noted  in  Ackerman  and  Sundquist  
[2008]  are  biases  as  opposed  to  random  uncertainty.  They  represent  the  difference  
between  the  mean  of  two  distributions.    
Ackerman  and  Sundquist  [2008]  found  a  1.4%  signed  difference  in  the  total  U.S.  CO2  
emissions  for  those  facilities  that  utilized  ETS/CEMs  devices  and  this  group  of  
facilities  accounted  for  ~70%  of  the  CO2  emissions.  Combination  facilities  
(accounting  for  ~20%  of  emissions)  had  signed  differences  of  9.9%.  Finally,  facilities  
utilizing  fuel  calculations  in  both  datasets  (accounting  for  the  remaining  ~10%  of  
CO2  emissions)  had  signed  differences  of  3.9%.  A  weighted  average  of    these  three  
categories  comes  to  a  signed  difference  of  3.3%  very  close  to  the  overall  signed  
difference  for  all  fossil  fuel  facilities  of  3.4%.  This  hi  bias  is  confirmed  by  industry  

23
studies  which  repeatedly  suggest  a  hi-­‐bias  associated  with  the  ETS/CEMs  
measurements  [Zimmerman  et  al.,  2010;  Berry  et  al.,  1998;  ICF  consulting]  .  
The  only  studies  available  regarding  random  uncertainty  are  non-­‐peer  reviewed  
industry  analysis.  Zimmerman  et  al.,  [2009]  analyzed  ETS/CEMs  data  and  concluded  
that  random  uncertainties  were  “at  least  ±4%-­‐5%”.  This  was  due  to  uncertainties  in  
the  determination  of  the  mass  flow  rate  of  CO2  (a  combination  of  CO2  flow  rate  and  
concentration).  
Hence,  we  utilize  two  forms  of  uncertainty  in  our  sensitivity  analysis.  We  consider  
that  all  of  the  emissions  estimates  in  the  ETS/CEMs  dataset  to  be  biased  high  by  
3.4%.  In  addition  we  include  random  uncertainty  of  5%  (assumed  a  one-­‐sigma  
error).  Hence,  we  have  a  “hi”  case  and  a  “low”  case.  The  hi  case  increases  all  
emissions  by  +1.6%  and  the  lo  case  decreases  emissions  by  -­‐8.4%.    
 

24
5.0  NEI  Nonpoint  CO2  Emissions  
The  area  or  nonpoint  source  emissions  (dominated  by  residential  and  commercial  
economic  sectoral  categories  though  industrial  and  utility  sector  emissions  exist)  
are  stationary  sources  that  are  not  inventoried  at  the  facility-­‐level  and  can  be  
thought  of  as  representing  “diffuse”  sources  within  a  geographic  area.  The  EPA  
provides  recommendations  to  state/local  agencies  on  how  to  collect  nonpoint  
source  emissions  information  and  the  state/local  agencies  are  given  a  number  of  
options  in  forming  the  basis  of  the  reported  information  [ERG  2001b].  The  EPA  
prefers  emissions  to  be  estimated  by  extrapolating  from  a  sample  set  of  data  for  the  
activity  to  the  entire  population,  but  a  number  of  other  approaches  are  allowed  
including  material  balance,  mathematical  models,  and  emission  factors.  This  means  
that  the  method  employed  will  vary  by  location  and  this  generally  implies  that  the  
nonpoint  source  emission  information  has  more  intrinsic  variability  in  terms  of  
quality  and  consistency  than  either  the  mobile  or  point  sources  emissions  estimates.  
5.1  Data  reduction  
The  NEI  nonpoint  database  is  comprised  of  a  file  structure  similar  to  the  point  
sources  noted  in  Figure  2.1.and  is  comprised  of  five  related  files  [USEPA  2006c].  
These  five  nonpoint  files  are:  1)  transmittal  (TR),  2)  emission  process  (EP),  3)  
emission  period  (PE),  4)  control  equipment  (CE),  5)  emission  (EM).  The  majority  of  
analysis  is  performed  with  the  emission  (EM)  data  file.  
The  fundamental  nonpoint  “unit”  as  defined  for  the  Vulcan  calculations  is  the  
“process”  which  identifies  a  single  SCC  in  a  single  county  using  a  single  fuel  and  with  
a  unique  Mat  IO.      
As  with  the  point  NEI  data,  the  nonpoint  database  contains  information  on  
processes  that  do  not  consume  fossil  fuels  or  processes  that  contain  emissions  from  
fossil  fuel  combustion  other  than  NOx  and  CO.  Hence,  the  database  is  reduced  to  
only  that  data  relevant  to  the  CO2  emissions  problem.  Currently,  the  Vulcan  
inventory  utilizes  CO  emissions  in  order  to  compute  fuel  throughput  and  
subsequent  CO2  emissions.  A  total  of  126,680  processes  were  retrieved  from  the  
nonpoint  NEI  that  report  CO  emissions.  
As  with  the  point  source  data,  a  series  of  reductions  are  made  to  this  NEI  nonpoint  
CO  emissions  dataset  before  processing  for  CO2  emissions.  
5.1.1.  Material  and  pollutant  qualifiers  
The  nonpoint  NEI  was  reduced  by  narrowing  the  database  by  the  process  
material/fuel  and  the  pollutant  produced  by  that  process.  Only  records  that  had  the  
following  combination  were  considered:  
1)  the  pollutant  code  indicated  CO  emissions  present  
      AND  
2)  the  material  can  be  found  in  the  Vulcan  fossil  fuel  list  (Table  2.1)    
      AND    
3)  the  Mat  IO  identifier  was  set  to  “input”  (“I”)  or  “null”  
 

25
The  material  is  identified  though  a  combination  of  examining  the  provided  NEI  
material  code  and  SCC  code.  As  with  the  point  NEI  data,  many  material  codes  were  
absent  (“null”  values).  In  order  to  explore  emission  processes  for  which  the  fuel  or  
input/output  identifier  was  listed  as  “null”,  the  NEI  input  format  (NIF)  source  
classification  code  (SCC)  lookup  table  was  used  to  fill  in  the  missing  information  and  
confirm  the  material  classifications  provided  by  the  NEI  material  code.14  This  
exercise  further  identified  how  the  material  was  used  in  the  emitting  process.  For  
materials  listed  in  Table  2.1,  only  actions  identified  as  “burned”    in  the  SCC  lookup  
table  were  retained  in  the  Vulcan  nonpoint  inventory.  Other  actions  such  as  
“processed”,  “shipped”,  or  “produced”  were  not  considered  the  purview  of  the  
Vulcan  CO2  inventory  and  these  processes  were  removed.    
The  SCC  was  also  used  to  identify  the  economic  sector  (residential,  commercial,  etc).  
If  the  sector  was  not  readily  identifiable,  the  process  was  designated  to  “unknown”.  
These  were  later  assigned  based  on  final  state-­‐level  mass  balance  considerations  
(see  section  5.2.3).  
5.1.2.  Time  period  consistency  (presentation  identical  to  section  2.1.2)  
Emissions  reporting  in  the  NEI  is  made  for  a  small  set  of  different  reporting  periods  
or  time  “types”  as  follows:  
o Type  27:  average  weekday  
o Type  28:  average  weekend  day  
o Type  29:  average  day  in  period  
o Type  30:  entire  period  total  
A  given  process  can  report  emissions  for  more  than  one  of  these  time  period  types.  
Only  processes  which  identify  time  type  30  are  retained  and  all  others  are  
removed.15    In  most  cases  the  time  type  30  is  a  complete  calendar  year  total  amount.  
These  annual  emissions  are  initially  divided  equally  amongst  total  number  of  days  
and  hours  in  the  year  (for  the  gridded  hourly  output).  Section  8.0  describes  further  
temporal  conditioning  of  the  emissions.  Though  most  facilities  with  emission  time  
type  30  estimate  the  emissions  for  a  period  of  365  days  or  8760  hours  per  year,  
certain  facilities  report  timespans  for  a  specific  portion  of  the  year  making  the  
effective  operational  number  of  days  in  the  year  less  than  365.  In  such  cases,  the  
annual  emissions  reported  by  the  facility  are  equally  divided  amongst  the  reported  
number  of  days/hours  rather  than  365  days  (8760  hours).16    
Hence,  the  effective  calculation  is  as  follows:  

14  Material  codes  are  actually  supplied  in  multiple  fields  in  the  NEI  which  are  often  contradictory.  The  

material  codes  are  associated  with  each  pollutant  field  in  addition  to  provided  as  an  independent  
field.  The  materials  identified  through  the  SCC  lookup  are  used  to  override  all  other  material  
classifications  and  form  the  basis  of  the  fuel  considered.  
15
Version 2.0 of the Vulcan inventory will utilize the multiple time types to further structure emissions
during the emitting period.
16
However, as noted in Section 8.0, the emissions are forced to be constant for the year prior to performing
monthly and hourly downscaling.

26
   (5.1)  

Where  E  is  emissions,  t  is  hourly  timestep,  p  is  the  reported  emissions  period,  and  
Δtp  is  the  number  of  days  in  the  reported  time  period  (most  commonly  365).    
There  are  also  cases  in  the  input  NEI  dataset  where  the  operational  start/end  date  
of  a  process  is  reported  as  a  year  other  than  2002.  These  are  a  mixture  of  typos  by  
the  reporting  agency  or  examples  where  a  previous  year  emissions  have  been  
“carried  over”  to  the  2002  database.  Such  records  are  modified  to  start  on  1/1/2002  
and  end  on  12/31/2002.  
5.2  Quantifying  CO2  emissions  
With  the  data  reduction  complete,  each  process  is  examined  in  order  to  retrieve  
information  by  which  an  amount  of  emitted  CO2  can  be  produced.  The  CO2  emission  
quantity  is  determined  from  the  provided  CO  emissions  amount  in  combination  with  
a  CO  emission  factor  (EF)  and  an  emission  factor  for  CO2.    
The  basic  process  by  which  CO2  emissions  are  created  is  theoretically  identical  to  
the  point  source  process:  
 

  (5.2)  
 
where  C,  is  the  emitted  amount  of  carbon,  PE  is  the  equivalent  amount  of  
uncontrolled  CO  emissions,  p  is  the  combustion  process,  f  is  the  fuel  category,  PF  is  
the  emission  factor  associated  with  the  criteria  pollutant,  and  CF  is  the  emission  
factor  associated  with  CO2  (provided  in  Appendix  A,  Table  A.3).  
5.2.1  CO  Emission  factor  retrieval  
The  CO  EF  used  is  chosen  from  two  different  alternatives  (in  the  following  order):  1)  
the  EF  provided  in  the  NEI  data  itself  for  the  particular  CO-­‐emitting  process,  2)  a  
default  EF  value  (provided  in  Appendix  A,  Table  A.1).  CO  emission  factors  provided  
in  units  other  than  mass  per  unit  energy  (applies  only  to  those  EFs  provided  within  
the  NEI)  are  converted  using  the  standard  fuel  heat  contents  provided  in  Appendix  
A,  Table  A.3.  Standardization  of  fuel  inputs  to  the  combustion  processes  is  essential  
to  maintain  numerical  integrity.    
5.2.2  CO2  emissions  estimation  
Once  the  material/fuel  throughput  has  been  produced,  a  CO2  emission  factor  is  
applied.  Emission  factors  for  CO2  are  based  on  the  fuel  carbon  content  and  assume  a  
gross  calorific  value  or  high  heating  value,  as  this  is  the  convention  most  commonly  
used  in  the  US  and  Canada  [URS,  2003].  Variation  in  the  carbon  content  of  fuels  is  
not  accounted  for  in  this  method  and  hence,  these  US-­‐average  values  can  introduce  
error  (discussed  in  section  5.4).  Emission  factors  are  reported  as  units  of  carbon  
dioxide  as  opposed  to  units  of  carbon  and  assume  100%  oxidation  of  fuel  carbon  to  
CO2  for  natural  gas,  99%  for  coal  and  oil  [IPCC  1996;  DOE/EIA  2007b].  

27
 5.2.3  Suspected  database  errors  and  corrections  
The  state  of  Illinois  provided  some  CO  emission  factors  for  LPG  use  in  the  
commercial  sector  that  were  different  from  all  other  emission  factors  for  this  fuel  in  
this  sector.  They  listed  some  values  as  “0.19  lbs  CO/e3gals”  versus  the  consistent  
reporting  in  all  other  states  of  1.9  lbs  CO/e3  gals.  This  latter  value  is  also  consistent  
with  the  default  value.  These  instances  were  changed  from  0.19  to  1.9  lbs  CO/e3gals.  
The  state  of  Alabama  provided  CO  emission  factors  for  bituminous/subbituminous  
coal  use  in  industrial,  residential,  and  commercial  of  0.6  lbs  CO/e6ft3,  6  lbs  CO/e6ft3  
and  11  lbs  CO/e6ft3,  respectively.  This  was  the  only  instance  of  reporting  for  coal  
that  utilized  volumetric  units  in  the  denominator.  Attempts  to  convert  these  units  to  
these  to  mass  units  returned  emission  factors  that  were  clearly  in  error.  In  these  
cases,  the  Vulcan  default  CO  emission  factors  were  used.  
Emissions  for  SCC  210300500  utilizing  residual  oil  in  the  commercial  sector  within  
the  state  of  Alabama  report  emission  units  in  “tons”.  Comparison  to  other  state  
values  for  the  same  fuel  and  technology  suggest  that  this  is  an  input  error  and  the  
correct  units  should  be  “lbs”.  The  reporting  unit  for  these  emissions  has  been  
changed  systematically  to  lbs.  
Data  reported  to  the  nonpoint  NEI  from  across  the  residential  sector  in  the  state  of  
Alabama  has  been  discovered  to  contain  errors  [Cole,  2008].  It  remains  unclear  what  
caused  the  reporting  error  but  CO  emissions  were  discovered  to  be  roughly  5x  too  
large  which  translated  into  CO2  emissions  also  being  roughly  5x  too  large.  Hence,  all  
Alabama  residential  emissions  originating  in  the  nonpoint  data  files  have  been  
reduced  by  a  factor  of  5.  It  is  unclear  whether  or  not  other  reporting  anomalies  
occurred  for  the  state  of  Alabama  (other  than  those  specifically  denoted  here  and  in  
other  sections),  but  the  Vulcan  team  recommends  caution  when  interpreting  the  
Vulcan  CO2  emissions  for  Alabama.  
The  state  of  Connecticut  reported  incorrect  units  on  their  CO  emission  factors  for  all  
natural  gas  processes.  They  reported  as  lbs/e3ft3  when  the  only  rational  
denominator  would  be  e6ft3.  
The  nonpoint  dataset  included  emission  factors  that  were  identified  as  having  
“parsing”  errors.  Emission  factors  were  clearly  identified  as  having  a  leading  “30”  in  
the  first  two  positions  in  the  provided  number.  These  were  parsed  incorrectly  from  
the  time  type  (the  previous  field)  and  this  error  showed  up  consistently  within  a  
state/sector/fuel  combination.  In  these  cases,  the  leading  “30”  was  stripped  from  
the  provided  emission  factor  and  the  remaining  emission  factor  used  in  the  
calculations.  There  was  one  case:  LPG  (mat  code  178)  in  which  the  leading  “30”  was  
real  and  not  an  artifact  of  parsing.  This  was  determined  from  knowledge  of  the  
typical  emission  factor  for  LPG.  In  this  instance  the  leading  “30”  was  not  stripped  
from  the  provided  emission  factor.  
Emissions  for  SCC  2104006000  in  FIPS  45045  (county  Greenville,  South  Carolina)  
constitute  a  variation  on  the  above  correction.  The  original  provided  emission  factor  
was  “30400  lbs/e6ft3”.  After  removing  the  leading  “30”  the  resulting  emission  factor  
is  400  lbs  CO/e6ft3.  Comparison  to  other  state  values  for  the  same  fuel  and  

28
technology  suggest  that  this  is  an  input  error  and  the  correct  emission  factor  value  
should  be  40  lbs  CO/e6ft3.  It  is  worth  noting  that  the  Vulcan  default  emission  factor  
would  be  65  lbs  CO/e6ft3  further  strengthening  the  conclusion  that  400  lbs  CO/e6ft3  
is  an  input  error.  
Emissions  for  SCC  2104006000  (residential  NG;  all  combustor  types)  in  the  state  of  
Utah  report  an  emission  factor  of  40  CO  lbs/e6ft3.  Normalization  by  population  
clearly  shows  a  problem  with  this  emission  factor  and  suggests  that  the  emission  
factor  is  too  low.  It  is  not  currently  known  what  method  the  state  of  Utah  employed  
to  quantify  their  nonpoint  source  emissions  of  CO.  In  order  to  generate  per  capita  
values  that  are  consistent  with  surrounding  states,  the  Vulcan  default  emission  
factor  of  65  lbs  CO/e6ft3  has  been  used.  
A  number  of  records  had  no  sectoral  assignment.  Sectoral  assignments  were  made  
through  comparison  of  the  state  totals  constructed  here  with  those  coming  from  the  
DOE  EIA  (reference).  All  unknown  sectoral  emissions  are  assigned  to  the  
commercial  sector  for  the  states  of  FL,  MI,  and  NM  except  the  unknown  emissions  in  
TN  are  assigned  to  the  industrial  sector.  The  unknown  emissions  in  California  are  
assigned  to  the  nonroad  sector  (5.9624  MtC/year)  and  must  be  performed  offline  to  
the  main  code  infrastructure  due  to  the  fact  that  the  nonraod  sector  is  not  fully  
incorporated  into  the  Vulcan  code.    
5.3  Spatial  Processing  
Nonpoint  CO2  emissions  are  defined  within  the  NEI  at  the  county-­‐scale  and  the  
annual  temporal  scale.  Downscaling  of  the  residential  and  commercial  emissions  (in  
addition  to  the  small  amount  of  industrial  sector  and  electricity  production  
emissions)  reported  in  the  nonpoint  NEI  files  are  performed  through  use  of  census  
tract-­‐level  spatial  surrogates  prepared  by  the  Environmental  Protection  Agency  
[DynTel,  2002].  The  spatial  surrogates  used  are  a  combination  of  different  spatial  
datasets  such  as  Landsat  7  land-­‐use  classification  and  Federal  Emergency  
Management  Agency’s  “HAZUS”  data.  For  the  purposes  of  downscaling  the  Vulcan  
emissions,  multiple  residential,  multiple  commercial  and  multiple  industrial  
building  classes  were  combined  into  a  single  total  floor  square  footage  quantity  for  
the  residential,  commercial  and  industrial  class  at  the  census  tract.  Each  county’s  
CO2  emissions  are  allocated  to  the  US  Census  tracts  within  the  county  according  to  
weighting  by  the  amount  of  residential/commercial/industrial  building  square  
footage  within  each  Census  tract.    
A  small  amount  of  electricity  production  was  present  in  the  nonpoint  data  source.  
This  occurred  in  the  states  of  California,  New  York,  New  Mexico  and  Nevada.  These  
county-­‐level  emissions  were  assigned  to  the  centroid  of  the  county  as  emission  
points.  
This  can  be  further  transformed  to  a  10  km  x  10  km  grid  (see  section  7.0)  by  further  
allocating  the  Census  tract  CO2  emissions  to  the  10  km  x  10  km  grid  through  area-­‐
based  weighting  (the  area-­‐based  percent  share  of  sub-­‐portions  of  each  grid  cell  
residing  in  different  tracts).  This  provides  each  10  km  x  10  km  gridcell  with  a  

