DECEMBER 1973
             Prepared for
ENVIRONMENTAL PROTECTION AGENCY
              REGION  IX
  SAN  FRANCISCO, CALIFORNIA 94111
                  ENVIRONMENTAL
                  SERVICES

-------
     This report was furnished to the Environmental Protection Agency by
TRW Transportation and Environmental Operations in fulfillment of Contract
Number 68-02-0048.  The contents of this report are reproduced herein as
received from the contractor.  The opinions, findings, and conclusions
are those of TRW and not necessarily those of the Environmental Protec-
tion Agency.  Mention of company or product names does not constitute
endorsement by the Environmental Protection Agency.
                                      t  '

-------
                            TABLE OF CONTENTS


                                                                  Page


1.0  SUMMARY AND INTRODUCTION   	       1

2.0  BACKGROUND INFORMATION  	       9

     2.1   General  Contingency  Plan Elements    	      10

          2.1.1  Episode Criteria 	      13

          2.1.2  Control Strategies  	      15

          2.1.3  Surveillance  Procedures    	      17

          2.1.4  Operational  Procedures  	      17

     2.2  Description of the Los  Angeles  Problem	      18

          2.2.1  Geography,  Demography,  and  Climate ....      18

          2.2.2  Overall Air Quality Levels    	      27

          2.2.3  Air Pollution Control Agencies   	      31

          2.2.4  Emission Source  Inventory 	      34

          2.2.5  The Potential  for Air Pollution Episodes  .   .      37

          2.2.6  Transportation Data Base	      47

          2.2.7  An Approximate Air Quality-Emission
                   Level Relationship	      53

     2.3  Special  Contingency  Plan Considerations  for the
            Los Angeles  Area	      60

References - Chapter 2	      65

3.0  REVIEW AND EVALUATION OF  EXISTING EMERGENCY EPISODE
       PLANS	      67

     3.1   Local Air Pollution  Control District Emergency
            Plans	      68

          3.1.1  Plan Description .     	      68

          3.1.2  Evaluation:   Preventing  Significant  Harm
                              Levels	      76
                                   11

-------
                                                                 Page
     3.2  Air Resources Board -  Air Pollution  Emergency
            Contingency Plan	79

          3.2.1   Plan Description	79

          3.2.2   Evaluation:  Preventing Significant  Harm
                   Levels	88

     3.3  EPA Region IX Traffic  Abatement Strategies   ....    92

          3.3.1   Plan Description	92

          3.3.2   Evaluation:  Alternative Traffic  Abatement
                   Measures	95

References - Chapter 3	98

4.0  ALTERNATIVE CONTINGENCY PLAN ELEMENTS   	    99

     4.1  Episode Criteria 	    99

          4.1.1   Number of Episode Stages	99

          4.1.2   Feedback  vs.  Forecasting Criteria .....   100

          4.1.3   Air Pollution Forecasting Capabilities  for  the
                   Los Angeles Basin	102

          4.1.4   A Basic Dilemma	106

     4.2  Control Strategies  	   107

          4.2.1   Degree of Emission Reduction  Required    .   .   .   108

          4.2.2   Locational Considerations   	   110

          4.2.3   Alternative Control  Measures    	   Ill

     4.3  Surveillance Procedures   	   113

          4.3.1   Surveillance of Air Quality  and Meteorology   .   113

          4.3.2   Surveillance of Strategy Implementation .   .   .   121

     4.4  Operational  Procedures 	   125

          4.4.1   Legal  Considerations  	   125

          4.4.2   Implementation  Organization  	   131

-------
                                                                  Page


          4.4.3  Communication Channels  	    138

          4.4.4  Activity Phasing  	    140

References   Chapter 4 	    145

5.0  SUMMARY OF ALTERNATIVE CONTROL MEASURE EVALUATIONS ...    147

     5.1  Evaluation of Stationary Source Episode Control
            Measures	    149

     5.2  Evaluation of Traffic Abatement Episode Control
            Measures	    156

          5.2.1  Summary of Traffic Abatement Control
                   Measures	    185

          5.2.2  Recommended Control  Measures for Inclusion in
                   Control  Plan	    186

References - Chapter 5	    189

6.0  A PROPOSED EPISODE CONTINGENCY PLAN  	    190

     6.1  Recommended Episode Criteria    	    191

     6.2  Recommended Control  Strategies  	    194

          6.2.1  Control  Strategy Applicability    	    198

          6.2.2  Control  Strategy Effectiveness    	    199

          6.2.3  Implementation and Enforcement    	    199

          6.2.4  Institutional  Problems and Socio-Economic
                   Impact	    204

     6.3  Recommended Surveillance Procedures  	    206

     6.4  Recommended Operational  Procedures   	    207

     6.5  Limitations of the Analysis	    208

APPENDICES

     A.   EVALUATION OF INTERIM EMERGENCY ABATEMENT ACTIONS   .    A-1

          A.I   The Interim  Emergency  Abatement Actions   ...    A-l

          A.2   Survey of Agency Participation	    A-8

                                   iv

-------
                                                                 Page


     A.3  Survey of Agency Employee Participation  	  A-13

     A.4  State and Local  Actions	A-28

     A.5  Impact of Abatement Action on Vehicle Mileage
            Travelled (VMT)   	  A-30

     A.6  Public Reaction and Media Involvement 	  A-37

References - Appendix A	A-49

B-.   THE PRELIMINARY EPA EPISODE CONTINGENCY PLAN FOR THE
       METROPOLITAN LOS ANGELES AQCR	B-l

     B.I  Effectiveness Evaluation of the Preliminary EPA
            Contingency Plan	B-4

     B.2  Comparison of the EPA Plan to the Plan Proposed
            by this Study	B-20

References - Appendix B	B-29

C.   OXIDANT EPISODES IN THE METROPOLITAN LOS ANGELES AIR
       QUALITY CONTROL REGION 	  C-l

D.   CALCULATION OF VMT REDUCTIONS ASSOCIATED WITH VARIOUS
       TRAFFIC ABATEMENT MEASURES	  .  .   .  D-l

References - Appendix D	"	D-21

-------
                              LIST OF TABLES
                                                                      Page

2.1    Population and  Land  Area  -  Metropolitan Los  Angeles  AQCR 	  19

2.2    Population Projections:   Counties  Within the Metropolitan
      Los Angeles AQCR	  21

2.3    Employment in the  Metropolitan  Los Angeles AQCR	23

2.4    Air Quality at  Selected Sites in the  Metropolitan
      Los Angeles Air Quality Control Region, 1972 .....  	  29

2.5    An Approximate  Emission Inventory  for the Metropolitan
      Los Angeles AQCR,  1973 -  1975	35

2.6    Alternative Hydrocarbon Reactivity Assumptions  	  36

2.7    Federal  Significant  Harm  Levels for Oxidant,
      Carbon Monoxide, and Nitrogen Dioxide	37

2.8   Comparison of Maximum Carbon Monoxide Levels in the
      Metropolitan Los Angeles  AQCR to the  Federal Significant
      Harm Levels, 1970  through 1972	40

2.9    Comparison of Maximum Nitrogen  Dioxide Levels in the
      Metropolitan Los Angeles  AQCR to the  Federal Significant
      Harm Levels, 1970  through 1972	.  .  ,	42

2.10  Oxidant Episodes Exceeding  Federal Significant  Harm  Levels
      in the Metropolitan  Los Angeles Air Quality  Control
      Region (1-70 -  6-73) 	  44

2.11  Severe Oxidant  Episodes in  the  Metropolitan
      Los Angeles Air Quality Control Region 	  45

2.12  Average Weekday VMT  in the  LARTS Region - Total,
      Freeway, Non-Freeway, and Rush-Hour	49

2.13  Distribution of Vehicle Trips and  Total VMT  by
      Trip Type:  Weekdays and  Weekends	'50

2.14  Average Weekday vs.  Weekend Total  Trips and  VMT; 1973	51

2.15  Comparison of Weekday - Weekend VMT According to Trip Type ....  52

2.16  Current Automobile Occupancy Rates in the South Coast Basin.  ...  53

2.17  Average Weekend Day  and Weekday Max.  Oxidant:
      Four Stations,  7-62  to 6-64	.56

2.18  Average Daily Maximal One Hour  Carbon Monoxide at
      Three Stations, Weekends  vs. Weekdays:
      December - February, 1969-1972	 ....  57

                                     vi

-------
                               LIST OF TABLES (COND'T)                p
2.19  Average Daily Maximal One Hour Oxidant at
      Five Stations, Weekends vs.  Weekdays:
      June - September, 1970-1972 ....................  57

2.20  The Effect of VMT Reductions on Total Emission Levels
      (For Base Year:  1974) ......................  58

2.21  The Approximate Effect of Contingency Plan Emission
      and VMT Reductions on Air Quality .................  59

3.1   L. A. County APCD Alert Stages for Oxidant,
      Carbon Monoxide, and Nitrogen Oxides ...............  69

3.2   Types of Stationary Sources Requiring Shutdown Plans -
      Los Angeles County APCD ......................  71

3.3   The Effect of Los Angeles County Shutdown Procedures
      for Stationary Sources ......................  72

3.4   Comparison of L. A. County APCD Alert Levels to the
      Federal Significant Harm Levels ..................  77

3.5   ARB Contingency Plan Episode Criteria Levels ...........  80

3.6   ARB Contingency Plan - Example Traffic Abatement Measures .....  83

3.7   OES Evaluation of Traffic Abatement Measures ...........  87

3.8   Comparison of the ARB Contingency Plan Alert Levels to the
      Federal Significant Harm Levels ..................  89

3.9   Summary of EPA - Region IX Evaluations of Alternative
      Traffic Abatement Measures ....................
4.1   Air Pollution Forecasting Capabilities for the
      Metropolitan Los Angeles AQCR ................... 103

4.2   Brief Description of Alternative Control Measures ......... 112

4.3   Sources of Real -Time Air Quality and Meteorological
      Data in the Metropolitan Los Angeles AQCR ............. 114

5.1   Description of the Categories for Control Measure Evaluation .  .   .148

5.2   Evaluation of Stationary Source Episode Control Measures ..... 150

5.3   List of Traffic Abatement Control Measures ............ 157
                                    vi i

-------
                              LIST OF TABLES (COND'T)
                                                                      Page

5.4   Evaluation of Traffic Abatement Episode Control Measures .... ,158

6.1   Episode Criteria for Oxidant	192

6.2   Episode Criteria for CO and N02	193

6.3   Proposed Control Strategies:  VMT Reduction and
      RHC Emission Reduction	  . .196

6.4   Proposed Control Strategies:  VMT Reduction and
      CO and NO  Emission Reductions	197
               A

6.5   Episode Levels Which Can Be Reduced To
      Below Significant Harm Levels By The
      Proposed Control Strategies	200


A-l   Federal Agencies Participating in July 26
      Air Episode Abatement Action	A-2

A-2   Major Federal Agencies Without Plans -
      Participation in July 26 Abatement Action	A-10

A-3   Federal Agencies of the Employee Survey	A-l4

A-4   Federal Employee Questionnaire Response Rate	A-l7

A-5   Usual Mode of Travel - Federal Employees on
      Administrative Leave July 26	A-l9

A-6   Change in Mode of Travel  - Federal Employees
      Which Were Urged to Share Transportation July 26	 .A-20

A-7   Change in Vehicle Mileage Traveled - Employees on
      Leave and Employees Encouraged to Share Transportation	A-21

A-8   Nature of Trips Made by Federal Employees
      on Leave July 26	A-23

A-9   Means by Which Federal Employees Were Notified to
      Take Emergency Smog Abatement Actions	A-25

A-10  Employee Reaction to Emergency Air Episode Plan -
      Based on July 26 Action	A-27

A-ll  Total Daily Vehicle Count
      Selected Stations in "42 Mile Freeway Loop"	A-34

A-12  Morning Rush Hour (6-9 AM) Vehicle Count
      Selected Stations in "42 Mile Freeway Loop"	A-36


                                     viii

-------
                            LIST OF TABLES (COND'T)
                                                                   Page

B-l  Preliminary EPA Contingency Plan for
     Los Angeles - Episode Criteria 	 B-2

B-2  Preliminary EPA Contingency Plan for
     Los Angeles - Control Strategies 	 B-3

B-3a Effectiveness of the Proposed EPA Plan:
     VMT Reductions and RHC Emission Reductions 	 B-5

B-3b Effectiveness of the Proposed EPA Plan:
     VMT Reductions and CO and NOV Emission Reductions	ts-b
                                 X

B-4  Episode Levels Which Can Be Reduced To Below Significant
     Harm Levels By The EPA Control Strategies	B-8

B-5  Comparison of the EPA Control Strategies and the
     Control Strategies Recommended by This Study	 ..•  . B-22

C-l  Oxidant Episodes in the Metropolitan Los Angeles
     Air Quality Control Region 	 C-2

-------
                               LIST  OF  FIGURES
                                                                       Page

2.1  Feedback vs. Feedforward Episode Control Systems ..........  11

2.2  General Contingency Plan Elements for Air Pollution
     Episode Control ..........................  15

2.3  Metropolitan Los Angeles AQCR ...................  20

2.4  1970 Employment in Southern California ...............  22

2.5a Air Flow Patterns - South Coast Air Basin (October 0000-0700 PST)  .  25

2.5b Air Flow Patterns - South Coast Air Basin (October 1200-1800 PST)  .  26

2.6  Comparison of Seasonal Pollution Patterns
     Carbon Monoxide vs. Oxidant ....................  28

2.7  Location of Air Quality Monitoring Stations
     (Metropolitan Los Angeles AQCR - 1970) ...............  30

2.8  Total  RHC,  CO, and NO  Emissions in the Metropolitan
     Los Angeles AQCR Under the Present Control Program .....  ....  38

3.1  County APCD Emergency Episode Operations ..............  74

3.2  ARB Emergency Episode Operations Organization  ...........  85

4.1  Location of Real-Time Sources of Air Quality Data ......... 115

4.2  Los Angeles APCD - Organizational Structure
     for Episode Control ....................... ..132

4.3  Argonne National Laboratory - Organizational Structure
     for Episode Control ........................ 134

4.4  EPA Region  IX - Organizational Structure
     for Episode Control ........................ 137

4.5  Communications Channels as Outlined in
     REOCC  Operations Manual
4.6  Possible Modifications to Existing
     Communications Channels ...................... 142


A-l  Survey Questionnaire for Federal  Employees on
     Administrative Leave July  26 .................... A_1
A- 2  Survey Questionnaire for Federal  Employees of
     Agencies Encouraging Transportation Sharing ............ A-l 6

-------
                         LIST OF FIGURES (COND'T)

                                                                  Page

A-3  "42 Mile Loop"	A-33

A-4  Los Angeles Times Report on Abatement Action,  Appearing
     July 26, 1973	A-40

A-5  Los Angeles Times Report on Abatement Action,  Appearing
     July 27, 1973	A-41

A-6  Editorial from Evening Outlook,  Appearing Saturday,
     July 28, 1973	A-42

A-7  Editorial from Los Angeles Times,  Appearing  Friday,
     July 27, 1973	A-43

A-8  KNXT Editorial Broadcast August 1  and 2,  1973	A-*S

A-9  Report on City Officials' Reaction, Valley News  and
     Green Sheet, Appearing Friday, July 27,  1973    	A-46

A-10 Column from Washington Post, July, 1973   	  A-48

B-l  Hourly Distribution of Recreational Trips in the
     LARTS Study Area	 .  B-27

-------
                    1.0  SUMMARY AND INTRODUCTION

     This report presents the results of a study to develop an air
pollution episode contingency plan for the Metropolitan Los Angeles Air
Quality Control Region.  The study is directed at formulating a plan
which will prevent oxidant, carbon monoxide, and nitrogen dioxide from
reaching the "significant harm levels" established by the Environmental
Protection Agency.  An appropriate plan, consisting of episode initiation
criteria, control strategies, surveillance procedures, and operational
procedures, has been developed.  This proposed plan is presented at the
conclusion of  the report.
     The present chapter summarizes the principal findings, conclusions,
and recommendations which have resulted from the study. The main text of
the report covers three basic areas.  First, background information nec-
essary for formulating a contingency plan is compiled.  Second, existing
emergency contingency plans are reviewed.  Finally, a recommended contin-
gency plan is developed.
     Chapter 2 provides the background information for the project.
General contingency plan requirements are reviewed, and specific informa-
tion relevant to the Los Angeles area is compiled.  The latter includes
regional statistics, air quality data, an emission inventory, a transpor-
tation data base, and an air quality-emission level relationship.
     Chapter 3 reviews and evaluates existing episode control programs.
Included are the county APCD emergency episode programs, the preliminary
California Air Resources Board contingency plan, the State Office of

                                  1

-------
Emergency Services report on traffic control measures, and the EPA Region-IX
Task Force report on traffic abatement during episodes.  A preliminary
episode control plan recently proposed by the EPA Region-IX is reviewed in
Appendix B.  An effectiveness evaluation of the EPA control stategies is
performed, and a detailed comparison is made between the EPA plan and the
plan recommended by this study.
     Chapter 4 discusses the alternatives which are available in
formulating a contingency plan for Los Angeles.  Possible episode
criteria, control strategies, surveillance procedures, and operational
procedures are examined.  A detailed review of alternative control
measures is performed in Chapter 5.   Four stationary source controls
and twenty-four traffic abatement measures are evaluated as to enforce-
ment and implementation requirements, technical effectiveness.,
institutional problems, and socio-economic impact.
     In Chapter 6, on the basis of the discussions  of the previous two
chapters, a complete episode contingency plan is recommended.   The
proposed episode criteria depend on  air quality forecasts rather than
air monitoring feedback.  The recommended control plan consists of
four stages which attain 0%, 16% 39%, and 64% RHC  emission control,
respectively.  The proposed surveillance and operational procedures
rely basically on existing capabilities.
     Four appendices follow the main body of the report.  Appendix A
provides detailed documentation on the results of the EPA actions taken
on 26  July 1973,  when certain federal  agencies instituted their traffic
abatement plans.   Appendix B was described above.  The other two
appendices  contain support material  for the main text.

-------
     The findings, conclusions and recommendations  which  have  resulted  from

this study are listed below.   The ordering of the  list  corresponds  generally

to the organization of the text.


   Findings and Conclusions

        G   The .40 PPM—4 hour average harm level  is the binding
            constraint of the three significant harm levels for
            oxidant.It is exceeded, on the average, around 4 to 5
            times per year in the early 1970's.  The .60 PPM--2 hour
            average and .70 PPM--1 hour average harm levels are very
            rarely exceeded (possibly once every 5 to 10 years).
            Furthermore, if the one or two-hour harm levels are
            exceeded, the four-hour harm level is  virtually certain
            to also be violated.

        o   Violations of the CO or NO- signficant harm levels are
            likely to never occur in tnis decade,  (especially in
            view of decreasing CO and NO  emissions due to ongoing
            control programs).  The only harm levels which are even
            remotely threatened of being exceeded are the eight-hour
            average, 50 PPM CO level  and the twenty-four hour average,
            .50 PPM N02 harm level.  The one and four-hour CO harm
            levels and the one-hour N02 harm level  are far below maximal
            levels now found in the Los Angeles basin.

        e   An analysis of weekend/weekday air quality data and
            emission levels reveals that to attain a given percentage
            reduction in air pollutant concentrations on a certain
            day requires a disproportionately greater reduction in
            emissions for that day.  For primary contaminants, this
            is the result of residual  pollutants remaining from the
            previous day.   For oxidant, this disproportionality
            is even greater due to nonlinearities  in the oxidant
            precusor relation as well  as to residual precursors.

        •   Local  APCD plans have not been designed around federal
            significant harm levels-and are not capable of preventing
            harm levels from being exceeded for oxidant, CO, and
            NO?.   Oust lowering the APCD alert criteria would not
            make their plans adequate for averting federal harm
            levels.   This  is because  the plans are based on feedback
            criteria,  while forecasting criteria are required,
            especially for oxidant.  The APCD emergency plans have
            well-organized operational  procedures  for stationary
            source control,  but the mobile source  program should be
            made more definitive.

        •   The preliminary ARB contingency plan basically consists
            of a guide for APCD's in  reconstructing their own plans.
            The ARB alert  criteria are well  below  present APCD
            criteria but they still  are of the feedback variety.

-------
 (The OES Report recommended forecast criteria,  but
 evidently the ARB has concluded that prediction
 capabilities are inadequate.)   Since the criteria
 are of the feedback type,  oxidant will  very likely
 not be prevented from exceeding harm levels.   The
 MF plan, by leaving jurisdiction basically up  to
 local APCD's, does not explicitly allow for region-
 wide control.  This seems  to neglect the distinct
 regional-transport character of the  oxidant problem.
 Finally, the actual control  programs of the ARB plan
 require further definition (this may be a result of
 the preliminary nature of  the  ARB plan).

 The OES Report gives a comprehensive survey of  traffic
-abatement measures.  It arrives at several  general  con-
 clusions about episode plans which are  in concurrence
 with the present study. A few improvements which are
 possible in the OES analysis of traffic abatement
 measures are discussed in  Chapter 3.

 The EPA Region IX Report on Traffic  Abatement Strategies
 is not a complete contingency  plan but  rather a guide
 for certain elements which should be included in a plan.
 It also conducts a comprehensive analysis of  episode
 controls for traffic.  Possible improvements  in this
 analysis are discussed in  Chapter 3,

 Feedback episode criteria, which call for controls when
 air pollution actually reaches certain  levels,  are not
 adequate for oxidant control.   Forecasting  criteria are
 required because some lead time is needed to  implement
 most traffic abatement measures and  because oxidant
 levels do not respond well to  shutting  off  emissions
 after the early morning.   However, obtaining  reasonably
 accurate forecasts on episode  type days may be  a difficult
 problem.  Measures which may help to increase forecast
 reliability are discussed  in Chapter 4.

 Air monitoring data generated  -in the Los  Angeles Basin
 is very adequate for air quality surveillance during
 episodes.  The main problem in air quality  surveillance
 is obtaining the measured  data on a  real-time basis.

 The air monitoring, enforcement, source surveillance,
 and.other capabilities of  the  local  APCD's  would be
 very valuable in carrying  out  an episode contingency
 plan.   The traffic monitoring  stations  of the State Depart-
 ment of Transportation would be able to provide useful
 data on mobile source activity levels.

-------
The organizational structure of the EPA-Region IX
Regional Emergency Operations Control  Center appears
to be generally adequate to administer the type of
contingency plan recommended by this study.

Of the three types of traffic abatement measures,
(direct controls on traffic, controls  on traffic
support facilities, and controls on incentives to
drive), incentive control  appears to be the  best
form.  Controls on traffic and/or traffic support
facilities are either ineffective or they present
extremely difficult enforcement and implementation
problems.  Controls on incentives (e.g. work holidays)
can be effective and they do not present great enforce-
ment or implementation problems.  However, the one
extremely negative aspect of incentive controls is the
economic disruption involved.,  Nearly  all effective
traffic controls will  rank low as to socio-political
acceptability.

Substantial  controversy has resulted from the July 26,
1973 execution of an Interim Emergency Air Episode
Program in the Los Angeles Basin.  Based on  a survey
study of both agencies and employees participating in
this action (Appendix A),  the following conclusions
seem appropriate:

   1)  The impact of the current plan  (as measured by
       the survey of the agenices and  by an  examination
       of traffic data) is essentially insignificant in
       in the sense of achieving reductions  in total
       vehicle mileage traveled.

   2)  The relatively low level of agency participation
       in the plan stems from its unauthoritative
       execution approach  (voluntary and unilateral
       among the participants).

   3)  Much  of the adverse reaction to the action of the
       26th  was due to a minimal awareness of the interim
       stature of the plan and the mechanisms now in
       work  to adopt a more comprehensive emergency plan
       this  fall.

   4)  Federal employees participating in administrative
       leave plans appear to have made a conscious effort
       to reduce vehicle travel during the abatement
       action, (unlike the typical holiday pattern, no
       substitute driving  was indicated).

-------
         The  recently  proposed  EPA-Region  IX contingency  plan
         provides  for  sufficient emission  reductions  to prevent
         significant harm levels.  The three stages of the plan,
         (voluntary abatement,  certain mandatory controls, and
         a  work  holiday), result in 0%, 22%, and 64%  RHC  emission
         reductions, respectively.  A comparison of the EPA  plan
         to the  plan proposed in this study points to neither one
         being unambiguously superior to the other.
Recommendations
         Consideration should be given to Devaluating the  .40  PPM--
         four-hour  average oxidant harm level.  All other federal
         harm  levels are rarely exceeded (at most once every few
         years).  The present four-hour harm level changes  the  nature
         of the  problem from an emergency plan to be instituted
         very  rarely to an episode control plan which is implemented
         several  times per year.  It should be noted that there may
         be a  public outcry against disruptive control measures
         which are  instituted several times per year to prevent air
         pollution  episodes that have been experienced regularly
         in the  past with no major health problems generally
         perceived  by the public.

         Predictive (forecasting) type episode criteria should be
         used.  They are a must for oxidant control.   Since pre-
         sently available forecasts can contain considerable error
         and since  they may not give results in the appropriate
         form, some resources should be devoted toward improving
         prediction capabilities.  A preliminary forecasting scheme
         based on previous day's air quality measurements is pro-
         posed in Section 6.1.

         Episode controls should be implemented for both stationary
         and vehicular sources.   Gasoline powered motor vehicles
         must  be controlled because they constitute the one major
         source of  RHC,  NO ,  and CO emissions.   Stationary sources,
         although minor contributors, should be controlled because
         curtailment is often much easier to implement than for
         mobile sources.   The control program recommended in Section
         6.2 can be summarized  as follows:

           Stage I   1.  Media  Announcements and Voluntary Abatement
           Stage II  1.  Stage  I
                     2.  Ban  Filling of Underground Service Station
                          Tanks
                     3.  Control Specified Stationary RHC Sources
                     4.  Ban  Non-Critical  Fleet Vehicles
                     5.  Close  Non-Critical Government Agencies
           Stage III 1.  Stage  I
                     2.  Operate on Sunday Status

-------
       Stage IV  1.  Stage I
                 2.  Complete Work Holiday (Including Closing
                      Recreational Facilities)

•   The vehicular episode controls for the advanced stages
    basically consist of controls on incentives to travel.  If
    advanced alerts  will be called very frequently, the economic
    disruption of incentive controls may be intolerable.   In
    such a case, direct traffic controls (say by a "windshield
    sticker" scheme) might be considered as an alternative.

•   Cooperation and  communication between County APCD's,  the
    State ARB, and the Federal  EPA should be promoted.   Each
    type of agency has certain  special  capabilities that  would
    be valuable in operating an effective contingency plan.
    If communication is not maintained, actions by one  agency
    could be duplicated by or even be counterproductive to
    actions of another agency.

e   Existing surveillance capabilities  should be used whenever
    possible.  Local state, and federal air monitoring  sites
    provide adequate air quality data.   More reliable procedures
    should be formalized for obtaining  the aerometric data
    during episodes.  Appropriate arrangements should be  made
    to obtain traffic count data from the California Department
    of Transportation.  Stationary source surveillance  might
    be conducted by  existing, experienced local and state per-
    sonnel, or a special EPA team could be organized.

•   The EPA-Region IX REOCC appears capable of handling the
    necessary operational procedures.  However, a local EPA
    communications center in the Los Angeles Basin during the
    episode would prove useful.

e   Recommendations  evolving from the study (Appendix A)  of
    the interim emergency abatement action executed in  the
    Los Angeles Basin on July 26, 1973  are as follows:

       1)  Deliberate measures  should be instituted to
           shape positive public awareness and public
           opinion with respect to the  intentions and
           objectives of the emergency  air episode program.

       2)  Emergency air episode actions should not be
           instituted when only partial and ineffective
           abatement results are expected.  This type of
           result appears non-constructive in terms of
           heightening any type of public awareness, and
           in fact,  appears to  invite controversy and
           negative  reaction.

-------
3)  The interim emergency plan should be strengthened
    to include a significant number of agencies.
    Abatement measures  should be  well-defined  and a
    forceful  commitment to  execute  the measures  accord'
    ing to  plan should  be made.

-------
                       2.0  BACKGROUND INFORMATION

     This chapter establishes the background information related to the
study.  Section 2.1  presents an introductory overview of general
contingency plan elements, describing briefly how episode criteria,
control strategies,  surveillance procedures and operational  procedures
form the basis for an episode control plan.
     Section 2.2 presents a detailed description of the Los  Angeles air
pollution problem, citing many of its unique characteristics.   Some of
this information is  intended for readers unfamiliar with the Los Angeles
problem.  In addition to reviewing much of the published data for the
region (e.g. population statistics, climatology, and historical  air
quality data),this section provides an updating of certain necessary
data bases (e.g. emission inventories and traffic data).  The potential
for air pollution episodes in the basin is thoroughly analyzed.   Finally,
in section 2.2.7 a very approximate emissions - air quality  relationship
is developed for episode control.  This relationship is derived from an
analysis of weekday-weekend air quality data and emission levels.  To
our knowledge, this  is the first time such a relationship has been
formulated.
     Section 2.3 summarizes the considerations which are most important
in the development of a contingency plan for the Los Angeles basin.

-------
2.1  GENERAL CONTINGENCY PLAN ELEMENTS





     The purpose of an air pollution episode contingency plan is to prevent



damage to public health from excessive short term air pollution levels.



Contingency plan action is necessary when meteorological conditions cause



the concentration of air pollutants to reach levels which could result in



significant harm to residents of an area".  In general, there are two ways



to lessen the health danger:  first, by preventing the pollutants from ever



reaching danger levels, or second, by protecting the public from the effects



of such levels.  Obviously, the second approach would be a monumental  trsk



during severe episodes since it would be necessary to ensure that each and



every citizen were protected.  Thus, practical  considerations require pre-



venting air pollutants from ever reaching significant harm levels.   This



can be accomplished by either controlling emissions sources or by attempt-



ing to disperse contaminants once they have entered the atmosphere.   However,



it seems very unlikely that practical dispersion methods could be devised;



this would amount to changing the weather.   For the present, emission con-



trols appear to be the most effective method for dealing with short-term



episodes.   Appropriately,  the EPA Administrator, in 40CFR 51.16, requires



state implementation plans to "provide for taking emission control  actions



necessary to prevent ambient pollution concentrations from reaching levels



which would constitute imminent and substantial endangerment to health..."




     In general, the basic configuration of an  episode control  plan can



assume one of two principle forms.   The first involves feedback control,



the second is based on forecasting.   The difference between these two types



of control  can  be best understood by viewing the air pollution problem as a



"control  system",  [19].  As shown in Figure 2.1, the air shed is the system,
                                     10

-------
                     FEEDFORWARD
                       CONTROL
                      MECHANISM
    CONTROL VARIABLES:
  Emission Quantities,
    Location & Timing
                         UNCONTROLLABLE INPUTS
                                                              /I                                    \
                                                               METEOROLOGY:   (Wind,  Stability,  Sun, etc.)
AIR SHED SYSTEM
                                                       FEEDBACK
                                                       CONTROL
                                                       MECHANISM
AIR  POLLUTION
CONCENTRATIONS
                                       Figure 2.1  Feedback vs.  Feedforward
                                                  Episode Control  Systems
Source:  Seinfeld and Kyan, Reference [19].

-------
and pollutant concentrations are the basic system variables.  The manipu-
lable inputs of this system are source emissions.  The unmanipulable inputs
are the meteorological variables.   The object of control  policy is to pre-
vent air pollution concentrations  from reaching a "dangerous" level.
     Feedback control is based on  air quality (system variable) measurements;
control mechanisms are activated when pollutant concentrations actually reach
certain levels.  Measures then are instituted to prevent  the concentrations
from going any higher.  However, if the system does not respond quickly
enough, the concentrations might continue to increase and exceed significant
harm levels.  On the other hand, feedforward (forecasting) control is based
on measuring inputs to the system (e.g. meteorology); controls are activated
according to predicted pollution levels.   This allows more time for preven-
tive action.
     The determination of which type of configuration is  best for a particular
air pollution system depends entirely on  the characteristics of that system.
For example, if the air shed responds relatively quickly  to alterations in
emissions, an effective feedback plan can be established  by setting the con-
trol trigger low enough to ensure  that significant harm levels will never be
reached.   However, if the control  system  is  sluggish, then a plan using
feedback control  would have to set trigger levels so low  that emergency con-
centrations  would be declared much more often than necessary.  In such a
case,  forecasting methods should be developed.
     Although  the actual  configuration of an episode contingency plan
depends  principally on the characteristics of the air quality control region
for which  it is  designed, there are several  basic components which are
necessary  to any  episode  plan based on emission control.   These are
                                    12

-------
 1)  episode  criteria, 2) control strategies, (groups of control measures plus



 enforcement and  implementation specifications), 3) survei11 a n c e procedures,



 and 4) operational procedures.  Each of these basic components will be dis-



 cussed below.  Their place in the entire episode plan structure is summarized



 in  Figure 2.2.




 2.1.1  Episode Criteria




     A contingency plan must contain a set of criteria which cause the plan



 to  be put into action.  These are known as episode criteria.  These criteria



 should be organized in stages of severity, more stringent controls being



 activated for more advanced stages.  EPA requirements specify that emergency



 contingency parts of state implementation plans must have at least two stages,



 [40 CFR 51.16(6)].




     The episode criteria can take one of two forms depending on whether the



 contingency plan is of a feedback or forecasting variety.  Feedback criteria



 cause the episode to be initiated once a certain level of pollutant concen-



 tration is reached.  Feedforward criteria cause control action to be initiated



 once it is determined that conditions are such that pollutants will reach or



 exceed significant harm levels.  Feedback criteria are based on air quality



 measurements, while feedforward criteria are usually based on meteorological



 measurements.




     Besides initiating action, episode criteria also establish the severity



 of the episode.  Feedback criteria determine the severity by setting up



 different successive episode levels based on higher degrees of pollutant



 concentration.   Every time one of these levels is reached, a new stage or



 phase of the episode begins.   On the other hand, feedforward criteria deter-



mine severity according to how high uncontrolled pollutant concentrations






                                     13

-------
  CONTROL STRATEGY

(GROUP OF MEASURES)
6U
        LU

        CO
CO
LU
cr

LU
o
        o
        CO
       CO

       CO
    UJ
    CO


    i
    CO
    UJ
    ce:
       3
o


|




CO

o
1—4


g
               o
                  2:
                  LU
                          I
                                  EPISODE CRITERIA


                                     (INITIATION)
                                        OPERATIONAL


                                        PROCEDURES
                                       SURVEILLANCE


                                        PROCEDURES
                                                                :g jaBTMam^MaiigJEa^^aai i^^ajs^-sfc^ j tTS
                                                                    METEOROLOGY
                                                               FEED FORWARD
'fj"iiTffilrlm*-V:y-'p
J***Waj4lL-IHID-^f1'TC!- -V-JjJ^E,
| AIR QUALITY
il
;
•
1
i
                                                                                    i
                                                                                    LU
                                                                                    U,
                                                                                                 UJ

                                                                                                 CO
                                                                                                 CO

                                                                                                 UJ

                                                                                                 OL
                                Figure 2.2  General  Contingency  Plan Elements

-------
are expected to be.  This predicted magnitude stipulates the episode stage.




     As mentioned above, the actual plan configuration and the type of episode



criteria to be used depend upon the nature of the specific air basin under



consideration.  However, it is possible to note some of the difficulties



which the two types of criteria encounter in general.  First they must both



deal not just with maximal concentrations but also with various time averages



of concentrations, such as the .40 PPM (4 hr. avg.), .60 PPM (2 hr. avg.),



and .70 PPM (1 hr. avg.) federal oxidant harm levels.  Feedback criteria must



insure that control stages are implemented in time to prevent exceeding any



of the harm levels for all the various averaging times.  Feedforward mecha-



nisms must include forecasts for each time average.   Second, there are ques-



tions of availability of measurements and/or technical  problems.   This is



much less of a problem for feedback criteria than for forecasting criteria.



Forecasting requires considerable meteorological data,  and even with the best



available technical capabilities it may be inaccurate.   Third, there is the



problem of establishing actual criteria levels which will  make the plan ef-



fective without causing undue socio-economic disruption.  This could be a



particularly important problem for feedback configured plans.  The response



of the system may be so slow as to severely restrict the possibility of



effective action once the episode has started.




2.1.2  Control Strategies




     At the heart of an episode contingency plan are the control  strategies.



A control  strategy consists of a set of control measures to be taken at a



particular stage of the episode along with implementation and enforcement



procedures for those measures.  There are as many control  strategies as there



are episode stages.
                                     15

-------
     Each measure associated with a control  strategy is designed to alter



or control source emissions.  Emission control  can take one of several



forms, [27].   First, it can change the temporal  distribution of emissions



as, for example, would a ban on waste incineration during daytime hours.



Second, it can change the spatial  distribution  of emissions.  An example



would be the curtailment of vehicular traffic  in the downtown area.  Finally,



it can decrease total  source emissions.   The latter effect is the main thrust



of most control measures now in practice.




     Besides  a description of the actual  control  technique, each control



measure should stipulate how the control  is  to  be implemented and enforc:d.



For example,  a partial  ban on automobiles  could  be implemented through a



license plate lottery,  a windshield sticker  program,  or perhaps even by the



color of the  car.   Enforcement could be  voluntary or  involve penalties and



could be carried out by the police, national guard,  or  EPA field personnel.



All of these  procedures must be specified  in the measure.




