-------
      3.5
      3.0
      2.5
  flC

  OG
      2.0
      1.5
  X
  o
  CJ
  E
      1.0
      0.5
1       I      I       T
        1.0
i      i      i       r
                                                           — SMOKE >1
                                                          — SMOKE <1
            APPROXIMATELY 80% OF EXISTING OIL
            FURNACES PRODUCE CYCLE-AVERAGED
            EMISSIONS OF NOX IN THIS RANGE (REF.J
                   PROTOTYPE FURNACE WITH
                   RETENTION HEAD BURNER
                           y    /   /
                                  'PROTOTYPE LOW-EMISSION FURNACE
                     . TARGET LEVEL, CONTRACT 68-02-1819       	
                       NOMINAL OBJECTIVE OF FURTHER NOX
                       REDUCTION EFFORTS
               I       I      I       I      I      I       I      I       I
                          1.5

                  STOICHIOMETRIC RATIO
                          2.0
Figure  1.   Cycle-averaged  flue gas NO  concentrations, comparison of  the
          prototype  optimum furnace with  other existing units.
                                    1013

-------
12.
13.



14.


15.
Martin, G. B. and M. P. Heap.   Evaluation of N0x  Emission  Characteristics
of Alcohol Fuels for Use in Stationary Combustion Systems.   Presented at
the Symposium on Impact of Methanol  on Urban Air  Pollution,  80th  National
AIChE Meeting, Boston, Massachusetts,  September 7-10,  1975.
Axworthy, A. E., G. R. Schneider, M.  D.  Schuman,  and V.
istry of Fuel Nitrogen Conversion to  Nitrogen Oxides in
EPA 600/2-76-039, NTIS No.  PB250-373/AS, February 1976.
    H. Dayan.  Chem-
    Combustion.
Bowman, C. T. and L. S. Cohen.   Influence of Aerodynamic Phenomena on
Pollutant Formation in Combustion.   EPA-650/2-75-061a,  July 1975.
Martin, G. B.  NO  Considerations in Alternate Fuel
Symposium Proceedings:  Environmental Aspects of Fuel
II, EPA-600/2-76-149, June 1976.
Combustion.  In:
  Conversion Technology,
                                     1014

-------
                                                                         21-2
                 EMISSION ESTIMATES OF NO? AND ORGANIC COMPOUNDS
                     FROM COAL-FIRED FLUIDIZED BED COMBUSTION

                                 P.  F.  Fennelly*

ABSTRACT

      fluidized bud combustion *y*tem*  operate at significantly lower tempera-
ture* than conventional system*  (about 900°C versus 1500°C, respectively].
Question* have, ban raised concerning  the organic and nitrogen oxides
emissions which could result at these  lower operating temperatures.  Lining
available data ^rom bench scale on pilot plant fluidized bed combustion
experiments and/on Dimple thermodynamics and empirical correlations with
conventional combustion systems, estimates have been made o^ the concen-
trations o^ various compound!) in the &lue gas.  The&c estimate* are probably
good to within an order o& magnitude.

     The results indicate that with respect to formation ofa photochemical
ozidant precursors, no special environmental problem* should result farom
{,luidized bed combustion.  In &act, this combustion technique offers
significant advantages with respect to reductions in nitrogen oxides and
sulfur oxide* emission*.

                                  INTRODUCTION

     A fluidized bed boiler can be simply depicted as an enclosed Cavity
containing boiler tubes and a bed of granular solids, to which fuel is
added.  As shown in Figure 1, the solids are supported on a grid, at the
bottom of the boiler, through which combustion air is passed at high
velocities, typically 2 to 5 feet per  second (0.6-1.5 m/s).  The solids are
held  in suspension by the upward flow  of the air and a quasi-fluid is
created which contains many properties of a liquid.  The most important
liquid-like property is that the bed material is exceptionally well mixed
and flows throughout the system without mechanical agitation.

     Fluidized bed coal combustion systems for the production of steam
and/or electricity have several advantages over conventional combustion
systems.

     •  High heat transfer coefficients and volumetric heat release
        rates will reduce the boiler size by one-half to two-thirds
        or more compared to a conventional unit.
*GCA Corporation, Bedford, Massachusetts 01730.

                                     1015

-------
                                                        g
                                                        
-------
     •  Capital  costs will  be reduced due to the size reduction and
        the potential for shop fabrication instead of field construction.

     •  The use of limestone as bed material provides a means for i_n_
        situ sulfur dioxide (S02) removal.

     •  The high heat transfer coefficients permit lower operating
        temperatures (850 to 950°c) which can potentially decrease
        nitrogen oxides (N0x) emission.

     Questions have been raised concerning the organic emissions that could
result at these lower operating temperatures.  Using available data and/or
simple thermodynamics and empirical correlations, this oaper provides esti-
mates of the concentrations of various compounds that could result from
coal-fired fluidized bed combustion (FBC).  These estimates are probably
good to within an order of magnitude.

                            STABLE ORGANIC COMPOUNDS

     Potential organic pollutants are those compounds which could form from
the incomplete combustion of coal.  One can view the burning of a coal
particle as a sequential process.  The particle vaporizes or volatilizes;
these volatile compounds can react among themselves in a chemically reducing
atmosphere.  Next, they react with oxygen within the system in a diffusion
flame.  After devolatilization is completed, the char continues to burn.
In specifying the potential organic compounds which could form, the basic
question is what types of chemical species are produced during coal pyrolysis
and to what extent will they survive in the reactive environment of a
fluidized bed combustor.

     The chemical structure of coal can be viewed as a network of inter-
connected aromatic hydrocarbon compounds (1),  Small hydrocarbons (HC) can
form directly from cleavage of substituted alkyl groups.  Polynuclear
aromatic hydrocarbons (PAH) can form directly via bond cleavages in the
structural network or they can form via condensation reactions of various
HC decomposition products.   These condensation reactions even though generally
endothermic, based on thermodynamic estimating techniques, at FBC temperatures
of 1500°F probably proceed at a significant rate (2,3).  Branched or cyclic
HC are seldom found as products of coal pyrolysis carried out at similar
temperatures  (1); hence, only two classes of HC should be of any significance
in fluidized bed coal combustion: small HC (i.e., less than three carbon
atoms) and PAH.  Both could form and survive within the bed at temperatures
on the order of 1500°F.

     Similar arguments apply to organic nitrogen and sulfur compounds.
Species such as pyridine decompose at temperatures of 1000°F to form HCN and
small HC (4).  Thiophenes and mercaptans can also decompose to form small
HC and hydrogen sulfide (H2S).
                                     1017
-------
                  CONCENTRATION  ESTIMATES OF ORGANIC COMPOUNDS
     For a rough estimate of the concentrations  at which some of the small
HC and reduced sulfur and nitrogen compounds might exist, one can use
equilibrium calculations based on free energy minimization.   An upper limit
can be obtained from calculations performed in conjunction with coal
gasification experiments, where highly reducing, fuel-rich conditions  •
exist (5).  Even with only 60 per cent stoichiometric oxygen present, the
data in Table 1 indicate that HCN, COS, CS:,, etc.  are less than 10 parts per
million (ppm).  Extrapolation of the calculations  to typical operating
conditions such as 20 percent excess air, where S02 and NOX become the
predominant sulfur and nitrogen compounds, indicates that compounds such

           TABLE  1.  CALCULATED EQUILIBRIUM CONCENTRATION FOR SELECTED
                SPECIES  PRODUCED BY  INCOMPLETE COMBUSTION OF COAL
               Coal analysis:  C-68.5%, H-5.3%, 0-8.5%, N-l.4%,
                               S-4.1%
               Oxygen present:  59% of stoichiometric requirements
               Temperature:     1400°F
                                                  Mole
                         Species                fraction
                         N2                       0.36
                         CO                       0.26
                         H2                       0.15
                         H?0                      0.06
                         C0?                      0.10
                         CH,t                      0.06
                         H?S                      0.008
                         COS                      0.0005
                         NH3                      0.0007
                         C(g), HCN, CS?         < 10"5
                         C?H?, C7Hlf,  (CN)?
                         S?, S07, S03
                         N-,, NO,
                  Based  on  Reference 5
                                     1013
-------
as HvS, HCN, COS, (CN)>,  etc.  should be present in concentrations less than
1  pprn.   Free energy minimization calculations for the more complicated PAH
are impractical.   To estimate  the concentrations  at which these types  of
compounds might exist in  the FBC flue gas, one can use empirical  correlations
between benzo(a)pyrene and methane (CHj concentrations from measurements  in
conventional coal-fired combustion systems.   Figure 2 indicates the variation
of total  hydrocarbons (as CHt() as a function of oxygen (0-) concentration
from one set of FBC experiments.  Under normal operating conditions, about
3 per cent 02 in  the flue gas  (20 per cent excess air), the concentration
of HC (as CHi+ ) is about  100 ppm (vol ume/vol ume-V/V) (6).  Although emissions
can often vary between different fluidized bed systems, 100 ppm provides
a convenient reference value.   Previous measurements with conventional
coal-fired systems indicate that compounds such as benzo(a)pyrene are
     aiiy iQ-5 times less than the concentration of total HC as CH4(7).
         3000
        ex
        o.

       V)
         2500
       §2000
z
Ul
u 1500
O
UJ
z

jE  1000
UJ
2
       z
       o
       CD
       tr
       o
       tr
       o
           500
                             TEST  CONDITIONS
                      RE ACTOR'FLUIDIZED  BED MODULE
                      SUPERFICIAL  VELOCITY =12-14  fps
                      BED  TEMPERATURE = 1600-I800°F

                      BED DEPTH' 12"- 13"

                                    TYPE  OF COAL

                        O WASHED.OHIO^B  SEAM, 2.6%  S

                        X  UNWASHED, OHIO #8 SEAM, 4.5% S
                                                                    30
                                                                       0)
                                                                       o
                                                                    20
                                                             10
Q.


tt




V)
                                                                X
                                                                UJ
       Figure 2.
                  1.0        2.0        3.0        4.0        5.0
                 FLUE  GAS  OXYGEN  CONTENT, percent

           Variation in hydrocarbons concentration  with  flue
             gas  oxygen content  in the  FBM  (6).
                                    1019
-------
Using our reference value of 100 ppm (CH..),  this  implies  that in a
fluidized bed system, PAH could exist in the flue gas  at  concentrations  (V/V)
on the order of 1  part per billion (ppb).   Considering that flue gases  are
eventually diluted by a factor of a thousand when they are emitted  from
the stack (8), this implies that ambient concentrations of PAH near FBC
facilities would be on the order of 1 part per trillion.   This corresponds
to about 0.6 ng/tn3 which is roughly comparable to the  natural background
concentration ranges found in rural areas  (7). Accordingly, it seems that
PAH concentrations should not be high enough to cause  problems.

     Recently, some conventional coal-fired flue  gases have been tested for
the presence of polychlorinated biphenyls  (PCB) and trace concentrations
have been reported (8).  Experience in coal  combustion and coal pyrolysis
indicates that at temperatures similar to  that in coal-fired FBC, chlorine
(when present) predominatly exists as HC1  (1). Accordingly, one would  not
expect significant concentrations of PCB's In coal-fired  FBC.  If oresent
at all, they should be in concentrations less than those  of the unsubstituted
PAH (i.e., <1 ppb (V/V)).

                            NITROGEN OXIDES EMISSIONS

     There is considerably more experimental data available for NOX emissions
in FBC compared with that for organic compounds (9-12).  These emissions can
vary from system to system and work is underway to summarize the pertinent
fluidized bed parameters which influence NOX emissions.  Of special interest
are the relative contributions of thermal  NQX and fuel  NOX.

     At the temperatures encountered in FBC (1500 to 1600°F), equilibrium
calculations based on thermal fixation of  atmospheric  nitrogen indicate
nitric oxide (NO)  levels should be 100 to  200 ppm, depending on the amount
of 02 present.  In fact, NO levels in most coal-fired  fluidized bed experi-
ments conducted at atmospheric pressure are on the order  of 300 to  500  ppm
(9-12).  Two plausible explanations are (1)  the excess NO results from  hot
spots within the bed which are far in excess of the equilibrium temperature,
or (2) the higher NO levels result from substantial conversion of the nitrogen
in coal.  There is some interesting experimental  evidence supporting the
latter mechanism.

     Workers at Argonne Laboratories in a  set of  experiments on fluidized
bed coal combustion substituted argon for  the nitrogen in combustion air
and found that NOX levels did not change significantly.  In one experiment,
the initial NO concentration of 580 ppm did not vary during the argon sub-
stitution by more than +_ 20 ppm (9,10).

     NOX concentrations in fluidized bed systems, however, can be very
apparatus dependent, and significant concentration profiles have been found
in some reactors (11).  Accordingly, one must be  wary  of  the manner in  which
emission measurements are made.

     In a recent review, workers at Battelle compared  NOX data from a number
of laboratories and, in almost all cases,  the emissions are less than the
                                     1020
-------
EPA New Source Performance Standard which is  0.7 pound per million  Btu  heat
input,  maximum 2-hour average expressed as  N02  (equivalent to  approximately
550 ppm N02 at 3 per cent 0?  in the flue gas)  (13).

     At higher pressures  (6 to 8 atm),  experiments  indicate NOX  levels  are
reduced to values on the  order of 100  to 200  ppm, although the reason for
lower values is not yet known (14,15).

                             SULFUR OXIDES  EMISSIONS

     Only brief attention will be paid here to  SOX  emissions as  their role
in oxidant formation is considerably less well  defined than that of HC  or
NOX.  The emissions of SOX depend markedly  on the calcium/sulfur (Ca/S) mole
ratio which in turn is dependent on the amounts and types  of coal and lime-
stone used in the bed.  In general, however,  with Ca/S mole ratios  greater
than 2.0, 90 per cent control of S02 emission is achievable.  By way of
example, experimental results indicate that a bituminous coal  with  a heating
value of 13,700 Btu per pound and a 2.8 per cent sulfur content  when burned
in a fluidized bed at velocities of 2.1 feet  per second and a  Ca/S  ratio of
2.9 has an emission factor of 0.23 pound S02/106 Btu compared  with  EPA's New
Source Performance Standard of 1.2 pounds S02/106 Btu (14).
                                   CONCLUSIONS

     Concentration estimates for the various types of pollutants that could
form in coal-fired fluidized bed combustion are provided in Table 2 (16).   As
mentioned previously, there is only sparse experimental information available
for most of these compounds and these estimates which are based on simple
thermodynamic considerations and empirical correlations with conventional
combustion systems are probably good only to within an order of magnitude.
The only compounds generated in quantities high enough to be of concern in
oxidant formation are NO, S02, and carbon monoxide (CO).  The concentrations
of these compounds, however, are significantly less than those generated in
conventional combustion; hence, with respect to reducing oxidant formation,
fluidized bed combustion is a promising alternative to present boiler systems,

     Most of the existing data, however, has been gathered on bench scale
equipment or small pilot plant systems.  Final acceptance of FBC will prob-
ably await tests on larger facilities; but, presently, no emission problems
are anticipated in the scaling up of the smaller facilities.
                                     1021
-------
           TABLE 2.   ESTIMATED CONCENTRATION  RANGES  OF  POTENTIAL
          POLLUTANTS FROM COAL-FIRED FLUIDIZED  BED COMBUSTION  (16)
 Gas  phase

   Several  hundred parts/million:     CH,+ , CO, HC1, S02, NO

   Ten  parts  per million:             S03, C?H4, C?H6

   One  part per million:              HF, HCN, NH3, (CN)2, COS, H2S, h^SO^,
                                     HN03, F, Na

   One  part per billion:              Diolefins, aromatic hydrocarbons,
                                     phenols, azoarenes, As, Pb, Hg, Br, Cr,
                                     Ni, Se, Cd, U, Be

   One-tenth  (0.1) part per billion:  Carboxylic acids, sulfonic acids,
                                     polychlorinated biphenyls, alkynes,
                                     cyclic hydrocarbons, amines, pyridines,
                                     pyroles, furans, ethers, esters,
                                     epoxides, alcohols, ozone, aldehydes,
                                     ketones, thiophenes, mercaptans

 Solids

   One  part per million:              Al, Ca, Fe, K, Mg, Si, Ti, Cu, Zn, Ni,
                                     U, V

   One  part per billion:              Ba, Co, Mn, Rb, Sc, Sr, Cd, Sb, Se, Ca

   One-tenth  (0.1) part per billion:  Eu, Hf, La, Sn, Ta, Th
        —	"~  "  '            ACKNOWLEDGEMENTS

     This work was performed under contract to  the  U.S.  Environmental  Pro-
tection Agency under the guidance of Mr.  D.  Bruce Henchel  as  Project  Officer.
The opinions, findings,  and conclusions  are those of the author  and not
necessarily those of the Environmental  Protection Agency.
                                   REFERENCES

1.  Lowry, H.H. (ed.).   The Chemistry of Coal  Utilization.   John  Wiley and
    Sons, New York, 1963.

2.  Franklin, J.L.   Ind.  Eng.  Chem. ,  41:1070,  1949.
                                     1022
-------
 3.   Reid,  R.C.  and  T.K.  Sherwood.  The  Properties of Gases and Liquids -
     Their  Estimation  and Correlation.   McGraw-Hill, New York, 1958.

 4.   Sternling,  C.V. and  J.O.L.  Wendt.   Kinetic Mechanisms Governing the
     Fate of Chemically Bound  Sulfur  and Nitrogen  in Combustion.  EPA-
     650/2-74-017,  U.S. Environmental  Protection Agency.

 5.   Stinnett,  S.J., D.P. Harrison, and  R.W.  Pike.  Fuel Gasification.
     The Prediction  of Sulfur  Species  Distribution by Free Energy Minimiza-
     tion.   Environ. Sci. and  Technol . ,  8:441,  1974.

 6.   Robinson,  E.B., A.H. Bagnulo,  J.W.  Bishop, and S.  Ehrlich.  Characteri-
     zation and Control of Gaseous  Emissions  from  Coal-Fired  Fluidized Bed
     Boilers.   Pope, Evans, Robbins,  Inc.,  Alexandria,  Va.  Report  prepared
     for Division  of Process Control  Engineering,  National Air Pollution
     Control Administration (now U.S.  Environmental Protection Agency).
     1970.   p.  103.

 7.   Hangebrauck,  R.P., D.J. von Lehmden, and J.E. Meeker.  Emissions of
     Polynuclear Aromatic Hydrocarbons and  Other Pollutants from Heat
     Generation and  Incineration Processes.   J. Air Pollu. Control  Assoc.,
     14:267, 1964.   See also:   Hangebrauck,  R.P.,  D.J.  von Lehmden, and
     J.E. Meeker.   Sources of  Polynuclear Hydrocarbons  in the Atmosphere.
     PHS 999-AP-33,  U.S.  Department of Health,  Education and  Welfare, 1967.

 8.   Cowheard,  C.,  M.  Marcus,  C.M.  Guenther,  and J.L. Spigarelli. Hazardous
     Emission Characteristics  of Utility Boilers.   EPA-650/2-75-006, U. S.
     Environmental  Protection  Agency,  July  1975.

 9.   Jarry, R.L.,  L.J. Anastasia, E.L. Carls, A.A. Jonke, and G.J.  Vogel.
     Comparative Emissions of  Pollutants During Combustion of Natural Gas
     and Coal  in Fluidized Beds.   In:  Proceedings of Second  International
     Conference on  Fluidized Bed Combustion,  Hueston Woods, Ohio, 1970.
     U.  S.  Environmental  Protection Agency  Publication  AP-109.

10.   Jonke, A., et  al. Reduction of  Atmospheric Pollution by the Application
     of Fluidized  Bed  Combustion.  ANL/ES-CEN-1001 , Argonne National Labora-
     tory,  1969.

11.   Pereira, F.J.,  J.M.  Beer, B. Gibbs, and A.B.  Hedley.  NOX Emissions  from
     Fluidized Bed Coal Combustion.   Fifteenth Symposium  (International)  On
     Combustion, 1974, The Combustion Intitute, Pittsburgh, Pa.
                                    1023
-------
                                                                           21-3
     EMISSIONS ASSESSMENT OF THE  CHEMICALLY  ACTIVE FLUID BED (CAFB) PROCESS

  A. S. Werner, R. M. Bradway,  D.  F.  Durocher,  S.  L.  Rakes, and R. M. Statnick*

ABSTRACT
     Stack ga(> and poAticLilate.  miAtxionA  &Aom the. Ckz.m-ic.atly Active, fluid
Bed pilot plant at the, EMO  Rue.ax.ch dntAe.,  Abtngdon, MeA.e, me.at>uAe,d during
Q0i> location o^ ^u.e.1 oiJL  and b-ituman.   Qua.ntLU.eA me.aAuAe,d include.d total
paAticulate. , poAticulatz.  &i.ztAibutsionA ,  nitAoge^n oztdu, Aul&uA
and gaAe.ouA organic compounds.   Both paAti.caI.ate- and goxSeou/S e,m-i!>£
analyzed &OA oAgan-ic faunctsional gAou.pt,  uA-ing  Liquid chAomatogAaphy and tn
Ae.d ApzctAOAcopy.  In addition,  bulk and  AuA&ace. tAace. mutal compoA-itionA
paAticatate. weAe. date,Am^ne.d  &Aom &paAk  Bounce. mo44 t>pe.ctAot>copy and X-Aay
pkoto-eJL&ctAon £pe.ctAo-t>copy.  Result* aAe. pAZAnntzd uihtch indicate, that
^actoAy con&iol o& ga&couA Apnci
-------
sure the environmental acceptability of this process.   To attain these goals,
a systematic evaluation of all waste streams from the  CAFB was made and a
process emissions inventory was compiled.   These data  were derived from
engineering estimates and from an extensive pilot plant field sampling and
laboratory analysis program.   Emission rates determined for the pilot plant
were then used to predict pollutant loadings for the CAFB demonstration plant
and proposed commercial units and to make  recommendations for additional
control requirements.

     The experimental phase of the project consisted of two parts::  a field
test program carried out at the ERCA pilot plant during the period December 3 -
11, 1975, which consisted of on-site determinations of air emissions and
collection of gaseous and solid samples; and a laboratory analysis program to
identify and, where possible, quantify chemical  constituents of samples
collected at the pilot plant.

                                THE CAFB PROCESS

     Figure 1 is a schematic diagram of the CAFB pilot plant and Table 1
lists mass flow rates and temperatures of  internal and external streams.
Limestone and oil are fed continuously into the  gasifier at a calcium (Ca)/
sulfur  (S) (limestone/oil) molar ratio of  unity.  In the gasifier,, limestone
(CaC03) is rapidly converted to lime (CaO) and carbon  dioxide (C0?), and the
CaO is maintained in a fluidized state by  a preheated  air/flue gas mixture.
The air input rate is roughly 20 percent stoichiometric with respect to oil.
Fuel oil is consecutively vaporized, oxidized, cracked and reduced at 870°C
(1600°F) to produce a low Btu gas.  Over 80 percent of the input feed sulfur
is removed by the lime.  The gas travels from the gasifier through cyclones
for particulate removal and then into a boiler for combustion.  The boiler
flue gas encountered a knockout baffle and another cyclone before entering
the stack.  Lime is continuously cycled between  the gasifier and the regen-
erator where the roughly 7 percent of the  lime which is sulfided in the
gasifier is oxidized to CaO.   At the pilot plant, sulfur dioxide (S02) pro-
duced in the regenerator is fed to the boiler stack.  (In the demonstration
plant, S02 will be fed to another unit which will reduce the sulfur (S) to its
elemental form.)  Some spent CaO is continuously withdrawn from the regenera-
tor and retained for disposal.  To maintain S removal  efficiency, an equiva-
lent amount of CaCQ3 is continuously added to the gasifier.  Proper disposal
of spent CaCOs is a major environmental concern.

                               FIELD TEST  PROGRAM

     Coincident with the planning and pre-test site visit was the announce-
ment of the "multilevel phased approach" to source sampling and analysis by
the Process Measurement Branch of the Industrial Environmental Research
Laboratory, EPA (4).  Because this project was a preliminary environmental
assessment of a facility which heretofore  had not been subjected to a com-
prehensive emissions assessment, the decision was made to combine measurements
of the so-called criteria pollutants using the standard EPA methods with the
new procedures for broad identification of organic and inorganic emissions.

     The field test program was carried out at the ERCA pilot plant location


                                     1026
-------
                                                                E
                                                                (O

                                                                Q.


                                                                •!->
                                                                O

                                                                •I	 •»
                                                                Q-
                                                               c_>   S-
                                                                ro -W>
                                                                S-  W
                                                                cu
                                                                Q.    o
                                                               •r-  O
                                                                E  S-
                                                               => O.
                                                                O)
1027
-------
                  TABLE 1.   ERCA PILOT  PLANT  MASS  FLOW RATES

No.
Fig
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
on
. 1 Process stream
Oil feed to gasifier
Limestone feed to gasifier
Gasifier to regenerator stone transfer
Regenerator to gasifier stone transfer
Product gas to cyclone
Cyclone solids return to gasifier
N? gas to solids transfer lines
Product gas to boiler
Air to regenerator
Spent solids from regenerator
Regenerator off gas to cyclone
Regenerator off gas, cyclone to stack
Flue gas from boiler
Flue gas recirculated to gasifier
Flue gas to Tuyere Blower
Recycled flue gas from cyclone
Flue gas and air to gasifier
Flue gas to stack
Solids from boiler flue gas cyclone
Solids from recycled flue gas cyclone
Solids from regenerator off gas cyclone
Start up kerosene to gasifier
Stack emissions
Fuel injection air
Mass fl
kg/sec
0.04
0.003
0.11
0.11
0.16

0.0006
0.16
0.01
0.002
0.01
0.01
0.50
0.03
0.02
0.02
0.10
0.50



0.0005
0.50
0.01
ow rate,
(Ib/hr)
(228)
(25)
(860)
(850)
(1,279)

(4.5)
(1,279)
(65)
(14)
(63)
(63)
(4,000)
(250)
(125)
(125)
(800)
(4,000)



(4)
(4,000)
(45)
Temperature,
°C ' (°F)
88 (190)



850 (1,560)

850 (1,560)


1,050 (1,920)
1,050 (1,920)










43 (110)


during December, 1975.   During this period, two fuel  oils were fired:  bit-
umen (vacuum bottoms) and residual  oil  (atmospheric bottoms).   Samples were
collected during seven  separate runs:   four residual  oil  gasification runs;
two bitumen gasification runs; and one  combustion/startup bitumen run.  The
field measurement program consisted of  on-site quantification of S02> sulfur

                                     1028
-------
trioxide (S03),  nitrogen oxides (N0x),  hydrogen sulfide (HrS),  total  parti-
culate and participate size distributions.   Flue gas was sampled and  analyzed
for N0x by Method 7, S0.9/S03 by Method  8 and H?S by Method 11  (5).   An Orsat
analyzer was used to measure carbon monoxide (CO), C0? and oxygen (02).
Particulate sampling was accomplished using a standard RAC train constructed
according to the procedures outlined in EPA Method 5 (5).  Due  to the positions
of the installed ports, eight point traverses were taken on two diameters
120 degrees apart.  The train was modified slightly to allow for sampling of
gaseous organic species (see below).  Particulate size distribution measurements
were  taken with a University of Washington eight stage instack impactor using
ungreased foil substrates.  A  single point was sampled isokinetically for
sufficient time (15 to 30 minutes) to collect a weighable quantity on each
stage.

      In addition  to collection of particulate and inorganic gases, organic
vapors were collected  in a  specially designed adsorbent trap placed between
the filter and the  impingers on the Method 5 train.  The polymer used in
this  adsorbent cell, Tenax, reportedly retains all organic gases larger than
C5.
                         LABORATORY ANALYSIS PROCEDURES

     Three general types of analytical procedures were applied to oil, flue
gas, and particulate samples collected during the field test program:  organic
functional group identification; trace element quantification; and surface
element and inorganic compound quantification.

ORGANIC FUNCTIONAL GROUP ANALYSIS*

     In this procedure, sample extracts are separated into eight fractions
by  liquid chromatography (LC), evaporated to dryness, weighted, redissolved,
and analyzed by infrared spectroscopy  (IR) (6).  Methylene chloride was used
to  extract oil and particulate samples whereas pentane was used to extract
organic vapors adsorbed on the Tenax polymer.

     Liquid chromatographic separation into eight fractions is accomplished
by  transferring the extract to an LC column and eluting sequentially with a
series of solvent mixtures.  After collection from the LC column each frac-
tion is reduced to dryness using a Kuderna-Danish evaporator and air evapora-
tion and then weighed to determine the amount of organic material in each
fraction.

     The dried fractions are then redissolved in methylene chloride and
subjected to IR analysis.  The IR spectra are then scanned for functional
group peaks.
 *These  analyses were  performed  by  Battelle  Columbus  Laboratories  under  sub-
  contract  and  by  the  Process  Measurements Branch  of  EPA.

                                     1029
-------
TRACE ELEMENT ANALYSIS*

     Stack particulate was analyzed for bulk elemental composition using low
precision (±200 percent) spark source mass spectrometry (SSMS).   This techni-
que is sensitive to 70 elements.  To calibrate the SSMS results, some elements
were quantified by higher precision atomic absorption (AA) spectroscopy.
Interference of organic ions with low atomic weight elements is  well known
in SSMS as are losses of volatile compounds.  Thus uncertainties of values
derived for light elements such as fluorine, sodium and sulfur may be higher
than the indicated precision.

SURFACE ANALYSIS

     A number of particulate samples were investigated for surface elements
and inorganic compounds by X-ray photoelectron spectroscopy (XPS or ESCA).
All samples analyzed by ESCA in this study were first scanned over the entire
electron binding energy range (broadband scan) to identify those elements
present in concentrations greater than 0.1 percent.  These broadband spectra
were then analyzed to yield surface concentrations of all identifiable
elements.  The compound forms of surface vanadium and sulfur are also of in-
terest in this study.  They were investigated by scanning the binding energy
ranges corresponding to the ejection of the 2p electron of vanadium and the 2p
electron of sulfur.  To supplement the bulk SSMS analyses, high  energy argon
ions were used to etch away surface layers exposing strata 20 to 100 ft deep.
The exposed sample layers were then rescanned over the entire binding energy
range and the resultant elemental concentrations were compared with the surface
values and the SSMS analyses.

                               FIELD TEST RESULTS

     Table 2 summarizes the stack emission rates of N0x, S02, S03, H2S, and
total particulate.  The average NO concentration of 53*5 ppm is  considerably
lower than the low end of the concentration range found for conventional oil-
and gas-fired boilers.  The measured emission rate is also about one-fourth
of the New Source Performance Standard (NSPS) for oil-fired boilers and one-
third of the NSPS for gas-fired boilers (7).  Furthermore, the relative in-
variance of the three measurements suggests a low correlation between the NO
emission rate and temperature, excess oxygen, bed stone history, and fuel.

     Several factors may contribute to the low absolute NO  emission rate.
The reducing atmosphere in the gasifier may severely inhibft oxidation of
fuel nitrogen.  It has also been suggested that CaC03 might aid catalytically
in the decomposition of NO or react directly with NO (8).  The presence of
nitrogen on the surface of some of the smaller particulate (which was noted
in the ESCA analysis) is consistent with this latter mechanism.   It is also
possible that reduced nitrogen species, such as ammonia (NH3), formed in the
gasifier, pass through the boiler without reacting.  It is more  difficult to
explain away the apparently small amount of NO formed by thermal fixation in
*This work was performed by Battelle Columbus Laboratories and Aculabs under
 subcontract to GCA and by Northrop Services under contract to EPA.

                                     1030
-------
                    TABLE 2.   SUMMARY OF STACK EMISSIONS
N0/ SO; S03 H;S Total
particulate
Run
1
2
3
4
5
fi

7
Fuel ppm lb/10" Btu ppm lb/10 Btu ppm lb/10 Btu ppri lb/10 Btu ppn lb/10 Btu
Fuel Oil 53.5 0.0851 292 0.6431 8.3 0.023 <0.05
Fuel Oil 45.7 0.0671 305 0.6191 7.9 0.0201 <0.05
Fuel Oil 0.23 2.6xlO~'4
Fuel Oil
Bitumen 58.4 0.0791 828 1.5624 11.1 0.0263
Bitumen and
Stone Feeding
Bitumen
0.0971
0.0561
0.063
0.0921
0.101

0.1046
0.1921

the boiler.  The use of the thermal fixation reaction is strongly affected by
boiler design and firing characteristics which cannot easily be evaluated.

     Two uncertainties are thus apparent.  What is the fate of the fuel bound
nitrogen which is not converted to NO ?  How will the rate of the thermal
fixation reaction be affected by the particular characteristics of the boiler
to be used in conjunction with the demonstration plant?

     The S02 emission rates for residual oil gasification were about 20 percent
below NSPS for oil-fired steam generators and correspond to retention of 80
percent of the fuel sulfur in the gasifier lime (7).  The higher S02 rate
measured for bitumen gasification is due to two factors:  the sulfur content
of bitumen is roughly 50 percent higher than in fuel oil and, more importantly,
stone present in the gasifier during this run had been unregenerated and unre-
plenished for some time because of clogging in the gasifier-regenerator transfer
duct.

     The primary source of particulate emissions from the CAFB is gasifier
bed stone which passes through the two internal  cyclones, the knockout baffle,
and the stack cyclone.  The NSPS for particulate emissions from oil-fired
boilers is 0.1 lb/106 Btu (7).  Table 2 shows that during oil gasification
two of the four runs produced emissions only a few percent below the standard.
The NSPS was exceeded during bitumen gasification and combustion; the final
bitumen gasification run exceeded the particulate standard by a factor of two.

     Two factors must be invoked to understand the variation in and magni-
tude of particulate emissions:  cyclone efficiency and fresh stone feed.
The cyclones used at the pilot plant were not designed specifically for the
CAFB and operated at an efficiency of only 50 percent.  The high emission
rate from Run 7 can be attributed to an unusually high fresh stone feed rate.
Fresh stone undergoes attrition and calcination as it enters the gasifier.
High stone feed rate normally occurs during start up (combustion) but was
employed during Run 7 to compensate for the buildup of saturated stone due
to the clogging of the gasifier-regenerator stone transfer duct.  The parti-


                                     1031
-------
culate sizing distributions measured during Runs  3  and  5  indicate  that a
substantial fraction of the particulate emissions is in the respi^able range
and hence of primary concern.   The large respirable fraction is typical  of
conventional cyclones and may be expected in emissions  from the demonstration
plant which will also employ cyclones for particulate control.

                           LABORATORY ANALYSIS RESULTS

     The laboratory program consisted of organic  and inorganic  analyses  of
stack particulate and organic analyses of flue gas, bitumen, and spent regene-
ration stone.   Stack gas emissions collected during bitumen gasification were
found to contain aliphatic hydrocarbons, alcohols,  ketones, aromatics  and
carboxylic acid salts.   The total  gaseous organic emission rate (C>5)  was 0.022
lb/106 Btu.  Particulate from this same run contained aliphatic hydrocarbons,
carbonyl compounds, esters, aromatics, phthalates,  phenols and  alcohols  to-
taling 7 x TO'5 lb/106  Btu.

     Elemental  analyses of these stack particulates indicated that the most
abundant species are:  calcium, 14.1%; sulfur, 3.9%; vanadium,  1%; silicon,
0.5%; sodium, 0.4%; magnesium, 0.3%; nickel, 0.2%; and iron, 0.1%.  Compari-
son of these values with EPA's Multi-Media Environmental  Goals  (MEG's)(9),
which are derived from TLV's, showed that only the vanadium emission rate is
of potential concern.  Compound identification by ESCA showed that particulate
vanadium exists as a mixture of oxides, V205, V02 and V203.

     Particulate from fuel oil gasification was found to contain aliphatic
hydrocarbons, carbonyls, esters, ketones, aldehydes, alcohols and carboxy-
lates.  The total organic particulate emission rate was approximately equal
to that from bitumen gasification.  The elemental composition of these parti-
culates is similar to the bitumen particulates but contains half as much
sulfur and roughly 0.15% of chlorine and fluorine.