29
residential/commercial/industrial  CO2  emission  amount  that  is  based  on  the  share  
of  residential/commercial/industrial  building  square  footage.    
5.4  Sources  of  uncertainty  
The  computation  of  CO2  emissions  in  the  non-­‐point  data  source  includes  a  number  
of  self-­‐reporting  uncertainty  sources  which  we  designate  here  as  “categorical”  and  
“numerical”  uncertainties.  Categorical  uncertainties  include  the  following:  
1. Time  period  designation  
2. Fuel  designation    
3. SCC  designation  
Errors  in  these  information  sources  imply  that  the  state  or  county  office  tasked  with  
estimating  CO  emissions  mis-­‐categorized  the  time  period  for  which  emissions  were  
estimated,  the  fuel  being  consumed  or  the  SCC  code  for  which  the  pollutant  
emissions  were  estimated.  Estimating  the  liklihood  that  categorical  errors  were  
made  is  difficult.  Quantifying  how  that  categorical  error  would  impact  the  final  CO2  
emission  estimate  is  also  difficult.  Given  the  nature  of  the  reporting  (county  and  
state  environmental  professionals  tasked  with  complying  with  air  quality  regulatory  
law)  and  the  difficulty  associated  with  estimating  the  potential  errors,  this  study  
considers  these  sources  of  uncertainty  to  be  low.  Nonetheless,  uncertainty  
associated  with  category  3.  is  attempted  below.    
The  numerical  sources  of  uncertainty  in  the  CO2  calculation  include  the  following:  
1. Pollutant  emission  quantity  reported  
2. Provided  pollutant  emission  factor  
3. Default  pollutant  emission  factor  
4. CO2  emission  factor  (carbon  content  of  fuel)  
5. Heat  content  conversions  
Among  these  uncertainty  sources,  3  through  5:  the  CO2  emission  factor,  the  heat  
content,  and  the  default  pollutant  emission  factor,  can  be  quantified  with  available  
data.  The  first  two  uncertainty  sources  are  difficult  to  quantify.  Unlike  the  
categorical  uncertainties,  however,  these  are  both  much  more  likely  to  contain  
errors  and  those  errors  would  have  a  direct  and  potentially  large  impact  on  the  CO2  
emissions  estimation.  
In  order  to  provide  a  first  order  sense  of  the  impact  of  the  quantifiable  components  
of  the  last  three  sources  of  numerical  uncertainty,  we  take  a  sensitivity  approach.    
The  sensitivity  approach  asks  the  question:  how  wrong  could  the  CO2  emissions  
estimate  be,  given  typical  variations  in  the  underlying  sources  of  uncertainty?  These  
variations  are  conservatively  interpreted  as  a  one  sigma  spread  on  the  central  
estimate  of  the  CO2  emissions  (though  the  variations  described  below  are  likely  
higher  than  a  true  one-­‐sigma  spread  of  an  actual  sample  of  underlying  factors).  
5.4.1  Pollutant  emission  factor  
For  the  default  pollutant  emission  factors,  a  range  of  values  is  used  as  a  form  of  
sensitivity.  The  range  reflects  values  in  the  WebFire  database  as  well  as  a  range  of  

30
values  that  are  self-­‐reported  in  the  NEI  point  database  itself.  For  example,  for  
commercial  non-­‐point  natural  gas  combustion,  values  ranging  from  15  lbs  CO/e6ft3  
to  84  lbs  CO/  e6ft3  are  included  in  the  sensitivity  test.  These  represent  the  highest  
and  lowest  possible  values  for  CO  emissions/unit  fuel  available  in  the  webfire/NEI  
combined  datasets.  The  lower  CO  emission  factor  will  lead  to  a  greater  amount  of  
fuel  consumed  and  a  greater  CO2  emission.  Whereas  the  high  CO  emission  factor  will  
do  the  opposite  (result  in  lower  CO2  emissions).  These  extreme  ranges  are  
considered  2-­‐sigma  errors  and  hence,  the  distance  between  the  central  EF  and  the  hi  
and  lo  extremes  are  halved  to  arrive  at  a  one-­‐sigma  value.  
5.4.2  Heat  and  carbon  content  
As  described  in  section  5.2.1,  pollutant  emissions  that  are  reported  in  mass  or  
volume  units  are  first  converted  to  emission  per  unit  thermal  content  (per  106  btu).  
This  requires  the  use  of  a  heat  content  conversion  which  is  dependent  upon  the  fuel  
considered  as  provided  in  Appendix  A,  Table  A.3.  This  alters  equation  5.2  as  follows,  

  (5.3)  

where  HCf  is  the  heat  content  which  is  only  dependent  upon  the  fuel  consumed  in  
the  combustion  process.    
Fuel  heat  contents  can  vary  substantially  and  are  generally  associated  with  the  
parent  fuel  formation/location  (coal  mine,  oil  well,  etc).  Variation  (one  standard  
deviation)  in  heat  content  is  derived  from  fuel  samples  and  is  described  and  
quantified  in  DOE/EIA  [2007b].  The  largest  variation  in  heat  content  is  found  for  
coal  and  is  derived  from  sampling  coal  from  each  producing  state  destined  for  
power  plants  in  the  United  States.  Depending  upon  coal  rank,  variation  (standard  
deviation  about  the  mean  value)  in  heat  content  ranges  from  4  to  12%.  Additional  
analysis  was  performed  here  by  quantifying  the  variation  in  coal  delivered  to  power  
plants  using  the  DOE/EIA  form  423  database  and  consistent  results  were  found  
[DOE/EIA,  2002a;  2002b].    
Variation  in  heat  content  for  the  remaining  fuel  categories  are  partly  derived  from  
the  DOE/EIA  form  423  database  or  quantitatively  identical  to  the  variation  assigned  
for  the  carbon  coefficient  (CO2  emission  factor).  Variation  in  the  CO2  emission  factor  
is  similarly  derived  from  DOE/EIA  [2007b].  The  largest  variation  is  for  refinery  
gases  (33%).  Variation  in  the  heat  content  and  carbon  content  of  fuel  are  generally  
correlated.  We  treat  them  as  uncorrelated  and  additive.  This  is  likely  a  conservative  
approach.    These  stated  variations  are  considered  a  one-­‐sigma  spread.  
5.4.3  Utilizing  only  default  pollutant  EFs  
Finally,  the  provided  pollutant  emission  factor  can  be  tested  somewhat  by  
substituting  all  provided  pollutant  emission  factors  with  default  factors  in  all  
instances.  This  bypasses  both  the  acceptance  of  provided  emission  factors  and  the  
SCC-­‐specific  WebFire  emission  factor  lookup  and  defaults  to  the  values  in  Appendix  
Table  A.1  and  Table  A.2.  

31
5.4.4  Summary  of  sensitivities  
Hence,  we  have  four  sensitivity  tests:  
1. vary  the  default  pollutant  emission  factors  (hi  and  lo  cases)  
2. vary  the  fuel  heat  content  (hi  and  lo  cases)  
3. vary  the  fuel  carbon  content  (hi  and  lo  cases)  
4. utilize  only  default  emission  factor  
The  first  three  can  be  quantified  in  a  directional  sense  to  arrive  at  a  “hi”  and  “lo”  CO2  
emissions  estimate  whereas  test  4  will  cause  results  to  vary  in  both  numerical  
directions.  Results  are  produced  which  isolates  the  impact  of  each  of  these  tests  and  
the  combination  of  all  four  sensitivity  tests.  The  combination  sensitivity  test  is  as  
follows:  
Low-­‐end  pollutant  emission  factors  +  hi-­‐end  heat  content  +  hi-­‐end  CO2  EF.  
This  combination  senstivity  test  is  run  with  and  without  utilization  of  default  
emission  factors.  

32
6.0  Transportation  CO2  Emissions  
The  transport  sector  contains  three  separate  components:  onroad  emissions  
(mobile  transport  using  designated  roadways),  nonroad  emissions  (e.g.  boats,  
trains,  ATVs)  and  emissions  associated  with  air  travel  (airports  and  airplanes).    
6.1  Onroad  Sources  
The  onroad  mobile  portion  of  the  Vulcan  CO2  emission  inventory  is  constructed  
from  a  series  of  existing  databases  and  modeling  efforts  to  generate  monthly  carbon  
dioxide  (CO2)  emissions  for  the  year  2002  at  the  spatial  scale  of  a  U.S.  county  for  the  
entire  U.S.  The emissions are based on a combination of county-level data from the
National Mobile Inventory Model (NMIM) County Database (NCD) and standard
internal combustion engine stoichiometry from the Mobile6.2 combustion emissions
model [USEPA 2005b; USEPA 2001; Harrington 1998; Gurney et al., 2009]. The  NMIM  
NCD  is  part  of  the  NMIM  software  package  produced  by  the  EPA  [USEPA  2005d].  In  
addition  to  estimating  CO2  emissions  from  transportation,  the  NMIM  provides  the  
information  necessary  to  estimate  criteria  air  pollutant  emissions  and  much  of  the  
data  volume  is  devoted  to  this  objective.    
Further  spatial  allocation  is  performed  in  order  to  place  these  emissions  onto  U.S.  
roads  and  onto  the  common  10  km  x  10  km  spatial  grid  (see  Section  7.0).  Temporal  
allocation,  based  on  traffic  count  data,  is  performed  to  place  these  emissions  into  
hourly  patterns  [Mendoza  et  al.,  in  preparation].  
6.1.1.  Vehicle  Miles  Traveled  
The  Vulcan  onroad  transportation  emissions  calculation  utilizes  the  total  vehicle  
miles  traveled  (VMT)  from  the  National  Mobile  Inventory  Model  (NMIM)  County  
Database  (NCD)  in  which  the  data  is  provided  for  each  combination  of  vehicle  type,  
road  type,  county,  and  month  for  the  year  2002  (see  Appendix  B  for  tabular  
information  describing  these  elements).    
The  VMT  data  has  been  compiled  from  historical  data  obtained  from  the  Federal  
Highway  Administration’s  (FHWA)  Highway  Performance  Monitoring  System  
(HPMS)  [FHWA  2005].  The  data  contained  in  HPMS  is  obtained  from  a  collaboration  
between  State  Highway  Agencies  (SHAs),  local  governments,  and  metropolitan  
planning  organizations  (MPOs).  The  VMT  data  is  a  mixture  of  “universe”,  “expanded  
sample”,  and  “summary”  data.  Universe  data  refers  to  a  limited  set  of  data  items  
reported  for  the  entire  public  road  system,  either  as  individual  or  grouped  road  
length  sections.  For  example,  the  data  for  the  entire  interstate  system  would  be  
considered  universe  data.  Sample  data  is  defined  as  data  reported  for  a  randomly  
selected  sample  of  roadway  links  in  a  road  system.  This  is  the  case  for  minor  
arterial,  and  collector  roads  in  both  urban  and  rural  systems.  These  sections  are  
generally  a  fixed  set  of  road  segments  that  are  monitored  year  to  year  to  create  a  
sample.  Summary  data  is  data  reported  in  aggregate  form  by  road  type.  In  the  case  
of  minor  collector  and  local  roads,  states  are  not  required  to  report  Annual  Average  
Daily  Traffic  (AADT)  except  for  National  Highway  System  (NHS)  sections.  Table  B.5  
(Appendix  B)  shows  the  data  categories  for  selected  HPMS  data.  

33
Reported  HPMS  data  represent  conditions  as  of  December  31st  of  the  data  year  and  
State  highway  agencies  are  required  to  submit  Linear  Referencing  System  (LRS)  
data  and  any  updates  on  a  yearly  basis.  An  LRS  is  used  to  obtain  the  length  of  road  
sections.  While  there  may  be  other  participants  in  the  collection  and  reporting  
process,  the  ultimate  responsibility  for  the  accuracy  and  timely  reporting  of  HPMS  
data  lies  with  each  individual  State  highway  agency.  Sample  Daily  Vehicle  Miles  
Traveled  (DVMT)  are  obtained  by  multiplying  standard  sample  section  AADT  by  the  
section  length  and  by  the  standard  sample  expansion  factor.  The  expansion  factor  is  
an  annual  growth  factor  used  if  the  AADT  is  not  current  for  the  particular  data  year  
and  older  AADTs  are  used.  As  outlined  in  FHWA  [2005],  the  AADT  submitted  for  
each  road  section  as  part  of  HPMS  reporting  must  meet  the  following  criteria  
(quoted  from  document):  
a. Classification  data  are  representative  of  specific  functional  systems.  
b. Each  season  of  the  year  is  represented  in  the  development  of  axle  correction  
factors.  
c. Classification  sessions  are  long  enough  to  account  for  the  changes  in  vehicle  mix  
from  day  to  day.  The  Traffic  Monitoring  Guide  (TMG)  recommends  that  vehicle  
classification  sessions  be  at  least  48-­‐hours.  Data  for  less  than  24  continuous  
hours  is  not  appropriate.  
d. The  total  volume  of  vehicles  observed  is  at  least  equal  to  that  for  an  average  day.  
e. Classification  counts  are  well  distributed  among  rural  and  urban  locations.  
f. Classification  counts  are  collected,  at  a  minimum,  over  a  3-­‐year  cycle,  one-­‐third  
of  the  counts  per  year.  
g. There  are  sufficient  classification  categories  to  represent  vehicles  with  two  to  
seven  axles.  
Though  the  NCD  reports  VMT  at  the  county  level,  the  county  values  are  often  an  
estimate  derived  from  state-­‐level  data  which  is  allocated  to  the  counties  by  road  
type  and  vehicle  type.  
Roads  can  first  be  broadly  classified  into  “rural”  and  “urban”  road  types.  Rural  VMT  
is  quantified  at  the  state  level  for  the  following  six  road  types:    
1) interstate    
2) other  principal  arterial    
3) minor  arterial    
4) major  collector    
5) minor  collector    
6) local    
The  county-­‐level  rural  interstate  VMT  is  derived  from  the  state  level  total  via  a  
simple  fractional  apportionment  based  on  the  relative  mileage,    

  (6.1)  

where    is  the  rural  interstate  (RI)  VMT  in  county  C,    is  the  total  rural  
interstate  VMT  in  state  S,    is  the  total  rural  interstate  mileage  length  in  county  C,  
and    is  the  total  rural  interstate  mileage  in  state  S  [FWHA  2003].  

34
All  other  rural  road  type  VMT  is  derived  from  the  state  level  total  via  a  fractional  
apportionment  based  on  relative  population,  

  (6.2)  

where    is  the  VMT  on  rural  road  type  X  in  county  C,      is  the  total  VMT  
on  rural  road  type  X  in  state  S,  PC  is  the  rural  population  in  county  C  (county  must  
have  some  length  of  road  type  X,  otherwise  PC  is  zero),  PS  is  the  total  rural  
population  in  state  S  (the  sum  of  only  those  counties  with  non-­‐zero  mileage  from  
rural  roadway  type  X)  [USCB  2004].  
The  2002  rural  population  was  estimated  at  the  county  level  by  multiplying  the  
Census  Bureau’s  2002  county-­‐level  intercensal  population  estimates  by  the  ratio  of  
each  county’s  rural  population  in  the  2000  Census  to  its  total  rural  plus  urban  
population.  
Urban  VMT  is  quantified  for  the  following  six  roadway  types:    
1) interstate    
2) other  freeways  
3) other  expressways    
4) other  principal  arterial    
5) collector    
6) local    
The  approach  to  quantifying  county-­‐level  urban  VMT  by  road  type  considers  urban  
areas  in  two  different  classifications:  1)  “large”  –  greater  than  50,000  residents,  and  
2)  “small”  –  less  than  50,000  residents.  Table  HM-­‐71  in  FHWA  [2003]  provides  the  
VMT  from  all  large  urban  areas,  by  road  type,  in  the  U.S..  Many  of  these  large  urban  
areas  stretch  across  multiple  states  and  multiple  counties.  Hence,  in  order  to  
quantify  the  county-­‐level  VMT  from  this  large  urban  area  data,  the  EPA  distributes  
the  large  urban  area’s  VMT  according  to  the  fraction  of  the  urban  area’s  population  
in  each  county,    

  (6.3)  

where    is  the  VMT  of  large  urban  area  UA  on  road  type  X  in  county  C,  
 is  the  total  VMT  of  large  urban  area  UA  on  road  type  X,    is  the  
population  of  large  urban  area  UA  in  county  C  for  road  type  X  (the  county  must  have  
some  length  of  road  type  X  in  large  urban  area  UA,  otherwise    is  zero),    is  
the  population  of  large  urban  area  UA  for  road  type  X  (the  sum  of  only  those  counties  
with  non-­‐zero  mileage  in  large  urban  area  UA  from  road  type  X)  [FHWA  2003;  USCB  
2004b].  
In  order  to  quantify  VMT  at  the  county-­‐level  for  the  small  urban  areas,  the  EPA  first  
quantifies  the  total  small  urban  area  VMT  within  each  U.S.  state  by  subtracting  the  
state-­‐total  large  urban  area  VMT  (the  sum  of  all  VMT  in  large  urban  areas  from  table  

35
HM-­‐71  in  FHWA  [2003])  from  the  total  urban  area  VMT  within  each  state  (found  in  
table  VM-­‐2  in  FHWA  [2003]).  This  provides  a  state  total  VMT  for  small  urban  areas  
disaggregated  by  the  different  urban  road  types.  
The  county’s  share  of  the  small  urban  VMT  on  road  type  X  is,    

  (6.4)  

where    is  the  VMT  of  small  urban  areas  in  county  C  on  road  type  X,    
is  the  VMT  of  small  urban  areas  in  state  S  on  road  type  X  (calculation  described  in  
previous  paragraph),   is  the  small  urban  population  in  county  C  for  road  type  X  
(the  county  must  have  some  length  of  road  type  X  in  small  urban  areas,  otherwise,  
 is  zero)  and    is  the  state-­‐level  small  urban  population  in  state  S  for  road  
type  X.  
In  both  the  large  urban  and  small  urban  VMT  allocation  schemes,  urban  population  
values  are  needed  at  different  scales  and  for  the  year  2002.  Hence,  the  EPA  utilizes  
the  following  approach  in  order  to  estimate  2002  small  and  large  urban  population  
values.  
The  census  2000  state-­‐level  large  urban  population  was  obtained  by  summing  the  
large  urban  area  population  for  all  counties  within  a  state  [USCB  2004b].  This  
population  was  then  subtracted  from  the  census  state-­‐level  total  urban  population  
in  2000  to  obtain  the  state-­‐level  small  urban  population  [UCSB  2004a].    
                    (6.5)  
Where    is  the  state-­‐level  small  urban  center  population,   is  the  state-­‐level  total  
urban  population,  and    is  the  state-­‐level  large  urban  center  population.      
The  county-­‐level  small  urban  population  in  2002  was  calculated  as  the  total  county-­‐
level  urban  population  in  2002  multipled  by  the  ratio  of  small  to  total  urban  county-­‐
level  population  in  2000:    

                  (6.6)  

Where   is  the  2002  small  urban  population  for  county  C,   is  the  2002  
intercensal  total  population  for  county  C,    is  the  2000  small  urban  population  
for  county  C,  and   is  the  2000  total  county  population  for  county  C  for  2000.  
In addition to VMT designation by county and road type, the NCD contains the 2002
VMT allocated to the 28 MOBILE6 vehicle types. The allocation uses the distribution of
the 2002 VMT among the six HPMS vehicle types (found in Table VM-1 of FHWA
[2003]) and a mapping of these HPMS vehicle categories to the 28 MOBILE6 vehicle
types, provided by the EPA’s Office of Transportation and Air Quality (OTAQ) [OTAQ
2007]. The VMT totals for each of the six HPMS vehicle categories (passenger cars,
motorcycles, other 2-axle 4-tire vehicles, single unit 2-axle 6-tire or more trucks,
combination trucks, and buses) were calculated as a fraction of the total VMT. This