     There are several  considerations  which  should be taken into account in



formulating control  strategies.   First,  all  control  measures should be capable



of rapid institution and fairly immediate  effects.   Second, all  measures



should be implementable from a legal  point of view.   There should be no



question of legality which would have  to  be  tested in courts during the



episode.   Third, for given effectiveness,  measures should be designed to



minimize economic  disruption and social-political  impact.   Finally, the



design of the strategy  should be such  that it will  not  fail  (let pollutant



concentrations reach significant harm  levels) under  any forseeable circum-



stances.   This necessarily means rating  the  effectiveness of a control



measure with  a large degree of conservatism.
                                    16

-------
2.1.3  Surveillance Procedures
     The surveillance procedures necessary for the effective operation of an
air pollution episode control plan can be divided into two groups:   air
quality and meteorology surveillance  and strategy implementation surveillance
The first group of procedures provides the information needed to determine if
episode conditions exist and the severity of those conditions.   It  partly
consists of routine sampling of appropriate air quality and meteorological
parameters.  During an episode it again consists of air quality and meteoro-
logical measurement, but on a much more frequent basis, so that the necessary
information is available to make use of in the contingency plan.  The second
group of procedures is designed to determine the extent to which the imple-
mentation of control measures is successful.  It includes observing other
agencies involved in the plan to ensure that they are fulfilling their
responsibilities as well as observing the effects of control measures on
various sources.  These surveillance procedures help to provide the informa-
tion inputs necessary for the effective operation of the plan.
2.1.4  Operational Procedures
     In order to ensure rapid and smooth execution of an episode contingency
plan, there must be contained within the plan itself a set of procedures
which explain how it is to be made operational during an air pollution
episode.  These procedures should explain how responsibility is to be
delegated, how communications are to be handled, and how enforcement is to
be conducted.  Further, they should indicate the necessary hardware needs,
like communications equipment to manage the implementation of various control
measures.
                                     17

-------
     Although much of the details of the operating procedures will depend
 upon the actual makeup of the control strategy, it is possible to outline
 some general areas which must be considered.  Under the topic of responsi-
 bility, clear lines of authority will have to be outlined so that there is
 no question of who is responsible for enforcement, who is in charge of
 communication, and particularly, who is responsible for confirming that
 indeed an episode exists and insuring that controls are selected and
 implemented.  Also, the location of various personnel  should be specified.
 In the area of communication, guidelines must be drawn up explaining how
 and when information concerning the episode is to be handled, how communica-
 tion among various personnel is to be conducted, how questions concerning
 the alert are to be dealt with, and how the public is  to be made aware of
 actions it must take.  Under enforcement, the operational  procedures should
 indicate which agencies will be involved, how offenders will be determined
 and dealt with, and when, if necessary, injunctions will  be obtained.   A
 final  area of importance involves personnel  and equipment needs.   The
 necessary personnel  requirements should be indicated,  and equipment needs
 should be listed along with  other items which are to be prepared beforehand
 (e.g.  news releases).

 2.2  DESCRIPTION OF THE LOS  ANGELES PROBLEM
 2.2.1   Geography,  Demography,  and Climate
     Geography
     The  Metropolitan  Los Angeles Air Quality Control  Region, also known
as the  South  Coast Air-Basin,  includes all  of Orange and Ventura Counties
and portions  of Los  Angeles, Riverside,  San  Bernardino, and Santa Barbara
Counties.   Geographically, the  region is bounded by the Pacific Ocean  on
                                     18

-------
the west, by San Diego County on the south,  and by a  series  of  mountain



ranges on the north and east.  Figure 2.3 gives a map of the basin  and



illustrates its location within California,  [2].




     Demography



     Around 10 million people, roughly one-half of California's population,



reside in the Metropolitan Los Angeles AQCR.   Table 2.1  gives the distribu-



tion, by county, of population and land area  within the  region, [2],  [7].



Los Angeles County, the most urbanized area  of the region,  includes  69%  of



the regional population within 32% of the regional  area.  Orange County  ranks



a distant second in population, with 16% of  the regional  total; however,  its



population density is nearly that of Los Angeles.





    Table  2.1.  POPULATION AND LAND AREA - METROPOLITAN  LOS  ANGELES AQCR

County
Los Angeles
Orange
Riverside
San Bernardino
Santa Barbara
Ventura
Totals
California
Sources: State

County's Popul
in Basin - 1
6,890
1,565,000
342,000
565,000
153,000
411,000
9,926,000
20,218,000
of California,
% of County's
ation Population in
972 Basin
98.8
100.0
70.4
79.3
56.9
100.0


The Resources Agency,
Area in
Basin
(Mi2)
2,768
782
1,851
1,135
281
1,863
8,680

Population
Densi ty
(Persons/Mi^)
2,480
2,000
185
497
545
220


Air Resources Board,
The State of California Implementation Plan for Achieving and
Maintaining the National Ambient Air Quali
30, 1
State
Stati
972.
of California,
stical Abstract,

Department of Finance
1972.
ty Standards

, California

, January



                                     19

-------
                                              Location Of

                                                 Basin
SANTA
    BARBARA
                                     LOS  ANGELES
                     £>
                        c
                          /
                            /
                              c
                   Figure 2.3.     Metropolitan Los Angeles AQCR
        Source:  State" of~California,  The  Resources Agency, Air Resources Board,
                The  State of  California Implementation Plan for Achieving and
                Maintaining the  National  Ambient Air Quality Standards, Jan. 30, 19'I
                                      20

-------
      Table  2.2  summarizes  the most  recent  population projections for the
 region,  [6].  As  has  been  the typical pattern for Los Angeles, growth is
 expected  to be  much greater  in satellite areas than it will be in the central
 Los Angeles area.
           Table  2.2.   POPULATION  PROJECTIONS:  COUNTIES WITHIN THE
                       METROPOLITAN LOS ANGELES AQCR  (in millions)
Los Angeles
Orange
Riverside
San Bernardino
Santa Barbara
Ventura
Source: Population
Projections

1970
7.034
1.430
.461
.689
.258
.380
Research Unit,
of California

1975
7.255
1.688
.499
.750
.280
.451
Department
Counties to

1980
7.681
1.940
.573
.851
.315
.575
Approximate
Growth Rate
0.9%
3.1%
2.2%
2.2%
2.0%
4.1%
of Finance, Provisional
2000,

September 15, 1971.
     The location and density of employment within Southern California is
illustrated in Figure 2.4, [25].  As was the case with population,  the central
Los Angeles area contains the greatest concentration of employment.   However,
in comparison to other metropolitan areas, the change in population  and
employment density between Los Angeles and outlying areas is small.   In
general, land use in Los Angeles has led to a sprawling, rather than concen-
trated, metropolitan area.  The percentage of those employed by categories
is depicted in Table 2.3, [7].  Data are presented, for example, on  the
percentage of workers in various levels of government.
                                    21

-------
rv>
                                       Figure 2.4.    1970 Employment in Southern California

                                                                    I
L- ^-T,, 	 	 	 	 „„„„,„„„
|i|3 \f|:
V| ' . * «-.«.«« » ;:••::
J .: : li 1^ 	 _.^ _>^
:' - . ;•:; •:...<: • • • \ L '"' ...i— L!
1 	 "i$| «:
s
j
-. *
,1,
_.a. y _
T
                 REGIONAL STATISTICAL AREAS
                 SUB-COUNTY AfltAS
                Stffif* Cft'fKIH jjttfitfrf* tf ffrtf/vr*/irt
                  Sub-Cou*, Af«c

                  BowAdcr? Lint
                                                            EMPLOYMENT

                                                   (persons employed/gross acre)
D . 01 -. 10
Ll3 .11-1.00
d 1.00-5.00

   >s.oo
                           Source:   Southern  California  Association of  Governments,  Employment Projections.

-------
                               Table  2.3
              EMPLOYMENT IN THE METROPOLITAN LOS ANGELES AQCR
TOTAL CIVILIAN EMPLOYMENT 1973:   4,200,000

      CATEGORY                          PERCENTAGE OF TOTAL EMPLOYMENT

      CONTRACT CONSTRUCTING                    3%
      MANUFACTURING                           25%
         Nondurable Goods                                      8%
         Durable Goods                                        17%
      TRANSPORTATION & UTILITIES               5%
      TRADE                                   20%
         Wholesale                                             6%
         Retail                                               14%
      FINANCE, INSURANCE & REAL ESTATE         6%
      SERVICES                                17%
      GOVERNMENT                              14%
         Federal                                               2-1/2%
         State & Local                                        11-1/2%
      TOTAL                                  100%
 SOURCE:   State of  California, Department of Finance,
          California  Statistical Abstract, 1972
                                    23

-------
     Climate

     The Metropolitan Los Angeles  AQCR can be climatically classified as a


sub-tropical region.   The total  annual  rainfall  in  the Basin is moderate


and occurs predominately in the  winter season;  oftentimes, the summer is


completely void of any precipitation.   Although  the coast can be moderately


cool, the inland valleys frequently experience  very hot summers.


     Wind patterns in the basin  vary as a  function  of both time of day and


season.  Typically, during the late spring,  the  summer, and the early fall,


a strong Seabreeze (4-8 mph)  exists from late morning to evening; winds flow

from the west and sweep across the basin in  an  easterly manner, [22].   In


the evening, the winds stagnate  and undergo  a slight reversal; a very weak

land breeze (0-3 mph) exists.   Figures 2.5a  and  2.5b illustrate this  pattern
                                                                        \
for a typical  October day, [2].


     During the winter, the typical condition is a  moderate daytime Seabreeze

and a moderate nighttime la^d  breeze.   Because  of this balance, air masses


can remain in  the basin for up to  three days in  the winter, whereas in the


summer, most of the basin on  any day receives marine air from the Seabreeze,

[17], [22].   From day to day,  the  wind intensities  and directions in  these


typical patterns can  vary.  Also,  atypical  patterns, such as the strong

Santa Ana land breeze, periodically occur.


     The differences  in the winter and summer winds and precipitation are


just two aspects of a general  winter-summer  meteorological  dichotomy that


exists in Southern California.   Summer is  characterized by high temperature,


intense solar  radiation, and a persistent  marine layer temperature inversion


that restricts vertical  mixing.  Winter has  lower temperature and solar


radiation and  a low,  but weak  nocturnal  temperature inversion.  As illustrated
                                    24

-------
ro
en
                            Figure 2.5a.  Air Flow Patterns - South Coast Air Basin  (October  1200-1800  PST)
                               Source:  State of California, The Resources Agency,  Air Resources  Board,
                                        The State of California Implementation Plan for Achieving
                                        and Maintaining the National Ambient Air Quality Standards,
                                        January 30, 1972.

-------
                 X.
ro
cr>
                                                                                              »     -a	.        \\y
                                                                                        'naf-df/j^^^L^^    ^
                   i   -it^            ~-» iQi_^
                   •^^-^Jiis,^^^.—^^^ifc^^^^^^^^te^y-^v^j^^-^/
                   ^_^,	---^^ _ --- _ ^- J^S^>_ ~ -^ ^^ -^ »-^,^ 3
                   ^-~-—^ -—*"*—(T^—''^^C—"^—•~**4^-^--~s~ —^J^_^
                 ^-^_i» -—- ••—^^c-—-~^-—*~~—^ov^-^^_!^»-'^—-^—^^-.—>v
                 --s	^"Jw-^^-^^rf-^^^---^—^^_xv^^—'-O«^._^v_^wx^i^.^.^^--w--^_^^_^O-"^j»H-.' *^s^y^jy--~—?ti^T_-'-*»-rf
^N^^^^^_^-x_«^^lI-^i^II---Il-^^        ' "<;'!LV_^-^^^^Z^^^~^ ^ ^"^-^-~-^^_^~ ^_^~^_^-s_x-
            Figure  2.5b.  Air Flow Patterns -  South  Coast Air Basin  (October  0000-0700  PST)
                Source:  The State  of California, The Resources Agency, Air Resources Board
                           The State  of California  Implementation Plan  for Achieving  and
                           Maintaining the National Ambient Air Quality  Standards,  January  30,  1972.

-------
in Figure 2.6 these weather patterns make winter the season  for  primary  air
contaminants,* such as carbon monoxide, and summer the  season  for  secondary
photochemical contaminants,* such as oxidant.
2.2.2  Overall Air Quality Levels
     Air quality measurements taken in the Metropolitan Los  Angeles  AQCR
show the air pollution problem to be severe for all three of the contaminants
to be considered in this study, (oxidant, CO,  and N02).  The severity is
demonstrated by three indicators:   (1) the geographic extent of  the  problem,
(2) the frequency of standard violations, and  (3) the magnitude  of the
standard violations.
     Table 2.4 compares 1972 air quality levels for oxidant, CO, and N02 at
several selected stations in the basin to the  National  Ambient Air Quality
Standards.   Figure 2.7 depicts the  location of these air monitoring  stations.
In 1972, oxidant levels exceeded the one-hour  standard at every  station.  In
the central-eastern portion of the  basin, (Azusa-Riverside), values  5 to 6
times the oxidant standard occurred several times.  At these stations,
violations usually have occurred as many as 150 to 200 times per year.  In
the basin as a whole, the oxidant standard typically has been exceeded 250
to 300 days  per year.
     The situation is only slightly better with CO.  All stations except
Ventura exceeded the eight hour CO  standard, and Ventura is just in compliance
In the central-coastal area,  (Lennox-Reseda-Downtown-Anaheim-Pasadena), values
 * Primary pollutants are those which are emitted directly into the atmo-
   sphere, e.g. CO, HC, NO, SOp.  Secondary pollutant are those which
   result from chemical transformation processes involving primary pollutants,
   e.g., N02, 03.
                                     27

-------
                          30 ,
ro
CO
                                                                           -—\
                  Q- 0.
                  D- 0_
               f—  -o
               2:  -1-
               uj  x
               c_>  o
               2:  c:
               O  o
                  c  c
                  O  KS
                  ^a  -a
                  S_  •!-
                  ra  x
                  o  o
10
                                                                                                   CO
                                                                         Oxidant
                                  FIGURE  2.6.    Comparison of Seasonal Pollution  Patterns
                                                 Carbon Monoxide vs. Oxidant
                                                 (Average of Daily 1-hr. Maximum by Month:

                                                 Pasadena, 1970)

                    Source:   Air Monitoring Logs  of  the  Los Angeles County APCD

-------
                 Table 2.4. AIR QUALITY AT SELECTED SITES IN THE METROPOLITAN LOS  ANGELES
                                         AIR QUALITY CONTROL REGION,  1972

Station
Downtown L.A.
Lennox
Santa Barbara
Ventura County
Anaheim
Reseda
Pasadena
Azusa
Riverside
OXIDANT
Days Violating
Federal Standard
138
20
19
85
80
157
81
199
158
Maximum One
Hour Average
(PPM)
.25
.17
.10
.15
.35
.29
.38
.49
.50
CARBON MONOXIDE
Days Violating
Federal Standard
104
144
10
.0
50
98
43
6
98
Maximum Eight
Hour Average
(PPM)
34
33
f
16
8'6
34
37
34
i
I
14
25 1
NITROGEN DIOXIDE
Yearly Average
(PPM)
.078
.064
.02 - .03*
.02 - .03*
.046
.068
.076
.062
.04*
1 11
.08 1 I 9 1 .050
BU^aiga^s^Mu^&i^.'^^
* Estimated
       Source.-   California Air Resources Board, California Air Quality Data,  Volumes  II,  III,  IV,  January  1971
                 through December 1972.
                 Air Monitoring Logs of the Los Angeles County APCD,  Ventura County  APCD,  Riverside County  APCD,
                 and Orange County APCD.

-------
co
o
  Santa

Barbara
                   ^ Full  Station

                   * Partial Station

                   •fr NASN-EPA
                                                                     Location Of

                                                                        Basin

                                                                                 Pasadena

                                                                                      Azuza
                                Figure 2.7  Location of Air Quality Monitoring Stations
                                            (Metropolitan Los Angeles AQCR - 197Q)

-------
up to 4 times the standard occurred.  Lennox and Downtown Los-Angeles



exceeded the CO standard over TOO days in 1972.




     The NCL pollution problem is the least severe of the three.   The  annual



average NCL standard was exceeded at most stations, but the excess was less



than 60%.  The highest t^ levels occur in the central  area, (.Downtown



Los Angeles - Pasadena).




     The spatial distribution of each pollutant  is consistent  with the known



meteorology and emission source distribution.   Since the marine layer  inver-



sion ceiling is lowest near "the coast and since  the central area  has the



highest concentration of sources, carbon monoxide (a primary pollutant)  is



greatest in the coastal to central area.  N0? (an intermediate pollutant)*



is highest in the central basin, and oxidant (a  secondary pollutant) is  most



severe in the central to eastern portions of the basin.  Both  of  these obser-



vations are consistent with the daily sea breeze blowing the polluted  air



across the basin as photochemical reactions proceed.  Santa Barbara and



Ventura experience the least pollution because they are relatively low in



source density and because they are not downstream of the most dense areas  for



usual wind patterns.




2.2.3  Air Pollution Control Agencies
                                                           «,



     There are basically three levels of government involved with air



pollution control in the Metropolitan Los Angeles AQCR:  local Air Pollution



Control Districts (APCD's)at the county level, the California Air Resources
* N02 plays an intermediate role since it  is first  produced from the  primary


  contaminant, NO, but is then consumed as the photochemical reaction reaches

  the end product stage.
                                    31

-------
Board (ARB) at the state  level,  and the United States Environmental  Protection
Agency (EPA) at the national  level.   Brief discussions of the major  res-
ponsibilities and special  capabilities  of these agencies follows.
     County APCD's
     In California, County Air Pollution Control  Districts (APCD's)  have
local responsibility for  air  pollution  problems.   In the Metropolitan
Los Angeles AQCR there are six such agencies,  corresponding to the six
counties partially or completely within the basin (see Table 2.1).  The
oldest and  largest of these is the Los  Angeles County APCD, created by
the Air Pollution Control  Act of 1947.   The L.A.  APCD's regulations  as
well as its monitoring, calculational, permit, reporting, and enforcement
techniques  have served as models for other California APCD's as well as
other agencies nationwide.
     The primary responsibility of the  local APCD's is to control
stationary  sources of air pollution.  This is  done through a system  of
regulations arid permit procedures.  Inspection (source testing) and
enforcement operations are carried out  to insure  that the regulations
and permits are complied  with.  An engineering department is maintained to
support these efforts, [26].
     Another APCD responsibility is to  operate an air pollution emergency
alert system.   (The next chapter will  describe and evaluate this  system
in detail.)  Air monitoring programs have been established in each county
to support  the alert function.  The L.A. APCD maintains an extensive
monitoring  system, with 12 full stations now  in operation.  The monitoring
data gathered by APCD stations provides important background data for
general air pollution control policy as well  as the basic data for
alert warnings.
                                    32

-------
     The L.A. APCD also has become involved in air pollution research.   It
presently maintains rather unique air quality analysis and meteorology
sections.  The meteorology section has been concentrating recent efforts
on developing prediction capabilities for various pollutants.   The oxidant
forecasting program has progressed very well, but predictions  of other
pollutants, e.g. CO and NCL, are not yet reliable, [13].
     The California ARE
     The California Air Resources Board has responsibility for implementing
and enforcing the statewide air pollution control program.  One ARB
function is  to  review, evaluate, and change  (if necessary) local APCu
programs.  At times, the ARB also provides certain APCD's with technical
services (air monitoring set-up, emission inventory compilation, etc.)
While the APCD's are concerned basically with stationary sources, the
ARB has primary responsibility for motor vehicle  emission standards
and .accredits  devices  for  the  extensive  California Used  Car Retrofit
Program.
     Other ARB  functions include supporting research into the causes
and effects of  air pollution, evaluating various plans as to air quality
impact,  and  performing  certain basic technical  studies.  The ARB  is
presently formulating a statewide air pollution episode contingency plan.
This preliminary plan will  be  described and  evaluated in the next chapter.
     The United States EPA
     The Environmental Protection Agency is  responsible for national environ-
mental programs.  In air pollution, this responsibility basically trans-
lates to carrying out the provisions of the  1970  Federal Clean Air Act
(as amended).   Accordingly, the EPA has promulgated national ambient air
quality standards  (NAAQS) and  has given the  states a  time table  for
developing and  implementing plans to achieve  these standards.   EPA
regulations  also call for episode contingency plans to prevent  "significant
                                  33

-------
h-arm" air pollution levels from being reached.  If the states do not
prepare satisfactory implementation plans, the EPA has the authority
to develop and carry out its own plan.  The EPA is presently in the pro-
cess of evaluating and, where necessary, developing implementation
strategies for various air basins, including the Metropolitan Los Angeles
AQCR-.

2.2.4  Emission Source Inventory

     Table 2.5 presents an approximate emission source inventory for
reactive hydrocarbons (RHC), CO, and NOV in the Metropolitan Los Angeler
                                       A
AQCR.  Percentage emissions for each source are given to the nearest %%.
Data are given for 1973.  and projections are made through 1975 for the
following:
   e   Stationary Sources:  Present APCD controls.
   ®   Ljght Duty Motor Vehicles (LDMV's):  The existing Federal
       and California new car emission standards and the present
       California ARB used car retrofit program for 1955-65 and
       1966-70 vehicles.
   «   Heavy Duty Motor Vehicles (HDMV's) and Diesels:  The present
       Federal control program.
   e   Aircraft:   The present Federal  control  program.
   e   Motorcycles:  No controls.

     Basically, Table 2.5 has been derived by projecting the 1970 California
ARB inventory, [3], and the 1970 L.A.  County APCD inventory [21], to 1973
and 1975, while including the effect of various control  programs.  The data
for the projections, for  the control  effects, and for any corrections to the
inventory have been taken  from references [14], [21], [27], [28], [29].  The
only very significant change in the ARB-APCD inventory is a change in hydro-
carbon reactivity assumptions;  recent EPA results, 11],  have been substituted
                                    34

-------
for ARB-APCD reactivities in the mobile source and petroleum marketing
categories.  The reactivity differences are summarized in Table 2.6.
• 	 	 r* 	 —
TABLE 2.5 AN APPROXIMATE EMISSION
- LOS ANGELES AQCR, 1973



STATIONARY SOURCES:
Petroleum Prod. & Refining
Petroleum Mrkting: Auto Tank Fill.
Petroleum Mrkting: Undgrd Tanks & Oth
Organic Indust. Surface Coating
Solvents kes. & uomm. burface loating
Dry Cleaning
Degreasing
Other
Steam Power Plants
Res., Comm., & Ind. Fuel Combustion
Other Stationary Sources
MOTOR VEHICLES:
Light Duty Motor Vehicles (LDMV's)
Heavy Duty Motor Vehicles (HDMV's)
Motorcycles
Diesels
OTHER MOBILE SOURCES:
Jet Aircraft
Piston Aircraft
Ships & Railroads

TOTAL PERCENTS
INVENTORY FOR THE
- 1975
PERCENT OF TOTAL
1973
RHC TCT NOx

k - 5k
8
.4 - k
2
2
k
Ik - -
2k - -
Ik
9k
1 1% k

58 87 65
8 7k 9
3 1%
1 k Ik

5 k k
2k ^k k
k - -

100% 100% 100%
METROPOLITAN



(to nearest 2%)

RHC

k
9
5
2k
2k
k
2
3
_
-
1

50
9
4
Ik

Sk
3
k

100%
1975
-or

_
-
-
-
-
—
-
-
_
-
2

83
Sk
2k
k

1
2
k

100%

NOX

6k
-
1
-
—
-
-
-
9
n
i

56
n
-
2

1
1
k

100%

TOTAL EMISSIONS, (TONS/DAY) 1
250 9000 1450
1100
7400 1
300
     Table 2.5 shows that gasoline-powered  motor  vehicles  (LDMV's,  HDMV's
and motorcycles),  constitute the largest  source of all  three  emitted  contam-
inants.   In 1973,  gasoline-powered vehicles account for 69% of  RHC; 96%  of  CO;
                                     35

-------
and 74% of NO .   In  1975,  they  constitute  63%  of  RHC;  94%  of  CO;-and  67% of
             X


NO    LDMV's alone contribute  58%  RHC;  87% CO;  and  65% NOY in 1973.   This
  x                                                     A


shrinks to 50% RHC;  83% CO;  and 56%  N0v in 1975,  because of the present
                                     X


Federal and California  new car  controls and the present California  ARB



retrofit program.   Total  emissions reduce  from 1250 tons/day  RHC;  9000



tons/day CO; 1450  tons/day NO   in  1973  to  1100 tons/day RHC;  7400  tons/day



CO; and 1300 tons/day NO  in 1975.   This change is  basically  due  to the
                        X


motor vehicle control programs.
  TABLE-2.6.   ALTERNATIVE  HYDROCARBON  REACTIVITY ASSUMPTIONS

SOURCE TYPE
Stationary Sources
Petroleum Production
& Refining
Petroleum Marketing
Organic Solvents
Others
Motor Vehicles
LDMV Exhaust
HDMV Exhaust
LDMV & HDMV Evaporative
Motorcycles
Diesels
Other Mobile Sources
Jet Aircraft
Piston Aircraft
ARB-APCD
REACTIVITY
10%
55%
20%
0-20%
75%
75%
55%
•t *•
10%
30%
75% Exhaust
55% Evaporative

REACTIVITY
USED HERE
(EPA)
10%
93%
20%
0-20%
77%
79%
93%
96% (2-Stroke)
86% (4-Stroke)
99%
90%
77% Exhaust
93% Evaporative
                                     36

-------
     Figures 2.8a, 2.8b, and 2.8c, summarize the trends  in RHC,  CO,  and  NO
                                                                         A

emission levels from 1965 to 1980 under the present control  program,  [14],

[21], [29].  Notable reductions are being made for each  contaminant  during
                           /
the 1970's.  To attain the NAAQS in the Los Angeles basin  by 1975  or  1977

will require even greater reductions than indicated in  Figure 2.8, especially

for RHC.  Thus, one would expect even more controls to  be  initiated  and  even

further emission reductions to be achieved during this  decade.


2.2.5  The Potential for Air Pollution Episodes

     The purpose of this study is to formulate an episode  contingency plan

which will insure that the federal "significant harm"  levels for oxidant,

carbon monoxide, and nitrogen dioxide are "never reached"  in the Metropolitan

Los Angeles AQCR.  These levels, established by the EPA  in the Code  of Federal

Regulations - 40 CFR 51.16(a), are summarized in Table  2.7.   The specifica-

tions include:  one, two, and four hour oxidant averages;  one, four,  and
 Table  2.7    FEDERAL  SIGNIFICANT  HARM
             LEVELS FOR OXIDANT,  CARBON
             MONOXIDE, AND NITROGEN DIOXIDE
       Pollutant              Harm Level             Averaging Time


        Oxidant                 .70 PPM                One Hour
                                .60 PPM                Two Hours
                                .40 PPM                Four Hours


        Carbon                  125 PPM                One Hour
        Monoxide                 75 PPM                Four Hours
                                 50 PPM                Eight Hours


        Nitrogen               2.0  PPM                One Hour
        Dioxide                 .50 PPM                Twenty-four Hours
                                     37

-------
 CO .
 Z2!
 O
 co
 CO
            2000
 o
    00
    ~ZL

    O
     4 1000
 CO
 •z.
 o
 CO
 CO
     CO
     z:
     o
            65



            20,000
                   70
                              75
                                               80
                                                85
          4 10,000
            65
                   70
                         75
                                                80
                                                      —4

                                                       85
 CO
 •z.
 o
 CO
 to
  X
o
Q


CO
z:
O
           I- 2000
-L looo
             65
                    70
                          75
80
-H

 85
Figure 2.8  Total RHC, CO,  and  NO  Emissions in the  Metropolitan Los Angeles
            AQCR Under the  present Control Program  (See Text,) 1965-1985
                                      38

-------
eight hour carbon monoxide averages; and one, twenty-four hour nitrogen
dioxide averages.
     Before formulating a contingency plan, it is useful  to examine the
potential for air pollution episodes in the Los Angeles basin  by answering
the question:  How frequently would the "never to be reached,  significant
harm" levels be attained without emergency controls?  The answer to this
question may be very different for each contaminant; hence the three
pollutants will be discussed separately below.
     Carbon Monoxide
     It was shown earlier, (Table 2.4), that violations of the federal
air quality standard for carbon monoxide are large, frequent,  and wide-
spread in the Metropolitan Los Angeles AQCR.  The 9 PPM,  8 hour CO standard
typically has been violated around 150-225 days per year in the L.A. basin
during the late 60's and early 70's.*  However, the "significant harm"
levels are much higher than federal air quality standards.  Table 2.8
compares the significant harm levels to maximal carbon monoxide con-
centrations in the Metropolitan Los Angeles AQCR for 1970 through 1972.
     Table 2.8 shows that the CO significant harm levels  have  not been
exceeded in the Los Angeles basin from 1970 through 1972.  In  fact, the
maximal recorded one hour CO level (54 PPM) is well below one  half the  one
hour harm level (125 PPM) and even well below the four hour harm level
(75 PPM).  The eight hour CO harm level is the only one that appears to
be at all in danger of being exceeded.  The nearest occurrence to a
*  Actually, L.A.  APCD records show a consistent violation frequency of
   365 days per year prior to 1968.  This is due to a flaw in air monitoring
   practices for CO prior to April, 1968.  Readings were high (by around
   4 PPM on the average), with the actual error depending on absolute
   humidity, [15].
                                     39

-------
violation  in  1970  -'1972  is a 36 PPM 8 hour CO concentration, only 28%

below' the  8 hour harm level  (50  PPM).

     It  would  be useful to examine monitoring data for CO even further in the

past,  (say the 1960's  and the late 1950's).  However, prior to April 1968,

recorded CO data is  considerably in error,  [15J.  Thus, we are restricted to

the recent past.   The  greatest one hour CO concentration since April 1968 is

65 PPM at  Lennox in  January 1969  [12].  This is still about one half the one

hour harm  level.   The  greatest 8 hour recording is 42 PPM at Burbank in

December 1969, 16% below  the 8 hour harm  level.
        Table  2.8    COMPARISON OF MAXIMUM CARBON  MONOXIDE  LEVELS
                     IN THE METROPOLITAN  LOS  ANGELES  AQCR TO THE
                     FEDERAL SIGNIFICANT  HARM LEVELS, 1970  THROUGH
                     1972
Averaging CO "Significant Maximum Five Values. 1970 - 1972
Time Harm" Level
54 PPM -
49 PPM -
1 m. 125 PPM 49 PPM -
48 PPM -
47 PPM -
4 m. 75 PPM
36 PPM -
33 PPM -
8 m. 50 PPM 33 PPM -
32 PPM -
32 PPM -
Reseda
Lennox
Lennox
Lennox
Lennox
October 1970
January 1971
November 1971
February 1970
December 1971
NOT AVAILABLE
Burbank
Reseda
Lennox
Burbank
Reseda
February 1970
November 1970
December 1972
January 1970
December 1972
Data Sources:
California Air Resources Board, California Air Quality Data
Volumes II, II & IV, January 1970 through December 1972.
Air Monitoring Logs of the Los Angeles County APCD.
                                     40

-------
     Based on Table 2.8, it appears that a violation  of the  one  hour  or  four
hour CO harm level  would be extremely rare at the CO  emission  levels  of  the
early 1970's.  A violation of the eight hour CO harm  level  is  more  possible,
but even this would be very infrequent, (possibly once  every 5 to 20  years).
As was shown in Figure 2.8, CO emission levels are rapidly  declining, (it
may be significant that the maximum CO concentrations for 1970 - 1972 occurred
in 1970).   By 1975 and 1980, CO emission levels should  be 60%  and 30%,
respectively, of the 1970 emission level.   Furthermore, the  emissions will
be slightly less concentrated since growth is occurring more in  outlying
areas of the basin.  These considerations  indicated that it  is highly
improbable that significant harm levels for CO will be  reached in the
Metropolitan Los Angeles AQCR during this  (or subsequent) decades.
     Nitrogen Dioxide
     The situation for nitrogen dioxide is similar to that  for carbon
monoxide;  although the annual mean NOp air quality standard is commonly
violated in the Los Angeles basin, exceeding the significant harm levels
appears highly unlikely.  Table 2.9  compares the significant  harm  levels
to the maximal N02 levels recorded in the Metrpolitan Los Angeles AQCR
from 1970  to 1972.
     Table  2.9  shows that N02 significant harm levels have not  been
exceeded in the Los Angeles basin from 1970 to 1972.   The maximal  one
hour N02 concentration  (.83 PPM) is less than one half the  one hour
harm level.  The maximum 24 hour N02 value (.35 PPM)  is. 30% lower than the
twenty-four hour harm level (.50 PPM).   On this basis,  violations of  the one
hour N0? level appear extremely improbable.  Violations of  the twenty-four
hour NOg harm level are more likely, but these would  be very infrequent.
     An examination of air monitoring data for the 1960's and  late  1950's

                                     41

-------
reveals  that  maximal one hour and twenty-four hour N02 values were near
or below those  for  the  1970-1972 period with six exceptions, [12]:
     ©  1.29  PPM  for one hour (Downtown L.A., December 1958)
     ©  1.23  PPM  for one hour (Burbank, December 1960)
     ©  1.00  PPM  for one hour (Burbank, December 1959)
     •  .61 PPM for 24  hours (Burbank, January 1960)
     g  .47 PPM for 24  hours (Burbank, December 1959)
     i  .42 PPM for 24  hours (Lennox, October 1967)
         Table 2.9.   COMPARISON OF MAXIMUM NITROGEN DIOXIDE LEVELS
                     IN  THE  METROPOLITAN LOS ANGELES AQCR TO THE
                     FEDERAL SIGNIFICANT HARM LEVELS, 1970 THROUGH
                     1972
 Averaging
    Time
N0? "Significant
  ^Harm" Level
   Maximum Five Values,  1970 - 1972
    1  hr.
     2.0 PPM
.83 PPM -  Pasadena
.80 PPM -  Downtown L.A.
.74 PPM -  Downtown L.A.
,71 PPM -'Downtown L.A.
.67 PPM -  Whittier
January 1970
January 1970
March   1971
July    1970
January 1972
   24 hr.
     ,50 PPM
.35 PPM -  Pasadena
.32 PPM -  Whittier
.31 PPM -  Burbank
.28 PPM -  Downtown L.A.
.27 PPM -  West L.A.
January 1970
January 1972
January 1970
January 1971
January 1971
 Source:  California Air Resources  Board,  California  Air  Quality  Data,
          Volumes_II, III  & IV,  January  1970  through  December  1972.
          Air Monitoring Logs  of the  Los Angeles  County APCD.
                                    42

-------
These are, respectively, the three highest one hour and twenty-four hour



NOp averages ever recorded in Los Angeles.  The one hour maxima are still



far below the 2.0 PPM one hour harm level.  The .61 PPM 24 hour average



in January 1960 is the only reported data exceeding the 24 hour harm



level (.50 PPM).




     Based on these considerations > a violation of the one hour NO,, harm



level would be expected to occur almost never at the NO  emission level of
                  ^   " ~~~ J *"-—""'- J J —  •"•           j_... -.   j. _   .A   "~" '  ~    J


the early 1970's.  Violation of the twenty-four hour harm level would be



extremely infrequent (possibly once every 5-20 years).  As was shown in



Figure 2.9, NO  emissions will decline considerably in the Los Angeles basin.
              A


By 1975 and 1980, NO  emission levels should be respectively 75% and 55% of
                    X


the 1970 emission level.  Furthermore, the emission spatial distribution



will be slightly less concentrated.  This leads to the conclusion that it is



very improbable that significant harm levels for NOg will be reached in the



Los Angeles basin during this (or subsequent) decades.




     Photochemical Oxidant




     The situation is much different with oxidant; the Los Angeles Metropolitan



AQCR is notorious for extremely high oxidant concentrations.   Oxidant values



up to seven or eight times the federal air quality standard are known to



occur.  Data has been gathered and is presented in Appendix C which compares



maximal one, two and four hour oxidant averages in the Metropolitan Los Angeles



AQCR to the federal  "significant harm levels."  Appendix C includes all days



in 1-70 through 6-73 with maximal one hour oxidant greater than or equal to



.40 PPM.   In the Appendix, violations of significant harm levels are marked



with an asterisk.  The data for these violation days are summarized in



Table 2.10.
                                     43

-------
Table  2.10
OXIDAHT EPISODES EXCEEDING
FEDERAL SIGNIFICANT HARM LEVELS
IN THE METROPOLITAN LOS ANGELES
AIR QUALITY CONTROL REGION
(1-70 - 6-73)
Federal Significant Harm Levels
Date Station
Tuesday, 2 June 1970 Azusa
Pamcna
Tuesday, 2 June 1970 Azusa
Parnona
Wednesday, 24 June 1970 Redlands
Saturday, 4 July 1970 Pasadena
Thursday, 6 August 1970 Riverside
Azusa
Friday, 7 August 1970 Azusa
Riverside
Thursday, 10 Sept. 1970 San
Bernardino
Friday, 11 Sept. 1970 Azusa
Monday, 13 Sept. 1971 Riverside
Thursday, 29 June 1972 Riverside
Tuesday, 11 July 1972 Azusa
Wednesday, 6 June 1973 Upland
Friday, 8 June 1973 Fontana
Thursday, 21 June 1973 Downtown
Los Angeles
* Significant Harm Level Exceeded
Maximum
One Hour
Average
(PPM)

.70

.47
.47
.47
.47
.52
.45
.62
.49
.58
.48
.42

.47
.56
.45
.47
.60
.44
.52


Maximum
Two Hour
Average
(PPM)

.60

.46
.45
.46
.45
.47
.44
.59
.49
.54
.45
.42

.47
.47
.44
.44
.53
.44
.52


Maximum
Four Hour
Average
(PPM)

.40

.42*
.42*
.42*
.42*
.43*
.41*
.53*
.44*
.46*
.42*
.44*

.42*
.40*
.40*
.40*
.46*
.41*
.47*


NOTE:
 Source:   California Air Resources Board, California Air Quality Data.
          Volumes  II. Ill & IV, January 1970 through December 1972.

          Air Monitoring Logs of the Los Angeles County APCD
                          "   "   "  San Bernardino  "   "
                          "   "   "  Riverside       "   "
                          "   "   "  Orange          "   "

:   Data in 1970 and 1971  from the UC Riverside station at Upland have
   not been included in this table.  That station frequently experienced
   the highest oxidant concentrations in the basin; however, a data
   analysis by the ARB has lead to the conclusion that the UCR Upland
   data are unreliable.
                               44

-------
     Table 2.10 indicates that in 1970-1973,  the  1  hour  harm  level  (.70 PPM)



and the 2 hour harm level (.60 PPM)  were not  exceeded.   The maximum  1 hour



average was .62 PPM, and the maximum 2 hour average was  .59 PPM, both at



Riverside, 6 August 1970.  The 4 hour harm level  (.40  PPM) was exceeded 14



times, (8 in 1970, 1 in 1971, 2 on 1972 and 3 in  1973).   The  maximal 4 hour



average was .53 PPM, again at Riverside, 6 August 1970.