                                   CONCLUSIONS

     The CAFB process was developed to control sulfur oxide emissions from
combustion of high sulfur heavy fuel oil.  The results presented here show
that this goal  can be met under proper system operating conditions.  Under
the upset mode in which gasifier stone becomes saturated, control effective-
ness of the process is impeded.

     Particulate emissions present the most pressing air pollution problem.
Even under normal operation the particulate emission rate, caused by attrition
of gasifier bed stone, approaches the NSPS for oil-fired units.  Improved
particulate collection efficiency is required and is being developed, in the
form of more highly efficient cyclones, for the demonstration plant presently
in the final design state  (2).

     Nitrogen oxide emissions from the CAFB are very low and appear to be
independent of fuel, gasifier stone history, and detailed operating conditions,
The reasons for this observation are not understood completely and additional
studies should be undertaken to uncover the fate of fuel  nitrogen.
                                     1032
-------
     Gaseous and participate organic stack emission rates do not appear to
present any significant problem.  Vanadium is the only trace element whose
emission rate is potentially harmful.   However, control Of total participate
emissions will significantly reduce vanadium output.

                                 ACKNOWLEDGMENTS

     This work was supported by the U.S. EPA, Industrial Environmental Research
Laboratory-Research Triangle Park, under contract number 68-02-1316, T.O.  No.
14.  The authors thank Dr. Graham Johnes, Mr. Z. Kowszun and Dr. Gerry Moss
and their staff at Esso for their cooperation during the field measurement
program; Dr. Peter Jones of Battelle for the loan of the gas adsorption column;
and John Langley and Stephen Brenan of the GCA staff for their assistance
in the sampling effort.

                                   REFERENCES

1.    Craig, J.W.T., G.L. Johnes, Z. Kowszun, G. Moss, J.H. Taylor, and D.E.
     Tisdall.  Chemically Active Fluid-Bed Process for Sulphur Removal During
     Gasification of Heavy Fuel Oil - Second Phase.  EPA-650/2-74-109, U.S.
     Environmental Protection Agency, Research Triangle Park, N.C., November
     1974.

2.   Foster-Wheeler Energy Corporation.  Chemically Active Fluid Bed Process
     (CAFB) Preliminary Process Design Manual.  U.S. Environmental Protection
     Agency, Research Triangle  Park, N.C.  EPA Contract Number 68-02-2106.
     December 1975.

3.   Werner, A.S., C.W. Young,  M.I. Bornstein, R.M. Bradway, M.T. Mills and
     D.F. Durocher.  Preliminary Environmental Assessment of the CAFB.
     U.S. Environmental Protection Agency, Research Triangle Park, N.C.
     Contract Number 68-02-1316.  Task Order Number 14.  June 1976.

4.   Dorsey, J.A., C.H. Lochmuller, L.D. Johnson, and  R.M. Statnick.  Guide-
     lines  for  Environmental Assessment  Sampling and Analysis Programs.  U.S.
     Environmental Protection Agency, Research Triangle Park, N.C., March  1976.

 5.   Standards  of Performance  for  New  Stationary  Sources.   Code  of  Federal
      Regulations, 40  CFT,  Part  60,  May  23,  1975.

 6.   Jones, P.W., A.P.  Graffeo, R.  Detrick,  P.A.  Clarke,  and  R.J.  Jakobsen.
      Technical  Manual  for Analysis  of  Organic  Materials  in  Process  Streams.
      EPA-600/2-76-072,  Environmental  Protection  Agency,  Research Triangle
      Park,  N.C.,  March  1976.   81  pp.

 7.   Title  40-Protection  of the Environment.   Part  60  Standards of Performance
      for New Stationary Sources.   Federal  Register, 36(247):24876,  December 23,
      1971.

 8.   Jonke, A.  A.,  E.L.  Carls,  R.L.  Jarry,  M.  Haas, W.A.  Murphy, and  C.B.
      Schoffstoll.  Reduction of Atmospheric Pollution  by the  Application  of
      Fluidized-Bed Combustion.   ANL/ES-CEN-1001,  Argonne  National  Laboratory,
      Argonne,  Illinois, June 1969.   65 pp.

                                     1033
-------
9.   Eimutus, E.G. et al.   Source Assessment, Prioritization of Stationary
     Air Pollution Sources.  Model Description.   EPA 600/2-76-032'a,
     U.S. Environmental Protection Agency, Research Triangle Park, N.C.,
     February 1976.
                                     1034
-------
                                                                           21-4
                             NO  CONTROL BY ABSORPTION
                               X

                                    G.  Sakash*

ABSTRACT
     The, elimination  ofa  nitrogen oxides ^ume4 exhausted ^fiom a. uAanium
nccov&iy operation -U>  accompLti>kcd by the, che.rn.ical. conveAAi.on o£ nitrogen
oxides to nitrogen dioxide, uAing puAe oxygen, and a AubAequcnt ab&o?iption
0$ the, nitrogen  dioxide,  by ia.pi.dty neciAcuJLating waten to faofw nWtic acid.
An exce^-6 ofa pome oxygen fad ^fiom a cryogenic &y&tcm iA tued to compJL&tnly
convafut aUL o^ the, nitsiogun oxA.de, QOA to nitAogun dioxi.de,>  The, exce-ii  oxy-
gen it> the, onty  QO*> UiLtwateJiy dit>c.haSLge.d to the, atmoApheAe..  The, nitsiic. acid
{fO.vne.d by the, ab^oftption pioceA* /u ?ie.cinc,uJtate,d untit the. deAisind  c,on.ce,ntna-
tion u> obtained.  The, acid u> then, pumped back to the 4cAap ticcovcfiy &y&tcm
faon. tieube.

                                   INTRODUCTION

     The high toxic hazard rating associated with oxides of nitrogen  (NO  )
are of major concern  to  all manufacturing facilities where the potential
of NO  fume emission  exists.

      It  is  the  intent of this paper to describe a process which  has  success-
fully eliminated  the N0x fumes emitted from the dissolution  of uranium dioxide
in  nitric  acid.
      The  process  chemically converts NO  to nitrogen dioxide  (N02)  using
 pure  oxygen;  the  N02 is concurrently absorbed by rapidly  recirculated scrubbing
 water to  form nitric acid.  The process utilizes an excess  of pure  oxygen to
 completely  convert all  of the NO  gases to N02 which is very  soluble  in  water.
                                 X
      This  process  has been in use at the General Electric  Company  in  Wilming-
ton,  North Carolina,  for three years.  It has completely eliminated all  visual
N02  fumes  and  reduced the N02 emission level to less  than  5  ppm  which is the
current  Threshold  Limit Value (TLV) as published by ACGIH.

                                    DISCUSSION

      In  the reclamation of scrap uranium dioxide, the  scrap  material  is
dissolved  in nitric acid.  During this dissolution process,  large  quantities
of gaseous vapors, in the form of nitric oxide  (NO) and N02,  are evolved.
These  gases are  identified as N0x and are exhausted as a brownish  plume,
creating a very  undesirable atmospheric pollution condition.
*General Electric Company, Wilmington,  North Carolina.

                                      1035
-------
     The TLV for NO, as published by ACGIH, is listed at 25 ppm or 30 mg/m3
of air.  The N0x fumes exhausted during the uranium dioxide dissolution pro-
cess is far in excess of this; therefore, it was necessary to devise a process
which would reduce this emission to an acceptable level.  This was accomplished
by the utilization of a device (Figure 1) in which oxygen and water convert
the NO  fumes into nitric acid.  This device consists of a conversion chamber
      X
where the NO  are converted to N02 and an absorption chamber where the NO
                                                                         2
is converted to nitric acid.  An excess of pure oxygen, fed from a cryogenic
system, is used to completely convert all of the NO  gases to N02.   The ex-
cess oxygen is the only gas ultimately discharged to the atmosphere.

     The absorption column, constructed of 317 stainless steel, measures 20
feet high and 4 feet in diameter and is divided into three compartments.  The
top five-foot section is used as the conversion chamber where the chemical
conversion of the N0x to N02 takes place.  The oxidation reaction occurs
readily at room temperature without a catalyst.  The middle ten-feet of the
column is used as a mixing chamber.  This portion of the column is filled with
ceramic Intelox packing.  In this portion of the column, the water and N02
mix to form nitric acid.  The bottom five-feet of the column is the acid
accumulation chamber.

     The scrubbing water, flowing at a rate of 250 gpm, is continually re-
circulated through the absorption column until the acid concentration is
increased to a reusable product.  The heat of reaction is controlled by
passing the recirculation solution through a stainless steel heat exchanger.
Continuous monitoring of the acid concentration is done at a main control
panel.  After the required acid concentration is reached, the acid is pumped
back into the dissolvers for reuse.

     The off gas consisting mainly of excess oxygen is drawn up through a
high efficiency filter housing, then through a Buffalo gas absorber scrub-
ber, and finally through another bank of high efficiency filters.  A Leeds
and Northrup N0x emission monitor is installed in the off gas stack and
continuously monitors and records the visible NO  fumes being exhausted out
the stack.  An alarm automatically sounds if a malfunction in the absorption
system occurs.

     Because the system is used in the reclamation process of radioactive
uranium dioxide scrap, constant monitoring for radioactive material is re-
quired.  This is done both on the recirculation solution by periodic labora-
tory analysis and on the exhaust gases by continuous stack exhaust monitoring.

     Chemically, the major reactions which take place during the absorption
are as follows:

          3NO + 1% 02 	»  3N02

          3N02 + H20  	» 2HN03 + NO
          2ND + ~\h 02 + H20        2HN03
                                    1036
-------An error occurred while trying to OCR this image.
-------
     The process (Figure 1) occurs as follows:

1.   Uranium dioxide is dissolved in nitric acid in the dissolution chambers.

2.   NO  gas generated by dissolution leaves the top of the dissolution cham-
     bers and enters the absorption column at A.

3.   Pure oxygen from the cryogenic oxygen system enters the absorption cham-
     ber at B.  When additional oxygen is required, it is automatically
     added through the pressure sensing V2, C2, T2 system.

4.   A water-acid mixture recirculating through pump, P, and heat exchanger,
     E, enters the absorption chamber through a spray nozzle at C.  Roto-
     meter, R, indicates flow.

5.   The excess oxygen is exhausted as spent gas through D and goes through
     absolute filters prior to exhausting to the atmosphere.

6.   As the concentration of the nitric acid formed is increased, it is
     sensed through T3; and when the desired concentration is reached, it is
     automatically pumped through control valve V3 into the dissolvers.

7.   Make-up deionized water enters the acid storage tank at G.  A level is
     automatically maintained by a float controller at F.

8.   A Leeds and Northrup N0x emission monitor, I, senses visual emission and
     records it at a remote control panel.

9.   A monitor, H, continuously monitors the off gas for radioactive
     particulate.

     It is conceivable that this system could be adapted to any facility
where control of NO  emission is a problem.
                                     1038
-------
                                                                              21-5
                    DEVELOPMENT  OF  A LOW EMISSIONS PROCESS
                      FOR ETHYLENE  BICHLORIDE PRODUCTION

           W.S.  Amato, B. Bandyopadhyay, B.E. Kurtz,  and  R.H. Fitch*
ABSTRACT
     Thii>  wofik coveAS the initial phase ofa the. development of, a low- emissions
ethylenc  oxyhydAochloAA.nation pAocess faA pn.oda.cA.nQ  1,1 dichloAo ethane
{ethylene dichloAide) .

     Firstly,  experimental woA.k on. an existing pilot plant scale, once.-thsiou.gk
ph.oc.Ui> was  used to obtain basetine emission data.  in mass o{> hydJLOc.aA.bon plus
ethylene  dic.htotu.de. pen. mass  Ofj hydA.oge.n cktofcide.  ^ed aj> a function o^  ie.acJ:o>i
tmpeAotuJKi  and peAc.nnt exce^-6 zthylzne. to the. nz.act.on.; and to ie4c£ve  potan-
&jjJi pnobtwA  which may OAAAZ. -in a >ie.c.ycle. opoAation, viz., whether catibon
monoxide.    oxtdize-d to cafibon di.oxA.dn by the. n.ii-
nation rniakt be. ne.c.u&an.y -in  oideA to 4aue &thyle.ne.  uoltie in the. tie&ctot
Jifa a c.ah.bon  monoxx.de oxidation unit weAe. ne.c&>Aariy ,  and whether, the. Ae.actoA.
wouJLd oxidize, methane, and ethane., theAe.by p>ie.ve.nting tkeJJi btu/dup -en the.
     Secondly,  tke. e.xi^ting onc.e.-thnou.gh piJLot plant IMOA  convened to a le.-
cycle. opeAation which then {^u.nctione.d 4ucce4-6 fatly and  yi.ztde.d em-ibtxion data
in moi4 Oj$  hydA.ocoA.bon plu* e.thyle.ne. dichloAide. pejt mo64  o& hydA.oge.n chic Aide.
fae.d 06 a  function o{, n.o,actoA tempeA.atuA.e. and pe.Ace.nt e.xceA& eJ.hyle.ne. to the.
A&actoA.   In poAticalaA, the. pn.oje.ct objective, o^ A~e.du.cing by 90 peA.ce.nt the.
hydAocaAbon piuA ethyte.ne. dichloAide. emiA&iont, ^Aom an  ethyle.ne. oxyhydAo-
chloAination pAoceAA, thAough the. Ae.cycting o^ A&actoA  o^-gat>eA, hai> been
            demom,tAate.d.
      Thirdly, vaAioitt,  operating di^icaltie^  weAe. aA-t>ej>Ae.d which womld be.
 impontant fan. fautuAe, control appLication*  and 4co£e-up e.^faAtA,  viz.,  the.
 incAe.aAe.d sensitivity  o^ the. pAocus to upsets in filom, tempeAotuAu, and
 conce.ntAotA.ons .

      Lastly, economic  analyses aAe pAesented  to dmonstAote the  competitive
 position oft the improved pAocess.
*W.S. Amato,  B.E.  Kurtz and R.H.  Fitch, Allied Chemical  Corporation, Syracuse
 Technical  Center,  Solvay, New  York.
 B. Bandyopadhyay,  Allied Chemical  Corporation, Syracuse Works, Solvay,  New
 York.


                                       1039
-------
     ThLt> Monk wo6 p&i&ofme.d in fiulfi'illme.n.t o& Project Number R0AP 27 AXM  20,
Contract NmbeA 68-02-1835, by Alti&d Chmic.al Corporation, lnd.(U>&uaJL Chwi.-
eo£6 V• C?HtfCl2 + H?0, arid

   cracking of ethylene di chloride:

                            2 C,H4C12 -- * 2 C2H3C1 + 2 IIC1

The overall reaction in this balanced process is:

                2 C2H4 + C12 + 1/2 02 -- * 2 C?H3C1 + H?0.

The oxyhydrochlori nation step is the most technically difficult part of  the
overall process and has had a profound impact on the industry since its  de-
velopment.  The first commercial ethylene oxyhydrochlorination process appear-
ed in 1964 and within the next few years nearly 90 percent of the classical
process for vinyl chloride from acetylene was displaced.

EXISTING PROCESS

     There are currently two basic variants of ethylene oxyhydrochlorination
processes, those employing fixed catalytic beds and those employing fluid
catalytic beds.  Most of the major producers employ a fluid bed.  Ethylene
(CzHiJ, hydrogen chloride (HC1), and air are introduced into the bottom  of
the fluid bed where the oxyhydrochlorination reaction takes place.  The  heat
released by the reaction is substantial, so the reactors are equipped with
internal cooling coils which are submerged in the fluid bed.

     Downstream product recovery involves cooling of the reactor exit gases
by either direct quench or a heat exchanger, followed by condensation of
the ethylene dichloride (EDC) product and water by-product which are separat-
ed by decantation.  The remaining gases still contain 1 to 5 percent by
volume of EDC, so they are further processed in a secondary recovery system
employing either solvent absorption or a refrigerated condenser.  The off-
gas is then vented to the atmosphere.  (See Figure 1.)
                                    1040
-------
f
\

1
V


x o
O h-

                                                         to
                                                         S-
                                                         o
                                                         o
                                                         o
                                                         X
                                                         o

                                                         O)

                                                         O)

                                                        ^
                                                        _c:

                                                         0)

                                                         en
                                                         X
                                                         
-------
      The  vent  stream  contains 85  to  95 percent nitrogen from the air employ-
 ed  as  a source of  oxygen  for the  oxyhydrochlorination reaction.  It also
 inc"[udes_  a  variety of other compounds which originat_e__ejther iin_the feeds
to the reactor or as by-products of the oxyhydrochlorination,  such  as  carbon
dioxide, carbon monoxide, EDC,  C2H4, methane (CHi+), ethane (C2H6),  ethyl
chloride and absorber solvent.

     The aspect of most current ethylene oxyhydrochlorination  processes
which contributes the most to hydrocarbon (HC) emissions is the use of air
as the source of oxygen.  The inerts (nitrogen)  accompanying the air neces-
sitate a once-through process.   If pure oxygen were used it would be possible,
at least in theory, to recycle  the reactor off-gases.   Only one plant  cur-
rently employs pure oxygen, but it does not recycle the reactor off-gases.

PROPOSED PROCESS

     Recycle of reactor off-gases will  eliminate the need  for  a secondary
recovery system and the resultant losses of solvent which  typically amount
to 0.001 ton per ton of EDC.   It will also eliminate unreacted C2H4.  Some
means will have to he employed  to limit build-up of feed impurities (CHa
and C2H6)  and by-products (C02,  CO, and ethyl  chloride), probably a combina-
tion of removal by chemical or  physical  processing and  a purge to atmosphere.
The purge, however, will amount to only a small  fraction of the vent stream
which characterizes present processes and it is  certainly  reasonable to expect
a 90 percent reduction in HC emissions  over present processes.

     The detailed technical approach currently visualized  may  best  be  explain-
ed by reference to Figure 2.
     A fluid bed oxyhydrochlorination reactor employing a conventional copper
chloride on alumina catalyst is fed with C?Hi+, HC1, and recycle gas contain-
ing added oxygen.  The C?hK is in 5 percent excess over that required by the
reaction stoichiometry, the oxygen in 60 to 80 percent excess.  Near the top
of the fluid bed a small stream of chlorine, amounting to 2 to 5 percent of
the HC1 feed, is added to convert the unreacted C?H4 to additional EDC.

     The gases leaving this reactor contain the products of reaction, EDC and
water, and other substances originating as unreactive impurities in the feeds
or as by-products.  The gases pass to a quench column where they are cooled
by circulating water, then to a condenser where EDC and water are recovered.
These are separated in a decanter and the non-condensable gases flow through
a recuperative heat exchanger where they are re-heated.

     The hot gases pass to a fixed bed oxidation reactor (shown dashed) em-
ploying a platinum on alumina catalyst where the CO will be converted to
C02 by reaction with the excess oxygen from the oxyhydrochlorination reactor.
The gases exiting the reactor are cooled against the entering gases in the
recuperative heat exchanger and passed to a booster compressor.  It may be
possible to eliminate the separate CO oxidation step if it can be shown that
CO can be oxidized to C02 by recycle to the oxyhydrochlorinator.  If this is
possible, then the need for post-chlorination is obviated.


                                    1042
-------
LJ
>~
CJ
LU
cc:
                                                                                       o.
                                                                                       ex
                                                                                      o:
                                                                                      I—
                                                                                      oo
                                                                          > o
                                                                                   ca
                                                                                   C£
                                                                                   o
                                                                                   oo
                                                                                   D3
/
^
_


V
\
,1 	
                                                                                     oe.
                                                                                     eC
                                                                                     Q_
                                                                                     UJ
                                                                                     CO
                                                                                                                           cu
                                                                                                                           o
                                                                                                                           O
                                                                                                                           s_
                                                                                                                           CL
 o
•I—
-M
 ro
                                                                                                                           S-
                                                                                                                           o
                                                                                                                           o
                                                                                                                           o
                                                                                                                          -a
x
o
                                                                                                                           0)
                                                                                                                           c
                                                                                                                           OJ
                                                                                                                          CD

                                                                                                                          TD
                                                                                                                          OJ
                                                                                                                          I/)
                                                                                                                          O
                                                                                                                          Q_
                                                                                                                          O
                                                                                                                          s-
                                                                                                                          o
                                                                                                                          M-
                                                                                                                          (1)
                                                                                                                          CU
                                                                                                                          CD
                                                                                                                          i-
                                                          1043
-------
     The majority of the discharge from the booster compressor is  recycled
to the oxyhydrochlorinator.   This recycle gas will  be predominately CO? with
the process as described here.   A small portion of the recycle gas is  purged
in order to limit the build-up  of unreactive impurities which  enter the sys-
tem with the reactor feeds,  for example:   CHit and C2He which  are  typical
impurities in the C2Hl+.

     This purge stream is first cooled in a recuperative heat exchanger and
then further cooled in a refrigerated vent condenser which removes most of
the EDC and ethyl chloride.   These are collected in a receiver from which
the purge gases, consisting of 80-90 percent C02 and 10 percent CH4 and C2H6,
are discharged through the recuperative heat exchanger and may be disposed
of via incineration.

     The net effect of this process will  be at least a 90 percent reduction
in HC emissions compared to typical present processes.

OBJECTIVE AND SCOPE OF PROJECT

     The objective of the project was to demonstrate that emissions from the
production of EDC by oxyhydrochlorination of 02^ could be reduced in a modi-
fied process by at least 90 percent from the levels encountered with typical
existing processes.  The modified process was to employ recycle of reactor
exit gas, oxygen feed, and whatever additional processing steps were deter-
mined to be necessary to control build-up of by-products in the recycle stream
and was to be economically competitive with present-day processes.  The pro-
cess performance was to be evaluated on a laboratory scale and a  preliminary
study of technical and economic feasibility was to be carried out.

              PRELIMINARY EVALUATION OF ALTERNATIVE APPROACHES
     The preliminary evaluation of alternative approaches considered the
following process alternatives:  oxygen feed with vent gas recycle on the
existing process; catalytic oxidation of vent gas; and thermal  incineration
of the vent gas.  The results can be summarized as shown in Table 1.

     The results of the economic evaluations of these alternatives show
clearly that the oxygen addition with vent gas recycle on the existing
process is by far the most economic choice.   The estimates are  based on a
700 MM Ib/year EDC plant.

                            EXPERIMENTAL PROGRAM
     The main chemical  reactions which occur in fluid-bed oxyhydrochlorination
of CHt  are:
                                    1044
-------
                       TABLE  1.   PROCESS  ALTERNATIVES

Economic Consideration
Net Capital Increase
Net Increase in
Manufacturing Cost
Unit Cost Increase,
it/lb EDC

Oxy Feed -
Recycle
($1 ,300,000)*
$377,000/y
0.054
Process
Thermal
Incineration
$1,750,000
$695,000/y
0.099

Catalytic
Oxidation
$2,250,000
$985,000/y
0.141

  *  Figures  in  parentheses  refer  to  savings  in  capital.



Principal Reaction:

             C?H4 + 2 HC1 + 1/2 02 - *C2H4C12 + 2 H20         (1)
                                Cupric Chloride
                                   Catalyst
Main Side Reactions:
                       C2H,f + 2 02          » 2 CO + 2 H20            (2)


                       C2H^ + 3 02 -        » 2 C02 + 2 H?0           (3)


                        C2H4 + HCI^         »C2H5C1                  (4)


                         2 CO + Op          » 2 CO 2                   (5)


QUESTIONS NEEDING RESOLUTION BEFORE FINAL DEVELOPMENT OF A RECYCLE PROCESS

Question 1:  Will CO build-up in the recycle system reach levels which will
             cause explosive conditions?

             Will the ratio of C02/C0 remain at about 2 in the recycle system
             as it does in the once-through system?

             Can the reactor system effectively oxidize CO to C02?
                                    1045
-------
Question 2:  If CO is not oxidized in the reactor, but is oxidized in a
             separate unit, how do we avoid the loss of CoH^ value by oxida-
             tion?

Question 3:  Can post-chlorination effectively work on the exiting reactor
             gases to convert unreacted C^Hi4 to EDC?

Question 4:  What is the effect of the reactor system on HC impurities, such
             As CH^ and C2H6, which might occur in the C2H1+ feed?

     With these questions before us, the scope of an experimental program
was evolved as follows:

SCOPE OF THE EXPERIMENTAL PROGRAM

1.    Once-Through Runs

     (a)  Baseline runs at 230-240°C with 5, 10, 15 percent excess C2H,4 and
          60-70 percent excess 02 (as air).
     (b)  Runs with CO and C02 additions.
     (c)  Hydrocarbon additions:  CH^ and C2H6.
     (d)  Post-chlorination work.

2.    Recycle Runs (02 Feed)

     (a)  Runs at 230, 240, 250°C with 5-20 percent excess C^ and 50-115
          percent excess 02 with vent rates of less than 10 percent of total
          off-gases.

RESULTS

     The results of correlation of emissions data from the once-through pro-
cess are:

   (Mean - 231°C) EM0 = 2.47(lQ-?) + 1.35(10-3) XSETH                 (6)

   (Mean - 240°C) EM0 = 1.32(lQ-?) + 1.35(1Q-3) XSETH                 (7)

     Where

   EM0 = Emissions in    VQ ur-i—  f°r once-through operation.


XSETH = Percent excess ethylene to reactor.

     The results of correlation of emission data from the recycle process are:

   (Mean - 230°C) EMR = -1.08(10^) + 2.05(1Q-I+) XSETH                (8)

   (Mean - 242°C) EMR = -G.ASdQ-1*) + 2.05(10-4) XSETH                (9)

   (Mean - 251°C) EMR = -8.0300-14) + 2.05(1Q-'4) XSETH               (10)


                                    1046
-------
     Where
   CM    c •   •     •   kg(HC + EDC) f        -,        . .
   EMp = Emissions in -|/ — rrpi — - for recycle operation.
XSETH = Percent excess ethylene to reactor.

     Comparison of the emissions from the once-through and recycle operation
at the same reactor temperature and percent excess ethylene is shown in Table
2.  Table entries show % reductions in emissions -n of recycle process com-
pared with once-through process at similar values of % excess ethylene and
reactor temperature, T°C.
       TABLE 2.   COMPARISON OF EMISSIONS DATA, RECYCLE AND ONCE-THROUGH
                Table entries show % reductions in emissions -n
                of recycle process compared with once-through
                process at similar values of % excess ethylene
                and reactor temperature, T°C.
                       % Excess Ethylene   230°C   240°C
5
10
15
20
97.2
95.1
93.6
92.4
97.4
94.2
92.3
91.1

     This result is also shown in Figure 3.  Thus, it can be seen that the
original  objective of this investigation, viz., the reduction by 90 percent
of the emissions from the EDC process through implementation of a recycle
scheme has been positively demonstrated.

                                 CONCLUSIONS
EXPERIMENTAL RESULTS OF ONCE-THROUGH OPERATION

     •    Crude ethylene dichloride (EDC) purity varied inversely
          linearly with reactor temperature.

                                    1047
-------
    100
   CO
   52;
   o
   H
   CO
   CO
   w

   ^
   H

   55
   O

   £
   o
   t>
   o
   w
   w
   o
   f£
   w
     90
0
2
                FOR  R/R+V ^0.9

SUPERFICIAL VELOCITY       1.2  ft/sec

        EXCESS OXYGEN -  50fo TO  100^  4=

                R = RECYCLE  RATE

                V = VENT RATE
6
  8     10    32    1



% EXCESS  K'HIYLKNE
                            16    18
                                                                20
          Most  commercial plants  opcrtite tit  50-70%  excess

          air and 5-10% excess C2II.} .
Figure  3.  n - Percent reduction in  emissions recycle  compared to once-through

          vs. % excess ethylene with  reactor temperature as parameter.
           Kilograms of  HC plus EDC per  kg of HC1  emissions could be

           correlated as a linear function of excess  C2Hi+ to the re-

           actor with reactor temperature as parameter.

           Ethylene efficiency decreases with higher  operating tempera-

           ture due to increased oxidation of ethylene.
                                  1048
-------
      •    Addition  of CO  to  the  feed  gases  indicated  that  CO was
           oxidized  to C02; also  large additions  of C02  to  the feed
           gases  did not adversely affect the  reaction.
      •    Methane passed  through the  reactor  system unchanged and
           C2H6 was  significantly reduced (80  percent  in this case)
           over inlet values.
      •    Post-chlorination  of excess C2H4  in the  reactor  exit gases
           was feasible provided  chlorine was  injected into the gas
           phase  and not the  catalyst  itself.

 EXPERIMENTAL RESULTS OF RECYCLE  OPERATION

      •    Crude  EDC purity varied inversely linearly  with  reactor
           temperature, but was a few  tenths of a percentage point
           lower  than the  once-through purity  at  the same temperature.
      •    The C2H4  efficiency  for the recycle operation was higher
           than that for the  once-through process due  to the recycling
           of off-gases.
      •    Ethylene  efficiency  decreases  with  higher operating tempera-
           ture due  to increased  oxidation of  ethylene.
      •    All runs  of this report involved  the recycling of greater
           than 90 percent of the reactor off-gases.
      •    Kilograms of HC plus EDC per kg of  HC1 emissions could be
           correlated as a linear function of  excess C2H4 to the reactor
           with reactor temperature as parameter.
      •    The recycle operation  is more  sensitive  than  the once-through
           operation and requires closer  control  for successful  operation.

 COMPARISON

      Comparison  of  the emissions data for once-through  and recycle operation
 indicates  that the  percent reductions of HC plus EDC  emissions have exceeded
 90  percent for all  runs conducted in  this work.
      It was  possible  to develop  a  computer  simulation program,  based on
 actual experimental  data, which  could  duplicate  vent  gas compositions and
 superficial  velocities at given  reactor  temperatures  and pressures for
 given  excess CaH^/oxygen  values  to  the reactor.  Based  upon  the optimum expe-
 rimental conditions  from  recycle  operation, viz.,  230°C, 5-7  percent excess
 C2Hit  and 50  percent  excess oxygen,  together with this simulation program, it
 was found  that the  recycle process  was competitive  with  existing processes to
 within a tenth of a  percent  of total  incremental  capital and  manufacturing
 costs.

      The work of this  report has  revealed that the  newly-developed recycle
 process is much more  sensitive to  upsets  than  the conventional  once-through
 pjigcess.

     Before a commercial prototype of the proposed  process  can be  designed,
these sensitivity factors  must  be more fully explored along with the  estab-
lishment of well-defined requirements necessary for stable  operation.   Thus,
it is recommended that further  work be performed  on the newly-developed
recycle process  through the  use of computer  control technology.

                                     1049
-------
          SESSION  22
AIR QUALITY AND EMISSION  TRENDS

    ChaiVima.yi'  R.E.  Neligan
Environmental Protection  Agency
              1051
-------
                                                                             22-1
        THE  IMPACT  OF EMISSIONS CONTROL TECHNOLOGY  ON PASSENGER
               CAR HYDROCARBON EMISSION RATES  AND  PATTERNS

                                    F.  Black*

ABSTRACT
     A bnJLu-h kiAtony o& the. Ae.guJLation o& pat>£e.ngeA cent hydAocaAbon mi&&
tnciading the.  contAol. AtjAtzmA and de.vi.ceA utitize.d &OA e.mi^^i.on Ae.daction, AJ>
pAeAe.nte.d.  ThAne.  typeA Of} e.mit>Ai.oni> cum. diAcuAAe.d'-  cAankca&e. ve.ntsULati.on,
e.vapoAative. ga&otine.,  and combustion ex^iaaAi,  OnJLy e.x.hauAt and e.vapoAative.
          aAe.  o^ Asignifai-cance, and d&taitad kydAoc.an.bon patt&inA ^on. th&>£ oJiAi.on&  and the. degree o{, control
            hydA.ocaA.bont> OA gfioa.p& o^ hydAocaAbonb  JJ>  vafuabte. with the. contAot
         HydAocaAbon emiAb-ion. patteAnt, afte. pAeJ>e.nte.d at, a function ofi contAol
           a  va?u.&ty o£ catalyst and noncatalyAt  ve.hic£e^>.   1ncAe.cu,e. -in
      wth ve.hicte. ag
-------
                             EMISSION TRENDS

     In 1971, the National Petroleum Council projected decreases in annual
emissions from motor vehicles as shown in Figure 1 (1). These estimates were
based on emission rates predicted using the 1968 Federal Test Procedure and the
promulgated schedule of emissions reduction. The current test procedure would
increase these estimates, but the trends illustrated are valid.  The currently
utilized urban driving simulation and analytical procedures are considered more
accurate than those used previously.  One must be careful when evaluating
historical data.  Driving patterns typically used in vehicle testing have been
changed and these changes affect the emissions significantly.

     Table 1 illustrates the magnitude of control called for by the legislation.
                 TABLE 1.  EXHAUST EMISSION REDUCTION
                         Hydrocarbons,
                         g/mi.   g/km.
                    Carbon Monoxide,       NOx
                      g/mi.  g/km.    g/mi.  g/km.
Pre-1968 vehicles
(uncontrolled)*

Statutory levels
(highest degree of
control called for)
 9.1   5.7


0.41   0.25
  97     60.2     3.34   2.07
 3.4     2.11     0.4    0.25
Percent reduction
95.5
96.5
89.1
*low altitude, 49-state passenger cars (excluding California) (calender year
 1976 values) (2).

The emission rates for uncontrolled vehicles were derived from actual tests with
320 vehicles, 1963-1967 model years, using the current Federal Test Procedure
(2, 3).

     Manufacturers responded to this call for reduction in emissions with many
engine modifications and varied emission control packages.  They have been
effective in reducing emissions and, although not called for by law, many
production engine groups are already emitting exhaust hydrocarbons at rates less
than the statutory 0.41 g/mi level in low mileage vehicles.

     In pre-emission control automobiles, there were three major areas of
hydrocarbon emission:  combustion exhaust, crankcase blowby, and evaporative
loss from the fuel system.    Table 2 illustrates the relative contributions of
each.
                                     1054
-------
    cc
    <
    LU
    >
    CC
    Ul
    Q.
    CO
    o
    CO
    CO
    CO
    CO
    LU
              CARBON
             MONOXIDE
                               TOTAL VEHICLES IN USE
                                               HYDROCARBONS
            NITROGEN OXIDES

                           PARTICULATE MATTER
                  1960      1965
         10
1955
1980     1985
              Figure 1.  Mobile source annual  emission  rates  (1).
CONTROL OF HYDROCARBON  EMISSIONS

     The advent of emission  control has eliminated crankcase blowby emissions  J
low mileage cars,  and exhaust  hydrocarbons have been reduced greater than 95
percent for many vehicles.   Crankcase blowby has been eliminated with Positive
Crankcase Ventilation Systems.  These systems provide for circulation of air
through the crankcase and  for  drawing the circulated air and blowby gases into
                                    1055
-------
                  TABLE  2.   PASSENGER  CAR  HYDROCARBON  EMISSIONS
                                Exhaust        Blowby          Evaporative
                             g/mi.  g/km.    g/mi  g/km.       g/mi.   g/km.
  Pre-control  *               9.1     5.7     4.1   2.5        2.53     1.57
  Percentage  of  total        57.9            26.1              16.0
  *     Low  altitude, 49-state  passenger  cars  (2)

the intake manifold, where they are carried to the combustion chambers.  The
systems perform very well when maintained, but are subject to total failure if
the critical control valve becomes clogged.

     Engine modifications, catalytic converters, and combinations of both have
been used to reduce exhaust emissions.  Generally the engine modifications
relative to exhaust hydrocarbon emissions permit operation on the lean side of
the stoichiometric air-fuel ratio.  Lean operation is attractive from an emis-
sions standpoint as illustrated in Figure 2.  Emissions of hydrocarbons (HC), CO
and NO  are favored by lean operation.  Engine performance can suffer under lean
operation, however, and most of the modifications are designed to maintain
drivability throughout normal  driving patterns.  The catalyst systems are
designed to complete combustion of the hydrocarbons emitted from the engine to
C02 and H20 prior to their being emitted from the tailpipe.  Tim's permits
operations at air-fuel ratios  favoring performance (12.5 best power, 15.7 best
economy).  The variety of catalyst systems employed is numerous and no attempt
will be made to detail them.  Hydrocarbon emission patterns from several systems
are presented later in the discussion.