36
calculation was performed separately for five groups of roadway classes. EPA assigned
each of the 28 MOBILE6 vehicle types to one of the 6 HPMS vehicle categories (see
Appendix B, Table B.7). Using the default MOBILE6 VMT fractions for 2002, the
MOBILE6 VMT fractions were renormalized among all MOBILE6 vehicle types mapped
to a given HPMS vehicle category. Then the HPMS VMT fractions for each roadway
group were separately multiplied by the renormalized MOBILE6 VMT fractions for all
MOBILE6 vehicle types included within a given HPMS vehicle category. Each of the
VMT records in the 2002 VMT database, at the county/roadway type level of detail was
multiplied by the fraction of VMT in each of the corresponding MOBILE6 vehicle type
categories to obtain total annual VMT at the county/roadway type/MOBILE6 vehicle
type level.
The  VMT  for  twenty-­‐eight  MOBILE6  vehicle  classes  are  aggregated  to  the  more  
commonly  used  twelve  Source  Classification  Code  vehicle  classes.  The  aggregation  
map  is  shown  in  Appendix  B,  Table  B.8.  
Monthly  values  of  the  VMT  for  each  county/vehicle/road  type  combination  are  
achieved  by  multiplying  the  annual  VMT  (in  million  of  miles  traveled)  by  the  
county/vehicle/road-­‐specific  monthly  allocation  factors  (twelve  fractions)  supplied  
within  the  NCD.  157  counties  out  of  3,142  (4.99%)  have  a  specific  VMT  monthly  
allocation  structure.  These  proportions  are  obtained  by  local  transit  authorities  and  
estimate  the  traffic  volume  and  disaggregate  by  road  and  vehicle  type.  If  no  county-­‐
specific  values  are  found,  a  standard  NCD  monthly  allocation  table  is  used.  This  
standard  allocation  table  is  produced  from  accepted  national  average  AADT  values  
(monthly  allocation  specific  to  road  class  and  vehicle  type)  for  a  particular  road  
section  multiplied  by  the  road  section  length  if  a  state  did  not  report  specific  values.    
Appendix  B,  Table  B.9  shows  the  seasonal  VMT  factors  describing  the  VMT  
allocation  by  season  and  Appendix  B,  Table  B.10  shows  the  distribution  of  these  
seasonal  factors  into  monthly  percentages  of  total  annual  VMT  weighted  by  length  
of  month.  
Little  county-­‐specific  monthly  structure  is  available  and  the  average  AADT  values  
are  used  in  nearly  all  counties,  contributing uncertainty to the monthly VMT time
structure..    Uncertainty  in  VMT  allocation  arises  due  to  the  use  of  national  average  
monthly  allocation  for  over  95%  of  the  counties.  Uncertainty  in  the  VMT  itself  is  due  
to  estimation  methods  used  by  local  and  federal  agencies.  Factors  such  as  
malfunctioning  measuring  devices,  heterogeneity  of  the  spatial  allocation  of  
measuring  devices,  and  data  gaps  play  a  role  in  the  errors  associated  with  the  VMT  
and  its  allocation.  
6.1.2  CO2  Emission  Factors  
To  obtain  mobile  CO2  emission  factors  (grams/mile  driven),  EPAs  MOBILE6.2  
mobile  combustion  model  was  utilized  [USEPA  2001;  Harrington  1998].  MOBILE6.2  
uses  inputs  comprising  different  transport  scenarios  in  order  to  obtain  the  
appropriate  mobile  CO2  emission  factors.  A  scenario  consists  of  a  particular  vehicle  
type  combined  with  a  particular  road  type  (which determines mean travel speed; see
Appendix B, Table B.3).  MOBILE6.2  emission  factors  are  derived  from  emissions  

37
tests  conducted  under  standard  conditions  such  as  temperature,  fuel,  and  driving  
cycle.  Emission  factors  further  assume  a  pattern  of  deterioration  in  emission  
performance  over  time  based  on  results  of  standardized  emission  tests  [USEPA  
2003].  There  are  twenty-­‐eight  vehicle  types  and  twelve  road  types  and  in  order  to  
encompass  all  of  them,  168  MOBILE6.2  scenario  runs    would  be  required  for  every  
US  county.  Instead,  eighteen  scenarios  were  run  which  have  been  historically  used  
in  NEI  datasets  and  encompass  the  entirety  of  the  possible  scenarios  while  retaining  
flexibility  (Appendix  B,  Table  B.3).    
Out  of  3,141  counties  in  the  US,  468  counties  have  fleet  information  based  on  state  
vehicle  registration  data.  In  addition  to  these  individual  county-­‐level  reports,  234  
counties  reported  fleet  data  that  utilize  statewide  average  fleet  estimates  rather  
than  county-­‐by-­‐county  estimates.  States  for  which  either  the  entire  or  individual  
counties  reported  fleet  information  are  Arizona  (4),  Delaware  (10),  DC  (11),  Illinois  
(17),  Iowa  (19),  Kentucky  (21),  Maryland  (24),  Massachusetts  (25)*,  Minnesota  
(27),  New  Jersey  (34),  New  York  (36),  Ohio  (39),  Oregon  (41),  Rhode  Island  (44),  
Tennessee  (47),  Texas  (48),  Utah  (49),  Vermont  (50)*,  Virginia  (51)*,  Washington  
(53)*,  Wisconsin  (55)*.  The  asterisk  denotes  those  states  for  which  only  statewide  
average  fleet  information  was  available.  Consequently  about  78%  of  the  counties  
use  a  default  fleet  based  on  a  national  average  which  has  a  fixed  proportion  of  age  
cohorts  for  each  vehicle  class  [USEPA, 2001].  
The  CO2  emission  factors  calculated  above  represent  the  estimated  average  grams  
per  mile  of  CO2  emitted  by  a  vehicle  in  a  particular  road  type  for  a  county.  Each  
county  has  a  VMT  value  for  each  available  road  type  and  vehicle  type  combination.  
The  product  of  VMT  and  the  corresponding  CO2  emission  factor  yields  the  county  
CO2  emissions  for  each  road  type  and  vehicle  type  combination.  The  twenty-­‐eight  
vehicle  classes  are  then  collapsed  to  a  simpler  and  more  commonly-­‐used  twelve  
classes  using  Appendix  B,  Table  B.4.  Six  of  the  vehicle  types  are  light  duty  and  six  are  
heavy  duty.  Five  vehicle  types  use  gasoline  and  seven  use  diesel  as  their  fuel.  Each  
county-­‐specific  fleet  is  therefore  defined  as  the  combination  of  the  vehicle  type  mix  
and  their  respective  VMT.  The  combination  of  the  fleet,  VMT  and  emission  factors  
results  in  a  unique  set  of  CO2  emissions  for  each  vehicle  type,  road  type,  and  month  
within  each  county.  
6.1.3  Time  structure  
The  monthly/county/road/vehicle-­‐specific  CO2  emissions  are  further  subdivided  in  
time  using  traffic  count  data  from  the  Federal  Highway  Administration.  
6.1.3a  Traffic  data  records  
Hourly  traffic  data  at  monitoring  stations  were  obtained  from  the  Federal  Highway  
Administration’s  (FHWA)  permanent  automatic  traffic  recorder  (ATR)  network.  
Permanent  traffic  recorder  data  is  submitted  by  the  state  managing  the  ATR  to  the  
FHWA  within  20  days  after  the  closing  of  each  calendar  month  
[www.fhwa.dot.gov/ohim/tvtw/tvtfaq.cfm].  The  data  from  the  ATRs  are  compiled  
into  a  monthly  publication,  Traffic  Volume  Trends  (TVT)  by  the  FHWA  Office  of  
Highway  Policy  Information  [FHWA  2001b].  The  data  records  from  the  TVT  are  

38
divided  into  four  types:  station  description  data,  traffic  volume  data,  vehicle  
classification  data,  and  truck  weight  data.  Each  type  of  data  has  its  own  
individualized  record  format  and  certain  data  items  are  common  to  all  four  types  of  
records.  For  example,  all  records  contain  a  six-­‐character  station  identification.  This  
allows  States  to  use  a  common  identification  system  for  all  traffic  monitoring  
stations.  This  identification  system  combined  with  the  latitude  and  longitude  values  
enable  geolocation  of  the  stations.  This  allows  traffic  data  to  be  overlaid  on  the  
National  Highway  Planning  Network  (NHPN)  and  similar  systems  [FHWA  2001a].  
In  the  Vulcan  Project,  we  utilize  the  ATR  data  from  the  years  2007  and  2008  –  two  
recent  and  relatively  complete  years  of  data.  These  data  are  combined  as  described  
below  to  create  a  “climatology”  of  traffic  space  and  time  distribution  allocation.  
Two  of  the  four  data  types  present  in  the  ATR  data  are  used  in  the  Vulcan  Project  to  
distribute  onroad  emissions  over  time:  the  station  description  data  and  the  traffic  
volume  data.  The  station  description  data  contains  all  the  information  required  to  
identify  a  station’s  location  such  as  FIPS  State  Code,  Station  ID,  Direction  of  Travel,  
Lane  of  Travel,  Latitude,  and  Longitude  coordinates.  Other  information  such  as  the  
sensor  types  is  also  present.  The  full  list  of  fields  can  be  found  in  Table  6.1.  

Table  6.1:  Station  Description  Record  

Field Columns Width Description
1 1 1 Record Type
2 2-3 2 FIPS State Code
3 4-9 6 Station ID
4 10 1 Direction of Travel Code
5 11 1 Lane of Travel
6 12-13 2 Year of Data
7 14-15 2 Functional Classification Code
8 16 1 Number of Lanes in Direction Indicated
9 17 1 Sample Type for Traffic Volume
10 18 1 Number of Lanes Monitored for Traffic Volume
11 19 1 Method of Traffic Volume Counting
12 20 1 Sample Type for Vehicle Classification
13 21 1 Number of Lanes Monitored for Vehicle Class
14 22 1 Method of Vehicle Classification
15 23 1 Algorithm for Vehicle Classification
16 24-25 2 Classification System for Vehicle Classification
17 26 1 Sample Type for Truck Weight
18 27 1 Number of Lanes Monitored for Truck Weight
19 28 1 Method of Truck Weighing
20 29 1 Calibration of Weighing System
21 30 1 Method of Data Retrieval
22 31 1 Type of Sensor
23 32 1 Second Type of Sensor
24 33 1 Primary Purpose - NEW
25 34-45 12 LRS Identification - NEW
26 46-51 6 LRS Location Point - NEW
27 52-59 8 Latitude - NEW
28 60-68 9 Longitude - NEW
29 69-72 4 SHRP Site Identification - NEW
30 73-78 6 Previous Station ID
31 79-80 2 Year Station Established

39
 32 81-82 2 Year Station Discontinued
33 83-85 3 FIPS County Code
34 86 1 HPMS Sample Type
35 87-98 12 HPMS Sample Identifier
36 99 1 National Highway System - NEW
37 100 1 Posted Route Signing
38 101-108 8 Posted Signed Route Number
39 109 1 Concurrent Route Signing
40 110-117 8 Concurrent Signed Route Number
41 118-167 50 Station Location

The  traffic  volume  data  contains  the  actual  vehicle  count  from  each  station.  The  FIPS  
State  Code  and  Station  Identification  fields  are  used  to  identify  the  location  of  the  
station.  The  other  fields  identify  the  direction  of  travel,  lane  of  travel,  year,  day,  
month,  and  the  hourly  traffic  counts.  Tables  6.2  and  6.3  show  the  possible  values  for  
direction,  and  lane  of  travel  respectively.  Table  6.4  shows  the  full  list  of  fields.  

Table  6.2:  Direction  of  Travel  

Code Direction
1 North
2 Northeast
3 East
4 Southeast
5 South
6 Southwest
7 West
8 Northwest
9 North-South or Northeast-Southwest combined (ATR stations only)
0 East-West or Southeast-Northwest combined (ATR stations only)

Table 6.3: Lane of Travel

Code Lane
0 Data with lanes combined
1 Outside (rightmost) lane
2-9 Other lanes

40
Table 6.4: Hourly Traffic Volume Record
Field Columns Length Description
1 1 1 Record Type
2 2-3 2 FIPS State Code
3 4-5 2 Functional Classification
4 6-11 6 Station Identification
5 12 1 Direction of Travel
6 13 1 Lane of Travel
7 14-15 2 Year of Data
8 16-17 2 Month of Data
9 18-19 2 Day of Data
0 20 1 Day of Week
11 21-25 5 Traffic Volume Counted, 00:01 - 01:00
12 26-30 5 Traffic Volume Counted, 01:01 - 02:00
13 31-35 5 Traffic Volume Counted, 01:01 - 02:00
14 36-40 5 Traffic Volume Counted, 03:01 - 04:00
15 41-45 5 Traffic Volume Counted, 04:01 - 05:00
16 46-50 5 Traffic Volume Counted, 05:01 - 06:00
17 51-55 5 Traffic Volume Counted, 06:01 - 07:00
18 56-60 5 Traffic Volume Counted, 07:01 - 08:00
19 61-65 5 Traffic Volume Counted, 08:01 - 09:00
20 66-70 5 Traffic Volume Counted, 09:01 - 10:00
21 71-75 5 Traffic Volume Counted, 10:01 - 11:00
22 76-80 5 Traffic Volume Counted, 11:01 - 12:00
23 81-85 5 Traffic Volume Counted, 12:01 - 13:00
24 86-90 5 Traffic Volume Counted, 13:01 - 14:00
25 91-95 5 Traffic Volume Counted, 14:01 - 15:00
26 96-100 5 Traffic Volume Counted, 15:01 - 16:00
27 101-105 5 Traffic Volume Counted, 16:01 - 17:00
28 106-110 5 Traffic Volume Counted, 17:01 - 18:00
29 111-115 5 Traffic Volume Counted, 18:01 - 19:00
30 116-120 5 Traffic Volume Counted, 19:01 - 20:00
31 121-125 5 Traffic Volume Counted, 20:01 - 21:00
32 126-130 5 Traffic Volume Counted, 21:01 - 22:00
33 131-135 5 Traffic Volume Counted, 22:01 - 23:00
34 136-140 5 Traffic Volume Counted, 23:01 - 24:00
35 141 1 Restrictions

The  last  field  in  Table  6.4,  “Restrictions”,  was  used  to  evaluate  the  quality  of  the  
data.  A  value  of  “0”  means  that  the  data  from  the  station  has  no  restrictions,  while  a  
value  of  “1”  or  “2”  show  that  there  was  either  construction  or  a  malfunction  of  the  
device.  For  the  years  2008  and  2007,  all  the  data  had  a  value  of  “0”  for  this  field  and  
none  was  discarded.  
6.1.3.b  Data  conditioning  and  gap  filling  
The  ATR  data  for  2007  and  2008  had  5809  and  5774  unique  stations,  respectively.  
There  were  a  small  number  of  unique  stations  in  each  year  with  most  being  
identical.  Furthermore,  after  combining  the  two  years,  there  were  only  4772  
stations  that  could  be  geolocated  using  latitude  and  longitude  coordinates.  Some  
stations  are  located  at  intersections  and  have  data  for  more  than  one  road  type,  
raising  the  number  of  unique  station/road  type  combinations  to  4883.  

41
The  raw  data  was  present  in  the  format  outlined  in  Table  6.4.  There  were  3,407,991  
individual  records  for  2007  and  3,662,160  individual  records  for  2008.  A  record  
details  a  full  24-­‐hour  cycle  of  vehicle  counts  for  a  specific  lane  of  traffic  traveling  in  a  
specific  direction  for  a  particular  station  in  a  particular  day.  The  traffic  counts  for  all  
the  lanes  for  each  direction  and  all  the  directions  for  a  single  station  were  summed  
in  order  to  obtain  a  24-­‐hour  cycle  for  all  lanes  and  directions  combined.  For  each  
station  the  maximum  number  of  lanes  and  directions  was  found  and  any  daily  sum  
record  not  containing  data  from  all  lanes  and  directions  was  removed.  This  was  
done  in  order  to  only  take  days  that  had  a  complete  24  hour  traffic  cycle.  This  
process  resulted  in  an  annual  file  containing  the  traffic  counts  only  for  days  that  
were  fully  populated  with  respect  to  the  maximum  number  of  lanes  and  directions.  
An  external  file  was  created  that  listed  the  days  that  were  present  for  each  station  as  
a  look-­‐up  table  that  will  be  used  in  the  following  step.    Some  stations  are  located  at  
the  intersection  of  two  road  types  and,  as  such,  have  data  for  two  different  road  
types.  Consequently,  each  station  is  uniquely  identified  by  the  state  FIPS  code,  
station  ID,  and  road  type.  The  data  for  the  different  road  types  within  a  station  are  
kept  separate.  
Once  the  station  totals  file  was  created,  an  annual  file  of  hourly  totals  for  each  
station  using  data  from  both  2007  and  2008  was  created.  This  file  is  used  to  create  
the  2002  hourly  traffic  pattern.  There  are  two  temporal  allocation  challenges  when  
combining  2007  and  2008  data  to  make  a  2002  hourly  file;  the  starting  day  for  each  
year,  and  the  leap-­‐year  extra  day  for  2008.  Both  of  these  problems  were  solved  
using  a  day  offset  method.  
Both  2007  and  2002  have  365  days  except  2007  starts  on  Monday  (January  1st)  
while  2002  starts  on  Tuesday.  To  account  for  this,  2007  data  is  offset  by  one  day  
which  means  that  the  data  for  January  1st  is  not  used  and  instead  the  first  day  is  a  
Tuesday  just  like  2002.  This  offset  is  kept  constant  throughout  the  year  which  
means  that  the  last  day  of  data  from  the  2007  dataset,  December  31st,  thus  matches  
2002’s  December  30th.  The  year  2008  also  starts  on  a  Tuesday,  like  2002,  and  there  
is  no  offset  for  January  and  February.  However,  there  is  a  1-­‐day  offset  due  to  the  
leap  year  which  means  that  February  29th  of  2008  maps  to  March  1st  2002.  This  
means  that  the  data  for  December  31st  2008  is  not  used  at  all  like  January  1st  2007.  
The  hourly  totals  for  each  individual  month  are  calculated  at  each  station  by  looking  
at  each  day  of  each  month  and  obtaining  hourly  data  for  each  day,  when  available,  
from  either  2007  or  2008.  If  a  certain  day  of  a  month  has  data  for  only  one  of  the  
two  years,  that  data  is  directly  imported  into  that  month’s  hourly  data.  If  there  is  
data  present  from  both  years,  the  average  hourly  value  is  calculated  for  each  hour  of  
that  day  and  imported  into  the  month’s  hourly  data.  Once  the  month  is  filled  with  
available  data  from  2007  and  2008,  a  sample  week  is  created  from  the  data  
collected.  This  sample  week  is  generated  from  the  average  hourly  values  from  the  
hourly  data  collected  in  the  previous  step.  The  sample  week  is  then  used  to  fill  in  
any  missing  days  in  the  month  in  order  to  obtain  a  full  month  worth  of  data.    
In  the  case  where  there  is  not  a  full  week’s  worth  of  data  that  can  be  used  to  create  
the  sample  week,  a  linear  interpolation  gap-­‐filling  method  is  employed.  A  station  

42
can  not  have  more  than  six  months  of  data  missing  in  order  for  gaps  to  be  filled.  The  
six  months  can  be  continuous  or  there  can  be  several  groups  of  missing  months.  If  
there  are  more  than  six  months  missing,  the  station  is  not  used  in  the  analysis.    
Only  4561  stations  fulfill  the  criteria  of  having  six  or  less  months  of  missing  data.  
Once  a  station  is  accepted,  each  gap  is  marked  by  finding  the  month  prior  to  the  
beginning  of  the  gap  and  the  subsequent  month  to  the  last  month  in  the  gap.  At  
times  this  may  involve  “wrapping-­‐around”  the  year.  For  example,  if  a  gap  extends  
from  January  to  March,  the  prior  month  would  be  December  and  the  posterior  map  
would  be  April.  A  sample  week  is  created  for  the  months  bracketing  the  gap  in  order  
to  have  a  basis  for  the  linear  interpolation.  This  sample  week  is  obtained  by  
averaging  the  values  for  each  hour  of  the  week  for  each  week.  This  means  that  for  
each  month  the  1st  hour  of  Monday  will  consist  of  the  average  of  the  1st  hour  of  all  
the  Mondays  in  the  month.  
Once  the  sample  weeks  for  the  prior  and  posterior  months  are  created,  the  number  
of  each  day  of  the  week  missing  within  the  gap  is  obtained.  For  example,  a  month  
such  as  March  with  31  days  would  have  3  days  that  are  missed  5  times  and  the  
remaining  4  days  will  be  missed  only  4  times.  The  linear  interpolation  is  formed  by  
taking  the  difference  in  traffic  counts  for  each  hour  of  the  day  of  the  sample  week  for  
the  month  prior  and  posterior  to  the  gap,  and  dividing  that  value  by  the  number  of  
missing  days  and  creating  the  linear  “step”.    The  missing  days  are  then  filled  for  each  
subsequent  same  day  and  hour  (such  as  the  1st  hour  of  each  Monday)  by  increasing  
or  decreasing  the  value  of  the  prior  month’s  weekly  traffic  count  by  the  linear  
interpolation  step.  Once  all  the  gaps  are  filled  for  a  station,  fractions  are  created  for  
each  hour  by  taking  the  value  of  each  hour  and  dividing  by  the  sum  of  values  for  all  
the  hours.  
6.1.3c  Application  of  ATR  data  
The  onroad  mobile  fossil  fuel  CO2  emissions  obtained  from  NMIM  NCD/Mobile6.2  
process  are  provided  at  the  monthly  and  county  scale  disaggregated  by  road  and  
vehicle  type.  The  ATR  data  is  used  to  further  downscale  these  estimates  in  space  and  
time.  
For  each  road  type  category  (“functional  classification”  in  Table  6.5),  the  spatial  
distribution  of  traffic  monitoring  stations  is  unevenly  distributed  across  the  country.  
In  order  to  objectively  allocate  monthly  CO2  emissions  to  individual  hours,  a  
“nearest-­‐neighbor”  algorithm  utilizing  Thiessen  polygons,  shown  in  Figure  6.1,  was  
utilized.  Due  to  the  GIS  road  layer  having  only  six  road  type  classifications  while  the  
ATR  stations  have  twelve  road  types,  ATR  road  type  classifications  were  combined  
to  make  six  road  classifications  as  shown  in  Table  6.5.  The  available  number  of  
usable  stations  is  also  listed.  As  can  be  seen,  the  number  of  stations  located  in  urban  
collector  roads  is  very  small  compared  to  the  other  road  types.  Consequently,  it  was  
decided  to  combine  the  stations  for  urban  arterial  and  urban  collector  roads  and  
create  Thiessen  polygons  based  on  the  combination  of  the  two  road  types.  As  a  
result  both  road  types  have  the  same  time  structure.  Several  of  the  stations  in  these  
two  road  classifications  are  present  in  both  the  arterial  and  urban  collector  class  

43
because  they  are  located  at  intersections.  Hence,  combining  these  two  road  
classifications  the  purposes  of  allocating  the  time  distribution  has  limited  impact  on  
the  results.  