     Monitoring data prior to 1970 also has been  examined,  [12].  The three



highest one hour oxidant values that have been  recorded  prior to 1970 are



shown in Table 2.11.
I r\ui_i_ <- . i i .
SEVERE OXIDANT EPISODES IN
THE METROPOLITAN LOS ANGELES
AIR QUALITY CONTROL REGION
Federal Significant Harm Level
Station
Pasadena - July 1955
Downtown LA - Sept. 1955
Azusa - Aug 1967
*Significant Harm Level Exceeded
Maximum
One Hour
Average
(PRM)
.70

.74*
.68
.65

Maximum
Two Hour
Average
(PPM)
.60

.69*
.56
.58

Maximum
Four Hour
Average
(PPM)
.40

.58*
.44*
.48*

     Appendix C and Tables  2.10 and 2.11 point toward the following conclusions:



     •    Violations of the 1 hour and 2 hour oxidant harm levels  would tend



to be rare  (say about every 5 to 10years)at 1970-1973 emission levels.  The only



recorded violation of each occurred back in 1955,although the 2 hour harm level



was almost  reached in 1967  (.58 PPM: Azusa, August) and 1970 (.59  PPM:  Riverside,



August).
                                     45

-------
     e    Violations  of the  4  hour  oxidant  harm  level  (.40  PPM)  are  much

more common;  they should be  expected,  on  the  average,  almost  3 to  5  times

per year at the 1970  -  1973  emission  level.   In  the  past, the greatest  excess

of the 4 hour harm level  occurred at  Pasadena  in  1955,  .58  PPM  (45%  excess).

In recent years,-the  greatest  excess  has  been  in  Riverside  in 1970,  .53 PPM

(33% excess).   The second greatest  level  in recent years occurred  in Downtown

Los Angeles in 1973,  .47 PPM (18% excess].


     ©    Even though the one  and two  hour  harm  levels might  be  (rarely)

exceeded, the 4 hour  harm level  appears to  be  the binding constraint^   My

days that approach the  one or  two hour harm level easily exceed  the  four hour

level.  For instance, all  recorded  days with  one  hour  level above  .60 PPM or

two hour level  above  .55 PPM show a four  hour  level  in excess of .44 PPM.


     e    As  was  shown  in Figure 2.8,  RHC and  NO  emission  levels  are
                                                A

decreasing in the Los Angeles  Basin.   Violation  of the one  and two hour harm

levels will be even less likely  in  the future; they  might not even occur.

Violation of  the  four hour harm  level  will  continue  to occur, but  the fre-

quency and magnitude  of these  violations  should  decrease through this decade.


In summary, including consideration of decreasing emissions,  the potential

for exceeding the federal  "significant harm levels"  for oxidant, CO, and

N02 in the Metropolitan Los  Angeles AQCR  is as follows:


     ®®@    It is extremely  probable  that the  one and  four  hour  CO harm levels

and the one hour  N02  harm level  will  never  be  exceeded  in this decade.

(subjectively one chance in  100  to  1000).*
 These probabilities  could  be  actually  estimated  by  doing  a  statistical
 analysis  of past  air monitoring  data and  by  allowing  for  emission  level
 changes.   However, such  a  detailed  calculation is not within  the scope
 of this  study.


                                    46

-------
            It is 1ikely that the eight hour CO harm level  and twenty-four

hour N02 harm level  will never be exceeded in this decade,  (subjectively one

chance in 10 to 100).*

     •e»    It is possible that the one or two hour oxidant harm levels will

be exceeded in this  decade, (subjectively one chance in two).
            It is highly probable that the four hour oxidant harm level  will

contin.ue to be exceeded.  Violations are occurring an average of 3 to 5  times

per year in the early 1970's.  The violation frequency should decrease con-

siderably by the end of the decade.


2.2.6  Transportation Data Base


     Gasoline powered motor vehicles, (including auto tank filling at

service stations), account for 77% of RHC, 96% of CO, and 74% of NO
                                                                   /^

emissions in the Metropolitan Los Angeles AQCR.  An effective episode

contingency plan must include measures to abate traffic emissions.  Later

in this study, a series of alternative transportation controls will be

evaluated.   In order to determine the effectiveness of these controls, a

transportation data base is required.


     The most extensive transportation data for Southern California has  been

generated in the ongoing Los Angeles Regional Transportation Study (LARTS)

by the California Department of Transportation (DOT)**, [8], [9].  The LARTS

study area  includes all of Los Angeles, Orange, and Ventura counties and the

western, most developed portions of Riverside and San Bernardino counties.
 *These probabilities could be actually estimated by doing a statistical
  analysis of past air monitoring data and by allowing for emission level
  changes.  However, such a detailed calculation is not within the scope
  of this study.
**Formerly the Division of Highways.

                                     47

-------
 The boundaries  do  not  coincide  exactly with the Metropolitan Los Angeles
 AQCR,  (Santa  Barbara County  is  omitted while  portions of Los Angeles,
 San Bernardino,  and Riverside County are added).  However, on the whole the
 two regions are  nearly equivalent; the population contained in each  is
 almost  idential  [28],  It is therefore assumed here that the LARTS trans-
 portation  data  is  representative of the Metropolitan Los Angeles AQCR.
     The LARTS  project, as reported in the LARTS 1967 Base Year Report,
 (1971),  [8] ,  and the 1966 Base  Year_ Calibration Report, (1972), [9] , has
 surveyed approximately 2% of total daily trips in the LARTS region.  This
 survey  has provided information on total  trips taken, with breakdowns for
 various types of trip categories.  When combined with estimates of average
 length  for different trip types, [28], results can be obtained for vehicle
 miles travelled  (VMT).
     Table 2,12  presents results for average weekday VMT.   Total  auto VMT
 is  presented  along with percentages for freeway vs non-freeway and for
 morning and afternoon rush hours.  Values are given for 1967 (LARTS base
 year), and projections  have been obtained for 1980, [16],  [28].   Linear
 interpolation provides  the values for 1973 and 1975.   These results are
 current in the sense that previous estimates,  [28], have been modified to
 include the latest correction factors,  ([20],  July 1973).
     In formulating a contingency plan  for Los Angeles air pollution
 episodes, this study will  later consider, as potential  controls,  measures
which eliminate various incentives  to  drive ("trip purposes").   In order
to ascertain  the overall  effect of these  controls, it will  be necessary
to know how total VMT  is  distributed  according to  various  trip purposes.
Basic data  on  percentage  of  total  trips by trip purpose has been  provided
                                     48

-------
    by LARTS studies, [8], [9], [20], and by a System Development Corporation
    survey, [18].  This data has been reviewed and reworked in other transpor-
    tation studies, [16], [28],  Unfortunately, very significant disagreements
    exist among these various reports.  After careful analysis of these differ-
    ences, it has been concluded that, on the whole, recently corrected LARTS
    data is most reliable.  The SDC survey is based on a very small  sample  and
    it contains a considerable bias since only employed drivers were included
    in the survey.  The other studies, [16], [28], have not benefited from  very
    recent LARTS correction factors.   Thus LARTS data will  serve as  the basis
    for most results given below.
                Table  2.12   AVERAGE WEEKDAY  VMT IN  THE  LARTS  REGION  -
                            TOTAL,  FREEWAY,  NON-FREEWAY,  AND  RUSH-HOUR
Year
1967
1973C
1975C
1980b
Total VMTa
•(Million Miles\
\ Per Day /
153
183
194
219
Freeway
45%
48%
49%
52%
Non-Freeway
55%
52%
51%
48%
• d
Morning
Rush Hour
(7:24 - 8:24)
10.5%
10.5%
10.5%
10.5%
Evening
Rush Hour
(16:24 - 17:24)
11.8%
11.8%
' 11.8%
11.8%
  Computed from references [8],  [20],  and [28]
b Reference [28].
  Linear interpolation.
  Computed from reference [28].
                                        49

-------
      Table 2.13 presents the distribution of total vehicle trips and total

 VMT by trip purpose.  Values are given for average weekdays as well as

 for average weekends.  The notable and expected difference between weekdays

 and weekends is a considerable decline in the percentage of work related

 VMT, 42% to 18%.
           Table 2.13
DISTRIBUTION OF VEHICLE TRIPS AND TOTAL VMT
   BY TRIP TYPE: WEEKDAYS AND WEEKENDS
j '
Trip
Home -
Work -
Home -
Home -
Other -
Type
Work
Otherd
Shop
Other
Otherd
Average
Length
(Miles)
10
7 1/2
3
6 1/2
6
1
AVERAGE WE
i •
Percent Of
Vehicle Trips
21%
9%
14%
28%
28%
IEKDAY
Percent Ofc
Total VMT
32%
10%
6%
27%
25%
AVERAGE W
Percent Ofb
Vehicle Trips
9%
2%
20%
41%
28%
EEKEND
Percent Of(
Total VMT
15%
3%
10%
44%
2B%
Based on  References  [18]  and  [28].

Updated LARTS  data,  [8]  and  [20].

Calculated  from average  trip  length  and  percentage  by  trip  type.

Fleet vehicle  travel  is  accounted  for  in these  two  categories.
      Table  2.14  compares  average weekday  total  auto  trips  and VMT to average

 weekend  values.   These  results  have  been  obtained  by combining data from

 the  LARTS and  SDC surveys  and by obtaining  actual  traffic  counts  from the
                                     50

-------
Division of Highways (data was used from eight widely separated stations),

[11].  Table 2.14 shows that total auto trips decrease around 22% on weekends

while total VMT decreases around 30%.   VMT declines more than total  trips

because weekend trips generally occur more often in shorter trip categories.

Table 2.14 also shows that VMT reductions are significantly greater  on

Sundays (36%) than on Saturdays (24%).
               Table 2.14  AVERAGE WEEKDAY  vs.  WEEKEND  TOTAL
                                  TRIPS AND VMT;  1973
     Average
     Weekday
     Average
     Saturday
     Average
     Sunday
     Average
     Weekend
                         Total  Autod
                        Vehicle Trips
                        Million Trips
                           Per Day
21.9
                   Percentage of Weekday
 84%
 72%
 78%
                       Total  Auto VMT
                       Million  Miles
                          Per Day
183
                   Percentage of Weekday
                                                                    b  i
76%
64%
70%
       Corrected LARTS  data  [3]  and [20J-

       Based  on  LARTS data,  [8]  and [20],  SDC  survey  data  [18]   and
       [28],  and traffic  counts obtained  from the  Department  of  Trans-
       portation,  [11].
                                    51

-------
     By combining Tables  2.13  and  2.14  into  Table  2.15,  some  approximate

but significant conclusions  can  be made about  substitute driving.   To make

weekend and weekday values  more  comparable,  Table  2.15 gives  weekend and

weekday VMT in various trip  categories, both as  %  of average  weekday VMT.

Table 2.15 shows that VMT in the Home-Work and Work-Other categories decreases

by 30% (as a percentage of total weekday VMT)  from weekdays to weekends.

VMT in the Home-Shop and Home-Other categories increases by a total of 5%.

Basically due to a decrease in business travel,  VMT in the Other-Other^

category decreases by 5%.  Now,  the weekend decrease in total "commuter"

travel is probably around 26-28% (some of the decrease in Work-Other is

due to business).  The resulting increase in Home-Shop, Home-Other, and

Other-Other trips resulting from this is probably around 5-7% (the 5%

mentioned above  plus some increase in the Other-Other category that does
              Table 2.15  COMPARISON-OF WEEKDAY  -  WEEKEND
                            VMT ACCORDING TO  TRIP  TYPE

Trip Type
Home - Work
'Work - Other3
Home - Shop
Home - Other
Other - Other9
TOTAL
VMT As % Of Average Weekday
Weekday
32%
10%
6%
27%
25%
100%
Weekend
10%
2%
7%
31%
20%
70%
Change
-22%
- 8%
+ 1%
+ 4%
- 5%
-30%
        Fleet vehicles are accounted for in these two categories,
                                     52

-------
not appear because of lessened business travel).   Thus,  as  a first
approximation, a decrease in X% of commuter travel  brings  about a
corresponding Increase of around I/5X in substitute driving, so that
the overall decrease is only around 4/5X.
     Table 2.16 presents LARTS data on average automobile  occupancy rates,
[8].   Values are given for weekdays and weekends  and for each  of the  trip
type categories.  The striking feature of this table is  the low average
occupancy rate, 1.1, of Home-Work and Work-Other  trips.
           Table  2.16  CURRENT AUTOMOBILE OCCUPANCY RATES IN THE
                                   SOUTH COAST BASIN
Average
Weekend Day
Weekday
Home-
Other
2.0
1.7
Other-
Other
1.6
1.3
Work-
Other
1.1
1.1
Home-
Work
1.1
1.1
Home-
Shop
1.5
1.4
Total
1.8
1.4
  Reference  [ 8]•

 2.2.7   An  Approximate  Air Quality-Emission  Level  Relationship
      In order  to  ascertain the  effectiveness of an episode contingency plan
 (in  preventing "significant harm"  air pollution levels) one must know the
 relationship between air  quality and the emission level changes brought
 about  by the plan.  This  section develops an approximate relationship for
 the  Metropolitan  Los Angeles AQCR.
     In general,  air quality-emission level relations for episode control
 tend to be  less than proportional.  For example, a 30% reduction in emissions
 on a given  day  tends to bring about less than a 30% improvement in air

                                    53

-------
quality.  This effect is due to residual pollution left from previous days

emissions.  The greater the "background" pollution, the less will air quality

on a given day respond to emission reductions on that day.  In Los Angeles,

the residual effect may be especially significant for CO and N02 episodes.

The highest CO and N02 levels occur during the winter months, (see Figure 2.6),

and an  air mass can remain in the Los Angeles basin for up to two or three

days during the winter season, [17].


     Photochemical oxidant is greatest during the summer and early fall.

Fortunately, during the summer, the strong sea breeze tends to sweep the

polluted  air nearly completely across the basin on a daily basis.  Although

oxidant episodes sometimes occur on successive days, this is usually because

the same  meteorological pattern occurs on successive days and not because the

same air  mass remains in the basin for long periods.  However, residual

pollution is still significant.  Further, with oxidant, there is a separate

important factor which leads to a less than proportional air qual ity-emis'sibn

level relationship.  The non-linear chemical  nature of the oxidant-hydrocarbon

relationship appears to be inherently a less  than proportional dependence

when both RHC and NOX are reduced by the same percentage, [21],  [24], [27].


     In this study, in order to estimate how  air quality will  respond to

episode emission reductions, an analysis will be made comparing weekday

pollution levels for CO and oxidant to weekend levels.   Table 2.14 showed

that VMT is  24%, 36%, and 30% lower on average Saturdays, Sundays, and

weekends than it is on weekdays*.   Since CO emissions result almost entirely
  There is also a spatial redistribution of traffic.  On weekdays  the
  pattern is concentrated at central business districts.  On weekends
  the pattern is concentrated more at the beaches and outlying areas
                                    54

-------
from traffic, it will be assumed that they are reduced as VMT.  How
stationary sources  of hydrocarbons change on weekends has not been docu-
mented.   As a first approximation, it will be assumed here that total  RHC
emission levels also change as VMT.
     Only two existing sources were found comparing weekend to weekday
pollution levels in Los Angeles, [4], [23].   A 1966 paper by Shuck,  Pitts
and Wan, [23], indicates that morning peak carbon monoxide levels  in Downtown
Los Angeles were around 20% lower on weekends than on weekdays from 7-62
to 6-64.  This paper also presents some data on average oxidant by day of
week.  Calculations with this data have yielded the results shown  in
Table 2.17 .  Table 2.17 indicates that near the coast (with low max. oxidant)
there is little, if any, improvement on weekends.  Possibly this is  due to
a shift in traffic toward the coast.   However, further inland (where
oxidant is greatest) there is significant improvement on weekends, around
10 to 15%.  At each station except Lennox, pollution reductions are
slightly better on Sundays than on Saturdays.
     The California Air Resources Board, [4], indicates that oxidant values
are around 15-17% lower on average Sundays than on average weekdays  at
inland stations.  These results are significant since they include ten years
of data for July through September.
     To check these conclusions, data for 1970 through 1972 were examined
in the present study, [5].   CO data were taken for December, January,  and
February, the months with maximal primary contaminants.  Oxidant data  were
taken for June through September.  Tables 2.18 and 2.19 summarize  the
results.  Table 2.18 shows that for Lennox, Downtown Los Angeles,  and
Pasadena, (three of the stations which experience the highest values for

                                     55

-------
            Table  2.17  AVERAGE WEEKEND DAY AND WEEKDAY MAX. OXIDANT:
                                FOUR STATIONS, 7-62  to 6-64
Station
Lennox
(Inglewood)
Downtown
Los Angeles
Pasadena
Azusa
Average Weekday
Value (PPM)
.08
.13
.17
.17
Average Week*
Percentage of A
Saturday
106%
96%
87%
90%
2nd Day As
^erage Weekday
Sunday
106%
94%
83%
86%
Average
Weekend-
Weekday
Change
+ 6%
- 5%
-15%
-12%
-a
c:
fO
QJ
O
(/)

Q
 rO
  V
      Reference  [23].


   primary pollutants), weekend CO values averaged from 13-27% lower than weekday


   values during this period.  This tends to confirm the previous estimate of


   20%.  However, Table 2.19 shows that weekend oxidant values at inland stations

   averaged only 6-11% lower on weekends than on weekdays.  There is less of an


   effect than the previously reported 10-17% change.  Of course, the results in


   Tables 2.18 and 2.19 can contain significant statistical error since the

   sample size (around 50 weekend-weekday comparisons) is not very large.


      Summarizing, at stations which experience maximal pollution levels,


   CO concentrations tend to average approximately 20% lower on weekends and


   oxidant concentrations tend to average approximately 15% lower on weekends.


   Assuming a 30% reduction in emissions on weekends this indicates that, due

   to residual pollution, episode control is only around 2/3 effective for


   primary pollutants.  Due to residual pollution and nonlinearities in the
                                      56

-------
          Table 2.18 AVERAGE  DAILY MAXIMAL ONE HOUR  CARBON MONOXIDE AT
                           THREE  STATIONS, WEEKENDS vs. WEEKDAYS:
                               December  -  February, 1969-1972
Station
Lennox
Downtown
Los Angeles
Pasadena
Average Weekday
Value (PPM)
21
17
15
Average Weeker
Percentage of A\
Saturday
91%
73%
90%
d Value As
^erage Weekday
Sunday
83%
73%
82%
Average
Weekend-
Weekday
Change
-13%
-27%
-14%
   Data  obtained  from  California  Air  Resources  Board,  California Air Quality
   Data,  Vol.  II,  III,  and  IV.
               Table  2.19  AVERAGE  DAILY  MAXIMAL ONE  HOUR OXIDANT AT
                             FIVE STATIONS, WEEKENDS  vs. WEEKDAYS:
                                 June  -  September, 1970-1972
Station
Downtown
Los Angeles
Reseda
Pasadena
Azusa
Riverside
Average Weekday
Value (PPM)
11
15
21
23
23
Average Week
Percentage of Av
Saturday
105%
92%
94%
96%
92%
,end As A
'erage Weekday
Sunday
113%
94%
90%
92%
86%
Average
Weekend-
Weekday
Change
+ 9%
- 7%
- 8%
- 6%
-11%
"O
Q)
O

ro
-t->
CO
•i—
Q

CD
  ^
      Data obtained from California Air Resources Board, California Air Quality
      Data, Vol. II, III, and IV.
                                       57

-------
oxidant-precursor relation,  episode control  is  only around 1/2 effective for


photochemical  pollutants.  Accordingly,  it will  be  assumed in this  work that


on a given day,  an X% reduction in  CO  or N0,,emissions  will  result in only a
.. .	. ,t* .. .—  .. -. . j/	•	— • •	—	     *          A
                                                          An
reduction in
2/3X% reduction in CO and [^concentrations on that day.


RHC emissions on a given day will  be assumed to bring about only 1/2X% reduc-


tion in maximal oxidant values on  that day.  It should be emphasized that


this is a very approximate relationship.   It is only intended to be a "first


order", linear correction to account for  some obvious and important non-
                                                   t

linearities in the air quality relationship for episode control.



    By using the emission inventory presented in Table 2.5, VMT reductions


can be translated into emission reductions.  In making this calculation the


effect on RHC emissions from auto  tank filling at service stations should be


included, (i.e. reducing VMT by a  given % reduces auto tank filling emissions


by the same %).  These calculations have  been performed and are presented in


Table 2.20.   Using the approximate air quality-emission level  functions




                                Table  2.20.

          THE  EFFECT  OF VMT  REDUCTIONS  ON  TOTAL  EMISSION  LEVELS*

                          (For Base  Year:   1974)
VMT Reduction
Reduce LDMV VMT by X
Reduce HDMV VMT by X
Reduce both LDMV and HDMV
VMT by X
Correspond
RHC
.62X
.10X
.72X
ing Emission
CO
.85X
.08X
.93X
Reduction
NOX
.60X
.10X
.70X
      *This  includes  the  effect on  RHC  emissions  at  service  stations
        from  vehicle tank  filling.
                                   58

-------
above, VMT reductions can also be translated into air quality improvements,

These results are presented in Table 2.21.
                                Table'2.21
            THE APPROXIMATE EFFECT OF CONTINGENCY PLAN EMISSION
                     AND VMT REDUCTIONS ON  AIR QUALITY
Emission Reduction
Reduce RHC Emissions by X
Reduce CO Emissions by X
Reduce NO Emissions by X
X
VMT Reduction
Reduce LDMV VMT by X
Reduce HDMV VMT by X
Reduce LDMV and HDMV
VMT by X
Correspond!' r
Oxidant
%X
--
--
g Pollutio
CO
__
2/3X
—
n Reduction
N02
—
—
2/3X

.SIX
.05X
.36X
.57X
.05X
.62X
.40X
.07X
.47X
                                     59

-------
2.3  SPECIAL CONTINGENCY PLAN CONSIDERATIONS  FOR THE LOS  ANGELES AREA
     Section 2.1  described the main elements  of general  contingency plans.
These were episode criteria,  control  strategies, surveillance procedures
and operational  procedures.   Of course,  the details  of these elements should
depend on the specific situation for which  a  contingency  plan is formulated.
For instance, the details would be much  different for an  S02 episode plan in
Chicago than they would be for an oxidant plan  in San Francisco.
     Section 2.2 reviewed the air pollution problem  in the Metropolitan
Los Angeles AQCR.  This review has established  certain special  considerations
which should be accounted for in a contingency  plan  for Los Angeles.   These
special considerations are summarized below.
Episode Criteria
     •    For oxidant, the plan should center around preventing the
          .40 PPM four hour  average harm level.*  This level  is exceeded
          in the Metropolitan Los Angeles AQCR  on the average around
          3 to 5 times per year.   The .60 PPM two hour average and the
          .70 PPM one hour average are exceeded much less frequently.
          Furthermore, it is  virtually certain  that  if the one or two
          hour averages are  exceeded  that the four hour average also will
          be exceeded.
     e
Forecasting (feedforward)  type  criteria  should  be adopted,
especially for oxidant.   Vehicular  traffic  is the principal
source of all  three  types  of  emissions under consideration  (RHC,
N0x and CO),  and effective traffic  abatement seems to require
*This assumes the EPA four-hour-average  significant oxidant harm level  for
 oxidant is not changed.
                                    60

-------
          that an  alert be called  before  the  morning  rush  hour.  For oxi-



          dant alerts,  forecasting is  absolutely  necessary.   Even  if sources



          could be instantaneously shut down  when oxidant  started  to reach



          danger levels, oxidant levels still  might continue  to  increase.




     •    Feedback criteria might  be  acceptable for CO  and NQp.  Forecasts



          of CO and NCL levels  are less accurate  than forecasts  of oxidant



          levels.   Also, CO and NOp levels  respond more quickly  to a source



          shutdown.  If an acceptable  method  of significantly reducing



          traffic  with  little advanced notice could be  derived,  feedback



          criteria for  CO and NO^  could be  sufficient.   These issues, as



          they pertain  to CO and N0?,  are not very critical,  however, since



          it is extremely unlikely that CO  or N02 will  ever reach  signifi-



          cant harm levels in this decade even without  a contingency plan.




Control  Strategies




     o    Control  measures should  be  instituted for both stationary and



          mobile sources.   Vehicular  emissions must be  controlled  because



          they constitute the greatest single source  category.   Gasoline



          powered  motor vehicles (LDMV's, HDMV's,  and motorcycles), account



          for 77%* of RHC, 96%  of  CO,  and 74% of  N0v  emissions in  the
                                                  X


          Los Angeles area.  Stationary sources account for less of the



          total; however,  they  are significant and they often represent



          very concentrated sources of emissions.  It is often simpler to



          control  a few concentrated  sources  than  it  is to control many,


                  j/

          disperse'automobiles.  Thus, stationary sources  should be
   is  includes  automobile  tank  filling at gas stations





                                    61

-------
included although their aggregate contribution is much less
than mobile sources.
As indicated in Section 2.2.7, to attain a given percentage reduction
in air pollutant concentration on a given day, a disproportionately
greater percentage reduction in emissions is required on that day.
This is due to leftover or residual pollutants remaining from the
previous day's emissions.   The magnitude of this residual  and of
the disproportionality vary from day to day, depending on  meteoro-
logical conditions.   As a  first approximation, it may be assumed
that an X% reduction  in CO or NO  emissions on a given day will
           ~       ~J~~  "~           X*  """*~"~"—————
result in a 2/3X% reduction in CO and NO,, air pollution levels.
For oxidant, due to  residual pollutants as  well  as to the  non-
linearity of the precursor relationship (RHC—>-Oo formation) s
an X% reduction in RHC emissions on a given day will  result in
only a 1/2X% reduction in  oxidant levels.
Control measures should be implemented on.a region-wide basis,
with the exception of Ventura and Santa Barbara counties.   Air
pollution in Ventura  and Santa Barbara Counties is much less
severe than air pollution  in the other four counties  of the
Metropolitan Los Angeles AQCR, (See Table 2.3).   In fact,  air
pollution alerts should not be expected to  occur in those  two
counties.  Furthermore, Ventura and Santa Barbara counties do
not contribute significant pollutants to the other four counties.
Thus, a division of  the total region into a four county control
region (Los Angeles,  Orange, Riverside and  San Bernardino), with
Ventura or Santa Barbara to be added only under special conditions,
seems called for.  However, any further subdivision of the
                           62

-------
          region for episode control  does  not seem practical.   Implementing



          control  on a subregion basis would require  prediction of  the



          general  trajectories of air parcels that will  reach  alert levels,



          (control  could then be initiated in the  regions  along the



          trajectories).  Reliable forecasting of  this  type  seems doubtful.



          The four county region should be treated as a  whole  (unless an



          acceptable trajectory forecasting scheme is devised).




Surveillance Procedures




     ©    Appropriate air monitoring  surveillance  can be maintained with



          existing air monitoring stations.  As shown in Figure 2.7,



          numerous air monitoring stations exist in the  Metropolitan



          Los Angeles AQCR.   The local APCD's, in  particular,  conduct



          an extensive, continuous monitoring program.   Adequate air



          monitoring data is generated in  the basin.  The  key  problem



          as to air quality  surveillance might be  to  promote enough



          cooperation between the various  agencies involved  so that



          reliable communication channels  are established.




     9    For source activity surveillance, it may be possible to rely  on



          expanding present  local and state programs  during  episodes.



          The county APCD's  have considerable experience in  monitoring



          stationary source  activities.  The California  Department  of



          Transportation (DOT) maintains continuous traffic  counts  on



          certain  freeways and state  highways.  These facilities would



          be extremely useful  in monitoring source response  during  an



          episode.   Again, this will  depend on establishing  a  cooperative



          working  association with local and state agencies.
                                    63

-------
Operational  Procedures
     s    A  cooperative working  relationship with  local  and  state  agencies
          would  be  valuable  in operating an episode  contingency  plan.
          As  mentioned above, county and state capabilities  would  be very
          useful  in monitoring air quality as well as surveying  stationary
          and mobile  source  activity levels.  Cooperation also may be
          fruitful  in air quality forecasting and  in control measure
          enforcement and implementation.
     ©    The overall episode program should be flexible and should be
          changed as  new information is obtained by its implementation.
          The technical analysis which supports the contingency  plan will
          contain many uncertainties.  This is especially true for the
          Los  Angeles area because of the complexity of photochemical
          smog (hydrocarbon  reactivities, oxidant-precursor relationships,
          etc.) and because of the necessity to adopt rather stringent
          and  untested traffic control  procedures.   Once the plan is
          actually tried, it may be found not to be effectual enough,
          (in  that it does not prevent significant harm levels).   Alterna-
          tively, it may be found overly effective (in that it reduces
          pollution far below the harm levels  but at a very high social
          cost).   The program should be flexible enough to allow changes
          in the plan as  warranted.
                                   64

-------
                         REFERENCES - CHAPTER 2


 1.  Altshuller, A. P., "Recommendations  on Reactive  Hydrocarbons  (Organics)
     as Related to Photochemical  Oxidant  Formation",  Working  Paper  - EPA
     National Environmental  Research Center, 1973.

 2.  California Air Resources Board, The  State of California  Implementation
     Plan for Achieving and  Maintaining the National  Ambient  Air Quality
     Standards, January 30,  1972.

 3.  California Air Resources Board, California  Emission  Inventory, 1970,
     July 1972.

 4.  California Air Resources Board, Personal  Communication with Technical
     Services Division, September, 1973.

 5.  California Air Resources Board, California  Air Quality Data, Volumes III
     and IV, January 1970 -  December 1972.

 6.  California Department of Finance, Provisional Projections of California
     Counties to 2000, September  15, 1971.

 7.  California Department of Finance, California Statistical Abstract, 1972,
     Sacramento, California, 1972.

 8.  California Department of Transportation,  LARTS Base  Year Report, 1967
     Origin - Destination Survey, December, 1971.

 9.  California Department of Transportation,  California  Statewide Transporta-
     tion Study, 1966 Base Year Calibration Report, May,  1972.

10.  California Department of Transportation,  Advanced  Planning Department,
     District 7, Personal  Communication,  November 1972.

11.  California Department of Transportation,  Data Processing Department,
     Personal Communication  of Data Recorded by  the Los Angeles Freeway
     Surveillance and Control  Project and by Monitoring of State Highways,
     September, 1973.

12.  County Air Pollution Control  Districts (Los Angeles, Orange, Riverside,
     and San Bernardino),  Air Monitoring  Logs  for Oxidant, Carbon Monoxide,
     and Nitrogen Dioxide.

13.  Davidson, Arthur, Meteorologist - Los  Angeles County Air Pollution
     Control District, Personal Communication, August,  1973.

14.  Environmental  Quality Laboratory, Smog:   A  Report  to the People of
     the South Coast Air Basin, EQL Report  No. 4, California  Institute
     of Technology, 1972.
                                    65

-------
15.   Foon,  Julian,  Data  Processing - Los Angeles County Air Pollution
     Control  District, Personal Communication, August, 1973,

16.   Gray,  G.  E., Watkins, R. 0., and Kirchner, F. F., Can Vehicle Travel
     Be Reduced  20  Percent in the South Coast Air Basin?  Report of the
     California  Deparment of Public Works to the California Air Resources
     Board, January,  1973.

17.   Holmes,  R.  G.,  Kauper, E. K., Street, A. B., and Taylor, J. R., "Air
     Flow Studies of  Days of Heavy Smog in Los Angeles", Presented at the
     49th Annual Meeting of the Air Pollution Control Association, Buffalo,
     New York, May,  1956.

18.   Kearin,  D.  H.  and Lamoureaux, R. L., A Survey of Average Driving
     Patterns in the  Los Angeles Urban Area, System Development Corporation
     Technical Memo  (TM(L) - 4119/000/01), February 1969.

19.   Kyan,  C.  P. and  Seinfeld, J. H. , "Real Time Control of Air Pollution",
     AICHE  Journal,  Vol. 19, #3, 1973.

20.   Los Angeles City Department of Traffic, Transportation Planning
     Division, Staff  Report:  LARTS Trip Type Distributions, Supplement
     Report,  July 6,  1973.

21.   Los Angeles County  Air Pollution Control District, Profile of Air
     Pollution Control,  1971.

22.   Neiburger,  Morris and Edinger, James, "Meteorology of the Los Angeles
     Basin, "Report  No.  1 of the Air Pollution Foundation (Los Angeles:
     Southern California Air Pollution Foundation, 1954).

23.   Schuck,  E.  A.,  Pitts, J. N., and Wan, J. K. S., "Relationships Between
     Certain  Meteorological Factors and Photochemical Smog", International
     Journal  of  Air  and  Water Pollution, Vol. 10, 1966.

24.   Schuck,  E.  A.,  et.  al., "Relationship of Hydrocarbons to Oxidants in
     Ambient  Atmospheres", Journal of the Air Pollution Control Association,
     Vol. 20,  May 1970.

25.   Southern California Association of Governments.  Employment Projections
     for the  SCAG Area,  (unpublished) November 1972.

26.   Thomas,  George,  Engineering Department of the Los Angeles County Air
     Pollution Control District, Personal Communication, August 1973.

27.   Trijonis, John,  An  Economic Air Pollution Control Model - Application:
     Photochemical  Smog  in Los Angeles County in 1975, Ph.D. Thesis,
     California  Institute of Technology, Pasadena, California, May 1972.

28.   TRW Transportation  and Environmental Operations, Transportation Control
     Strategy Development for the Metropolitan Los Angeles Region, EPA
     Contract No. 68-02-0048, January, 1973.

29.   TRW Transportation  and Environmental Operations, Air Quality Implementation
     Plan Development for Critical California Regions:  Summary Report,  EPA
     Contract No. 68-02-0048, July, 1973.

                                   66

-------
      3.0  REVIEW AND EVALUATION OF EXISTING EMERGENCY  EPISODE  PLANS



     The severity of the California air pollution situation,  especially  in

the Los Angeles region where it is most acute, has prompted the formulation

of several emergency action plans for use in case of episode  alert  condi-

tions.  Plans have been developed at the local, state,  and federal  levels,

by the APCD's, ARE, and EPA.  This chapter reviews each of these three

programs, presenting -both merits and shortcomings of the plans.

     The Los Angeles County APCD alert program is the oldest, and probably

most familiar, plan which has been developed for the region.   In Section 3.1,

it is reviewed as being representative of abatement plans proposed  by several

county APCD's.  The ARE plan, as of this writing, is in draft stages and has

not been officially adopted; it is based in large part, on a  technical support

study conducted by the California Office of Emergency Services  (OES).  Both

the preliminary ARE plan and the OES Report are discussed in  Section 3.2.

Recently, EPA-Region IX designated a task force to examine traffic  abatement

strategies and to submit a report on recommended actions.  This report is

reviewed in Section 3.3.
                        »
     In addition to the EPA-Region IX Task Force on Traffic Abatement During

Air Pollution Episodes Report, an actual contingency plan has recently been

formulated by EPA-Region IX.  This is described in Appendix B.   Also presented

in Appendix B are an effectiveness evaluation of the EPA strategies and  a

comparison of the EPA contingency plan to a plan recommended  by this study.
                                     67

-------
3.1  LOCAL AIR POLLUTION CONTROL DISTRICT EMERGENCY PLANS
3.1.1  Plan Description
     Presently, four counties in the South Coast Air Basin, Los Angeles,
Riverside, San Bernardino, and Santa Barbara have adopted contingency plans
for air pollution episodes, [1].  These plans are not designed to prevent
Federal significant harm levels but are rather designed "to prevent the
excessive buildup of air contaminants and to avoid any possibility of a
catastrophe caused by toxic concentrations of air contaminants", [6].
However, in the view of the APCD's, the possibility of such a "catastrophe"
really occurring is "extremely remote", [6].  The other two counties, Orange
and Ventura, are awaiting the preparation of the statewide contingency plan,
developed by the Air Resources Board, before they adopt emergency procedures,
[1J.
     The Los Angeles APCD plan was the first and is the most complete of the
four plans so far adopted.   It has served as the model, for both the San
Bernardino and Santa Barbara episode programs.   The Riverside plan contains
no control phases and is basically just a community warning procedure.  Thus,
the following discussion of local  control  plans will  concentrate on the L.A.
APCD plan.
     As mentioned above, an episode control  contingency plan consists of
four basic elements:   episode criteria, control strategies, operational
procedures,  and surveillance procedures.   The following is a description of
the local  APCD emergency plans in  terms of these four elements.
     Episode Criteria
     The L.A.  APCD emergency air pollution program divides an episode into
three degrees  of severity called alert stages,  [6].  Each alert stage is
                                    68

-------
defined by threshold concentrations of various air pollutants.   These
thresholds, given in Table 3.1 are indicative of the health  threat  to  the
general public.  For example, in a first stage alert, pollutant concentra-
tions have supposedly reached a level  "close to the maximum  allowable
concentration for the population at large, a point where preventive action
is required".  A second stage alert indicates that air contamination has
reached levels which are beginning to pose a serious health  threat.  A third
stage alert means pollutants exist in concentrations which are  a "dangerous
health menace", [6].  These three stages form the basis for  a simple feed-
back control system; various actions are taken whenever pollutant concentra-
tions actually reach certain levels.