     Evaporative emissions have been reduced by using adsorption-regeneration
systems which use canisters of activated carbon to trap fuel tank and carburetor
vapors and hold them until such time as they can be directed into the induction
system for burning in the combustion chamber. These systems first appeared on
motor vehicles in 1971.  The standard of 2 g/test appeared to be easily met by
the control systems utilized.   The test procedure adopted by the federal certi-
fication laboratory (charcoal  canister technique) (5), however, has been found
to dramatically underestimate  these emissions.  More accurate procedures (Sealed
Housing for Evaporative Determination, SHED,  technique) (6) indicate actual
emissions from post-control vehicles in the field to be near 12.0 g/test with
about 12.1 g/day diurnal  emissions (7).  This equates to about 1.76 g/mi (1.09
g/km) (2).  Ambient conditions (temperature and pressure) dramatically effect
evaporative losses.  Tests using the SHED technique with post-control vehicles
in Denver gave 34.8 g/test with about 47.2 g/day diurnal (8).  This equates to
5.51 g/mi (3.42 g/km).  This is probably due  to the lower atmospheric pressure
and different fuel composition used in the Rocky Mountains.  Exhaust hydrocarbon
emissions are also somewhat higher at high altitudes.

                                      1056
-------
                                  14    16    18

                                 AIR-FUEL RATIO
        Figure 2.  Effects of air-fuel ratio on exhaust composition (4).

     Heywood effectively projected the significance of the higher than expected
evaporative emissions on outyear aggregate emission rates as illustrated in
Figure 3.   The model  used by Heywood assumes deterioration of the crankcase
emission control  systems as predicted by Voelz, et al.  (9). It also assumes
applicable exhaust standards are met during the life of the vehicle.  The
potential  importance of evaporative emissions in outyears is readily apparent.
The Environmental Protection Agency (EPA) is currently revising the evaporative
emissions  certification procedure to incorporate the SHED technique for the
1978 model year.   Curve B is thus the most accurate prediction.  It is based on
95% control in the 1977 model year, however, and this probably will not be
realized until the 1978 model year.

Patterns of Hydrocarbon Emissions

     Recognizing the relative contribution of exhaust,  evaporative, and crank-
case hydrocarbon emission, an examination of the detailed patterns of each
ensues.  The mass and pattern of hydrocarbons emitted from motor vehicles are
dependent  on many factors.  The fuel;  the emissions control system; the vehicle
history (wear, deposits, vibration, and maintenance); ambient conditions
(temperature, pressure, and humidity); and driving patterns are all significant
controlling factors in vehicle emissions. Therefore, the conclusions of any
reported study must be tied to the specific conditions  under which the study was
conducted.  The trends indicated are generally true, but the absolute  numerical
quantities reported are of value only  to indicate the trends.
                                     1057
-------
                  LLJ
                                          EVAPORATIVE
                      1960        1970        1980
                                      YEAR
1990
  Figure 3.   Normalized urban  aggregate  HC  emission  rate  from  in-use  auto-
  mobile population,  baseline  case.   Crankcase,  exhaust,  and evaporative HC
  emissions indicated separately.   Curve A-Conventional  assumption that
  evaporative HC are  95% controlled  in post-1970  vehicles.  Curve  B-Baseline-
  field data used to  estimate  HC evaporative  emissions from 1971-1976 model
  year vehicles, with 95% effective  control assumed  in 1977 and subsequent
  model years.   Curve C-Field  data based estimate of evaporative HC used for
  all  post-1970  vehicles.  (8)


     Tailpipe hydrocarbon emissions  have been examined in some detail by gas
chromatography (GC)  for many new and older  cars  over the  past  several years in
the EPA mobile source program at Research Triangle Park  (RTF).  The exhaust
patterns to be discussed were  determined for, generally,  low mileage  automobiles
(<10,000 miles)  tested with the  1975 Federal  Test Procedure (FTP)  (5).  This
procedure calls  for  driving the  vehicle  on  a  dynamometer  over  a defined urban
driving cycle.

     Table  3 presents the  relative exhaust flow and  hydrocarbon  concentration
of the modes under which vehicles  are normally operated  and the  approximate
percentages of  the FTP  driving cycle in each  of these modes.

     The hydrocarbon patterns reported below  were determined  using a  consistent
batch  of lead-free gasoline the properties  of which  are  shown  in  Table  4.
                                     1058
-------
          TABLE 3.   EFFECT OF VEHICLE MODE ON  HYDROCARBON  EMISSIONS
                                Exhaust (10)          FTP  Time  In  (8)
           Mode              Flow        HC Cone.          Mode,%
       Idle              very low          high             18.2

       Cruise
low speed
high speed
Acceleration
moderate
heavy
Deceleration
low
high
high
very high
very low
low
very low
low
moderate
very high
30.2*
27.7
23.9

      *    The average speed of the 1975 FTP is 19.7 mph.
       TABLE 4.  FUEL USED FOR EXHAUST HYDROCARBON EMISSIONS TESTING
       Research octane number                                  93.2
       Motor octane number                                     84.7
       Reid vapor pressure, psig                               10.2
       Distillation, ASTM D-86, °F
                   10%                                         123
                   50%                                         199
                   95%                                         325
                  100%                                         383

       FIA analysis, %
                  aromatic                                     24.0
                  olefins                                       8.3
                  parafins                                     67.7
This fuel is reasonably representative of typically marketed lead-free gasoline
The detailed hydrocarbon composition of the test fuel  has been previously
reported (11).

     Tables 5a and 5b present the hydrocarbon exhaust  emission patterns for a
wide variety of catalyst and non-catalyst automobiles.   Detailed  GC  procedures
(12) were used to produce all listed data.   The totals  were based on summations
of appropriate individual  hydrocarbon concentrations.

                                     1059
-------











oo
•z.
OL
LU
1—
1—
=c
D-

"^
O
1 — I
oo
oo
h- -H
LU
^^
UJ
Q
"^~~
re
LU
a
	 i
-
1C
I—
oo
ID
C-T-
*--l.
rc
X
LU
LU
	 I
i — i
CQ
O
•^_
O
h-
ID
>
^;
CO
-a
i —
•<
+j
o
i—
ai
-o
>>
-C
CU
-a
s-
o
^9- LA-
'S C
*\ 4_
C 0)
O r-
-Q O
S-
ro
o ai
0 C
s- a)
-O r—
>^ ^
-C ^
4->
i — LjJ
ro
-M
0 0
+J
M- ro
0 g
o
cu s-
01 c£
ro
i ^
c cu
cu c
o a)
S- Nl
cu c
o- a>
CQ

cu
c
cu
">>
+->
cu
o
<
M-
L(—
ro
ro
Q_
cu
ro
-C
0)

-t->
1/1 ^
r; c
ro o
-C _Q •
X S- •!-
LU ro E
o -~-
i— O cn
ro s-
+-> ~o
0 >,
h- 31

'o E
S_ 0)
-*-> 4-J
cr to
0 >,
CJ CO

OJ
U
J=
ai
>




0 0
0 0
r-^ i—




co ro
CM CO
ro LO

0 0
CD r~-
LO ro
CM ro



r--- LO
CT. «*
CM r^



0 <*•
ro o
C\J CO
CM i —


ro **•
^r o
CM CM




•* 0
O CM
i— LD
CO CM
LO LO
0 r-
«* "*


cn cn
CM cn
cn co





LO r^~
i— «*•
_ i 	





+->
in r- s_
>> ro o
r— E 4-)
i ro s- cj
c: +-> cu ro
o ro ^r cu
C U 4-> ^,

a
1 — ' >i / — ^
cj ro s- i-»
CM -^- ~o ro o
r^ CD r^ N 4-J
cn LO cn ro o o
r- CO r- S l_^f




o *^
CM LD
rM ro




0 0
•— LO
1

LO r-- ^
CM ^T 53-
r^ o o
i — CM CM



CT) O LO
CM LO r-~
LO t-~. 00



r-^ 01 co
CO ej- LO
oo cr> cn
i i r—

co ro ^j-
CO f~- <3-





LO LO CM
cn ro r^
r— ro CXJ
LO LO CT)
CO LO O
i— LO r-
LO LO LO


CO I — • =3-
ro ro ^j-
cn LO LO
r-




O^ LO LO
^ LO CXJ
•





c c " c:
O 4-J O 4-J Q_ O 4-J
•i— (/I T- m E -i— t/)
QJ4->>> QJ 4-J >, ;s CU 4-> >>
CT ~O n3 CTi~a ro Cr>~O n3
CT -r- +-J C -i— 4-J S- C T- 4-J
•-— X rQ -f— X ro T- -r- X 03
t^OU LOOUra LOOCJ
>i "0
S- S- "
Z3 Q O rO CD " O
U- I-H U_ ~O •— 1 (T3 I-H
un • LO c LO > n3
r^ >iCO r^ ro c\j r-. cp 0.0
cn r— ,— cns-o cn_c;Ei-n
i — cxco ( — oro i — o i— H ro
1060
-------











00
~*^.
fy*
; i i

1 —
1—
CL


^^
o
t— -H
OO
OO
t— H
^—
UJ
„ — ,
i i i
Q
^
i i i
Q
— 1
)
L__ 2C




1 	
0 E
S- OJ
-M 4-1
C >
C-J OO

(V
r—
O
.) —
r-
01
^





CTl
CM
i-^



0
0

CO




CO
t^J-

00
ro



IH
LO
1 —



OJ

00
OO


CTi
I--.

i —



CTl
00
r^

00
CO
t~^
ro



CO

^j-




co
00





"O
CD
«i —
'1
•r- CD '—
+-> cn c
n3 S_ ro
S- ra CD
-U -C r—
OO O * — '


O CD

-a co o o
C 0 S_
o ;> O-i—
n: cj ^ — cn





! 	 1
LO
CO



^
CD

f—
i —



CD
CO

cn
ro



LO
, —
OO



ro
CO

cn
i —


LO
O

00



cn
LO
o
1
^
<*£>
cn
CXI



S

*^j-




01
ro





o

c~
0
S-
4_)
U C C
CD re !-
i— cu r;
CD r— -Q
i_
OJ '— -Q
r- O 1— 1
LO -U C_)
>> O
!- S- 0
r" o cj-
!_> 	 -•*





CD
cn
CO



o
ro

00




CD
ro

ro
•—



LO
LO
ro




CO

CO



co
LO





ro
OO
^«

LO
i—
CO
r*-



CTl

^}-
CM



ro





i
c c:
•a o o +-1
CD T- T- LO

C-) ro i —
i— 3 "O ra

Z3 QJ X ro
-a i- o o

^ — - / — 1
-M O t— i
CD +-I CJ
C O
CJ S- S- 00
s: o a-oo
cr rr^oo





LO
00
ro



^
CTl






o
r—

1 — ^
•—



^
LO




O
CO

1 —
CM


r- —
^3-

i —



ro
^
^3-

cn
LO
CO
LO



cn

cn
1



CD





1
>> = C
ra O O -!->
3 T- T- LO

CD O ro i—
CD Z3 ~a ra
S- "O -^ 4->
-£Z CD X ra
4-> S- O U

,^^
0
o +->
> o
r- S_ O
OQ.ro
> 	 	 r —
1061
-------
     In general, vehicles employing oxidation catalysts for exhaust hydrocarbon
control have a lower percentage of unsaturated hydrocarbons than do vehicles
without exhaust hydrocarbon control.  Also, the methane percentage is increased.
The net hydrocarbon photochemical reactivity is thus reduced.   Of interest to
those who use acetylene as a tracer for auto exhaust hydrocarbons in ambient air
studies, the acetylene relative abundance in auto exhaust has  been reduced
significantly.

     Four vehicle configurations of potential significance in  outyear automobile
populations are listed in Table 5b.  The stratified charge engine, in which net
lean combustion is used for emission  control, yields a higher  percentage of
unsaturated hydrocarbons than previous non-catalyst cars.  The Chrysler Corpo-
ration "lean-burn" engine also yields a higher percentage of unsaturated hydro-
carbons.  Two control systems are listed which are leading candidates for
combined HC, CO, and NOX control.  Both the dual-bed and three-way catalyst
systems yield lower percentages of exhaust unsaturated hydrocarbons than non-
catalyst cars.  This combined with the higher percentage yields of methane
results in significantly lower net HC photochemical reactivity.

     Exhaust aldehyde data are also listed.   Here too, the percentage yields are
lowest for vehicles employing catalysts for exhaust emission control.  In
general, the aldehyde rates are an order-of-magnitude lower than the hydrocarbon
rates.

     The detailed patterns of evaporative hydrocarbon emissions are dependent to
a great extent on the gasoline used and ambient temperature and pressure.  As
previously mentioned, evaporative emissions are much greater in Denver than in
Los Angeles,due primarily to the lower pressure and temperature of Denver and
the attendant greater percentage of low molecular weight hydrocarbons in the
fuel.  This situation should be considered when attempting to  define automobile
sourced hydrocarbons in ambient air studies.  In cities of higher altitudes the
contribution of evaporative gasoline  is greater.  In the vehicle there are two
primarily sources of evaporative losses:  the fuel tank and the carburetor.  The
carburetor bowl potentially reaches higher temperatures and consequently the
evaporative losses from this source reaches higher average molecular weight.
The bowl total volume is also of significance to total mass of emission. The
temperature, vapor volume, and pressure of the fuel tank are all of significance
to tank losses.  Many parameters impact the composition of evaporative losses,
but in  general it can be stated that  the emissions are very much dominated by
the Ci+ to C6 paraffins.  Normal-butane and isopentane generally account for in
excess  of 50 percent of the total evaporative losses.  Table 6 is representative
of typical evaporative emissions.  The dominance of n-butane and isopentane, and
C^-Cg paraffins in general is readily apparent.

     Crankcase blowby has been essentially eliminated with the positive crank-
case ventilation systems.  Voelz et al. estimated the systems  to be about  98%
effective  (9).  This estimate presumes proper maintenance of the system.   In a
study of over 75,000 vehicles in 15 metropolitan areas,  17 percent of the
vehicles tested needed PCV system maintenance and 3.6% were emitting some
crankcase fumes into the environment (9). The hydrocarbon composition of crank-
case emissions is dominated by carbureted mixture  (^85%) and combustion gases
(^15%) (15,16,17).  The patterns are thus very  dependent on the fuel being  used.

                                      1062
-------
Lubricating oil does not contribute to crankcase emissions in a significant
manner (18).

               TABLE 6.   PATTERNS  OF  EVAPORATIVE  HYDROCARBONS
                                      Fraction  of  total  evaporated  hydrocarbons,
                                      vol.  %
                                      Carburetor  (13)             Fuel  tank  (14)
 Paraffins
       n-butane                          11.9                         30.5
       isopentane                        45.3                         26.4
       U-C6                              66.2                         76.0

 Olefins
       Cit-Ce                               8.5                         14.0

 Aromatics
       C6-Cy                               1.0                          1.3
     The pattern of the aggregate hydrocarbon emissions from catalyst equipped
vehicles is very much dominated by the paraffins.   As discussed,  this is  due to
the increased relative abundance of evaporative emissions with its paraffinic
dominance and the greater activity of the catalyst for the exhaust unsaturated
hydrocarbons.  Table 7 is illustrative of the relative abundance  of the various
hydrocarbon classes.

     For the purpose of illustration the 1975 Chevrolet Impala exhaust class
weight percentages previously discussed were assumed to be typical of the
model year.  Also, it was assumed that 55% of the total evaporative emissions
were from the carburetor and 45% from the fuel tank.  About 75% of the aggre-
gate emissions are paraffin, 9% aromatic, 15% olefin, and 1% acetylene.  These
estimated are for the 1975 model year.  Outyear estimates vary as the relative
abundance of exhaust to evaporative emissions changes.  The 1975 model year
emission factors suggest that the aggregate is composed of 41% exhaust, 59%
evaporative  (2).  For the 1978 model year, the aggregate is estimated to be
35% exhaust, 65% evaporative (assumes exhaust standard is reduced to  .41  g/mi).

Emission Control Systems Deterioration

     Although mentioned, but not emphasized, deterioration factors are very
important when attempting to assess the importance of the automobile to atmos-
pheric pollution.  Most of the data found in the literature is for low mileage
(<10,000 miles) vehicles.  With deterioration of emissions control systems, both
hydrocarbon mass and patterns will change.  Figure 4 is indicative of typical
exhaust control deterioration factors.  These factors are based on linear
deterioration at the rate of 1  percent per year for the pre-1968 model years, 10

                                     1063
-------
percent per year for the 1968-1974 model  years,  and 20 percent per year for the
post 1974 model  years (2).  No inspection/maintenance program is assumed.   Such
programs would decrease the deterioration significantly.   The factors for post-
1974 model years were based on tests of prototype vehicles.   The actual dete-
rioriation factors for 1975 and later model  year production  vehicles should
likewise be higher than 1968-74 model due to the increased sophistication of the
control system,  i.e. catalysts, air pumps etc.

     The actual  emissions increases experienced  with post-1974 in-use vehicles
on aging may be  significantly different from that indicated  by the published
deterioration factors.   The deterioration factors reported were determined with
specific vehicles chosen for the durability  fleet.   They  were maintained  on
owner's manual prescribed maintenance schedules  and were  not abused as consumer-
owned vehicles might1be.  For example, these vehicles do  not have their catalyst
exposed to leaded gasoline, their carburetors are not tampered with to achieve
better drivability, the spark advance is not set variant from manufacturers'
specifications,  and their catalysts are not physically removed from the exhaust
system.

     Ongoing surveillance programs are carried out by EPA under which consumer
automobiles are tested and actual emissions from in-use vehicles are determined
(19).  Catalysts 'having been on automobiles for only two years, it is difficult
to accurately project outyear emissions increases.

                                  CONCLUSIONS

     In conclusion, the advent of emission control  has dramatically decreased
the rate of emission of hydrocarbons into the environment from motor vehicles.
Crankcase emissions have essentially been eliminated.  Exhaust emissions have
been reduced as much as 95% in some vehicles relative to uncontrolled vehicles.
Vehicles using catalysts have seen a significant shift in the pattern of emis-
sions with unsaturated hydrocarbons being reduced to a greater extent than
saturated.  The advanced catalyst systems also show significant increases in the
relative abundance  of methane in the exhaust.  Vehicles using lean-combustion
for emission control reduce saturated hydrocarbons to a greater extent than
unsaturated hydrocarbon.  Evaporative hydrocarbon emissions are dominated by the
Ct^-Cg paraffins with n-butane and isopentane being the most significant.

     In general, hydrocarbons emitted from outyear motor vehicles will show an
increased dominance of paraffins due to the impact of catalyst on exhaust
hydrocarbons and the increased relative abundance of evaporative hydrocarbons.
The relative abundance of acetylene will decrease.

     Ambient conditions significantly impact emissions.  Hydrocarbon emissions
are generally greater at high altitudes than at low altitudes, particularly
evaporative emissions. Emissions will in general increase with vehicle age,
potentially as much as a factor of three over ten years.

                                ACKNOWLEDGMENT

     The author wishes to acknowledge and express gratitude to Mr. Larry High
for the GC exhaust  hydrocarbon analysis and Mr.  Fred Stump and Mr. William Ray
for the aldehyde analysis.

                                     1064
-------
                                                00
                                                       o
                                                       DC
                                                       <
                                                       LU
                                                    co>
                                                    cc_
                                                    LU O
                                                    C3CC
                                                    NOISSIIAI3  I aV3A
JNOISSIIAI3 X aV3A
o
o
      aoiovd i\ioiivaoia3i3a
                                                                       I
                                                                      -l->
                                                                      ^r
                                                                      CT)
                                                                      S-
                                                                      o
                                                     co to
                                                     S- S_
                                                     O 13
                                                    +J O
                                                     u
                                                     03 a>
                                                    (+_ +j
                                                       to
                                                    •I-  I
                                                    4-> CTl
                                                    ra ^t
                                                    5_ * __ ^

                                                    O
                                                    •r- Ul
                                                    S- 0)
                                                    O) i—
                                                    •M U
                                                    >
                                                    -£= O)
                                                                       to i —
                                                                       Z3 O
                                                                       fO to
                                                                      _C ro
                                                                       X CD
                                                                       0)
                                    1065
-------
          TABLE 7.  ESTIMATE OF COMPOSITE HYDROCARBON EMISSIONS
                    FOR TYPICAL 1975 PASSENGER CAR *


Total Hydrocarbon, g/km (2)
Percent of total , wt.%
C^-Cg paraffins, wt.%
Olefin, wt.%
Aromatic, wt.%
Paraffin, wt.%
Acetylene, wt. %
Exhaust
.75
40.8
8.24
8.34
8.03
23.29
1.11
Evaporative
1.09
59.2
40.4
6.67
.66
51.9
-
Composite
1.84
100
48.6
15.0
8.7
75.2
1.1

*low altitude 49-state car
                              REFERENCES

1.   National Petroleum Council.   Environmental  Conservation:   The Oil
     and Gas Industries.  A Summary.   Vol.  1.   Washington,  D.C.,  1971.
     106 pp.

2.   U.S. Environmental Protection Agency.   Compilation of  Air Pollutant
     Emission Factors.  AP-42 part B.  2nd Edition,  third printing.
     OAQPS, Research Triangle Park, N.C. 27711,  February 1976.

3.   Fegraus, C.E., Domke, C.J.,  and  J.  Marzen.   Contribution  of  the
     Vehicle Population to Atmospheric Pollution.   SAE No.  730530.
     Society of Automotive Engineers,  Automobile Engineering Meeting,
     Detroit, Michigan, May 1973.

4.   Trayser, D.A.   A Study of the Influence of Fuel  Atomization,
     Vaporation, and Mixing Processes  on Pollutant  Emissions from Motor
     Vehicle Power Plants.  Battelle  Memorial  Institute. Columbus,
     Ohio.   April  1969. pp 16.

5.   Federal Register, Vol 37, No. 221,  Nov. 15, 1972, and   Vol.  39,  No. 101,
     May 23, 1974.

6.   SAE Recommended Practice, Measurement  of Fuel  Evaporative Emissions
     from Gasoline Powered Passenger  Cars and Light Trucks  Using  the
     Enclosure Technique - SAE J171a,  SAE. Handbook, SAE, Inc.  Two
     Pennsylvania  Plaza, New York, N.Y., 10001

                                   1066
-------
7.    Williams, M.E.,  White,  J.T.,  Platte,  L.A.  and  C.J. Domke.  Automobile
     Exhaust Emission Surveillance.   Analysis  FY-72 Program.  EPA,
     Ann Arbor, Michigan.Publication  EPA-460/2-74-001.  February, 1974.

8.    Heywood, J.B.  and M.K.  Martin.   Aggregate Emissions  from the Automotive
     Population^SAE No. 740536,  Society of Automotive  Engineers.   Combined
     Commercial Vehicle and  Fuels  and Lubricants  Meeting,  Chicago,  Illinois
     June 1974.

9.    Voelz, F.L., Onyon, E.W., Oust,  R.M.  Rusnack,  N.C.,  Segal, J.S.  and
     B.G. Gower.  Survey of  Nationwide  Automotive  Exhaust Emissions  and
     PCV Systems Conditions  - Summer  1970.  SAE 710834.   Society of
     Automotive Engineers, Combined Truck  Powerplant, and  Fuels and
     Lubricants Meeting, St. Louis, Mo., October  1971.

10.  Brehob, W.M.,  Control of Mobile  Sources,  Purdue University.  8th
     Conference on  Air Pollution Control.   Lafayette,  Ind.,  October 1969.

11.  Black. F.M., and R.L. Bradow, Patterns of Hydrocarbon Emissions  from
     1975 Production  Cars, SAE 750681, Society of Automotive Engineers,
     National Fuels & Lubricants Meeting,  Houston,  Texas,  June, 1975.

12.  Dimitriades, B., and D.E. Seizinger,  Environ.  Sci. Techno!., 5:223-
     229, 1973.

13.  Wade, D.T., Factor Influencing Vehicle Evaporative Emissions,  SAE
     670126, Society  of Automotive Engineers,  Automotive  Engineering
     Congress, Detroit, Michigan,  January  1967.

14.  McEwen, D.J.,  The Analysis of Gasoline Vapors  from Automotive  Fuel
     Tanks.  153rd  National  Meeting of the American Chemical Society,
     Miami Beach, Florida, April  1967.

15.  Bennett, P.A., Jackson, M.W., Murphy, C.K.,  and R.A.  Randall.
     Reduction of Air Pollution by Control of  Emissions from Automobile
     Crankcases.  Vehicle Emissions.   Part I.   SAE  Tech.  Prog. Sec.
     6:224-253, 1964.

16.  Patison, J.N.  and E.R.  Stephens.  The Composition of  Automotive
     Blowby Gases,   pp. 37-50.  Proceeding of  the 3rd Technical Meeting,
     West Coast Section, Air Pollution Control  Association.  Monterey,
     Call for., 1963.

17.  Rose, A.H., and  R.C.  Stahman.  The Role of Engine  Blowby in Air
     Pollution.  J. Air Pollut. Control  Associ.   11:114-117, 1961.

18.  Domke, C.J., Lindley, D.J., and  C.N.  Sechrist.  How  to  Study Effect
     of Blowby Gas.  Hydrocarbon Processing 45(9):303-306, 1966.

19.  White. J.T. Automobile  Exhaust Emission Surveillance.   FY-74 program:
     Contract No. 68-03-2183, 68-03-2184,  68-03-2185, 68-03-2229.
     FY-75 program:  Contract No.  68-03-2378,  68-03-2379,  Environmental
     Protection Agency, Office of  Mobile Source  Pollution  Control,
     Ann Arbor, Michigan 48105.
                                   1067
-------
                                                                             22-2
                          CONTROL OF OXIDANTS IN  SYDNEY

                           D.  Iverach and M.G. Mowle*
ABSTRACT

     The. YiatuAe.,  e.xte.nt and Ae.veAity o^ photochemical pollution hi Sydney aAe.
             Recently ob*eAve.d health e^ec-ts on school chiX.dn.e.n have,  empha-
-6-czed the. AeAiouAne.&A o^ the. problem.
     Vn.oQn.eAi> in. cnmbating the. problem i& n.e.viewed.   The. impoAtance.  o&  a. coop-
e.native. inve.Atigati.on known  oj> the. Sydne.y OxA.da.Yit Study i& &tAeJ>t>e.d,  The.
introduction o{, the. U.S.  1973  ve.hicle. e.n\ik&ion AtandaAdA JJ> atbo du>cuJ>&e.d.
Re.ce.nt te.AtA Ahow that thu>  AtandaAd MJJJ, not btung  about the. sie.du.ctLonA
ofu.gtnal^.y hope.d faofi.  Ton. e.x.amp£e., sieAtittA ofa ve.hicie. te.i>ti> tn Sydne.y  i>how
that e.x.hauj>t hydfioc.an.boYi £eveZ6 w^tt only be. n.e.du.ce.d appn.oxAjmate.ty 30%  com-
pan.e.d to  mon.e, than 60% pn.e.dtdte.d on the. baAtt, o&  U.S.  data,  A n.e.\)Lt>e.d
     tnve.nton.y L*> pn.ue.nte.d  taking the.be. n.e.4uItA  into account.
     \notheA dibtuAbing n-e.ce.nt finding AJ> that  poUute.d ain. ^n,om Sydne.y may be.
tnanAponte.d about 50 km to an  an.e.a whesie. con^ideAabte. population Qn.owth 
&on.e.caf>t to OCCUA.

     Fn.om  oJUL thu, it u> cie.an. that toweA e.x.hauAt  Atandan.d*> an.e. needed to
contAol the. pAobLem, but the. natuAe. o& the. vehicle. indut>tn.y in AuAtnaLLa wJUL
make. the. intAoduction ofa tow,  deteAioAation  ^ee emi^Aion te.chnology a chal-
lenging tat>k.


                                  INTRODUCTION

     With  three million people living in an  area of 1000 sq. km.,  Sydney is
Australia's largest city.  Located on the east  coast of the continent at lati-
tude 34°S,  the city lies across the east-facing mouth of the Parramatta River
basin.  Gently rising hills lie to the south, west, and north of the region.
Westward the land continues to rise and 90 km from the coast the hills are
about  1200  m high.

     In this region the use of 1.2 million motor vehicles combined with the
operations  of an extensive range of industries  result in the emission of more
than 400 tons/day of hydrocarbons (HC) and about 170 tons/day of nitrogen
*State Pollution  Control Commission,  Sydney, New South  Wales, Australia.
 Dr. Iverach  is  also presently  at the Sloan School  of Management, Massachu-
 setts Institute  of Technology,  Cambridge, Massachusetts.
                                     1069
-------
oxides (NOX).  It is not surprising therefore that photochemical oxidants form
in undesirable amounts whenever appropriate meteorological conditions prevail.

     The problem of photochemical pollution first came into prominence in
Sydney in November 1971 when widespread damage to petunia seedlings was identi-
fied as being caused by oxidants.  Full realization of the seriousness of the
problem came in March 1976 when thirteen school children were taken to the
hospital after suffering from chest pains and breathing difficulties while
exercizing. The symptoms have been attributed to oxidants; at the time of the
incident ozone (03) concentrations over 0.2' ppm were recorded in the vicinity
of the school.  Prior to this occurrence the main recorded effects of oxidants
had been vegetation damage, reduced visibility and accelerated rubber crack-
ing.   Few reports of eye irritation had been received.

     To put the problem in perspective it should be added that, the potential
for a Los Angeles scale problem does not exist in Sydney.  It is not likely
that hourly values as high as those measured in the Los Angeles basin (up to
0.7 ppm) could form in Sydney.  Perhaps more pertinent is the fact that mete-
orological conditions are such that Sydney should never experience the same
frequency of high levels as recorded in that city (150 days per year over 0.15
ppm).

     Typically in Sydney there are five to ten episodes per year, each lasting
one to five days when 03 concentrations in excess of 0.15 ppm (1 hr ave) are
measured.   Hourly values over 0.25 ppm and peaks over 0.4 ppm have been re-
corded infrequently.

     If nothing were to be done, 0.15 values would become 0.25 within a decade
and such values might occur on 20-30 days each year.  Although not compatible
to experiences in Los Angeles, such a situation would, on present day atti-
tudes, be considered a very serious one.  Indeed there is considerable concern
about the existing levels.  Fortunately the need for effective action has been
recognized.


                              PROGRESS TO DATE

     In 1971 when oxidant pollution was first recognized as a problem, very
few data existed on which to base a control strategy.  The extent of the
problem was not known; virtually nothing was known of the concentrations of
the precursors, reactive HC's and NOX; the important role of meteorology was
only suspected, and no emissions inventory was available.  Nor was there
provision in the Clean Air Act to regulate vehicle emissions.

     Progress in both data collection and analysis and in the legislative
areas  has  been substantial, but as is the case in the rest of the world, much
remains to be done.

     With  respect to data collection and analysis, perhaps the single most
important decision the commission has made was to enlist the aid of three ex-
ternal  organizations.   It is not common in Australia for a State Government
Department to contract outside organizations in the scientific area.  However,


                                    1070
-------
the study of oxidants in Sydney required a greater variety of measurements to
be made with more sophistication than had hitherto been attempted.  Contracts
have been entered into with specific groups within the University of Sydney,
Maquarie University and the Commonwealth Scientific and Industrial Research
Organization (CSIRO).

     At the same time, the Commission has increased its staff and developed
its facilities in order that it be still directly involved in the day to day
effort and not simply a contract manager or coordinator.  Approaching the
problem in this way allowed a much greater effort to be mounted more rapidly
than would have been possible by doing all the work in-house, but without
losing the essential information that derives from day to day involvement.

     The responsibilities and interests of the four groups are listed below.
The activities in parentheses are not financially supported by the Commission
but obviously contribute to the study.

     •   The State Pollution Control Commission - Coordination, development of
         control strategies, development of emission inventories, responsi-
         bility for fixed site monitoring and mobile monitoring, including use
         of ground, sea and air craft.

     •   The University of Sydney - Responsibility for ground based mobile
         monitoring along air mass trajectories, analyses of data and determi-
         nation of driving conditions in Sydney (development of p.c. smog
         models for Sydney and development of emission inventories).

     •   Maquarie University — Responsibility for meteorological measurements
         and analysis of wind data and air movements (development of models).

     •   CSIRO- Responsibility for detailed measurements of reactive HC's and
         development of a doubly compartmented smog chamber (measurement and
         analysis of aerosols).

Published results from the various groups are listed in the reference section.

     With respect to legislative developments, the most significant event was
the decision in November 1972 by the N.S.W.  Government to amend the Clean Air
Act to provide for the regulation of vehicle emissions.  Regulations were
gazetted in April, 1974 requiring, among other things, control of evaporative
emissions on new cars manufactured after 1 January 1975 and the U.S. 1973
exhaust emission standards to be met on new cars manufactured after 1 July
1976.   This action had been proposed in 1972.  Details of the development of
this and other vehicle related legislation and the anticipated effects may be
found in Murphy (1975) and Iverach and Bilger (1975).

     The important decision to adopt the U.S. 1973 standards and test pro-
cedures was based primarily on practical considerations.

     It has been pointed out that lack of data made it difficult to develop a
control strategy in the early years of the problem.   However certain qualita-
tive judgements could be (and were) made.   If the automobile were the major


                                   1071
-------
emitter of precursors in Los Angeles there were good reasons to suspect this
to be the case also in Sydney.  On a per capita basis vehicle ownership and
usage were somewhat comparable but there were fewer other major sources of
oxidant precursors (oil refineries and power stations) in the Sydney region.
Emissions of HC and NO  from vehicles had to be significant.  This was veri-
fied by emission inventory calculations which, when calculated on a two by two
mile grid square basis, also allowed non-methane hydrocarbons (NMHC) and NO
concentrations to be predicted. It was then possible to compare these predi£-
ted concentrations with the various oxidant NMHC NO  relationships that had
been developed.  These relationships had been used fn the U.S. and Japan to
determine whether HC or NO  should be controlled preferentially.,
                          X
     Iverach (1975) summarizes the situation.  Neither an HC only (U.S.
approach) nor an NO  only (Japanese approach at that time, 1973) strategy was
unmistakably indicated.  This suggested concentrating on HC control because
the technology for it was better developed both for mobile and stationary
sources.  Hence, adaptation of the U.S.  1973 standards was consistent with
both the strategy as it could be derived at the time and the philosophy of
using the "best practicable means" to control pollutants.

     Using a "best practicable means" approach to the problem rather than pro-
mulgating what may have been impossible, though justifiable, is typical of
pollution control in Australia and has been used since the first Clean Air
Acts were gazetted in the early 1960's.   Such an approach does not promise
that the lowest possible emissions will  be attained but neither does it
threaten to involve the Government in a "back-off" situation as has occurred
recently in the U.S. and Japan.  Perhaps the main disadvantage of the approach
is that it does not force, or at least encourage, the development of control
technology.


                   RECENT DEVELOPMENTS AND FUTURE OPTIONS

     A deal of data have been gathered since smog was first recognized as a
problem in Sydney and more will be obtained in the near future.  The most sig-
nificant developments are summarized below.

EMISSIONS  INVENTORY

     The emissions inventory for HC and NO  continues to become more accurate
as actual measurements of sources replace Emission estimates that were based
on the U.S. Environmental Protection Agency (EPA) emission factors.  Signifi-
cant differences have been found between the EPA factors and actual emissions
in Australia.   For example NO  emissions from stationary sources have been
found to be approximately one-half the EPA factors.  Preliminary measurements
of pre-1977 vehicle emissions have indicated average exhaust HC emission rates
of about one-half the pre-1968 U.S.  model  (uncontrolled) value.  NO  emission
rates are about 15-20% higher than pre-1968 vehicles in the U.S.   x

     Although  evaporative emissions have not been measured in Australia using
the SHED method the recent discovery in  the U.S.  of high evaporative emissions
from "controlled" vehicles will most likely carry over to Australia because


                                   1072
-------
identical control techniques have been used.  This has caused a further re-
vision in the estimates of the benefits of existing controls.

     Current estimates of emissions from vehicles in Sydney compared to U.S.
data and U.S. 73 standards are shown in Table 1.