Table  6.5:  Functional  Classification  Code  

Code Functional Classification GIS Road GIS Road Layer Available
Type
RURAL 01 Principal Arterial - Interstate 1 Rural Interstate 610
02 Principal Arterial - Other 2 Rural Arterial 1642
06 Minor Arterial
07 Major Collector 3 Rural Collector 400
08 Minor Collector
09 Local System
URBAN 11 Principal Arterial - Interstate 4 Urban Interstate 992
12 Principal Arterial - Other
Freeways or Expressways
14 Principal Arterial - Other 5 Urban Arterial 821
16 Minor Arterial
17 Collector 6 Urban Collector 96
19 Local System

Figure  6.1  –  Thiessen  Polygons  for  Road  Type  1  (principal  arterial  –  interstate)  

44
The  Thiessen  polygons  determine  the  spatial  extent  of  each  ATR  station’s  influence  
on  the  surrounding  roads.  A  superposition  of  these  polygons  with  a  county  map,  the  
2008  Census  GIS  road  atlas,  and  the  Vulcan  10km  x  10km  grid,  determines  what  
fraction  of  the  emissions  from  a  particular  county  is  affected  by  a  particular  
polygon’s  time  structure.  This  superposition  also  determines  allocation  of  emissions  
into  the  Vulcan  10km  x  10km  grid.  Table  6.6  shows  a  sample  of  the  fraction  table  
used  to  allocate  county  emissions  to  the  10km  x  10km  gridded  data  product.

Table 6.6: Sample of the fractional distribution of county emissions using the Thiessen
polygons to distribute the ATR station influence.
ATR Grid i Grid j State County Length in Length in Weight
Station FIPS FIPS Grid County
56000106 199 176 56 1 5.204939 93.9007 0.055430
56000106 199 175 56 1 7.13245 93.9007 0.075957
56000106 200 176 56 1 12.0616 93.9007 0.128449
. . . . . . . .
. . . . . . . .
. . . . . . . .
1000050 364 266 1 125 20.9728 134.101 0.156395
1000050 365 266 1 125 23.614 134.101 0.176091
1000050 366 266 1 125 2.84498 134.101 0.021215

In  order  to  use  the  ATR  data  to  distribute  emissions  over  time,  the  county-­‐level  
monthly  CO2  emissions  are  first  summed  to  obtain  annual  county  totals  (still  
disaggregated  by  road  and  vehicle  type).  The  hourly  fraction  of  the  annual  traffic  
(defined  as  the  sum  of  all  lanes  and  all  directions)  at  each  ATR  station  is  then  used  
to  allocate  the  annual  CO2  emissions,  which  are  spatially  allocated  via  the  Thiessen  
polygons  as  perviously  explained.    
The  first  row  of  Table  6.6  demonstrates  the  allocation  influence  or  weight  of  the  
time  structure  for  a  particular  station  (56000106)  on  a  gridcell  (199,  176)  for  
county  1.    This  is  obtained  by  taking  the  length  of  road  contained  in  the  gridcell  
(5.204939)  and  dividing  that  value  by  the  total  length  of  road  contained  in  the  
county  (93.9007)  yielding  a  value  of  0.055430.  This  means  that  about  5.5%  of  the  
emission  values  from  county  1  will  be  placed  on  cell  199,176  and  follow  the  time  
structure  dictated  by  station  56000106.  
The  hourly-­‐resolved  CO2  emissions  in  each  10km  x  10km  grid  cell  are  therefore  
defined  as:    
  (6.7)  

where    is  the  annual  CO2  emissions  for  road  type  (f)  in  particular  county  
(c),   ,  is  the  hourly  ATR  traffic  volume  fraction  at  hour  (h),  and  
,  is  the  weight  function  (last  column  of  Table  6.6)  which  denotes  the  
fractional  amount  of  road  type  (f)  from  county  (c)  present  on  cell  (x,y)  due  to  the  
superposition  of  the  polygon’s  shape  on  the  county  and  grid  cell  in  question.  This  
fraction  determines  the  effect  of  the  polygon’s  time  structure  on  the  CO2  emissions  
present  in  the  10km  x  10km  cell.    

45
6.1.4  Spatial  Rendering
6.1.4a  Roadway  rendering  
The  first  rendering  allocates  the  hourly/county/road/vehicle-­‐specific  CO2  emissions  
that  are  available  from  the  Vulcan  fossil  fuel  CO2  inventory  onto  roadways  using  a  
GIS  road  atlas  [NTAD  2003]  which  has  all  twelve  road  types  (six  rural  and  six  urban  
subdivisions).  The  crosswalk  table  that  places  the  twelve  road  types  onto  six  road  
types  is  found  in  Appendix B, Table B.13.  The  hourly  sum  of  all  vehicle  classes  on  a  
single  road  class  within  a  county  are  distributed  evenly  over  the  total  road  class  
distance  in  the  county.  This  results  in  a  per  kilometer  amount  of  CO2  emissions  that  
remains  constant  over  space  within  a  county  and  road  class.  Time  variations  are  as  
described  in  section  6.1.3  using  ATR  data.  
Certain  road  classes  in  the  currently-­‐used  GIS  road  atlas  are  not  present  in  all  
counties.  In  some  locations  the  following  road  classes  are  often  missing:  rural  major  
collector,  rural  minor  collector,  rural  local,  urban  minor  arterial,  urban  collector,  
and  urban  local.  Hence,  there  is  a  mismatch  between  the  road  classes  identified  by  
the  Vulcan  onroad  CO2  emissions  and  the  available  road  types.  In  order  to  solve  this  
problem,  we  moved  the  road-­‐specific  rural  CO2  emissions  from  rural  major  collector  
(32.06  MtC/year),  rural  minor  collector  (9.46  MtC/year),  and  rural  local  (21.98  
MtC/year)  to  the  next  coarsest  road  class  -­‐  rural  minor  arterial  in  rural  areas.  
Similarly,  we  moved  the  road-­‐specific  urban  CO2  emissions  from  urban  minor  
arterial  (49.37  MtC/year),  urban  collector  (20.01  MtC/year),  and  urban  local  (33.72  
MtC/year)  to  next  coarsest  road  class  -­‐  urban  principal  arterial-­‐other.  Through  this  
method,  we  are  able  to  render  all  of  the  road-­‐specific  CO2  emissions  to  the  roads  
present  in  the  GIS  road  atlas.  Roughly  168  MtC/year  out  of  our  total  440  MtC/year  
were  moved  upscale  via  this  method.  
This  approach  can  lead  to  some  unrealistic  spatial  anomalies  in  the  vehicle  
emissions.  A  given  road  type  traversing  a  county  boundary  can  exhibit  “jumps”  or  
large  changes  in  CO2  emissions  by  virtue  of  the  fact  that  the  county  emissions  are  
distributed  evenly  on  a  given  road  type  within  each  county  separately  even  though  
the  road  segment  traverses  county  boundaries  with  no  emission  shift  at  the  
boundary.  Similarly,  a  single  roadway  that  changes  from  urban  to  rural,  for  example,  
at  the  edge  of  a  city  or  densely  populated  area  will  also  exhibit  a  sudden  change  in  
CO2  emissions  within  the  Vulcan  inventory,  which  likely  does  not  occur  as  
dramatically  in  the  real  world.    
6.1.4b    Rendering  to  regular  grid  
The  second  rendering  of  the  county-­‐level  mobile  emissions  features  both  a  
geoprocessing  and  visualization  component.  In  order  to  aggregate  mobile  emissions  
into  a  common  10  km  x  10  km  grid  (see  section  7.0),  road  segments  with  their  
emissions  values  must  be  fractured  by  the  edges  of  the  grid  cells,  then  collected  into  
the  cells  to  which  they  belong.  Using  the  border  of  a  grid  cell  to  split  a  road  segment  
with  emission  value  V  results  in  two  road  segments  with  value  V.  If  those  two  
segments  were  then  aggregated  into  their  parent  grid  cells,  drastic  
overmeasurement  would  occur,  with  value  V  being  added  to  the  gridded  sum  twice.  

46
In  order  to  account  for  this,  emissions  values  must  be  smeared  to  road  segments  per  
kilometer  so  that  when  the  segment  is  split  by  a  cell  border  each  resulting  segment's  
total  emission  can  be  recalculated  by  its  new,  shorter  length.  
Within  a  GIS,  all  road  segments  have  their  lengths  calculated  per  kilometer.  The  
emissions  value  assigned  to  each  segment  (based  on  its  road  class  and  parent  
county)  is  then  divided  by  kilometers  to  reach  a  per-­‐kilometer  emissions  value  for  
each  segment.  The  road  segments  are  then  physically  split  by  the  10  km  x  10  km  
grid  cells.  New  length  values  are  calculated  for  each  road  segment  and  new  total  
emissions  are  calculated  by  factoring  the  original  segment's  total  emissions  value  by  
the  percentage  of  its  original  length  now  represented  by  its  fractured  pieces.  A  road  
segment  with  original  emissions  V,  and  length  of  100km  would  have  a  per-­‐kilometer  
value  of  V/100.  Split  at  kilometer  40  by  a  grid  cell,  each  of  the  two  resulting  
segments  would  have  length=40  km,  length=60  km,  respectively.  Knowing  the  
original  value,  V  of  the  segment's  emissions  while  intact,  the  new  segments'  per  
kilometer  values  can  be  calculated  using  the  percentage  of  length  of  the  intact  
segment  now  represented  by  the  fragment.  This  per-­‐kilometer  value  is  then  used  to  
aggregate  into  the  10  km  x  10  km  grid  cells  all  road  segments  now  found  within  
each  cell,  each  of  which  represents  x  kilometers  of  road/road  type  (fragments  from  
one  or  more  counties  that  happen  to  fall  within  the  cell)  that  carry  with  them  a  
certain  per-­‐kilometer  value  of  emission  output.    
6.2  Nonroad  mobile  emissions  
The  nonroad  mobile  emissions  are  derived  from  NMIM  NCD  and  represent  mobile  
sources  that  do  not  travel  on  roads  such  as  trains,  boats,  snowmobiles,  and  
lawnmowers  [USEPA  2005d;  USEPA  2005e].  The  original  446  vehicle  classes  (few  
counties  contain  all  possible  classes,  however)  were  reduced  to  12  through  
grouping  of  like  processes.  Each  can  utilize  4  different  fuel  types  and  some  variation  
by  engine  configuration  is  retained.  As  with  onroad  mobile  emissions,  the  
space/time  resolution  of  the  incoming  data  is  at  the  county  level  and  at  monthly  
timesteps  within  the  year  2002.  
The  SCC  for  nonroad  equipment  always  falls  under  only  one  of  the  segments  in  
Appendix  B,  Table  B.11  corresponding  to  its  most  typical  application,  although  it  
may  be  used  in  other  segments  as  well.    As  an  example,  skid  steer  loaders  are  in  the  
construction  segment,  but  they  may  also  be  used  in  agriculture.  The  fuel  types  
present  in  the  NONROAD  sector  are  gasoline,  diesel,  LPG,  and  CNG.  
The  nonroad  emissions  are  calculated  as  the  product  of  four  provided  data  
elements,  
  (6.8)  

 is  the  monthly  CO2  emission  in  county  c  for  vehicle  type  v,  EFv  is  the  CO2  
emission  factor  in  grams  of  CO2  per  operating  hour  for  vehicle  type  v,  P  is  the  
population  (number  of  individual  vehicles)  of  vehicle  type  v  in  county  c,    is  the  

47
activity  level  (in  hours  per  year)  for  vehicle  type  v  in  county  c,  and   is  the  
 
seasonality  for  each  month  for  vehicle  type  v  in  county  c.  The  seasonality  is  defined  
as  the  fraction  of  the  total  number  of  hours  in  a  year  that  is  allocated  for  each  
month.  
The  emission  factor  data  is  obtained  from  the  USEPA’s  NONROAD  model  (USEPA  
2005f).    The  activity,  seasonality,  and  population  tables  are  obtained  from  the  NMIM  
NCD  which  represents  extensive  data  collection  from  S/L/T’s  and  estimation  
performed  by  the  USEPA  [USEPA  2005e].    
The  NCD  contains  fields  that  may  be  populated  with  the  file  names  of  external  data  
files  containing  state  or  county  data  specific  to  nonroad.  If  alternate  data  files  are  
not  provided,  NMIM  uses  the  default  NONROAD  model  data  files.  NONROAD  
external  data  files  include:  
1. Activity  rates  (including  annual  hours  of  use  and  load  factor)  
2. Temporal  (monthly  and  daily)  allocations  
3. Source  populations.  
4. Growth  indexes  
5. Geographic  allocations  by  equipment  category  
Many  of  the  nonroad  specific  parameters  are  contained  in  the  NONROAD  model  
itself  as  defaults.  Appendix  B,  Table  B.12  details  the  state-­‐specific  data  provided  by  
S/L/T  agencies  used  to  replace  the  NONROAD  model  default  national  average  
values.  
Currently,  the  nonroad  sources  do  not  include  railroad  or  commercial  marine  vessel  
(CMV)  emissions  as  these  were  not  included  in  the  NMIM  NCD.  The  will  be  included  
in  future  versions  of  the  Vulcan  inventory.  
6.3  Aircraft  emissions  
Aircraft  emissions  in  the  Vulcan  inventory  are  derived  from  two  different  datasets.  
The  first  is  the  NEI  airport  datafile  that  reports  emissions  of  CAPs  at  geocoded  
airport  locations  in  the  U.S.  [USEPA  2005e].  As  with  the  other  NEI  datasets,  
emissions  are  classified  according  to  key  fields  such  as  SCC  and  fuel.  The  NEI  airport  
datafile  includes  information  on  3865  airport  facilities.  The  NEI  airport  emission  
data  is  reported  in  units  of  short  tons  of  CO  for  either  the  entire  year  or  a  daily  
average  of  CO  emitted,  also  in  units  of  short  tons.  The  majority  of  airports  operate  
year-­‐round  and  have  emissions  reported  as  an  annual  total  but  some  airports  
operate  only  during  the  months  of  June  through  August  and  the  emissions  are  
reported  as  a  daily  average  value.  The  CO  emissions  are  converted  to  CO2  emissions  
using  the  following  expression,    

  (6.9)  
where  C,  is  the  emitted  amount  of  carbon,  PE  is  the  equivalent  amount  of  
uncontrolled  aircraft  CO  pollutant  emissions,  p  is  the  aircraft  type,  f  is  the  fuel,  PF  is  

48
the  CO  emission  factor  (provided  in  Appendix  A,  Table  A.1),  and  CF  is  the  emission  
factor  associated  with  CO2  (provided  in  Appendix  A,  Table  A.3).  
There  are  five  main  aircraft  types:  General  Aviation,  Military  Aircraft,  Business  
Turboprop,  and  Air  Taxis,  and  Air  Carriers.    Each  of  these  have  specific  emissions  for  
each  airport  and  the  total  emissions  for  an  airport  location  is  the  sum  of  the  
emissions  from  each  aircraft  type  that  uses  the  facility.  All  aircraft  are  assumed  to  
use  jet  fuel  and  the  CO2  emission  factor  used  is  0.0702  tonnes  CO2/1x106  btu.    
Three  CO  emission  factors  are  used:  1.082,  0.944,  and  0.056  lbs  CO/1x106  btu  used  
respectively  for  SCC  codes  containing  "reciprocating"  or  "turbine  aircraft"  in  their  
name,  SCC  codes  containing  "engine"  in  their  name,  and  all  other  SCC  codes  for  
aircraft  using  jet  fuel,  respectively.  
The  second  dataset  utilized  is  the  Aero2K  database  that  quantifies  global  airborne  
emissions  (including  take-­‐off/landing)  on  a  1°  x  1°  x  500  ft  grid  and  includes  
information  on  fuel,  CO2,  CO,  NOx,  H2O,  soot,  hydrocarbons,  and  particulates  for  
commercial  aircraft  and  all  but  CO2  for  military  aircraft  [Eyers  2004].  The  emissions  
are  based  on  flight  path  information  collected  from  commercial  and  military  aircraft.  
The  aircraft  population  was  obtained  from  commercial  airline  data  which  provides  
fleet  information  in  terms  of  aircraft  and  engine  type.  In  order  to  keep  the  database  
to  a  manageable  size,  forty  representative  aircraft  types  were  chosen  which  fit  into  
four  broad  categories:  Large  Jets,  Regional  Jets,  Turboprops,  and  Bizjets.  The  CO2  
emissions  were  obtained  by  multiplying  the  fuel  consumption  of  each  
aircraft/engine  type  by  the  amount  of  distance  travelled  and  the  take-­‐
off/climb/cruise/descent/landing  cycle.  The  fuel  usage  predictions  were  calculated  
using  PIANO  for  the  year  2002  [Piano  2002].  
Fuel  profiling  and  prediction  takes  place  within  the  AERO2k  Data  Integration  Tool  
[Eyers  2004].  The  method  for  assigning  fuel  data  to  the  flight  profiles  in  the  flights  
relies  on  a  series  of  data-­‐tables  as  follows:  
1.  Take-­‐off.  Using  60.9%  of  maximum  payload,  estimate  the  take-­‐off  weight  for  
the  mission  range  to  be  flown.  Taxi,  take-­‐off  and  climb  out  (to  3000  ft)  data  
from  emissions  databank  and  airport-­‐specific  departure  times-­‐in-­‐mode  look-­‐
up  table.  
2.  Climb  (>3000ft).  Determine  initial  cruise  altitude  from  the  profile  data,  
calculate  fuel  used  in  climb  from  climb  data  tables,  re-­‐calculate  aircraft  mass  
at  top  of  climb,  and  calculate  distance  flown.  
3.  Cruise.  Select  appropriate  cruise  fuel  flow  data  from  the  cruise  data  tables,  for  
the  altitude,  Mach  number  and  aircraft  mass.  Continue  to  re-­‐calculate  distance  
flown  and  aircraft  mass  through-­‐out  the  cruise  segment.  
4.  Step-­‐climb  or  mid-­‐cruise  descent  if  appropriate,  then  repeat  Cruise  step.  
5.  Descent  (to  3000  ft).  Descent  fuel  from  final  cruise  altitude  to  3000  ft  
calculated  from  descent  data  tables.  