          TABLE 3.1.  LA COUNTY APCD  ALERT STAGES  FOR OXIDANT,
                   CARBON MONOXIDE, AND NITROGEN  OXIDES*
Pollutant
Oxidant
Nitrogen Oxides
Carbon Monoxide
First Alert
(PPM)
0.5
3
50
Second Alert
(PPM)
1.0
5
100
Third Alert
(PPM)
1.5
10
150
     *The  sum of nitrogen  dioxide  and  nitric  oxide.
     Control Strategies
     At each phase of the alert, certain actions are taken in order to reduce
pollutant concentrations.  These control measures, which constitute a "control
strategy" for each stage, are presented below in outline form, [6]:
                                    69

-------
     I.    First Stage Alert
           1.   Prohibit refuse incineration.
           2.   Request via the news  media  for voluntary cessation
               of all  non-essential  vehicle  use.
     II.   Second Stage Alert
           1.   Continue actions outlined  under first stage alert.
           2.   Notify industries emitting  contaminants  responsible for the
               alert to institute (pre-planned)  shutdown procedures.
           3.   Close additional industrial and commercial  establishments
               and/or stop vehicular travel  if the  Air  Pollution Control
               Board and Emergency Action  Committee deem it necessary.
     III.  Third Stage Alert
           1.   Continue actions outlined  under first- and  second-stage
               alerts.
           2.   Take actions as deemed necessary and appropriate by the
               Governor of California.
     IV.   Alert Termination
           1.   Verify that contaminant  concentrations have fallen below
               first stage alert criteria.
           2.   Verify that available meteorological and scientific data
               indicate that pollutant  concentrations will not immediately
               rise above first stage levels.
           3.   Terminate episode if  the above is  true.
     Stage II  (measures 2. and 3.) and  Stage  III  (measure  2.)  institute the
principal active (as opposed to voluntary) control  measures in the APCD plan.
Stage II, measure 2.,  implements pre-planned  shutdown procedures for certain
                                    70

-------
stationary sources which emit the alert-causing contaminants.   The types  of
industries requiring shutdown plans in Los Angeles County are  summarized  in
Table 3.2, [6].
            TABLE 3.2.  TYPES OF STATIONARY SOURCES REQUIRING
                 SHUTDOWN PLANS - LOS ANGELES COUNTY APCD
SOURCE TYPE EMISSIONS

Petroleum Refinery
Asphalt saturators, asphalt paving
man. plants
Bulk loading facilities for tank vehicles
Chem. plants reacting or producing
organic liquids or gases
Paint laquer, or varnish man. plants
Rubber tire man. or rubber machinery
plants
Automobile assembly
Metal melting, refining or smelting
Rock wool manufacturing plants
Glass or frit manufacturing plants
Fossil fuel fired steam generating plants
Container man. or container decorating plants
Large dry cleaning
Large printing plants
RHC
X
X
X
X
X
X
X




X
X
X
CO
X
X





X






NO
X
X


X



X
X
X
X



References:   [6],  [7]
Table 3.3 compares total emission reductions from these plans to total
stationary source  emissions, [10].  Noting that much of total stationary
                                     71

-------
     source  emissions result from disperse commercial and residential sources,

     which are not controlled in the plan, the stationary shutdown procedures

     have a  considerable effect.
Table 3.3  THE  EFFECT  OF  LOS ANGELES COUNTY SHUTDOWN
           PROCEDURES  FOR STATIONARY SOURCES
                                                    RHC
Contaminant

 CO            NO
                                                                                x
Total Emissions from all mobile and
Stationary Sources (Metropolitan Los Angeles
AQCR, 1973)
Percentage of Total Emissions Resulting
from Stationary Sources, (Not counting auto
tank filling)
Percentage of total emissions reduced, by APCD
Shutdown Plans, (assuming LA County plan is
applied to whole basin)
1-rntons
1250day
14%
4~%
~ Q nnn tons
9)000 day
1 1/2%
1%
1450 JSDi
day
23 \m
5%
         The  emergency vehicular control program is not as well organized as the

     stationary source program.  At Stage II, the 300 companies which require

     stationary source shutdown procedures will  also be asked to implement a pre-

     planned transportation control program (basically car pooling), [6].  This

     is  of  very limited effectiveness, especially since action is not taken for

     oxidant alerts until the afternoon of that alert.  Also in Stage II, a "work

     holiday"  could be declared (this suffers from the same limitation), [10].

     In  Stage  III, at the request of the Governor, a ban on all non-emergency

     traffic might be called.  Enforcement would be monitored by the Sheriff's

     office and other police departments.

         As to location, strategies are implemented strictly on a county by

     county basis.  That is, control action is taken in a  given county when

     pollutant levels in that.county reach an alert stage.  There is presently

     no  mechanism for regional cooperation in episode control procedures.  As

                                         72

-------
mentioned previously, the strategies for San Bernardino and Santa Barbara
counties are essentially the same as for Los Angeles County.
     Surveillance Procedures
     The discussion of general contingency plans, (Section  2.1), described
three general surveillance requirements:  air quality monitoring, source
surveillance, and overviewing other agencies.  The surveillance procedures
of the APCD plan fall into two of these categories:   air quality and source.
Each APCD is not concerned with the responses of other sub-agencies, since
there are none under it with air quality responsibilities.
     The L.A. APCD monitors air quality through a permanent network of
continuous atomospheric sampling stations.  Routinely, these stations
continually provide quite detailed data on air quality and meteorology.
No special monitoring techniques are automatically instituted in the event
that an episode occurs, [9],  However, the APCD does possess the capability
to establish temporary stations, using mobile equipment, if it is deemed
necessary, [8].
     Routine source surveillance is also conducted by the L.A. APCD.  This
is the responsibility of the enforcement division.  This division includes
uniformed enforcement officers in patrol cars as well as plain clothes
inspection engineers, [9].  During an alert, the APCD enforcement group
is put on special duty, and sheriff's officers as well as other police
can be called on for assistance.
     Operational Procedures
     The APCD emergency episode plan, as described in the LA County APCD
Rules and Regulations,  contains fairly explicit operational procedures,
[6].   Figure 3.1 has been  constructed to summarize the basic organizational
                                    73

-------
HIGHWAY
PATROL
                                         EMERGENCY ACTION
                                            COMMITTEE
                                         REQUEST
                                         RECOMMENDATIONS
                                         (phone)
                                      A
                      ENFORCEMENT ASSISTANCE
f
                       (phone)
                  COUNTY
                 SHERIFF
           (teletype,
                     V
    radio)
 LOCAL
POLICE
                                     ANNOUNCE
                                     (phone
                                             W
                           *
                           *
                            ALERT
                       PUBLIC
                        MEDIA
                 SHERIFF
              SUBSTATIONS
                                                     * *
                                                      GOVERNOR
                                                         QF
                                                       CALIF,
                                                     REQUEST STATE
                                                     OF  EMERGENCY
                                                     (phone)
                                                                     * * *
                                                           APCU
                                                          VOLUNTARY
                                                         ABATEMENT
                                                           CONTROL BURNING (media)
  * *   SHUTDOWNS
       (private  radio)

   * #
    POSSIBLE
    DIRECT
    MEASURES
  W (unspeci fi ed)
                                                        MOTOR
                                                       VEHICLES
                                                                                     ^ '
                      STATIONARY
                        SOURCES
   *  FIRST STAGE ALERT
  **  SECOND STAGE ALERT
* * *  THIRD STAGE ALERT
                        Figure  3.1   County APCD Emergency  Episode  Operations

-------
structure and the communication channels.  The phasing of operational



procedures is outlined below.




     I.    Preparation




           1.  Request shutdown plans from business, industrial  and



               commercial establishments emitting toxic contaminants.



           2.  Develop shutdown plans and vehicle abatement plans into



               a coordinated control program.



           3.  Appoint a Scientific Committee to serve as a consulting



               body to the APCD.



           4.  Appoint an Emergency Action Committee to advise the



               Air Pollution control officer during an episode.




     II.   First Stage Alert




           1.  Notify sheriff and other law enforcement officers.



           2.  Notify stationary source of possible shutdown if



               second stage alert is reached.



           3.  Notify public via media of alert to request cooperation.



           4.  If second stage alert appears eminent, notify Emergency



               Action Committee and media.



     III.  Second Stage Alert




           1.  Notify appropriate stationary sources.



           2.  Supply necessary information to Emergency Action



               Committee and receive recommendations.



           3.  Notify media.



           4.  Notify sheriff and other law enforcement officers of



               necessary enforcement steps.
                                    75

-------
    IV.     Third Stage Alert




           1.  Notify Governor of California to receive his recommendations



           2.  Notify media.




     V.     Termination




           1.  Notify sheriff.



           2.  Notify public via media.




3.1.2  Evaluation:   Preventing Significant Harm Levels




     The APCD emergency episode plan will  not prevent oxidants, carbon



monoxide, and nitrogen dioxide from reaching federal  "significant harm"



levels during an air pollution episode.   The reason for this is simple.



The plan was  not designed with the federal  significant harm levels in mind.



Rather,  it was constructed to avoid what local  officials term an air



pollution "catastrophe"  or "disaster".   These  levels differ considerably



from the significant harm levels established by the EPA.




     Table 3.4 compares the APCD alert  levels to the  federal significant



harm levels.   It can be seen that the APCD concentration specifications



are of a different  type than federal  specifications.   The APCD criteria



are given in  terms  of instantaneous concentrations  while the EPA's are



measured in  terms of one to twenty-four  hour averages.  Also, the nitrogen



oxide levels  are specified differently.   The EPA uses NOp while the APCD



uses NO  .   In order to make the specifications  more comparable, the APCD
       A


levels have  been adjusted to reflect a  two-thirds conversion rate from



total  NO  to  N09.
        X      c.
                                    76

-------
                 TABLE  3.4.  COMPARISON OF L. A.COUNTY APCD

            ALERT LEVELS  TO THE  FEDERAL SIGNIFICANT  HARM LEVELS



. Pollutant
Oxidant


Carbon
Monoxide

Nitrogen
Dioxide
LEVELS
Federal
Significant Harm
Level
.70 PPM 1 hour average
.60 PPM 2 hour average
.40 PPM 4 hour average
125 PPM 1 hour average
75 PPM 4 hour average
50 PPM 8 hour average
2.0 PPM 1 hour average
.50 PPM 24 hour average
APCD ALERT LEVELS
(PPM)*

Stage 1

0.5


50

**
2.0

Stage 2

1.0


100

**
3.3

Stage 3

1.5


150

**
6.6
 *  APCD Alert Levels are for instantaneous  values.

**  APCD criteria  are for total  nitrogen oxides-.   It is assumed here
    that maximal nitrogen dioxide  levels are  two-thirds of maximal
    total  nitrogen oxide  levels.   This  assumption  is supported by
    air  monitoring data  in  the  Los Angeles region.
      Examining Table 3.4 and noting that the first mandatory control  stage

 of the APCD plan is the second stage., (the first prohibits burning and calls

 for voluntary traffic abatement), it becomes obvious that in most cases, the

 APCD plan cannot prevent pollutants from reaching significant harm levels.

 It is almost certain that if the oxidant concentration exceeds an instanta-

 neous level of 1.0 PPM (the APCD second stage alert level) that it will have

 already exceeded the .7 PPM-one hour, .6 PPM-two hour, or .4 PPM-four hour

 significant harm levels.   The same is true for carbon monoxide; the four

 hour 75 PPM or the eight hour 50 PPM level will  likely be exceeded by the
                                     77

-------
 time  a  second  stage alert is declared.  In fact, the only significant  harm
 level adequately covered by the second stage alert criteria appears to  be
 the 125 PPM  CO concentration for one hour.  This does not necessarily  mean
 the APCD plan  is a poor plan; it is just a reflection of the fact that  the
 APCD  plan was  not designed to prevent federal significant harm levels.
 However, it  is doubtful that the APCD plan could be made much more effective
 in preventing  air pollutants from reaching significant harm levels by  simply
 lowering the criteria  levels.  The APCD plan is based on feedback criteria
 rather  than  a  forecasting system.  As noted in Chapter 2, feedforward
 (predictive) initiation criteria are essential for oxidant episodes;
 reducing emissions after the concentration of oxidants has reached a
 given level  does not have significant effect on the continued short-term
 accumulation of oxidant.  Further, most transportation control measures
 require a certain amount of time to implement; this again requires predictive
 initiation criteria.
      The APCD  plan also has room for improvement since control measures are
 presently well  defined and operational only for certain stationary sources.
 As noted in Table 3.3, preplanned shutdown procedures are available for
 stationary sources which account for 4% of total RHC, 1% of total CO and
 5% of total  NO    However, 78% of RHC, 99% of CO and 76% of NO  are emitted
                                                              A
 from transportation related sources, (including auto tank filling).  Presently,
 emergency control  procedures for mobile sources are not as well organized
operationally as those for stationary sources.  The mobile source control
 procedure should improve however, due to present ARB-APCD work on contingency
 plan  development.
     The APCD does  have certain  capabilities which would be very useful in
any contingency plan.   For instance, it has an extensive, continuous air
                                    78

-------
monitoring system for atmospheric surveillance.  Also, the APCD maintains



an enforcement division, with a staff of approximately 100 officers, to



monitor source activity,  [9].  Further, the APCD has an established, well



organized communications system along with a well defined set of operational



procedures for stationary source control.  These capabilities are very



valuable assets for contingency plan action.




3.2  AIR RESOURCES BOARD - AIR POLLUTION EMERGENCY CONTINGENCY PLAN




3.2.1  Plan Description




     The purpose of the ARB's emergency contingency plan is "to provide the



basis for taking action to prevent or abate air pollutant concentrations



from reaching levels which could endanger the public health", [2],   Under



the plan, local air pollution control districts would have the primary



responsibility for taking the necessary steps to curtail polluting  activities,



However, if the episode is so severe that it is beyond the local  districts'



capability to control, then it becomes the responsibility of the state to



alleviate the conditions.




     The Air Resources Board plan is based in part on a study conducted by



the California Office of Emergency Services, [3].  At present, the  plan is



still in preliminary draft and as yet has not been adopted.  The discussion



below is based on Preliminary Draft No. 2, dated July 23, 1973, [2].




     Episode Criteria




     The ARB emergency contingency plan breaks up an air pollution  episode



into four stages of severity.   As in the LACAPCD plan, these stages are



defined by threshold concentrations being reached (with expectancy  of



persisting or increasing).  Thus, the episode initiation criteria are



basically feedback criteria with only slight forecasting involved.




                                    79

-------
     Of the three pollutants  to  be  considered  in  this  report,  only two,
oxidant and carbon monoxide,  are considered  in the  ARE plan.   N02  is  omitted
The alert concentrations  for  oxidant and  CO  as well  as the  corresponding
"stages" are given in Table 3.5, [2].

                                TABLE  3.5
               ARB CONTINGENCY PLAN EPISODE  CRITERIA LEVELS

POLLUTANT

Photochemical
Oxidant
Carbon Monoxide
AVERAGING
TIME

1 Hour
1 Hour
12 Hours
HEALTH
ADVISORY

.20 PPM
40 PPM
20^PPM
ABATEMENT
STAGE 1
(ALERT)
.25 PPMa
50 PPMa
25 PPM
ABATEMENT
STAGE 2
(WARNING)
.40 PPMa
75 PPMa
35 PPM
ABATEMENT
STAGE 3
(EMERGENCY)
.60 PPMb
100 PPMb
50 PPM
 Reached and  expected  to  approach  the  next  stage  unless  abatement action
 taken.
'Reached and  expected  to  persist for 1  additional  hour  unless  abatement
 action  taken.
General  descriptions  of the  episode  stages  are  outlined below:
     1.   Health Advisory -  Pollutant concentrations  have reached or are
         predicted to reach  a  level  at which  persons with respiratory or
         coronary diseases  should  take precaution against exposure.
     2.   Alert - An alert is called  when  the  concentration of pollutants
         specified for this  stage  is reached  and when it has been predicted
         that air pollutant  concentrations  may  reach the next stage.
                                    80

-------
     3.   Warning - A warning is called when the concentration  of pollutants



         specified for this stage is reached and when  it has been predicted



         that air pollution concentrations may reach the next  stage.




     4.   Emergency - An emergency is called when the conditions  specified



         for this stage are reached, (and in certain instances are expected



         to persist for one hour).




         A stage is assumed to be terminated whenever  the concentration  of



         pollutants which caused the declaration of the episode  has been



         verified to have fallen below the threshold level  for that stage



         and pollutant concentrations are expected to  continue down.




     Control_ Strategies




     The control strategies of the  ARB contingency plan consist  only  of  a



general  outline of what types of control  measures are  to be taken at  each



stage of the episode.  It is left up to the local APCD's to come up with



actual  control measures.   The control strategies can be summarized as



follows:




     I.    Health Advisory (No real  control action)



          1.  Notify persons with respiratory disease  or coronary



              ailments to take precautions.




     II.  Alert (Stage 1  - Voluntary Control)




          1.  Continue actions for  first stage.



          2.  Initiate voluntary traffic and source abatement  procedures.




     III.  Warning (Stage  2 - Mandatory Controls)




          1.  Continue actions of first and second stages.
                                    81

-------
          2.   Initiate curtailment actions  for stationary sources  in
              accordance with pre-established  plans.
          3.   Initiate mandatory traffic  abatement.
     IV.  Emergency (Stage 3 - Further  Mandatory  Controls)
          1.   Continue actions outlined under  first,  second  and  third
              stages.
          2.   If necessary,  close or  shut down all  industry  and  business.
          3.   If necessary,  take action under  provisions  of  Peace  Time
              Emergency Plan of the Emergency  Services  Act.
     V.   Termination
          1.   Terminate control  action  as stages  are  terminated.
     In addition to outlining what actions  should be  taken at  each stage  of
the episode,  the ARB strategy further details  what king of controls it
expects the local  APCD's to  develop.  These can be divided into  two groups;
stationary source curtailment controls  and  traffic abatement plans.
Stationary source controls are to provide for  the immediate  reduction of
industrial, business,  or commercial activities which  emit 100  tons/year
or more of pollutants.   They must include source  identification,  location,
and emission  estimates  along with designation  of  whom to  contact,  shutdown
procedures, and shutdown time estimates.  Traffic abatement  plans  are to
include specific actions to  be taken  at both the  first  and second  stages.
They are to be directed at reducing the reasons for vehicular  traffic rather
than direct traffic control.  Examples  of traffic abatement  measures are
given in the  ARB plan  and are presented in  Table  3.6.
                                    82

-------
                                TABLE  3.6.
         ARB CONTINGENCY PLAN - EXAMPLE TRAFFIC ABATEMENT MEASURES
              STRATEGY
 EPISODE STAGE
Voluntary Reduction in Traffic
1.  Alert
Ban Government Vehicles
Close Admission to Public Recreation
Facilities
Close Government Offices*
Ban Fleet Vehicles - Excluding Gaseous
Fueled Vehicles
Close Admission to Private Recreation
Facilities
Close Admission to Regional Shopping
Centers
Close Schools and Colleges
Close Admission to "Downtown" Retail and
Service Business
Ban Delivery Service of all Non-Perishables
Ban Non-Essential Service Calls
Close Establishments with 100 or more
Employees*
2.  Warning
Close Admission to All Other Retail and
Service Establishments
Close Industrial  Sources
3.  Emergency
Reference:  [2]
   Additional  activities taken when prediction is available.
                                     83

-------
     Surveillance Procedures
     In the area of surveillance procedures, the ARB provides only a minimum
amount of detail.  It does indicate, however, that each local APCD should
maintain at least one air monitoring station within its jurisdiction.  It
also states that ARB will monitor and provide all  necessary meteorological
information.  This lack of detail may be due to the preliminary nature of
the plan.
     Operational Procedures
     The state ARB plan provides for the organizational structure as
diagrammed in Figure 3.2.  Again, the essential  elements are the local
APCD's.  Each APCD is the responsibility of an air pollution control officer
who may or may not be advised by an  Emergency Action Committee.   Coordina-
tion among the various APCD's is achieved through  an Air Basin Coordinating
Council.   Finally, all of this is overseen by the  ARB and  OES.  They advise
the Governor and recommend what actions  he should  take, if any,  under the
Emergency Services Act.
     The  ARB plan has outlined the following operational  procedures  in  the
event of  an  episode.
     I.    Health Advisory
           1.   Notify all  affected APCD's.
           2.   Notify all  monitoring  stations.
     II.   Alert
           1.   Activate ARB emergency action staff.
          2.   Notify  all  necessary  personnel.
          3.   Monitor actions  of local  APCD.
                                   84

-------
CO
01
                       OTHER LOCAL
                          APCD S
                                                      OES
                                                                     MARE RECOMMENDATIONS
                                                                       S EVALUATIONS
                                                                      * *
                                                                   NOTIFY
                                                                         w
INFORM OF CURRENT STATUS
                                          REOJJEST COOPERATION
                                          EMERGENCY ACTIOi'J
                                             COMMITTEE
                                                                            OF  EPISODE
                                                          GOVERNOR
                                                             OF
                                                           CALIF.
                                     ARB
                                                           Recomrnendat ions
                                       ENFORCEMENT ASSISTANCE
                                                        ANNOUNCE
                             ENFORCEMENT
                               AGENCIES
                                                                  ALERT
                       PUBLIC
                        MEDIA
                                                                REQUESTS STATE
                                                                op EMERGENCY
                                           *
                                           * *
                                           * * #
  NOTIFY BOARD
OF EPISODE STATE
                                                                             AP CiJ
                                     VOLUNTARY
                                                                            ABATEMENT
                                                                                                   CURTAIL SOURCES
                                                                                                   CONTROL  OPEN  BURNING
                                                          POSSIBLE  DIRECT
                                                          MEASURES
                                                          (UNSPECIFIED)
     MOTOR
    VEHICLES
STATIONARY
  SOURCES
                                                                                      «  ADVISORY
                                                                                     #*  ABATEMENT STAGES 1 AND 1  (ALERT ii WARNING)
                                                                                   * * *  ABATEMENT STAGE 3  (EMERGENCY)
                                               FIGURE  3.2.  ARB EMERGENCY  EPISODE  OPERATIONS ORGANIZATION

-------
     III.   Warning -  same  as  Alert




     IV.    Emergency




           1.   Notify Governor.



           2.   Same as above.




     More  detailed procedures  are not well  defined  as yet,  probably  because



of the preliminary nature  of  the plan and  because the burden  of  responsi-



bility is  placed on the local  APCD's.




OES Traffic Abatement Plan




     Preliminary to the ARB formulating  an  episode  contingency plan,  a  study



was conducted  by the  Air Pollution  Emergency Traffic Control  Planning



Committee  of the California Office  of Emergency  Services,  (OES),  [3].  The



purpose of this study was  to  consider all  possible  traffic  abatement



measures,  to determine the possible effect  of  each  upon  the air  pollution



problem,  to organize  the most  promising  measures into a  traffic  abatement



plan, and  to make recommendations regarding implementation  of such a  plan.



The OES report served as part  of the technical basis for the  ARB plan,  and



it will be reviewed briefly below.




     The major portion of  the  OES study  was devoted to an  evaluation  of the



cost and effectiveness as  well as the enforcement and implementation  prob-



lems associated with  various  traffic abatement measures.   This evaluation,



summarized in  Table 3.7, resulted in the following  conclusions:




     1. Prediction of air pollution episodes  is essential  to instituting



        effective traffic abatement measures.




     2. The most effective approach to  traffic  abatement  is  to  eliminate



        the  incentive for travel.   Significant  enforcement problems  would



        be associated with any direct traffic controls.



                                    86

-------
                             TABLE 3.7   OES EVALUATION OF TRAFFIC ABATEMENT MEASURES
co
STRATEGY
Work Holiday
Simple San on


Facilities 6.
Reta 1 1 Stores
Licence Plate
Lottery
Close Large
Business
Cormisrcial
Abatement
Plans
Windshield
Stickers
Offices (S.
Contractors)
Vehicles
Gas Rationing
Voluntary
Abatement
Mandatory Car
Pool Ing
Parking Tax
L\-'\t PtrHr
Close Gas


Close Soma Freeway
(Casol In*
Powered)
ENFORCEMENT
PROBLEMS

High



High
Some
High
High
Gov't
Contr.
Maybe



High
High
"1 • '
Public
Put sta.
high
High


ADMINISTRATION
COSTS




"
Low
"
High
High





High


Low


SAFETY
HAZARDS







"-


"
Storage
--


High gas

Low
"",""

3:
E
i- cy
Prod.
Prod.



Prod.
Pred,
Pred.
Pred.
Pred.
Pred.
Preferred




Pred.

Pred. pre
ferred
.ferred
ferred
PROBLEMS OF
INSTRUCTION





High


High bee.
of di f f.
opt ions





-TT7 /





-i
a-
o o
^ 0
< z





!








- -





o
z
z
o
< -1
< 3
Q- Q-




"

—
Some



Sorie


"5rr 	





OVERALL PUBLIC'S
^^ACCEPTABILITY




Med. to
High
Some
Some
Some
Some


High -e.<:
Present
EPA Hear-
ing


Hiqh
Mi " h


HInh
Mi gh

X
o
Q I—
- 31
S O
guaranteed
pay for

guaranteed
wages

Restrict
store -
put bus.
fami 1 les
Yes If no
guaranteed
wage ~lrg.
Against
large bus.
- guar.woge
Some
Wage

System of
A) lotment

wage

Or,; car
fami ly -
Gas. sla.
operator


Sonic
ADVERSE IMPACT
ON NON-MOTORISTS




High

Yes













o
^ s:
0 (—
L st Ess-
ential
services



List of
Exempt
Stores
Months
List of
Exempt
Year
List of
El igible
for rtg,

Exemp t
Year


•T-, 	 •


Months
Month
.in.
-J <
< i—
-1 O
Pay - Hln.
Fine-No
Susp.
V F-
riages

Cuar.pay
no Susp.
- Mln, f inc
Guar.pay
•1 i n . F 1 ne
-Guar.wage
-Mln. fine
susp.
-Wo auth.
close
Agreement
with Fed.



-RT7T 	
T

""


LAWS
REQUIRED









Obi Iga-
tlon



lin. fin
W/out *u
, ,





MAX. POTENTIAL
V.H.T. REDUCTION
70*



6*
Var.
si -
85*
Var.
5* -
201
Var,
» -
70Z
Var. 85S
Max,
5* -
51 -
20*
m
Var.
60t
,55* >t.
p[home
5M
~7
Var.

5J LA
103; SF
n«,.
5'.
o
< 0
< K
70»



M
2J -
tot
5* -
20»

50> »t
MaK.
51 - o — o
M - 1 E <-
2o» ; "
s s
0 —
1/2*

5» - 10*
,. Pool -
151 >t.y
TW
• "hW "o"r
Scg.


10% SF



-------
      3.   When  implemented,  any  traffic abatement
          over  a  relatively  broad traffic influence area, rather than upon a
          smaller area  from  which pollutants are being emitted.
      4.   A  traffic  abatement plan should be progressive in nature  both  in
          its anticipated benefits on pollutant levels and its anticipated
          impact  on  the community.  It should be implemented in stages as
          necessary.
      5.   Local levels  of government should initiate the first stages of
          control.   Only the most severe stage, involving declaration of an
          emergency, would be instituted at the State level.
      The  OES report proposed a  traffic abatement plan that is very
 similar,  in several respects, to the ARB plan, (particularly the
 "example" traffic control strategy of Table 3.6).   First, the OES
 report, as  the ARB  plan, divides the episode strategies into four
 levels of severity.  Second, it used the same threshold criteria as the
 ARB  plan.   Finally, the control  actions to be initiated at each level
 fit  into  the same kind of control philosophy as expressed by the ARB
 (e.g., eliminate  incentives to travel), with voluntary abatement at the
 lowest level and  progressing through increasingly disruptive measures
 until at  the highest level  of severity, a complete work holiday is
 declared.
 3.2.2  Evaluation:  Preventing  Significant Harm Levels
     A full  evaluation of the ARB contingency plan, as to its potential  for
 preventing federal significant  harm levels, is difficult since the plan still
 is under preparation and parts of it  have not yet  been defined.   Many details
 in the ARB contingency plan  have been  left up to  the local  APCD's.   After the
ARB and APCD's  complete the  plan, a more  thorough  evaluation  will  be possible,
                                   88

-------
      However,  on  the  basis  of  the  episode criteria,  (compared to federal

 significant  harm  levels  in  Table  3.8),  it seems very unlikely that the ARB

 contingency  plan  will  prevent  federal harm  levels.  First, there are no

 episode criteria  for  N02.   Second, although  episode criteria exist for

 oxidant and  carbon  monoxide which  call  for  active control before federal

 harm  levels  are exceeded*  (this  is ARB  Stage  2; ARB Stage 1 has only volun-

 tary  control), these  criteria  are  basically  of the feedback variety.  As

 noted in the OES  support-document  for the ARB plan ancL.in Section 2.4 of

 this  report, forecast (feedforward) episode  criteria are necessary to

 effectively  abate traffic.   Furthermore, even if traffic could be reduced

 on  an instant's notice,  a forecasting system would still be needed for

 oxidant pollution.  Because of the chemical-physical nature of photochemical

 smog, once oxidant  has started to  increase,  it can continue to increase even

 if  all  emission sources  are subsequently shut down.


      Another area-where  the current ARB draft plan could be improved is in

 the provision  of  regional cooperation during alert conditions.  Currently,

 it  appears this control  is  left up to the APCD's to initiate.  As discussed

 elsewhere, this neglects the regional character of the problem.   For example,

 it  is  doubtful that even drastic actions taken by Riverside County or
*  The first active (as opposed to voluntary) control phase, Stage 2, is
called when the one hour oxidant average exceeds .40 PPM.  This is below
the .40 PPM four hour, .60 PPM two hour, and .70 PPM one hour averages
specified as significant harm levels by the EPA.  Stage 2 is called when
one hour average CO reaches 75 PPM or 12 hour average CO reaches 35 PPM.
These criteria are below the 125 PPM one hour average, 75 PPM four hour
average, and 50 PPM eight hour average CO harm level.
   It should be noted that the second active control phase, Stage 3, is
below the one and two hour oxidant harm levels and the one hour CO harm
levels.  However, Stage 3 is not activated in time to protect against the
.40 PPM 4 hour oxidant harm level nor the 75 PPM-4 hour and 50 PPM-8 hour
CO harm levels.
                                    89

-------
                                TABLE  3.8


                  COMPARISON  OF THE  ARB  CONTINGENCY  PLAN

           ALERT LEVELS TO THE  FEDERAL SIGNIFICANT HARM LEVELS
POLLUTANT
Oxidant
Carbon
Monoxide
Nitrogen
Dioxide
FEDERAL SIGNIFICANT
HARM LEVELS - PPM
(Avg.
Time)
.70 - 1 Hour
.60-2 Hours
.40 - 4 Hours
125 - 1 Hour
75-4 Hours
50-8 Hours
2.0 - 1 Hour
.50 -24 Hours
ARB EPISODE CRITERIA (PPM)
HEALTH
(ADVISORY

.20/1 Hr.
Avg.
40/1 Hr.
Avg.
20/12 Hr.
Avg.
(1)
ALERT
Voluntary
Controls
.25a/l Hr.
Avg.
50a/1 Hr.
Avg.
25/12 Hr
Avg.
(2)
WARNING
Mandatory
Controls
.40a/l Hr.
Avg.
75a/l Hr.
Avg.
35/12 Hr.
Avg.
(3)
EMERGENCY
Mandatory
Controls
.60b/l Hr.
Avg.
100b/l Hr.
Avg.
50/12 Hr.
Avg.
None
aReached and expected to approach  the  next  stage  unless  abatement action
 taken.

 Reached and expected to persist for 1  additional  hour unless  abatement
 action taken.
San Bernardino County alone could prevent acute episode conditions which

originated from Los Angeles and/or Orange Counties.   It is critical for

adequate episode control  that a concerted and coordinated abatement plan be

instituted in all  the counties either significantly  contributing to the

emissions or. actually experiencing the alert conditions.


      The ARB proposal  presents only a general  outline of the control

measures to be taken at each alert stage.  In order  to be complete, this

outline needs specific  control  measures to be chosen by the APCD's.  However,

                                     90

-------
the general control outline itself is well organized.  A health advisory at



the first episode  level is followed by voluntary abatement at the alert



level, mandatory control at the warning level, and drastic mandatory controls



at the emergency level.  Controls for both stationary and mobile sources are



included.  The recommended mobile source controls (Table 3.6) basically



involve eliminating incentives for travel.  This is consistent with the OES



study which concluded that shutting off reasons for traveling would be the



most effective and enforceable type of traffic abatement strategy.




     The OES study is not really a complete contingency plan and will  not



be evaluated here  as such.  It is actually more of an analysis of various



alternative traffic abatement strategies.  The OES analysis is a comprehen-



sive and useful study and has served as the basic technical support for the



ARB plan.




     The OES evaluation of traffic controls includes consideration  of factors



such as potential  VMT reduction, preparation time requirements, implementa-



tion problems, socio-political effects, fairness, legal  requirements and



problems, enforcement difficulties, safety hazards, and administration costs.



As noted above, this analysis was quite comprehensive.  Only three  slight



deficiencies are apparent.  First, more detail might have been given on the



enforcement and implementation procedures to be attached to the traffic



abatement measures.  Measures were evaluated as to implementation and enforce-



ment problems, but little description was presented on actual procedures.



Second, total  economic costs associated with various control measures, (e.g.



the cost of shutting industrial  or commercial  establishments), were not



included as a  point of evaluation.  In many cases, this cost is order of



magnitudes greater than administrative costs (which were included).  Third,



the sample calculations of VMT reduction in the OES report were slightly




                                    91

-------
optimistic.  This was partially the result of not allowing for "substitute

driving" when certain types of trips were foregone,  (e.g.  a 10% reduction

in commuter trips will  really be partially cancelled by an increase in

recreational and shopping trips, this was not allowed for  in the OES report).

Including these effects would lower certain OES VMT  reduction estimates by

around 20%.* (e.g.  instead of obtaining a 10% VMT reduction, the actual

reduction would only be around 8%).

     The results of the present study basically agree with the first four

conclusions of the OES report (listed on page 86).   However, the fifth con-

clusion, that local levels of government should initiate all but the last

(most severe) control stage seems to contradict the  need for a regional

approach to oxidant control.   As noted in the evaluation of the ARB plan,

controls initiated only by an individual  county often will  not be sufficient

to abate oxidant episodes in  that county.
                      4

3.3  EPA REGION IX TRAFFIC ABATEMENT STRATEGIES


3.3.1  Plan Description


     The Preliminary Report of the EPA Region IX Task Force on "Traffic

Abatement During Air Pollution Episodes" does not present  per se an episode

contingency plan, [4].   Rather, it is more a discussion of the actions and

procedures necessary to put a contingency plan together.  For example, while

there is a detailed description of possible alert criteria and procedures

for establishing such criteria, no actual episode criteria are presented.

The same is true for enforcement and communication.   Thus, it is very
*  In defense of the OES report,  it should be noted that the OES study was
more interested in "order of magnitude effects".   Twenty percent is not
a large correction, considering the error in estimates of VMT reduction.
                                   ,92

-------
difficult to analyze the report in the context of an actual  air pollution
episode contingency plan.  However, the report does present  a detailed des-
cription and evaluation of proposed traffic abatement measures, and,  based
on this, makes recommendations as to which abatement measures to use  in a
contingency plan.   This review of the Region IX Traffic Abatement Report
will concentrate on the analysis of proposed vehicle abatement measures.
     The EPA Task Force has divided the evaluation of traffic abatement
measures into five categories.  These are:  effectiveness, enforceability,
cost, socio-political  impact and time needed to implement.   In all, fourteen
different measures were evaluated using the above criteria.   The results of
the evaluation are summarized in Table 3.9.   On the basis of this evalua-
tion, the Task Force arrived at the following recommendations:
     1.  Individual measures that can be implemented almost  immediately
         (in order of decreasing recommendation)
         a.  Mandatory car pooling on freeways.
         b.  Request closure of government agencies and contractors.
         c.  Limit availability of parking.
         d.  Special parking tax.
         e.  Close gas stations.
         f.  License plate lottery.
         g.  Ban all non-emergency vehicles.
         h.  Request voluntary abatement.
     2.  Individual measures which can be  implemented by Spring 1973
         (in order of decreasing preference)
         a.  Mandatory car pooling on freeways.
         b.  Limit availability of parking.
         c.  Request closure of government agencies and contractors.
                                    93

-------
                                   TABLE 3.9 SUMMARY OF EPA - REGION IX EVALUATIONS
                                             OF ALTERNATIVE TRAFFIC ABATEMENT MEASURES
CONTROL MEASURE
DIRECT CONTROLS ON TRAFFIC
(1) Close major thoroughfares to
vehicles carrying less than
2 passengers
(2) Restrict road use according
to number of passengers in
vehicle
(3) Windshield sticker program
(4) License plate lottery
(5) Ban all non-emergency traffic
(6) Voluntary carpooling &
,work holidays
(7) Voluntary traffic reduction
INDIRECT CONTROLS ON TRAFFIC
(1) Limit availability of parking
(2) Close all public on and off
street parking
(3) Emissions abatement plans
(4) Close gas stations
(5) Special parking tax
HIGHLY INDIRECT CONTROLS ON TRAFFIC
(1) Closure of government agencies
to contractors
(2) Work holiday
EFFECTIVENESS

Good
Good
Good
Good
Good
Poor
Poor

Poor
Fair
Fair
Fair- Poor
Fair-Poor

Fair- Good
Good
ENFORCEABILITY

Good if Fines
are Sufficient
Good if Fines
are Sufficient
Dependent on
Amount of Fine
Dependent on
. Fine
Dependent on
Fine
Good
Good

Poor
Dependent on
Fine
Fair
Good
Good

Good
Good
COST

Minimal
Minimal
High
Minimal
High
High
Moderate

Minimal
Minimal
High
Moderate
to High
Minimal

Moderate
High
SOCIO-POLITICAL
IMPACT

Minimal
Minimal
Fair
Fair to Good
Poor
Good
Good

Poor
Poor
Good
Poor
Fair

Fair
Poor- Fair
TIME (NEEDED)
TO IMPLEMENT

Minimal
(few days)
Minimal
At Least a Year
Minimal
Minimal
Considerable
Minimal

Minimal
Minimal
Considerable
—
Minimal

Minimal
Minimal
VD

-------
         d.  Windshield sticker program.
         e.  Abatement plans for indirect emissions.
         f.  Close gas stations.
         g.  Special parking tax.
         h.  License plate lottery.
         i.  Ban on non-emergency vehicles.
         j.  Voluntary abatement only.
     3.  Combination of measures that can be used simultaneously
         (in order of decreasing preference)
         a.  Mandatory car pooling on freeways and limit availability of
             parking.
         b.  Work holiday and close gas stations.
         c.  Request closure of government agencies and contractors,  special
             parking tax, and limit availability of parking.
         d.  Request closure of government agencies and contractors  and
             close gas stations.
         e.  Mandatory car pooling on freeways, request closure of government
             agencies and contractors, and limit availability of parking.
         f.  Mandatory car pooling on freeways, special parking tax,  and
             limit availability of parking.
         g.  Mandatory car pooling on freeways only.
3.3.2  Evaluation:  Alternative Traffic Abatement Measures
     Since the Preliminary Report of the EPA Region IX Task Force on  Traffic
Abatement does not consist of an actual air pollution episode contingency
plan, this evaluation of the report will center on the task force's  analysis
of alternative traffic abatement measures.  In general, the analysis  is both
fair and well  organized.   All principle measures which have been proposed up
                                    95

-------
to now are thoroughly reviewed, with particular attention being paid to the
implementation and enforcement requirements of each.  Also the criteria upon
which the task force judged each measure are well defined and meaningful.
However, there are several areas in which the report could be expanded.
These are discussed below.