     The present best estimates of Sydney's emissions are summarized in Table
2. Estimates for 1971 are shown in parentheses.   Most of these estimates
should be given 'B1  or 'C'  rating in terms of AP42 (EPA, 1975).

                     TABLE I.  VEHICLE  EXHAUST EMISSIONS

         Vehicle Fleet or Standard      Emissions      g/Km
                                       CO             HC        NOx

         Ave. of 81 Sydney Vehicles
tested in 1976
Uncontrolled U.S. (pre 68)
U.S. 1973 Standard
32
54
24.2
3.0
5.5
2.1
2.6
2.25
1.9

PRECURSOR MEASUREMENTS

     Much effort has gone into measuring early morning concentrations of
reactive HC and NO .  The detailed results may be found in Allen et al. (1976).
In general it has Been found that NMHC/NO  values are in reasonable agreement
with values predicted from the present emission inventory.  In addition, con-
centrations are rarely measured that are as high as the 'worst case1 predic-
tions made by Iverach et al. (1976).   High oxidant day precursor concentra-
tions are 0.02-0.15 ppm NO  and 0.2-1.5 ppm NMHC.  Perhaps most importantly,
precursor concentrations have been related to oxidant levels downwind on
several occasions, and fair agreement has been found with currently accepted
oxidant-precursor relationships.

AIR MASS TRAJECTORIES

     Hyde and Hawke (1976) describe the various kinds of meteorological condi-
tions that prevail on high oxidant days.  Their most important conclusion,
based on modelling air mass trajectories, is that on a high proportion of high
oxidant days photochemical oxidants may be transported some 50 km. to the
south west of Sydney — an area where considerable population growth is
anticipated.
                                    1073
-------
TABLE 2.   EMISSIONS OF NON-METHANE HYDROCARBONS AND
    NO  IN SYDNEY (TONNES/DAY), 1976 AND (1971)

Source
Light Duty
Vehicle-Exhaust
Vehicle-Evap.
Vehicle-C' case
Petrol Trucks,
all sources
Diesel Truck
Motor Cycles
Utility Engines
Air, Rail & Shipping
Total Mobile
Industrial Combustion
Incineration
Refineries
Storage, Transfer
of Gasoline
Solvent Evap.
Service Stns.
Dry Cleaning
Total Stationary
Grand Total
Hydrocarbons

110
63
10

18
1
14
16
3
235
8
9
53
21
67
14
10
173
418

(176)
(51)
(39)

(ID
(1)
(12)
(13)
(3)
(306)
(7)
(7)
(45)
(18)
(56)
(12)
(8)
(152)
(458)
N0x

103 (86)
-
-

8 (6)
13 (10)
-
-
4 (3)
128 (90)
33 (58)
1 (2)
7 (13)
_
-
-
-
41 (72)
170 (161)
                    1074
-------
                               FUTURE CONTROLS

     Future control will focus on vehicles, the refineries and associated
gasoline transfer operations that exist to serve the needs of motorists, and
on solvents.  This discussion presumes that relatively straight-forward steps
such as legislation of the Rule 66 type can be taken to reduce the importance
of solvent emissions.  Achieving significant  reductions in vehicle emissions,
however, is likely to be tougher.

     The total Australian market is relatively small -about 600,000 units/yr.
The N.S.W.  market is about 200,000 (170,000 cars and station wagons).  This
market is split between some twenty-five companies offering some 150 models.
The local manufacturers, General Motors-Hoi den, Ford (Aust), Chrysler Australia,
and Ley!and (small) have the largest share of the market but have lost ground
in recent years to Japanese imports (mainly Toyota, Datsun, Mazda and Honda).
The top ten companies achieve about 98% of sales in the light duty vehicle
market.

     Some automobiles are manufactured in Australia; some producers simply
import the assembled vehicle but may have to install seatbelts or, as happened
in NSW, evaporative emission controls, depending on local  regulations and the
ability of the overseas principal to comply with them.

     All companies rely heavily on their overseas principals for emission con-
trol system design.  Even local manufacturers have only limited design author-
ity. The importers and assemblers rely absolutely on their overseas companies
for virtually all emission control work.   And to reiterate:  often the
Australian (and N.S.W.) market accounts for only a small fraction of their
total sales. Consequently many of the companies are slow to adapt and react to
new standards. Few are actively pursuing new engine designs or modifications
in anticipation of future standards.   Indeed only the local manufacturers and
the local carburetor manufacturer (Bendix Corp.) have emission testing facil-
ities in Australia and these are located in their development laboratories.
Only Leyland of all the vehicle companies has an emission test facility in
N.S.W.

     Despite these difficulties this industry has to be persuaded, in some
manner, to adopt what preferably should be a new, low emission, technology.
It must be a new technology because,  even assuming evaporative emissions can
be properly controlled, the need for exhaust emissions substantially lower
than the current standards can be demonstrated.  And it is most important that
essentially non-deteriorating technology be used.  The use of technologies or
techniques that suffer the currently observed rates of deterioration (EPA,
1975 and Stork, 1976) will not allow even modest improvements to be made in
air quality.

     The only technologies that offer real hope in this regard are the diesel
and the stratified charge engine.  Both these technologies are new to the
mass-produced light duty vehicle scene.   However, influencing the adoption of
these and possibly other, yet to emerge,  engines is one of the great chal-
lenges facing governments in Australia today.
                                    1075
-------
                                  REFERENCES

 1.   Allen,  Bilger,  Post  and  Haynes  (1976), present proceedings.

 2.   EPA (1975),  A Compilation  of  Air  Pollutant Emission Factors, AP42 4th
     edition,  U.  S.  Environmental  Protection Agency.

 3.   Hyde and  Hawke  (1976), present  proceedings.

 4.   Iverach (1975),  Planning for  Oxidant Control in Sydney, Jnt'l Clean Air
     Conference,  Rotorua, N.  Z., Feb.  1975.

5.    Iverach and  Bilger (1975), The  Status of Air Pollution and  Its Control in
     Australia,  in Advances in  Environmental Science and Technology, Pitts &
     Metcalf,  Vol  5,  1975.

 6.   Iverach,  Mongan,  Neilson and  Formby  (1976), Vehicle Related Air Pollution
     in Sydney,  J.A.P.C.A., Jan.,  1976.

 7.   Murphy  (1975),  Motor Vehicle  Legislation in NSW,  Int'l Clean Air Conference,
     Rotorua N.Z., Feb.,  1975.

 8.   Stork (1976), Preliminary  Data  on Emission Performance of 1975 model year
     cars, internal  EPA memo, Feb. 25, 1976.
                                     1076
-------
                                                                            22-3
                       OXIDANT  AND  PRECURSOR TRENDS  IN
                     THE METROPOLITAN  LOS  ANGELES  REGION

                  J. Trijonis,  T. Peng,  G.  McRae,  L.  Lees*


ABSTRACT
          papeA. de^cribeA fte.ce.nt  historical. &ie.ndt>  in oxA.da.nt and precursors
in the. Lot, \ngeZu fis.gi.on.   Control strategies  and  basin-wide emission trends
^or nitrogen oxides and rcacti.vc  hydrocarbons are documented year by year ^rom
1965 to 1974.  Trends in the geographic diAtsu.buti.on  o^ emissions are illus-
trated by computing nit percentage  emission changes over the. de.ca.de. ^or indi-
vidual, counties.  The. changes in  emissions  are  compared to change.* in ambie.nt
psie.cuuosi concentrations and oxx.dant concentration^.   It i& ^ound that many o{,
the. changes in monLtotutd aifi quaJLity can be. e.x.p&iine.d by tn.e.ndt>  in total
e.mu>£ion and in the. bpatiaJt distribution o& zmiAAion*.


                                INTRODUCTION

     The control of photochemical oxidant is a  complex,  dynamic  process involv-
ing major technical uncertainties.   An  effective  way  to deal with the dynamic
and uncertain aspects of the problem is  to  adopt  a  "feedback" approach in
control strategy planning.   Progress  (or lack of  progress) in improving ambient
air quality should be continually reviewed.  Control  strategies  should be
periodically reexamined and, if necessary,  reformulated.   Control plans should
be flexible so that changes  can be  made  as  new  information is developed.

     Trend analysis constitutes a fundamental part  of the feedback approach to
air pollution control.  Air  quality and  emission  trends  should be major factors
in determining whether or not control strategies  need reformulation.   Trend
studies also can help to reduce technical uncertainties.   For instance, the
analysis of ambient precursor trends  can indicate the adequacies, or deficien-
cies, in emission inventories and emission  reduction  estimates.   Comparing
historical oxidant changes to precursor  emission  changes  can provide insight
into the relationship between emissions  and air quality.   This improved know-
ledge can be used in evaluating future  control  strategies.

     This paper documents trends  for oxidant and  the  precursors  (nitrogen
oxides (NOX) and reactive hydrocarbons  (HC)) in the Los  Angeles  region from
1965 to 1974. Changes in basinwide  precursor emissions and in the geographic
*J.  Trijonis, Technology Service  Corporation,  Santa  Monica,  California; T.  Peng,
 Jet Propulsion Laboratory, Pasadena,  California:  G.  McRae  and L.  Lees, Environ-
 mental Quality Laboratory, California  Institute  of  Technology,  Pasadena, Calif.


                                      1077
-------
distribution of emissions are compared to changes in ambient precursor concen-
trations and ambient oxidant levels.   Most of the results presented here are
based on a recent report of the Caltech Environmental Quality Laboratory (1).
The reader is referred to that report for detailed discussions of the methodol-
ogy and results.
                          PRECURSOR EMISSION TRENDS

     This section describes trends in control strategies and emissions over
the last decade for the two oxidant precursors, NOX and the reactive HC (RHC).
The trend estimates are based on much of the latest information available for
emission sources in the Los Angeles region.   Included are new reactivity
factors for HC emissions (2, 3), recent measurements of auto exhaust emission
and deterioration factors (4, 5, 6, 7, 8), test data on the unexpectedly large
contributions made by evaporative emissions  from light-duty vehicles (9, 10),
the latest stationary source NOX inventory (11), and recent data on traffic
(VMT) growth and fuel usage patterns (12, 13, 14, 15, 16).   A detailed descrip-
tion of the methodology for determining the  emission trends can be found in
Reference 1).

     Basin-wide emissions of each pollutant  are documented year by year from
1965 to 1974 in order to characterize overall emission trends and the changing
role of various source categories.  To illustrate the spatial distribution of
emission trends, net percentage changes in emissions over the decade are
determined for each of the six counties within (or partially within) the
Metropolitan Los Angeles Air Quality Control  Region (AQCR).  Figure 1 shows
the area covered by the Los Angeles air basin and the geography of the six
county sub-areas.
 TRENDS  IN  TOTAL  EMISSIONS

      The purpose of  documenting  basin-wide  trends  in  total  emissions  is  to
 help  explain  trends  in  average,  basin-wide  air  quality.   Total  emission  trends
 are also useful  for  illustrating changes  in  emission  levels  for various  source
 categories.   These changes  result from emission  control  actions in  competition
 with  the growth  of source  activity levels.   Since  different source  categories
 are associated with  different  rates of growth  and  degrees of control, the
 roles of various source categories have undergone  continual  alteration.

      Figure 2 presents  estimates of basin-wide  NOx emission  trends  over  the
 past  decade.  The basin-wide total  is  represented  by  the top line,  while the
 distances  between the other lines  illustrate the contributions  of various
 source  categories. Figure  2 shows  that total NOx emissions  in  the Metropolitan
 Los Angeles AQCR increased  by  about 36% from 1965  to  1974.   The major contri-
 butor to this increase  was  the rise in NOX  emissions  from light-duty  vehicles;
 total NOX  emissions  from light-duty vehicles increased by 75%  over  the decade.
 This  was partially due  to  a 41%  increase  in basin-wide VMT  and  partially due
 to the  large  increase in NOx emissions from  1966 — 1969  model year  vehicles.
 The control techniques  used to reduce  HC  and carbon monoxide (CO) in  1966 —
 1969  vehicles had the side  effects  of  raising MOX  emissions.   Mew car emission


                                     1078
-------
                                                     O
                                                     O)

                                                     'QJ
                                                     01
                                                     £Z
                                                     l/l
                                                     o
                                                      o
                                                      CL
                                                      O
                                                      
-------
 standards for NOX took effect in 1971, but as yet these standards have not had
 a great impact on basin-wide NOX emissions.
      YEARI Y  1000

      
-------
 an  even  greater rate  after 1975  because of the stringent new car NOX emission
 standards  which started in 1975.

      Figure  3 summarizes basin-wide trends in reactive HC emissions  from 1965
 to  1974.   The top  line represents  total RHC emissions, and the distances
 between  the  lines  represent the  contributions from various source categories.
 Here, the  definition  of reactive hydrocarbon emissions is based on a two-group
 reactivity scheme  proposed by Dimitriades (2) of the Environmental Protection
 Agency and on reactivity factors calculated in a recent TRW report (3).
Figure 3.   Total  reactive hydrocarbon emission trends in the Los Angeles region.

      The  net  change  in  basin-wide  reactive  hydrocarbon  emissions  over  the
 decade was  a  decrease of  18%.   Most  of  this  reduction was  due  to  controls  on
 light-duty  vehicles.  Light-duty vehicle  crankcase  emissions were  reduced  in
 the early and middle 1960's.   Slight reductions  in  evaporative  emissions were
 obtained  in the  1970's  from  the new  car evaporative controls.   Exhaust emission
 standards for new automobiles  resulted  in significant control  of  exhaust
 emissions.  The  net  change in  RHC  emissions  for  the average vehicle  over the
 decade was a  52% decrease.   In  the presence  of a 41% increase  in  VMT,  the  net
 reduction in  total RHC  emissions from light-duty vehicles  was  only  32%.
                                      1081
-------
      Gasoline-powered,  heavy-duty  vehicles  experienced  a  23%  reduction  of per-
 mile  RHC emissions  over nine years,  most  of the  reduction  coming  from crank-
 case  emission  control.   With the 41% growth in VMT,  the net change was  a 7%
 increase over  the period from  heavy-duty  vehicles.   Emissions  from other
 mobile  sources have increased  by 56%.

      The organic chemicals  category, the  largest contributor  of RHC  in  the
 stationary  source category, has been reduced by  approximately  33% from  1966  to
 1974  due to APCD Rule 66 in the late 1960's and  to substitution of water-based
 coatings for organic solvent coatings  in  the early 1970's.  Emissions from
 gasoline marketing,  which is a major source of organic  fuel emissions,  have
 increased by 26%, although  filling of  underground gasoline station tanks has
 been  partially controlled by local APCD rules.   An important  part of the
 organic fuels  and combustion category  is  the "geogenic  source" that  was re-
 cently  pointed out  by Mayrsohn and Crabtree (17).  It is  assumed  that this
 source  remained constant over  the  ten-year  period.


 GEOGRAPHICAL DISTRIBUTION OF EMISSION  TRENDS

      The spatial distribution  of emissions, as well  as  the total  amount of
 emissions,  is  important to  ambient air quality.   Accordingly,  trends in the
 spatial  distribution of emissions  should  be considered  in  analyzing  air qual-
 ity trends.  This section illustrates  large-scale trends  in the geographic
 distribution of emissions within the Los  Angeles  air basin by  documenting net
 emission changes over the decade on  a  county-by-county  basis.

      There  is  reason to expect that  the spatial  distribution  of emission
 trends  has  been very nonuniform in the Los  Angeles region.  Although similar
 emission control policies apply to each county,  growth  rates  vary considerably
 from  county to county.   Figure 4 illustrates the  spatial  pattern  of  VMT and
 population  growth from  1965 to 1974.   Growth rates for  each individual  county
 are presented  within the boundaries  of those counties,  while  basin-wide fig-
 ures  are given in the lower left-hand  corner of  the  figure.

      As  indicated in Figure 4, the average  VMT and population  growth rates for
 the entire  basin are 3.9% per  year and 1.1% per  year, respectively.  Los
 Angeles  County has  been growing at a rate considerably  slower  than the  other
 five  counties  within the Basin.  The fastest growing county,  Orange  County,
 has a VMT growth (7.5 percent  per  year) triple that  of  Los Angeles County  (2.8
 percent per year) and a population growth rate (4.3% per  year) more  than ten
 times that  of  Los Angeles county (0.3% per  year).  These  differential growth
rates, compounded year by year over  the decade, led  to  a spreading out  of
emissions away  from  Los  Angeles County to the outlying  counties..   Los Angeles
County accounted for about 75% of basin-wide emissions  of NOX and RHC in 1965,
but only  for about 65% in 1974.

      Figure 5  illustrates the  spatial  distribution of NOX emission changes
over  the decade.  The net percentage change  in NOX emissions over the decade
is presented for each of  the six counties  and for the entire basin.   Each
county experienced  an increase in NOX  emissions over the decade,  with the
basin-wide total increasing by 36%.  Following the spatial trends in popula-

                                       1082
-------
                                                      01
                                                      i—I
                                                       I
                                                      LD
                                                      i-D
                                                      (13
                                                      S-
                                                      -I-J
                                                      s
                                                      O

                                                      en
                                                      Zi
                                                      Q.
                                                      O
                                                      O.

                                                      -a
                                                      a
                                                      (T3
                                                      rd
                                                      O)
                                                      O)
                                                      CD
                                                      OJ
                                                      >
                                                     
-------
                                                CT>
                                                LO
                                                IT)
                                                cr>
                                                 o
                                                •I—

                                                 I/)

                                                 £
                                                 QJ

                                                  >
                                                O
                                                ^

                                                 c


                                                 l/l


                                                 O)
                                                 cn
1084
-------
 tion  and VMT  growth,  Los Angeles  County  NOX  emissions  grew by  the smallest
 amount, 25%.  NOX emissions  in  Orange  County rose  drastically,  an estimated
 89% increase  over the decade.

      Figure 6 presents  the county-by-county  distribution  of  RHC emission
 changes. Over the decade, basin-wide RHC emissions decreased 18%.   Following
 the order  of  population and  VMT growth rates,  Los  Angeles County showed the
 largest decrease, 24%, while  the  outlying  counties experienced  lesser  decreases,
 In fact, Orange  County evidently  underwent an  actual increase  in RHC emissions
 of six percent over the decade.


                         AMBIENT  AIR QUALITY TRENDS

      Having documented emission trends in  the  Los  Angeles region over  the past
 decade, the next logical step is  to compare  the emission  trends to  air quality
 trends.  This section presents  air aua1ity__trends  over the decade for  both
 oxidant and its precursors and  analyzes  these  trends in terms of the changes
 in total emissions and the spatial distribution of emissions.   Comparing
 ambient precursor trends to emission trends  provides a  test  of the adequacy of
 simple proportional (rollback) models for relating emissions and precursor
 concentrations.  This juxtaposition also serves as a partial check on  both the
 emission and  air quality data bases.   Comparing oxidant trends to precursor
 trends yields insight as to the oxidant/precursor relationship in various
 parts of the  Los Angeles region.

     The data base for determining air quality trends  in the Los Angeles
 region consisted of computer tapes of hourly average pollutant measurements
 obtained from the California Air  Resources Board (ARB).  These tapes contained
 data from  all stations in the Los Angeles region which  reported to the ARB
 during the years 1963 to 1974.  A search was conducted  among personnel of the
 ARB and the County APCD's to determine whether monitoring changes occurred
which could affect air quality  trends.    Data were included from each site only
 if they were  taken with the same  measurement method each year.   It was deter-
 mined that the locations of monitors in  a few  cities had been altered  during
 the period of interest; data from these  locations were  eliminated from the
 study.  Some  of the monitoring  sites that were included in the  trend analysis
 are shown  in  Figure 1.

     In order to compare air quality changes to emission changes over  the
 decade, net changes in ambient  concentrations were determined from  1965 to
 1974.   For each pollutant (NOX, nonmethane hydrocarbons (NMHC), and oxidant),
 a least-squares trend line was  fit to the yearly averages of daily maximum
 one-hour concentrations.  The percentage change in this trend line from 1965
 to 1974 was used as the measure of overall air quality  change for each site.
 Yearly averages of daily maxima were eliminated from the  analysis if the
 averages were based on a very limited number of sampling days.  The analysis
was performed only for stations with at  least  eight years of adequate  data.


 AMBIENT PRECURSOR TRENDS

     Figure 7 compares changes  in NOX air quality to changes in NOX emissions.

                                     1085
-------
                                                CTl
                                                 I
                                                un
                                                en
                                                 co
                                                 to
                                                 0)

                                                O
                                                 1/1
                                                T3
                                                 O)
                                                
-------
                                             I
                                            in
                                             03
                                             13
                                             CT

                                             S-


                                             fO
                                             o

                                             co
                                             O)
                                             c:
                                             O)
                                             en
1087
-------
 The  circled  numbers  in  Figure 7 represent the net percentage emission changes
 for  each  county  over the  decade (also  given  in  Figure  5).   For  instance,  Los
 Angeles County experienced  a 25% increase in NOX emissions  over the decade;
 Orange County experienced an 89% increase; Riverside underwent  a 69% increase,
 etc.  The net percentage  changes in ambient  NOX concentrations  over the decade
 are  also  plotted in  Figure  7 for eleven monitoring  sites.   For  instance,
 ambient NOX  levels increased by 8% at  downtown  Los  Angeles, 76% at Anaheim,
 47%  at San Bernandino,  etc.  Basin-wide averages in emission trends and ambient
 concentration trends  are  presented in  the lower left-hand corner of Figure 7.

      The  agreement between  calculated  NOX emissions changes and measured  NOX
 air  quality  changes  is  strikingly good.  Total  basin-wide NOX emissions in-
 creased by 36% over  the decade, while  average NOX concentrations at the eleven
 monitoring sites increased  by 35%.  Even the spatial distribution of emissions
 and  air quality  trends  agree on a county-by-county  basis.   NOX  emissions  grew
 by 89% in Orange County,  while the two Orange County stations showed an
 ambient NOX increase of 93%.  NOX emissions  increased by 25% in Los Angeles
 County, while the eight Los  Angeles  County stations  experienced a 19% increase
 in measured NOX  concentrations.

     Figure 8 compares  changes in NMHC concentrations to changes in RHC
 emissions.  The  circled numbers  again represent net  percentage emission changes,
 for each county  and for the  entire basin.   Only four monitoring sites  provide
 data sufficient  for determining  ambient HC trends.   The net percentage  changes
 in NMHC* levels  over the decade  are  plotted  at the  locations of these  four
 sites.

     It is not possible to draw  definitive conclusions  concerning ambient NMHC
 trends because of the sparsity of sites providing data  on changes in HC concen-
 trations.   However, it  is encouraging that the average  change in ambient NMHC
 concentrations among the four sites,  a 15% decrease, is close to the basin-wide
 change in RHC emissions, an  18%  decrease.   Also, the spatial pattern of concen-
 tration changes  seems to agree qualitatively with the spatial pattern  of
 emission changes (at least in the sense that Los Angeles, San Bernandino,  and
 Orange counties  fall  in the  same order for concentration trends and emission
 trends).  The only puzzling aspect of  the ambient trend  data for NMHC is the
wide disparity between trends at the  two sites in Los  Angeles County,  Downtown
 Los Angeles and Azusa.

 TRENDS IN OXIDANT AIR QUALITY

      Figure 9 illustrates percentage changes in oxidant concentrations over
 the  past  decade  and  compares these to  changes in RHC emissions.  The circled
 numbers represent RHC emission trends  (county by county and basin-wide) while
 the  other numbers represent trends in  ambient oxidant  levels at thirteen
 monitoring sites. Basin-wide, the agreement between RHC emission trends and
 oxidant trends is quite good; total  RHC emissions decreased by  18% over the
 decade, while the average improvement  in oxidant levels at thirteen stations
 was  19%.   Also,  the  spatial  distribution of air quality changes agrees quali-
*Nonmethane hydrocarbon trends are computed from total hydrocarbon trends
 using empirical formulas relating THC to NMHC (18).

                                      1088
-------
   BASIN'.HDE KHC EMISSION CHANGE:  (-18%
                            V	x

   AVERAGE NMKC CONCENTRATION CHANGE (4 STATIONS):   -15%
         Figure  8.   Trends  in  RHC  emissions and air quality, 1965-1974.

tatively with the spatial  distribution  of  emission changes, at least in  the
sense that Los  Angeles  County  experienced  the  greatest  improvement  in both  RHC
emissions and oxidant concentrations.

     The spatial pattern of oxidant  air quality  trends  can  be  even  better ex-
plained by considering  what has been  termed  the  "dual role" of NOX  emissions
in oxidant formation.   From the present understanding of photochemistry,  it
can be argued that,  at  the existing  ratio  of RHC to NOX, increases  in NOX
emissions would tend to decrease  oxidant concentrations  in  the source-
intensive areas of  the  Basin  (the  central  and  western-coastal  areas).  In
these areas, the highest oxidant  concentrations  occur early in the  day  (around
noon) before the pollution is  transportad  inland by the  daily  sea breeze.   The
most effective  strategy for reducing  oxidant near the source areas  early  in
the day is to inhibit the  reaction by reducing the RHC/NOX  ratio.   This  can  be
done by decreasing  RHC  emissions  and/or increasing NOX  emissions.   The  trans-
ported oxidant  problem  in  the  eastern-inland areas should depend more 9n  the
absolute levels of  RHC  and NOX than on  the ratio.   The  key  to  controlling the
transport problem is less understood, but  indications are that reductions are
needed in overall levels of RHC and  (possibly) NO .

     These observations help to explain  why  the  coastal  areas  of Los Angeles
County show the greatest improvement  in  oxidant  concentrations (a 30 to  40%
decrease from 1965  to 1974).   These source-intensive areas  have  experienced a
24% decrease in RHC emissions, and they  have been  aided  by  the NOX  emission
increase.  The apparent anomaly at Anaheim,  a  31% decrease  in  oxidant within a
county with a 6% increase in RHC emissions,  can  be partially accounted  for  by
the large NOX emission  increase (and  resulting RHC/NOX  ratio decrease)  at that
source-intensive location.   The lesser  rates of  improvement in the  eastern-
                                      1089
-------
                                                  r--
                                                  en
                                                  LO
                                                  l£>
                                                  CTi
                                                  ro
                                                  13
                                                  CT

                                                  S_
                                                  •i—
                                                  ro
                                                  ro
                                                  -o
                                                  •r-
                                                  X
                                                  O

                                                  T3
                                                  c:
                                                  ra

                                                  00
                                                  c
                                                  o

                                                  00
                                                  oo

                                                  E
                                                  CD
                                                  n:
                                                  o:
                                                  0-)
                                                 -a
                                                  c:
                                                  ai
                                                  S-.
1090
-------
inland portions of the basin may be explained by noting that basin-wide RHC
emissions have been reduced only slightly and that total NOX emissions have
increased.   In particular, Riverside may have experienced a 23% increase in
oxidant concentrations because it is downwind of Orange County, which had a 6%
increase in RHC emissions and a large increase (89%) in NOX emissions.


                                 CONCLUSIONS

     Basin-wide emissions of NCX have increased by 36% in the Los Angeles
region from 1965 to 1974, while basin-wide RHC emissions have decreased by
18%.  Controls on crankcase and exhaust emissions from light-duty vehicles
were largely responsible for the total RHC emission decrease.  VMT growth and
the increase in NOX exhaust emissions from 1966 to 1969 model year vehicles
were the major contributors to the total NOX increase.

     Projections of emissions made in the early 1970's  (19) have turned out to
be overly optimistic when compared to emission trends which we now determine
in retrospect.  The projections made in the early 1970's indicated a  20%
increase in NOX emissions and a 40% decrease in RHC emissions from 1965 to
1974.  The error in the earlier projections is caused partly by the fact that
test data for automobile3 do not show as large emission reductions as would
have been expected from the time-table of tightening emission standards.
Evaporative emissions, for example, are much higher than called for by federal
and state standards.  For RHC, the new reactivity scale we are using  now leads
to a more pessimistic picture because it gives greater weight to source cate-
gories that have been reduced little in the last ten years.  For NOX, the
shortage of natural gas for industrial interrupt]'bles and power plants and the
resultant unforeseen switch to fuel oil have produced higher emissions than
originally projected for stationary sources.  The occurrences of these unfore-
seen circumstances should serve as a warning in making  future emission projec-
tions.  A sensitivity analysis of air quality to the possible range of emissions
should be included as part of future studies.

     Emission trends reflect a competition between reductions from emission
control and increases from source growth.  Since the geographical distribution
of source growth, in particular traffic growth, has been very nonuniform over
the Los Angeles region, the geographical distribution of emission trends is
also very nonuniform.  Among the counties in the basin, Los Angeles County has
grown the slowest, while Orange County has expanded at  the greatest rate. This
nonuniform spatial distribution of emission trends is important in inter-
preting air quality trends.  What appear to be contradictory air quality
trends at different locations in the basin can often be explained by  the
geographical distribution of emission trends.  In analyzing  future strategies,
the geographical distribution of emission changes should be  accounted for.

     The basin-wide trend in ambient NOX concentrations  (about a 35%  increase)
agrees well with the basin-wide trend in NOX emissions  over  the past  decade  (a
36% increase).  The geographical distribution of NOX emission trends  and NOX
air quality trends on a county-by-county basis also agree.  Ambient data on HC
trends are very sparse.  However, the average trend in NMHC concentrations at
                                     1091
-------
four locations (a 15% decrease)  is  consistent  with  the  basin-wide change  in
RHC emissions (an 18% decrease).

     The basin-wide improvement  in  oxidant  air quality  (19% over the decade)
agrees well with the basin-wide  reductions  in  RHC emissions.  The geographical
trends in oxidant agree,  qualitatively,  with  the geographical trends in  RHC
emissions, especially for Los  Angeles  County.   If we  account  for the dual  role
of NOX we obtain an even  better  clarification  of the  geographical distribution
of oxidant trends.   The increase in NOX  emissions helped  to decrease the  RHC/
NOX ratio; this resulted  in substantial  oxidant reductions  in the source-
intensive, central-coastal  areas.   Little  or  no improvement in  transport
related oxidant occurred  in the  downwind,  eastern-inland  areas  because overall
RHC emissions were reduced  only  slightly and  (possibly) because NOX emissions
increased.
                                  REFERENCES

 1.   J.  Trijonis, T.  Peng,  G.  McRae,  and  L.  Lees,  "Emissions  and Air Quality
     Trends in the South Coast Air Basin,"  EQL Memorandum  No.  16,  Caltech
     Environmental Quality  Laboratory,  Pasadena, California,  January 1976.

 2.   B.  Dimitriades,  "The Concept of  Reactivity and  Its  Possible Applications
     in  Control," Proceedings  of the  Solvent Reactivity  Conference, EPA
     650/3-74-010, November 1974.

 3.   TRW Environmental  Services, "Impact  of Reactivity Criteria on Organic
     Emission Control  Strategies in Los Angeles,"  EPA Contract No,
     68-02-1735, Redondo Beach, California,  July 1975.


 4.   Automotive Environmental  Systems,  Inc.., "A Study of Emissions from
     Light-Duty Vehicles in Six Cities,"  EPA Contract No.  68-04-0042,
     Westminster, California,  June 1972.

 5.   Automotive Environmental  Systems,  Inc., "A Study of Emissions from
     1967-1974 Light-Duty Vehicles in Los Angeles  and St.  Louis,"  EPA
     Contract No. 68-03-0390,  Westminster,  California, October 1974.

 6.   California Air Resources  Board,  "Exhaust Emissions  from  Privately
     Owned 1966-72 California  Automobiles -  A Statistical  Evaluation of
     Surveillance Data," El  Monte, California, March 1973.

 7.   Environmental Protection  Agency, "Compilation of Air  Pollutant
     Emission Factors,"  AP-42,  Second Edition, April 1973.

 8.   Private  communication,  D.  Bratton, California Air Resources Board;
     "Interim Estimates  of  Emissions  from Mobile Sources in California,"
     unpublished draft  by the  California  Air Resources Board  staff, dated
     April  1975.
                                     1092
-------
 9.  C.  D. Paulsell,  "Mobile  Source  Evaporative  Emissions," an EPA dis-
     cussion paper,  June 1974.

10.  California Air Resources Board,  "Public  Hearings to Consider Fuel
     Evaporative Emission Regulations for Light-Duty Vehicles," No. 75-7-6,
     April 1975.

11.  D.  R. Bartz, et al., "Control of Oxides  of  Nitrogen from Stationary
     Sources in the  South Coast Air  Basin," prepared for the California
     Air Resources Board (ARB 2-1471),  Report No.  5800-179, KVB, Tustin,
     California, September 1974.

12.  P.  K. Mazaika,  "Trends of Energy Use in  California and the South
     Coast Air Basin,"  Environmental  Quality  Laboratory, California
     Institute of Technology, Pasadena,  California, May 1975.

13.  Los Angeles Regional Transportation Study,  "LARTS Base Year Report
     1960," California  Division of Highways,  Los Angeles, California,
     December 1963.

14.  Los Angeles Regional Transportation Study,  "LARTS Base Year Report
     1967 Origin Destination  Survey," California Division of Highways,
     Los Angeles, California, December 1971.

15.  Private communication, G.  Bennett,  Transportation Studies Analysis
     Group, LARTS branch, California  Department  of Transportation, Los
     Angeles, California, September  1975.

16.  Private communication, H.  Linnard,  California Air Resources Board,
     October 1974; "Motor Vehicle Population  and Model Year Distribution,"
     a California Air Resources Board working paper, February 1974.

17.  H.  Mayrsohn and J.  Crabtree, "Source Reconciliation of Atmospheric
     Hydrocarbons,"  Atmospheric Studies  Section, California Air Resources
     Board, El  Monte, California, March  1975.

18.  J.  Paskind and  J.  Kinosian, "Hydrocarbon, Oxides of Nitrogen, and
     Oxidant Pollutant  Relationships  in  the Atmosphere over California
     Cities," Presented at the  67th Annual Meeting of the Air Pollution
     Control Association, Denver, Colorado, June 1974.

19.  L.  Lees, et al., "SMOG:   A Report to the People," Environmental
     Quality Laboratory, California  Institute of Technology, Pasadena,
     California, 1972.
                                    1093
-------
                                                                           22-4
             COMPARISON  OF  PAST  AND PROJECTED TRENDS IN OXIDANT
                   CONCENTRATIONS AND HYDROCARBON EMISSIONS

              R. M. Angus,  E.  W.  Finke,  and J.  H.  Wilson, Jr.*


ABSTRACT
         qaoLity projection*  ^or photochemical oxx.dan.t6 have, been made fioti the.
year* 1980,  1990 and  2000.  Thus, projection*  con*-ider alternative level* o&
hydrocarbon  control, ^or mobile and *tatA.onary  *ource*.   In order to make. the.
projection*, a**wmption*  concerning background concentration* and the. hydro-
carbon /oxA.dant relation*hip Mere ne.ce.MOAy.  The6e a**umption* Mere te*ted
u*i.ng pa*t trend* -in  air.  qaotity and emi^*on*.   The. projection* indicate, that
{,or the. mo*t optani*tic level* o& con&iot,  txignifa-icant xjnp-tovemen^ti in CUA.
qaatity are.  pAOj'e.cte.d &OA many afie.au> within the. United States.  Howe.veA,  even
(tiJMi theJ>e. Jj/npAoveinentis many  o& the.-i>e. 4ajne  aA.eai may not attain the. Motional.
Ambient AxA  QaaLity Standard  &or oitidantb  by the. ye.aA. 2000.


                                 INTRODUCTION

     The U.  S. Environmental  Protection  Agency (EPA) makes projections of
photochemical oxidant concentrations in  order  to compare the effectiveness of
proposed hydrocarbon  (HC) control strategies.   Recently an interagency task
force on motor vehicle goals  beyond 1980 analyzed the effect on air quality of
different pollutant control strategies through the year 2000 (Tuerk 1976).
One of the effects studied by the task force was the expected change in future
year 0  quality.
      /\

     Oxidant air quality  projections were made utilizing two models relating
HC emissions and 0  concentrations.   These  air quality projection models
assume growth rates and future emission  levels for both mobile and stationary
sources. Uncertainties abound in the projections because of these assumptions.
For photochemical oxidant projections, there are two additional uncertainties
caused by inaccuracies in determining:  (1) the proper relationship between HC
emissions and ambient 0   concentrations, and (2) the appropriate background
concentration for typical urban areas.  These  uncertainties are assessed by
studying past trends  in 0 air quality and  estimated HC emissions.  Finally,
using the two air quality models, projections  are made to the year 2000.