49
 6.  Landing.  Data  from  emissions  databank  and  airport-­‐specific  arrival  times-­‐in-­‐
mode  lookup  table.  
Aero2k  CO2  emissions  above  3000  feet  are  allocated  to  US  airports  through  a  
proportional  allocation  scheme.  This  procedure  involves  selecting  a  rectangular  
region  encompassing  the  area  in  question  and  integrating  the  emissions  contained  
in  that  area  to  obtain  emissions  aloft.    For  the  Continental  United  States  this  
rectangle  was  drawn  from  50N,  124W  to  23N,  65W.    The  region  for  Alaska  was  
drawn  from  72N,  172E  to  51N,  130W.    Both  of  these  closely  match  the  boundaries  of  
each  region.    Hawaii  had  a  10  degree  buffer  on  all  four  sides  to  account  for  travel  
outward  and  into  the  state;  the  region  was  drawn  from  39N,  171E  to  39N,  144W.    
Only  the  AERO2k  emissions  above  3000  ft  from  both  the  Commercial  and  Military  
aircraft  sector  were  summed  over  these  regions  to  obtain  the  total  for  each  part  of  
the  country.  These  individual  aloft  regional  CO2  emission  totals  were  allocated  to  
surface  airports  via  each  surface  airport’s  share  of  the  regional  total.  AERO2k  
provides  CO2  emissions  estimates  for  commercial  aircraft  but  CO  emission  estimates  
for  military  aircraft.  The  latter  are  converted  to  CO2  emissions  using  default  
emission  factor  values  for  jet  fuel  of  0.963  lbs  CO/1x106  btu  and  0.071  tonnes  
CO2/1x106  btu,  respectively.  This  allows  for  a  direct  comparison  to  independent  
state-­‐level  estimates  that  track  fuel  sales,  such  as  that  performed  by  the  State  
Energy  Data  System  (SEDS)  of  the  DOE/EIA  [DOE/EIA  2007].  However,  for  the  
purposes  of  atmospheric  modeling,  the  emissions  above  3000  ft  are  maintained  as  a  
separate  inventory  in  three  dimensions.  Hence,  the  gridded  hourly  Vulcan  emissions  
surface  files  have  only  those  emissions  associated  with  the  LTO  cycle  as  represented  
in  the  NEI  data.  
6.4  Sources  of  Uncertainty  
TBD  

50
7.0  Sectoral  assignment  and  geospatial  representation  
The  Vulcan  CO2  emissions  are  reported  following  a  number  of  categorical  divisions.  
The  most  common  are  emissions  reporting  by  broad  economic  sectoral  division  
(industrial,  residential,  commercial,  mobile,  utility,  and  cement).    Initially,  a  small  
proportion  of  the  incoming  data  (~7  MtC/year)  could  not  be  classified  as  one  of  the  
six  sectors  but  these  have  since  been  assigned  and  are  notated  elsewhere  in  this  
document  (see  Section  5.2.3).  All  of  these  sectors  are  reported  in  both  the  NEI  point  
and  nonpoint  source  data.  Nearly  all  of  the  onroad  mobile  emission  reporting  is  
found  in  the  NMIM  NCD  data  and  similarly,  nearly  all  of  the  electricity  production  
emissions  are  derived  from  the  geocoded  ETS/CEM  data.    
Geospatial  representation  of  the  Vulcan  inventory  is  performed  in  two  different  
ways.  The  first  is  representation  in  a  “native”  format  or  at  the  spatial  resolution  
most  resembling  the  incoming  data  (points,  county,  etc).  The  second  is  
representation  on  a  common  10  km  x  10  km  grid  to  facilitate  atmospheric  modeling.    
When  representing  the  sectoral  emission  in  a  “native”  format,  a  mixture  of  
resolutions  occur.  For  example,  industrial  sources  are  represented  as  both  geocoded  
points  (as  derived  from  the  NEI  point  source  data  files)  and  as  emission  spread  over  
census  tracts  (in  the  case  of  industrial  emissions  reported  in  the  NEI  nonpoint  
source  data  files  -­‐  see  section  5.3).    A  similar  result  occurs  for  the  electricity  
production  sector  in  which  the  ETS/CEMs  data  is  geocoded  but  some  electricity  
production  emissions  are  present  in  the  nonpoint  source  data  files  and  these  are  
downscaled  similarly  to  the  industrial  sources.    
The  residential  sector  is  derived  from  nonpoint  source  data  only  and  is  therefore  
represented  within  census  tracts  per  section  5.3.  Commercial  emissions  are  derived  
from  both  the  point  and  nonpoint  source  data  and  are  hence,  a  mixture  of  geocoded  
point  locations  and  within  census  tracts  per  section  5.3.  
Nonroad  transportation  emissions  are  distributed  evenly  over  the  county  where  
emissions  are  reported  and  are  hence,  represented  as  county  totals.  Further  spatial  
allocation  will  be  performed  in  future  Vulcan  releases.  
The  NEI  airport  emissions  are  represented  as  geocoded  locations.  However,  
emissions  associated  with  the  airborne  portion  of  this  category,  as  derived  from  the  
Aero2K  inventory  above  3000  feet  are  allocated  to  the  airport  locations  based  on  
each  airport’s  share  of  total  airport  emissions  in  the  airport  NEI.  Aero2K  emissions  
below  3000  feet  are  not  included  as  these  are  considered  the  take-­‐off/landing  
component  of  the  aircraft  emissions  and,  hence,  are  already  included  in  the  NEI  
airport  database.  The  allocation  of  airborne  emissions  to  airport  locations  is  
performed  in  order  to  compare  the  Vulcan  inventory  to  independent  sources  that  
quantify  emissions  according  to  fuel  sales.  From  a  visualization  perspective,  the  
reduction  of  the  airborne  emissions  to  airports  simplifies  the  two-­‐dimensional  
representation  of  the  Vulcan  inventory.  However,  for  the  purposes  of  atmospheric  
modeling,  the  Aero2K  inventory  is  also  maintained  as  a  separate  3D  emission  
dataset  as  a  partner  to  the  NEI  airport  emissions.  

51
All  of  the  sectoral  emissions  are  also  represented  on  the  common  10  km  x  10  km  
grid,  Point  values  are  placed  in  the  grid  cell  occupied  by  the  geocoded  point  source  
while  sources  distributed  across  roads  or  census  tracts  are  placed  within  10  km  x  10  
km  gridcells  via  area-­‐weighted  proportions.    The  center  of  the  first  gridcell  is  
located  at:    -­‐137.16°  W,  51.95°  N  and  the  map  projection  is  Lambert  Conformal  
Conic  with  standard  parallels  of  33.0°,  45.0°,  a  central  meridian  of  -­‐97.0°,  and  a  
latitude  of  projection  origin  of  40.0°.  The  Vulcan  results  have  also  been  transformed  
to  a  0.1°  x  0.1°  grid  and  regridding  information  can  be  found  on  the  Vulcan  website.  
 

52
  
8.0 Temporal Processing
The Residential and Commercial annual emissions as derived from the NEI reflect a mix
of annual level data and portion-year emissions as was noted in section 2.1.2 and 5.1.2
which describe the time period consistency in the point and area source data,
respectively. Though only time type 30 data is retained, some of the incoming data
contains start and end dates that cover sub-portions of the year. The result is that the
initial Vulcan emissions output for these two sectors is not completely “flat” in time but
contains some temporal structure. Given that we utilize independent data (fuel
sales/consumption, heating degree day, etc) to perform temporal structuring, we
“override” the implied time structure provided by the NEI data and spread it evenly over
each hour of the year.
8.1 Monthly downscaling
The next step in conditioning the temporal structure is the monthly downscaling. This is
achieved through the use of monthly, state-level residential and commercial natural gas
sales/consumption fractions based on the Department of Energy/Energy Information
Administration’s (DOE/EIA) form EIA-857 surveys [DOE/EIA 2009].
We focus on natural gas use as a temporal proxy for all space heating because it is the
dominant fuel used in space heating at the end-user point. At the national level, the
Vulcan results indicate that natural gas constitutes 72% of the CO2 emissions in the
residential sector and roughly 65% in the commercial sector. Some fuel oil (distillate –
18% of residential CO2 emissions) and LPG (9% of residential emissions) is used in
isolated portions of the United States and it is assumed that the time structure of that fuel
use for space heating is no different than that constructed for natural gas space heating.
Hence, these temporal proxies are imperfect to the extent that the remaining fuel
consumption in these sectors has a different monthly time structure (currently under
investigation). Natural gas is used in this way because the DOE does not report at the
state/month/sectoral level for liquid or solid fuels. This monthly temporal allocation will
have no sub-state spatial footprint as the EIA data is resolved only at the state level.
The DOE/EIA form-857 surveys are designed to collect data on the quantity and cost of
natural gas delivered to distribution systems and the quantity and revenue of natural gas
delivered to residential and commercial end-user consumers, separately. A sample of
approximately 400 natural gas companies, including interstate pipelines, intrastate
pipelines, and local distribution companies, report to the survey. The form DOE/EIA
form-857 comprises reporting by companies statistically selected by the DOE from a list
of all companies in the US that deliver natural gas to consumers, including pipeline
companies that serve consumers directly. The selection provides a representative sample
of natural gas deliveries to states.
The classification of consumers are as follows:
1) Residential:
o master-metered apartments
o mobile homes

53
 o multi-family dwellings that are individually metered
o and single-family dwellings
uses: natural gas for space heating, water heating and cooking
2) Commercial:
o businesses (eg. Restaurants, hotels, retail)
o federal, state and local governments
o other private and public organizations such as religious, social, and fraternal
groups
uses: natural gas for space heating, water heating cooking and a wide variety of
other equipment.
Commercial use of natural gas is complicated by the fact that a higher percentage is used
for needs other than space heating. However, there is insufficient data to apportion
natural gas in the commercial sector among various uses and it is assumed that the time
structure of total commercial natural gas consumption is an accurate portrayal of the
space heating component. Figure 8.1 presents residential and commercial natural gas
consumption for 2002 in a series of states.

Figure 8.1. Monthly residential and commercial natural gas consumption in a series of states for
the year 2002. Units: 1x106 ft3/month. These are not “adjusted” values (no month length
adjustment performed).
The state/month DOE/EIA natural gas residential and commercial sales/consumption data

54
is converted into a monthly fraction. Details can be found in either (on KRG macintosh):
/KRG_Work/Carbon_Cycle/fossil/datasets/EIA/monthly_fuel/NG.state.month.commerci
al.fracs.xlsx
or
/KRG_Work/Carbon_Cycle/fossil/datasets/EIA/monthly_fuel/NG.state.month.residential.
fracs.xlsx
which are both currently also found on baja3 in a ‘/build_xxx/10k/EIA_time/’ folder as
text files.
These state-level monthly fractions are applied to the hourly, gridded Vulcan fossil fuel
CO2 residential and commercial emissions that emerge from the NEI data (“flattened” to
remove any vestigial temporal structure as noted in sections 2.1.2 and 5.1.2). In order to
apply state-level values to 10k gridcell values, a weightfile outlining what portion of each
10k gridcell resides within a given state domain, is utilized. This file can be found on
Baja3 in ‘/build_xxx/10k/EIA_time/10k_2_state_0813.sorted.prn’. The processing of this
is performed within the ‘make.all.f’ programs in each build.
8.2 Sub-monthly downscaling
In order to reflect sub-monthly temporal variations in space heating fossil fuel CO2
emissions, we relied upon the well-established relationship between space heating needs
and external surface temperature via the heating degree day relationship (Ruth and Lin
2006) defined as:
(8.1)
Where denotes the gridcell, HDDsp is the set point temperature and T represents the
surface air temperature. The set point temperature was chosen as 68 °F, the commonly
established set point from the literature on the topic for a US-average [Ruth and Lin
2006; Amato et al., 2005] and the surface air temperature was taken from the NCEP
North American Regional Reanalysis (NARR) [Mesinger et al., 2006].
The NARR contains surface air temperature every 3 hours on a roughly 0.3°x0.3° (32.46
km in Lambert Conformal) grid for the contiguous US and this was regridded to the
10km x 10km Vulcan grid. This allowed for the computation of an HDD value every 3
hours for every gridcell on the Vulcan grid.
In generating the 3-hourly fractional allocation, two different fuel uses were assigned
based on the categories outlined in the previous section: 1) space heating and 2) other
uses (sum of water heating, cooking and all other miscellaneous uses). For the
commercial sector, the other fuel uses are assumed to be small. Space heating was
defined as varying according to the HDD computation while the other uses were deemed
constant over time based on the observation that water heating is not directly related to
external temperature but to occupancy, shower frequency, etc [Mansur et al., 2008]. The
portion of monthly fossil fuel CO2 emissions resulting from other uses, as a percentage of
the monthly total residential and commercial emissions, was derived from the HDD
calculation:

55
 (8.2)
where Po represents that proportion of the monthly fossil fuel residential or commercial
CO2 emissions allocated to the other uses, m denotes the month, and t denotes the hour.
This assumes that the proportion of fuel devoted to space heating in a month is equal to
the number of hours the surface air temperature falls below HDDsp out of the total
number of hours in a month. Hence, locations where there were many hours below the
HDDsp (e.g. Wisconsin) would have a large proportion of the monthly fuel use devoted to
space heating while locations in which few hours were below the HDDsp (e.g. Florida)
would have relatively small proportions of the month fuel use devoted to space heating.
The proportion of monthly fossil fuel residential or commercial CO2 emission devoted to
space heating is then:
(8.3)
where Psh denotes the space heating proportion.
With these proportions defined, one can calculate the hourly emissions based on the sum
of the hourly CO2 emissions devoted to uses other than space heating and the hourly CO2
emissions devoted to space heating. The latter quantity has a time varying quality which
we reflect by quantifying the variation of the HDD at a given hour about the mean HDD
value for the month. This hourly adjustment factor can be expressed as,

; for t when T(x,t)<HDDsp

(8.4)
where this is only defined at hours where T( ,t) is below the set point value. Hours
where T( ,t) is above the set point value are assigned an adjustment factor value of 0.
This adjustment factor can then be incorporated into the complete hourly calculation to
produce a final hourly CO2 emissions amount:

(8.5)

An entire month in which the surface air temperature never falls below the HDDsp will
exhibit a constant emission throughout the month. Months in which at least a single hour
fell below the HDDsp will have hours in which the fractional allocation value reflects the
constant fraction devoted to the other uses and hours in which the fractional allocation
values represent the sum of a time varying portion (devoted to space heating) and the
constant amount from other uses.

56
Figure 8.2 shows examples of gridcell level emissions in four locations around the US.

Figure 8.2. Daily fossil fuel CO2 emissions in the residential sector at four locations in
the United States. Units: million tonnes C/day.

 
8.3  Multiyear  time  structure  
In  order  to  produce  emissions  for  years  other  than  the  base  year  of  2002,  annual  
sales/consumption  data  from  the  DOE/EIA  is  utilized  [DOE/EIA  2007c].  The  SEDS  
sales/consumption  data  is  organized  by  state,  sector  and  fuel  and  spans  the  1960  to  
2007  time  period.  As  with  the  monthly  state-­‐level  residential  natural  gas  data  
referred  to  previously,  the  basis  of  the  SEDS  sales/consumption  data  are  derived  
from  survey  data  collection  efforts.  Principal  among  these  are  the  data  outlined  in  
the  Annual  Coal  Report,  the  Natural  Gas  Annual  and  the  Petroleum  Supply  Annual  
document  series.  The  strategy  is  to  construct  ratios  of  a  given  year’s  
state/sector/fuel  sales/consumption  relative  to  2002.  These  ratios  are  then  applied  
to  the  2002  Vulcan  hourly  gridded  output  to  construct  a  multiyear  data  product.  
This  implies  a  number  of  approximations:  
1) this  assumes  that  the  time  structure  of  SEDS  sales/consumption  at  the  
state/sector/fuel  level  can  be  directly  mapped  to  the  time  structure  of  the  
resulting  CO2  emissions.  This  would  be  violated  if,  for  example,  the  carbon  
content  of  fuel  at  the  state/sector/fuel  level  varies  over  time.  

57
 2) this  assumes  no  variation  of  sub-­‐state  spatial  distibution  of  CO2  emissions  
over  time  due  to,  for  example,  housing  development,  new  point  sources,  etc.  
The  SEDS  sales/consumption  data  includes  production  within  a  state  that  is  
exported  to  locations  outside  the  state.  When  exports  exceed  in-­‐state  consumption,  
a  negative  sales/consumption  value  results.  However,  without  details  on  
import/export,  it  is  not  possible  to  ascertain  how  much  domestic  consumption  
occurred  in  the  instances  of  negative  entries.  The  exported  sales/consumption  
quantity  will  be  captured  correctly  in  the  entries  for  the  importing  states.  Because  
negative  entries  are  not  useable  for  temporal  structuring,  these  values  are  replaced  
by  zero  entries  wherever  they  occur.  This  is  acknowledged  as  creating  a  potential  
negative  bias  for  the  temporal  structure  in  those  state/sector/fuel  cases  in  which  
negative  entries  occur.    
Because  stockpiling  of  fuel  can  occur  over  time,  the  sales/consumption  values  can  
exhibit  significant  interannual  variiablity  that  is  not  an  refelctive  of  actual  
combustion  in  a  given  year.  This  is  particularly  noticeable  in  the  coal  data.  In  order  
to  attempt  to  account  for  potential  stockpiling,  a  “backward  looking”  exponential  
smoothing  filter  is  applied.  This  filter  transforms  each  year’s  sales/consumption  of  
coal  to  represent  a  diminishing  proportion  of  previous  year’s  original  
sales/consumption  values.  A  five  year  backward-­‐looking  window  is  used.  The  
expression  is  as  follows:  

  (8.6)  

where  E(t)’  is  the  new  emissions  at  timestep  t,  and  E(t)  is  the  original  emissions  at  
timestep  t.  The  window,  w,  designates  the  number  of  years  in  arrears  that  
contribute  to  the  current  year  sales/consumption.  Currenlty,  this  value  is  5.  The  
logic  is  that  a  given  year’s  sales/consumption  is  a  diminishing  contribution  from  
previous  year  values.  
With  a  smoother  in  placed  the  annual  state/sector/fuel-­‐specific  fractions  are  
constructed.  In  instances  in  which  the  baseyear  of  2002  contains  a  zero  value,  we  
simply  transfer  the  2002  vulcan  value  to  all  other  years.  This  is  being  reviewed  for  a  
superior  approach  and  will  be  available  in  future  releases.  
Because  the  Vulcan  fuel  list  is  far  more  detailed  than  the  categories  available  in  the  
SEDS  sales/consumption  data,  a  crosswalk  file  is  constructed  that  maps  every  
Vulcan  fuel/sector  combination  to  a  fuel/sector  combination  in  the  SEDS  
sales/consumption  datafile.  This  is  shown  in  Table  8.1.  