     Perhaps the major shortcoming to the task force's evaluation is that it
fails to quantify (even approximately) VMT reductions for most control  measures
For example, in analyzing the effectiveness of mandatory car pooling, it
states that "a very significant reduction in the amount of traffic is bound
to result."  On the other hand, it classifies the effectiveness of limited
parking availability as "doubtful."  While these observations may be appropri-
ate, there is no indication as to how they were obtained.   This, and the
general  absence of quantified estimates make comparisons with other studies
difficult.

     Another possible  point of criticism is that the report seems to over
estimate the socio-political  acceptability of using  heavy fines as a whip
to force public cooperation with  proposed abatement  measures.  It is usually
true that the stiffer  the fine, the more effective the regulation will  be.
However,  an enforcement policy based on extremely heavy penalties may be very
unpopular.   A case in  point is the following statement on the enforceability
of mandatory car pooling, a measure which the report highly recommends:

     "If a  substantial  number of  vehicle operators violate these
     driving restrictions, it will  of course be impossible to
     apprehend all  of  them.   However, if 'enough'  are apprehended
     and  end up paying  heavy fines ($100+), there is bound
     to  be  a significant  deterrent effect in the future."
                                    96

-------
     Finally, the task-force report does not always include all  of the



economic consequences of many of the proposed abatement measures, both in



terms of direct and indirect costs.  For example, the report concludes that



the direct costs to implement mandatory car pooling on freeways  is "almost



nil if sufficient cooperation is obtained from state and local  law enforce-



ment officials."  This may be true for the EPA, but it is certainly not the



case for state and local agencies which may already be hard put  to meet the



demands placed on them by routine assignments.  Another instance is its



assessment of the costs involved in closing governmental agencies:




     "(a)  Direct:  There will be no direct cost for this strategy



           other than the differential pay for essential employees.



     "(b)  Indirect:  Some public services will be put off temporarily,



           but flexibility in deadlines can be added to mitigate this



           cost."




     It is not mentioned that someone, either the public employees or  the



tax payer, must bear the burden of wages and/or product foregone on those



days when a holiday is declared.  This cost can be very large.   There  is



also the cost of increased flexibility, additional staff hours,  etc.
                                    97

-------
                        REFERENCES - CHAPTER THREE


 1.   California Air Resources Board, The State of California Implementation
     Plan for Achieving and Maintaining the National Ambient Air Quality
     Standards, 30 January 1972.

 2.   California Air Resources Board, Proposed Air Pollution Contingency Plan,
     Preliminary Draft #2, 23 July 19717'

 3.   California State Office of Emergency Services,  Air Pollution Emergency
     Traffic Control Planning Committee, Traffic Abatement Plan for Air
     Pollution Episodes, 1973.

 4.   Environmental Protection Agency, Region IX, Preliminary Report of the
     EPA Region IX Task Force on Traffic Abatement During Air Episodes,
     15 September 1972.

 5.   Foon, Julian, Data Processing - Los Angeles County Air Pollution Control
     District, Personal Communication,  August 1973.

 6.   Los Angeles County Air Pollution Control  District, Rules and Regulations,
     Part VII, Rules 150 - 164.

 7.   Los Angeles County Air Pollution Control  District, Profile of Air
     Pollution Control, 1971.

8.   Los Angeles County Air Pollution Control  District, Smog Alerts - School
     and Health  Smog Warnings,  Special Services Division Publication #2,
     May 1972.

9.   Los Angeles County Air Pollution Control  District, The APCD Enforcement
     Program,  Special  Services  Division  Publication  #3, May 1972.~

10.   Los Angeles County Air Pollution Control  District, Personal  Communica-
     tion with  Engineering Division,  August  1973.
                                   98

-------
                4.0  ALTERNATIVE CONTINGENCY PLAN ELEMENTS







     Episode contingency plans contain four general elements:  episode



criteria, control strategies, surveillance procedures, and operational



procedures.  In formulating an actual contingency plan for a specific air



basin, certain alternatives will exist for each component.  This section



discusses the principal alternatives for a South Coast Basin plan.




4.1  EPISODE CRITERIA




4.1.1  Number of Episode Stages




     A contingency plan must contain a set of episode criteria which cause



various parts of the plan to be put into action.  These criteria should be



organized into stages of severity, more stringent controls being associated



with more advanced stages.  For a particular plan, the actual number of



stages should be chosen by balancing two opposing factors:  adding more



stages increases flexibility; however, it decreases simplicity and may



make the plan more cumbersome to operate.




     EPA requirements, [CFT 420.16(6)], specify that contingency plans



must have at least two stages.   To choose only two stages in a plan seems



inflexible.  This is particularly true for Los Angeles.  It will be shown



later that a nearly complete shutdown of activities is required in the most



advanced stage if pollutant concentrations in excess of significant harm



levels are to be prevented.   Five or more stages seem to violate the con-



straint of operational  simplicity.  Thus, three or four appears to be the



appropriate number of stages for a Los Angeles episode plan.




                                    99

-------
     In the plan recommended by this study (Chapter 6), the first stage will



essentially be voluntary control  (no emission reduction assumed), while the



last stage will involve a shutdown of recreational  facilities and all non-



critical business.   To provide a  more gradual transition between these



extremes, two (rather than one) intermediate stages will be included.  Thus,



the proposed plan will contain four episode stages.




4.1.2  Feedback vs. Forecasting Criteria




     Section 2.1.1  noted that episode criteria can  be of two varieties,



feedback and forecasting.  Feedback criteria are based on pollutants a.:tually



having reached certain levels.  Forecasting criteria are based on prediction



that pollutants will  reach emergency levels.   Within the latter category, a



further distinction can be made.   Some forecasts, (usually longer term ones,



say 1  to 2 days in  advance), are  based on predicting weather conditions and



then using certain  procedures to  determine expected pollutant concentrations



from expected weather conditions.   Other forecasts, (usually shorter term



ones,  say less than a day in advance), use present  pollution levels and



expected changes in weather conditions (i.e.  better, same, worse) to predict



upcoming pollution  levels.   The first might be called a future-weather type



forecast, while the second might  be called a weather-change type forecast.



The weather-change  type forecast  tends to be more reliable since it deals



with predicting increments in meteorological  conditions rather than predicting



absolute meteorological  conditions.




     Section 2.3 stressed the need for using forecasting (as opposed to



feedback) episode criteria in a contingency plan for Los Angeles



since  advanced notice is required  to implement most control measures.
                                   100

-------
This is particularly true for traffic abatement measures, and motor vehicles
are the dominant source of pollution.  For the control of oxidant, the
pollutant likely to cause nearly all episodes, forecasting appears to be
an absolute necessity.  Due to the complex nature of photochemical smog,
oxidant levels can continue to rise (while sunlight is available) even if
all emission sources are instantaneously shut-down.
     The one major problem with the forecast approach is that predictions
of air pollution levels are not very accurate.  Forecast errors will
mean that on some days more controls than necessary will be implemented;
an alert may be called when no controls are needed.  On other days, in-
sufficient controls will be put into effect; no alert may be called and
significant harm levels may be exceeded.
     One possible way to eliminate the latter decision error would be to
use feedback criteria as a backup system.  If the forecast fails  to pre-
dict an alert, the feedback system could still  activate controls  when
pollution starts to reach danger levels.  Such a system might be  possible
for primary contaminant alerts, although it still  would have to contend
with the great difficulty of implementing controls on short-term  notice.
For oxidant there is the added consideration that oxidant can continue to
accumulate even if all  sources are shut-down.   All things considered, feedback
criteria seem totally inappropriate for oxidant control*
* Under certain conditions,  feedback controls  might actually  be  counter
  productive to oxidant control!   Nitric oxide,  the bas-j c emitted form of
  NO ,  actually consumes ozone,  the basic component of total  oxidant,  when
     the former is  emitted tnto  the atmosphere after intense  sunlight  has
  waned.  This  accounts for the  fact that overnight, ozone levels are
  suppressed below  the natural background level  in  the populated sectors
  of the Los Angeles  basin,  while  ozone  remains  high in the air  that has
  blown out of  the  basTn and into  the desert.   Thus, reducing the evening
  traffic on the day  of an episode might allow more oxidant to persist
  into  the evening.   (On the other hand, reducing the evening traffic
  would help reduce pollutant levels on  the following day.)
                                   101

-------
     The overall  viability  of  the  contingency  plan may be  very  sensitive

to the available  forecasting capabilities.   In order  to assess  the  present

capabilities  in the Los  Angeles  Basin,  a  study was made of agencies  now

issuing air pollution  forecasts.   The  results  of  this analysis  are  pre-

sented in the next section, 4.1.3.

4.1.3  Air Pollution Forecasting Capabilities  for the Los  Angeles Basin

     Nearly all air quality predictions for  the Metropolitan  Los Angeles

AQCR are done on  the local  (county)  level.   The California Air  Resources

Board conducts only a  very  limited amount of pollution forecasting  for

Los Angeles.   Although the  National  Weather  Service issues "Air Stagnation

Advisories,"  no actual predictions of  air quality levels are  put out by

federal agencies.

     Table 4.1 summarizes the  air  quality forecasting capabilities which

are now available for  the Los  Angeles  Basin.   This table supports the

following general conclusions:


      9  Experience -  A  great  deal of  experience  has been  gained in
         forecasting instantaneous oxidant maxima (peaks)  slightly
         more than a day in advance.   In  particular, the Los  Angeles
         County APCD has conducted such a program for more than
         seven years.  There are indications that all of the  fore-
         casters, relying on knowledge  gained  in  the L.A.  County APCD
         program  as well as their  own  experience, seem to  have
         approached an optimum "state  of  the art" for predicting
         peak oxidant  26-30 hours  in advance within the constraint
         imposed  by availability of  meteorological data, [9 ],
         [31].
         Experience is rather  limited  in  forecasting  primary  pollutan^t
         concentrations  and time averages of oxidant.  The L.A. APCD
         has  been conducting an  experimental program  for predicting
         peak primary  contaminant  levels  but has  found that such fore-
         casts tend to be much less  reliabile  than peak oxidant fore-
         casts.  Predicting time averages of oxidant  (say  1,  2, or  4
         hours) should not  be  particularly difficult; wtth sufficient
         time to  adjust, meteorologists may"be able to forecast time averaged
         oxidant  with,  less  error than  peak oxidant, [9],  [15],  [31].
                                    102

-------
    TABLE 4.1  AIR POLLUTION FORECASTING CAPABILITIES FOR THE METROPOLITAN LOS ANGELES AQCR
ORGANIZATION
AND PRINCIPAL
METEOROLOGISTS
Los Angeles County
APCD
Ralph Keith
Arthur Davidson
David Hartman
San Bernardino
County APCD
Mel Zeldin
Riverside County
APCD
Erwin Kauper
(Consul tant)
Ventura County APCD
Earl Taylor
California Air
Resources Board
Spencer Duchworth
Robert McMullen
EXPERIENCE:
YEARS HAVING
ISSUED FORECASTS
7 years for oxidant
An experimental
program for CO
and NO- has been
underway 2-1/2
years
2 years
1-1/2 years
1/2 year
1/2 year
LOCATIONS IN
BASIN FOR
WHICH FORECASTS
ARE ISSUED
10 monitoring locations
within the basin:
Downtown, Long Beach,
Reseda, West L.A. ,
Lennox, Burbank,
La Mirada, Pasadena,
Azusa, and Pomona
4 locations: San '
Bernardino, Upland, ;
Chi no, and Redlands
4 locations in Riverside
County within the basin:
Riverside, Hemet, Prado
Park, and Pern's
2 locations in Ventura
County: Ventura,
Inland Val leys
One Basin-wide figure
(predicted maximum)
FORECASTED
QUANTITIES
Instantaneous
peak oxidant
(Experimental
Program, In-
stantaneous
peak CO and NOj
Instantaneous
peak oxidant
Instantaneous
peak oxidant
Maximum 1 hour
average oxidant
Maximum 1 hour
average oxidant
TIME OF
FORECAST
10:00 in the
morning of the
previous day
10:00 in the
morning of the
previous day
2:00 in after-
noon of the
previous day
10:00 in the
morning of the
previous day.
Update at 4:00 pir
2:00 in after-
noon of the
previous day
ACCURACY
Oxidant -- Average error, all stations:
Yearly: .028 ppm
July-Aug.-Sept. : .05 ppm
"Worst days" : .1-.2 ppm
CO and N00 -- Relative error is much
"~ greater than for oxidant.
Accuracy is not much better
than using monthly averages
for the forecast. Average
yearly CO error at all sta-
tions is 5 ppm.
Oxidant -- Average error at
San Bernardino station:
Yearly: Approx.* .03 ppm
July-Aug.-Sept. :*. 05 ppm
"Worst days": . 1-.2 ppm
*The average error reported by the
San Bernardino APCD is less than the
figures given above. However, the
San Bernardino APCD uses a -OB ppm
prediction range, i.e. 0-.05 ppm, .05-
.10 ppm, etc. The reported errors have
been adjusted upward to approximately
indicate the error to be expected from
an exact prediction.
Oxidant -- Average error at
West Riverside:
Yearly: .05 ppm
Oxidant -- 85-95% of all predictions
are within ± .05 ppm
Oxidant -- Approximately 90% of all
predictions are within
t .05 ppm.
REFER-
ENCES
[3], [9],
and [29].
[31]
[15], [22]
&6]
[10]
o
u>

-------
       o  Locations - From the various APCD's, forecasts are available
         at~20 locations throughout the basin.  The only significant
         omission is Orange County which uses the L.A. County forecast
         for areas near the Orange County border, [9 ].

       e  Forecasted Quantities - Presently, the basic predicted
         quantity is peak oxidant level; three APCD's, which account
         for 18 of the 20 locations, forecast this parameter.
         Ventura County and the State ARB use max. 1-hour oxidqnt.
         The L.A. APCD is the only organization significantly  involved
         with forecasting primary pollutants; their experimental  pro-
         gram predicts instantaneous peaks  for primary contaminants.

       e  Time of Forecast - Basic forecasting times range from 10:00 a.m.
         (Los Angeles, San Bernardino, and  Ventura) to 2:00 p.m.
         (Riverside and the State ARB) of the previous day.  One  very
         significant aspect about these times (as compared to  later in the
         afternoon) is that the basinwide maximal oxidant of the  previous
         day is not a known quantity to be  used in the forecast.   If the
         previous day maximum were available, it should improve forecasting
         accuracy, [9], [15].

      e  Accuracy -  As  noted above,  most forecasters  seem to have  approached
         a  best state of the art for predicting  peak  oxidant one day in
         advance  given  the available weather  data.  On  a  yearly basis, the
         average  error  is  slightly  less  than  .03  PPM  in Los  Angeles  County.
         However,  this  includes  many days of  very  low  oxidant which  are
         very easily  forecasted  on  an  absolute basis.   For the months
         of highest oxidant,  July-August--September, the average error
         is around  .05  PPM.   For the "worst days",  say maximum oxidant
         values above  .40  PPM, the average error  tends  to be around
         .10-.20  PPM, [15],  [29], J31J.

         The  L.A. APCD  experimental  program for forecasting primary  con-
         taminants has  shown  that the  relative error in predicting maxi-
         mum  primary  contaminant concentrations is  considerably greater
         than  that in predicting maximum oxidant  concentrations.  This is
         essentially  because maximum primary contaminant concentrations
         depend much  more  on  local,  small scale, meteorological fluctuations
         than do oxidant concentrations; some of  these fluctuations,
         (e.g. wind direction away from high traffic areas), are difficult,
         if not impossible,  to predict a day in advance, [9j.


     Since  this study is  concerned  with  episode  control,  the forecasts of

most interest are  those for days of extremely high pollutant concentrations.

For these days, the  existing forecasting capabilities  lack the  accuracy

which would be desired  for instituting episode control  measures which have
                                   104

-------
great socio-economic impact.  For episode type days, peak oxidant forecasts

tend to be around .1 to .2 PPM in error.  Peak primary contaminant forecasts

are not reliable in general?  As part of the overall episode control  program,

ways should be sought to reduce these forecast errors.  Possible methods  of

increasing accuracy are the following:

     e  Instead of forecasting instantaneous peaks, time averaged con-
        centrations, (one hour, two hour, four hour, etc.)  might be
        used.  This should eliminate some of the error due  to unpre-
        dictable "noise" type fluctuations which affect the peak value.
        For days of high oxidant, switching to time averages might
        reduce the forecast error by as much as 10-25%.  Using
        long term averages would improve the relative accuracy of pri-
        mary contaminant forecasts even more, [9], [31],

        Switching to time averaged concentrations should not present
        significant problems to the forecasting meteorologists,  [9], [31].
        Essentially, they would just need to gain the experience of
        dealing with a new, and slightly different, parameter.

     •  Instead of using the forecast of the late morning or early
        afternoon of the previous day, an oxidant episode control
        plan might be based on a late afternoon forecast.  The
        meteorologist would then have the benefit of knowing the
        previous day maximum.  In the terminology of section 4.1.2
        this would be a change from a future-w_eatKer type forecast
        to more of a weather-change type forecast.   Switching to
        late afternoon should improve forecast accuracy somewhat,  but
        it should not be expected to result in a tremendous reduction in
        error.  The one institutional problem associated with using
        a late afternoon forecast is that a significant change will
        be required in the present operational time-table of the
        forecasting agencies, [9], [14], [15].  Also, later forecasts will
        imply less time for implementation of control strategies,  (See
        Section 4.4.4.).

     •  Prediction accuracy would also be improved if more  detailed
        meteorological  information were made avilable to the fore-
        casters, [9].   An example would be taking vertical  soundings
        at more locations and more frequently.  However, instituting
        a program to provide more comprehensive data on air pollution
        weather factors would involve considerable costs.
*  Fortunately, primary contaminant forecasts are less  critical  since primary
contaminants are extremely unlikely ever to reach significant harm levels  in
Los Angeles.

                                   105

-------
     It should be emphasized that,  although improvements in forecasting


accuracy can be obtained by the means outlined above,  these improvements


will tend to be marginal.  Even with all  of the above  measures taken, one


cannot realistically expect the oxidant prediction error on episode type


days to be reduced to .05 PPM or less.   It appears that significant fore-


casting uncertainty on episode type days  simply cannot be eliminated in


the short term.




4.1.4  A Basic Dilemma



     The preceding sections indicate that episode control in Los Angeles


involves a fundamental dilemma.  In order to prevent significant harm


levels, in particular for oxidant,  forecasting-type episode criteria are


necessary.  However, even if a major effort is undertaken to improve the


already advanced and experienced forecasting capabilities, very significant


errors will  remain in forecasts for episode type days.  These errors will


lead to corresponding errors in the amount of control  which is called for


as compared  to the amount of control which is necessary.  On days when the


forecast proves to be overly optimistic,  less stringent controls than neces-


sary may be  implemented and the significant harm levels may be exceeded.
                                 «

One way to avoid this would be to "over control" the system by instituting


very strict  controls whenever there is  any threat of high oxidant levels.


However, since threats of high oxidant  levels are frequent in Los Angeles


and since the socio-economic impact of  effective control measures is very


great (see Chapter 5), and "over control" policy might be totally unacceptable


to the public.  With a more realistic policy, it seems that exceeding the


significant  harm levels because of  forecast errors will just have to be an



accepted possibility.



                                    106

-------
     In conclusion, it appears that an emergency control  plan for Los  Angeles

should be based on forecasting type episode criteria.   This  type of plan  will

be recommended by the present study (Chapter 6).  An attempt to improve  pre-

diction accuracy should be made, but significant forecast error is inevitable

and exceeding the harm levels must be accepted as a possibility.  If the  fore-

casting system fails and significant harm levels are reached, it would appear

best to concentrate efforts on measures for protecting public health,  (e.g.

by broadcasting warnings to stay indoors, by keeping children late at school,

etc.) rather than to attempt to reduce emissions during the  episode.*



4.2  CONTROL STRATEGIES

     Associated with each stage of an episode contingency plan is a control

strategy to be implemented at that stage.  Each control strategy consists

of a group of control measures along with implementation and enforcement

procedures for those measures.  Before adopting specific control strategies

for the Los Angeles contingency plan, certain issues must be resolved.

First, the degree of emission control required to prevent significant harm

levels must be determined.  Second, location aspects of the  plan, (i.e.,

basin-wide vs. subregional control), must be decided upon.   Third, an

evaluation of various alternative control measures as  to effectiveness

(VMT reduction and/or emission reduction), institutional  obstacles (imple-

mentation and enforcement), and socio-economic impact  (cost  and equity)
* This was the major point made in a testimony by the San Bernardino
  County APCD to the Air Resources Board on September 19, 1973 in re-
  viewing the State Air Pollution Emergency Contingency Plan.  The
  San Bernardino APCD noted that some control measures taken during the
  middle of an episode, (e.g. closing retail stores and sending people
  frome or stopping traffic), might actually increase the public exposure
  to pollutant levels.
                                    107

-------
should be performed.  Of the control measures which are effective enough

to prevent harm levels, those should be chosen which minimize institutional

and socio-economic problems.

     Section  4.2.1  calculates the degree of emission control required  for  a

Los Angeles episode plan.  Section 4.2.2 deals with locational  considerations.

Alternative control measures are thoroughly evaluated in Chapter 5, leading

to recommendations for adopting specific strategies.

4.2.1  Degree of Emission Reduction  Required

     A contingency plan for the Los  Angeles Basin should provide for

sufficient reductions in RHC, NO , and CO emissions to  prevent  oxidant, N0?,
                                s\                                         ^
and CO from exceeding significant harm levels  during  air pollution episodes.

Below,  the degree  of RHC  emission  reduction  required  to  prevent  harm levels dur-

ing the most  severe anticipated  oxidant  episodes  will be  calculated.*   This will

serve as  the  emission  reduction  objective for  the most advanced  control  strategy

stage.  It will be assumed that control  strategies  which yield  sufficient

RHC reductions to  prevent oxidant harm levels  will  also  yield sufficient

NO  and CO reductions (through the vehicular control  part of the strategies)
  A

to prevent N02 and CO harm levels.   This assumption should surely be valid

since ambient CO and NOp levels  are  very much  lower than oxidant levels

relative to the significant harm standards  (see section  2.2.5)  and since

CO and NO^ pollution levels are  more sensitive to emission level changes

during episodes than are oxidant levels  (see section  2.2.7).
* Theoretically,  the maximum emission reduction to insure always preventing
  harm levels is  a 100% reduction because  theoretically, meteorological con-
  ditions could be so adverse as  to produce  harm level  concentrations from
  negligible emissions.   In the text above,  a more practical  and operational
  concept of maximum degree of control will  be  used.
                                    108

-------
     Sections 2.2.5 and 2.2.7 provide the basic information for calculating

the maximum degree of RHC control  required to prevent significant harm

levels for oxidant.  Section 2.2.5 noted that the .40 PPM - 4 hour harm

level is the "binding constraint"  of the three oxidant harm levels (.40 PPM

4 hour, .60 PPM - 2 hour, and .70  PPM - 1 hour).   That section also noted

that the greatest recorded excesses of all three harm levels occurred  in

July 1955 at Pasadena, (.58 PPM -  4 hours, .69 PPM - 2 hours, and .74  PPM  -

1 hour).  From January-1970 through June-1973, only the 4 hour harm level

has been violated; the greatest excess has been .53 PPM - 4 hours at River-

side on August 6, 1970.

     Figure 2.8  showed that overall  emission  levels  are  now  undergoing

a considerable decrease in the Los Angeles Basin  due to the  new car and used

car retrofit and other control programs.   Oxidant episodes  in the future

are expected to be less severe than in the past.   It should  surely be

sufficient to plan for a  maximum degree of control  capable  of reducing

a .58 PPM - 4 hour oxidant episode to the .40 PPM -  4 hour  harm level

standard.*

     Section 2.2.7 derived the very approximate relationship that an X%

reduction in RHC emissions on a given day would result in only about a

%X% reduction in oxidant levels on that day.   This less than proportional

effect results from carryover precursors  remaining from the previous day

as well as certain basic  nonlinearities in the oxidant-hydrocarbon
* On an equivalent percentage basis, such a plan could reduce a ,86 PPM -
  2 hour episode or a 1.0 PPM - 1  hour episode to the .60 PPM - 2 hour
  and .70 PPM - 1  hour harm levels.
                                   109

-------
 relationship.  Thus, a plan capable of abating a .58 PPM - 4 hour oxidant
 episode,  (the harm level  is 30% less than this), must provide for a 60%
 reduction in RHC emissions.
 4.2.2  Locational Considerations
     An air pollution control  plan can be instituted on either a basin-wide
 or subregional basis.  For air pollution episodes in the Metropolitan Los
 Angeles AQCR, it would seem useful to make special  regional  considerations
 for Ventura and Santa Barbara  counties.   Air pollution in these counties is
 much less severe than in  the other four  counties of the Los  Angeles Basin
 (see Table 2.4); in fact, it is not expected that air pollution emergencies
 would ever occur in these two  counties.   Furthermore, because of lower emis-
 sion source density in Santa Barbara and Ventura and because of the typical
 wind flow pattern, Santa  Barbara and Ventura do not contribute significantly
 to air pollution in the other  four counties.  The effect of  Santa Barbara and
 Ventura on oxidant episodes, which typically occur in the central-eastern parts
 of the basin, is practically nil.   It does not appear worthwhile to make these
 two counties institute control  measures  with severe socio-economic impact dur-
 ing air pollution alerts  occurring elsewhere in the basin.   A division of the
 total  region into a four  county control  region (Los Angeles, Orange, Riverside,
 and San Bernardino),  with Ventura  and Santa Barbara to be added only under
 special  circumstances, seems called for.
     However, any further regional subdivision for episode control does not
 seem practical.   Air  pollution  alerts might occur in any of  the other four
counties,  and very significant  interaction exists between emissions in the
coastal  counties (Los Angeles  and  Orange) and pollution levels in the
inland  counties  (San  Bernardino and Riverside).  To try to implement con-
trol measures on a subregional  basis within the four county  area would
                                    110

-------
require prediction of trajectories for air parcels that will  reach alert



levels, (so that control could be initiated in the regions along the trajec-



tories).  Reliable trajectory forecasting of this type is doubtful.   Further



investigation into region subdivision deserves consideration, but until  sub-



division is demonstrated as practical, it appears that the four county region



should be treated as a whole in an episode control plan.





 4.2.3  Alternative  Control  Measures



      This  study will  consider potential  episode  control  measures  for  both



 stationary and  mobile sources.   Control  of vehicular sources  must be  in-



 cluded since they constitute the greatest single source category.   Gasoline



 powered motor vehicles (LDMV's,  HDMVs,  and motorcycles), account  for  77%*



 of RHC, 96% of  CO,  and 74%  of NO  emissions in the  Los  Angeles  Basin  in
                                 /\


 1973 (see  Table 2.5).  Although they produce less total emissions,



 stationary sources  are significant.   Further,  they  often are  very con-



 centrated  sources,  and it is usually easier to control  a few  concentrated



 sources then it is  to control  many disperse vehicular sources.  Thus,



 stationary source control  should also be considered.



      Table 4.2  lists  the alternative episode control  measures which will



 be considered in this study.  These  control  measures will  be  thoroughly



 evaluated  in Chapter 5 as  to effectiveness  (VMT and/or emissions re-



 duction),  institutional problems (implementability  and enforceability),



 and socio-economic  impact (cost  and  equity).  The  most promising measures



 will be organized into proposed  control  strategies  in Chapter 6.
 * This includes  vehicle tank filling at gas  stations.
                                     Ill

-------
  TABLE  4.2    BRIEF  DESCRIPTION  OF  ALTERNATIVE  CONTROL  MEASURES


STATIONARY SOURCE  CONTROL  MEASURES
     1.  Ban Filling of Underground Service Station Tanks  -  Filling underground service station  tanks prohibited.
     2.  Shut Down Major Point  Sources - All operations at EPA  defined major point sources  would be closed.
     3a. Shut_Down_Majq£ Point  Sources and Other APCD_Desia_nated Categories (Rule 155.2) -  No  Exceptions Granted - All  operations  covered by above
                                                                                                                definitions  would  be closed.
     3b. Shut Down Major Point  Sources and Other APCD Designated Categories (Rule 155.2) -  Exemptions Allowed - All operations  covered by above
         definitions -  subject  to shut down, unless previously  granted an exemption from a  review committee established for this  function.

TRAFFIC ABATEMENT  CONTROL  MEASURES
   Direct Controls on Traffic
     1.   Voluntary Traffic Abatement - An appeal  for voluntary  reductions in unnecessary travel.
     2.   Windshield Sticker Program - Traffic restricted on  the basis of previously issued  windshield sticker colors.
     3.   License Plate  Lottery  - Traffic restricted on the basis of license plate numbers.
     4.   Ban  Non-Essential Government Vehicles -  Travel by governmental vehicles restricted,  unless 1t  is of an essential nature.
     5.   Ban  (Private)  Fleet Vehicles - Travel by private  fleet vehicles prohibited.
     6.   Ban  (Private)  Fleet Vehicles and Use of Vehicles  for Commercial Purposes - Travel  by  private fleets and other vehicles 1n commercial  use
                                                                                  prohibited.
     7.   Ban  All Non-Emergency  Traffic - All travel not of an emergency nature, e.g. health and  safety  related, would be banned.
     8.   Mandatory Car-Pooling  on Freeways - Travel on freeways restricted on the basis of  vehicle occupancy.
     9.   Mandatory Car-Pooling  on Freeways and Major Thoroughfares - Travel on freeways and major thoroughfares restricted according  to  vehicle
                                                                   occupancy.
    10.   Commercial  Abatement Schemes - Commercial  enterprises  required to reduce travel directly or  indirectly generated by their operations.
   Controls on Traffic  Support  Facilities
    11.   Close Freeways, (or Certain Freeway Lanes) - Travel on freeways (or certain lanes  of freeways) would be prohibited.
    12.   Limited Availability of Parking - On-street and/or  off-street parking would be restricted.
    13.   Special Parking Tax -  Tax for parking instituted  wherever feasible.
    14.   Gasoline  Rationing - Program of gasoline rationing  instituted.
    15.   Close Gasoline  Stations - All  retail  gasoline outlets  closed.
   Control on  Incentives to Drive
    16.  ' Close (Non-Essential)  Federal  Agencies - All  federal agencies engaged in non-essential  activities  would be  closed.
    17.   Close (non-Essential)  Government Agencies  (Federal, State 5 Local) - All governmental agencies engaged  in non-essential activities would be
                                                                            closed.
    18.   Close Government  Contractors - All  governmental contractors engaged In non-essential  ooerations  would  be  closed.
    19.   Close Shopping  Facilities at Regional Shopping Centers and Central Business Districts - Shut down  all  retail  and  business activities at regional
                                                                                               shopping centers  and  central  business districts.
    20.   Close Recreational Facilities - Shut down  all recreational facilities, such as amusement parks,  bowling  alleys,  sporting  events.
    21.   Uork Holiday for Major Firms - Declare a work holiday  for all firms larger than a  certain size.
    22.   Rotating  Work  Holiday  for Major Firms -  Formulate a rotating schedule of work holidays  for firri larger  than  a  certain size.
    23.   Operate on  Sunday Status - All businesses  would assume a Sunday status for work.
    24.   Complete  Uork  Holiday with Recreation Closed - All  businesses and recreational facilities would  be closed.

-------
4.3  SURVEILLANCE PROCEDURES
     As explained in section 2.3, two different types of surveillance
procedures are necessary for the effective operation of an air pollution
episode control plan.  The first, monitoring of air quality and meteoro-
logy, provides the basic atomspheric data for the plan.  The atmospheric
surveillance should consist of routine measurements between episodes
with stepped up activity during alert situations.  The second, control
strategy surveillance, provides information on how well  the control
plan is being implemented.  This consists of monitoring emission sources
to determine the effectiveness of control measures and observing other
agencies to determine the extent to which they are fulfilling their
responsibilities.  For the Los Angeles area, there are several  alternative
ways of providing the appropriate surveillance.  These are discussed
below.
4.3.1  Surveillance of Air Quality and Meteorology
     Air quality and meteorology is currently monitored in the
South Coast Air Basin by agencies on three different governmental
levels - local, state, and federal.  At the local level are the various
county Air Pollution Control Districts.  At the state level is the
Air Resources Board.  At the federal level are the Environmental Pro-
tection Agency and the National Weather Service.  The air quality and
meteorology monitoring capabilities of these organizations are reviewed
below and are summarized in Table 4.3.
     County APCDs
     In the overall  California state air pollution control plan, county
APCD's have the principal responsibility for air quality monitoring.
In the Metropolitan Los Angeles AQCR there are five APCD's maintaining
                                    113

-------
   Table 4. 3     SOURCES OF  REAL-TIME AIR QUALITY AND METEOROLOGICAL  DATA
                          IN THE METROPOLITAN LOS ANGELES AQCR
SOURCE
Los Angeles County APCD
Orange County APCD
Riverside County APCD
San Bernardino County
' APCD
Ventura County APCD
California Air Resources
Board
Env. Protection Agency
(CHESS)
National Weather Service
TYPE OF DATA
OX CO NO NO? N0v HC S0? Meteorological
C- A £•
X X X X XX X
XX X X X X X
X
XX X X X X
XX XXX X
X
XX X X X X X
X
monitoring networks, Los Angeles County,  Orange County,  Riverside County,
San Bernardino County, and Ventura  County.   These districts  routinely provide
various air monitoring data,  and some conduct special  services which would be
available during an air pollution episode.
     The Los Angeles APCD maintains a telemetry network  of twelve air quality
sampling stations located throughout the  county (See Figure  4.1).  They con-
tinuously monitor concentrations of the following contaminants:  ozone (OJ,
nitric oxide (NO), nitrogen dioxide (N02),  sulfur dioxide (SO-), carbon
monoxide (CO), and total hydrocarbons (THC).  In addition, the stations also
monitor wind speed and direction, temperature, solar radiation, and relative
humidity.  At the present time,  however,  the telemetry system is not considered
                                    114

-------
                                               FIGURE  4.1   LOCATION OF REAL-TIME SOURCES OF AIR QUALITY DATA
 1.
 2.
 3.
 4.
 5.
 6.
 7.
 8.
 9.
10.
11.
12.
13.
14.
15.
16.
     Aiuaa
     Burbanlc
     El Monte
     Glesdora
     La Hirada
     Lennox
     Long Beach
     Loa Angeles
     Nevhall
     Pasadena
     Pomona
     Reseda
     Santa Monica
     Torrance
     West Covlna
     West Loa Angele
ORAMCE COUNTY

17.  Anaheim
18.  Garden Grove
19.  La Uabra
RIVERSIDE COUNTY

20.  Prado Park
21.  Riverside- Magnolia
22.  Rubidoux-Ml»«ion Blvd.

SAN BESJJAfiDlNO COUNTY
                                     APCD  •
                                     ARB   *
                                     EPA   B
23.
24.
25.
26.
27.
     Chlno
     Foncana
     Red land.
     Sao Bernardino
     Upland
VENTUTIA COUNTf

28.   CaiMrillo-Pal»
29.   OJai
30.   Santa Paula
31.   Thousand Oaks
32.   Ventura
                                                                 115

-------
reliable, and data is normally read by hand off strip-chart recorders located

at the stations, [11].


     Presently, the Orange County APCD maintains two air quality monitoring

stations which provide real-time (immediately available) data, one at Anaheim

and the other at La Habra (see Figure 4.1), [30].   Each measures the concen-

tration of oxidants, nitrogen dioxide, oxides of nitrogen, carbon monoxide,

and sulfur dioxide.  In addition,the Anaheim station provides data on total

hydrocarbons (THC).  This information along with wind speed and direction,
                                                                    /
temperature, and humidity, is usually read and relayed over the phone by

technicians several times a  day.   During an episode more technicians could be

sent into the field to other stations to increase  the flow of data.


     The Riverside APCD has  two air quality monitoring stations (see

Figure 4.1) which manually telemeter oxidant concentrations, [13],   The

system is queried by phoning each station.   In addition, these stations  also

monitor carbon monoxide,  nitrogen dioxide,  oxides  of nitrogen, hydrocarbons,

sulfur dioxide, wind (speed  and direction)  and temperature.  However, at this

time they are not included in the telemetry system.


     The San Bernardino APCD is presently installing a telemetry system  to

relay air quality data gathered from five stations  (see Figure 4.1)  on a real-

time basis, [31].   Information on the concentrations of the following contam-

inants will  be transmitted:   oxidants, carbon monoxide, oxides of nitrogen,

nitrogen dioxide,  hydrocarbons, and sulfur dioxide.   At this time,  no ,

telemetry of meteorological  data  is planned.


     The Ventura APCD is  now installing a telemetry system which will provide

both real-time air quality and meteorology data, [27].  Information  will be

relayed on the concentrations of oxidants,  carbon  monoxide, oxides of nitrogen,



                                   116

-------
nitrogen dioxide, and hydrocarbons, plus wind speed and direction.  At the



present time three stations are included in the system.  In addition, two



other stations also report data on a real-time basis (see Figure 4.1).




California ARE




     At the state level, the California Air Resources Board monitors air



quality and meteorology.  The ARB has just installed a pilot telemetry system



with connections to three stations in the South Coast Basin:  Upland, Riverside



and El Monte.  These stations measure oxidants, nitrogen dioxide, oxides of



nitrogen, carbon monoxide, total hydrocarbons, plus wind speed and direction.



While the system has the capability to relay eight pollutants along with wind



data, it is presently only being used to transmit oxidant concentrations, [16].