*R. Angus, University of  Missouri,  Columbia, Missouri.   This paper was prepared
 while he was on leave with the Office of Air  Quality Planning and Standards,
 U.S.  Environmental Protection Agency, Durham,  North CarolinaA_
  E.  Finke,  formerly  of U. S.   Environmental   Protection  Agency.
  J. Wilson,  Office  of Air Quality Planning  and  Standards, Environmental
  Protection  Agency,  Durham,  North Carolina.
                                      1095
-------
                  TECHNIQUES FOR PROJECTING CONCENTRATIONS


AIR QUALITY MODELS

     The effects of projected changes in HC emissions on 0  concentrations
have been analyzed by using two different means of relating the HC emissions
and the resulting 0  concentrations.
                   X

     The first HC/0  relationship is based on a nonlinear upper limit curve
(Schuck and Papettix'1973).  The upper limit, curve enclosed the highest 0
values observed for a given HC concentration.   The relationship is appro$i-
mated by Equation 1:

         HC = 2.049 0  = 4.792 0 2 + 5.036 0 3                             (1)
                     XX           X

For a given base year 0  concentration, Equation 1 is used to calculate a
corresponding HC concentration (6-99 am).   Assuming a proportional relation-
ship between HC emissions and the HC concentration, a new HC concentration is
projected for the year of interest.   Then, the relationship implied by Equa-
tion 1 is used to estimate the 0  concentration corresponding to the new HC
concentration for the projection year.

     The second HC/0  relationship assumes that the 0  concentration is a
linear function of HC emissions.   Therefore, a change in HC emissions is
assumed to result in a proportional  change in 0  concentrations.
                                               X

     Both of the above calculation procedures assume a zero background concen-
tration of HC and 0 .
ASSUMPTIONS USED IN MODELING

     Projections of 0  air quality were made for 37 areas in the United States.
The assumptions on wh¥ch the 0  projections are based are summarized in Table
1.                            x

     Four assumptions are made which immediately affect the quantity of esti-
mated HC emissions from mobile sources.  First, for light-duty vehicles one of
several possible emission standards for HC is assumed to take effect in 1978.
Second, the replacement rate for light-duty vehicles is assumed to result in a
complete turnover in the vehicle population every 13 years.   Next, the deteri-
oration rate of control devices for mobile sources is reduced from the present
rate by assuming that inspection and maintenance plans will  be in operation.
Credits for inspection and maintenance are calculated using the assumptions
listed in AP-42, Supplement 5 (EPA 1976).   Lastly, the exhaust emissions for
light-duty trucks and gasoline and diesel-powered heavy-duty vehicles are pro-
jected based on the emission factors of AP-42.   It should be noted that the
emission reduction assumed for these latter vehicles is generally less than
that assumed for light-duty vehicles.
                                     1096
-------










GO
o
1 — 1
I—
UJ
o

Q_
>_
| —
i — i
	 1

=}
o-

1 — 1
et;

1 —
5p
Q
I — i
x
o

^
o
CO


0-3:
E 0
0 S-
<_) CJ3




0 S-
|Q
O CO
1- >-
re co
Q£ to
re
S- CO
o
4J O
O +J
re
1 i f™
UL_ t_
0
c *t~
O (O
•r— {/)
•}
s-
o
CD
Q}
4-)
(O
0

CO
0
S—
0

o
•r—
V)
°i
UJ











o
o
CM


O
en
r^



o
CO
Cn
1


o
CT^
,_



^J-
to
en

































CO
CO
r—
0

C"
O)
>
$
3
Q
^j
01
•t—
— 1


^5
O

CO





o r-^ CM oo to
CM •— f— O O



i — i*-- CM en to
CM r— r— O O




r— o co r~^ to
^f *^~ CO CO CO




o o o o o





00 00 00 00 CO
«d- ^j- ^j- ^j- ^}-













+
•4- *^
o LO en ^t~ > —
• • • • •
CM r—











to
^
re
CO
>>

to i —
CO TJ CO "O
II i— J. -0 C
•r- re o o
to E T3 E >»
C ^v, C CO
o to re o .a
•r- E •»-> 4J
 re co ~o
to S- -o c
T- 05 c co re
e o T-
UJ CO T- r— 00
• (/> Q. r^
o o to o_ en
r-- i — i- re r—
en E
r— UJ



^5
O
•
co




^c
r—



^J-
•— |




en
co
*



o
"




CO

j= cn
•*-> T-
O _l











LO
«-^



00
"— ;




LO
to
*



0
"




^~
I —


































O)

•r—
r—
O
to
re
o
£>
3
O

re
CO












co
CM



r-.
CM




fN^
CO
*



o
"





o




































f—
0)
to
CO
•r-
Q
-»->
3
O
1
>,
>^
re
O)




^^
CM
*
co




LO
CM



O
LO




LO
CO
*



O
1— J_
3
0
U
0

O (/)
• CO
c
re
.c
o

CO
>
o
-0
re

r—
•r-
4^
C
3

^

Q.
0.
re

(/)
•o
i.
re
TJ
c
re
4-)
to

C
to O
CO i-
o to
S- to
3 •!-
O E
CO CO
i. c
« £
0 J-
•<- 3
4J CJ
re -K
CO




















































•
T3
s_
re
-a
c:
re
4^ "O
co s_
re
C XJ
o c
•i- re
to 4->
to CO
•r—
E c
UJ O
•r-
E t/)
•f- IO
S- i-
CO E
4-> UJ
C
~ t
4-> O
^— 1 \
CO 3
to 4->
co re
i- 4->
Q. CO
+ +

1097
-------
     Growth rates have a major effect on HC emissions projected to future
years. A growth rate of three percent compounded annually was selected for all
mobile sources.  This growth rate is representative of the historical growth
in vehicle miles traveled in metropolitan areas.  Stationary source growth
rates were estimated from economic indicators from four groupings of sources:
electric generation units, industrial activities, area sources and miscellaneous
point sources.  The stationary sources of HC are expected to grow at a 3.2
percent compounded rate.


PAST TRENDS

     To help assess some of the assumptions used in the protective air quality
models a study has been made of past trends in 0  air quality and estimated HC
emissions.  For this comparison, the most effective sets of data come from
California.  In general, trends in air quality have followed the trends in HC
emissions.  For example, in Los Angeles HC emissions have been decreasing
steadily since the mid-1960's.  It was estimated that the average emissions of
HC have decreased by 15 to 18 percent between 1964 and 1974 in Los Angeles;
during this same period peak 0  concentrations for six locations in Los Angeles
have decreased an average of 20 percent (Trijonis et al.  1976).

     Figure 1 presents a summary of the observed changes in 0  concentrations
in the Los Angeles Air Quality Control Program (AQCR).  In addition, 0
concentration projections are shown.  The upper trend line represents £he
monitoring site with the least improvement in 0  concentration from 1964 to
1974, and the lower line represents the site with the most improvement.  The
middle trend line is the median of six sites in the AQCR.

     Noting that the projection lines are in agreement with the general trend
in air quality, it appears that the assumptions concerning the relationships
between HC emissions and 0  concentrations are reasonable.
                          X

                             PROJECTION RESULTS

     For the areas that were modeled, the base year concentration was the
second highest one-hour annual average for 1972-1974.  Table 2 shows the
results of the upper-limit curve projections for an HC emission standard of
0.4 grams/mile.  These calculations showed an average improvement in 0  air
quality of 15 percent by 1980, 42 percent by 1990 and 50 percent by 2o5o.  As
mentioned previously, these projections depend on the growth assumptions and
the assumption concerning implementation of the emission standard in 1978.

     In spite of these improvements in Oy air quality many areas "are not pro-
jected to meet the ambient standard for u  by 2000.  Table 2 shows that 26 of
the 37 areas modeled will have concentrations in excess of the 0.08 ppm stand-
ard in 2000.  In addition, none of these cities is projected to have 0
concentrations which are at least twice the standard.

     It should also be noted that there is a significant difference in the
results obtained for the two models used.   A comparison of the effect for
                                     1098
-------
luepixo UL




        1099
-------
                  TABLE 2.  PROJECTED OXIDANT AIR QUALITY*

BASE YEAR
CONCENTRATION

004 Birmingham
005 Mob.-Pensacola
013 Clark-Mohave
015 Phoenix-Tucson
024 Los Angeles
028 Sacramento Valley
029 San Diego
030 San Francisco
031 San Joaquin
033 SE Desert
036 Denver
043 NY-NJ-Conn.
045 Philadelphia
047 Nat. Capitol
067 Chicago
070 St. Louis
079 Cincinnati
105 S Lou-SE Texas
115 Baltimore
119 Boston
131 Minn-St. Paul
153 El Paso
160 Gnsse-Fngr Lks.
173 Dayton
193 Portland
197 S.W. Penna.
214 Corpus -Christi
215 Dallas-Ft. Wrth
216 Houston-Glvstn
229 Puget Sound
991 Region 1
992 Region 2
993 Region 3
994 Region 4
995 Region 5
996 Region 6
997 Region 7
(PPM)
.29
.29
.22
.19
.54
.28
.32
.33
.27
.32
.33
.36
.20
.38
.30
.39
.19
.32
.23
.21
.10
.13
.12
.16
.15
.17
.15
.19
.38
.13
.19
.19
.19
.17
.17
.20
.11
Average Percent Decrease
No. Cities Above Std.

1980
CONCENTRATION
(PPM)
.24
.24
.18
.15
.46
.25
.25
.29
.24
.28
.26
.30
.15
.30
.29
.36
.15
.28
.22
.17
.08
.09
.09
.13
.13
.14
.15
.16
.36
.09
.16
.16
.19
.16
.14
.19
.09
15.
36
1990
CONCENTRATION
(PPM)
.17
.17
.12
.10
.34
.18
.16
.22
.18
.21
.16
.20
.09
.19
.23
.28
.10
.20
.17
.11
.05
.05
.05
.08
.08
.09
.12
.10
.29
.05
.11
.11
.15
.11
.09
.15
.06
42.
30
2000
CONCENTRATION
(PPM)
.15
.,15
.10
.08
.30
.15
.15
.18
.15
.17
.15
.18
.08
.18
.18
.23
.09
.17
.14
.10
.04
.05
.04
.07
.07
.08
.09
.09
.24
.05
.09
.09
.12
.09
.08
.12
.05
50.
26
*Light Duty Vehicle Emission Standard = 0.4 grams/mile

Projected Cone:  Second highest 1-hour concentration in ppm
Base Year:  Base Year concentration is second highest 1-hour annual
            average recorded from 1971 to 1974
                                   1100
-------
various assumptions concerning HC emissions and 0  background concentrations
on the long-range projections to the year 2000 isxpresented in Table 3.   At
the statutory emission standard of 0.4 grams/mile the empirical  upper-limit
relation projects an average air quality improvement of 50 percent by the year
2000.   The linear assumption projects an average improvement of 61 percent if
background 0  concentrations are neglected and a 46 percent improvement if a
0.05 ppm background is assumed.

     These differences represent the average improvement for 37 areas.   Indi-
vidual cities may show improvements that differ from these average changes
because of the different magnitudes of base-year concentrations.   For example,
a background concentration will  have a smaller effect on the projected improve-
ment in cities with high 0  concentrations than for cities with low concentra-
tions.  In addition, there will  be less difference between the results  of the
two projective models for cities with low concentrations than for cities with
high concentrations.  These effects are illustrated by comparing the number of
cities which are projected to violate the ambient air quality standard in
2000.  With an emission standard of 0.4 grams/mile the upper-limit curve pro-
jects  that 70 percent of the cities considered will still  be in violation of
the air quality standard in 2000.  The linear assumption with zero background
projects that this number will be reduced to 49 percent.  However, when the
background of 0.05 ppm is used,  the number of cities in violation of the
standard is projected to be 86 percent.


                                   SUMMARY

     The methods used to predict 0  concentrations indicate that even under
stringent control measures the 0  standard will not be met in many cities.  It
is predicted, however, that the 6  concentrations can be reduced by as  much as
46 to 61 percent under the most optimistic assumptions.


                                 REFERENCES

 1.  Schuck, E.  A.  and R.  A.  Papetti.   Examination  of the  Photochemical  Air
    Pollution Problem in  the Southern  California Area,  Region  IX,  U. S.
    Environmental  Protection Agency,  San  Francisco,  California,  October 30,
    1973.

 2.  Trijom's, J.  C.;  T.  K.  Peng;  G.  J.  McRae;  and  L.  Lees.   Emissions and
    Air Quality Trends  in the South  Coast Air  Basin,  Environmental  Quality
    Laboratory, California  Institute  of Technology,  Pasadena, California,
    January 1976.

 3.  Tuerk,  E.  Air Quality,  Noise and  Health,  Report of a  Panel  of the
    Interagency Task  Force  on Motor  Vehicle  Goals  Beyond  1980, U.  S.
    Environmental  Protection Agency, Washington, D.  C., March 1976.

 4.  U.  S.  Environmental Protection Agency,  Compilation  of  Air Pollutant
    Emission Factors  (AP-42), Second  Edition,  Research  Triangle  Park,
    N.  C.,  February 1976.

                                     1101
-------






oo
I—
2^
-
1 —
t— H
	 1
— i
«=£
CO
UJ
— 1
CO
<
1 —














CO

r-.

vo

co *
0) LO
CO
(C
CJ

*d-


co



CM


•—














CO
c
o
T-
P
Q
E
3
CO
CO
<





+
+
<3~ CO LO CM
r-. _J
+
*± r- LO _l
r- ID
+
r— OO LO _1
r>« rs
+
«*• co LO _i
f>. 3


4.
cn oo LO _J
• r*. 3

+
*J- CO LO r—
r>» _j


+
LO CO LO _l
t-^ =3
r—

^- CO O — 1
LO ID








CO
i- fc« »S
^•"•H 1 1
O)
, 5
n)
-o o:
S-
(0 .E r- TJ
T3 -P O O
C 3 i- .C
(O O P P
P i- C • >• OO T-J
o
>>>>>> s-
& ff £ 0.
3 3 to
Q Q C P
o c
P -P -i- «
x: J= P XJ
O> CD  X
_1 _J CO O

VO
•*
LO
CM
LO

0
LO



VO
<3"

vo
«s-



5

Cf>
CM











•P
C
n
CO
>
2
CL
B. *
t-t *
O
CO O
cno
IO CM
•P
c o
CO P
o
i-o
cor^
o_ cr>
^_
CO
ens
gg
cot-
«a:
a>
•si-
CT>
LO
CM
vo

0
1^.



VO
r>»

vo
00



 O
CO C
a: o
o
t-
CO T3
cn c
i. 3
to p "o
_i £- c
cn 3
1^ ^: o
0 t-
•o  p
•i- i. S- LO S~
•P  • N
UJ C
IO -P JC .C
CO -Q i- -P 4->
r— S- E T- T-
-Q Z3 -r- 5 2
(0 —1
CO S- S-
O CO S- IO IO
CO CO CO CO
10 c: a. c c
a> o Q. i- T-
o: ZD _j _i
•a
•P CO II II II
CO CO
o 
-------
                                                                             22-5
                       TRENDS  IN  AMBIENT LEVELS OF OXIDANT
                  AND THEIR  POSSIBLE UNDERLYING EXPLANATIONS

                  E. L. Martinez,  N.  C.  Possiel, E. L. Meyer
                  L. G. Wayne,  K.  W.  Wilson, and C. L. Boyd*
ABSTRACT
           papeA deACAibu woAk  andeAway within the. Ofifii.ce.  ofi  kin. Quality Plan-
ning and StandoAdii, U. S. EnviAonme.ntal Protection Age.nct/,  to Ae.v-iew and eval-
uate, trend*  ofi ambient o aidant /ozone. concentration* at Aite*  -in. the nation
that have. suitable amount!, ofi h4AtoAA.cal data..  Among the.  tAend paAametcAA
being examined annually aAe-  (1}  the. second htghe^t hourly concentration;  (2)
the. average.  dajJLy maxAjnum hourly concentration; and  (3) the. peA.ce.nt o{, day-i,
Mith vatid kouAJLy A&adingA exceeding the. National. Ambie.nt  MJI Quatity StandaAd.
In aAe.aA having moAe. than one. A-ite.,  •dpe.ctal emphaAiA it> gi.ve.n to the. spatial
vaAiation. o
     TAe.ndi>  -in oxA.dant/ ozone. le.ve.lA  aAe. oJU,o beting e.xamine.d in Ae-lation to
hydAocoAbon  and nttAoge.n oxx.de-6  e.miAton friend*.  A majoA.  objective. o& evo€u-
ating the.&e.  AeJLation&hipA Lt> to  de^teAmtne. L^ tAe.nd anaJtyt>ti> can be o6ed to
e-valaate. the. e^ec^cvene^-6 o&  ox^.dant control pAogAam*.  Ambtent oxA.da.nt/ ozone.
tAe.ndi> and coAAe^ponding ern^ci-i-con  tAe.nd data have, been examined in Ae.veAal
CaLihonnJja. MA BOAJM.A; Ve.nveA, ColoAado;  Philadelphia, Pe.nnt>ylvanta;  and a
non-uAban  bite, -in wuteAn Maryland.
      In  oAe.oi,  wheAe. Atatii>tically ^tgni^-icant &ie.ndA in amb-ie.nt ox4.dantf ozone.
le.veJLi> have, been faound, theAe. WOA Ae.a&onable. coAAUpondcnce. to  tAendb -in
hydAoroAbon e.mu>AtonA.  Howe.veA,  the. combined hydAocaAbon  and oxtde-6 o^ nitAo-
gen e.rniA-!>i.on trundi, w&ie. fiound to be. moAe. Atgnifiicantly coAAclate.d to oxA.da.nt/
ozone. &ie.ndA than tho&e. oft hydAocaAbon e.miAA'ionA alone..
                          INTRODUCTION  AND BACKGROUND

     In  several  locations across  the United States, where  attainment of the
National  Ambient Air Quality Standards (NAAQS) for photochemical  oxidants (Ox)
(0.08 ppm — one  hour average)  is  unlikely, additional measures  to reduce
*E. L. Martinez  and N.  C. Possiel, U.S.  Environmental Protection  Agency,
 Research Triangle  Park, North Carolina;  on  assignment from the National
 Oceanic and Atmospheric Administration,  U.S.  Department of Commerce.
 E. L. Meyer,  U.S.  Environmental Protection  Agency, Research Triangle  Park,
 North Carolina.
 L. G. Wayne,  K.  W.  Wilson, and C. L.  Boyd,  Pacific Environmental  Services,
 Inc., Santa Monica,  California.


                                      1103
-------
precursor emissions are necessary.   Since these measures may sometimes be
expensive or difficult to implement, the question arises, "Have past emission
controls been effective in improving ambient concentrations of photochemical
Ox or ozone (Os)"?  Perhaps an effective means for answering such a question
is through ambient air quality trend analysis.

     Observed ambient, Ox/03 trends are likely to depend upon a number of
factors. Among these factors are population and industrial growth patterns,
trends in non-methane hydrocarbons (NMHC) emissions, changes in reactivity of
the organic mix, trends in emissions or ambient concentrations of oxides of
nitrogen (NOx) and annual variations in the frequency of meteorological con-
ditions conducive to pnotochemical  Ox formation.  It should be noted that in
the U.S., Ox/03 control strategies have stressed reductions in non-methane or
reactive NMHC or RHC emissions.

     Results of Ox/0o trend studies have been reported in the literature.
Notable among these have been those of Altshuller (1) and Trijonis et al. (2).

     Altshuller (1) reviewed data from six Continuous Air Monitoring Program
(CAMP) sites in the United States.   These were central city sites located in
the following U« S. cities:  Chicago, Cincinnati, Denver, Philadelphia, St.
Louis, and Washington, D,  C.  He noted significant reductions both in observed
levels of ambient Ox/03 and in the frequency of high levels between the two
periods studied, 1964-66 and 1971-73, at five of the six sites,.  The reduc-
tions in the frequency of the highest concentrations (>0.125 ppm) appeared to
be the most dramatic, ranging from 67 to 94 percent at the stations exhibiting
downward trends.  Perhaps a portion of these long-term trends could have
resulted from downward trends in NMHC emissions and upward trends in local
nitric oxide (NO) emissions from motor vehicles between the tv/o periods.
However, no concerted effort was made to relate ambient Ox trends to precursor
emission trends.

     A comprehensive analysis of air quality and the relation to emission
trends in the California South Coast Air Basin (i.e., Los Angeles area) was
prepared by Trijonis et al, (2) in a California Institute of Technology re-
port.  Table 1 provides a synopsis of emission and air quality trends pre-
sented in the Cal.  Tech.  report.  The analysis covered the 10-year period
1965-74.  An approximation of the prevailing wind direction can be thought of
as oriented from west to east or left to right in the table.   In general, the
large geographical  differences in HC and NOX emission trends were found to
agree reasonably well with the observed geographical variations in ambient 0
(average daily 1-hour maximum) and N02 (annual average) concentration trends?
In evaluating geographical agreement between ambient Ox and precursor emission
trends, consideration of local NO depletion and transport must be taken into
account.   Overall,  the Basin-wide trends indicated:

     (1) a 36 percent increase in NOX emissions was  accompanied by a 35 per-
         cent increase in nitrogen dioxide (N02) concentrations; and

     (2)  an 18 percent reduction in reactive HC emissions was accompanied by a
         19 percent reduction in ambient levels of Ox.   The Cal. Tech. analysis
         provides strong evidence that reduction of organic pollutants reduces


                                    1104
-------
                 TABLE 1.  EMISSIONS AND AIR QUALITY TRENDS  IN
                     THE SOUTH COAST AIR BASIN, 1965-1974
PREVAILING WIND DIRECTION i^
COUNTY
% CHANGE IN
RHC EMISSIONS
% CHANGE IN
NOX EMISSIONS
% CHANGE IN
COUNTY-WIDE
OXIDANT
(# OF STATIONS)
% CHANGE IN
COUNTY-WIDE
N02(#OF
STATIONS)
%OF BASIN
POPULATION
(1970)
SANTA
BARBARA
- 9%
+69%
NO
DATA
NO
DATA
2.6
VENTURA
- 2%
+42%
NO
DATA
NO
DATA
3.7
LOS
ANGELES
-24%
+25%
-31%
(8 SITES)
+19%
(8 SITES)
68.7
ORANGE
+ 6%
+89%
- 12%*
(2 SITES)
+93%
(2 SITES)
13.9
RIVERSIDE
- 8%
+69%
+12%
(2 SITES)
NO
DATA
4.5
SAN
BERNARDINO
-17%
+38%
+ 2%
(1SITE)
+47%
(1 SITE)
6.7
BASIN-WIDE
AVERAGE
-18%
+36%
-19%
(13 SITES)
+35%
(11 SITES)

 *ONE SITE REPORTED AN INCREASE OF +78%, THE OTHER A DECREASE OF -31%

  (FROM TRIJONIS ET AL. • REF 2)
         ambient levels of Ox.  However, an increase in NOx emissions  espe-
         cially in the vicinity of the Ox monitoring sites may  have  influenced
         the ambient 0  reductions.
                      X

     More recently, Pacific Environmental Services, Inc.  (PES),  in a study
done for the U. S.  Environmental Protection Agency  (EPA), performed  an analysis
of trends in ambient Ox/03 and precursor emission levels  for  areas where  these
trends have not previously been extensively evaluated.  The remainder  of  this
paper describes the PES Study and its principal findings  (3)  supplemented by
related trend analyses performed within EPA.


                         DESCRIPTION OF TRENDS  DATA

     In the PES Study, consideration was limited to locations other  than  the
Los Angeles Basin with 5 or more consecutive years of  rather  complete  data.
At least a 5-year period was considered to be necessary to minimize  possible
meteorological biases.  If the period of record is much shorter,  the meteoro-
logical influences on air quality trends are likely to be dominant.  Over a
sufficiently long period, variations due to meteorological anomalies are
likely to be identified and the effects of emission changes,  growth, etc.
probably will have a more consistent influence  on ambient trends.

     Initially, only the years 1970-74 were considered.   This period had  the
broadest common data base.  Yet nearly all the  suitable ambient  data sites
were located in California.  Changes in analytical method, missing data,
                                     1105
-------
changes in monitoring locations or, as was often the case, a brief history of
monitoring for Ox/03 limited the base of suitable trends data outside of
California.  Suitable sites were located in the San Francisco Bay Area (9
sites), San Joaquin Air Basin (Stockton, Modesto), Southeast Desert Air Basin
(Lancaster), Sacramento Valley Air Basin (Redding), North Central Coast Air
Basin  (Salinas), all in California; and in Denver, Colorado; and Philadelphia,
Pennsylvania.  In addition to the ambient data, HC and NOX emissions data for
the individual years 1970 through 1974 or information that would have indi-
cated  the relative trend during the period were obtained, as available, from
EPA and state/local air pollution control agencies.

     Ambient trend data were also examined for a non-urban site, Garrett
County, Maryland, located in the extreme western part of the state.  This non-
urban  site was the only one found to have a rather lengthy period of record (4
years).  However, measurements were taken only during the summer Ox season.
For this particular site, only ambient data and apparent trends, if any, were
to be  noted. No attempt was to be made to evaluate the relation to emission
trends since there is presently no reasonable, non-arbitrary means of assign-
ing an area of emissions that would significantly affect a remote non-urban
site.

     A map of the California Air Basins and station locations outside the Bay
Area Air Basin are shown in Figure 1.  Figure 2 shows station locations in the
Bay Area Air Basin.  Site locations outside of California are shown in Figure 3.


                      DATA ANALYSIS AND INTERPRETATION

     The following ambient 0 /03 trend parameters were analyzed at each urban
site annually, the:         x

     •   2nd high hourly concentration

     •   annual average daily maximum hourly concentration

     •   third quarter average daily maximum hourly concentration

     •   frequency of days with hourly concentrations greater than 0.08 ppm
         (NAAQS)

     •   frequency of days with hourly concentrations greater than 0.16 ppm
         (2 x NAAQS).

     Table 2 shows the indicated direction of the first four trend parameters
at all sites during 1970-74.   The fifth parameter, values greater than 0.16
ppm, occurred too infrequently to evaluate trends.  Note that different trend
parameters may indicate contrasting trend directions.  Only the trends of the
pooled values of the 2nd high concentration in the San Francisco Bay Area and
the average daily maximum in Denver proved to be statistically significant at
the 5 percent level.   The trends were downward in both cases.
                                     1106
-------
      STATIONS

 SF   SAN FRANCISCO
 RC   REDWOOD CITY
 SJ   SAN JOSE
 FR   FREMONT
 SL   SAN LEANDRO
 OK   OAKLAND
 Rl   RICHMOND
 PT   PITTSBURG
 LV   LIVERMORE
   Figure  1.  Bay area  air basin station locations.
TABLE  2.  TREND  DIRECTIONS FOR  VARIOUS OXIDANT INDICES,
       AT  INDICATED MONITORING  STATIONS, 1970-74

SECOND-HIGH
AVERAGE MAX.
3RDQTR. AVG. MAX.
FREQUENCY "• 0 08 ppm
SF ^
-
-
-
-
RC
-
+
,-
-
SJ
+
+
-t-
+
FR
-
+
-
-
SL
-
+
+
-
OK
-
+
-
-
Rl
-
-
-
-
PT
-
-
-
-
LV
-
+
+
+
RD
0
+
+
-t-
ST
+
4-
+
-f
MO
-
-
-
-
SA
-f
+
+
+
LN
+
+
+
+
DV
-
-
-
+
PH
+
+
+
-
SF   SAN FRANCISCO       SL
RC - REDWOOD CITY       OK
SJ   SAN JOSE           Rl
FR - FREMONT           PT
                           STATIONS

                 SAN LEANDRO      LV  LIVERMORE
                 OAKLAND         RD - REDDING
                 RICHMOND         ST  - STOCKTON
                 PITTSBURG         MO  MODESTO
SA - SALINAS
LN - LANCASTER
DV  DENVER
PH - PHILADELPHIA
                            1107
-------
          NORTH
          COAST
                      STATIONS
                   RD  REDDING
                   ST  STOCKTON
                   MO  MODESTO
                   SA  SALINAS
                   LN  LANCASTER
               SAN
            FRANCISCO
            BAY AREA
                    SA

                  NORTH
                 CENTRAL
                  COAST
                       SOUTH
                      CENTRAL
                       COAST
                                SOUTH
                                COAST
                                            SAN
                                            DIEGO
               Figure 2.  California  air basin station locations
                          (Outside  San  Francisco Bay area).
     Since the emission  data from the Denver vicinity were not available  in
sufficiently detailed  form,  evaluation of relation between ambient data and
emissions could  not be made.   Therefore, the San Francisco Bay Area was the
only area (outside of  Los  Angeles)  with sufficient information to analyze  for
the relation of  HC and N0v emission trends or control programs to definitive
0  trends.
 X
x
     The ambient Ox data  showing  significant trends in the Bay Area are shown
in Table 3.  Note the  large  improvement in western areas of the Basin  in  con-
trast to eastern areas  (e.g.  San  Jose  and Livermore).   Table 4 is indicative
of HC and NOX emissions trends  in the  Bay Area during the 1970-74 period  (in
box).  Table 4 shows that  overall emissions have not changed much in the  most
recent five years.  Only  HC  emissions  show a slight but gradual decrease.
                                     1108
-------
            Figure 3.  Station  locations outside  California.
TABLE 3.  OXIDANT LEVEL BY YEAR AND
TREND DIRECTION (2ND HIGHEST DAILY
MAX. HOURLY AVG., PPM)
TABLE 4.   ESTIMATED EMISSIONS* AND
COMBINED EMISSION RATIOS FOR HYDRO-
CARBON AND OXIDES OF NITROGEN IN
SAN FRANCISCO BAY AREA, 1962 -
1974

SAN FRANCISCO (SF)
REDWOOD CITY (RC)
SAN JOSE (SJ)
FREMONT (FR)
SAN LEANDRO (SL)
OAKLAND (OK)
RICHMOND (Rl)
PITTSBURG (PT)
LIVERMORE(LV)
'70
.14
.16
.16
.27
.21
.20
.18
.18
.29
71
.15
.16
.14
.24
.24
.21
.14
.18
.21
'72
.07
.17
.17
.29
.14
.09
.10
.18
.18
'73
.08
.12
.18
.18
20
.14
.10
.10
.20
'74
.06
.16
.26
.21
.17
.09
.08
.10
.25
TREND
-
-
4-
-
-
-
—
-
—
YEAR
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
HC
1300
1300
1400
1300
1300
1300
1300
1300
1140
1110
1130
1070
1030
NOX
490
500
550
550
620
620
670
700
740
740
790
820
720
HC/NOx
2.65
2.60
2.55
2.36
2.10
2.10
1.94
1.86
1.54
1.50
1.43
1.30
1.43
                                               *TONS PER DAY
                                  1109
-------
However, more detailed emissions data by county (not shown) and source type
(stationary or mobile) indicate that:  (1) significant HC emission reductions
have been achieved in the western part of the Bay Area; and, (2) mobile sources
NOX emissions have increased in most places.  Air quality and emissions trends
reflect new auto and stationary source emission control programs and differential
area growth patterns.  Notably, some eastern areas of the Basin have undergone
rapid growth, thus overshadowing beneficial effects of emission control pro-
grams .

     The correlation coefficient between pooled second highest Ox concentra-
tions in the Bay Area and regionwide HC emissions was only + 0.58, not statis-
tically greater than zero at the 5 percent  level.  The significance test
involves the Student _t test, as explained by Moroney (4).  The correlation was
greater when both HC and NOX emission trends were related to oxidant trends.
The relation between HC/NOX emission ratio and second highest Ox levels was
considerably higher at + 0.88 and statistically significant at the 5 percent
level.  The improved results are indicative of the important role of NOX
in the photochemical process.  In these results the variable effect of meteoro-
logical factors was not considered.

     A much clearer picture for the Bay Area emerges when 13 years of data are
examined and normalized by considering only days with photochemical Ox conducive
meteorological conditions.  Table 5 denotes the normalized Ox trends.  Over
the 13-year period for which data are available, the downward trend was found
to be highly significant (p<.001).   Thus, it is virtually certain (i.e. 99.9
percent) that improvement in the Bay Area oxidant levels is attributable to
some physical phenomena rather than chance.  The possibility that the improve-
ment was unduly influenced by meteorological factors was minimized by the long
period of record and consideration of ambient Ox levels  only for days with
similar meteorological conditions.  Table 4 shows basin-wide HC arid NOX
emission trends for  1962-74. Note  the much wider range of emissions occurring
during this  longer period.  When HC emissions (see Table 4) were correlated
with normalized Ox levels pooled for the six stations  haying complete records
over the 13-year period  (1962-74), a correlation coefficient of + 0.77, statis-
tically significant  at the  5 percent level, was found.   Moreover, the correla-
tion coefficient between Ox and HC/NOX emission ratio  trends was even more
highly significant, + 0.92, significant at  the 0.1 percent  level.

     The available 03 trend parameter data for the non-urban station at
Garrett County, Maryland, is given in Table 6.   No apparent trends were noted.
What seems evident, however, are considerably large fluctuations in some of
the statistics from year to year.   For instance, the percent of hours that the
hourly concentration exceeded the ambient standard varied by more than a
factor of 3.  Probably, these variations in the large-scale regional 0 /03
levels may be reflected in the levels experienced in urban areas.      x

     An obvious deficiency of this  paper was the lack of suitable trends
information outside of California.   Improvement in  the trend data base, how-
ever,  is forthcoming.  At least ten states  outside  of California now have
sites  with three consecutive years  of Ox/03 measurements.  These data will
eventually provide a data base for  assessing nationwide 0  trends  and their
probable causes.                                          x

                                     1110
-------
  TABLE 5.  AVERAGE HIGH-HOUR OXIDANT CONCENTRATIONS  FOR  DAYS WITH COMPARABLE
  TEMPERATURE AND INVERSION CONDITIONS  (APRIL  THROUGH OCTOBER OXIDANT SMOG
  SEASONS, 1962 - 1974)
AVERAGE HIGH-HOUR OXIDANT CONCENTRATION
(Kl PARTS PER MILLION)
MONITORING
STATIONS
SAN FRANCISCO (SF)
SAN LEANDRO (SL)
SAN JOSE (SJ)
REDWOOD CITY (RC)
WALNUT CREEK (WCI
SAN RAFAEL (SR)
BAAPCD
AVERAGE*
LIVERMORE**
'62
.14
.13
.11
.13
.10
08
.12
--
•63
.12
.16
17
.10
.11
.09
.12
--
'64
.15
.19
14
.10
10
.07
13
--
'65
.09
.19
.16
.14
.11
.08
.13
--
'66
.08
.14
.11
.10
.10
.07
.10
--
•67
.08
.12
.13
.09
.13
07
.10
.13
'68
.05
.11
.13
.08
.10
.06
.09
.18
•69
.04
.12
.13
.09
13
.07
.10
.18
70
.07
.12
.12
.08
.09
08
.09
13
71
.05
11
.08
.07
.09
07
.08
.11
'72
.03
10
.10
.08
09
05
.08
.09
'73
.04
.11
11
.07
.08
05
.08
.12
'74
.05
.10
.16
07
.08
.06
.09
.13
13
YEAR
.08
.13
.13
.09
10
.07
.10
.13
OXIDANT TREND
DIRECTION
1970-74
ONLY
-
-
+
-
-
—
—
+
ALL
DATA
-
-
—
—
—
—
—
—
   •FOR BENCHMARK STATIONS ABOVE, WITH 13 YEARS OF RECORD

   •STATION WITH 8 YEARS OF RECORD
                                                   	NO DATA
(SOURCE:   INFORMATION BULLETIN 3-25-75:  A STUDY OF OXIDANT  CONCENTRATION  TRENDS;
 TECHNICAL SERVICES DIVISION, BAY AREA AIR POLLUTION CONTROL DISTRICT)

                          TABLE  6.   OXIDANT/OZONE TREND
                          STATISTICS FOR  GARRETT CO.,
                          MARYLAND
STATISTIC
MAX HOURLY CONC. (ppm)
MEAN HOURLY CONC. (ppm)
% DAYS HOURLY CONC. > 0.08 ppm
% HOURS CONC. > 0.08 ppm
YEAR
'70
0.13
0.05
40
27
71*
--
--
--
--
'72
0.12
0.05
40
11
'73
0.15
0.07
78
37
'74
0.15
0.06
43
13
                 *NO SAMPLING PERFORMED IN 1971
                             REVIEW AND CONCLUSIONS

      The following statements summarize trends information as described  in
 this  report.   Long-term trends in oxidant in the Bay Area Air Basin  are  signi-
 ficantly related to trends in HC emissions.  Oxidant concentrations  in the  Bay
 Area  are more highly related to HC/NOX emission ratios.  This indicates  re-
 ductions in  NO  emissions may have a tendency to increase Ox levels  within  the
 Basin.   Gxidanr problems in the Bay Area, due to changing patterns of precur-
 sor emissions, are becoming more severe in inland locations than in  Bayside
 locations.   Monitoring data for a 5-year period may not, in general  be suffi-
 cient to establish statistically significant Ox trends in an area.   And  concen-
 trations of  Ox/03 in non-urban locations exhibit large fluctuations  from year
 to year that may have an influence on urban oxidant levels.
                                      1111
-------
                             REFERENCES

Altshuller, A. P.,  "Evaluation  of Oxidant  Results at CAMP sites in the
United States,"  JAPCA 25 (Jan.,  1975).