Table  8.1  Fuel  mapping  from  Vulcan  fuel  categories  to  the  SEDS  sales/consumption  
fuel  categories  

Vulcan Fuel Vulcan Fuel Description Sector SEDS Fuel code SEDS Fuel Description
2 Waste Oil COM 279 Residual Oil
2 Waste Oil IND 216 Oil
2 Waste Oil UTL 279 Residual Oil
44 Diesel MOB 56 Distillate Oil

58
 44 Diesel UTL 56 Distillate Oil
44 Diesel IND 56 Distillate Oil
57 Distillate Oil (Diesel) COM 56 Distillate Oil
57 Distillate Oil (Diesel) IND 56 Distillate Oil
57 Distillate Oil (Diesel) UTL 56 Distillate Oil
58 Distillate Oil (No. 1 & 2) IND 56 Distillate Oil
126 Gas IND 209 Natural Gas
159 JetFuel RES 162 Kerosene
159 JetFuel IND 162 Kerosene
159 JetFuel UTL 56 Distillate Oil
160 Jet Naphta COM 162 Kerosene
160 Jet Naphta IND 162 Kerosene
173 Lignite IND 717 Coal
216 Oil COM 56 Distillate Oil
216 Oil RES 56 Distillate Oil
251 Process Gas IND 717 Coal
251 Process Gas COM 717 Coal
251 Process Gas UTL 717 Coal
255 Propane IND 178 LPG
255 Propane COM 178 LPG
255 Propane RES 178 LPG
255 Propane UTL 209 Natural Gas
256 Propane/Butane IND 178 LPG
256 Propane/Butane COM 178 LPG
279 Residual oil RES 56 Distillate Oil
323 Subbituminous Coal IND 717 Coal
323 Subbituminous Coal COM 717 Coal
323 Subbituminous Coal UTL 717 Coal
374 Crude Oil IND 279 Residual Oil
425 Coke Oven Gas IND 279 Petroleum Products
640 Antracite RES 717 Coal
640 Antracite COM 717 Coal
640 Antracite IND 717 Coal
640 Antracite UTL 717 Coal
663 Bituminous Coal IND 717 Coal
663 Bituminous Coal COM 717 Coal
663 Bituminous Coal UTL 717 Coal
664 Bituminous/Subbituminous Coal RES 717 Coal
664 Bituminous/Subbituminous Coal COM 717 Coal
664 Bituminous/Subbituminous Coal IND 717 Coal
664 Bituminous/Subbituminous Coal UTL 717 Coal
675 Butane IND 178 LPG
675 Butane COM 178 LPG
724 Coke IND 717 Coal
809 Coke Oven or Blast Furnace Gas IND 717 Coal
818 Diesel/Kerosene IND 162 Kerosene
823 Distillate Oil (No. 1 & 2) IND 56 Distillate Oil
823 Distillate Oil (No. 1 & 2) UTL 56 Distillate Oil
823 Distillate Oil (No. 1 & 2) COM 56 Distillate Oil
825 Distillate Oil (No. 4) IND 56 Distillate Oil
825 Distillate Oil (No. 4) UTL 56 Distillate Oil
825 Distillate Oil (No. 4) COM 56 Distillate Oil
864 Jet A Fuel IND 162 Kerosene
865 Jet A Kerosene IND 162 Kerosene
922 Residual Oil (No. 5) COM 279 Residual Oil
922 Residual Oil (No. 5) IND 279 Residual Oil
922 Residual Oil (No. 5) UTL 279 Residual Oil
923 Residual Oil (No. 6) IND 279 Residual Oil
923 Residual Oil (No. 6) COM 279 Residual Oil
923 Residual Oil (No. 6) UTL 279 Residual Oil
924 Residual/Crude Oil IND 279 Residual Oil

59
References  
Ackerman,  K.  V.,  and  E.  T.  Sundquist  (2008),  Comparison  of  two  U.S.  power-­‐plant  carbon  
dioxide  emissions  data  sets,  Environ.  Sci.  Technol.,  42(15),  5688  –  5693,  
doi:10.1021/es800221q.    
Amato,  A.D.,  M.  Ruth,  P.  Kirshen,  and  J.  Horwitz  (2005),  Regional  energy  demand  responses  
to  climate  change:  methodology  and  application  to  the  Commonwealth  of  Massachusetts,  
Climatic  Change,  71,  175-­‐201  
American  Petroleum  Institute  (2004),  Compendium  of  Greenhouse  Gas  Emissions  
Emthodologies  for  the  Oil  and  Gas  Industry,  February  2004.  
Andres,  R.J.  Marland,  G.,  Fung,  I.  and  E.  Matthews  (1996),  A  1  x  1  distribution  of  carbon  
dioxide    emissions  from  fossil  fuel  consumption  and  cement  manufacture,  1950-­‐1990,  GBC,  
10  (3),  419-­‐429.  
Andres,  R.J.,  D.J.  Fielding,  G.  Marland,  T.A.  Boden,  N.  Kumar  and  A.T.  Kearney  (1999),  Carbon  
dioxide  emissions  from  fossil-­‐fuel  use,  1751-­‐1950,  Tellus,  51B,  759-­‐765.  Blasing,  T.J.,  C.T.  
Broniak,  and  G.  Marland  (2005),  the  annual  cycle  of  fossil-­‐fuel  carbon  dioxide  emissions  in  
the  United  States,  Tellus,  57B,  107-­‐115.  
Berry,  R.S.,  J.C.  Martin,  and  C.E.  Dene  (1998)  CEMS  analyzer  bias  and  linearity  effects  study  
(CABLE  study),  RMB  Consulting  &  Research,  Inc.  
Code  of  Federal  Regulations  (2000),  Protection  of  Environment,  Environmental  Protection  
Agency,  40  CFR  Part  75,  Continuous  Emission  Monitoring,  Code  of  Federal  Regulations,  
Revised  as  of  July  1,  2000.  
Cole,  Lisa  Personal  Communication,  Alabama  Department  of  Environmental  Management,  
Alabama,  2008.  
Crisp,  D.,  R.M.  Atlas,  F.-­‐M.  Breon,  L.R.  Brown,  J.P.  Burrows,  P.  Ciais,  B.J.  Connor,  S.C.  Doney,  
I.Y.  Fung,  D.J.  Jacob,  C.E.  Miller,  D.  O'Brien,  S.  Pawson,  J.T.  Randerson,  P.  Rayner,  R.J.  
Salawitch,  S.P.  Sander,  B.  Sen,  G.L.  Stephens,  P.P.  Tans,  G.C.  Toon,  P.O.  Wennberg,  S.C.  Wofsy,  
Y.L.  Yung,  Z.  Kuang,  B.  Chudasama,  G.  Sprague,  B.  Weiss,  R.  Pollock,  D.  Kenyon  and  S.  Schroll  
(2004),  The  Orbiting  Carbon  Observatory  (OCO)  mission,  Advances  in  Space  Research,  34,  
(4),  700-­‐709.  
Denning,  A.S.  et  al.  (2005),  Science  Implementation  Strategy  for  the  North  American  Carbon  
Program.  Report  of  the  NACP  Implementation  Strategy  Group  of  the  U.S.  Carbon  Cycle  
Interagency  Working  Group.    Washington,  DC:  U.S.  Carbon  Cycle  Science  Program,  68  pp.  
DOE/EIA  (2002a),  Form  FERC-­423  Database,  Monthly  Cost  and  Quality  of  Fuels  for  Electric  
Plants  Data;  EIA:  Washington,  DC,  2002;  
www.eia.doe.gov/cneaf/electricity/page/ferc423.html.  
DOE/EIA  (2002b),  Form  FERC-­423  Database,  Monthly  Non-­utility  Fuel  Receipts  and  Fuel  
Quality  Data;  EIA:  Washington,  DC,  2002;  
www.eia.doe.gov/cneaf/electricity/page/eia423.html.  
DOE/EIA  (2003),  Steam-­Electric  Plant  Operation  and  Design  Report,  Form  EIA-­‐767,  data  files  
for  1985-­‐2002,  United  States  Department  of  Energy,  Energy  Information  Administration,  
Washington,  D.C.  

60
DOE/EIA  (2007a),  Emission  of  Greenhouse  Gases  in  the  United  States  2006,  Energy  
Information  Administration,  Office  of  Integrated  Analysis  and  Forecasting,  U.S.  Department  
of  Energy,  Washington,  DC  20585,  DOE/EIA-­‐0573.  
DOE/EIA  (2007b),  Documentation  for  Emissions  of  Greenhouse  Gases  in  the  United  States  
2006,  Energy  Information  Agency,  Office  of  Integrated  Analysis  and  Forecasting,  U.S.  
Department  of  Energy,  Washington,  DC  20585,  DOE/EIA-­‐0638.  
DOE/EIA  (2007c),  State  Energy  Data  System  2007:  Consumption,  Energy  Information  
Administration,  Washington,  DC  20585,  DOE/EIA-­‐0638.  Data  portal:  
http://www.eia.doe.gov/emeu/states/_seds.html    
DOE/EIA  (2008),  Form  EIA-­‐906  and  EIA-­‐920  Databases,  derived  from  reporting  in  the  
Electric  Power  Monthly  series,  
http://www.eia.doe.gov/cneaf/electricity/page/eia906_920.html.  
DOE/EIA  (2009),  Natural  Gas  Monthly,  March  2009,  Energy  Information  Administration,  
Office  of  Oil  and  Gas,  U.S.  Department  of  Energy,  Washington,  DC  20585,  DOE/EIA-­‐0130  
(2009/03).  
DOE/EIA  (2010),  Electric  Power  Annual  2008,  U.S.  Energy  Information  Administration,  
Office  of  Coal,  Nuclear,  Electric  and  Alternate  Fuels,  U.S.  Department  of  Energy,  Washington  
DC  20585,  DOE/EIA-­‐0348(2008),  January  2010.  
DynTel  (2002)  Spatial  Allocation  Information  Improvements,  Technical  Memorandum,  
Review  of  Existing  Data  Sources,  Work  Order  25.6,  January  18,  2002.  Accessed  at  
www.epa.gov/ttn/chief/emch/spatial/newsurrogate.html.  
Eastern  Research  Group,  Inc  (2001a),  Introduction  to  Stationary  Point  Source  Emissions  
Inventory  Development,  Volume  II,  Chapter  1,  prepared  for:  Point  Sources  Committee,  
Emission  Inventory  Improvement  Program,  May  2001.  
Eastern  Research  Group,  Inc  (2001b),  Introduction  to  Area  Source  Emissions  Inventory  
Development,  Volume  III,  Chapter  1,  prepared  for:  Area  Sources  Committee,  Emission  
Inventory  Improvement  Program,  January  2001.  
Eastern  Research  Group,  Inc  and  E.H.  Pechan  &  Associates,  Inc  (2004),  Documentation  for  
the  2002  Electric  Generating  Unit  (EGU)  National  Emissions  Inventory  (NEI),  prepared  for  
Emission  Factor  and  Inventory  Group  (D205-­‐01),  Emissions  Monitoring  and  Analysis  
Division,  U.S.  Environmental  Protection  Agency,  Research  Triangle  Park,  NC  27711,  
September  2004.  
Eyers,  C.J.,  Norman,  P.,  Middel,  J.,  Plohr,  M.,  Michot.,  S.,  Atkinson,  K.,  Christou,  R.A.  (2004),  
AERO2k  Global  Aviation  Emissions  Inventories  for  2002  and  2025,  Center  for  Air  Transport  
and  the  Environment,  QINETIQ/04/01113,  December.  
Federal  Highway  Administration  (2001a)  Traffic  Monitoring  Guide,  U.S.  Department  of  
Transportation,  Federal  Highway  Administration,  Office  of  Highway  Policy  Information  
http://www.fhwa.dot.gov/ohim/tmguide/tmg6.htm#tab631  
FHWA  (2001b)  Traffic  Volume  Trends,  U.S.  Department  of  Transportation,  Federal  Highway  
Administration,  Office  of  Highway  Policy  Information.    December  2008.  
Federal  Highway  Administration  (2003),  Highway  Statistics  2002,  U.S.  Department  of  
Transportation,  Washington.  DC,  2003.  
Federal  Highway  Administration  (2005),  Office  of  Highway  Policy  Information,  Highway  

61
Performance  Monitoring  System  for  the  Continuing  Analytical  and  Statistical  Database,  OMB  
No.  21250028,  May  2005.  Accessed  June,  2008:  
www.fhwa.dot.gov/ohim/hpmsmanl/hpms.htm  
Gregg,  J.S.  and  R.J.  Andres  (2008),  A  method  for  improving  temporal  and  spatial  resolution  
of  carbon  dioxide  emissions  in  the  United  States,  Tellus  68B,  1-­‐10.  
Gurney,  K.R.  et  al.  (2007),  “Research  needs  for  process-­‐driven,  finely  resolved  fossil  fuel  
carbon  dioxide  emissions,”  EOS  Trans.,  88  (49),  doi:  10.1029/2007EO490008.  
Gurney,  K.R.,  R.M.  Law,  A.S  Denning,  P.J.  Rayner,  D.  Baker,  P.  Bousquet,  L.  Bruhwiler,  Y.H.  
Chen,  P.  Ciais,  S.  Fan,  I.Y.  Fung,  M.  Gloor,  M.  Heimann,  K.  Higuchi,  J.  John,  T.  Maki,  S.  
Maksyutov,  K.  Masarie,  P.  Peylin,  M.  Prather,  B.C.  Pak,  J.  Randerson,  J.  Sarmiento,  S.  Taguchi,  
T.  Takahashi,  C.W.  Yuen,  (2002),  Towards  robust  regional  estimates  of    CO2  sources  and  
sinks  using  atmospheric  transport  models,  Nature,  415,  626-­‐630.  
Gurney,  K.R.,  Y.H.  Chen,  T.  Maki,  S.R.  Kawa,  A.  Andrews,  Z.  Zhu,  (2005),  Sensitivity  of  
Atmospheric  CO2  Inversion  to  Seasonal  and  Interannual  Variations  in  Fossil  Fuel  Emissions,  
J.  Geophys.  Res.  110  (D10),  10308-­‐10321.  
Harrigton,  W.,  (1998),  A  Behavioral  Analysis  of  EPA’s  MOBILE  Emission  Factor  Model,  
Discussion  Paper  98-­‐47,  Resources  for  the  Future,  September  1998,  Washington  DC.    
ICF  Consulting  (date  unknown)  Emissions  inventory  reconciliation  utility  sectory,  prepared  
for  Western  Governor’s  Association  Western  Regional  Air  Partnership.  
IPCC  (1996),  Revised  1996  IPCC  Guidelines  for  National  Greenhouse  Gas  Inventories,  
Reference  Manual  (Volume  3),  United  Nations  Environment  Programme,  the  Organizaqtion  
for  Economic  Cooperation  and  Development,  the  International  Energy  Agency,  and  the  
Intergovernmental  Panel  on  Climate  Change,  1996.  
Lieberman,  J.  and  M.  Warner  (2007),  America’s  Climate  Security  Act  of  2007,  S.2191,  110th  
Congress,  U.S.  Senate.  
Mansur,  E.T.,  R.  Mendelsoh,  and  W.  Morrison,  (2008)  Climate  change  adaptation:  A  study  of  
fuel  choice  and  consumption  in  the  US  energy  sector,  J.  Env.  Econ  Mngmnt.,  55,  175-­‐193.  
Marland,  G.  and  R.M.  Rotty  (1984),  Carbon  dioxide  emissions  from  fossil  fuels:  a  procedure  
for  estimation  and  results  for  1950-­‐1982,  Tellus,  36B,  232-­‐261.  
Marland,  G.,  R.M  Rotty,  and  N.L.  Treat  (1985),  CO2  from  fossil  fuel  burning:  global  
distribution  of  emissions,  Tellus,  37B,  243-­‐258.  
Marr,  L.  C.,  D.  R.  Black  and  R.  A.  Harley  (2002).  Formation  of  photochemical  air  pollution  in  
central  California  -­‐  1.  Development  of  a  revised  motor  vehicle  emission  inventory  -­‐  art.  no.  
4047.  Journal  of  Geophysical  Research-­Atmospheres  107(D5-­‐D6):  4047.  
Mesinger,  F.  et  al.  (2006)  North  American  Regional  Reanalysis,  Bull.  Amer.  Met.  Society,  
March,  doi:  10.1175/BAMS-­‐87-­‐3-­‐343.  
NTAD  (2003).  A  collection  of  spatial  data  for  use  in  GIS-­‐based  applications.  Washington,  
D.C.:  The  Bureau.  worldcat.org/oclc/52933703&referer=one_hit
OTAQ  (2007)  Documentation  for  the  final  2002  Mobile  National  Emissions  Inventory,  
Version  3,  Assessment  and  Standards  Division,  Office  of  Transportation  and  Air  Quality  U.S.  
Environmental  Protection  Agency  and  E.H.  Pechan  &  Associates,  Inc.,  Research  Triangle  
Park,  NC  27711,  September  2007,  EPA68-­‐D-­‐02-­‐063.  

62
Portland  Cement  Company,  Economic  Research  Department  (2006)  U.S.  and  Canadian  
Portland  Cement  Industry  Plant  Information  Summary,  Portland  Cement  Association,  Skokie,  
IL.  
Petron,  G.  P.  Tans,  G.  Frost,  D.  Chao,  and  M.  Trainer  (2008),  High  resolution  emissions  of  CO2  
from  power  generation  in  the  USA,  J.  Geophys.  Res.,  113  doi:10.1029/2007/JG000602.  
PIANO  (2002)  Project  Interactive  Analysis  and  Optimisation  aircraft  performance  tool,  see  
http://www.lissys.demon.co.uk  
Rotty,  R.  (1983),  Distribution  of  and  changes  in  industrial  carbon  dioxide  production,  JGR,  
88,  C2,  1301-­‐1308.  
Ruth,  M.,  and  A-­‐C.  Lin,  (2006)  Regional  energy  demand  and  adaptations  to  climate  change:  
Methodology  and  application  to  the  state  of  Maryland,  USA,  Energy  Policy,  34,  2820-­‐2833.  
URS  (2003),  Greenhouse  Gas  Emission  Factor  Review,  Final  Technical  Memorandum,  URS  
Corporation,  Austin,  Texas,  February  3.  
USCB  (2000).  Census  Regions  and  Divisions  of  the  United  States,  U.S.  Department  of  
Commerce,  Economics,  and  Statistics  Administration,  U.S.  Census  Bureau.  
USCB  (2004a),  U.S.  Census  Bureau,  Population  Division,  Population  Estimates,  Subcounty  
population  dataset,  downloaded  at  
www.census.gov/popest/archives/2000s/vintage_2002/CO-­‐EST2002-­‐01/    
USCB  (2004b),  U.S.  Census  Bureau,  Census  2000  Summary  File  1  (SF  1)  100-­Percent  Data,  
Table  P2  Urban  and  Rural  Population,  downloaded  at  
www.census.gov/support/SF1ASCII.html    
USEPA  (1997),  Procedures  for  Preparing  Emission  Factor  Documents,  Office  of  Air  Quality  
Planning  and  Standards,  Office  of  Air  and  Radiation,  U.S.  Environmental  Protection  Agency,  
Research  Triangle  Park,  NC  27711,  November  1997,  EPA-­‐454/R-­‐95-­‐015  Revised.  
USEPA  (2001),  Fleet  Characterization  Data  for  MOBILE6:  Development  and  Use  of  Age  
Distributions,  Average  Annual  Mileage  Accumulation  Rates,  and  Projected  Vehicle  Counts  for  
Use  in  MOBILE6,  United  States  Environmental  Protection  Agency,  EPA420-­‐R-­‐01-­‐047,  
September.    
USEPA  (2002),  Code  of  Federal  Regulations,  40  CFR  Part  51,  Consolidated  Emissions  
Reporting,  Final  Rule,  Federal  Register,  Vol  67  (111),  June  10.  
USEPA  (2003),  User’s  Guide  to  MOBILE6.1  and  MOBILE6.2  Mobile  Source  Emission  Factor  
Model,  United  States  Environmental  Protection  Agency,  EPA420-­‐R-­‐03-­‐010,  August.  
USEPA  (2004a),  2002  Emission  Tracking  System/Continuous  Emissions  Monitoring  
(ETS/CEM)  Annual  and  Ozone  Season  Data  Files,  U.S.  Environmental  Protection  
Agency/Clean  Air  Markets  Division  (CAMD),  Washington,  D.C.  
USEPA  (2004b),  Readme  file  for  February  2004  version  of  SCC  List,  downloaded  from  www.  
epa.gov/ttn/chief/codes/readme_scc.txt  
USEPA  (2005a),  Emissions  Inventory  Guidance  for  Implementation  of  Ozone  and  Particulate  
Matter  National  Ambient  Air  Quality  Standards  (NAAQS  and  Regional  Haze  Regulations),  
Emissions  Inventory  Group,  Emissions,  Monitoring  and  Analysis  Division,  Office  of  Air  
Quality  Planning  and  Standards,  U.S.  Environmental  Protection  Agency,  Research  Triangle  
Park,  NC  27711,  EPA-­‐454/R-­‐05-­‐001,  August.  