     Besides the information received via the telemetry system,  the



ARB routinely obtains data from several  other sources, [21],  The ARB sub-



scribes to the National  Weather Service  service A, service C and PONY



service.  Reports on all hourly average concentrations  for each  pollutant



are received from all  APCD's in the state.   There is, however, a con-



siderable lag in processing the APCD data,  15 to 45 days.   The ARB also



conducts its own aircraft soundings at Riverside to measure the  vertical



stability of the atmosphere over the Los Angeles basin.  In addition,



during an actual  air pollution episode,  the ARB maintains  phone  contact  with



Us own stations  In the  affected area (on an hourly basis)  and with the  APCD



in the county where the  episode is occurring.







EPA and NWS



     At the federal level, the Environmental Protection Agency and the



National Weather Service provide air quality and meteorological  data
                                    117

-------
 for  the  Los Angeles basin.  The EPA operates a nationwide system of air

 quality  monitoring stations with six in the Los Angeles basin  (see Figure 4.1).

 These  stations,  known as CHESS stations, relay information concerning atmos-

 pheric concentrations of ozone, carbon monoxide, nitrogen dioxide, oxides of

 nitrogen, sulfur dioxide, and total hydrocarbons along with temperature and

 dew  point to Durham, North Carolina, [2].   From there, the data is available

 by phone.*  The NWS offers several  meteorological services (via teletype) with

 information relevant to the South Coast Basin, [20].  They are service A,

 service  C, and PONY service.  Service A, while principally for aviation, pro-

 vides  hourly information on temperature, pressure, winds, etc., which is use-

 ful  in determining the overall meteorological  condition of the basin.   Service

 C, which originates in Washington,  discusses regional meteorological  conditions

 and  issues broad based advisories,  such as that for air stagnation.  PONY

 service, on the other hand, is a regional  service (catering to the FAA)

 designed to alert subscribers to the latest special  observations, including

 local  air stagnation advisories.   In addition  to these services the NWS also

 conducts balloon soundings twice daily at  El Monte and Los Angeles International

 Airport to determine vertical  stability.


     Having described the available monitoring capabilities, it is now

 possible to review the atmospheric  surveillance procedures of each

 government agency as they pertain directly to  air pollution episodes.

 At the local  level, air quality surveillance is conducted primarily

 tn conjunction with the health warning programs of the county APCDs.
   However, due to the time zone difference, data is normally available
only until  1:30 PM Pacific Time, when the EPA North Carolina operations
close.   In  unusual circumstances, this time limit can be extended, [2].
                                    118

-------
Atmospheric pollutant concentrations are observed throughout the day to



insure that conditions have not deteriorated to alert levels.   During an



episode, air quality levels are simply checked more frequently.




     Essentially the same is true of the state procedures.   The  ARB continually



checks the telemetry from its three stations in the basin to determine if



pollutant concentrations are approaching alert levels and surveys meteorologi-



cal conditions to check if an episode is probable.   If this data indicates



that an alert is imminent (or if the ARB is informed by an  APCD  that an episode



has occurred or is likely to occur), then an effort is made to obtain more



complete and up-to-date air monitoring data.  Although there exist no formal



alert procedures, at least once per hour each ARB station in the basin is



called, and less frequently, the affected APCDs are phoned.




     Surveillance at the federal level is conducted by EPA  Region IX in



San Francisco.  Each day, the Region IX meteorologist notes the  peak pollut-



ant readings for the previous day from newspaper reports and calls the station



with the highest value to obtain relevant data, [14].  From the  National



Weather Service, the Region IX office obtains a special forecast summary of



atmospheric conditions in the basin.  If an episode appears probable, oxidant



data from the ARB is relayed by telecopier and telephone as it becomes avail-



able.  In addition, data from the CHESS stations are obtained from Durham,



North Carolina, and the affected APCDs can be called to obtain data not



available from the ARB.



      The  above  description  of air monitoring  capabilities  indicates



 that existing  air  quality  and meteorological  measurements  in  the South Coast



 Bastn are quite  adequate for  supporting  an  episode  contingency  plan



 operation;  atmospheric monitoring data  are  generated on  a  continuous



 basis  at  numerous  sampling  sites-.   However,  the  discussion of surveillance




                                    119

-------
procedures indicates that reliable methods for quickly obtaining all this
data  (especially for the EPA) have not been instituted.  Too much seems to
depend on extemporaneous phone calls to agencies which may not have adequate
time  to deal with these calls due to their own activities with regard to the
episode or due to their routine duties.
      EPA Region IX has indicated that the major problem in air quality
surveillance is to obtain the existing data.   There are several possible
ways  this could be improved.   For one, a system could be established to
telemeter data directly from the monitoring stations to EPA headquarters in
San Francisco.  Second, a regional  center could be established in Los Angeles.
Air quality and meteorological  data from all  stations (APCD, ARB, and EPA)
could be telemetered to this  station for processing and then relayed to EPA
Region IX headquarters over the telephone.   Third, the EPA could attempt to
maintain phone contact with all  affected APCDs.   Finally the ARB could be
encouraged to establish real-time informational  channels with the APCDs. in
the South Coast Basin and to  act as the principal  disseminator of data.   EPA
Region IX could then establish  well defined arrangements for the ARB to relay
appropriate data on a regular basis, (say by telephone).
     Of these four alternatives, the last appears  preferable.  The first two
would be enormously expensive and would force the  EPA to do all of its own
data analysis and hence greatly increase its  work  load.   The third presents
problems of reliability and availability especially during episode conditions
when APCD manpower is already hard pressed  to meet its own demands.   The last
alternative would keep EPA Region IX out of data processing and encourage
closer cooperation and coordination between the ARB and APCDs.  Further,  it
would provide a large body of data all  at one location and would simplify
overall  communication problems.
                                   120

-------
     In summary, the problem of air quality and meteorology surveillance in
the Metropolitan Los Angeles AQCR is not so much due to lack of data  as  it
is due to difficulties experienced in attempting to obtain the data  on a
real-time basis.  It would appear that new procedures such as those  mentioned
above are necessary in order to facilitate smoother dissemination of
information.
4.3.2  Surveillance of Strategy Implementation
     Strategy implementation surveillance consists of two parts,  agency
surveillance and source surveillance.  Agencies with responsibilities for
air pollution episode control must be observed to determine the extent to
which they are fulfilling these obligations.  Source surveillance determines
the degree of compliance with the controls which have been instituted and
provides basic information for enforcing these controls as well as for
planning additional actions.
Agency Surveillance
     Under present state regulations and plans, the primary responsibility
for episode control lies with the county APCDs.  Until  a Third Stage  Alert
is called (defined by the extremely high and unprobable atmospheric  levels
of 1.5 PPM for oxidant or 150 PPM for CO, Table 3.1), at which time  the
Governor is requested to declare a state of emergency and take appropriate
action, no basin wide cooperation is provided for.  No single APCD has super-
visory responsibilities concerning the actions of any other APCD, and
although the various APCDs sometimes cooperate in the exchange of informa-
tion during an air pollution episode, there are no established procedures
for reporting activities to one another.  A Basin Coordinating Council exists
for the basin, but this Council is not normally called into action during an

                                   121

-------
 episode.  Thus, on the regional level, there are no procedures for monitoring
 the actions of the various APCDs during an alert.
     On the state level, the Air Resources Board has general authority to
 survey the activities of the individual APCDs.  However, during episodes
 there is no actual requirement for APCDs to report to the ARB.  In practice,
 during alert situations, the ARB keeps in fairly close telephone contact
 (approximately on an hourly basis) with the affected APCDs and acts in an
 advisory role, [1].   If it is determined that an APCD is not adequately
 coping with the situation, the ARB can take direct control, but only after
 24 hours notice.   Thus, even at the state level no formal procedures presently
 exist for appraising the actions of all APCDs in the South Coast Basin during
 an alert.
     It should be noted that regulations and legislation are pending concerning
 regional air pollution control  in Los Angeles and concerning episode contin-
 gency plans in California.   These may significantly change certain responsi-
 bilities and thus alter the surveillance pattern outlined above.
     At the federal  level,  the Environmental  Protection Agency, through the
 Regional Offices, has responsibility to take action during an episode either
 in the absence of an acceptable State or local program or in the case that
 the State or local program is not applied with sufficient effectiveness.
 EPA Region IX must observe  the activities of the APCDs and ARB during an air
 pollution episode in the South Coast Basin to insure that adequate controls
 are instituted.   At  least three alternative forms of agency surveillance are
 open to Region IX.  First,  Region IX could use field personnel for on-the-
 scene observations of actions by all APCDs and the ARB.  These personnel
would report back to EPA headquarters at regular intervals.  Second, Region  IX
could maintain telephone contact with all the APCDs and the ARB and obtain verbal
                                    122

-------
descriptions of all activities from the state and local personnel.
Finally, the EPA might leave it up to the ARB to monitor local APCD
activities and obtain reports of both state and local actions from the
ARB alone.  For the EPA, the last approach is most practical  in terms
of communications simplicity and administrative cost.
Source Surveillance
     The source surveillance procedures of an air pollution episode plan
must deal with both stationary sources and mobile sources.   Since the
former basically involve emissions at certain well defined locations
while the latter involve emissions from diffuse traffic throughout the
basin, different procedures might be applicable to the different
source types.   Thus, the existing surveillance alternatives for stationary
and mobile sources will be discussed separately below.
     In the California air pollution control program, County APCDs   have
the responsibility  for stationary source surveillance  and enforcement,
both in the overall air quality control program and during episodes.
The surveillance procedures are fairly uniform among all the APCDs  in
the Los Angeles Basin.   Each APCD maintains a staff (from 4 to 100) of
trained enforcement personnel,  [7].   These investigators routinely patrol
particular areas within their county, make unannounced spot checks, and
respond to complaints filed by citizens.   To supplement this capacity,
the ARB has a  staff of approximately fifty people who respond to sur-
veillance requests by individuals or APCDs, [17].
     The APCD  and ARB enforcement personnel are trained in visual
observation and source  sampling techniques.  During an episode, this
level  of expertise is not really required since sources basically have
to be  checked  for shut down and not for compliance with standards.  Thus,
                                    123

-------
although the familiarity with source locations would make APCD and ARB
personnel very valuable in stationary source surveillance during an episode,
source surveillance could be conducted by any well-trained personnel.
     Presently, the only agency in the Los Angeles Metropolitan Air Quality
Control Region capable of moni'toring traffic on a real-time basis is the
California Department of Transportation (DOT) - District 07.  As part of the
Los Angeles Area Freeway Surveillance and Control Program, it conducts
traffic counts on a forty-two mile loop encompassing portions of the
Santa Monica, San Diego, and Harbor Freeways (see Figure A.3 in Appendix A),
[4].  These counts are used to compute average speeds which  are in turn used
to detect accidents and congestion.  During the day, this traffic data is
accumulated at five minute intervals.  Around midnight, all  five minute
traffic data for the entire loop is printed out.   This is the only volume
data currently available except by operator interrogation of a particular
system station, [5].  It would not be exceedingly difficult  to set up a
procedure for obtaining instantaneous traffic volume estimates through an
appropriate computer interrogation subroutine.   The DOT has  indicated that
it should be able to provide the EPA with qualitative estimates of traffic
volume on the loop.  Further, officials  have indicated that  these estimates
should give a fairly good representation of the total  traffic picture in
the L.A.  area.
     If cooperation can be obtained from county and state officials, it
would be  more practical  to rely on the existing surveillance procedures of
the APCDs for stationary sources and the DOT for traffic rather than to
attempt to  institute a whole new surveillance program.  The  programs
discussed above are very expensive to implement,  and duplication of effort
should be avoided.
                                    124

-------
4.4  OPERATIONAL PROCEEDURES
 4.4.1  Legal Considerations
      Preliminary to discussing alternative operational procedures for an
 episode contingency plan,  it is useful to review the existing structure
 of legal authority and responsibility.  In the Metropolitan Los Angeles
 AQCR, episode control action can by law be exercised at three different
 governmental levels - local, state, and federal.  The nature and extent
 of the legal authority at  each level to implement emission control measures
 will be discussed below.   Particular attention will be given to whether
 or not the power exists to regulate vehicular traffic, especially through
 such indirect methods as closing gas stations, regional shopping centers,
 recreational facilities, etc., and to the question of liability for lost
 wages and profit.
 Local
      Section 24198 of the California State Health and Safety Code establishes
 the county Air Pollution Control  District as the focus for local  air pol-
 lution control, [23].   Section 24260 authorizes the governing boards of
 these districts, (e.g.,  county supervisors), to "make all  needful  orders,
 rules and regulations  necessary or proper to accomplish the purposes of the
 [Act] .  .  .  ".   It further provides that "... the air pollution control
 districts, (e.g. , their governing board,  in general the County Boards of
 Supervisors), to "make and enforce such orders, rules, regulations as will
 reduce the amount of air contaminants  released within the district."
 Violations of any such rule or regulation are made misdemeanors under Health
 and Safety Code Section  24281.   Thus,  in view of this broad authority to
 reduce air pollution,  it appears  that  a rule or regulation enacted to imple-
 ment an  emission reduction measure whether it involved direct or indirect
 control  of traffic would be valid.
                                    125

-------
     However, the actual  regulations which have been enacted under the



provisions of the Health  and Safety Code may not have sufficient scope for



large scale, indirect vehicle control.   Typical is Rule 158 of the Los Angeles



Air Pollution Control District which states that "The Air Pollution Control



Officer may .  .  .  order the closing of  any industrial, commercial, or busi-



ness establishment and stop vehicular traffic . .  .  , if, in the opinion of



the Air Pollution Control  officer,  the  continued operation of such establish-



ment or vehicle contributes to the  further concentration of any air contami-



nant, the concentration of which caused the declaration of the 'alert'.",



[19].  This separation of causation (i.e., there is  no acknowledgment of the



link between the operation of an establishment and the operation of motor



vehicles) along with the  provisions in  Rule 154.1  that episode control  plans



should not interfere "with the operations  of public  utilities or other pro-



ductive, industrial, business, and  other activities,  which are essential  to



the health and welfare of the public or result in  irreparable injury to any



means of production or distribution" makes it questionable if any large scale



form of indirect control,  such as a work holiday,  could be locally instituted.




     However,  APCD rules  do not apparently rule out  the possibility of using



more specific  forms of indirect control.   For example, closing heavy



industry would certainly  be permissible under Rule 154.1, paragraph a,



which states the Air Pollution Control  Officer may give written notice



to the owner or operator  of such industrial, business, or commercial



establishment  or activity to submit .  . .  plans for  immediate shutdown



or curtailment,  in the event of an  air  pollution emergency, all of the



sources of hydrocarbons .  .  .  , including  vehicles owned or operated



by such person,  his agents or employees in the scope of the business or



operation of such  establishment or  activity."  Further, since control




                                    126

-------
action is being taken in the context of curtailing a specific short
term public health menace, there is no indication that local government
would be liable for any lost wages or profit, £24].
State
     At present, the State of California may implement air pollution
episode control in one of two ways.  The first is by superceding the
authority of a local Air Pollution Control District through the Air
Resources Board.  The second way is to have the Governor declare a
state of air pollution emergency.  Both are discussed below.
     Pursuant to the Mulford-Carrell  Air Resource Act, Chapter 7,
section 39273, each Basinwide Air Pollution Control Coordinating Council
(composed of members or appointees from each APCD board)  is required to
submit a basinwide air pollution control plan to the ARB  designed to
meet or exceed air quality standards  established by the board for that
basin, [24].  If this plan is not submitted or if it is not enforced,
section 39274 of the act provides that "the board may take any appropriate
legal action to enforce such a plan,  including emission standards and
enforcement procedures therein."  Further, section 39274  states "the
Board shall also have the authority to take in any county air pollution
control district any action or function the district is authorized to  take."
Thus, if the Board requires as part of the basinwide plan an episode
control program, it may assume control  over the program whenever the
program is not being implemented or enforced by local authorities.  More-
over since the ARB is empowered with  all the authorities  of the local  Air
Pollution Control  Districts, it should be able to implement both direct
                                    127

-------
and indirect types of vehicle regulation  and  should  not be  liable  for lost

wages and profits.

     As mentioned above,  the second  way in which  the State  may  promulgate

episode control  is through  the declaration of an  air pollution  emergency.

According to Government Code Section 8625, the Governor may proclaim a  state

of air pollution emergency  "in any area affected  or  likely  to be affected

thereby when:",  [25]

      a)   He makes a  determination that conditions of disaster
          or extreme  peril  to human  safety and  property  caused
          by air pollutants  are  likely to be  beyond  the  control
          of the services,  personnel, equipment and  facilities
          of any local  government body; and either

      b)   He is  requested to do  so (i) by the  mayor  or  chief
          executive of  a city (ii) or the board of supervisors
          or the chief  executive of  a county;  or

      c)   He finds that local  authority is inadequate to cope
          with the air  pollution emergency.

 It should be noted that section 8625 states  clearly the Governor may

 declare  an air  pollution emergency  not only when the conditions above

 exist, but also when they  are likely to exist.  This allows action  to

 be taken on a forecast basis.

      Once an air pollution  emergency has been  declared, the Governor may

 assume  "complete authority  over all  agencies  of  the state  government and

 the right to exercise  within the area designated,all police power vested

 in the state" (section 8627).   Furthermore,  he may  commandeer  or use

 private  property or  personnel,  if necessary,  to  aid in  abating the

 emergency.   However, the state  must then pay  the reasonable value of

 property or personnel  services  used (section  8572).

      Thus,  it appears  that  the  Governor's extraordinary powers during

 an air pollution emergency  clearly  would support both  direct and  indirect
                                    128

-------
traffic control measures.  In addition,if private property and services

are not actually taken over or commandered by local  or state govern-

ment officials  in order to implement an episode control  plan, then,

no compensation is required for lost wages or profits.   However,  should

a particular type of business be singled out for control,  or if a

person's (or business') service or property are commandered by the

State\ reasonable compensation is probably necessary, [3].

Federal

     The Federal government, through EPA Region IX,  may implement

emission reduction measures on the basis of two different  sections  of

the Clean Air Act as amended (December 31, 1970), [6].   They are:

A.  Section 113, (a)(2) - Federal Enforcement

     Section 113 (a)(2) allows the Administrator  of EPA to enforce a

state implementation plan (including episode control programs)" when-

ever, on the basis of information available to him,  . . .  [he] finds

that violations .  .  .  are so wide spread that such violations appear  to

result from a failure of the State ... to enforce  the plan effactivity

..."  He would do so by:

     1)  notifying the state,

     2)  allowing  thirty  days for the state to begin enforcing
         the plan; and if the state fails to do so

     3)  announcing  this publicly and commencing federal  enforce-
         ment

Section 113 (a)(2) also provides that ". . . the Administrator may

enforce any requirement of such plan with respect to any person -

     (A)  by issuing an order to comply with such requirement,

     (B)  by bringing  a civil  action . . . ."
                                   129

-------
B.  Section 303   Emergency Powers




     This section states "...  the Administrator upon receipt of



evidence that a pollution source  or combination of sources  (including



moving sources) is presenting an  imminent and  substantial  endangerment



to the health of persons, and that appropriate State  and Local  authorities



have not acted to abate such sources, may bring suit  on behalf  of the



United States ...  to immediately restrain  any person  causing  or



contributing to the  alleged pollution to  stop  the  emission  of air



pollutants causing or contributing to such pollution  or to  take such



other action as may  be necessary."




     From the above  descriptions,it is clear that  the actual  legal



authorities which sections  113 and 303 grant the EPA  are somewhat



different in nature.   The authority provided under section  113  is more



long term.  It does  not require convincing a judge, and allows  action



to be taken before there is an imminent and  substantial  endangerment to



health.   However, it  also requires a longer  time to invoke  (30  days).



In addition, the EPA  would  be enforcing a state plan  which  may  or may



not allow indirect forms of VMT reductions and which  may or may not



have certain kinds of compensatory obligation  connected with  it.



(Section 110(c) does, however, allow the  EPA to promulgate  plans  or



portions thereof when it finds the State's plan unacceptable).   Section  303,



on the other hand, grants the EPA short-term authority  to  implement



episode  controls. Under its provisions,  the EPA may  take action



against  both stationary and moving sources.  Further, it appears  to



allow for both direct and indirect forms  of  traffic regulation.  How-



ever, it also means  that a  judge  must be  convinced that not only  does
                                   130

-------
an imminent and substantial endangerment to health exist but that state and



local officials have failed to act properly and that the injunctive action



being sought is indeed appropriate and correct.




     Thus, it appears that the EPA has the legal  authority to implement



both direct and indirect episode control measures in the South Coast Basin.



This authority is certainly granted under Section 303 and most likely could



be incorporated by means of Section 112 in conjunction with Section 110.



Further, it also appears, at least as far as the  Clean Air Act is  concerned,



that the EPA will not be liable for lost wages and profits.  In addition,  this



exemption from liability also seems to extend down to state and local  levels.




4.4.2  Imp!ementation Organization




     In order for an air pollution episode control plan to be implemented



effectively, it must be supported by an organizational  structure capable of



sustaining four different kinds of operations:  surveillance, strategy



selection, enforcement, and communications.  This section briefly  describes



three alternative approaches to the design of just such an organizational



structure.  They were developed by, respectively, the Los Angeles  County



Air Pollution Control District, the Argonne National  Laboratory (for the



City of Chicago), and Region IX of the Environmental  Protection Agency.




Los Angeles APCD




     The supporting organization for the Los Angeles APCD episode  control



plan consists of six different parts (see Figure  4.2).   They are the air



pollution meteorologist, the Air Pollution Control Officer, the Emergency



Action Committee, the Air Pollution Control Board, the Director of



Enforcement, and the Special Services Division.  The function of each



is outlined below, [18], [19]:





                                   131

-------
                 AIR POLLUTION
                 METEOROLOGIST
co
                                                     AIR POLLUTION
                                                     CONTROL  BOARD
AIR POLLUTION CONTROL
       OFFICER
                                      -Ess,
                                          DIRECTOR OF ENFORCEMENT
                                                                        I
                                               SPECIAL SERVICES DIVISION
EMERGENCY ACTION
    COMMITTEE
                                                        Figure 4.2
                             Los Angeles APCD  	 Organizational  Structure for Episode Control

-------
     1.   Air Pollution Meteorologist -- The task of the meteoro-
         logist is to provide a real-time assessment and short
         range (up to 36 hours) forecast of the state of the
         atmosphere.

     2.   Air Pollution Control Officer (APCO) -- As shown in
         Figure 4.2 , the Control Officer is the key figure in
         the control operation.  He must select the appropriate
         control  strategies and supervise all personnel  to in-
         sure that the controls are being implemented properly.

     3.   Emergency Action Committee -- This committee advises
         the APCO as to which control strategies are most
         appropriate.   It consist of ten members with training
         or knowledge in air pollution, medicine, business,
         and law enforcement.

     4.   Air Pollution Control Board -- The Control Board
         (L.A. County Board of Supervisors) reviews the
         actions  of the APCO and must grant concurrence
         before he may take certain steps to abate the
         episode.

     5.   Director of Enforcement -- The Director is responsible
         for initiating episode abatement and activities and oversees
         surveillance and enforcement.

     6.   Special  Services Division -- This Division handles all
         the necessary communication with the media, public
         agencies, schools, etc.

Thus, in this particular organizational configuration, surveillance  is

handled by the air pollution meteorologist and the Director of Enforce-

ment, strategy selection by the Air Pollution Control Officer, enforce-

ment by the Director of Enforcement, and communications  by the Special

Services Division.

Argonne National  Laboratory

     Argonne National  Laboratory has modelled its episode control

organization for  the City of Chicago after the dispatching setup  used

by many utilities.  As Figure 4.3 shows, there are three functional  roles

the air pollution meteorologist, the control  strategist, and the  tactical

dispatcher.   A brief description of each role is presented below,  [8]:
                                    133

-------
                      AIR POLLUTION
                      METEOROLOGIST
   AIR POLLUTION
CONTROL STRATEGIST
                                                       1
CO
                                                       AIR POLLUTION
                                               TACTICAL RESPONSE COORDINATOR
                          Argonne National Laboratory
        Figure  4.3

      Organizational Structure for Episode Control

-------
     Air Pollution Meteorologist -- The task of the meteorologist
     is to produce a real-time assessment and short-range
     (12-36 hours) forecast of the state of the atmosphere
     in the basin and to translate these evaluations into
     air quality predictions which can be employed directly
     in the selection of control strategies.

     Control Strategist -- The control strategist seeks to
     impose the minimum degree of control that is practical
     and consistent with the prevention of significant
     harm levels.  He does this by integrating the air quality
     report and forecast provided by the meteorologist with an
     assessment of the current status of significant emission
     sources within the basin.  He is,in effect.master planner
     and policy maker.

     Tactical Coordinator — The primary responsibility of the
     tactical coordinator is to insure that the control strategy
     selected by the strategist is carried out.  He must monitor,
     direct, and deploy the various surveillance and enforcement
     units available to him.  He must also manage communications,
     acting as a liason between the control strategist and the
     public.

It is obvious from the above description that Argonne has taken  a more

simple approach to the problem, preferring to concentrate responsibility

in three key positions, rather than delegate it-to a larger number of

people as done by the LAAPCD and (as will be seen below) by EPA  Region

IX.  All four key types of operations can be handled by the Argonne

organizational structure.


Region IX EPA

     Region IX has chosen to take a team approach to the problem of organi

zation structure.  Responsibility is divided among five specialist teams,

[28]:
                                    135

-------
     1.  Surveillance Response Team

     2.  Technical Control Strategy Response Team

     3.  Public Affairs Response Team

     4.  Communications Response Team

     5.  Legal Enforcement Response Team

The activities of these teams are supervised and coordinated by the

Emergency Response Coordinator whose responsibility it is to decide

what actions if any the EPA will take during the episode, (see Figure 4.4).


     The functions of the various Special  Response Teams are as follows:


     Surveillance Response Team -- During  episode operations
     this team will  be responsible for data acquisition.  It
     will analyze data and prepare special  advisory reports.

     Technical Control Strategy Response Team -- The function
     of this team is to develop and maintain control strategies
     and emission reduction plans.  It will also help determine
     the need for legal action.

     Public Affairs  Response Team -- The Public Affairs Response
     Team will handle all  releases of information to the public.

     Communications  Response Team -- This  team will coordinate
     all communications for EPA operations  during the episode.

     Legal  Enforcement Response Team -- The Legal Enforcement
     Response Team will handle all legal aspects of the episode
     control operation, including legal briefs, depositions,
     expert witnesses, etc.


In addition to this, the Surveillance, Technical Control Strategy, and Legal

Enforcement Response Teams are responsible  for providing a representative

who will act for his team.


Evaluation


     Since  none of the organizational  structures has operated under actual,

prolonged episode conditions, it is very difficult to evaluate,and thus,

compare relative effectiveness.  All that  can be said at this time is that

                                    136

-------
                                     EMERGENCY RESPONSE
                                         COORDINATOR
PUBLIC AFFAIRS
RESPONSE TEAM
LEGAL ENFORCEMENT
  RESPONSE TEAM
TECHNICAL CONTROL
  RESPONSE TEAM
SURVEILLANCE
RESPONSE TEAM
                                COMMUNICATIONS RESPONSE TEAM
                                          ON  SCENES
                                         COORDINATOR
                                        RESPONSE TEAM
                                       REPRESENTATIVES
                                        Figure 4.4
                  EPA Region IX   Organizational  Structure  for Episode  Control
                                            137

-------
on paper at least,each of the three structures  does contain provisions for
handling the four essential  operations  mentioned above:   surveillance,
strategy selection,  enforcement,  and communication.
4.4.3  Communication Channels
     Several different types  of communication are essential  for EPA
Region IX to manage  effectively an  air  pollution episode  plan for the
Metropolitan' Los  Angeles  AQCR.  They are:
     a)  Communication with  headquarters
     b)  Communication with  personnel at  the scene of the
         episode
     c)  Communication among  personnel  at the scene of the
         episode
     d)  Communication with  the ARB and local APCDs
     e) .Communication with  other government agencies,
         commercial  firms,  industry, and  business
     f)  Communication with  the public.
     This section will  discuss possible channels for dealing with these
communications and methods  for establishing them.
     First, it is useful  to  review  the  provisions for communications that
Region IX has already incorporated into  its Regional  Emergency Operations
Control Center (REOCC).   According  to the REOCC Operation Manual, all
communications are to be  handled  by the Center's Communications Response
Team (CRT), [28].  It is  its  job  to receive and route all incoming calls,
maintain a log for incoming  calls,  coordinate dissemination of information,
and provide a buffer zone for internal  REOCC Operations.  In addition to the
CRT, the REOCC also  maintains a Public  Affairs  Response Team (PART) whose
function it is to disseminate information to the press and  the public and
to answer inquiries  from  them.  Finally,  there  are provisions in REOCC for
                                    138

-------
two On-Scene Coordinators (OSC).  During an episode,one of them is sent to

the ARB and the other to the local episode area.  One of their principal  jobs

is to assist in the transmission of data and information back to the REOCC.


     Generally speaking, the REOCC Operations Manual  defines fairly well  the

possible channels of communication.  However, it does leave open the question

of whether the channels should exist at the regional  (EPA Region IX) level,

the local level, or both.  It also fails to specify exactly what means  of

communication should be used, (radio, telephone, etc.).  These possibilities

are considered in more detail below for each type of required communication:


     a)  Communication with headquarters -- Since this will involve
         only the highest level  decisions (i.e., federal versus non-
         federal control), it should-be conducted at the regional
         level only, via the Communications Response Team.   Telephone
         should be sufficient.

     b)  Communications with personnel  at the scene of the  episode --
         Presently field personnel would maintain direct telephone
         contact with the CRT.  For a small number of field personnel
         this seems adequate.  However, for any substantial field
         operations program, it would appear more practical to send
         representatives of the CRT into the field to establish a
         local -communications center.  At this time,  direct radio
         contact seems unnecessary and too expensive.

     c)  Communications among personnel at the scene  of the episode --
         Currently, any formal communications among field personnel
         must be routed through  the CRT at the regional control  center.
         If a local communications center were established, this could
         be handled through the  local center.

     d)  Communication with the  ARB and local APCDs -- At present,
         the CRT maintains telephone contact with the APCDs and ARB.
         This is enhanced by the presence of On-Scene Coordinators,
         both at the state level and at affected APCDs.  In the
         past, it has been noted that as the episode  becomes more
         severe, it is increasingly difficult to maintain reliable
         contact with those agencies.  More reliable, pre-planned
         telephone communication should be arranged.
                                    139

-------
      e)   Communication with other government agencies, commercial
          firms,  industry, and business -- Two options appear  to
          exist.   First, communication can be channeled through a.
          field  representative or second, it can go directly to
          the  regional control center via the CRT with regional
          operations dispatching field representatives only when
          necessary.  The REOCC Operations Manual is unclear as
          to what  is intended.  However, the second option appears
          preferable, since the control operation is being run at
          the  regional level.  Telephone seems to be the only
          feasible means of establishing communications.  A toll
          free number for inquiries seems advisable.

      f)   Communication with the public -- Presently, all
          communication with the public would be handled by
          the  PART at the regional control center.  The problem
          with this is that public action is taken at the local
          level, not the regional level.  The dissemination of
          information may be unnecessarily slow since the impor-
          tant media representatives are at the local level.
          An alternate approach would be to send representatives
          of PART  into the field to hold press conferences and
          to locally coordinate communications to the media and
          to the public (via the media).  Again a pre-established
          toll free number for inquiries seems to be advisable.

      Figures  4.5 and 4.6 schematically show communications ch'annels as they

now exist along with the improvements suggested above.   It should be noted

that  a local  communications center necessarily implies  extra planning and

effort.   In all  likelihood, prior arrangement would have to be made with

agencies  in potential  trouble areas to provide for certain needs (e.g.,

office space).  However,  the benefits of such a center  may outweigh the

difficulties.


4.4.4  Activity Phasing


     A final  aspect of the  air pollution episode contingency plan which needs

to be considered is the  actual  sequencing of the control  operation activities.

Since the plan must be implemented  on short notice, twenty-four to thirty-six

hours, proper  timing  is  critical;  failure to complete one step could negate

the whole operation.
                                    140

-------
                    REGIONAL EMERGENCY
                    OPERATIONS CONTROL
                          CENTER
                                                    EPA
                                               HEADQUARTERS
                          PUBLIC
                          AFFAIRS
                          RESPONSE
                           TEAM
                                                                   ON-SCENE
                                                                 COORDINATORS
                                                    (   ON-SCENE  A
                                                    I COORDINATORS/
     FIELD  X   f    FIELD
REPRESENTATIVE)   (REPRESENTATIVE
                                                          LOCAL
                                                          APCDs
           THER GOV't
           AGENCIES,
          INDUSTRY, ETC
                        Figure  4.5

Communications Channels As Outlined in REOCC Operations  Manual

-------
ro
                                                REGIONAL EMERGENCY
                                                OPERATIONS CONTROL
                                                      CENTER
                                          EPA
                                     HEADQUARTERS
                                                      PUBLIC
                                                     AFFAIRS
                                                     RESPONSE
                                                       TEAM
    LOCAL
COMMUNICATIONS
    CENTER
                                                                                          ON-SCENE
                                                                                        COORDINATORS
                                                                                   ON-SCENE
                                                                                 COORDINATORS
             FIELD
       REPRESENTATIVE
    FIELD
EPRESENTATIVE
                                                   THER GOV't
                                                   AGENCIES,
                                                 INDUSTRY, ETC
                                   LOCAL
                                  APCD's
                                                   Figure 4.6
                                Possible Modifications to Existing Commum"cations Channels

-------
     Episode control can be divided into three phases - the initiation phase,


the control phase, and the termination phase.  Each of these phases consists


of a series of activities which must be completed before the operation can


advance to the next phase.  In addition to this, there are also certain time


constraints which the various phases must meet.  The activities and time


constraints for each phase are descussed below.



Initiation Phase



     When the air pollution meteorologist forecasts that air quality will


deteriorate to episode levels, control operations enter the initiation


phase.  In this phase, the severity of the episode must be determined and


an appropriate control strategy chosen.  Once this has been done, the choice


must be made public so that those affected may take the necessary actions.



     Strategy selection should be completed, at the latest, by 4:00 to 5:00 PM


the previous evening in order to make the evening news as an emergency


announcement.  However, even earlier strategy selection would be very helpful


in control implementation.  Selection at 1:00 to 2:00 PM would allow more


time for proper notification of radio and television broadcasters.   An


11:00 AM selection would be able to make the evening newspapers.  If


employee/employer coordination is required in the control strategy, a
                                       •4

2:00 - 3:00 PM deadline appears reasonable, [12].



     A very important trade-off is involved in scheduling the strategy


selection phase.  With a very late deadline, (say late afternoon),  imple-


mentation and proper public notification will be difficult; however, a


weather-change type forecast (based on knowing the previous day maximum


oxidant level) can be used.  With an early deadline, (say around noon),


many implementation and public notification problems will be alleviated,




                                    143

-------
but a  less accurate, future-weather type forecast must be used.   It  is
recommended that this trade-off be investigated to determine if the  gains
in forecast accuracy are worth the loss in operating time associated with
a late afternoon deadline.  A compromise alternative, which might prove to
be more effective, would be to start control implementation based on a
late morning forecast and then to alter certain aspects of the strategy
according to the late afternoon forecast.
Control Phase
     With the announcement of the control  strategy, operations enter the
control phase.  The first half of this phase is devoted to the implementation
of control measures.  By the time the work day begins (six A.M.) all necessary
activities required to implement the measures should be completed.  In the
second half, attention is turned to enforcement and surveillance which con-
tinue into the afternoon of the episode day.
Termination Phase
     In the morning or afternoon of the episode day,  as the meteorologist
compiles his forecast for the next day, the third  phase of the operation
begins.  In this phase,  officials  must determine whether to continue control
into the next day or to  terminate  control.   At the same time,  they should
continue to oversee the  enforcement of the  measures already implemented.
The time constraints for this phase of the  activities are similar to those
of the initiation phase.
                                    144

-------
                        REFERENCES - CHAPTER FOUR
1.   Adrian, Robert, California Air Resources Board,  Personal  Communication,
     September 1973.

2.   Ayerson, R., Surveillance and Analysis Division, EPA-Region  IX,
     Personal Communication, September 1973.

3.   California State Office of Emergency Services, Air Pollution Emergency
     Traffic Control Planning Committee, Traffic Abatement Plan for Air
     Pollution Episodes, 1973.

4.   California Transportation Agency, Department of  Public Works, Division
     of Highways-07, Los Angeles Area Freeway Surveillance and Control
     Project, 1973.

5.   California Transportation Agency, Department of  Public Works, Division
     of Highways-07, Los Angeles Area Freeway Surveillance and Control
     Project Service Programs, 1972

6.   Clean Air Act of 1970, "Clean Air Amendment," (P.L.  91-604).

7.   County Air Pollution Districts, Personal Communications,  September  1973.

8.   Croke, E.  J., e_t aj_, Chicago Air Pollution Systems Analysis  Program,
     Fourth Quarterly Progress Report, Argonne National Laboratory,
     ANL/ES-CC-004, pp.  226-250.

9.   Davidson,  Arthur, Meteorologist - Los  Angeles County Air  Pollution
     Control District, Personal  Communication, September 1973.

10.  Duckworth, Spencer, Meteorologist - California Air Resources Board,
     Personal Communication, September 1973.

11.  Foon, Julian, Data  Processing - Los Angeles County Air Pollution
     District,  Personal  Communication, September 1973.

12.  Garetz, William, Surveillance and Analysis Division, EPA-Region  IX,
     Personal Communication, September 1973.

13.  Hernandez, Tony, Acting Director of Technical Services -  Riverside
     County Air Pollution Control  District, Personal  Communication,
     September  1973.
                                    145

-------
14.  Hopper, Charlotte,  Meteorologist  -  EPA-Region  IX,  San  Francisco,
     Personal  Communication,  September 1973.

15.  Kauper, Erwin,  Consulting  Meteorologist  -  Riverside  County  Air
     Pollution Control  District,  Personal  Communication,  September  1973.

16.  Kinosian, John, Director of  Technical  Services,  California  Air
     Resources Board, Personal 'Communication, September 1973.

17.  Leonard,  Robert, Enforcement Division -  California Air Resources
     Board, Personal Communication,  September 1973.