Trijonis, J. C.,  T.  K. Peng,  C. J.  McPae and L. Lees,  "Emission and Air
Quality Trends in the South  Coast Air  Basin,"  Environmental Quality
Laboratory, California Institue of Technology  (Jan., 1976).

Wayne, L. G., K.  W.  Wilson,  and C.  L.  Boyd,  "Detection and  Interpretation
of Trends in Oxidant Air Quality,"  DRAFT REPORT Prepared Pursuant to EPA
Contract 68-02-1890, Task Order 1  (July 1, 1976).

M. J. Moroney, "Facts from Figures," Third Edition, p. 311, Penquin Books,
Baltimore, Maryland  (1956).
                                1112
-------
                 SESSION 23
ON THE OZONE/OXIDANT CONTROL STRATEGY IN U.S.

                      A.B. Bromley
                OECD, France
                     1113
-------
                                                                           23-1
               TRENDS  IN PHOTOCHEMICAL OXIDANT  CONTROL  STRATEGY

                                  J.   Padgett*
ABSTRACT
     The national, ambient air quality *tandard  far ox.ic evidence ofi health and other adverse  e^ect* *upport thiA
*tandard although data Limitation* even today provide ample opportunity to
chatting?, it.  The ba*ic Environmental Protection Agency *trategy  ^or the,
control Of} oxiidant i& to contnot anthropogenic. e.mJJ>Aion& ofa organic compound*
that can react photochnmicatiy to ^orm ox.idantt>.  It 4eemi c£eoA that thi&
approach i& bound eve.n though much additional, research  -06 needed to answer
question* aAtociated with the e^ecti, o^ ox.idanti> and the be&t approach to
their control.  Oxidantb originally were o& concern only in large  citi&>,
but recent technical, finding* *hoio that high oxA,dant level* (exceeding the
standard] can occur over large geographical area* because ofa the transport
oft ox.idant* and their precursor* over a *everal day period.  Given enough
time, even low reactive organic* can faorm ox.idant* and  are o&  concern.
The emerging oxi.dant control *trat.egy .therefore place*  more empha*i* on
the control o{> ati organic* emitted within large geographical  area*.  In
parall.el with the continuing implementation o&  oxidant  control plan* in
many urban area*, the Environmental Protection  Agency i* developing  new
*ource regulation* to control organic cmi**ion* ^rom chemical,  plant*.
Guideline* and regulation* are al*o being developed to  effectively reduce
organic emi**ion* firom other new and ex.4*t^ng *ource*,  and State*  are
encouraged to apply reasonably available emi**ion contAol* over  broad
geographical, area* where emi**ioyi den*itie* are relatively high.

                                 INTRODUCTION

     To protect the public from the harmful  effects of  photochemical oxidant,
the U.S.  Environmental  Protection Agency (EPA) promulgated a national ambient
air quality standard in April, 1971(1).   This standard  placed a limit on the
amount of photochemical  oxidant that could be present in the air after July,
1975 (with some provisions for extending this deadline  for two years), based
on the best available scientific evidence of health and other adverse ef-
fects(2).   Many difficulties have been encountered as federal, state, and
local governments have attempted to implement air pollution control programs
which would ensure that the standard would be met.  Even today, there are
many places in the country that experience photochemical oxidant levels sub-
stantially in excess of the standard.
 *U.  S.  Environmental  Protection  Agency,  Research  Triangle Park,  North Carolina,

                                      1115
-------
     Our basic oxidant control strategy is to reduce anthropogenic emissions
of organic compounds that can react photochemically to form oxidants.  Even
though the rationale for this approach is well documented(S), questions still
are raised by community and industry leaders responsible for implementing it.
Typical questions are:  Why do we need to meet the oxidant standard?  Are
we concerned only with ozone  (which is the pollutant we monitor) or with all
oxidants?  Is the basic strategy of reducing hydrocarbons a valid one?  What
is the impact on control strategies of hydrocarbon reactivity, background
oxidant levels, and the transport of oxidants and their precursor into and
out of a community?

    There are as yet no simple non-controversial answers to many of the above
questions.  And in some cases, as you have heard at this conference, scienti-
fic opinion is sharply divided on the answers to these and other questions we
could ask.  But if we are to  succeed in our national goal of oxidant control,
it is important for the public, as well as the scientific community, to have
answers to these questions and to be convinced that the control strategies we
employ are needed and will be effective.  Conferences such as this are a
vital part of this process.   Also important is the support of independent
scientific organizations such as the National Academy of Sciences, which
stated in its comprehensive 1976 report on Ozone and Other Photochemical Qxi-
dants(4) the following:

          Despite uncertainties concerning the causative agents
     and their effects, we must proceed with the regulation of
     emissions that lead to the formation of photochemical smog.
     At the same time, research should continue on identifying
     the individual harmful agents in photochemical smog and
     their effects.

     This paper first will consider briefly the scientific support for the
oxidant standard.  We then will describe the current EPA control strategy
for oxidant control and how this strategy is evolving as a result of new re-
search findings.  We will not, however, attempt in this brief paper to try
to answer all of the questions we have posed.

                       SUPPORT FOR THE OXIDANT STANDARD

     The U. S. national ambient air quality standard for photochemical oxi-
dants is 0.08 parts per million (ppm).  This is an hourly average, not to
be exceeded more than once per year.  While few will disagree that high con-
centration levels of ozone and other oxidant cause serious adverse health
effects, some people question the need for such a restrictive standard, and
others question the basis on which the current standard was established.  A
summary of the health implications of oxidants is given in Reference 5.
Questions continue to arise because the data on human health effects from
exposure to low levels of photochemical oxidant still  are limited and hence
may never be completely settled even though many independent scientific
reviews by medical experts support the present standard.
                                      1116
-------
      One such study was commissioned in 1974 when (,on<)rtss asked trie National
 Academy of Sciences (NAS) to review and update the information on health
 effects from air pollution, including those effects caused by photochemical
 oxidant.  After reviewing and interpreting the available data, these highly
 qualified scientists from the various scientific and technical fields con-
 cluded:

           In general, the evidence that has been accumulated since
      the promulgation of the Federal ambient air quality standards
      by the EPA Administrator on April  30, 1971, supports those
      standards.  Hence on the balance,  the (NAS) panels found no
      substantial basis for changing the standards(6).

    Although  this  conclusion  applied to  all  national standards,  the  NAS,
 in  specifically addressing  photochemical oxidant,  noted that  even  with  the
 current standard,  the  risk  to  the  general  population may not  be  negligible,
 and that there  is  evidence  which suggests  that  there may not  be  a  completely
 safe  level  of this air pollutant for all people.   A similar conclusion  has
 more  recently been reached  in  a  report  by  the  World Health Organization
 (WHO),  which  states:

          It  is apparent that any primary protection standard  (for
     oxidant) between  .05 and  .10 ppm will  provide the narrowest margin
     of  safety against some possible detrimental effects in the more
     susceptible segments of the population.  There would seem to be
     little justification for exceeding  .10 ppm for a primary protection
     standard(7).

                           OXIDANT CONTROL STRATEGY

     Our understanding of the photochemical oxidant formation process has
grown over the years because of both improvements in the theory of photo-
chemical oxidant formation and continuing empirical research.   However,
from a regulatory point of view, the most significant fact remains that
oxidants are formed from two types of precursors:  nitrogen oxides and
organic  compounds  (of which hydrocarbons are a major component).  In most
urban areas, the ratio of the organic concentrations to the concentration
of nitrogen oxides is such that the nitrogen oxides are in surplus and
hence the organics are the limiting factor in the set of reactions which pro-
duce oxidants.  Therefore, hydrocarbon and other reactive organic emissions
must be  reduced in order to lower oxidant levels.

     The reductions in emissions of hydrocarbons and other organic compounds
are to be achieved through Federal  and State programs that have been formal-
ized in  regulations promulgated under the Federal Clean Air Act of 1970(8).
The Federal programs provide for the reduction in emissions nationwide
through  the Federal Motor Vehicle Control Program, the Federal program for
control of aircraft emissions, and the development of New Source Performance
Standards.  The State programs provide for addit-onal control  measures through
State Implementation Plans in those areas of the country where the Federal
programs will not be sufficient to meet the air quality standard for oxidants.
                                     1117
-------
     Because roughly half of the hydrocarbons emitted to the ambient air
in the United States currently is attributable to motor vehicles, a large
portion of hydrocarbon reduction is to be achieved through the Federal
Motor Vehicle Control Program.  This program has required progressively
stricter hydrocarbon controls on all light-duty motor vehicles since the
1968 model year.  In 1970, the Clean Air Act required that hydrocarbon
emissions from light-duty vehicles and engines be reduced by at least 90%
from emissions permissible in 1970 model year vehicles.  This will eventually
require emissions to be reduced to 0.41 grams per mile.  Interim standards
of 1.5 grams per mile (g/m) are now required on 1975 through 1977 model
year vehicles.  Regulations also exist to limit hydrocarbon emissions from
other types of vehicles such as motorcycles, trucks, buses, and aircraft.

     The Clean Air Act also authorizes EPA to promulgate "Standards of
Performance for New Stationary Sources".  These are standards for new sour-
ces which reflect the best demonstrated system of emission control, taking
the cost of emission reduction into account.  New sources are sources that
are constructed or modified after the standard has been proposed.

     When Federal programs are not sufficient to meet the oxidant. standard,
the State is required to develop plans for additional reduction in emission
of organic compounds in those areas that will not meet the oxidant standard.
The State Implementation Plans include regulations for existing stationary
sources as well as the implementation, where necessary, of Transportation
Control Plans.

     For the development of Implementation Plans, it is necessary to deter-
mine how much the emissions of hydrocarbons and other organics must be re-
duced.  It was realized that the relatively simple methods used for directly
emitted pollutants were not appropriate for the complex oxidant formation
processes.  Therefore, an attempt was made to quantify the amount, of hydro-
carbon reduction required to meet the oxidant standard.  In 1970, the data
to relate non-methane hydrocarbon to oxidant concentrations were available
only from Denver, Los Angeles, Philadelphia, and Washington, D.C.  The
6 to 9 a.m. average non-methane hydrocarbon concentrations were plotted
against the peak hourly oxidant concentrations observed later in the day
at the same measurement sites.  The points scattered because of the varia-
tions in the conditions that affect oxidant formation.  A curved line was
drawn to enclose the data points within an upper boundary.  This upper limit
curve depicted the highest oxidant levels observed for a given hydrocarbon
concentration.  From this curve the minimum hydrocarbon reductions required
to meet the oxidant standard could be calculated.  EPA published the result
of this calculation as Appendix J in the August 14, 1971 Federal Register.
Given a measured maximum oxidant level in an urban area, Appendix J is to
be used to determine the amount of hydrocarbon control required in that area.
In some cities the measured oxidant levels were above (greater than 0.28 ppm)
the applicable range of Appendix J.  These cities were permitted to reduce
hydrocarbon emissions proportional to the amount that oxidant levels exceeded
the standard.

    The August 14, 1971 Federal  Register also gives emission reductions that
are attainable through the application of reasonably available emission con-

                                    1118
-------
trol technology.  These emission limitations emphasize reduction of total
organic compound emissions, rather than substitution of less reactive organic
compounds, because of the evidence that few organic compounds are photochem-
ically unreactive.  For organic solvent usage, however, such as in surface
coatings and dry cleaning, the guidelines present a list of low reactivity
compounds that may be considered for exemption for control.

     Transportation Control Plans provide for reductions in hydrocarbon
emissions beyond the reductions achieved by the Federal Motor Vehicle Control
Program and stationary source regulations set forth in approved State Imple-
mentation Plans.  Transportation Control Plans typically include measures  to
decrease automobile travel, require inspection and maintenance of vehicles,
and limit vapor emissions during gasoline transfer.

                                RECENT FINDINGS

     Recent field investigations and laboratory studies of oxidant formation
and control have provided several new insights.  It is now apparent that
oxidants are a rural as well as an urban problem.  Oxidants can be formed
over long time periods during stagnant conditions in high pressure systems
or during transport of oxidants and precursors.  This implies that the long-
term behavior of oxidants and precursors is an important contributor to
oxidant concentrations.  It also implies that less reactive organic compounds,
as well as the more reactive compounds, can contribute to observed oxidant
levels.  There are also natural sources of oxidants which at times may con-
tribute to oxidant concentrations reaching levels near the oxidant standard.
The studies indicate, however, that man-made emissions of hydrocarbons are the
predominant source of the highest levels of oxidant.

     Non-methane hydrocarbon (NMHC) and nitrogen oxides (NO ) measurements
made in relatively remote, sparsely populated areas of the nation show ex-
tremely low NO  levels and NMHC/NOx ratios that are sufficiently high that
control of NMHC may be of little benefit in these areas.  For urban areas,
however, these ratios appear to be in the range where a hydrocarbon control
strategy is effective.

     Although the theory of a hydrocarbon control strategy is well establish-
ed and has been validated in smog chamber tests, only confirmation in an ur-
ban environment, which requires years of data because of the predominant
year-to-year influence of meteorological conditions, will be convincing to
many.  Emissions and air quality data are becoming available over a suffi-
cient number of years for a few locations—notably Los Angeles and San Fran-
cisco—to allow meaningful statistical analyses of the effectiveness of
hydrocarbon emission reductions in reducing oxidant levels.  Separate ana-
lyses recently completed for these cities do show good correlations between
reductions in hydrocarbon emissions and ambient oxidant concentrations.


                       IMPLICATIONS FOR OXIDANT CONTROL

     Based on findings to date, implications for oxidant control include
the following:


                                    1119
-------
     1.   Recent findings support the basic EPA strategy, which emphasizes
the reduction of organic emissions in urban areas.   In many urban areas,
the required reduction in emissions is far greater than can be achieved by
use of all reasonably available control technology and it is likely that
some cities will not achieve the oxidant standard for many years.  This
problem has been recognized by Congress as well as the EPA, and legislative
amendments have been proposed that grant extensions in time to meet the
standard if the city has itself committed to employ all reasonably available
control measures to reduce emissions.

     2.   EPA policy has strongly encouraged the positive reduction of non-
methane organics but still recognizes as a useful interim control measure
the substitution of low reactivity for high reactivity solvents.  Since
even a low reactivity organic may contribute in time to oxidant formation,
the use of solvent substitution, even as an interim measure, is now being
reexamined.

     There is growing recognition that organic emission controls often should
extend far beyond the boundaries of a defined urban area.  Statewide or even
multi-state control of organic emissions may be effective in reducing oxidants
in populated regions such as the northeastern United States.  At the same
time, however, emission control in remote rural regions of the nation, such
as portions of the northwestern United States, where NO  emissions are very
low, may be counter-productive.  Further study and analysis is needed to
better define this problem.

     4.   Our improved understanding of the mechanisms for formation of
oxidant and the role of transport emphasizes the long existing need for im-
proved predictive models to replace the present Appendix J curve that relates
peak oxidant concentrations to non-methane hydrocarbon emissions.  The use
of improved models will not necessarily affect our current control programs,
particularly in urban areas where required emission reductions are far greater
than can be achieved, but they will establish more accurately long-range
program goals and facilitate the development of more effective area wide con-
trol programs.

            ACTIVITIES IN SUPPORT OF EPA's OXIDANT CONTROL STRATEGY

     In addition to the extensive research programs needed to provide better
answers to basic questions such as those posed earlier in this paper, a
number of other activities associated with control  implementation are under-
way.  These include:

     1.   Model Regulations to limit organic emissions from existing plants,
based on reasonably available control technology, are being completed for the
paint and coating industries.  New regulations for inspection and maintenance
programs and the recovery of emissions during gasoline transfer also are in
preparation.   These will  provide guidance to states which may incorporate
such regulations in State Implementation Plans for oxidant control.

     2.   New Source Performance Standards.   A comprehensive study of the
synthetic organic chemical industry will result in federal standards to limit


                                    1120
-------
new plant emissions of many organic compounds now uncontrolled.  Guidance
documents for control measures for existing plants also will be provided for
use in State oxidant control plans.

     3.   Reactivity.  The role of photochemical reactivity in oxidant con-
trol strategies will be re-examined with the assistance of State and local
regulatory personnel, members of the technical community, and other concerned
groups.

     4.   Substitutes for the Appendix J Model are being developed to provide
a method for accounting for the effect of NO  as well as oxidant and precursor
transport.  These new calculation techniquesxshould provide more accurate
quantification of the required reduction in organic emissions needed to
achieve a desired oxidant concentration.

                                  CONCLUSIONS

     Recent technical findings indicate the need for oxidant control programs
that span much broader geographical regions than orginally believed.  Also,
because of the role of transport of both oxidants and their precursors over
time periods measured in days, even low reactive organics may have significant
time to form oxidants and should be controlled.   The EPA strategy, which
emphasizes the control of anthropogenic organic emissions, is basically sound
and programs are underway to establish more effective control of these emis-
sions.  The evolving oxidant control  strategy places increasing emphasis on
the control of emissions over broad geographical areas where emission densi-
ties are relatively high.  States are being encouraged to apply reasonably
available emission control  measures in these areas as needed to supplement
Federal standards for vehicle emissions and new stationary sources.

                                  REFERENCES

1.   Federal register, Vol. 36, No. 84, "National Primary and Secondary Ambient
     Air Quality Standards", April  30, 1971.

2.   "Air Quality Criteria for Photochemical Oxidants", U.S. Department of
     Health, Education, and Welfare,  Publication No. AP-63, March 1970.

3.   "Control of Photochemical Oxidants—Technical Basis and Implications of
     Recent Findings:, U.S. Environmental Protection Agency, Publication
     No. EPA-450/2-75-005, July 1975.

4.   "Ozone and Other Photochemical Oxidants", National Academy of Sciences,
     Washington, D.C., 1976.

5.   "The Health Implications of Photochemical Oxidant Air Pollution to Your
     Community", U.S. Environmental Protection Agency, August 1976.

6.   "Air Quality and Automobile Emissions  Control--A Report by the Coordinat-
     ing Committee on Air Quality Studies", Volume 2, National Academy of
     Sciences, Prepared for the Committee on Public Works, U.S.
                                     1121
-------
7.   "Environmental  Health Criteria  for  Photochemical Oxidants", World Health
     Organization, Draft Report EHE/EHC/WP/75.5, March  20,  1975.

8.   Clean Air Act (42 U.S.C.  1857 et. seq.),  Clean Air Act Amendments of  1970,
     P.L. 91-604,  December 31,  1970.
                                    1122
-------
                                                                            23-2
             PROBLEMS  WITH CONVERTING STATE-OF-THE-ART PHOTOCHEMISTRY
                         TO STATE LEVEL CONTROL STRATEGIES

                              W.  Bonta and J. Paisie*

ABSTRACT

     A bsiie.fi history  ofi  the, development ofi the. photochemical  oxidant  control
pfiOQfiam in the State,  ofi  Mafiyljand iA presented.  Included in thiA  hiAtony iA
a. diAcuss-ion ofi the. efifiect ofi changes -in Federal legiAlation, specifically,
the, 1967 Ciean AxA kct and 1970  Clean AXA kct ame.ndme.ntf> on the, photochemical
oxidant control.
     In addition to  thiA,  the e.&fie.ctA o& changes -in  the. & tote.- o&- the.- an£ photo
          aAe. cfx^ca44e,d i.n fioJLaution to the. de.veJLopme.nt ofa the. photoche.mic.ai
oxA.dant control. pfioQUtm -in the, State. o& Maftytand.  Spe.ci^icatty , attejnptA
on the. pant ofa the. State, to de.ve£op an alternative, to the. Fe,de.?ial£y piomuJL-
gate-d "Appendix J" one, alAo dii>ciit>&e.d.  Included in  tku, jj> a. di&cu&Aion o£
the. tieAditb ofa the. aie ofi aeAomztAsic data ancULy&iA and analytical, simulation
o£ smog cliambet e.x.pe.tume.ntt> .
     Finally, obseAvationi*  on the. e,^
-------
THE 1967 FEDERAL ACT

     The accumulated photochemical oxidant data prior to the Clean Air Act
of 1967 was compared to the existing California oxidant standard of 0.10
ppm to determine the severity of the Maryland problem.   Although sparse,
the data indicated that both Baltimore and Washington exceeded the standard
less than twenty times per year and peak values were thought to be around
0.15 ppm.  This problem when compared with that known to exist in Califor-
nia was given a second priority behind suspended particulates (TSP) and oxides
of sulfur (S0x).

     According to the 1967 Federal Clean Air Act amendments, the issuance
of criteria and control techniques documents by the Federal agency
triggered the development of a regulatory program within state governments.
In March of 1970, these documents were issued for photochemical oxidants
and for oxides of nitrogen (N0x).  Following this event, the States were to
establish ambient air quality standards protective of the public health and
the appropriate control programs required to implement these standards.
Throughout the summer of 1970, the available documents were reviewed and
a preliminary assessment of control technology was assembled.  The primary
reference point was that of the compiled information contained in both the
Federal criteria and control techniques documents; there had been no general
awareness within the State of the problems or technological solutions associa-
ted with photochemical oxidants until 1970.

     At the end of November in 1970 the State held public hearings at which
the proposed ambient air quality standard to photochemical oxidants of 0.10
ppm was presented.  The presentation report (1) also contained an assessment
of the current levels and a preliminary finding of the necessary control
program to implement these standards.  As there were no modelling data  shown
in the criteria documents that would apply to ambient levels as low as those
found in the Washington Region, the empirical diagrams showing the associa-
tion between 6-9 a.m. non-methane hydrocarbons (NMHC) and afternoon oxidant
peaks were used.  One of these diagrams, found in the hydrocarbons criteria
document (2), showed that peak oxidant values above 0.10 ppm were not observed
when 6-9 a.m. NMHC levels were not above 0.30 ppm.  Thus, a control program
was sketched out that would reduce the mean levels of NMHC such that the peak
value expected was less than 0.30 ppm NMHC.  The available NMHC data showed
this level was exceeded 97% of the time.  Therefore, it was concluded that a
major reduction in mean NMHC emissions was required in order to achieve the
oxidant standard.  The reduction of the mean 6-9 a.m. NMHC level from 0.65
ppm to 0.10 ppm equalled 85% control and was thought to be available through
controls on solvent usage, gasoline marketing losses, dry cleaning vapor
losses, and the adoption of the then proposed Federal new vehicle control
program which provided a 90% reduction from precontrolled autos.  The control
strategy considered for solvent controls was based on California Rule 66, and
was used simply because it was the only adopted regulation available at the
time.  The gasoline marketing controls were developed by the State as a system
for reclaiming fuel vapors from the retail station filling process.  The
vapors were collected in the tank trucks and when the trucks returned to the
gasoline terminals the processing took place.  Terminal controls were pat-
terned after those used in southern California and the controls were applied
only to newly installed tanks.

                                     1124
-------
     In summary, the State thought that it had the beginning of a viable con-
trol program for obtaining the oxidant goal established pursuant to the 1967
version of the Federal Clean Air Act.

THE 1970 FEDERAL ACT AMENDMENTS

     When the Federal Act was changed on the last day of 1970, the State
postponed adoption of its ambient standard for photochemical oxidants un-
til the Administrator of the newly created EPA could promulgate the nation-
al ambient air quality standards required by the amendments.  During the year
1971, the majority of State effort centered around the development of new
capabilities for measuring oxidants using a continuous, telemetered system
of monitors of the chemiluminescence type.  In addition, detailed inventories
were compiled of all sources of hydrocarbons and extensive studies of mobile
and stationary source technologies were begun.  Land use and transportation
studies were also begun.  Throughout the summer and fall of 1971, regulations
implementing the above controls were designed and presented at public hearings
for both the Washington and Baltimore Metropolitan areas.  The Implementation
Plans which were required by EPA to demonstrate how the State intended to meet
the photochemical oxidant standard were forwarded to EPA in December of 1971
and the regulations were adopted as law in March of 1972.  In the process of
generating the 1971 Implementation Plan for photochemical oxidants, the inven-
tory of hydrocarbons was discovered to be less controllable than previously
thought.  This discovery was due to the facts found in the detailed technology
review which showed more realistic reduction potentials than had been origi-
nally contemplated.  The air quality forecasting model had changed, however,
to a linear rollback of NMHC which was assumed to be proportional to peak
oxidant.  The change was based on the guideline included as Appendix J to the
Federal regulations prescribing methods for preparation, adoption and submittal
of implementation plans under the 1970 Act (3).  In the range lower than 0.16
ppm oxidant, the guideline was so nearly linear that linearity was assumed.
The rollback requirement for the Washington area changed from 85% based on
hydrocarbons found in the preliminary survey to 50% based on the oxidant
oriented Appendix J.  Quite coincidentally, the revised estimate of the effec-
tiveness of the planned hydrocarbon control program also showed a 50% emission
reduction potential.  The Baltimore oxidant measurements, however, based on
the 10 minute grab phenolphthalein method prescribed a 30% reduction using
Appendix J.  Due to the greater industrial HC emission component in the in-
ventory, the Federal new car program had a smaller impact on the Baltimore
inventory.  The proposed control program was predicted to net a 26% reduction
in emissions which was considered sufficient for that time.  Thus, the program
designed under the 1967 version of the Federal Act was confirmed under the
1970 amendments.  The State added Rule 66 type solvent controls and gasoline
marketing controls to its arsenal of regulations.  The drycleaning vapor
losses were ignored as insignificant and also the drycleaning industry had
been phasing out the use of photochemically reactive solvents.  In addition,
nitrogen oxides emissions were regulated but only for the sake of meeting
N02 health standards, then thought to be exceedec.  There was no consideration
of NO  emissions in the oxidant control program.
     X
     During 1972, the planning phase of the program started studies of the
long term impact of alternative land use and transportation profiles.  These
studies were thought of as maintenance measures at that time.   The overall

                                    1125
-------
program design parameters remained unchanged until  the new continuous chemi-
luminescence system demonstrated peak ozone levels  over 0.20 ppm.   These new
data plus the Natural Resources Defense Council  vs.  EPA court decision (4)
forced development of an oxidant plan revision in mid-1973.   This  revision
showed that a 70% reduction in 6-9 a.m. HC emissions was necessary to meet
the oxidant standard, still using the Appendix J guideline.   As a  result,
the State regenerated previously set aside plans for controlling drycleaning
solvent losses and the emissions from gasoline marketing at existing stations.
Regulations were adopted for this purpose in October, 1973.   Two months later
a regulation was added which was designed to prohibit new major HC sources
and freeze the inventory of emissions from existing major sources  in the
Baltimore and Washington metropolitan areas.  The most significant feature of
the December, 1973 regulations from the ozone formulation standpoint, was the
use of the term photochemically reactive organic material (PROM) in the regu-
lation.  For the first time the program directly considered reactivity in its
oxidant control program, and such reactivity controls were based on reducing
the production of ozone, not eye lacrimators, as was the case in California.

     During the same period in the winter of 1972-1973, the State  was con-
sidering the air quality impacts of the newly proposed Brandon Shores power
plant to be located in the Baltimore metropolitan area.  Since this plant
was to be a large emitter of N0x, the State wanted to determine the impact on
photochemical oxidant levels within the region.   Unfortunately, the Federally
provided guideline was incapable of evaluating anything but tine effect of NMHC
on oxidant levels.  As a result, the State launched an empirical exercise
aimed at evaluating the aerometric data from the then operational  continuous,
telemetered system.  The objective of this study was to develop a  statistical
relationship between oxides of nitrogen and ozone for the Baltimore-Washington
metropolitan areas.

     The first of two attempts at developing a suitable relationship used
an eigenvector analysis which attempted to isolate  the major determinants
in the production of ozone.  This analysis which was similar in concept to
that of Peterson (5) used data from the six AIRMON  stations for selected days
during the summer of 1972.  The ozone data for a given day was arranged in
a 6 x 24 matrix (6 stations x 24 hours).  From this matrix,  it was possible
to derive a set of empirical orthogonal functions which could then be analyzed
for possible physical significance.  The orthogonal  functions consisted of
two matrices.  One matrix was 6x6 and the vectors in this matrix were re-
lated to station located parameters.  The second matrix was 6 x 24 and con-
sisted of vectors which were related to time of day.

     This process was repeated for a number of days during the summer of 1972.
The first orthogonal function, in most cases, accounted for 85-95% of the
variance in the data.  The first vector which was related to time  of day
correlated well with temperature or solar angle  for the test day.   Since the
first orthogonal function accounted for most of  the variance, no other signi-
ficant features of the data could be isolated.  Based upon these rather dis-
appointing results, this technique was abandoned.
                                     1126
-------
     Since the first attempt resulted in no information of value relating to
NO , a second analysis using multiple correlation was tried (6).  The analysis
technique was patterned after an exercise performed by Merz, et al.  (7) on
Los Angeles data.

     The Maryland data used in this analysis was for May through August, 1972
and came from the six AIRMON stations located in the Baltimore-Washington
metropolitan areas.  Four of the stations were located in the Maryland suburbs
of Washington approximately 8 miles from the central business district.  The
other two stations were located in Baltimore City.  One station was  approxi-
mately one mile from the central business district.  The other station was
approximately two miles from the central business district.  Each of the
stations measured the following parameters; nitric oxide (NO), nitrogen dioxide
(N02), methane, total hydrocarbons and ozone.  Due to instrument malfunction
and non-valid data, the total amount of data available for this analysis was
only 350 station-days.  This amount of data was further reduced based on
considerations of station location.  Two stations were located near  major
roadways and the corresponding levels of oxides of nitrogen were significantly
higher than for the other four stations.  One of the remaining stations was
located in a field, far removed from sources of oxides of nitrogen.   This
station was removed from the analysis due to the fact of significantly lower
levels of oxides of nitrogen.  Thus, three stations remained, two in the
Washington suburbs and one in Baltimore City.  The resulting number  of station-
days of data was 159.  Meteorological data for the analysis came from the
National Weather Service office at Baltimore-Washington International Airport.

     Since none of the three stations had sufficient data to do a station
specific analysis, all the available data was used and analyzed collectively.
No consideration was given to atmospheric transport.  The correlations,
also, assumed that measured levels of precursors at a given station  were
related to maximum ozone values measured later in the day at the same station.

     This analysis yielded the following results.  The most significant
variable in predicting ozone concentrations was temperature.  The next most
significant variable was the 6-9 a.m. average N02 concentration.  The least
significant variable was 6-9 a.m. average NMHC.  The following equation
summarizes the best correlation which this analysis of data produced.

  In (peak (03)) - -16.772 + 0.23 In (N02)6_g a^^ + 0.061 In (NMHC)6_9 a>m_ +

                             3.36 In (Max. temp.)

  Where 03, N02 and NMHC - concentration in ppm
        Max. Temp. = Baltimore-Washington International Airport
                     maximum temperature in degrees Fahrenheit.

The multiple correlation coefficient was 0.673 and the percent of the variance
explained by this correlation was 45.1.

     Based upon the correlation analysis, the Bureau concluded that  NMHC
control would have little or no effect on afternoon ozone levels.  From
this analysis it appeared that NO^ emission control would have a greater


                                     1127
-------
effect on afternoon ozone levels.   However,  it was recognized that the cor-
relations did not explain a large portion of the variance and further correl-
ations would be attempted as more data become available.

     This work was forwarded to EPA's National Environmental  Research Center
(NERC) for review and comment with the intention that the relationship derived
might be used in program planning work.   The response was critical of the
conclusions drawn from the study; that is, NERC did not believe that the
study justified controlling N0x instead of NMHC as a means for reducing ozone
levels.  Similar critical review by the California Air Resources Board found
flaws in that the data base was limited, the State did not consider transport
and other problems.

     During the confusion that attended the preparation and submittal of
Implementation Plans for the entire State, the photochemical  aspects of
hydrocarbon emissions were once again examined.  Even though  the mid-1973
submission was based on Appendix J, the State was interested  in the relative
degree of control afforded by taking reactivity into account.  As a result,
during the fall of 1973, the State prepared an inventory of HC emissions
based on reactivity factors for ozone formation.  These factors were being
given preliminary consideration in mid-1972 by the atmospheric chemistry
group in NERC and had been used by EPA in the demonstration project for trans-
portation control strategy development for Los Angeles completed in December,
1972 (8).  Following submission to the Philadelphia regional  office of EPA,
that agency chose not to use reactivity in its analysis and promulgated
transportation control strategies based on the use of Appendix J and NMHC
emissions.  This action immediately followed the same action  in Texas where
the EPA promulgated a plan based on a reactive inventory and  linear rollback.

     In the beginning of 1974 the State challenged EPA over its promulgation
in the fourth circuit court of appeals.   Thus, while the judicial system was
engaged to sort out the program design, the State continued to implement the
non-controversial elements and initiated other studies on the photochemical
phenomenon.  Since the continuous, telemetered air monitoring system had
produced a sizeable quantity of information on the subject, it seemed that
more statistical analysis of aerometric data was in order.  Thus, during the
first half of 1974, the earlier multiple regression analysis  was duplicated
using the more extensive data.  No significant change in the  relationship
of the parameters resulted.

     In addition, a hypothesis was advanced that was intended to shed light
on the EPA's hydrocarbon-only control guideline, Appendix J.   The theory
was advanced that if automotive traffic was significantly less prevalent
on weekends than on weekdays, it should follow that significant reductions
in ozone would also be observed.  Such a study was carried out and the results
reported in the literature (9).  It was found that significantly lower levels
of NMHC were measured on weekends with no commensurate reduction in ozone.
Indeed, it was observed that weekend ozone levels may even be higher than
the weekday measurements, however, the analysis did not demonstrate this
point.
                                    1128
-------
     During the early summer of 1974, the State took an opportunity to analyze
new aerometric data gathered at a rural site in Southern Maryland.   Although
a limited amount was available, wind directions were correlated with ozone and
precursors along with corresponding data in Washington, Baltimore and the
Tidewater, Virginia area.  For the first time the state had first-hand exper-
ience with rural ozone levels and indications at that time were that transport
of ozone was occurring from areas outside the State boundaries.  This obser-
vation was qualitatively compared with work performed for EPA by the Research
Triangle Institute which had measured ozone and precursor levels at McHenry,
Maryland in the westernmost portion of the State (10).