63
USEPA  (2005b),  Plain  English  Guide  to  the  Part  75  Rule,  U.S.  Environmental  Protection  
Agency,  Clean  Air  Markets  Division.  
USEPA  (2005c)  NEI  Quality  Assurance  and  Data  Augmentation  for  Point  Sources,  U.S.  
Environmental  Protection  Agency,  Emissions  Monitoring  and  Anlysis  Division,  Emission  
Inventory  Group,  Research  Triangle  Park,  NC,  27711,  February  1.  
USEPA  (2005d),  EPA’s  National  Mobile  Inventory  Model  (NMIM),  A  consolidated  emissions  
modeling  system  for  MOBILE6  and  NONROAD,  Office  of  Transportation  and  Air  Quality,  
Assessment  and  Standards  Division,  Office  of  Transportation  and  Air  Quality,  U.S.  
Environmental  Protection  Agency,  EPA420-­‐R-­‐05-­‐024,  December.  
USEPA  (2005e),  Documentation  for  Aircraft,  Commercial  Marine  Vessel,  Locomotive,  and  
Other  Nonroad  Components  of  the  National  Emissions  Inventory,    Volume  I  –  Methodology,  
Emission  Factor  and  Inventory  Group  (D205-­‐01),  Emissions,  Monitoring  and  Analysis  
Division,  U.S.  Environmental  Protection  Agency,  Research  Triangle  Park,  North  Carolina  
27711,  EPA  Contract  No.:  68-­‐D-­‐02-­‐063,  September.  
USEPA  (2005f),  User’s  Guide  for  the  Final  NONROAD2005  Model,  Assessment  and  Standards,  
Division  Office  of  Transportation  and  Air  Quality  U.S.  Environmental  Protection  Agency,  
December  
USEPA  (2006a),  Documentation  for  the  Final  2002  Point  Source  National  Emissions  Inventory,  
Emission  Inventory  and  Analysis  Group,  Air  Quality  and  Anlysis  Division,  U.S.  
Environmental  Protection  Agency,  Research  Triangle  Park,  NC  27711,  February  10,  2006.  
USEPA  (2006b),  Draft  Detailed  Procedures  for  Preparing  Emissions  Factors,  U.S.  
Environmental  Protection  Agency,  Office  of  Air  Quality  Planning  and  Standards,  Emissions  
Monitoring  and  Analysis  Division,  Emissions  Factors  and  Policy  Applications  Group,  
Research  Triangle  Park,  NC  27711,  June  29.  
USEPA  (2006c),  Documentation  for  the  Final  2002  Nonpoint  Sector  (Feb  06  version)  National  
Emission  Inventory  for  Criteria  and  Hazardous  Air  Pollutants,  Emissions  Inventory  and  
Analysis  Group,  Air  Quality  Division,  Office  of  Air  Quality  Planning  and  Standards,  U.S.  
Environmental  Protection  Agency,  Research  Triangle  Park,  NC  27711,  July  2006.  
USEPA  (2008a),  Clean  Air  Markets  –  Data  and  Maps,  U.S.  Environmental  Protection  Agency,  
http://camddataandmaps.epa.gov/gdm/index.cfm,  accessed  June  10,  2008.  
USEPA  (2008b),  Envirofacts  Multisystem,  U.S.  Environmental  Protection  Agency,  
http://www.epa.gov/enviro/html/multisystem_query_java.html,  accessed  June  10,  2008.  
USGS  (2003),  Minerals  Yearbook,  Vol.  1,  Metals  and  Minerals,  2002.  U.S.  Geological  Survey.  
U.S.  Department  of  the  Interior.  July  2003  
van  Oss,  H.G.  (2005)  Background  facts  and  issues  concerning  cement  and  cement  data,  U.S.  
Geological  Survey,  U.S.  Department  of  the  Interior,  report  2005-­‐1152.  
WebFIRE  (2005),  Technology  Transfer  Network  Clearinghouse  for  Inventories  &  Emissions  
Factor,  WebFIRE,  December  2005,  
http://cfpub.epa.gov/oarweb/index.cfm?action=fire.main,  accessed  June  10,  2008.  
World  Resources  Institute/World  Business  Council  for  Sustainable  Development  (2001),  
Calculating  CO2  Emissions  from  the  Combustion  of  Standard  Fuels  and  from  
Electricity/Steam  Purchase.  Guide  to  Calculation  Worksheets,  October  2001.  Files  accessed  
on  July  4,  2008  at  www.ghgprotocol.org/calculation-­‐tools/all-­‐tools/  

64
World  Business  Council  for  Sustainable  Development  (2005),  CO2  accounting  and  reporting  
standard  for  the  cement  industry,  version  2.0,  June  2005.  
Zimmerman,  R.S.  D.  MacTaggert,  P.  Wolff  (2009)  Increased  data  streatms  in  lowering  GHG  
uncertainty  bounds,  C-­‐lock  Technology,  extended  abstract.  

65
 Appendix  A  
Table  A.1.  Default  fuel/combustion  category  emission  factors  for  carbon  monoxide  
(CO)  
lbs CO/ lbs CO/
6
Mat id unit unit 10 btu material name modifier
663 TON 0.5 0.021 bituminous coal scc contains: "pulverized"
663 TON 0.5 0.021 bituminous coal scc contains: "cyclone"
663 TON 0.6 0.025 bituminous coal scc contains: "cogeneration"
663 TON 275 11.441 bituminous coal scc contains: "hand-fired"
663 TON 6 0.250 bituminous coal scc contains: "spreader stoker"
663 TON 6 0.250 bituminous coal scc contains: "overfeed stoker"
663 TON 18 0.749 bituminous coal scc contains: "atmospheric fluidized bed"
663 TON 11 0.458 bituminous coal scc contains: "underfeed stoker"
663 TON 5.94 0.247 bituminous coal all else
663 TON 6 0.250 bituminous coal all commercial nonpoint coal use
663 TON 275 11.441 bituminous coal all residential nonpoint coal use
663 TON 6 0.250 bituminous coal all industrial nonpoint coal use
323 TON 0.5 0.029 subbituminous coal scc contains: "pulverized"
323 TON 0.5 0.029 subbituminous coal scc contains: "cyclone"
323 TON 0.6 0.034 subbituminous coal scc contains: "cogeneration"
323 TON 275 15.705 subbituminous coal scc contains: "hand-fired"
323 TON 5.5 0.314 subbituminous coal scc contains: "stoker"
323 TON 18 1.028 subbituminous coal scc contains: "atmospheric fluidized bed"
323 TON 11 0.628 subbituminous coal scc contains: "underfeed stoker"
323 TON 6.02 0.344 subbituminous coal all else
323 TON 6 0.343 subbituminous coal all commercial nonpoint use
323 TON 275 15.705 subbituminous coal all residential nonpoint use
323 TON 6 0.343 subbituminous coal all industrial nonpoint use
664 TON 0.5 0.025 bituminous/subbituminous scc contains: "pulverized"
664 TON 0.5 0.025 bituminous/subbituminous scc contains: "cyclone"
664 TON 0.5 0.025 bituminous/subbituminous scc contains: "cogeneration"
664 TON 275 13.573 bituminous/subbituminous scc contains: "hand-fired"
664 TON 5 0.261 bituminous/subbituminous scc contains: "spreader stoker"
664 TON 6 0.296 bituminous/subbituminous scc contains: "overfeed stoker"
664 TON 18 0.888 bituminous/subbituminous scc contains: "atmospheric fluidized bed"
664 TON 11 0.543 bituminous/subbituminous scc contains: "underfeed stoker"
664 TON 5.94 0.295 bituminous/subbituminous all else
664 TON 6 0.296 bituminous/subbituminous all commercial nonpoint use
664 TON 275 13.573 bituminous/subbituminous all residential nonpoint use
664 TON 6 0.296 bituminous/subbituminous all industrial nonpoint use
717 TON 0.07 0.003 coal scc contains: "Oven Pushing"
717 TON 0.6 0.029 coal all else
717 TON 11 0.530 coal all commercial nonpoint use
717 TON 275 13.238 coal all residential nonpoint use
717 TON 6 0.289 coal all industrial nonpoint use
640 TON 90 3.609 anthracite scc contains: "hand-fired"
640 TON 0.6 0.024 anthracite all else
640 TON 275 11.028 anthracite all residential nonpoint use
639 TON 0.3 0.012 anthracite culm
173 TON 5.5 0.424 lignite scc contains "stoker"
173 TON 0.5 0.039 lignite all else
6 3
209 10 FT 1000 0.969 natural gas scc contains: "engine"
6 3
209 10 FT 150 0.145 natural gas scc contains: “engine” and “turbine”

66
 6 3
209 10 FT 400 0.388 natural gas scc contains: “engine” and "reciprocating"
6 3
209 10 FT 65 0.063 natural gas all else
6 3
209 10 FT 84 0.081 natural gas all commercial nonpoint
6 3
209 10 FT 84 0.081 natural gas all industrial nonpoint
6 3
251 10 FT 35 0.032 process gas
6 3
553 10 FT 35 0.032 refinery gas
6 3
310 10 FT 35 0.032 sour gas
6 3
126 10 FT 35 0.032 gas
3
255 E GAL 2.55 0.028 propane
3
832 E GAL 3 0.043 ethane
3
256 E GAL 3 0.031 propane/butane
3
675 E GAL 3 0.029 butane
3
178 E GAL 2.625 0.028 LPG
6
425 10 FT3 2912 4.936 coke oven gas scc is: 39000702, 39000789
6
425 10 FT3 1054 1.786 coke oven gas scc is: 10200707
6
425 10 FT3 18.4 0.031 coke oven gas all else
6
809 10 FT3 511 5.110 coke oven gas or blast furnace gas scc is: 39000701
6
809 10 FT3 185 1.850 coke oven gas or blast furnace gas scc is: 10200704
6
809 10 FT3 13.7 0.137 coke oven gas or blast furnace gas all else
3
44 E GAL 116 0.836 diesel
3
822 E GAL 5 0.036 distillate
3
56 E GAL 5 0.036 distillate oil
57 E3GAL 130 0.929 distillate oil (diesel) scc contains "engine” and “reciprocating"
57 E3GAL 113.5 0.811 distillate oil (diesel) scc contain “engine”
57 E3GAL 130 0.929 distillate oil (diesel) scc contains “reciprocating”
57 E3GAL 6.72 0.048 distillate oil (diesel) scc contains “turbine”
57 E3GAL 6.72 0.048 distillate oil (diesel) all else
823 E3GAL 5 0.036 distillate oil (no 1&2)
824 E3GAL 5 0.036 distillate oil (no 1)
58 E3GAL 5 0.036 distillate oil (no 2)
825 E3GAL 5 0.036 distillate oil (no 4)
818 E3GAL 130 0.949 diesel kerosene scc contains "engine” and “reciprocating"
818 E3GAL 113.5 0.828 diesel kerosene scc contain “engine”
818 E3GAL 130 0.949 diesel kerosene scc contains “reciprocating”
818 E3GAL 6.72 0.049 diesel kerosene scc contains “turbine”
818 E3GAL 6.72 0.049 diesel kerosene all else
279 E3GAL 130 0.867 residual oil scc contains "reciprocating"
279 E3GAL 5 0.033 residual oil all else
922 E3GAL 130 0.867 residual oil (no 5) scc contains "reciprocating"
922 E3GAL 5 0.033 residual oil (no 5) all else
923 E3GAL 130 0.867 residual oil (no 6) scc contains "reciprocating"
923 E3GAL 5 0.033 residual oil (no 6) all else
924 E3GAL 130 0.867 residual crude oil scc contains "reciprocating"
924 E3GAL 5 0.033 residual crude oil all else
272 E3GAL 130 0.867 refined oil using residual oil values
2 E3GAL 2.1 0.015 waste oil scc contains: "space heaters"
2 E3GAL 1.9 0.014 waste oil all else
216 E3GAL 5 0.036 oil
374 E3GAL 5 0.032 crude oil
181 E3GAL 5 0.036 lube oil
127 E3GAL 7900 60.82 gasoline
864 E3GAL 130 1.082 jet A fuel scc contains "reciprocating"
864 E3GAL 113.5 0.944 jet A fuel scc contains “engine”
864 E3GAL 6.72 0.056 jet A fuel all else

67
159 E3GAL 130 1.082 jet fuel scc contains "reciprocating"
159 E3GAL 113.5 0.944 jet fuel scc contains “engine”
159 E3GAL 6.72 0.056 jet fuel all else
865 E3GAL 130 1.082 jet kerosene scc contains "reciprocating"
865 E3GAL 113.5 0.944 jet keosene scc contains “engine”
865 E3GAL 6.72 0.056 jet kerosene all else
160 E3GAL 130 1.040 jet naptha scc contains "reciprocating"
160 E3GAL 113.5 0.908 jet naptha scc contains “engine”
160 E3GAL 6.72 0.054 jet naptha all else
162 E3GAL 5 0.037 kerosene
724 TON 6.6 0.236 coke scc is: 390000899
724 TON 0.6 0.021 coke all else
226 TON 6.6 0.220 raw coke scc is: 390000899
226 TON 0.6 0.020 raw coke all else
142 heat search scc desc for fuel then reference list
Default  emission  values  are  derived  from  the  FIRE  emissions  factor  database  [USEPA  1997;  USEPA  
2006b;  WebFIRE  2005].    

68
Table A.2. Default fuel combustion category emission factors for ntirogen oxides (NOx)
lbs NOx/ lbs NOx/
6
Mat id unit unit 10 btu material name modifier
663 TON 10 0.416 bituminous coal scc contains: "atmospheric fluidized bed"
663 TON 10 0.416 bituminous coal scc contains: "cogeneration"
663 TON 12 0.499 bituminous coal scc contains: "spreader stoker"
663 TON 7.5 0.312 bituminous coal scc contains: "traveling grate"
663 TON 9.1 0.379 bituminous coal scc contains: "underfeed stoker"
663 TON 9.1 0.379 bituminous coal scc contains: "overfeed stoker"
663 TON 9.1 0.379 bituminous coal scc contains: "hand-fired"
663 TON 30 1.248 bituminous coal all else
323 TON 15 0.857 subbituminous coal scc contains: "atmospheric fluidized bed"
323 TON 15 0.857 subbituminous coal scc contains: "cogeneration"
323 TON 11 0.628 subbituminous coal scc contains: "spreader stoker"
323 TON 7.5 0.428 subbituminous coal scc contains: "traveling grate"
323 TON 13.7 0.782 subbituminous coal scc contains: "underfeed stoker"
323 TON 13.7 0.782 subbituminous coal scc contains: "overfeed stoker"
323 TON 13.7 0.782 subbituminous coal scc contains: "hand-fired"
323 TON 25 1.428 subbituminous coal all else
664 TON 12.5 0.636 bituminous/subbituminous scc contains: "atmospheric fluidized bed"
664 TON 12.5 0.636 bituminous/subbituminous scc contains: "cogeneration"
664 TON 11.5 0.564 bituminous/subbituminous scc contains: "spreader stoker"
664 TON 7.5 0.370 bituminous/subbituminous scc contains: "traveling grate"
664 TON 11.4 0.581 bituminous/subbituminous scc contains: "underfeed stoker"
664 TON 11.4 0.581 bituminous/subbituminous scc contains: "overfeed stoker"
664 TON 11.4 0.581 bituminous/subbituminous scc contains: "hand-fired"
664 TON 27.5 1.338 bituminous/subbituminous all else
717 TON 0.03 0.00145 coal scc contains: “oven pushing”
717 TON 3 0.145 coal
640 TON 9 0.361 anthracite scc contains "traveling grate"
640 TON 3 0.120 anthracite scc contains: "hand-fired"
640 TON 18 0.722 anthracite all else
639 TON 1.8 0.075 anthracite culm
173 TON 15 1.157 lignite scc contains: "cyclone furnace"
173 TON 15 1.157 lignite scc contains: "traveling grate"
173 TON 6 0.463 lignite all else
6
209 10 FT3 3000 2.907 natural gas scc contains: "engine"
6
209 10 FT3 400 0.388 natural gas scc contains: “engine” and “turbine”
6
209 10 FT3 2840 2.752 natural gas scc contains: “engine” and "reciprocating"
6
209 10 FT3 140 0.136 natural gas all else
6
251 10 FT3 140 0.126 process gas
6
553 10 FT3 140 0.126 refinery gas
6
310 10 FT3 140 0.126 sour gas
6
126 10 FT3 140 0.126 gas
255 E3GAL 15 0.165 propane
832 E3GAL 15 0.216 ethane
256 E3GAL 15 0.154 propane/butane
675 E3GAL 21 0.204 butane
178 E3GAL 15 0.165 LPG
6
425 10 FT3 90.8 0.154 coke oven gas scc is: 39000702, 39000789
6
425 10 FT3 54 0.092 coke oven gas scc is: 10200707
6
425 10 FT3 80 0.136 coke oven gas all else
6
809 10 FT3 15.9 0.159 coke oven gas or blast furnace gas scc is: 39000701

69
 6
809 10 FT3 9.35 0.094 coke oven gas or blast furnace gas scc is: 10200704
6
809 10 FT3 23 0.230 coke oven gas or blast furnace gas all else
3
44 10 GAL 425 3.064 diesel
3
822 10 GAL 20 0.144 distillate
3
56 10 GAL 20 0.144 distillate oil
3
57 10 GAL 604 4.355 distillate oil (diesel) scc contains "reciprocating"
3
57 10 GAL 98 0.707 distillate oil (diesel) all else
3
823 10 GAL 20 0.144 distillate oil (no 1&2)
3
824 10 GAL 20 0.144 distillate oil (no 1)
3
58 10 GAL 20 0.144 distillate oil (no 2)
3
825 10 GAL 20 0.144 distillate oil (no 4)
3
818 10 GAL 604 4.355 diesel kerosene scc contains "reciprocating"
3
818 10 GAL 98 0.707 diesel kerosene all else
3
279 10 GAL 604 4.035 residual oil scc contains "reciprocating"
3
279 10 GAL 55 0.367 residual oil all else
3
922 10 GAL 604 4.035 residual oil (no 5) scc contains "reciprocating"
3
922 10 GAL 55 0.367 residual oil (no 5) all else
3
923 10 GAL 604 4.035 residual oil (no 6) scc contains "reciprocating"
3
923 10 GAL 55 0.367 residual oil (no 6) all else
3
924 10 GAL 604 4.035 residual crude oil scc contains "reciprocating"
3
924 10 GAL 55 0.367 residual crude oil all else
3
272 10 GAL 55 0.367 refined oil using residual oil values
3
2 10 GAL 16 0.116 waste oil scc contains: "space heaters"
3
2 10 GAL 19 0.138 waste oil all else
3
216 10 GAL 55 0.367 oil
3
374 10 GAL 55 0.367 crude oil
3
181 10 GAL 55 0.367 lube oil
3
127 10 GAL 200 1.599 gasoline
3
864 10 GAL 604 4.474 jet A fuel scc contains "reciprocating"
3
864 10 GAL 98 0.726 jet A fuel all else
3
159 10 GAL 604 4.474 jet fuel scc contains "reciprocating"
3
159 10 GAL 98 0.726 jet fuel all else
3
160 10 GAL 604 4.834 jet naptha scc contains "reciprocating"
3
160 10 GAL 98 0.784 jet naptha all else
3
162 10 GAL 18 0.133 kerosene
724 TON 14 0.466 coke scc contains: "cogeneration"
724 TON 21 0.698 coke all else
226 TON 14 0.466 raw coke scc contains: "cogeneration"
226 TON 21 0.698 raw coke all else
142 heat search SCC desc for fuel then reference list
Default emission values are derived from the FIRE emissions factor database [USEPA  1997;  USEPA  
2006b;  WebFIRE  2005].