18.  Los Angeles County  Air Pollution  Control District, Smog Alerts -  School
     and Health Smog Warnings,  May 1972.

19.  Los Angeles County  Air Pollution  Control District, Rules  and Regulations,
     Part VII.

20.  Lust, Don, Air  Pollution Meteorologist - National  Weather Service,
     Personal  Communication,  September 1973.

21.  McMullen, Robert,  Meteorologist,  California Air  Resources Board,
     Personal  Communication,  September 1973.

22.  Riverside County Air Pollution  Control District, Personal Communication,
     September 1973.

23.  State of California. Health  and Safety Code, Division  20, Chapter 2.

24.  State of California, Health  and Safety Code, Division  26,
     "Mulford-Carrel! Air Resources  Act".

25.  State of California, Government Code, Title 2, Division  1,  Chapter 7,
     "California Emergency Services  Act."

26.  Taylor, Earl, Meteorologist  - Ventura County Air Pollution  Control
     District, Personal  Communication, September 1973.

27.  Tuff, Douglas,  Ventura County Air Pollution Control  District,  Personal
     Communication,  September 1973.

28.  United States Environmental  Protection Agency, Region-IX, Regional
     Emergency Operations Control  Center Operations Manual.

29.  Wachtenheim, A. and Keith, R.,  "Forecasting Ozone  Maxima  for Los  Angeles
     County; Los Angeles County Air  Pollution Control District Paper #69-78,
     June 1969.

30.  Wells, Ora, Director of  Enforcement - Orance County  Air Pollution
     Control District,  Personal Communication,  September  1973.

31.  Zeldin, Mel, Meteorologist - San  Bernardino County Air Pollution
     Control District,  Personal Communication,  September  1973.
                                   146

-------
         5.0  SUMMARY OF ALTERNATIVE CONTROL MEASURE EVALUATIONS



     The following sections present summaries of the control  measure

evaluations performed during this study.  Included in each summary is an

assessment of the technical effectiveness, institutional  obstacles,  and

socio-economic impacts associated with each measure.  The control  measures

fall into two general classes — stationary source controls (Section 5.1)

and traffic abatement controls (Section 5.2).  In an attempt  to be concise

and to provide a uniform format for evaluating the control measures, each

summary has been given in tabular form.  Table 5.1 gives  a brief explana-

tion of the categories included in the evaluation.  Results are given in

Tables 5.2 and 5.4.   Basically, the assessments are organized as follows:

                    CONTROL MEASURE EVALUATION SUMMARY
  •
     I.    Control Measure Description

           A.  General

           B.  Implementation Procedure

           C.  Enforcement Plan

     II.    Technical  Effectiveness

           A.  Total  VMT Reduction (Vehicular Controls Only)

           B.  Total  Emissions Reduction

     III.   Institutional  Obstacles

           A.  Implementation

           B.  Administrative Cost

           C.  Enforcement Procedure
                                    147

-------
                      Table 5.1    DESCRIPTION  OF THE  CATEGORIES  FOR CONTROL MEASURE EVALUATION
            CONTROL MEASURE NO.
                                                       EFFECTIVENESS
                                                                                     INSTITUTIONAL  OBSTACLES
                                                                                                             SOCIO-ECONOMIC  IMPACTS
CO
          General Description:   Overall  des-
          cription of control measure, who
          it affects and under what con-
          ditions .
Implementation  Procedure:  Des-
cription of key steps  in  imple-
mentation of the control  measure.
For the most part,  actions  to be
taken to establish  the necessary
regulations or  in case of impend-
ing alert.
          Enforcement Plan:   Description
          of program envisioned.for sur-
          veillance, monitoring,  and
          enforcement of the controls.
                                    Total VMT Reduction:  Total  re-
                                    duction anticipated in VMT  and
                                    a breakdown of which types  of
                                    vehicles are affected by the
                                    control measure.
                                              Total Emissions Reduction:   Re-
                                              duction expected in emissions
                                              of RHC, CO, and NO .   Percentage
                                              expressed as a fraction of  the
                                              total emission inventory.   The
                                              effect on RHC emissions from the
                                              filling of vehicle gas tanks at
                                              service stations is included in
                                              the reduction estimate.
Implementation:  Discussion of
obstacles anticipated  in  imple-
mentation of the control  measure.
Depending on the measure, these
are expected to range  from minimal
to moderate to signi ficant to sub-
stantial , in increasing difficulty.
Direct Economic Cost:  Cost
incurred by  the  community in
terms of loss  wages, salaries,
Business, and  productivity.
Other costs  include various
taxes not collected.
 Administrative Cost:  Estimate  of
 cost required to establish  necess-
 ary personnel, paperwork, media
 campaign, etc. to administer  the
 program.  Time, money, and  labor
 costs will range from minor to
 moderate to high expenditures.
Other Cost Considerations:  In-
cludes decreased efficiency  of
business or governmental
operations, inconvenience  and
nuisance costs of foregone or
wasted trips, lost opportunities,
etc.
                                                                       Enforcement Procedure:  Problems
                                                                       envisioned  in enforcing control
                                                                       measure assuming prior implemen-
                                                                       tation.   Included are problems of
                                                                       effective surveillance and mon-
                                                                       itoring.  Range given from minimal
                                                                       to  substantial.
                                    Equity:   Deals with issue of dis-
                                    crimination  among certain groups,
                                    companies, etc.  Can be either
                                    direct  (intentional) or indirect
                                    (unintentional, subtle).

-------
     IV.    Socio-Economic Impacts




           A.   Direct Economic Impacts



           B.   Other Cost Considerations



           C.   Equity




     Critical  to the successfulness of an abatement strategy is  the  technical



effectiveness  of each control  measure.  Detailed calculations  of the VMT



reductions associated with traffic abatement control  measures  are given  in



Appendix D.




5.1  EVALUATION OF STATIONARY  SOURCE EPISODE CONTROL  MEASURES




     Stationary sources, excluding automobile tank filling,  accounted for



approximately  14 percent of the total  RHC emissions during  1973, (See Table  2.5)



For purposes of reducing these emissions during an episode  alert, a  number of



control  measures offer some potential.  The three evaluated  in this  section



appear to be the most promising:




     1)   Ban Filling of Underground Service Station Tanks



     2)   Shut  Down Major Point Sources



     3)   Shut  Down Major Point Sources and Other APCD Designated



         Categories (Rule 155.2)



         a)  No Exceptions Granted



         b)  Review Committee  Established - Exemptions Allowed Upon



            Approval  of Review Committee




     Table 5.2 presents a brief summary of these measures.   Although minor,



these source categories do contribute  a significant amount  of emissions  --  2



to 4 percent each of the total RHC.  Despite relatively equal  contributions



from these source categories,  substantial differences exist in both  the
                                   149

-------
                        Table  5.2    EVALUATION  OF  STATIONARY  SOURCE  EPISODE  CONTROL  MEASURES
        CONTROL MEASURE MO. 1
     Ban Filling of Underground
        Service Station Tanks
        EFFECTIVENESS
                                       INSTITUTIONAL OBSTACLES
                                                                        SOCIO-ECONOMIC IMPACTS
 General Description:

  Prior  to  and during an emergency
  episode,  the filling of under-
  ground service station tanks
  would  be  prohibited (except for
  those  with vapor recovery systems)
 Implementation Procedure:
 1)  Obtain a complete listing of
    all service stations in the
    area and all  bulk terminals
    serving them; notify both
    groups of the proposed control
    regulations.
2)  Exempt service stations (i.e.
    those with vapor recovery sys-
    tems) must apply for and have
    received permission  to fill
    tanks during  an  alert.
3)  Establish communication chan-
    nels to be used  during alert.
 Enforcement  Plan:

1)  Organize  teams  to be  used  for
    random surveillance and  moni-
    toring of the program.
2)  Violations of the filling  ban
    to be cited and subject  to
    fine.
Emission Reduction:
       (Percent of Total)
 RHC
CO
NOy
                                 Implementation:    Moderate.
                                                        Direct Economic Cost:   Minor.
                                 Dependent on large part  on  coope-
                                 ration of service station dealers
                                 and bulk terminal operators.
                                 Strong private interest  at  work,
                                 i.e. oil companies.   Complaints  of
                                 discrimination likely with  finger
                                 pointing to the automobiles as
                                 source of pollution.   Shut  down
                                 not expected to be overly dis-
                                 ruptive.
                                                         Only a fraction of service
                                                         stations would be affected,  and
                                                         presumably little reduction  in
                                                         gasoline sales would result.
                                                         Largest impact would be to spec-
                                                         ific stations "caught" without
                                                         gasoline for the day(s).
                                  Administrative  Cost:  Moderate.

                                 Mainly cost of obtaining lists  and
                                 informing operators of service
                                 stations and terminals of control
                                 measure.  Cost also associated
                                 with reviewing and granting
                                 exemptions and organizing required
                                 communication links.
                                                                                                           Other  Cost  Considerations:
                                                         Service  stations caught without
                                                         gasoline would also lose business
                                                         in  other retail areas (e.g.
                                                         servicing,  auto accessories,
                                                         repairs).
                                  Enforcement Procedure:

                                  With approximately 15,000  service
                                  stations in the basin,  monitoring
                                  could be a very formidable task.
                                  It appears easier to monitor  bulk
                                  terminals.  Overall, a  reasonable
                                  enforcement procedure could be
                                  developed.
                                                                                                           Equity:
                                                         Singles  out  service  stations as a
                                                         potential  stationary  source for
                                                         control.   Furthermore,  it affects
                                                         older service  stations  since only
                                                         newer stations are likely to be
                                                         equipped with  vapor  recovery sys-
                                                         tems.

-------
                 Table  5.2    EVALUATION  OF  STATIONARY  SOURCE  EPISODE  CONTROL  MEASURES  - continued
   CONTROL MEASURE NO.  Z
Shut Down Major RHC  Point  Sources
         EFFECTIVENESS
                                       INSTITUTIONAL  OBSTACLES
                                                                         SOCIO-ECONOMIC IMPACTS
 General Description:

 During an emergency episode, all
 major RHC point sources, as
 defined by the EPA and listed
 by the APCD, to shut down.
Implementation Procedure:

1)  Require all companies listed
    by the APCD as major RHC  point
    sources to submit emergency
    abatement plans for review
    and approval.
2)  Establish communication
    channels for use in notifying
    affected companies.
Emission Reduction:

       (Percent of Total)
                                      RHC
 Enforcement Plan:

 1)  Organize surveillance and
    monftorfng teams for random
    inspections during alert
    period to ensure compliance.
 2)  Offenders to be cited and
    subject to fines.   Blatant
    offenses subject to legal
    cease and desist orders.
                                                  CO
                         NOV
                                  Implementation:  Substantial.
                                 Major resistance to be expected  by
                                 nearly all firms.   Certain  indus-
                                 tries, e.g. refineries, require
                                 several days to start up and shut
                                 down and emit more pollutants  at
                                 these times than under normal
                                 operations.  Lobbying interests  of
                                 these groups should not be  under-
                                 estimated in the political  arena.
                                    Direct Economic  Cost: Variable.
                                    Dependent on compliance received
                                    from affected companies.  Strictly
                                    enforced, the economic impact of
                                    curtailed production at many of
                                    these  facilities would be quite
                                    high.  In some cases a. complete
                                    shutdown disturbs several  days
                                    because of down and up time.
                                                                       Administrative  Cost:  Moderate to
                                                                     Other Cost Considerations:
High.

Cost of enlisting  cooperation of
companies affected,  obtaining
 abatement plans,  reviewing and
approving abatement  plans, etc.
will be considerable.  Substantial
effort will  be  required to "sell"
this program.
In some instances,  the costs are
disproportionate  to actual shut
down time.   For example, a few
hours disruption  in a refinery may
upset the routine for several days.
For products already in short
demand, e.g.  gasoline, such an
inconvenience may be more costly
than dollar  value of the goods
lost.
                                  Enforcement Procedure: Substantial.
                                  While only about twenty companies
                                  are  affected, they represent many
                                  of the wealthiest and most influen-
                                  tial groups.  Two main problems
                                  will be:
                                  1) conscious violation of companies
                                  willing to pay fines and
                                  2) obtaining cease and desist court
                                  orders in time to make a percep-
                                  table difference in pollution.
                                   Equity:

                                   Singles out  the major point
                                   sources while  not controlling
                                   many other significant source
                                   categories.

-------
                           Table  5.2    EVALUATION OF STATIONARY  SOURCE EPISODE CONTROL MEASURES  -  continued
              CONTROL  MEASURE  NO.   3a
          Shut Down Major Point Sources  &  Othe
          APCODesignated Categories - No
                                             EFFECTIVENESS
                                                                           INSTITUTIONAL OBSTACLES
                                                                                                             SOCIO-ECONOMIC IMPACTS
cn
ro
            ueneral  Description:

            During an emergency episode,  the
            following facilities would  be
            shut down:

            1)  All  major RHC point  sources as
                defined by the EPA,  and
            2)  All  other facilities covered
                by the  LACAPCD's Rule 155.2
Implementation  Procedure:

1)  Obtain complete listing  of all
    affected facilities  from the
    local APCD, as well  as shut
    down plans  developed for
    APCD emergency alert program.
2)  Establish communication  chan-
    nels for use in notifying
    affected companies.   Effort
    would be minimized by coordi-
    nating efforts with  the  APCD.
           Enforcement Plan:

           1)  Organize surveillance and moni-
               toring teams for random
               inspections during alert
               period to ensure compliance.
           2}  Offenders to be cited and sub-
               jected to fines.  Obvious and
               intentional violations subject
               to legal cease and desist
               orders.
                                    Total VMT Reduction:
                                               Emission Reduction:
                                                      (Percent  of Total]
                                                 RHC
                                                              CO
NOV
                                                                      Implementation: Substantial.
                                             Direct Economic  Cost: Variable
                                                                      Major resistance expected from
                                                                      nearTy every firm and private
                                                                      interest  involved.  Problem of
                                                                      rapid shut down in some industries
                                                                       e.g.  refineries) which require
                                                                      several days to close down.   Close
                                                                      cooperation and support required
                                                                      from local APCD to effectively
                                                                      implement this control measure.
                                             Dependent  on  compliance and
                                             cooperation of  the affected
                                             facilities.   Strictly enforced,
                                             the economic  impact  of curtailed
                                             production at many of these
                                             facilities would be  quite high.
                                             In numerous instances, a complete
                                             shut down  disturbs several days
                                             due to start  up time.
          Administrative  Cost:  High.

          A large  number  of facilities are
          affected requiring a very good and
          rapid  communication system.
          Assisting and ensuring the
          development of  company abatement
          plans  will be a formidable task.
                                                                                                                     Other Cost Considerations:
                                                                                                                     Many of the disruptions  and  in-
                                                                                                                     conveniences far exceed  the  simple
                                                                                                                     dollar value of the goods.
                                                                       Enforcement  Procedure: Substantial
                                                                      Since so many companies are
                                                                      affected, successful implemen-
                                                                      tation is largely dependent on
                                                                      cooperation of the companies.
                                                                      Violations would be difficult to
                                                                      detect and monitor.  Many com-
                                                                      panies might be willing to pay
                                                                      the  fines imposed rather than
                                                                      shut down.
                                             Equity:

                                             Focuses on major point sources
                                             and other APCD categories  for
                                             control while not controlling
                                             many other smaller sources
                                             which also make significant
                                             contributions.

-------
                           Table  5.2     EVALUATION  OF STATIONARY SOURCE EPISODE CONTROL MEASURES  -  continued
   CONTROL MEASURE NO.   3b
Shut Down Major Potnt Sources  &
Qther APCD Designated Categories

                  t) esc nation:
                                                       EFFECTIVENESS
                                                                                      INSTITUTIONAL OBSTACLES
                                                                                                                        SOCIO-ECONOMIC IMPACTS
cn
CO
          During an emergency episode,  the
          following facilities would be
          shut down, unless prior approval
          for exemption had been granted.
          Such exemptions would be reviewed
          and granted by a review committee:
          1)  All major potnt sources as
              defined by the EPA, and
          2)  All other facilities covered
              by the LACAPCD's Rule 155.2
              (see LACAPCD Rules and
              Regulations).
 Implementation Procedure:
 1)  Obtain complete listing of all
     affected facilities from the
     local APCD, as well as shut
     down plans developed for APCD
     emergency alert program.
 2)  Establish communication chan-
     nels for use in notifying
     affected companies.  Effort
     would be minimized by coordi-
     nating efforts with the APCD.
 3)  Establish review committee for
     evaluating request for
     exemption to abatement actions
           Enforcement Plan:

           1)  Organize surveillance and moni
              toring teams for random
              inspections during alert
              period to ensure compliance.
           2)  Offenders to be cited and sub-
              ject to fines.   Obvious and
              intentional violations subject
              to legal cease and desist
              orders.
                                              Total VMT Reduction:
                                               Emission Reduction:
                                                      (Percent of Total]
                                                RHC
CO
N0y
                                                                                 Implementation:   Substantial
                                                                        Major resistance expected  from
                                                                        nearly every firm and  private
                                                                        interest involved.   Problem  of
                                                                        rapid shut down in  some  industries
                                                                        (e.g. refineries) which  require
                                                                        several  days to close  down.  Close
                                                                        cooperation and support  required
                                                                        from local APCD to  effectively
                                                                        implement this  control measure.
                                                        Direct Economic Cost:  Variable

                                                        Dependent on  compliance and
                                                        cooperation of the  affected
                                                        facilities.   Strictly enforced,
                                                        the economic  impact of curtailed
                                                        production  at many  of these
                                                        facilities  would be quite high.
                                                        In numerous  instances,  a complete
                                                        shut down disturbs  several days
                                                        due to  start  up time.   The impacts
                                                        are lessened  considerably if
                                                        exceptions  are granted  to hardship
                                                        cases,  e.g. cases where shutdown
                                                        for one day will disrupt activities
                                                        nvpr ypry Inng ppHnHc	.	
                     Administrative Cost:  High
                     A  large number of facilities are
                     affected requiring a very good and
                     rapid communication system.
                     Assisting and ensuring the
                     development of company abate-
                     ment plans will be a formidable
                     task, as will be the review
                     for exemptions requested.
                                                                                                                    Other Cost Considerations:

                                                                                                                    Many of  the  disruptions and  in-
                                                                                                                    conveniences  far  exceed the  simple
                                                                                                                    dollar value  of the  goods.
                                                                        Enforcement Procedure: Substantial

                                                                        Since so many  companies are
                                                                        affected, successful  implemen-
                                                                        tation is largely  dependent on
                                                                        cooperation of the companies.
                                                                        Violations would  be difficult to
                                                                        detect and monitor.   Many com-
                                                                        panies might be willing to pay
                                                                        the fines imposed  rather than
                                                                        shut down.
                                                        Focuses on major point sources and
                                                        other APCD categories for control
                                                        while not controlling many other
                                                        smaller sources which also make
                                                        significant contributions.
                                                        Questions of inequity are also
                                                        possible in granting of exemp-
                                                        tions.  Special interests and
                                                        lobbying groups may influence
                                                        review committee.

-------
institutional obstacles and socio-economic costs associated with the

respective control measures.

     Controlling the filling  of underground service station tanks, primarily

through control  at the 50 or  so bulk terminals  within the basin, appears as

one of the more  equitable and least disruptive  control  measures under consid-

eration.  While  the service stations are isolated for control, the impacts

are expected to  be minor for  several reasons:

     e - Among service stations, there is no intended discrimination,
        since the normal  filling schedules are  relatively fixed on
        varying  days and times of the week.

     »  The number of stations affected is expected to  be a small
        percentage of the total since under normal  conditions,
        service  stations get  filled about once  or twice a week.
        Most stations fill  their underground tank with  approxi-
        mately one day's supply still in the tanks.   Thus, in
        case of  an alert, it  is expected that only a fraction of
        stations would be scheduled for a refill  and significantly
        less would actually be completely depleted and  have to shut
        down during the day.

     e  Sufficient gasoline supplies would be available                ,  v  -
        basinwide so that any motorists wishing to purchase
        gasoline could do so.

     The control of "major RHC point sources,"  as defined by the EPA and

listed by the Los Angeles APCD, presents a number of problems related to

institutional obstacles, technical  difficulties,  and socio-economic impacts.

The companies affected represent many of the most influential groups in the

area; any plans  to shut down  their operations can be expected to stir up

major opposition from private interests.  In addition,  certain types of

operations, e.g. refineries,  require more than  a day to either completely

shut down or to  start up.   From a technical  standpoint, attempts to close

these facilities down on very short notice could result in more emissions

being emitted due to the inherent inefficiencies associated with these

procedures, [3].

                                    154

-------
     In addition to the "major RHC point sources" (of which there are
approximately 20), other smaller but significant RHC sources have been
listed by the L.A. County APCD as important emitters, (see Table 3.2).
If these source categories, named in Rule 155.2 (L.A. County's Rules and
Regulations), were also effectively controlled, approximately 4 percent
of the total RHC would be eliminated.   For reasons analogous to those
given under "major point source" control, institutional  and technical
problems arise with any attempts to completely shut down all the companies
affected by Rule 155.2 on short notice, i.e.  granting no exceptions.
Because of these difficulties, a more acceptable approach to controlling
these sources would be to establish an Emergency Review Board to review
applications for waiver or exemption to episode controls.  Since the
major sources and other listed companies presently file for permits to
operate, an additional question could be added to the normal permit
petition to allow certain facilities to request a waiver to the shut down
proceedings during an alert.  The requests would be reviewed and granted
only in case of technological problems (e.g.  lengthy shut down or start
up times, emitting more pollutants during a shut down) or for companies
providing essential goods and services.  While reducing the overall effec-
tiveness of this control slightly (estimated now to reduce total RHC by
approximately 3 percent), significant gains are made in reducing the socio-
economic impacts of the control measure.
     In conclusion, it is recommended that two stationary source control
measures be incorporated into the final episode control plan:
     •  Ban filling of underground service station tanks
     o  Shut down major point sources and other APCD designated
        categories (Rule 155.2) with exemptions granted upon
        approval of a review committee.
                                    155

-------
This recommendation is based on  an  attempt  to  impose  reasonably effective
technical  controls with as  modest  a disruption and  socio-economic impact as
possible.   It also attempts to trade-off the ease of  implementation (i.e.
institutional obstacles) with control  strategy complexity (i.e.  number of
control measures).  Successfully implemented,  it  is estimated  that approxi-
mately 7 percent of the total  RHC  in the basin can  be eliminated.
5.2  EVALUATION OF TRAFFIC  ABATEMENT EPISODE CONTROL  MEASURES
     Gasoline powered motor vehicles,  including auto  tank filling at service
stations, accounted for 77% of the  RHC emissions  in the Metropolitan
Los Angeles AQCR during 1973, (see  Table 2.5). Numerous alternative traffic
abatement measures can potentially  be  used  to  limit emissions  from this
major source during air pollution  episodes.  Among  these are twenty-four
which have been evaluated in this  study and which  include those considered
in the OES Report and the EPA Region IX Study, (see Table 5.3), [1], [2].
     Table 5.4 presents a qualitative  evaluation  of the alternative traffic
abatement measures.  The control measures are  separated into three groups:
Direct Controls on Traffic  (Control Measures   1-10), Controls on Traffic
Support Facilities (Control Measures 11-15),  and  Controls on Incentives
to Travel  (Control Measures 16-24).  As noted  previously, most of these
evaluations represent a qualitative summary of the  measures based upon
available information.  The VMT  reductions  associated with each measure are
determined in Appendix D.
     Table 5.4 indicates that nearly all traffic  abatement measures present
significant institutional obstacles and have  quite  disruptive socio-economic
impacts.  Clearly, though,  certain measures will  result in more problems  than
others.  The only measure with good socio-political  acceptability  is  "Voluntary
Abatement", but in reality this  turns out to   be no control measure  at  all.
                                    156

-------
 DIRECT CONTROLS ON TRAFFIC
 1.   Voluntary Traffic Abatement
 2.   Windshield Sticker Program
 3.   License Plate Lottery
 4.   Ban Non-Essential Government Vehicles
 5.   Ban (Private) Fleet Vehicles
 6.   Ban (Private) Fleet Vehicles and Use of Vehicles for Commercial  Purposes
 7.   Ban All Non-Emergency Traffic
 8.   Mandatory Car-Pooling on Freeways
 9.   Mandatory Carpooling on Freeways and Major Thoroughfares
10.   Commercial Abatement Schemes

 CONTROLS ON TRAFFIC SUPPORT FACILITIES
11.   Close Freeways (or Certain Freeway Lanes)
12.   Limited Availability of Parking
13.   Special Parking Tax
14.   Gasoline Rationing
15.   Close Gasoline Stations

 CONTROLS ON INCENTIVES TO DRIVE
16.   Close (Non-Essential) Federal  Agencies
17.   Close (Non-Essential) Government Agencies (Federal, State & Local)
18.   Close Government Contractors
19.   Close Shopping Facilities at Regional Shopping Centers (RSC's) and
     Central Business Districts (CBD's)
20.   Close Recreational Facilities
21.   Work Holiday for Major Firms
22.   Rotating Work Holiday for Major Firms
23.   Operate on Sunday Status
24.   Complete Work Holiday With Recreation Closed
                                    157

-------
                                  Table  5.4    EVALUATION OF  TRAFFIC  ABATEMENT EPISODE CONTROL MEASURES
               CONTROL MEASURE NO.  1
               Voluntary  Traffic Abatement
                                               EFFECTIVENESS
                                                                             INSTITUTIONAL OBSTACLES
                                                                                                              SOCIO-ECONOMIC IMPACTS
en
00
             General  Description:  An appeal
             to the  publicfor voluntary car-
             pooling,  use of public transit,
             and a  reduction in optional auto
             travel;  promotion of alternative,
             non-polluting  forms of travel --
             walking,  bicycling.
  Implementation Procedure:
1) Media compaign  to inform public
   of the pollution problem and
   the necessity of taking control
   action.   In particular, speci-
   fic and  detailed health warn-
   ings are necessary for  suscep-
   tible groups—young, old,
   asthmatics, etc.
2) Special  announcements informing
   the public of an impending
   episode  and urging them to  take
   appropriate steps.
3) Public service  spots on TV  and
   radio to assist people  in
   arranging carpools, etc,	
            Enforcement Plan:
            None - strictly a voluntary
            program.
                                      Total VMT Reduction:
                                          Negligible  to 5%.

                                          Most  likely  negligible.

                                          Applicable  to LDMV's.
                                                 Emission Reduction:
                                                      (Percent of Total..]
                                                   RHC
CO
                                                           negl.-
                                                                         NO
                                                                           X
        negl.-
                    Implementation:   Minimal.
Direct Economic  Cost:  Minor.
                    urrent APCD emergency  alert
                    Drogram needs only to be  modified
                    and extended.  Expansion  should
                    emphasize public awareness.
Since control  actions are
voluntary,  costs  incurred are
voluntarily "donated" to air
pollution  cause.   Voluntary
control  cannot be anticipated
to invoke  significant response
or economic cost.
                                                                                   Administrative Cost:
                                                                                                          Minor.
                                                                                                                      Other Cost Considerations.:  Minor.
                    Costs result primarily  from an
                    expanded media campaign.   Some
                    costs may be donated  by media
                    stations as a public  service.
                    Program can be readily  instituted
                    with, existing agency  organization.
                                                                                                                      Reasoning for  other  monetary,
                                                                                                                      opportunity, or  inconvenience
                                                                                                                      costs similar  to above.
                                                                        Enforcement Procedure:   None.

                                                                        Since the  program  is voluntary,
                                                                        no enforcement  procedures are
                                                                        required.
                                                        Equity:
                                                       Since the program 1s voluntary,
                                                       no legal discrimination is
                                                       involved.

-------
                        Table  5.4    EVALUATION  OF  TRAFFIC  ABATEMENT  EPISODE  CONTROL  MEASURES  -  continued
              CONTROL MEASURE NO.  2

              Windshield Sticker Program
                                             EFFECTIVENESS
                                                                           INSTITUTIONAL OBSTACLES
                                                           SOCIO-ECONOMIC  IMPACTS
in
10
            General Description:   Window
            stickers ot various colors would
            be issued to vehicle owners on the
            basis of emissions.  During an
            episode,, and depending on its
            severity, traffic bans would be
            placed on vehicles with certain
            colored stickers.  Alternatively,
            colors could be issued randomly,
            and color bans could be rotated so
            as to eliminate discrimination.
Implementation  Procedure:
1) Issue window stickers  at  some
   appropriate  time during the
   year, e.g.  during annual
   registration.
2) Inform the  public as  to the
   nature of the restrictions
   which will  be imposed  during an
   episode.
3) Restrict  auto travel  on the
   basis of  windshield sticker
   color and episode severity.
             Enforcement  Plan:   [_aw enforce-
            ment officers would see that
            stickers are indeed displayed
            and that vehicles are operated in
            accordance to the restrictions.
            Offenders subject to citations
            and/or fines.
                                    Total  VMT Reduction:

                                        Flexible.  Maximal is 50-85%
                                        depending on implementation
                                        and enforcement procedures.

                                        Applicable to LDMV's, HDMV's
                                        and diesels.
                                                 Emission Reduction:

                                                      (Percent of Total]
                                                   RHC
                                                 45 - 70$
                                                (maximal)
CO
                                                 50  -
                                                (maximal)
                                                                          NO,
          35-60^
         (maximal',
                    Implementation:   Moderate.
                    In addition to an extensive media
                    campaign to inform the  public of
                    the program,  an additional layer
                    of paperwork  is added to  the
                    State's Dept.  of Motor  Vehicles.
                    A staggered registration  system
                    would ease paperwork bottlenecks,
                    but create other problems —
                    adjusting to  new system.
Direct Economic  Cost:  Moderate.
 Dependent on number of vehicles
 and people affected which is
 dependent on the severity of the
 episode.  Impact primarily on
 lost wages and productivity of
 individuals and businesses with
 no alternative modes of travel.
                    Administrative  Cost:  Moderate to
                    High.   Additional  incremental
                    cost for DMV to execute this
                    program would be significant.
                    Cost involves organizing
                    classification  system and
                    exemptions to be allowed,
                    issuance of stickers and
                    insuring that the proper
                    sticker is attached to the
                    right vehicle.
                                                                                                                      Other Cost Considerations:
                                                                                                                       Will  have  some  impact on
                                                                                                                       reducing the  services and
                                                                                                                       efficiency of public and private
                                                                                                                       business operations.
                                                                       Enforcement Procedure: Significant

                                                                       Two major enforcement problems --
                                                                       identification  of violators and
                                                                       time required to cite violators.
                                                                       Identification  can only be made
                                                                       at close  range, must differentiate
                                                                       similar looking stickers, account
                                                                       for out-of-state vehicles, etc.
                                                                       Assuming  widespread violations,
                                                                       only a small  fraction could ever
                                                                       be cited.
                                                        Equity:   Sticker colors  based
                                                        on emissions discriminate
                                                        against poor who own the oldest
                                                        vehicles, drive less and for
                                                        more essential purposes.  Colors
                                                        issued randomly are clearly
                                                        more equitable in this sense.

-------
                         Table  5.4    EVALUATION OF  TRAFFIC  ABATEMENT EPISODE CONTROL MEASURES  -  continued
                CONTROL MEASURE NO.   3
                License Plate Lottery
                                             EFFECTIVENESS
                                                                           INSTITUTIONAL OBSTACLES
                                                  SOCIO-ECONOMIC  IMPACTS
cr>
O
              General Description:   All  vehicles
              with license plates having  speci-
              fied last digits  are banned from
              driving during the  episode  alert.
              The effect on traffic reductions
              is varied by banning more  than
              one number, in accordance  to  the
              severity  of the pollution  problem.
Implementation  Procedure:
 In  the event of an impending
 episode, the number(s) on those
 cars to be banned is widely
 publicized through the media.
 Conceivably this could be done
 at  the beginning of the alert
 season, similar to the military's
 draft lottery.
             Enforcement Plan:  Law enforce-
             ment  officers would ensure that
             only  allowed vehicles were in
             use  in  accordance to the restric-
             tions.  Offenders subject to
             citations and/or fines.  Addi-
             tional  staffing  required during
             alert period.
                                    Total  VMT Reduction:

                                        Flexible.   Maximal is 50-85%
                                        depending  on  implementation
                                        and enforcement procedures.
                                        Applicable to LDMV's, HDMV's
                                        and diesels.
                                                  Emission  Reduction:

                                                       (Percent of Totalj
                                                    RHC
                                                  45  -  70
                                                 {maximal
                                                  CO
                                                50-80*
                                               (maximal)
   NOX
 35  -  60%
(maximal)
           Implementation:
                                                                                      Minimal
Direct  Economic  Cost:  Moderate.
           Involves  informing the public
           sufficiently in advance so that
           alternate plans can be made for
           those  vehicles affected.
Dependent on  number of  vehicles
and people affected which  is
dependent on  the  severity  of the
episode.   Impact  primarily due
to lost wages  and productivity
of individuals  and businesses
with no alternative modes  of
travel.
                                                                                    Administrative Cost:  Minor.
                                                                                                                       Other Cost Considerations:
                                                                                    Cost of media campaign  to  inform
                                                                                    the public of the  program.
                                               Some  impact on  reducing
                                               efficiency of business and
                                               governmental operations.
                                                                       Enforcement  Procedure: Significan
                                                                       Two major problems  are identifica-
                                                                       tion of violators and time re-
                                                                       quired to issue  citations.  Only
                                                                       an insignificant number of viola-
                                                                       tors could ever  be  cited due to
                                                                       the manpower available for enforc
                                                                       me nt.
                                               Equity:    Subtle  form  of  discri-
                                               mination to one  car  families,
                                               who are  more likely  to be banned
                                               from auto use than multiple  car
                                               families.  Since  lower income
                                               groups are more  likely to own
                                               only one car, this discriminates
                                               against  the poor

-------
                         Table 5.4     EVALUATION  OF  TRAFFIC ABATEMENT  EPISODE  CONTROL  MEASURES  -  continued
               CONTROL MEASURE NO.  4

             Ban Non-Essential Govt. Vehicles
                                             EFFECTIVENESS
                                                                           INSTITUTIONAL OBSTACLES
                                                                                                              SOCIO-ECONOMIC IMPACTS
CTl
             General Description:
             All  travel  by  government-owned
             vehicles  would be banned except
             for  emergency
Implementation  Procedure:
1)  A complete description  of all
    vehicles covered under  the
    proposed ban would be pub-
    lished.  Exemptions would be
    carefully rev-iewed and  allowed
    only for critical  needs or
    services.
2)  Operators of all affected
    vehicles would be  notified in
    event of an impending episode.
             Enforcement  Plan:
             1)   Cooperation with all agencies
                 involved would be solicited
                 and  established well in
                 advance.
             2)   During an alert, all agencies
                 would be contacted again to
                 ensure cooperation and pro-
                 gram effectiveness.
             3)   Offenders would be reported
                 to responsible governmental
                 agency.
                                    Total  VMT Reduction:

                                        1-1/2 percent.

                                        Applicable to LDMV's and
                                        HDMV's.
                                                 Emission Reduction:

                                                      (Percent of Total!
                                                   RHC
CO
                                                                         NOX
                    Implementation:    Moderate.
                    Small, well-defined vehicle
                    population is involved with
                    established lines of communica-
                    tion.  Key to success of program
                    is enlisting cooperation of  all
                    agencies involved initially.
                    Direct communication channels
                    during alerts should be readily
                    available and pose no major
                    problems.
Direct Economic  Cost:   Minor to
  Moderate.

  Dependent on number and types
  of agencies closed or slowed
  down.  Losses primarily those
  of wages and productivity of
  affected individuals.
                    Administrative Cost:   Moderate.

                     All affected agencies required
                     to establish contingency  or
                     action plans for alerts.   Plans
                     must be widespread to individuals
                     actually affected.   List  of
                     exempt vehicles  or categories
                     must be established.
                                                                                                                      Other Cost Considerations:
 Decrease in efficiency of
 governmental operations.
 Secondary impacts accrue  to
 private businesses who have
 to deal with governmental
 operations.   Inconvenience
 and opportunity costs felt
 by patrons of affected services.
                                                                       Enforcement Procedure:  Moderate.

                                                                       Dependent on good faith coopera-
                                                                       tion and enforcement by local
                                                                       agencies involved.   Consistent
                                                                       and effective action by agencies
                                                                       involved appears to pose some
                                                                       problems
                                                         Cost borne primarily  by
                                                         governmental  agencies.   Some
                                                         cost borne by groups  which
                                                         must deal  with the  govern-
                                                         mental  operations.

-------
                         Table 5.4   EVALUATION OF  TRAFFIC ABATEMENT EPISODE CONTROL MEASURES  -  continued
              CONTROL MEASURE NO.    5
            Ban (Private) Fleet Vehicles
                                             EFFECTIVENESS
                                                                            INSTITUTIONAL OBSTACLES
                                       SOCIO-ECONOMIC  IMPACTS
Ol
ro
            General  Description:
            Travel by private fleet vehicles
            (10 or more vehicles registered
            in one name) would be banned
            unless previously exempted.
            Examples of exempt categories
            might be taxis and transit
            vehicles, deliveries of perish-
            ables, delivery to essential
            services (e.g., hospitals)  and
            emergency utility services.
implementation Procedure:
 1)  Arrive  at  a working definition
    of exempt  categories or fleets.
 2)  Establish  an exempt fleet
    vehicle  identification system
    e.g. window sticker.
 3)  Encourage  and promote contin-
    gency planning by affected
    fleets.
 4)  Establish  communication
    channels to be used in the
    event of an impending episode.
            Enforcement Plan:
            Law  enforcement  personnel would
            cite operators of fleet vehicles
            driving  illegally.  Other enforce-
            ment teams would spot check
            establ isnments maintaining large
            fleets to ensure compliance with
            ban.   Violators would be subject
            to  legal action.
                                    Total VMT Reduction:

                                         5% of LDMV VMT

                                         30% of HDMV VMT
                                                Emission Reduction:
                                                     (Percent  of  Total)
                                                  RHC
                                                              CO
                                                                        NO,
Implementation:   Moderate.
Direct  Economic  Cost: Moderate.
  Fleet  vehicle ownership data
  would  have  to be obtained and
  updated  from the DMV.  All
  affected fleets would have to
  be notified, exemptions given
  where  appropriate, and instructed
  as to  proper response. Soliciting
  cooperation from ail groups
  appears  to  pose some problems.
 Cost borne  by  companies whose
 fleets  are  banned  and  include
 loss of business,  productivity
 of employees,  salaries, etc.
 Wholesale  trade  in  Los Angeles
 County  amounts to  about $50
 mi 11 ion per day.
 Administrative Cost:   Moderate.