     Following the above work, a careful scrutiny of Appendix J revealed a
flaw that had not been mentioned by others.  Regardless of the lack of a
cause-effect relationship, the so-called "observational" approach from which
Appendix J was taken still showed that for given levels of hydrocarbons,
many ozone levels were observed.  Thus, the fundamental equation between the
two variables was an inequality with an upper boundary observed.  Thus, the
fundamental equation between the two variables was an inequality with an upper
boundary observed on the empirical diagram.  This boundary, slightly modified,
was transformed linearly by subtracting 0.24 from each hydrocarbon level and
dividing the difference by that hydrocarbon level.   Thus, the percent reduc-
tion necessary to achieve the Federal 6-9 a.m. NMHC guideline was given as a
function of peak afternoon ozone measurements.  The mathematical flaw is that
an equality was presented as the Appendix J guideline where an inequality
should have followed the transformation.  Correctly presented, the label for
the abscissae should have read Minimum Reduction in Hydrocarbon Emissions
Required to Achieve National StandarcTfor Photochemical Oxidant.  Because
of this error, the State of Maryland (and possibly-a number of other States)
were led to believe that a lesser degree of NMHC control was necessary in
order to achieve the national standard for photochemical oxidants.   Behind
Appendix J, after this mistake was realized, the goal remained the same as
that used during the first exercise in 1970, that is, the National  guideline
of 0.24 ppm 6-9 a.m. NMHC.

Program Revision Beginning in 1974

     In mid-year 1974, the State decided that not enough effort was being
devoted to understanding the basic state-of-the-art photochemistry of air
pollution.  As a result, a staff reorganization was ordered and two scientists
were assigned to work full-time on the problem.  The first task was to insti-
tute a very extensive literature survey of both theoretical and practical
aspects of photochemistry literature with the overall objective to determine
if the majority of observations were consistent among each other.  Up until
that time the results of State efforts seemed contradictory to the path of
National policies.

     One year later, the literature survey was essentially completed with a
summary paper having been written.  This paper was intended as an educational
exercise for members of staff routinely dealing with hydrocarbon abatement
activities.  It also was to serve as a base which staff can update yearly
to make sure that the State remains well aware of all latest studies.
                                     1129
-------
      It was  discovered also that when  taken  in  perspective  the  State  was
 not yet ready for the analysis  technique  being  highly  touted  by some,  that  is,
 the general  photochemical  oxidant simulation model.  These  models  which
 incorporate  both atmospheric diffusion processes  and atmospheric photochemis-
 try require  more input data than the  State  can  adequately provide  at  this
 time.   Therefore, while these models  provide a  very  useful  learning exer-
 cise,  the quantitative results  of a detailed strategy  analysis  would  be  of
 uncertain accuracy.   This  opinion was  based  on  both  the  paucity of reliable
 input  data which the State could provide  and the  reported analysis of the
 performance  of the models.

      The most reasonable second generation  planning  model appeared to be
 the use of the smog chamber simulation data  once  demonstrations were  avail-
 able that were carried out over the full  range  of ambient precursor concen-
 trations (11).  In order that these chamber  experiments  could be fully uti-
 lized  by the State, however, the agency personnel  had  to feel fully confident
 that State air quality would behave similarly.   In February of 1976,  there-
 fore,  the University  of Maryland was  engaged to  provide a  smog chamber
 simulator.  The main objective  is to  provide the  simulation tool such that
 State  personnel can evaluate the reactivities and dynamics  of Maryland-based
 emissions inventories and hopefully gain  "hands-on"  experience  with photo-
 chemical reactions.   It was also desired  that the feasibility of the  hybrid-
 analog computer be evaluated for these simulations since many linked  equa-
 tions  appear in the set of chemical kinetics under study (12).   The objective
 is to  provide inexpensive simulation  from short run  times,  rapid turnaround
 time and low cost facilities.  Available  digital  programs will  be  utilized
 if the use of the hybrid does not perform adequately.   In the meanwhile,
 the State has begun to collect  a detailed emissions  inventory to use  for in-
 put.  Emission questionnaires have been sent to approximately 80 stationary
 sources of organic emissions.  These  sources account for approximately 25%
 of the total anthropogenic emissions  of organic compounds in  the Baltimore
 metropolitan area.  The information which the State  is requesting  includes
 operating schedules in terms of shifts per  day, days  per week,  weeks  per
 year and information on percent of yearly production by  month.   Emissions
 information  is also being provided by the stationary sources  in terms of
 composition  of organic emissions.  This information  is in addition to the
 existing, non-species specific  organic emission inventory.   This data will
 provide a basis for future modelling  efforts.  Naturally, since this  is  the
 first  attempt in the State at compiling such an inventory,  problems are
 occurring.  However, it is envisioned that  this will  be  a continuing  program
 and the quality and quantity of the information will  improve  as time  progresses.

Observations                                                            _

     With the changes initiated by Congress  in the 1970 Clean Air Act  amend-
ments and subsequently promulgated "guidelines", there began a series  of
rapidly fluctuating program design changes.   The major changes were from a
program design based on observations of non-methane hydrocarbons during the
6-9 a.m. period to an oxidant control  strategy based on observed levels of
ozone.   With  the newly discovered levels of ozone measured using EPA approved
methodology,  the program design changed once again.  Much more restrictive
program design requirements were expected since the 1970 amendments provided

                                     1130
-------
for an auto control program that was not phased in with the ambient air quality
implementation deadlines contained in the Act.  Program design persons were
puzzled when long-term abatement needs disappeared.  Within five years the
hydrocarbon control requirements had made a complete cycle and those abatement
needs are perhaps more restrictive now, than originally contemplated in 1970.
Such a loop was the direct result of a mathematical error contained in the
Federal Appendix J guideline used in evaluating State proposals and promul-
gating State Implementation Plans.

     Throughout those areas of the Nation which were plagued by elevated
ozone levels, there appeared to be different techniques utilized by the EPA
in evaluating proposed plans and promulgating rules to abate such levels.
California, Texas and Colorado were allowed to use an approach based on
linear rollback of a reactive inventory while others, including Maryland,
were not.  No apparent reason for this is known to the State of Maryland.
The use of such a reactive inventory in Maryland would have removed the re-
quirements for those onerous strategies promulgated by the EPA that were
largely the impetus for the State seeking relief from the Federal 4th circuit
court of appeals.

     From the extensive literature survey, it was found that almost all of
the observations reported in the literature are consistent with one another.
The conclusions of some papers, however, were found to have been drawn in-
correctly due to a lack of overall perspective of the state-of-the-art.  Such
was the case with investigations made prior to the end of 1974 by the State
of Maryland.

     An emerging issue appears to be that of the occurrence of ozone levels
far removed from urban areas.  It seems that the extent of the ozone prob-
lem proportionally follows the expansion of the ambient air measurement sys-
tem.  Although there are schools of thought on the origin of those ozone
levels, much controversy remains over whether second or third day reactions are
responsible or the transport of "mature" ozone or even the reaction of surplus
nitrogen oxides with rural emissions of both natural and man-made reactive
hydrocarbon species.

                                 CONCLUSIONS

     Throughout the six year period that Maryland has been developing an
oxidant abatement program, much confusion has attended the cause-effect
relationships which result in observations of elevated levels of ambient
ozone.  Most of the confusion was resolved when it was realized that part-
time efforts to understand the photochemistry of air pollution were not suf-
ficient to understand the state-of-the-art.  It is recommended that States
with ozone problems fund and develop at least one full-time position for an
expert in such matters.  It is also recommended that the Regional offices  of
EPA do likewise.

     In addition there appears to be an emerging phenomenon relating observed
levels of ozone in rural areas to man's activities.  There are several dif-
ferent theories behind the cause of the phenomenon although none of the hypo-
theses have been adequately tested.  From the State level viewpoint, it is


                                     1131
-------
difficult to perceive if a uniform National policy exists and in particular,
if the Congress has been advised to relax nitrogen oxide controls without
regard for the rural problem.  Is it possible that the same confusion which
exists in the State ranks also pervades our national legislators?  Hopefully,
this is not the case.

                                   REFERENCES

1.   Andrew, D. P., Report for the Presentation of Ambient Air Quality Stan-
     dards for:  Carbon Monoxide, Hydrocarbons, Photochemical Oxidants, for
     the National Capital Interstate Air Quality Control Region.  Maryland
     State Department of Health and Mental Hygiene, Division of Air Quality
     Control, October 20, 1970.

2.   Air Quality Criteria for Hydrocarbons.  U.S. Department of Health, Edu-
     cation and Welfare, Public Health Service, National Air Pollution Con-
     trol Administration, Washington, D. C., March 1970.

3.   Requirements for Preparation, Adoption and Submittal of Implementation
     Plans, Vol. 36, No. 158 (Part II), (August 14, 1971  and amendments
     dated October 20,  1971).

4.   Natural Resources  Defense Council, Inc., et al. v. Environmental Pro-
     tection Agency, 475 F 2nd 986 (1973).

5.   Peterson, J. T., Distribution of Sulfur Dioxide over Metropolitan St.
     Louis, as Described by Empirical Eigenvectors, and Its Relation to
     Meteorological Parameters.  Atmospheric Environment 4, 501-518  (1970).

6.   Lebron, F., Paisie, J., Bonta, W. K., Analysis and Evaluation of Balti-
     more-Washington Photochemical Oxidant Data - Summer 1972.  Maryland
     State Department of Health and Mental Hygiene, Bureau of Air Quality
     Control, June 4, 1973.

7.   Merz, F. H., Painter, L. J., Ryason, P. R., Aerometric Data Analysis -
     Time Series Analysis and Forecast and an Atmospheric Smog Diagram.
     Atmospheric Environment 6, 291-342 (1972).
8.   Transportation Control Strategy Development for the Metropolitan Los
     Angeles Region, APTD-1372, U.S. Environmental  Protection Agency, Re-
     search Triangle Park, North Carolina,  December, 1972.

9.   Lebron, F., An Evaluation of Weekend/Weekday Ozone Concentrations in
     the Baltimore-Washington Areas.  Atmospheric Environment 9,  861-863
     (1975).

10.  Investigation of High Ozone Concentration in the Vicinity of Garrett
     County, Maryland and Preston County,  West Virginia, EPA-R4-73-019,  U.S.
     Environmental Protection Agency,  Research Triangle Park, North  Carolina,
     January 1973.
                                    1132
-------
11.   Dimitriades, B., Effects of Hydrocarbon and Nitrogen Oxides  on Photo-
     chemical Smog Formation.  Environmental Science & Technology 6, 253-260
     (1972).

12.   Meet,  T. A.  , Seinfeld,  J.  H.  ,  Dodge,  M.  C.,  Further Development of
     Generalized  Kinetic Mechanism for Photochemical Smog.   Environmental
     Science  & Technology 8,  327-339  (1974).
                                     1133
-------
                                                                           23-3
            OXIDANT  CONTROL UNDER SECTION 110 OF THE CLEAN AIR ACT

                                 J.  L.  Pearson*

ABSTRACT

     Section  110  oi  the Clean Ax>i Act fiequiAeA AtatcA to Aubmit p£cuu that
contain meaAuJieA  designed to attain and maintain the. national. Atanda^td 30 A
photochemical  ox^idant.   The original Stat.e Implementation PlanA weAe Aub-
mitted ,01  1972 and nave been Au.bje.ct to Aevi^ion Aince that time.  The.
approach to oK4.da.nt  control oft the. Env4Aonme.nt.ajL Pnote.cti.on Agency OA well
OA that ofa mo At States  haA been directed towatid con&iol o^ organic compound
e.mtAA4,onA.  A  ^.ecen^: Aeu-tetv ^104 Akovon that many State. J.mple.me.ntat^.on PtanA
aAe .inadequate, to attain and matntatn the. national Atandofid.  WkJJLe. -it 4^>
dt^^icult -in many OACOA to attain and matntatn the Atandatid because o& the.
Lcm4jte.d control tecbu.Qu.e4 ava.itable,,  e.^otit^ aAe continuing to reduce
ambient concentiattonA  ofi photochemical ozidant with a uxLew towand nve.ntu.al
attainment o^  the national Atandatid.

                                   INTRODUCTION

     Section 110  of  the  Clean  Air Act  (1)  of 1970 required submittal  of plans
for the implementation  of control  measures  to reduce emissions of specified
air pollutants  in order  to attain National  Ambient Air Quality Standards,
which were to  be  established by the  U.  S.  Environmental  Protection Agency
(EPA).   The Act required attainment  of the  national  primary, (meaning health-
related), standards  within three years of plan approval  by EPA.  However,
in certain circumstances,  i.e., where  one or more emission sources were
unable to comply  with the  requirements of the implementation plan because
the necessary  technology or other alternatives were  not or would not be avail-
able within the three-year period,  the Administrator could grant a two-year
extension.  In  most  cases  the  date  established for attainment of the national
primary standards was May 31,  1975.

     In addition  to  State  Implementation -Plans (SIP) required by section 110,
the Clean Air  Act also  required a broader control strategy to reduce organic
emissions and  oxidant air quality.   The Federal  program has primarily focused
upon the control  of  new  motor  vehicle  emissions  and  new stationary sources
of organic compounds.   Mobile  sources  are a major contributor to anthropo-
qenic emissions of organic compounds  and thus the U. S.  Congress, in Title II--
Kmissions Standards  for  Moving Sources—of  the Clean Air Act, mandated control
*United States Environmental  Protection  Agency,  Research Triangle Park,
 North Carolina


                                      1135
-------
of new mobile sources.  Specifically, the Act required that by 1978 hydrocar-
bon emissions from light duty vehicles be reduced by 90 percent, based upon
1970 light-duty vehicle emission rates.  Additionally, the Act authorizes the
Administrator of EPA to prescribe standards applicable to emissions from any
class of motor vehicles or motor vehicle engines which in his judgment causes
or contributes to air pollution capable of endangering the public'health or
welfare.  As a result, the Administrator has proposed or promulgated emission
limitations for diesel engines, motorcycles, and aircraft engines.   The Act
also required EPA to consider new stationary sources of air pollution.  Section
111 of the Act requires the Administrator to establish Federal standards of
performance for new stationary sources.  These standards of performance would
apply regardless of ambient concentrations of the specified pollutant.  While
these programs are required under Sections of the Act other than Section 110,
the emission reductions obtained due to these programs are an integral part
to implementation plan development.

                           HISTORY OF PLAN DEVELOPMENT

     On April 30, 1971, the Administrator promulgated in the Code of Federal
Regulations National Ambient Air Quality Standards for six air pollutants.
Included in this promulgation was the national primary standard for photochem-
ical oxidant.  This standard was established as a maximum one-hour ambient
concentration of 160 pg/m3 not to be exceeded more than once per year.  Another
of the National  Ambient Air Quality  Standards promulgated was a national
standard for hydrocarbons.  This standard, 160 yg/m3 maximum three-hour con-
centration (6-9  a.m.) not to be exceeded more than once per year, was  intended
to be used only  as a guide in devising implementation plans to achieve the
oxidant standard.  While implementation plans for oxidant were required,  no
implementation plans, ambient monitoring,  or other actions were required  as
a result of the  hydrocarbon standard.

    The State initiated programs were  included in the SIPs submitted  to
EPA in  1972.  The plans submitted at that time included a  variety of
stationary source control measures.   Regulations for controlling organic
emissions from stationary sources have generally followed  one of two  pat-
terns.  The  two  basic  patterns  are  the Appendix B to the  Part 51 regulations
 (2) and the  Los  Angeles organic  emission  control regulations.  EPA has
recognized deficiencies in both  of  these  approaches  to stationary source
control or organics  and has  undertaken a  task  to examine,  industry by
 industry, the availability of  controls.

    When  the  implementation  plans were submitted  in  1972,  EPA  recognized
that many were deficient  in  their ability to  attain  the national standard
for photochemical oxidant.   As  a result,  in  1973, several  transportation_
control measures  were  promulgated as  part of  the  SIPs  to  help  correct this
deficiency.   Since  that time a  major  concern  to States  in  their  development
of State  Implementation Plans  has been organic emissions  from  mobile  sources.
 Control of emissions  from mobile and  mobile-related  sources  has  been  a two-
 pronged attack.   The  first of  these is designed to  reduce organic emissions
 by the  imposition of controls  on the  source.   Among  these measures  are included:
                                     1136
-------
     t    Inspection/maintenance programs,

     •    Medium and heavy-duty vehicle retrofit,

     •    Ship and barge controls, and

     •    Vapor controls for gasoline marketing.

     The second part of the control program is designed to promote the effi-
cient use of transportation sources.  The following four measures are examples
of these types of regulations:

     •    Transit improvements,

     •    Employer/employee incentives,

     •    Parking management/restrictions, and

     t    Traffic management/restrictions.

                            REVIEW OF EXISTING PLANS

     The Environmental Protection Agency conducted a review, from late 1975
to early 1976, of State Implementation Plans to determine if they were sub-
stantially inadequate to attain and maintain the national standards.   This
review of recent ambient air quality data has shown that a number of Air
Quality Control Regions (AQCR's) had not attained the national  standard for
photochemical oxidant by June 1975 and are not anticipated, even with full
implementation of existing regulations, to attain that standard for several
years hence.  As a result of this review, in July 1976 EPA notified the appro-
priate State Governors and responsible air pollution control agencies of
the need for revisions to the control strategy portion of their implementation
plan.  These notifications of the need for revision were issued for 61 indi-
vidual Air Quality Control Regions and eight entire states without identifica-
tion as to specific Air Quality Control Regions (Figure 1).  In the case of
photochemical oxidant, these notifications included at least one Air Quality
Control Region in 30 states, which compares to the total of 55  states and
territories required to submit plans under Section 110 of the Clean Air Act.
These calls for State Implementation Plan revision do not, however, include
those areas where the implementation plans have recently been revised or are
currently being revised.

     Revised plans must be submitted to EPA for review and approval by July
1977.  These plans must include all measures necessary to provide for attain-
ment of the standard, and where attainment, of the standard cannot be demon-
strated, the plan must include all achievable emission limitations.  The
term "achievable" as used by EPA is intended to require the imposition of
reasonable "technology forcing" control measures, rather than simply requiring
"off the shelf" technology.  All other measures required to attain and main-
tain the National Ambient Air Quality Standard must be submitted by July 1978.
These other measures are intended to include, bjt are not limited to, trans-
portation and land use measures and programs designed to reduce organic con-

                                    1137
-------
                                                                   
                                                                   03
                                                                  ' rs
                                                                   cr
                                                                   CD
                                                                  T3
                                                                   "3
                                                                   10
                                                                   05
                                                                   HI
                                                                  41
                                                                   O)
                                                                  T3
                                                                  c
                                                                  fO
                                                                  I—
                                                                  n_

                                                                  E
                                                                  O
                                                                  •i—
                                                                  +->
                                                                  ro
                                                                  4->
                                                                  C
                                                                  OJ

                                                                  O)

                                                                  CL
                                                                  QJ
                                                                  S-
                                                                  O)
                                                                  01
                                                                  s_
                                                                  OJ
                                          <   9
1138
-------
pound emissions.  It is anticipated that some areas will  be unable to demon-
strate attainment of national  standards.  In such cases,  the submitted plan
should contain a discussion of the State's program to continue to identify
and explore additional  control measures and to adopt such measures when they
are determined to be effective and workable.

                         IMPLEMENTATION PLAN DEVELOPMENT

     During the first stages in the development of an implementation plan
it must be determined which areas have ambient air quality levels that
exceed the national  standard of 160 vg/m3.  At the time of initial implemen-
tation plan development, it was felt that the photo-chemical oxidant problem
was limited to major metropolitan areas.  Since that time EPA and others have
conducted several field studies and determined that ambient concentrations of
photochemical oxidant greater than the standard may be observed even in rural
areas of the country.  While EPA recognizes the transport of photochemical
oxidant (3, 4) and its precursors, it appears that the urban area emissions
are the major contributor to ambient concentrations in excess of the standard
in the vicinity of that urban area.  One of the basic reasons for this is
that a major portion of organic compounds emitted, motor vehicle exhaust and
solvents, are highly reactive and form photochemical oxidant within a short
period of time.  As  a consequence, EPA has generally recommended to the States
that during the current implementation plan development process, control of
emission sources in  the urban area be emphasized.

     The Clean Air Act dictates an air quality management approach.  As a
result, implementation plans must specify the reduction in emissions neces-
sary to attain the National standard and how these reductions will be obtained.
To assess the reductions necessary, a correlation between precursors to photo-
chemical oxidant and maximum oxidant levels is required.   The correlation
generally used for the initial plan development in 1971-1972 was based upon
the available non-methane hydrocarbon and oxidant air quality date (5) from a
limited number of cities.  This relationship, commonly referred to simply as
Appendix "J" because it was published as Appendix "J" to the "Requirements
for Preparation, Adoption, and Submittal of Implementation Plan,"  (6), has
been severely criticized.  Among the criticisms are the following points:

     t    The relationship is dominant at the upper bounds by data from Los
          Angeles, California,

     •    The  relationship does  not account for  the  effects  of  oxides  of
          nitrogen,

     •    The relationship does not account for the effects  of  transport, and

     •    The effect of background levels of oxidant due to  natural sources
          is not included.

     In an effort to develop a more comprehensive relationship, the Environ-
mental Protection Agency is continuing to study the correlation of maximum
oxidant levels in relation to these and other factors.  The  EPA is currently
attempting to develop several alternatives  to the original correlation that


                                    1139
-------
can be used by State and local air pollution agencies to determine the neces-
sary organic emission reductions to be used.

     For many areas of the U. S. where the implementation plan has been
identified as inadequate to attain the standard, these correlations indicate
a requirement for organic emission control greater than is available with
current control technology.  As a result, many Air Quality Control Regions
in the U.S. may not be able to demonstrate attainment of the national photo-
chemical oxidant standard in the near future.  Where this is the case, the
submitted implementation plan must contain requirements to implement all rea-
sonably achievable control technology for both mobile and stationary sources
and must indicate areas where further investigation and study will be conducted
to determine other mechanisms by which the oxidant standard may be attained.
Further, EPA requires that State programs for control Of mobile sources be
coordinated with appropriate state and local planning and transportation agen-
cies to ensure that air quality measures are implemented as part of the trans-
portation planning process.  In addition, the State will be required to adopt,
as they become available and are needed, other regulations that are identified
as representing reasonably achievable control technology.  These procedures
will likely require that the implementation plans submitted in 1977 and 1978
be revised in the future until the degree of control required to attain and
maintain the national standard has been achieved.

     EPA recognizes the important role of oxides of nitrogen in the photo-
chemical process (3, 7, 8).  However, recent studies have indicated that,
except in certain circumstances, i.e., rural areas with low ambient levels
of organics, the most efficient strategy for reducing ambient levels of
photochemical oxidant is through reductions in organic compound emissions.
In addition, EPA recognizes that as ambient levels of photochemical oxidant
and organic compounds are reduced, it may be necessary in the future to consider
the application of controls on sources of oxides of nitrogen emission (3).
It is not anticipated that control of oxides of nitrogen, for the purpose of
attaining the national standard for phof'ochemical oxidant, will be required
in the immediate future.  Additional research is continuing in this area.

     Natural sources of organic emissions are another problem which will have
to be accounted for in the future.  While natural sources of organic emissions
may account for oxidant levels in the range of 0.04 to 0.05 ppm, it is believed
that the contribution of natural sources is generally small in urban areas
in relation to peak oxidant concentrations.  Because EPA has recommended that
implementation plans be developed for urban areas, the impact; of natural
sources is not a major concern in plan development at this time.  However,
various members of the scientific community and EPA are conducting research
to determine the impact of natural sources of organics.

     The Environmental Protection Agency recognizes the difficulties that
will be faced as the U. S. progresses toward attainment of the national photo-
chemical oxidant standard and remains committed to the basic goal of reducing
ambient levels of photochemical oxidant.  The long-range goal is to attain
the national standard by reducing emissions of all organic compounds (except
methane), since virtually all organic compounds eventually participate in the
formation of photochemical oxidant when given sufficient time and ultraviolet


                                     1140
-------
energy.  However, due to the fact that many areas of the country will f\ot be
able to attain the photochemical oxidant standard for several years, EPA
currently requires that regulations requiring substitution of organic compounds
classified as less reactive for those classified as highly reactive be included
in the implementation plan.  This measure will be used only as an interim
measure until sufficient positive control measures for mobile and stationary
sources are available to reduce emissions of organic compounds to a level so
that the national standard will be attained.  The regulations of this type
currently in effect in many areas of the country are generally a modification
of the Los Angeles Rule 66, which was based upon the hydrocarbon reactivity
data available at the time of development.  The EPA is currently assessing the
need to (a) revise this regulation, or (b) develop an entirely new regulation
based upon the most recent reactivity data available.  It should also be
noted that EPA is currently reassessing its policy on the use of reactivity
and its requirement for adoption of substitution regulations.  The Agency is
also continuing to evaluate controls for existing sources and to develop
standards of performance for new sources of organic compound emissions.

                                     SUMMARY

     EPA believes that the combination of motor vehicle controls, transpor-
tation control measures, definition and imposition of achievable control
technology, regulations for substitution of organic compounds, development
of standards of performance for new sources, and other programs will make up
the total package of measures needed to attain and maintain the national
photochemical oxidant standard.  In addition, the U.S. Congress is currently
considering amendment of the Clean Air Act to account for the need for consid-
eration of differing types of control measures, the need to revise the Federal
Motor Vehicle Control Program, and the need for additional time to implement
control measures.  The Congress is considering additional land use and trans-
portation type measures for the control of all auto-related pollutant emissions
and will likely require EPA to provide information on many new and innovative
measures for control of mobile sources and to promote the most efficient use
of transportation systems.  While attainment of the national standard for
photochemical oxidant  in some areas of the U.S. may be difficult, continuing
efforts by industry and local, State, and Federal air pollution control
agencies will ensure that  there will be continuing progress  toward eventual
attainment of the standard.  Even though attainment of the  national  stan-
dard for photochemical oxidant may be several years away, EPA recognizes
that there are many benefits, both to health and welfare, to be gained
by reducing  both maximum oxidant levels and the number of times the  stan-
dard is exceeded.
                                      1141
-------
                               REFERENCES

1.  Clean Air Act (42 U.S.C.  1857 et seq.)  includes  the  Clean  Air  Act  of
    1963 (P.O. 88-206), and amendments  made by  the Motor Vehicle Air
    Pollution Control Act - P.O.  89-272 (October  20,  1965,  the Clean Air
    Act Amendments of 1966 -  P.O. 89-675 (October 15,  1966),  the Air
    Quality Act of 1967 - P.O.  90-148 (November 21,  1967),  the Clean Air
    Act Amendments of 1970 -  P.L. 91-604 -  (December  31, 1970), the
    Comprehensive Health Manpower Training  Act  of 1971 - P.O.  92-157 -
    (November 18, 1971) and the Energy  Supply and Environmental Coordina-
    tion Act of 1974 - P.L. 93-319 - (June  22,  1974).

2.  Code of Federal  Regulations,  Title  40 - Protection of the  Environment,
    Part 51 - Requirements for Preparation, Adoption,  and Submittal of
    Implementation Plans, Appendix B -  Examples of Emission Limitations
    Attainable with Reasonably Available Technology.

3.  Control of Photochemical  Oxidants - Technical Basis  and Implications
    of Recent Findings, July, EPA-450/2-75-005.

4.  Angus, R.M. and E.G. Martinez, "Rural Oxidant and  Oxidant  Transport,"
    October 22, 1975.

5.  Air Quality Criteria for  Hydrocarbons,  March  1970, U.S.  Public Health
    Service, AP-64.

6.  Code of Federal  Regulations,  Title  40 - Protection of Environment, Part
    51 - Requirements for Preparation,  Adoption,  and  Submittal of  Implemen-
    tation Plans, Appendix J  - Required Hydrocarbon  Emission  Control as a
    Function of Photochemical Oxidant Concentration.

7.  Air Quality Criteria for  Nitrogen Oxides, January  1971, U.S.  Public
    Health Service,  AP-84~

8.  Air Quality Criteria for  Photochemical  Oxidants,  March  1970,  U.S.
    Public Health Service, AP-63.
                                   1142
-------
                                                                          23-4
                 OXIDANT  CONTROL  STRATEGY:   RECENT  DEVELOPMENTS

                                B.  Dimitriades*

ABSTRACT
     Recent /teieotch  fai.ncU.ngA -f.eXe.vant to  aMnoAph&iic ox.-ida.nt occuAAence
         and examined ^on. tmpticationA  AegaAdtng  optimum -ittategt/ fan.
ox.tda.nt zontsioi.  ConcJtuAtonA o^eAed  aAe  that the Ae.ce.ftt evidence dee/6
tnvaiidate the Appendix-! method  1301 calculating  ccnfio^ Aequ-ciementA but
doe-i not pAov.-c.de gAound-i ^OA que/sttonxng the  validity o{j kydAocaJibon
ai an approach to ox.tda.nt Aeduction.

                                   INTRODUCTION

     The first oxidant control  strategy was put into effect in the U.S.  in
1970.  Experiences accumulated  since then  by  the  enforcement agencies have
brought to surface some enforcement problems  that were almost invariably
blamed on the non-defensibility of this first control  strategy.   Furthermore,
new research, conducted and reported in the last  3-4 years, raised additional
questions on the validity of the  current strategy and introduced numerous
speculative -- but reasonable --  suggestions  for  revisions or for altogether
new strategies.  Because of the complexity of the problem of oxidant control
and because of the confusing nature of some of the new research  findings it
was judged useful to  review these latter findings and closely examine them
for implications regarding optimum strategy for oxidant control.

                        CURRENT OXIDANT CONTROL STRATEGY

     To help explain  the relevance of  and  implications from the  new research
findings, the current oxidant control  strategy will  first be briefly described
in terms of its key steps and assumptions.

     The basic assumption underlying the current  strategy concerns the phy-
sical and chemical mechanism of the overall process  that starts  with the
discharge of emissions and terminates  with the occurrence of oxidant at
problem-concentrations.  Thus,  it has  been assumed that emissions are dis-
charged in the source areas and react  in-situ to  form oxidant that afflicts
the "receptor" areas,  the receptor areas being the very same source areas.
Consistent with this  assumption,  aerometric data  were gathered from several
urban centers and were analyzed to derive  a quantitative relationship between
HJ.S.  Environmental  Protection Agency, Research Triangle Park, North Carolina
                                     1143
-------
maximum-daily-oxidant concentrations and early-in-the-morning (6-9 a.m.)
non-methane hydrocarbon (NMHC) concentrations; both the oxidarit  (Ox) and
NMHC measurements were made  in the same location, in the core areas of the
urban centers.  This relationship is known as the "upper limit"  0 -HC curve
(Figure  1)  (1).
c/o
O
X
     0.30
     0.25
     0.20
     0.15
     0.10
     0.05
       0

            APPROXIMATE UPPER LIMIT
               OBSERVED OXIDANT
         0
                    V      m
                .• • i  • :•.••: ••••
                                I
0.5         1.0         1.5
       NONMETHANE HC, ppm C
2.0
2.5
  Figure  1.   "Upper limit"  curve relating ambient maximum 1-hour oxidant to
             6-9 a.m.  concentration of non-methane hydrocarbon.
                                   1144
-------
     Additional assumptions made were that all of observed Ox (and NMHC)
arises from anthropogenic sources -- the natural background levels being
zero -- and that ambient NMHC concentrations are proportional to NMHC emis-
sion rates.  These assumptions and the "upper limit" curve were used to de-
rive the Appendix-J curve relating ambient Ox air quality to NMHC emission
control requirements (Figure 2).

     A final key element in the current oxidant control strategy is that the
emission control effort is focused on the urban centers.  The non-urban areas
are exempted from control but not totally; they are subject to the controls
enforced in the form of national emission and new source performance standards.

     The validity of this control strategy has been questioned both because
of the questionable assumptions used and because of suspected ineffective-
ness in reducing rural oxidant.

                 RECENT RESEARCH FINDINGS AND THEIR IMPLICATIONS

     Briefly, recent research and findings can be summarized as follows:

1.   Field studies have revealed the occurrence of a wide-spread Ox problem
     in rural areas (3).

2.   Triggered by the findings on rural oxidant, extensive research has been
     initiated that has focused specifically on:

     •    occurrence of short, mid-, and long range Ox transport (3-7)

     t    meteorology related to Ox transport (6)

     •    chemistry of photochemical oxidant formation under pollutant trans-
          port conditions (8)

     t    role and importance of natural emissions as contributors to tropo-
          spheric oxidant (9)

     •    stratospheric contribution to tropospheric oxidant (10, 11).

3.   Smog chamber studies have resulted in an alternative (to Appendix J)
     method for calculating control requirements for urban Ox reduction (12,
     13, 14).  Key advantages of this latter method are the cause-effect na-
     ture of the chamber-derived 0 -to-precursor relationships, and the abil-
     ity to consider the role of the NO  factor quantitatively; the method,
     however, also has some serious limitations.

4.   Smog chamber and field studies have indicated that the reactivity-related
     approach to control  is less effective than previously thought (15, 16).

     These findings, as well as the extent to which and the manner in which
they can be used to support any immediate or future changes in the control
strategy for Ox reduction, are discussed next.
                                     1145
-------
 E

 CD
 3

  ft
•z.
O
t—i

g

I—
1Z.
LU
O

o
    i±i

                                                                                                     !    O
                                                                                                         o

                                                                                                         o
                                                                                                         nr
                                                                                                         Q-

                                                                                                         o:
                                                                                                         ZD
                                                                                                         O
 co
-I-J
                                                                                                                OJ
                                                                                                                OJ
                                                                                                                S-
                                                                                                                CT
                                                                                                                O)
                                                                                                                S-
o   •
S-  C
•4->  O
C •!-
o •*->
O  fO
 o  c
JD  CD
 S-  (J
 n3  c:
 O  O
 O  O
 i-
T3 4->
 >> C
jr  to
   -a
 OVr-
 C  X
•r-  O
4->
 (O "O
•—  OJ
 a>  v
 S-  3
    I/)
 a>    cu
 s-  E
 X  X
•i-  ro
-O  E
 sr
 01
 Q.
 CL
-------
     The occurrence of elevated (>0.08 ppm) oxidant levels in rural areas
is a fact and need not be discussed further.  What needs to be examined,
however, is the relative magnitudes of the rural and urban oxidant problems
in terms of effects and population exposure.  Both the make-up of the rural
oxidant mixture and the population exposed to rural oxidant are different
relative to the urban case, and it could be that the rural problem is in
effect less severe than suggested by the relative occurrence data.  An appro-
priate assessment here certainly is relevant to the question of how adequate
the current control strategy is, and therefore, needs to be made.

     The studies on oxidant transport have shown conclusively that single-
day, short-range transport does occur and may cause the oxidant problem to
be more severe in the receptor areas (suburbs) than in the source areas
(downtown) (3-7).  The same studies have also provided indirect evidence that
multi-day long-range transport occurs and impacts distant rural (and urban)
areas downwind from the source areas.  This latter evidence, however, is only
qualitative, the degree to which transported pollutants account for the
oxidant observed in the receptor rural (and urban) areas has not been
established.

     The control implications of the findings on short-range transport are
relatively simple.  These findings do not raise questions on the validity
of the current strategy, except for the obvious deduction that unless such
transport is considered the calculated degree of control needed will be too
low.  The relative roles of the HC and NOX precursors under such short range
transport conditions are not substantially different than those under no-
transport conditions.

     Unlike the short-range transport, the control implications of the find-
ings on multi-day, long range transport are extremely complex.  Thus, under
long range transport conditions both the source-receptor relationship and the
relative roles of the HC and NOX precursors are different than those under no-
transport conditions.  The implication related to the source-receptor relation-
ship factor is an obvious one.  It simply means that in areas affected by
transported oxidant, the Appendix-J curve is not valid because the source-
receptor relationship is not the same as the one assumed in the derivation
of the Appendix-J curve.  The implications related to the oxidant-to-precursor
dependency factor are not obvious and cannot be expressed simply in terms of
an overall judgment.  These latter implications merit a more detailed analysis
presented next.