70
Table A.3. Fuel combustion category emission factors for carbon dioxide (CO2) and fuel
heat content
mat tonnes heat
6
id CO2/10 btu material name modifier content units
1 3 6
663 0.0931 bituminous coal 24.04 10 BTU/TON
1 3 6
323 0.0967 subbituminous coal 17.51 10 BTU/TON
1 3 6
664 0.0949 bituminous/subbituminous Average of previous two 20.77 10 BTU/TON
1 3 6
717 0.0949 Coal Use previous row 20.77 10 BTU/TON
1 6
640 0.1032 Anthracite 24.94 10 BTU/TON
1 6
639 0.1032 anthracite culm Use previous row 24.94 10 BTU/TON
1 3 6
173 0.0961 Lignite 12.97 10 BTU/TON
3 6 6 3
209 0.0531 natural gas “natural gas pipeline” 1032 10 BTU/10 FT
1 6 6 3
251 0.0561 process gas “refinery fuel gas” entry 1068.6 10 BTU/10 FT
1 6 6 3
553 0.0561 refinery gas “refinery fuel gas” entry 1068.6 10 BTU/10 FT
1 6 6 3
310 0.0561 sour gas “refinery fuel gas” entry 1068.6 10 BTU/10 FT
1 6 6 3
126 0.0561 Gas “refinery fuel gas” entry 1068.6 10 BTU/10 FT
6 3
255 0.0625 Propane 90.42 10 BTU/10 GAL
1 6 3
832 0.0590 Ethane 69.43 10 BTU/10 GAL
6 3
256 0.0635 propane/butane Mix of propane and butane 93.82 10 BTU/10 GAL
6 3
675 0.0644 Butane 97.23 10 BTU/10 GAL
6 3
178 0.0620 LPG 94.0 10 BTU/10 GAL
2 4 6 6 3
425 0.0406 coke oven gas “coke (oven gas)” 574 10 BTU/10 FT
2 4 6 6 3
809 0.2063 coke oven gas or blast furnace gas “blast furnace gas” 92 10 BTU/10 FT
6 3
44 0.0735 Diesel “diesel/gas oil” entry 137.06 10 BTU/10 GAL
6 3
822 0.0725 Distillate “distillate fuel” entry 139.93 10 BTU/10 GAL
6 3
56 0.0725 distillate oil “distillate fuel” entry 139.93 10 BTU/10 GAL
6 3
57 0.0735 distillate oil (diesel) “diesel/gas oil” entry 137.06 10 BTU/10 GAL
6 3
823 0.0725 distillate oil (no 1&2) “distillate fuel” entry 139.93 10 BTU/10 GAL
6 3
824 0.0725 distillate oil (no 1) “distillate fuel” entry 139.93 10 BTU/10 GAL
6 3
58 0.0725 distillate oil (no 2) “distillate fuel” entry 139.93 10 BTU/10 GAL
6 3
825 0.0754 distillate oil (no 4) “fuel #4” entry 143.16 10 BTU/10 GAL
6 3
818 0.0725 diesel kerosene Mix of diesel and kerosene 135.98 10 BTU/10 GAL
6 3
279 0.0780 residual oil 149.97 10 BTU/10 GAL
6 3
922 0.0772 residual oil (no 5) 149.97 10 BTU/10 GAL
6 3
923 0.0803 residual oil (no 6) 153.20 10 BTU/10 GAL
6 3
924 0.0780 residual crude oil “residual oil” entry 149.97 10 BTU/10 GAL
6 3
272 0.0780 refined oil “residual oil” entry 149.97 10 BTU/10 GAL
1 6 3
2 0.0735 waste oil “unfinished oil” entry 138.69 10 BTU/10 GAL
1 6 3
216 0.0725 Oil “other oil” entry 138.69 10 BTU/10 GAL
6 3
374 0.0737 crude oil 142.26 10 BTU/10 GAL
1 6 3
181 0.0735 lube oil “lubricants” entry 138.1 10 BTU/10 GAL
6 3
127 0.0702 Gasoline 129.88 10 BTU/10 GAL
6 3
864 0.0702 jet A fuel “jet fuel” entry 120.19 10 BTU/10 GAL
6 3
159 0.0702 jet fuel 120.19 10 BTU/10 GAL
6 3
865 0.0709 jet kerosene Mix of jet fuel and kerosene 120.19 10 BTU/10 GAL
6 3
160 0.0721 jet naptha “special naptha” entry 120.19 10 BTU/10 GAL
6 3
162 0.0716 Kerosene 134.91 10 BTU/10 GAL
6
724 0.1011 Coke “petroleum coke” 27.96 10 BTU/TON
6
226 0.1011 raw coke “petroleum coke” 27.96 10 BTU/TON
696 Cement scc contains: "wet" process
696 Cement all else
715 Clinker scc contains: "wet" process
715 Clinker all else

71
729 Concrete scc contains: "wet" process
729 Concrete all else
search SCC desc for fuel
142 Heat then ref list
Notes:    CO2  emission  factors  and  heat  content  from  API  [2004]  unless  otherwise  noted.  This  
source  was  used  for  generating  internal  consistency  across  the  many  fuel  categories  
encountered.  The  values  are  within  1.5%  of  other  estimates  (eg.  DOE/EIA,  2007a,  
USEPA,  2008).    
1  CO2  emissions  factor  from  DOE/EIA  [2007b].    
2  CO2  emission  factor  from  IPCC,  [1996].    
3  Coal  heat  values  from  2006  data  contained  within  the  Energy  Information  

Administration,  Form  EIA-­‐423,  "Monthly  Cost  and  Quality  of  Fuels  for  Electric  
Plants  Report"  Federal  Energy  Regulatory  Commission,  FERC  Form  423,  "Monthly  
Report  of  Cost  and  Quality  of  Fuels  for  Electric  Plants."  US  averages  for  coal  types  
were  used.  Bituminous  and  anthracite  coal  types  were  reported  in  one  category.  
4  http://www.engineeringtoolbox.com/heating-­‐values-­‐fuel-­‐gases-­‐d_823.html  

72
 Appendix B

Table B.1. Complete MOBILE6 Vehicle Classifications1

VClass VClassAbbr VClassDesc
1 LDGV Light-Duty Gasoline Vehicles (Passenger Cars)
2 LDGT1 Light-Duty Gasoline Trucks 1 (0-6,000 lbs. GVWR, 0-3750 lbs. LVW)
3 LDGT2 Light-Duty Gasoline Trucks 2 (0-6,000 lbs. GVWR, 3751-5750 lbs. LVW)
4 LDGT3 Light-Duty Gasoline Trucks 3 (6,001-8,500 lbs. GVWR, 0-5750 lbs. ALVW)
5 LDGT4 Light-Duty Gasoline Trucks 4 (6,001-8,500 lbs. GVWR, 5751 lbs. and greater ALVW)
6 HDGV2B Class 2b Heavy-Duty Gasoline Vehicles (8501-10,000 lbs. GVWR)
7 HDGV3 Class 3 Heavy-Duty Gasoline Vehicles (10,001-14,000 lbs. GVWR)
8 HDGV4 Class 4 Heavy-Duty Gasoline Vehicles (14,001-16,000 lbs. GVWR)
9 HDGV5 Class 5 Heavy-Duty Gasoline Vehicles (16,001-19,500 lbs. GVWR)
10 HDGV6 Class 6 Heavy-Duty Gasoline Vehicles (19,501-26,000 lbs. GVWR)
11 HDGV7 Class 7 Heavy-Duty Gasoline Vehicles (26,001-33,000 lbs. GVWR)
12 HDGV8A Class 8a Heavy-Duty Gasoline Vehicles (33,001-60,000 lbs. GVWR)
13 HDGV8B Class 8b Heavy-Duty Gasoline Vehicles (>60,000 lbs. GVWR)
14 LDDV Light-Duty Diesel Vehicles (Passenger Cars)
15 LDDT12 Light-Duty Diesel Trucks 1 and 2 (0-6,000 lbs. GVWR)
16 HDDV2B Class 2b Heavy-Duty Diesel Vehicles (8501-10,000 lbs. GVWR)
17 HDDV3 Class 3 Heavy-Duty Diesel Vehicles (10,001-14,000 lbs. GVWR)
18 HDDV4 Class 4 Heavy-Duty Diesel Vehicles (14,001-16,000 lbs. GVWR)
19 HDDV5 Class 5 Heavy-Duty Diesel Vehicles (16,001-19,500 lbs. GVWR)
20 HDDV6 Class 6 Heavy-Duty Diesel Vehicles (19,501-26,000 lbs. GVWR)
21 HDDV7 Class 7 Heavy-Duty Diesel Vehicles (26,001-33,000 lbs. GVWR)
22 HDDV8A Class 8a Heavy-Duty Diesel Vehicles (33,001-60,000 lbs. GVWR)
23 HDDV8B Class 8b Heavy-Duty Diesel Vehicles (>60,000 lbs. GVWR)
24 MC Motorcycles (Gasoline)
25 HDGB Gasoline Buses (School, Transit and Urban)
26 HDDBT Diesel Transit and Urban Buses
27 HDDBS Diesel School Buses
28 LDDT34 Light-Duty Diesel Trucks 3 and 4 (6,001-8,500 lbs. GVWR)
1
Reproduced here from USEPA [2005d], Table 5a.

Table B.2. Complete MOBILE6 Road Classifications

RoadType RoadDesc
11 Interstate: Rural
13 Other Principal Arterial: Rural
15 Minor Arterial: Rural
17 Major Collector: Rural
19 Minor Collector: Rural
21 Local: Rural
23 Interstate: Urban
25 Other Freeways and Expressways: Urban
27 Other Principal Arterial: Urban
29 Minor Arterial: Urban
31 Collector: Urban
33 Local: Urban

73
Table B.3. The 18 Mobile6.2 vehicle class-road type combinations
Vehicle Types Road Types Mobile6.2 Ftype Mean Travel
Speed (MPH)
LDV Rural Interstate Freeway 60
LDT Rural Interstate Freeway 55
HDV Rural Interstate Freeway 40
LDV Urban Interstate Freeway 45
LDT Urban Interstate Freeway 45
HDV Urban Interstate Freeway 35
LDV Urban Freeways & Expressways Freeway 45
LDT Urban Freeways & Expressways Freeway 45
HDV Urban Freeways & Expressways Freeway 35
LDV, LDT Rural Principal Arterial Arterial 45
LDV, LDT Rural Minor Arterial Arterial 40
HDV Rural Principal Arterial Arterial 35
LDT, LDT Rural Major Collector Arterial 35
LDV, LDT Rural Minor Collector, Rural Local Arterial 30
HDV Rural Minor Arterial Arterial 30
LDV, LDT Urban Principal Arterial, Urban Minor Arterial, Urban Collector Arterial 20
HDV Rural Major Collector, Rural Minor Collector, Rural Local Arterial 25
HDV Urban Principal Arterial, Urban Minor Arterial, Urban Collector Arterial 15

LDV = Mobile6.2 vehicle types 1 and 16

LDT = Mobile6.2 vehicle types 2-5
HDV = Mobile6.2 vehicle types 6-15

Table B.4. Rural/Urban and Light/Heavy Duty Characterization

Category Vehicle Type (Table B.1) Road Type (Table B.2)
Light Duty Urban 1, 2, 3, 4, 5, 14, 15, 24, 28 23, 25, 27, 29, 31, 33
Light Duty Rural 1, 2, 3, 4, 5, 14, 15, 24, 28 11, 13, 15, 17, 19, 21
Heavy Duty Urban 6 - 13, 16 - 23, 25 – 27 23, 25, 27, 29, 31, 33
Heavy Dury Rural 6 - 13, 16 - 23, 25 – 27 11, 13, 15, 17, 19, 21

74
 Table B.5. Sources of selected HPMS Data
Rural Functional Systems
HPMS Data Other
Minor Major Minor
Interstate Principal Local
Arterial Collector Collector
Arterials
Interstate Lane Miles Universe
Interstate VMT Universe

Non-Interstate PAS Lane Universe

Miles Universe
Non-Interstate PAS VMT
1
FA Highway Lane Miles Universe Universe Universe Universe
1 2 2
FA Highway VMT Universe Universe Sample Sample

NHS Lane Miles Universe Universe Universe Universe Universe Universe

Miles Universe Universe Universe Universe Universe Universe

3 3
Lane Miles Universe Universe Universe Universe Universe Universe
2 2 4 4
VMT Universe Universe Sample Sample Summary Summary

Total Public Road Miles Certified Mileage ---------------------------------------------------------------------------------------------------

Urban Functional Systems

HPMS Data Other Other
Minor
Interstate Freeways & Principal Collector Local
Arterial
Expressways Arterial
Interstate Lane Miles Universe
Interstate VMT Universe

Non-Interstate PAS Lane Miles Universe Universe

Non-Interstate PAS VMT Universe Universe

FA Highway Lane Miles1 Universe Universe Universe Universe Universe

FA Highway VMT1 Universe Universe Universe Sample2 Sample2

NHS Lane Miles Universe Universe Universe Universe Universe Universe

Miles Universe Universe Universe Universe Universe Universe

Lane Miles Universe Universe Universe Universe Universe Universe3
VMT Universe Universe Universe Sample2 Sample2 Summary4

Total Public Road Miles Certified Mileage -----------------------------------------------------------------------------------------------------

1 Universe data are used to estimate lane-miles and VMT for the few miles of NHS that are on the minor collector and local
functional systems.
2 Expanded sample data are used.
3 Universe miles times 2 (lanes) are used. States are not required to report number of through lanes on these systems, except
for any NHS sections.
4 Summary data are used. States are not required to report section level AADT on these systems, except for any NHS
sections.

75
Table B.6. Census Bureau Regions and Divisions with State FIPS Codes
Region 1: Northeast
Division 1: New England Division 2: Middle Atlantic
Connecticut 09 New Jersey 34
Maine 23 New York 36
Massachusetts 25 Pennsylvania 42
New Hampshire 33
Rhode Island 44
Vermont 50
Region 2: Midwest
Division 3: East North Central Division 4: West North Central
Indiana 18 Iowa 19
Illinois 17 Kansas 20
Michigan 26 Minnesota 27
Ohio 39 Missouri 29
Wisconsin 55 Nebraska 31
North Dakota 38
South Dakota 46
Region 3: South
Division 5: South Atlantic Division 6: East South Central
Delaware 10 Alabama 01
District of Columbia 11 Kentucky 21
Florida 12 Mississippi 28
Georgia 13 Tennessee 47
Maryland 24 Division 7: West South Central
North Carolina 37 Arkansas 05
South Carolina 45 Louisiana 22
Virginia 51 Oklahoma 40
West Virginia 54 Texas 48
Region 4: West
Division 8: Mountain Division 9: Pacific
Arizona 04 Alaska 02
Colorado 08 California 06
Idaho 16 Hawaii 15
New Mexico 35 Oregon 41
Montana 30 Washington 53
Utah 49
Nevada 32
Wyoming 56

76
Table B.7. Fractions converting VMT by HPMS 2002 vehicle type to VMT by
MOBILE6 2002 vehicle type
HPMS 2002 VMT Fractions 2002 VMT Fractions by MOBILE6 Vehicle Type
HPMS Vehicle RInt ROPA, RMajC, UInt UOther MOBILE6 RInt ROPA, RMajC, UInt UOther
Category RMinArt RMinC, Vehicle Type RMinArt RMinC,
RLoc RLoc
Passenger Cars 0.4947 0.5485 0.5622 0.5951 0.6111 LDGV 0.4939 0.5476 0.5613 0.5941 0.6101
LDDV 0.0008 0.0009 0.0009 0.0010 0.0010
Motorcycles 0.0043 0.0037 0.0039 0.0041 0.0026 MC 0.0043 0.0037 0.0039 0.0041 0.0026
Other 2-Axle 4- 0.3034 0.3474 0.3592 0.3181 0.3425 LDGT1 0.0476 0.0545 0.0564 0.0499 0.0537
Tire Vehicles LDGT2 0.1585 0.1815 0.1876 0.1662 0.1789
LDGT3 0.0482 0.0552 0.0571 0.0505 0.0544
LDGT4 0.0222 0.0254 0.0262 0.0232 0.0250
LDDT12 0.0001 0.0002 0.0002 0.0002 0.0002
LDDT34 0.0010 0.0011 0.0012 0.0010 0.0011
HDGV2B 0.0195 0.0223 0.0231 0.0205 0.0220
HDDV2B 0.0063 0.0072 0.0075 0.0066 0.0071
Single-Unit 2-Axle 0.0312 0.0337 0.0361 0.0223 0.0216 HDGV3 0.0012 0.0013 0.0014 0.0008 0.0008
6-Tire or More HDGV4 0.0006 0.0006 0.0007 0.0004 0.0004
Trucks HDGV5 0.0013 0.0014 0.0015 0.0009 0.0009
HDGV6 0.0028 0.0031 0.0033 0.0020 0.0020
HDGV7 0.0013 0.0014 0.0015 0.0009 0.0009
HDDV3 0.0032 0.0034 0.0037 0.0023 0.0022
HDDV4 0.0028 0.0030 0.0032 0.0020 0.0019
HDDV5 0.0012 0.0013 0.0014 0.0009 0.0009
HDDV6 0.0068 0.0073 0.0078 0.0048 0.0047
HDDV7 0.0101 0.0109 0.0117 0.0072 0.0070
Combination 0.1630 0.0641 0.0340 0.0585 0.0206 HDGV8A 0.0000 0.0000 0.0000 0.0000 0.0000
Trucks HDGV8B 0.0000 0.0000 0.0000 0.0000 0.0000
HDDV8A 0.0357 0.0141 0.0075 0.0128 0.0045
HDDV9A 0.1273 0.0501 0.0265 0.0456 0.0161
Buses 0.0034 0.0025 0.0046 0.0020 0.0016 HDGB 0.0006 0.0004 0.0008 0.0003 0.0003
HDDBT 0.0011 0.0008 0.0015 0.0006 0.0005
HDDBS 0.0017 0.0013 0.0023 0.0010 0.0008
Total 1.0000 1.0000 1.0000 1.0000 1.0000 Total 1.0000 1.0000 1.0000 1.0000 1.0000

77
Table B.8. Mapping of the 28 MOBILE6 vehicle classes to 12 SCC vehicle classes and 8
MOBILE5 vehicle classes
Mapping of MOBILE6 to MOBILE5 and SCC Vehicle Classes
MOBILE6 MOBILE6 SCC-Level 12 Vehicle MOBILE5
Vehicle Class Vehicle Code Classes Vehicle Class
LDGV 1 LDGV (2201001) LDGV
LDGT1 2 LDGT1 (2201020) LDGT1
LDGT2 3
LDGT3 4 LDGT2 (2201040) LDGT2
LDGT4 5
HDGV2B 6 HDGV (2201070) HDGV
HDGV3 7
HDGV4 8
HDGV5 9
HDGV6 10
HDGV7 11
HDGV8A 12
HDGV8B 13
HDGB 25
MC 24 MC (2201080) MC
LDDV 14 LDDV (2230001) LDDV
LDDT12 15 LDDT (2230060) LDDT
LDDT34 28
HDDV2B 16 2BHDDV (2230071) HDDV
HDDV3 17 LHDDV (2230072)
HDDV4 18
HDDV5 19
HDDV6 20 MHDDV (2230073)
HDDV7 21
HDDV8A 22 HHDDV (2230074)
HDDV8B 23
HDDBT 26 BUS (2230075)
HDDBS 27

Table B.9. Seasonal VMT Factors

Vehicle Type Road Type Seasonal VMT Factors
Winter Spring Summer Fall
LDV, LDT, MC Rural 0.2160 0.2390 0.2890 0.2560
LDV, LDT, MC Urban 0.2340 0.2550 0.2650 0.2450
HDV All 0.2500 0.2500 0.2500 0.2500

Table B.10. Monthly VMT Factors

Vehicle Type Road Monthly VMT Factors
Type Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
LDV, LDT, MC Rural 7.44 6.72 8.05 7.79 8.05 9.42 9.74 9.75 8.44 8.72 8.44 7.44
LDV, LDT, MC Urban 8.06 7.28 8.60 8.33 8.60 8.65 8.94 8.94 8.09 8.36 8.09 8.06
HDV All 8.62 7.78 8.42 8.15 8.42 8.15 8.42 8.42 8.24 8.52 8.24 8.62

Table B.11. NONROAD Model Equipment Segments

Model  
Recreational  
Construction  
Industrial  
Lawn/Garden  
Agriculture  
Commercial  
Logging  
Airport  Support  
Underground  Mining  
Oil  Field  
Pleasure  Craft  
Railroad  

78
Table B.12. State Description File Type
State  
Colorado Oil production equipment allocations   oil  
 
Delaware Airport equipment allocations   air  
Delaware   Golf equipment allocations   gc  
Delaware   Household allocations   hou  
Delaware   Logging equipment allocations   log  
Delaware   Source populations   pop  
Delaware   Recreational vehicle park allocations   rvp  
Illinois   Nonroad activity   act  
Illinois   Growth rates   grw  
Illinois   Source populations   pop  
Illinois   Seasonal allocations   sea  
Illinois   Inboard watercraft allocations   wib  
Illinois   Outboard watercraft allocations   wob  
Indiana   Nonroad activity   act  
Indiana   Growth rates   grw  
Indiana   Source populations   pop  
Indiana   Seasonal allocations   sea  
Indiana   Inboard watercraft allocations   wib  
Indiana   Outboard watercraft allocations   wob  
Iowa   Nonroad activity   act  
Iowa   Source populations   pop  
Iowa   Seasonal allocations   sea  
Iowa   Inboard watercraft allocations   wib  
Iowa   Outboard watercraft allocations wob  
Michigan   Nonroad activity   act  
Michigan   Growth rates   grw  
Michigan   Source populations   pop  
Michigan   Seasonal allocations   sea  
Michigan   Inboard watercraft allocations   wib  
Michigan   Outboard watercraft allocations   wob  
Minnesota   Nonroad activity   act  
Minnesota   Growth rates   grw  
Minnesota   Seasonal allocations   sea  
Minnesota   Snowmobile allocations   snm  
Minnesota   Inboard watercraft allocations   wib  
Minnesota   Outboard watercraft allocations   wob  
Ohio   Nonroad activity   act  
Ohio   Growth rates   grw  
Ohio   Source populations   pop  
Ohio   Seasonal allocations   sea  
Ohio   Inboard watercraft allocations   wib  
Ohio   Outboard watercraft allocations   wob  
Rhode Island   Source populations   pop  
Washington   Inboard watercraft allocations   wib  
Washington   Outboard watercraft allocations   wob  
Wisconsin   Nonroad activity   act  
Wisconsin   Growth rates   grw  
Wisconsin   Source populations   pop  
Wisconsin   Seasonal allocations   sea  
Wisconsin   Inboard watercraft allocations   wib  
Wisconsin Outboard watercraft allocations wob  
 

79
Table B.13. Crosswalk table for road types
Roadway Type Vulcan Roadway Type Vulcan Road Classification
Interstate: Rural Interstate: Rural 1
Other Principal Arterial: Rural Arterial: Rural 2
Minor Arterial: Rural
Major Collector: Rural Collector: Rural 3
Minor Collector: Rural
Local: Rural
Interstate: Urban Interstate: Urban 4
Other Freeways and Expressways: Urban
Other Principal Arterial: Urban Arterial: Urban 5
Minor Arterial: Urban
Collector: Urban Collector: Urban 6
Local: Urban

80