     Present understanding regarding the relative roles of the HC and NOX
precursors is based almost entirely on smog chamber evidence and briefly is
as follows.  The difference in oxidant-to-precursor relationship between
transport and no-transport systems arises from differences in irradiation
time, and in extents of dilution and of interaction with fresh precursors.
Thus, under long range transport conditions the HC-!MOX precursor mixture is
irradiated, diluted and interacted by fresh emissions for several days, as
compared to a few hours only for the no-transport systems.  The combined
effects of prolonged irradiation and dilution upon simulated emission-control-
led and non-controlled atmospheres are illustrated by the data from a 3-day
smog chamber test  (Figure 3) (8).  The pertinent observation to be made here


                                     1147
-------
                                     X
                                   o
                                      CXJ
                                   4-J  S-
                                   (13 •!-
                                   •r- 4-
                                   T3
                                   03  CD
                                   S-  C.
                                   S- T-
                                   fO
                                   •O 4->
                                    I   C
                                   O>  d)
                                   cu  o
                                   i-  S-
                                   JT  O)
                                   -*->  CL
                                   4-  >,
                                      -Q
                                   (S)
                                   C "O
                                   O   -O
                                   E
                                   QJ to
                                   O 03
                                   C 2
                                   o
                                   o cu
                                      S-
                                   O) 3
                                   c: +->
                                   o x
                                   M -i-
                                   o s:
                                   O)
                                   OV
uiddto
                  1148
-------
is that in spite of reaction and dilution, the reacted pollutant mixture was
capable of generating significant concentrations of Ox during the 2nd and 3rd
days of irradiation.
ant oxidant but less
 Obviously, the effect
than proportionately.
       of dilution was to dilute result-
     The effect of hydrocarbon control  is depicted in Figure 4  (17).  It can
be seen that control of hydrocarbons had a drastic effect upon  the first day
oxidant levels and a much  lesser effect upon the second day oxidant levels.
   0.50
   0.25
 C*J
 o
                                     HC   2.90
                  12
       18
  24
TIME OF DAY
12
18
          Figure 4.   Second day 03 formation from "non-controlled" and
                          "controlled" HC-NOX mixtures.
     The effect of fresh precursor influx into the reacting system has not
been determined; it can be speculated reasonably that it would be a complex
function of the absolute as well as the relative levels of the reacted and
fresh precursors.

     The clear deduction from all these observations and expectations is that
under long range pollutant transport conditions the quantitative relation-
ships between oxidant and precorsors are indeed different than those assumed
in the derivation of the Appendix-J method.   Therefore, for areas affected
by multi-day transported oxidant, Appendix-J is not the appropriate method for
calculating numerical control requirements.   Nevertheless, there is no indica-
tion that hydrocarbon control will not benefit such areas.  What is somewhat
in question is the effect of NOx control upon such areas.  This question is
examined in some detail next.
                                     1149
-------
      Results from smog chamber experiments  have  shown  that irradiation  of
 an urban-!ike HC-NO mixture for one  solar day  results  in  a mixture  which  is
 virtually depleted of the reactive organics and  has  a  much higher HC-to-
 NOX ratio (8).   Irradiation for a second day will  further raise  the ratio to
 as high as  100:1  or higher.  At these high  ratios,  the reacted mixtures still
 retain their ability to form oxidant upon irradiation, even though  the  react-
 ant HC is at only a few pphm and the NOX at few  ppb  (8).   Furthermore,  injec-
 tion of small amounts of NOX seems to increase oxidant formation, suggesting
 that the role of NOX in such "aged"  systems, i.e.,  in  urban air  masses
 transported into downwind rural (or  urban)  areas,  is contrary to the NOX
 role in the atmosphere above the source area.  These results may suggest
 at first glance that for reduction of rural oxidant  it will be necessary  to
 control NOX along with HC.   In actuality, however,  such a conclusion may  or
 may not be  justified.  The  reasons for this uncertainity  are as  follows.
 Based on presently available evidence, control of  NOX  will have  two opposing
 effects:

      (1)  It will cause increased oxidant levels during the first day --  a
           part of the increment to be carried  into  the downwind  rural  (or
           urban)  areas during subsequent days.

      (2)  It will result in an aged  mixture containing less NOX  and, there-
           fore, with less potential  for oxidant  formation in the downwind
           rural (or urban)  areas.
      The net result from these two effects is  neither known  or predictable,
 at present.


     In conclusion, once again, the findings from recent research on oxidant
transport, so far at least, have not provided grounds for invalidating HC
control as the optimum approach to oxidant reduction in the source (urban)
areas.  Nevertheless, these findings do substantiate some of the criticisms
against the quantitative bases of the current strategy (i.e., Appendix J),
and raise some questions in regards to the effectiveness of the current
strategy in reducing oxidant in rural areas.  The next logical question is
whether there are any options and what these options are, insofar as the
rural oxidant reduction and replacement of the  Appendix-J method are concern-
ed.  Replacement of the Appendix-J method is- discussed elsewhere (13, 14).
A brief discussion of the rural oxidant problem is given here, next.

     Rural oxidant, conceivably, can arise from several sources, namely:

     •    From natural 03 (from stratosphere and electrical discharge);

     •    From transported (urban) anthropogenic pollutants;

     •    From natural emissions acting alone or in mixture with anthropogenic
          ones;

     •    From local (anthropogenic) emissions  acting alone or in mixture
          with transported ones.


                                      1150
-------
Relevant evidence -- accepted or not -- reported lately on the relative
importance of these sources is summarized as follows:

     (1)  Stratospheric 03 can be transferred to the upper troposphere
          through a "tropopause folding" mechanism (10).   However, the degree
          to which such Q% reaches ground level  has not been established.
          Based on radioactivity measurements, 24-hour-average estimates of
          ground level concentrations of stratospheric 63 are well below the
          NAQS level (11).  No estimates were possible, however, of the maxi-
          mum 1-hour concentrations.

     (2)  Oxidant transport and its potential for rural oxidant buildup have
          been demonstrated repeatedly so that its occurrence is no longer
          in question.  What is in question, however,  is:  (a) the importance
          of this rural oxidant source relative  to the other sources, and (b)
          the dependence of such oxidant on the  transported (presumably urban)
          HC and NOX, and on the locally emitted (rural)  fresh NOX.

     (3)  Some natural hydrocarbons, namely terpenes,  were found to be oxi-
          dant producers when mixed with NOX at  HC-to-NOx ratios between 15:1
          and 25:1 (9).  Such low ratios do not  occur in  rural areas unless
          there are significant sources of anthropogenic  emissions; in this
          latter case, terpenes could conceivably form oxidant.  However,
          the importance of such oxidant in terms of concentration level,
          extent and frequency of occurrence is  far from having been estab-
          lished.  A similar function was proposed for methane also, with
          the suggestion that anthropogenic NOX  would be  necessary for rural
          oxidant reduction (18).  This latter evidence and conclusions,
          however, were rejected by others as erroneous (19, 20).

     (4)  Based on the recent understanding of transport phenomena, one could
          hypothesize reasonably that rural oxidant formation could be caused
          by local anthropogenic emissions in at least two meteorological
          situations:

          •    Under conditions of severe stagnation,  in  which case the local
               emissions alone will enter the photochemical oxidant formation
               process.  Such conditions may occur, e.g., within the center
               of a high pressure cell.

          •    Within the back side of a high pressure cell, in which case
               the observed oxidant would reflect the sum total of urban
               and rural emissions dispersed within the high pressure cell.
               In such a case, obviously, the rural emissions will account
               for a fraction of the observed oxidant.

     The statistics of severe stagnation cases coupled with elevated oxidant
levels have been studied to some extent, and the connection between this
stagnation factor and elevated oxidant concentrations  has indeed been estab-
lished (6).  However, one would still have to consider that the severe stag-
nation conditions occur in a relatively small area only,  within the high pres-
sure cell.  In the severe stagnation case, therefore,  one cannot make a case
for strong effects from rural anthropogenic emissions.

                                    1151
-------
     In the back side of a high pressure cell,  the rural  and urban (anthro-
pogenic) emissions should have relative effects roughly equal  to their relative
levels.  Emission control, as prescribed by the current control  strategy,
will reduce mobile and stationary source emissions in the urban  areas and
mobile source emissions in the rural  areas.  Therefore, insofar  as the rural
problem is concerned, the current control strategy is inadequate only to the
extent that it does not call for control of the stationary source emissions
in the rural areas -- an inadequacy which cannot be serious.

     From the preceding analysis of the rural  oxidant problem and its con-
nection to the current control strategy, it can be concluded that (a) at
present, the rural oxidant problem is not sufficiently well  understood, and,
therefore, it is not possible to define the optimum approach to  rural oxidant
reduction, and (b) the current control  strategy, while it may not represent the
optimum approach, should nevertheless have a beneficial impact upon rural  air
quality.

                                   REFERENCES

1.  U.S. Environmental Protection Agency.  Air Quality Criteria  for Nitrogen
    Oxides.  Air Pollution Control Office Publication No. AP-84, Washington,
    D.C. , January 1971.

2.  Federal Register, 36_,  15486, August 14, 1971.

3.  Research Triangle Institute, "Investigation of Rural  Oxidant Levels as
    Related to Urban Hydrocarbon Control Strategies,"  Environmental
    Protection Agency Publication No. EPA-450/3-74-034, March, (1975).
    Research Triangle Park, N.C.  27711.

4.  Blumenthal, L.L., W.H. White, R.L.  Peace, and T.B. Smith.   Determination
    of the Feasibility of the Long-Range Transport of Ozone or Ozone
    Precursors.  EPA-450/3-74-061, Nov. 1974.   U.S. Environmental Protection
    Agency, Research Triangle Park, N.C.  27711.

5.  Cox, R.A., A.E.J. Eggleton, R.G.  Derwent, J.E. Lovelock, and D.H.
    Pack.   "Long-Range Transport of Photochemical  Ozone in North-Western
    Europe,"  Nature, Vol. 255, No. 5504, pp.  118-121, May 8,  1975.

6.  Research Triangle Institute.  "Study of the Formation and Transport of
    Ambient Oxidants in the Western Gulf Coast  and North-Central and
    North-East Regions of the United  States."   RTI Report pursuant to
    USEPA Contract 68-02-2048, Sept.  1976.   U.S. Environmental Protection
    Agency, Research Triangle Park, N.C.  27711.

7.  Research Triangle Institute.  "Ambient Monitoring Aloft  of Ozone  and
    Precursors in the Vicinity and Downwind of  a Major City  Using a
    Balloon-Borne Platform and Aircraft."  RTI  report on  part  of Project
    DaVinci under USEPA Contract 68-02-2341, 1976.   U.S.  Environmental
    Protection Agency, Research Triangle Park,  N.C.  27711.
                                     1152
-------
 8.  Ripperton, L.A., W.C. Eaton, J.E. Sickles, II, "A Study of the Oxidant-
     Precursor Relations Under Pollutant Transoort Conditions,"  Final
     Report to Environmental Protection Agency, EPA Contract No.  68-02-
     1296, January, (1976).  U.S. Environmental Protection Agency, Research
     Triangle Park, N.C.  27711.

 9.  Westberg, H.H. and R.A. Rasmussen, Washington State University, Pullman,
     Washington.  "Relationships  Between Oxidant Yield and the HC/NOX Ratio
     in Smog Chamber Irradiations of Natural Hydrocarbons."  To be issued as
     an USEPA report, 1976.

10.  Mohnen, V.A., P.E. Coffey, and W.N.Stasiuk.  "Ozone in Rural and
     Urban Areas of New York State.  Part II."  Paper presented at Interna-
     tional Conference on Photochemical Oxidant Pollution and Its Control,
     Raleigh, N.C., September 13-17, 1976.   Proceedings to be published as
     USEPA report.

11.  Reiter, E.R., "The Transport of Radioactive Debris and Ozone from the
     Stratosphere to the Ground,"  subcontract to Stanford Research Inst.
     EPA Contract No. 68-02-2084, November  (1975).  U.S. Environmental
     Protection Agency, Research Triangle Park, N.C.  27711.

12.  Dimitriades, B. "Oxidant Control  Strategies.  Part I.   An Urban Oxidant
     Control Strategy Derived from Existing Smog Chamber Data."  Accepted
     for publication in Envir. Science & Technol.  1976.

13.  Dimitriades, B. "An Alternative to the Appendix-J Method for Calcu-
     lating Oxidant-and N02-Related Control Requirements."  Paper presented
     at the International Conference on Photochemical Oxidant Pollution and
     Its Control, Raleigh, N.C., September  13-17, 1976.  Proceedings to be
     published as USEPA report.

14.  Dodge, M. "The Combined Use of Modeling Techniques and Smog Chamber
     Data to Derive Ozone-Precursor Relationships."  Paper oresented at the
     International Conference on Photochemical Oxidant Pollution and Its
     Control, Raleigh, N.C., September 13-17, 1976.  Proceedings to be
     published at USEPA report.

15.  U.S. Environmental Protection Agency.   "Proceedings of the Solvent
     Reactivity Conference,"  EPA-650/3-74-101 , Nov. 1974, U.S. Environmental
     Protection Agency, Research Triangle Park, N.C. 27711.

16.  Dimitriades, B.  "Application of Reactivity Criteria in Oxidant-Related
     Emission Control  in USA."  Paper  presented at the International  Con-
     ference on Photochemical  Oxidant  Pollution and Its Control,  Raleigh,
     N.C., September 13-17, 1976.  Proceedings to be published as USEPA
     report.

17.  Jeffries, H.E., D.L.  Fox, R.M. Kamens, "Outdoor Smog Chamber Studies:
     Effects of Hydrocarbon Reduction  on Nitrogen Dioxide,"  Environmental
     Protection Agency Publication No.  EPA-650/3-75-011, (1975).   U.S.
     Environmental  Protection  Agency,  Research Triangle Park, N.C.  27711.

                                      1153
-------
18.  Chameides, W., and D.H.  Stedman.  "Ozone  Formation  from  NOX in Clean
     Air,"  Envir.  Science &  Technol. ,  10,  p.  150,  Feb.  1976.
19.   Dimitriades, B.,  M.  Dodge,  J.  Bufalini,  K.  Demerjian, and A.P. Altshuller,
     Envir.  Science &  Technol.,  (in press)  1976.

20.   Weinstock, W., and T.Y.  Chang.  "Methane  and Nonurban Ozone."  Paper
     presented at National  APCA  Mtg.  in  Portland, Oregon, June 29, 1976.
                                    1154
-------
                                                                          23-5
          CONTROL  REGULATIONS  FOR  STATIONARY  SOURCES OF HYDROCARBONS
                             IN  THE  UNITED STATES

                                R.  T.  Walsh*
ABSTRACT

     Stationary *ource*  aac.ou.nt faor over 60  peA.ce.nt o& the. volatLte organic
compound* released to the  atmo*phere o&  the  United State*.   While em-is*ion*
^rom motor ve.hicJLe* are.  beting re.duce.d, hydrocarbon* ^rom *tationary *ource*
one. *til£ increasing in  many kin. Quality Control. Region*.   Recently, the.
Environmental  Protection Agency placed added emphasis  on photochemical oxi-
dant* traceable to stationary source* and it i*  actively engaged i.n *everai
program* aimed at reducing the*e hydrocarbon emi**
-------
   iuAtninA, Auch  OA gaAotine. masike^ing.   Ton. AuAfiace. coating and -some typ&A
   oAgantc ch&nicat manu.{iactLin.t,  ani^oAm n.o.QUitationA may not be pAacticabte..
   ie conAtdeAation may nave,  to  be gac.e. coating opuna-
tiont>, batk gaAoLine. t&untnalA, &klp and baige. &iant>pont o& po.tA.otum product!*,
        AtattonA, pntn.otvjm fLe.^n&u.&> ,  gaAotine. and c^tade o-UL & towage, and
        otkeA. 
-------
         50
         40
 OO fO
 UO O)
 UJ S-
   
-------
fully acceptable, there often has been much disagreement over effectiveness,
cost, and energy requirements of control  technology.   In some instances in
the gasoline marketing industry, control  devices not  only prevent air pollu-
tion but also save product and money for the operator.   Where control mea-
sures don't save enough product to offset control costs, acceptance has been
slower.   At the far end of the spectrum are measures  that require removal  of
organic vapors from dilute exhaust streams, usually with no product recovery
and at a measurable energy penalty.

     While there are dissimilarities in exhaust gases from industrial and
commercial sources, they usually can be categorized by the character and
concentration of the organic(s), presence of other pollutants, and point(s)
of emission.  Major groupings are:

     •    Mixtures too rich to burn (10 to 40 volume  percent hydrocarbon),
          such as those released from gasoline storage and transfer opera-
          tions and some refinery and chemical processes.

     t    Lean mixtures (less than 0.5 volume percent hydrocarbon) often
          contaminated with particulate matter, such  as those released from
          spray booths, baking and curing ovens, and  many chemical manu-
          facturing operations.

     •    Streams containing particulates, carbon monoxide, partially oxi-
          dized hydrocarbons, and/or sulfur compounds, such as those released
          from refinery catalyst regenerators, many chemical manufacturing
          processes, and a few metallurgical furnaces.

     •    Streams containing non-flammable halogenated compounds, such as
          those released from dry cleaning and degreasing operations.

     •    Dispersed, multi-point sources, such as pumps and valves in
          refineries and chemical plants.

     §    Waste disposal sites where solvents or other organics--often mixed
          with solids--are landfilled or otherwise discarded.

     The applicability of add-on control  options—incineration, condensation,
absorption, and adsorption—is strongly influenced by stream characteristics.
Exhaust streams containing rich hydrocarbon mixtures  or halogenated solvents
tend to be the most economical to control, often yielding a dividend through
product recovery.  For example, floating roofs on gasoline storage tanks
(FRT's)  usually pay for themselves in a short period.  Vapor recovery at
truck loading terminals is less cost effective than FRT's but recovered prod-
uct can  offset much  or all of the control cost.  However, several terminal
operators have opted for less expensive incineration  systems wherein dis-
placed vapors are burned in flares with no product recovery.

     The simple displacement or vapor balance system  has been employed at
gasoline service stations and bulk plants.  It has also been used in unloading
bulk solvents.  The transferred liquid displaces a nearly equal volume of air/
organic  vapor to the dispensing vessel with any excess vapor venting to the


                                    1158
-------
 atmosphere.  Where  vapor  tight  connections  can  be  used," the system  is effi-
 cient  and  highly  cost  effective.

     Carbon  adsorption  can  be used advantageously  with  halogenated  solvents
 because  of the  value of recovered product as well  as  the non-flammable char-
 acter  of the solvent.   Carbon finds frequent use in recovering chlorinated
 solvents in  the dry cleaning industry  and to some  extent in degreasing.
 However, adsorption is  seldom used with  petroleum  solvents.

     The most costly exhaust gases to  control are  dilute streams where the
 hydrocarbon  level is too  low to  sustain  combustion, has  little recovery
 value, and is frequently  contaminated  with  sticky  particulate.  For these
 lean streams, incineration  at 1200 to  1500°F is usually  effective for air
 pollution  control purposes  but  requires  appreciable energy input in the
 form of  natural gas or  distillate fuel oil.  Condensation and absorption
 are ineffective because hydrocarbon levels  are  too dilute.  Adsorption can
 provide  high removal efficiency  but requires that  particulates be removed
 beforehand to avoid contamination of the adsorbent.


       In some industries—particularly surface coating—the most promising
means of curtailing organic emissions  are process changes that back out much
of the input solvent or sharply reduce the volume of exhaust gases.   Water-
borne, high  solids  and powder coatings have all  been employed to a  limited
degree in  the manufacture of consumer goods.  Where they are used,  organic
emissions  have  been lowered by 85 to 100 percent.   Unfortunately, these repre-
sent well  under 10  percent of the industrial surface coatings in use today.
Acceptance of low solvent paints by the  coating  industry and its customers
has been far less than was predicted a few years ago.   Manufacturers'  reluc-
tance to commit to  untested coatings isn't surprising since the products often
are subject  to severe conditions.  Applications  range from food and beverage
containers to structural and mechanical  products subject to marine  atmosphere.

      Architectural  coatings are one area where  low solvent content coatings
have been  highly successful.  Today, water-borne paints account for about
70 percent of interior and exterior applications.   While these coatings still
contain  about 20 percent organic solvent, they represent reductions of 75 to
90 percent over conventional oil-base paints.

      In light of the diverse control  possibilities for stationary  hydrocarbon
sources, EPA recently directed additional resources to the review and docu-
mentation  of applicable technology.   The aim is  to publish control  technology
guidelines for principal sources and where applicable to promulgate new source
performance  standards and/or national  emission standards for hazardous pollu-
tants.    At present, the bulk of this effort is directed at the control of
oxidant  forming non-methane hydrocarbons.  At the same time, we are looking at
the toxic  effects of organic emissions,  particularly those associated with
chemical manufacturing processes.  If and when it becomes necessary to apply
more stringent controls to specific toxic materials or to organic halogens
that impact  the upper atmosphere, we should be in possession of sufficient
knowledge  to proceed with an effective control  program.
                                     1159
-------
      With regard to toxic organic emissions, our greatest concerns are the
chronic toxicants associated with carcinogenesis, mutagenesis, arid terato-
genesis.  A compilation of available data is nearing completion by the MITRE
Corporation that identifies over 600 organic compounds that are produced in
large enough quantity, are volatile enough., and exhibit known or potential
toxicity characteristics sufficient to warrant further study.

      For oxidant control purposes, the program is aimed at reducing emissions
of all volatile organic pollutants regardless of their photochemical reactivity.
This course of action is based on review of smog chamber studies and appraisal
of commercial solvents.   These studies show very few organic compounds to be
of "low reactivity"; furthermore, most of the latter are unacceptable as com-
mercial solvents.  Thus, solvent substitution strategies based on photochemical
reactivity appear questionable at best.  Nonetheless, this subject is being
further reviewed to determine whether, in fact, some low reactivity solvents
can be employed and whether solvent substitution would suffice as an oxidant
control strategy for isolated areas of the nation where transport phenomena
do not adversely affect downwind receptors.

     EPA is assembling control technology information for a variety of
petroleum-related industries, chemical manufacturing processes, and end-use
solvent industries, as listed in Table 2.  In some instances, a good deal
of information has already been assembled.  Briefly, developments and ex-
pectations are as follows:

     •    Gasoline Storage Tanks - The floating roof continues to be the
          dominant control device for large tankage, but EPA is making a
          renewed effort to evaluate vapor recovery systems.  Most of the
          effort by EPA and the industry is directed at more positive mea-
          surement techniques for emissions from floating roof tanks.  The
          current program should result in more definitive specifications
          for vapor seals and other components of floating roofs that
          affect emissions.

     •    Bulk Gasoline Terminals - Comprehensive tests are being conducted
          to more fully evaluate the effectiveness of vapor recovery systems,
          particularly in areas where service station controls are in effect.
          These data will be compared with the already completed evaluation
          of incineration devices.   Test procedures are being standardized
          and the feasibility of various monitoring devices analyzed.

     •    Bulk Gasoline Plants - Contract studies will soon be completed to
          evaluate the feasibility of both vapor balance and secondary re-
          covery systems at bulk plants.   Of particular importance is the
          economic impact of control regulations on small operators.  Large
          bulk plants are expected to utilize the same types of vapor recovery
          and incineration equipment that are being used at bulk terminals.

     •    Service Stations - Control systems for underground storage tanks
          (Stage I) have been installed in 1976 in many areas of the nation.
          For the most part, they are simple displacement systems that appear
          to provide well over 90 percent vapor recovery when operated pro-


                                     1160
-------
perly.  The effectiveness of existing systems and operating prac-
tices will be evaluated to determine if hardware improvements and/
or operating restrictions are required.

     The vehicle refueling (Stage II) program has proceeded much
more slowly and there is still much uncertainty as to the effective-
ness of the competing vapor balance and vacuum assist systems.  EPA
will repropose Stage II regulations this month and will continue to
evaluate systems installed or being installed in California and a
few other areas of the nation.  Under the revised Stage II regula-
tion, the first EPA mandated control systems would be installed
in 1978.
Ships and Barges - Most of the effort to date has been directed at
the loading of gasoline into tankers in Gulf Coast ports.   In addi-
tion, there is a great deal of interest in crude unloading both on
the Gulf Coast and the Pacific Coast.  Studies conducted by EPA and
the petroleum industry indicate that emissions are not as great as
originally predicted because the ullage space in the cargo tanks is
well below saturation levels.   Nonetheless, operating practices can
greatly influence saturation and the magnitude of emissions during
loading, ballasting of tanks,  and during inerting and certain main-
tenance operations.

Crude Petroleum Production - Studies of offshore drilling operations
were recently contracted.   Initial efforts will  characterize emissions
and review the more obvious control technology for storage, transfer,
valves, pumps, and separators.  A large fraction of emissions from
production operations are believed to be methane, but heavier frac-
tions are also released.   In at least some instances, air pollution
control equipment will recover substantial quantities of LPG
feedstocks.

Petroleum Refining - We plan to complete a guidelines document in
mid-1977 covering blowdown systems, cooling towers, flares, valves,
pumps, and other miscellaneous sources.  Processing units are being
evaluated under a major research and development contract.  It is
anticipated that equipment specifications and operating practices
will be a large part of any process unit regulations.  For those
processes with discrete exhaust streams, emission limits can be
expected.

Chemical Manufacturing - Carbon black and a few other chemical
processes have been evaluated, but EPA's major contract in this
field will be a comprehensive  study to be awarded within the next
few months.  Control measures  in this area are expected to be ex-
tremely source specific.   We are looking to incineration, adsorp-
tion, and absorption techniques as well as innovative process
changes that can reduce the need for pollution control.  Manufac-
ture of the nine chemicals listed in Table 2 are important because
of their toxicity as well  as their impact on photochemical oxidants.
                           1161
-------
     •    Surface Coating - Principal contract studies in this area are nearing
          completion.  We expect the eventual  answer to solvent control to be
          a combination of improved coating materials and better application
          methods.  Incineration and possibly adsorption are expected to sup-
          plement control programs, with incinerator designs to be maximixed
          for purposes of heat recovery and energy economy.

     t    Degreasing - Improvements in design  and operation  are needed to
          minimize losses and improve capture  efficiency at  the degreaser.
          We also expect that significant improvements will  be achieved
          through better operating practices.   Adsorption will continue
          to be the dominant stack gas clean-up technique.

     •    Dry Cleaning - Carbon adsorption is  expected to become standard
          practice for most perchloroethylene  plants.  Even  without control
          regulations, plants using fluorinated solvents will continue to
          maximize vapor recovery because of the cost of the solvent.  The
          future of petroleum solvent dry cleaning is uncertain because of
          related problems of safety, economics, and energy  requirements of
          available control technologies.

     For all of these sources, consideration will be given to the promulgation
of new source performance standards (NSPS) under Section 111 of the Clean Air
Act or national emission standards for hazardous pollutants  under Section 112.
New source standards are the more likely vehicle.  Hazardous pollutant stan-
dards are specific to the pollutant and apply  to both new and existing installa-
tions.  If the Section 112 route is chosen, emission limits  will be established
for all significant sources of the pollutant.   To date, hazardous pollutant
standards have been proposed for only one organic pollutant, vinyl chloride.
The final regulations are expected to be promulgated shortly.

     Where we are dealing with large identifiable sources of oxidant-forming
organics, NSPS will be the choice.  Thus, bulk gasoline terminals, refinery
process units, major surface coating operations, and many chemical manufac-
turing operations are principal candidates for NSPS.  In these instances,
control technology, costs, and economic and environmental impact will be
documented and the information made available  to interested  parties.

     To cover existing sources, we intend to publish what we term "guidelines
documents".   The purpose of the guidelines is  to apprise State and local
jurisdictions of the effectiveness and costs of available control options.
The documents will recommend no specific emission limits but will leave the
choice of control level to the States.  We anticipate that this approach will
lead to source specific hydrocarbon regulations for many different industries.
The greatest specificity probably will be applied to the chemical manufac-
turing and surface coating industries where both economics and feasibility
can vary with the product and the service to which the product is subjected.
Conversely,  such industries as gasoline marketing, dry cleanirg, and degreas-
ing are well standardized; here we look for relatively un^fonr regulations,
although degrees of stringency are expected depending on tne desired hydro-
carbon reduction needed in the particular area.
                                     1162
-------
     The first guidelines document was prepared for five surface coating
industries — can manufacturing, fabric coating, paper coating, coil coating,
and automobile assembly plants.   A draft document was distributed in July
to control agencies, affected industries, and other interested parties.
The finalized document is scheduled for release in November.   It will re-
flect the large number of comments we have received on the draft document.

     In summary, to combat the oxidant problem and a growing  toxic organic
chemical problem, EPA plans to enact a number of new source performance
standards and a few hazardous pollutant standards for sources of organic
pollutants.   State regulations will require increased control of organic
emissions from existing stationary sources.  All of these are likely to be
in the form of positive reductions rather than substitution of less reactive
solvents.  Source specific guideline documents will provide information
needed by the States to make decisions on emission limits, costs, and eco-
nomic and environmental impact.

     It is clear that a nationwide cleanup of organics is beginning and that
it will have a major impact on almost all industries that use, handle, or
produce organic materials.

          TABLE  1.   ANTHROPOGENIC  SOURCES  OF  ORGANIC AIR  POLLUTION
                          IN  THE UNITED  STATES,  1975	

                                         Emissions,      Fraction  of Total

                                        10  tons/yr         Emissions,  cl


     Transportation                        11.7                37.6

     Fuel Combustion                       1.4                4.5

     Forest, Agricultural
        and  Other Open  Burning              1.1                3.5

     Solid  Waste Disposal                  0.9                2.9

     Petroleum  Refining                    0.9                2.9

     Oil  and Gas Production
        and  Marketing                       4.2                13.5

     Chemical Manufacturing                1.6                5.1

     Metals Manufacturing                  0.2                0.6

     Other  Manufacturing                   0.8                2.6

     Organic Solvent Usage                 8.3                26.7

                                           31.1                99.9
                                     1163
-------
           TABLE 1A.  CHARACTERISTICS OF INDUSTRIAL EXHAUST STREAMS
                    CONTAINING VOLATILE ORGANIC POLLUTANTS


•     Mixtures too rich to burn - gasoline storage and transfer,  petroleum
      and chemical  processes.

•     Mixtures too lean to burn - surface coating, baking and curing ovens,
      plastics and rubber product manufacture.

•     Streams containing particulates, CO, HC,  SOX, H2S and partially oxidized
      organics - refinery, chemical  and metallurgical  processes.

§     Streams containing halogenated organics - dry cleaning and  degreasing.

•     Multipoint Sources - pumps, valves, blowdown system, cooling towers
      in refineries and chemical  plants.

t     Waste disposal sites for solvents, chemical  and  petroleum plant wastes.
                                     1164
-------
TABLE 2.  SOURCES OF ORGANIC AIR POLLUTANTS BEING EVALUATED

PETROLEUM RELATED                      SOLVENT USE
STORAGE TANKS                          SURFACE COATING
BULK TERMINALS                         DRY CLEANING
BULK PLANTS                            DECREASING
SERVICE STATIONS                       RUBBER PRODUCT MFG.
  VEHICLE FUELING                      PHARMACEUTICAL MFG.
  AND STORAGE TANKS                    ADHESIVE MFG.
SHIP AND BARGE TRANSFER                WOOD PRODUCT MFG.
PETROLEUM PRODUCTION                   PAINT MFG.
PETROLEUM REFINING
  PROCESSING EQUIPMENT
  FUGITIVE EMISSIONS FROM
  PUMPS, VALVES, ETC.
                  CHEMICAL MANUFACTURING
                  BENZENE
                  ACRYLONITRILE
                  CARBON TETRACHLORIDE
                  FORMALDEHYDE
                  MALEIC ANHYDRIDE
                  NITROBENZENE
                  PERCHLOROETHYLENE
                  TRICHLOROETHYLENE
                  VINYLIDENE CHLORIDE
                         1165
-------
                                                                     ADDENDUM
Prepared Comments cm Japanese Photochemical Air Quality Standards and Control
Strategies, by Professor Raisaku Kiyoura, Tokyo Institute of Technology.

     Following are some comments of mine and questions on the subject standards.
In the opinion of several biologists concerning the relative toxicities of
oxidants and N02, the adverse effects of oxidants are 10 to 20 times greater
than those of N02.  Therefore, control of oxidants is essential in air pollu-
tion control strategies.  Until now, N02 has not been controlled as stringently
as the hydrocarbon precursor of photochemical oxidants.  However, some people
believe that N02 may also have to be stringently controlled to achieve the air
quality standard for oxidant.

     The oxidant standards in Japan and U.S. have been set at the same level:
0.08 ppm (1 hour)*.  The U.S. N02 standard was set at 0.05 ppm for annual
average (equivalent to a 24-hour average of 0.14 ppm), and some biologists
feel that one-hour-maximum standard for N02 is also necessary.  The Japanese
N02 standard of 0.02 ppm for 24-hour average is five to seven times stricter
than that of the U.S., but there is some uncertainty whether this difference
is justified by biomedical evidence.

     Very stringent control of N02 without similarly stringent control of non-
methane hydrocarbons could result in increased oxidant formation.  Therefore,
these pollutant emission levels must be kept in balance consistent both with
the biomedical justification of the N02 and oxidant standards and with the
relative roles of the NO  and hydrocarbon precursors in the photochemical
oxidant formation process.
     Next, I submit that the following four questions need to be answered
with respect to the air pollution situation in Japan:

     1.   Under such a low Japanese N0> standard, can Japan reduce oxidant
          formation by controlling HC?

     2.   If so, what is the ambient HC level in ppm that would need to be
          achieved?
*The Japanese standard for oxidant was originally set at 0.06 ppm but was
subsequently changed to 0.08 ppm for reasons related to the analytical mea-
surement method.

                                     1167
-------
     3.   What is the relationship between the HC level and the natural
          background level?

     4.   Is that HC level attainable?

     A guide level limiting nonmethane HC between 0.20 and 0.31 ppmC (from
6 to 9 a.m.) was recommended in August 1976 by an expert committee of the
Japanese Central Council for Control of Environmental Pollution.
                                    1168
-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse bejore completing)
 1 REPORT NO
  EPA-600/3-77-001b
                                                            3 RECIPIENT'S ACCESSION-NO.
4 TITLE AND SUBTITLE
  INTERNATIONAL CONFERENCE  ON PHOTOCHEMICAL OXIDANT
  POLLUTION AND ITS CONTROL
  Proceedings:  Volume II                        	
                6. PERFORMING ORGANIZATION CODE
                5. REPORT DATE
                  January  1977
 7 AUTHOR(S)
                                                           8. PERFORMING ORGANIZATION REPORT NO.
  Basil  Dimj.triades, Editor
9 PERFORMING ORGANIZATION NAME AND ADDRESS
  Environmental Sciences Research Laboratory
  Office  of Research and Development
  U.S.  Environmental Protection Agency
  Research Triangle Park, N.C.   27711
                10. PROGRAM ELEMENT NO.
                  1AA603
                11. CONTRACT/GRANT NO.
 12. SPONSORING AGENCY NAME AND ADDRESS
  Environmental Sciences Research Laboratory
  Office  of Research and Development
  U.S.  Environmental Protection Agency
  Research Triangle Park, N.C.   27711
                13. TYPE OF REPORT AND PERIOD COVERED
                  In-house                	
                14. SPONSORING AGENCY CODE

                  EPA-ORD
 15. SUPPLEMENTARY NOTES
  This  Conference was sponsored by the US-Environmental Protection Agency
  with  the patronage of the  Organization for Economic Cooperation and Development.
 16 ABSTRACT

        The  proceedings consist  of 97 technical papers covering such areas  as
  analytical  methods for photochemical oxidants and precursors; causes of  urban,
  suburban, and non-urban oxidant; biological effects; oxidant control strategies;
  and  trends  in emissions and emission control technology.   The International
  Conference  was held in Raleigh, N.C. in September 1976.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS

  Mir pollution
  *Ozone
  * Photochemical reactions
                                              b.IDENTIFIERS/OPEN ENDED TERMS  C  COSATI Field/Group
                                13B
                                07B
                                07 E
13. DISTRIBUTION STATEMENT

  RELEASE  TO  PUBLIC
   19 SECURITY CLASS (This Report/
      UNCLASSIFIED
21. NO. OF PAGES
      620
   20. SECURITY CLASS (This page)
      UNCLASSIFIED
                              22 PRICE
EPA Form 2220-1 (9-73)
1169
-------An error occurred while trying to OCR this image.
-------