  Identifying  all fleets, classify-
  ing vehicles by exempt status,
  informing  companies involved of
  alert  responses.   Cost to
  companies  include  contingency
  planning with enforcement agency.
                                                                                                                     Other Cost Considerations:
 Loss  of goods  and  services  to
 individuals  and  companies
 dealing with companies  owning
 large fleets.
                                                                       Enforcement Procedure:   Moderate
                                                                        to Significant.
                                                                        Dependent on cooperation  of
                                                                        affected groups.   Problems exist
                                                                        in recognizing and citing banned
                                                                        vehicles.  Spot  checking  estab-
                                                                        lishment lots is  difficult,  due
                                                                        to the numerous  locations
                                                                        involved.
                                     Since the  program  is  directed
                                     at fleet owners, it discimi-
                                     nates against  this group of
                                     vehicle owners.

-------
            Table  5.4    EVALUATION OF  TRAFFIC  ABATEMENT  EPISODE  CONTROL  MEASURES -  continued
  CONTROL MEASURE NO.    6
Ban (Private) Fleet Vehicles and
Use of Vehicles for Comml.  Purpos^j
EFFECTIVENESS
                              INSTITUTIONAL OBSTACLES
                                                                 SOCIO-ECONOMIC IMPACTS
General  Description:

 Travel  by vehicles for commer-
 cial  purposes would be banned.
 This  measure essentially extends
 the fleet vehicle ban to all
 commercial travel.  Certain
 categories would be exempt (see
 "Ban  Private Fleet Vehicles").
                                    Total VMT Reduction:

                                         7-102  of LUMV  VMT

                                         50 -  75% of  HDMV VMT
                                         Depending on  enforcement
                                         procedures.
            Implementation Procedure:
            1)  Arrive at a  working defini-
	i             tion of exempt vehicles.
Q-,          2) Establish an  exempt vehicle
to             identification program.
            3) Encourage and promote contin-
               gency planning by the affected
               vehicle owners.
            4) Establish communication
               channels to be used in the
               event of an impending episode.
                                     Emission  Reduction:
                                          (Percent of Total)
                                       RHC
Enforcement Plan:
Law enforcement  officers would
site offenders.   Alternatively,
spot checks could be  made  at
select commerical  establishments.
      CO
                                                10  - 15%
                                                                        NOV
              9 - 13%
                         Implementation:   Significant.
                          A large scale effort would be
                          required to properly identify
                          the vehicles to be affected.
                          Informing these owners  of the
                          program and soliciting  their
                          cooperation would be a  formidable
                          task.
                                                                                                                      Direct Economic  Cost:   High.

                                                                                                                      Since most businesses would be
                                                                                                                      at least partially affected,
                                                                                                                      economic activity in the region
                                                                                                                      would be greatly reduced.
                                                                                                                      Daily economic activity in Los
                                                                                                                      Angeles SMSA is  approximately
                                                                                                                      $160 mi 11 ion.
                          Administrative Cost:   Moderate
                          to High.

                          Identifying all  vehicles  to  be
                          subject to control, granting
                          exemptions equitably,  informing
                          companies of proper response  in
                          event of an alert,  and encourag-
                          ing public support  of  the
                          program would require  a
                          significant level  od effort.
                                                                                                          Other Cost Considerations:

                                                                                                          Since both public and private
                                                                                                          business and governmental
                                                                                                          operations are affected, a
                                                                                                          high inconvenience cost  is
                                                                                                          incurred by the public  for
                                                                                                          loss of goods and services.
                          Enforcement  Procedure:  Significant)
                          Main difficulty is  one  of  identi-
                          fying violators.   In some  cases,
                          commercial vehicles are indis-
                          tinguishable from other vehicles.
                          Also, some commercial  vehicles
                          are used for private family  uses
                          as wcl1.
                                                                                                                      Singles out fleets and commer-
                                                                                                                      cial vehicles for ban, thus
                                                                                                                      affecting private businesses
                                                                                                                      heavily.  Businesses with large
                                                                                                                      fleets or auto dependence are
                                                                                                                      affected most.  Large businesses
                                                                                                                      may suffer more than small  ones.

-------
                          Table  5.4    EVALUATION  OF  TRAFFIC  ABATEMENT  EPISODE  CONTROL  MEASURES  -  continued
               CONTROL  MEASURE NO.   7
             Ban All  Non-Emergency  Traffic
                                             EFFECTIVENESS
                                                                           INSTITUTIONAL OBSTACLES
                                                                                                             SOCIO-ECONOMIC IMPACTS
O1
             General  Description:

              Travel  by all  non-essential
              vehicles would be  banned.
              Transportation related  vehicles,
              such as buses  and  taxis would
              be exempted.   Exceptions would
              be reserved primarily for
              vehicles used  for  health and
              safety  -- fire and police pro-
              tection, medical services
              (doctor's cars,  ambulances), etc
Implementation  Procedure:
 1) Define categories of vehicles
    to be exempted.
 2) Establish review committee
    for additional vehicles to
    be included.
 3) Establish identification pro-
    gram for all  exempt vehicles
    (e.g. stickers).
 4) Set up communication channels
    to be used in the event of
    an impending  episode.
             Enforcement  Plan:
              Only  those  vehicles identified as
              exempt would  be  allowed to drive
              during the  alert.  Offenders
              would be  cited by  law enforce-
              ment  officers and  subject to
              prosecution and/or fines.
                                    Total VMT Reduction:

                                          50  - 85%, depending on
                                          implementation and enforce-
                                          ment procedures.
                                          Applicable to LDMV's, HDMV
                                          and diesels.
                                                  Emission Reduction:
                                                       (Percent of Total!
                                                    RHC
                                                  45  -
CO
                                                            50 -
NOX
                                                            35  -  60%
                    Implementation:  Moderate to
                     Significant.
                     Since all  cars are  excluded
                     unless exempted,  one would anti-
                     cipate many individuals applying
                     for exemptions on the  basis of
                     their "essential" travel.  Review
                     and distribution  of exemptions
                     will  be a  difficult task.  Mas-
                     sive media campaign for public
                     would also be  required.
                                             Direct Economic Cost: Very High.
                                             Since nearly all  businesses
                                             and governmental  operations
                                             would be significantly cur-
                                             tailed, economic  activity
                                             would be only a fraction of
                                             what it normally  is.   Costs
                                             could conceivably exceed $100
                                             million for a one day shutdown.
                    Administrative Cost:   Moderate.
                     Consists mainly of  program  to
                     identify key exempt  categories,
                     review requests for  exemptions
                     and publicize program to  public
                     so that contingency  planning can
                     be made.
                                                                                                                       Other Cost  Considerations:
                                              Since public and private
                                              businesses and governmental
                                              operations are affected, a
                                              high social cost for incon-
                                              venience and lost opportunity
                                              is  incurred.   Impact of fore-
                                              going goods and services
                                              normally offered is very
                                              significant.
                                                                       Enforcement Procedure:  Moderate.

                                                                        Depends heavily on  public  co-
                                                                        operation.   !f widespread  viola-
                                                                        tions were to occur, enforcement
                                                                        manpower would be  grossly  inade-
                                                                        quate to cite more  than  a  few
                                                                        violators.   Many short trips
                                                                        could be expected  to be  in vio-
                                                                        lation of the ban.
                                                         Costs  borne  relatively evenly
                                                         by  public  in  general.  No
                                                         obvious  economic advantage
                                                         given  to any  individuals or
                                                         groups.

-------
                Table  5.4    EVALUATION OF TRAFFIC ABATEMENT EPISODE CONTROL MEASURES  -  continued
  CONTROL MEASURC  NO.    8

 Mandatory Carpooling on Freeways
         EFFECTIVENESS
                                       INSTITUTIONAL OBSTACLES
                                                              SOCIO-ECONOMIC IMPACTS
General  Description:

 Freeway access and use  would  be
 restricted on the basis of the
 number of occupants in  the
 vehicle.   For example,  use could
 be limited  to vehicles with
 four or more occupants.
         Implementation Procedure:

          1)  Inform public of details of
~-*           the  plan  (i.e., number of
y,           passengers required, vehicles
             exempted, etc.)
          2)  Encourage the public to make
             carpooling arrangements prior
             to the episode.
          3)  Establish communication
             channels  to be used prior to
             alert.
          4)  Organize  surveillance system
             to monitor program effective-
             ness .
 Enforcement Plan:

 1) Establish  enforcement  plan --
    personnel  required,  assign-
    ments,  access  points to be
    monitored.
 2) Offenders  would  be  subject to
    citations  and/or fines.
Total  VMT Reduction:

      2 -  5% (two people per
      vehicle),
      4 -  10%  (three people per
      vehicle), depending on
      implementation and enforce-
      ment procedures.

      Applicable to LDMV's
                                      Emission  Reduction:
                                           (Percent of Total]
                                        RHC
   1  - 37o
(2 people)

2-1/2 - 6%
(3 people)
                CO
1-1/2  -  47(
(2 people)

3-1/2  -
  8-1/2%
(3 people)
              N0y
  1  -  3%
(2 people)

2-1/2  -  £%
(3 people
                      Implementation:  Moderate.
                        Fairly  extensive media campaign
                        would be  required to inform the
                        public  initially and also prior
                        to  an impending episode alert.
                                                Direct Economic Cost:   Minor

                                                Sufficient flexibility  exists
                                                in the program that  most people
                                                would not be  significantly
                                                impacted economically.
                                                                       Administrative Cost:   Minor  to
                                                                       Moderate.

                                                                       Dependent on the  extent of pub-
                                                                       licity necessary  to adequately
                                                                       inform the public of the program.
                                                                                                          Other Cost Considerations:
                                                                      Inconvenience of finding  car-
                                                                      poolers or alternative travel
                                                                      route will probably exceed
                                                                      economic cost incurred.
                                                                               Enforcement Procedure:   Moderate
                                                                                to Significant.

                                                                                Extensive  monitoring of approxi-
                                                                                mately  1500 entrance points to
                                                                                freeway system needed for effec-
                                                                                tive enforcement.  Widespread
                                                                                violations could not be stopped
                                                                                due to  limited enforcement man-
                                                                                power
                                                                       No obvious or implied discri-
                                                                       mination associated with this
                                                                       control measure.

-------
                         Table 5.4     EVALUATION  OF  TRAFFIC  ABATEMENT  EPISODE  CONTROL  MEASURES -  continued
               CONTROL MEASURE NO.  9
              Mandatory  Carpooling on Freeways
             -and Major  Thoroughfares
                                             EFFECTIVENESS
                                                                           INSTITUTIONAL OBSTACLES
                                                             SOCIO-ECONOMIC IMPACTS
cn
cr>
             General  Description:

              Travel  on  freeways and major
              thoroughfares would be banned
              on  the  basis of vehicle occup-
              ancy.   For example, use of main
              arteries and freeways could be
              reserved for vehicles with three
              or  more persons.
Implementation  Procedure:

 1)  Inform public  of details of
    the  plan  (i.e.  number of
    passengers  required, vehicles
    exempted, etc.)
 2)  Encourage public to make car-
    pooling arrangements prior
    to the episode.
 3)  Establish communication
    channels to be  used for
    impending alerts.
 4}  Organize surveillance system
    to monitor program effective-
    ness.
             Enforcement Plan:

             1)  Determine  personnel  required,
                assignments  and  access points
                to  be  monitored  in case of
                alert.
             2)  Offenders  to  be  subject to
                citations  and/or fines.
                                    Total VMT Reduction:

                                          7  -  15% (2 persons per
                                          vehicle),
                                          15 - 35% (3 persons per
                                          vehicle), depending on
                                          implementation and enforce-
                                          ment procedures.

                                          Applicable to LDMV's.
                                                 Emission Reduction:
                                                      (Percent of Total)
                                                   RHC
  4 -  9%
(2 people)

  9 -  22%
(3 people)
   CO
 6 -  13%
(2 people)

 13 -  3C%
(3 people)
                                                                         NOv
 4 - .9%
(2 people)

 9 - 22%
(3 people)
                      Implementation:  Moderate to
                        Significant.

                        Extensive media campaign required
                        to inform public of program,
                        their  options during an alert,
                        etc.   Publicity required to make
                        program  elements and objectives
                        know as  well as during an impend-
                        ing alert.
                                               Direct Economic Cost:   Minor to
                                                Moderate.

                                                Dependent  largely  on  public
                                                response to control.   Flexi-
                                                bility still  exists in the
                                                system to  accomplish  most trip
                                                purposes,  if sufficiently moti-
                                                vated.  Costs likely  to  be
                                                proportional  to trips foregone
                                                by public.
                       Administrative Cost:   Moderate to
                        High.

                        In addition  to costs for media
                        campaign considerable effort
                        would  be required  to determine
                        which  main  arteries are to be
                        included in  the traffic ban.
                                                                                                                      Other Cost Considerations:
                                                Social costs incurred will  be
                                                high due to many "loopholes"
                                                which exist in this measure.
                                                Nuisances created will encour-
                                                age violations; limited enforce-
                                                ment capability will  largely
                                                be wasted since it will be
                                                grossly inadequate for the
                                                task.
                                                                       Enforcement Procedure: Substantial

                                                                        Monitoring this program would
                                                                        require  tremendous manpower
                                                                        requirements,  not available
                                                                        from  normal  law enforcement
                                                                        agencies.   Violations would be
                                                                        widespread and either largely
                                                                        undetected or  ignored.
                                                           No obvious or implied discrimi-
                                                           nation associated with this
                                                           control measure.

-------
                         Table  5.4    EVALUATION  OF  TRAFFIC  ABATEMENT  EPISODE  CONTROL  MEASURES   -  continued
               CONTROL MEASURE NO.   10

              Commercial Abatement Schemes
                                             EFFECTIVENESS
                                                                           INSTITUTIONAL OBSTACLES
                                                               SOCIO-ECONOMIC IMPACTS
CTl
              General Description:

              A  traffic abatement plan would
              be required from business,
              industrial, and commercial forms
              to reduce their work related VMT
              by a specified amount.   These
              plans could then be implemented
              on a programmed basis depending
              on the severity of the  episode.
                                    Total  VMT Reduction:

                                          Flexible.

                                          Maximum: 35%, calling
                                          "Work Holidays"
                                          0-15%*, keeping business
                                          operational
                                          *depending on enforcement
                                          and implementation
                                          procedures.

                                          Applicable to LDMV's  and
                                          HDMV's.
Implementation  Procedure:

 1) Establish criteria  for firms
    to be  affected  by control
    measure.
 2) Require  traffic  abatement plan
    be submitted  for review and
    approval according  to estab-
    lished criteria.
 3) Set up review committee to
    approve/disapprove  plans.
 4) Assist companies in writing,
    plans  and solicit their active
    support  in  case  of  implementa-
    tion.
                                                   Emission Reduction:
                                                        (Percent of Total]
                                                     RHC
              Enforcement  Plan:

              1) Set up enforcement teams  to  be
                 sent out into the field for
                 spot checking individual  firms
                 for compliance.
              2) Companies found  to be in  vio-
                 lation would be  subject to
                 citations and/or fines.
                                                    0-10%
                                                 (business
                                                  kept oper-
                                                  ational )
    CO
  0 -  13%
(business
 kept  oper-
 ational )
 0-10%
 business
 kept oper-
ational)
   NOX
                       Implementation:   Significant.
                        A  large effort would be required
                        to define companies affected,
                        inform them of plan requirements,
                        assist them in writing abatement
                        plans, review plans, correct
                        deficiencies, etc.
                                                 Direct  Economic  Cost:  Minor to
                                                 Moderate.

                                                 Dependent on good faith response
                                                 of companies affected.  Also,
                                                 dependent on the severity of
                                                 control required and the con-
                                                 tingency plans developed by
                                                 the firms.
                                                                                    Administrative Cost:   High.
                                                                                                                       Other Cost Considerations:
                         In  addition  to publicizing the
                         program,  the actual reviews,
                         approvals/disapprovals would
                         require considerable effort.  In
                         all  likelihood, many companies
                         would  not  submit plans placing
                         additional strain on the enforce-
                         ment agency.
                                                 Could reduce the efficiency of
                                                 business operations involved;
                                                 resulting in some degree of
                                                 inconvenience.
                                                                       Enforcement Procedure:

                                                                        Given  all  plans were submitted
                                                                        and  approved  (a formidable enfor-
                                                                        cement task in itself), actual
                                                                        monitoring during an episode
                                                                        would  be  extremely difficult.
                                                                        Implementation is largely depen-
                                                                        dent on good  faith action of
                                                                        companies.  Violators would be
                                                                        difficult to  detect.
                                                            Singles  out work  related travel
                                                            in commercial  institutions of
                                                            a given  size,  thereby  impacting
                                                            large businesses  the most.

-------
                         Table  5.4    EVALUATION OF TRAFFIC ABATEMENT EPISODE  CONTROL  MEASURES  - continued
CONTROL MEASURE  NO.    11
Close Freeways  (or Certain Free-
                                                        EFFECTIVENESS
                                                                                      INSTITUTIONAL OBSTACLES
                                                                                                                        SOCIO-ECONOMIC  IMPACTS
            General  Description:

              Access to  freeways  (or certain
              freeway lanes) would be closed
              to all  but  emergency and trans-
              portation  related vehicles.
           Implementation Procedure:

              1) Develop media campaign to
	,              inform public of control
en              proposed.
0°            2) Determine exempt categories
                or vehicles not subject to
                ban.
              3) Establish communication'
                channels to be used for
                impending episode.
              4) Organize surveillance  and
                monitoring program  --  check
                points, personnel  required,
                etc.
           Enforcement Plan:

             1) Establish enforcement  teams
                to either prevent  entrance
                to freeways or patrol  free-
                ways during alert  for  viola-
                tions.
             2) Offenders to be subject  to
                citations and/or fines.
                                  Total VMT Reduction:

                                       Doubtful,  very possibly an
                                       increase.
                                  Emission Reduction:
                                       (Percent of Total)
                                    RHC
CO
N0y
                    Implementation:   Moderate.
                                             Direct Economic Cost:   Minor.
                     Fairly extensive  media campaign
                     required to  inform  the public
                     initially of the  program and also
                     prior to an  impending alert.
                                             Closing freeways leaves  exten-
                                             sive street system available
                                             for travel.  Most travel  may
                                             take longer or be less  direct
                                             but will still  take place on
                                             surface streets, implying a
                                             minor economic impact will
                                             result.
                                                                    Administrative  Cost:  Minor to
                                                                     Moderate.

                                                                     Dependent  on the extent of pub-
                                                                     licity necessary to adequately
                                                                     inform the public of the program.
                                                                     Also,  cost of  setting up surveil-
                                                                     lance, monitoring, and enforce-
                                                                     ment procedures.
                                                                                                                     Other Cost  Considerations:
                                                        High social  cost  from  nuisance
                                                        and inconvenience of having
                                                        to use surface  streets for
                                                        travel.   Increased congestion,
                                                        noise and accidents  likely to
                                                        take place.
                                                                    Enforcement Procedure:   Moderate
                                                                     to Significant.

                                                                     Extensive  monitoring of approxi-
                                                                     mately  1500  entrance points or
                                                                     of 700  some  miles of freeway.
                                                                     More difficult to enforce if
                                                                     certain lanes are to be designate
                                                        Subtle form of discrimination
                                                        to those more dependent on
                                                        freeways, e.g., commuters
                                                        who travel  long distances,
                                                        businesses  who make frequent,
                                                        long deliveries, etc.

-------
            Table 5.4     EVALUATION  OF  TRAFFIC  ABATEMENT  EPISODE  CONTROL  MEASURES -  continued
  CONTROL MEASURE  NO.    12

Limit Availability of Parking
         EFFECTIVENESS
              INSTITUTIONAL OBSTACLES
   SOCIO-ECONOMIC IMPACTS
General  Description:

  The supply of on and off street
  parking would be limited to  a
  previously determined level  as
  an incentive for carpooling  and
  the use of public transit.   For
  example, off-street parking
  facilities could only fill  to
  1/4 their capacity and all  on-
  street parking would be pro-
  hibited.
Implementation  Procedure:

  1) Establish which facilities
     (and streets) are to be
     affected and to what extent.
  2) Notify off-street parkirvg
     lots of the proposed plans.
  3) Encourage alternative modes
     of travel.
 Enforcement Plan:

  Law enforcement personnel  to
  spot check parking facilities
  to ensure compliance.   Offenders
  to be subject to citations and/
  or fines.
Total  VMT Reduction:

      Doubtful, very possibly
      an  increase
 Emission  Reduction:

      (Percent  of Total}
                                       RHC
               CO
N0y
         Implementation:   Significant.
          Thousands of off-street facili-
          ties exist and many more spaces
          exist on street.   Determining
          how and what facilities should
          be controlled and to what extent
          will be a formidable task.
Direct Economic  Cost:  Moderate.

 Parking  revenues  in the Los
 Angeles  CBD exceed $100,000
 daily.   Adding  other parking
 revenues  collected as well as
 economic  activity curtailed
 due  to lack of  parking, the
 costs become significant.
                                   Administrative Cost:  High.

                                    Widespread publicity of the pro-
                                    gram,  contacting affected facili-
                                    ties,  and enlisting active  sup-
                                    port of  this program will be
                                    most difficult.
                                                                                                          Other  Cost Considerations:
                                             Lack of parking will  obviously
                                             decrease the efficiency of
                                             business operations,  who must
                                             find alternative transportation,
                                             forego the trip, or look for
                                             other parking.
                                   Enforcement  Procedure: Significant

                                    Due  to the  large number of faci-
                                    lities involved, compliance is
                                    heavily dependent on good faith
                                    cooperation.  Large scale resist-
                                    ance  to this measure would make
                                    enforcement practically
                                    impossible.
                                             Parking  lot owners  singled
                                             out to bear large  cost;  also,
                                             subtle form of discrimination
                                             to businesses which  place
                                             heavy reliance on  good park-
                                             ing facilities for attracting
                                             customers,  i.e., shopping
                                             centers.

-------
             Table  5.4    EVALUATION  OF TRAFFIC ABATEMENT  EPISODE  CONTROL  MEASURES -  continued
  CONTROL MEASURC NO.   13
  Special  Parking  Tax
         EFFECTIVENESS
                                       INSTITUTIONAL OBSTACLES
                                                SOCIO-ECONOMIC  IMPACTS
Genera] Description:

  A special  parking tax on sur-
  charge would  be  levied on
  motorists  during the episode.
  The  tax would be high enough
  to encourage  use of alternate
  transportation or to carpool.
Implementation  Procedure:

 1) All on-street and metered
    parking would be banned
    during an alert episode.
 2) All public and private lot
    operators would be required
    to collect a predetermined
    tax for each vehicle entering
    the parking facilities.
    Revenues collected would  go
    to the government, much like
    a sales tax.
Enforcement Plan:

 1) Enforcement and surveillance
    teams would be sent out  to
    spot check collection  proceed-
    ings and note  revenues being
    collected.
 2) Other teams would ensure
    revenues were  properly turned
    over to the government.
Total  VMT  Reduction:

       Very uncertain.  Possibly
       as  much as 20% but
       possibly negligible.
Emission Reduction:
     (Percent of Total;
  RHC
                                     0 -
              CO
            0 - 17%
NO,
                                                            0 - 12%
         Implementation:  Significant to
           Substantial.
           Many problems are foreseen for
           each phase  of the implementa-
           tion --   notifying all parking
           lots of  the  program, ensuring
           collection  of parking tax,
           turning  over all collected
           revenues.
Direct  Economic  Cost: Moderate.
 Dependent  on  public response
 to incentive.   If a significant
 VMT reduction was achieved, the
 economic  impact  on businesses
 could be  high.
                                  Administrative Cost:   High.

                                   Many additional layers of paper-
                                   work required to properly execute
                                   this program.  Could potentially
                                   become a bureaucratic nightmare.
                                                                                                          Other Cost Considerations:
                                              Dependent  on  public response
                                              to incentive.   For large VMT
                                              reductions,  the social costs
                                              of trips foregone or  lost
                                              opportunities would be high.
                                              If many  chose to pay  the tax,
                                              the economic  cost would be
                                              more appreciable.
                                   Enforcement Procedure: Substantial

                                   Many ways to violate this pro-
                                   gram, most of which are undetect-
                                   able with the normal enforcement
                                   agencies.  Widespread corruption
                                   could be expected with any signi-
                                   ficant sums of money collected.
                                              Clearly discriminates  against
                                              the poor who  are  least able
                                              to afford additional auto
                                              operating costs.

-------
              Table  5.4    EVALUATION OF  TRAFFIC  ABATEMENT EPISODE  CONTROL MEASURES  -  continued
  CONTROL MEASURE NO.
  Gasoline Rationing
General Description:   Prior to
and during an emergency episode,
the purchase and sale of gasoline
would be limited.  A variety of
actual implementation procedures
are available.
          EFFECTIVENESS
                                        INSTITUTIONAL OBSTACLES
                                                                                                             SOCIO-ECONOMIC IMPACTS
Implementation  Procedure:

1)   Establish distribution system
    for rationing  coupons to be
    used to purchase  gasoline
    during an alert.
2)   Initiate mass  media  campaign
    to acquaint public to controls
    proposed.
3)   Establish communication
    channels to gasoline dealers
    in case of  impending episode.
Enforcement Plan:
1)  Organize  surveillance  teams
    to ensure compliance to  regu-
    lations  during  alerts.
2)  Enlist support  by major  dis-
    tributors to implement controls
3)  Violators to be subject  to
    citations and/or fines.
 Total VMT Reduction:
   Flexible
   Maximum:  10% 1st day
             20% 2nd day
             50% 3rd day
 Emission  Reduction:
        (Percent of Total)
   RHC
 Maximum

 11  Day 1

14%  Day 2

35%  Day 3
  Maximum

  %  Day 1

  %  Day 2

47%  Day 3
                CO
                NOX
Maximum

 7% Day 1

14% Day 2

35% Day 3
                         Implementation:  Substantial.
                         Large  effort would be required
                         to  set up  the  program, distribute
                         coupons, inform the public and
                         service stations, etc.
                                               Direct Economic  Cost:  Largely
                                               dependent on public response  to
                                               the measure.  A short rationing
                                               program cannot be expected  to
                                               curtail much travel.   A longer
                                               rationing program could severly
                                               affect some businesses.   In any
                                               case, gasoline sales  would  suffer.
                                    Administrative Cost:  High.
                                   An entire new program would  have
                                   to be organized,  with significant
                                   amounts of bookkeeping,  paperwork,
                                   etc.
                                                            Other Cost Considerations:  High
                                                            social  cost expected from the
                                                            creation of illegal activities,
                                                            e.g.  counterfeiting coupons, black-
                                                            markets, hoarding gasoline.
Enforcement Procedure: Substantial
Prior experience  during World War
II has shown  the  difficulty of
attempting to enforce such a
measure.   More than 6000 service
stations  exist in L.A. County pre-
senting a formidable enforcement
problem.
Equity:  Dependent on implemen-
tation procedures.  Rationing
schemes generally impact the
poor most since  they are not
as likely to  resort to black
market supplies, bootleggers,
etc.  Service station owners are
clearly discriminated against.

-------
                 Table  5.4    EVALUATION OF TRAFFIC ABATEMENT  EPISODE  CONTROL  MEASURES  -  continued
   CONTROL MEASURC  NO.   15
   Close Gasoline Stations
                                         EFFECTIVENESS
                                                                            INSTITUTIONAL OBSTACLES
                                                                                                             SOCIO-ECONOMIC  IMPACTS
 General  Description:   Retail
 gasoline service  outlets would
 be closed during  an  episode.
                                Total VMT Reduction:

                                  10% 1st  day
                                  20% 2nd  day
                                  50% 3rd  day

                                  Applicable to LDMV's and HDMV's.
 Implementation Procedure:

 1)

 2}
3)
Obtain complete listing of
affected stores.
Have media campaign  to  these
owners to instruct them of
their required response during
an alert.
Establish communication pro-
cedures  to be used in case of
an impending alert.
Emission  Reduction:
       (Percent  of Total)
                                      RHC
Enforcement Plan:

1)  Organize surveillance  and
    enforcement teams  to be
    used to monitor compliance
    to the control  regulations.
2)  Violators  to be subject  to
    citations  and/or fines.
                                    7% Day 1

                                   14% Day 2
                                   35% Day 3
                                                   CO
                                            9% Day 1

                                           19% Day 2
                                           47% Day 3
                                                         NOX
                       7% Day 1
                      14% Day 2
                      35% Day 3
                                  Implementation:  Moderate to
                                  Significant.  Major effort required
                                  would be publicity campaign to  in-
                                  form service station owners and
                                  public of the program.
Direct Economic Cost:  Moderate.
Cost accrue  primarily from lost
of business  at  service stations --
gasoline  sales,  repair services,
auto accessories, etc.  Presumably
much of this would be made up when
the service  stations reopened.
                                   Administrative Cost: Moderate.
                                   Involves primarily cost of in-
                                   forming all the stations involved
                                   and establishing proper communi-
                                   cation channels for execution of
                                   the plan.
                                                                                                          Other  Cost Considerations:  Loss
of service  station related
services  offered.  Could result
in additional  safety hazards.
                                                                   Enforcement Procedure:  Significan
                                                                   Due  to  the large number of
                                                                   stations  involved, compliance
                                                                   would be  largely dependent on
                                                                   good faith, cooperation of
                                                                   service stations.  It would
                                                                   be difficult to detect more
                                                                   than a  few violations.
                                                                      Equity:   Singles out service
                                                                      station  owners  for control ; cost
                                                                      borne almost  exclusively by this
                                                                      group.

-------
                         Table  5.4    EVALUATION  OF  TRAFFIC  ABATEMENT  EPISODE  CONTROL  MEASURES  -  continued
              CONTROL  MEASURC NO.  16
              Close (Non-Essential) Federal
                        Agencies	
                                             EFFECTIVENESS
                                                                            INSTITUTIONAL OBSTACLES
                                                                                                             SOCIO-ECONOMIC IMPACTS
co
            General  Description:  A general
            work  holiday would be in force for
            federal  employees performing non-
            essential  services (i.e. not
            emergency, health, or safety
            related).
                                    Total VMT Reduction:
                                      Applicable to LDMV's.
Implementation Procedure:

1)   A working  definition of
    agencies and  personnel to be
    affected would be drafted.
    Agencies would be informed
    of the  emergency actions pro-
    posed.
2)   Cooperation  in the program
    would be actively solicited
    and encouraged to ensure pro-
    gram effectiveness.
3)   Establish  communication chan-
    nels to be used  in case of
    alert.
                                               Emission Reduction:

                                                      (Percent  of Total)
                                                  RHC
            Enforcement  Plan:

            Surveillance  teams would be
            organized  to  spot check com-
            pliance  during  the episode.
                                                              CO
                                                                         NOX
                                                                      Implementation:  Moderate.
Primarily a problem  of  notifying
and coordinating the implemen-
tation by the affected  agencies.
Good communication and  cooperation
key to successful execution of
plan.
 Direct Economic Cost: Moderate.
Heavily dependent  on  invoked re-
sponse during alert.   Cost would
be government goods and  services
not offered  during the alert.
Administrative Cost:   Moderate.
Encouraging  and  assisting each
affected agency  to formulate
abatement plans.  Cost primarily
one of publicity and communication
both prior to  and during an
emergency episode.
                                                                                                                     Other Cost  Considerations:  Social
                                                                                                                     cost  of decreased efficiency in
                                                                                                                     governmental operations.   Incon-
                                                                                                                     venience cost for trips foregone
                                                                                                                     or wasted due to closed facilities.
                                                                       Enforcement Procedure:  Moderate.
                                                                       Dependent on cooperation  of
                                                                       affected agencies.   Widespread
                                                                       violations, intentional  or other-
                                                                       wise, would be difficult  to  pre-
                                                                       vent.
                                   Equity:    Singles out federal
                                   faci 1 ities  for work "holiday!1
                                   Hay invoke  outcries of free
                                   loafing  of  federal employees
                                   at taxpayer's expense.

-------
            Table  5.4    EVALUATION OF TRAFFIC ABATEMENT  EPISODE  CONTROL  MEASURES  - continued
   CONTROL MEASURE  NO.    17
Close (Non-Essential) Government
Agencies (Federal,  State. & Local!
          EFFECTIVENESS
                                        INSTITUTIONAL OBSTACLES
                                               SOCIO-ECONOMIC IMPACTS
 General  Description: A general
 work holiday  would  be in force
 for all  governmental employees
 performing  non-essential services
 (i.e.  not related to emergency,
 health,  and safety).
 Implementation Procedure:

 1)  A working definition of
    agencies and personnel to be
    affected would be drafted.
    Agencies would be informed
    of the emergency actions
    proposed.
 2}  Cooperation in the program
    would be actively solicited
    and encouraged to ensure
    program effectiveness.
 3)  Establish communication chan-
    nels to be used in case of an
    alert.
 Enforcement  Plan:
Surveillance teams would be
organized to spot check com-
pliance during the episode.
 Total  VMT Reduction:

   5%*
   Applicable  to  LDMV's.

   *Add 1%  if  public school
    closure  is to be included.
Emission Reduction:
       (Percent  of Total ]
   RHC
   3%
                CO
NOv
        Implementation:  Significant.
        primarily  a problem of notifying
         nd  coordinating the implemen-
        tation  of  the many affected agencie
        iOod cooperation and communication
         re  key to the successful execution
         f the  plans.
Direct Economic Cost:   High.
Heavily dependent  on  invoked
response of  agencies  affected
during the alert.  Cost would
be loss of governmental goods
and services  normally offered.
                                                                       Administrative Cost:  Moderate
                                                                                                          Other Cost Considerations:  Social
                                   to High.   Encouraging  and assist-
                                   ing the affected  agencies to
                                   formulate abatement  plans.  Cost
                                   primarily one  of  publicity and
                                   coninunication,  both  prior to and
                                   during an emergency  episode.
                                            cost associated with decreased
                                            efficiency  in  governmental oper-
                                            ations.   Inconvenience cost for
                                            trips foregone or wasted due to
                                            closing  of  federal  facilities.
                                    Enforcement Procedure: Moderate
                                   to Significant.   Dependent on
                                   cooperation of affected agencies.
                                   Widespread violations,  intentional
                                   or otherwise,  would  be difficult
                                   to prevent.  Problem of enforce-
                                   ment, in part due to the  number
                                   of agencies and personnel affectec
                                             Equity:   Singles  out  governmental
                                             agencies for a work  holiday.   May
                                             invoke outcries from  private bus-
                                             inesses of wasting public  funds,
                                             taxes, etc.

-------
                         Table  5.4    EVALUATION  OF TRAFFIC ABATEMENT EPISODE CONTROL MEASURES  -  continued
               CONTROL MEASURE  NO.    18
               Close Government  Contractors
                                             EFFECTIVENESS
     INSTITUTIONAL  OBSTACLES
                                                                                                                          SOCIO-ECONOMIC IMPACTS
en
             General  Description:

             A work holiday would be declared
             for employees of  governmental
             contractors  performing non-essent-
             ial  services  (i.e. emergency,
             health,  safety).
Implementation Procedure:

1)  A working definition would
    have to be formulated  for
    which companies would  be
    affected.
2)  Companies covered  in traffic
    ban requested  to draw  up con-
    tingency plans.
3)  Communication  channels estab-
    lished for alerting companies
    affected.
4)  Publicity campaign developed
    for companies  employees
    affected.
              Enforcement Plan:

             1)   Include  additional clause in
                 government  contracts that
                 abatement plans be developed.
             2)   Surveillance  teams would be
                 organized to  spot check com-
                 pliance  during  the alert
                 periods.
                                    Total VMT Reduction:
                                      Applicable to LDMV's.
                                                 Emission  Reduction:
                                                        (Percent of Total)
                                                   RHC
                                                               CO
                                                                           NOv
Implementation:    Minimal  to
Moderate.
Some precedence available;  similar
provisions  (i.e.  job  discrimi-
nation) have been included  in
contacts before.
 Direct  Economic Cost:    High.
Estimated that about one  million
employees would be affected.  Cost
would either be borne by  individ-
uals or the government and  include
salaries, loss  of productivity.
services curtailed,  etc.
Administrative Cost:  Moderate.
 Cost associated with modifying
 contracts  to  include additional
 clause(s).  Time and effort to
 review  abatement plans of govern-
 ment contractors and establish
 necessary  communication links for
 program implementation.
                                                                                                                      Other Cost Considerations:
 Decreased efficiency of govern-
 men-contracted facilities, loss
 of  goods and services normally
 given, etc.
                                                                                    Enforcement Procedure:   Moderate.
                                                                                                          Equi ty:
                                                                       Enforcement would  be  tied  into
                                                                       contracts where penalties  for
                                                                       violations would be clearly
                                                                       delineated.  Enforcement teams
                                                                       would randomly spot check  con-
                                                                       tractors for violations.
                                    Singles  out employees of govern-
                                    mental  contractors.  Complaints
                                    of wasted  tax dollars likely to
                                    occur

-------
                          Table  5.4    EVALUATION OF TRAFFIC ABATEMENT EPISODE CONTROL  MEASURES  -  continued
               CONTROL MEASURE NO.  19
            Close Shop. Facilities at Regional
            Shop. Ctrs. & Central Bus. District
                                             EFFECTIVENESS
                                                                           INSTITUTIONAL  OBSTACLES
                                                                                                             SOCIO-ECONOMIC  IMPACTS
cr>
             General  Description:

             Prior  to and during an emergency
             episode, the regional shopping
             centers  and central business dis-
             tricts would be closed in an
             attempt  to reduce travel.
Implementation Procedure:

1}  Determine which  facilities are
    to be included  in  shut down.
2)  Organize publicity campaign to
    inform public and  businesses
    of areas to be  affected.
3)  Establish communication
    channels for informing busi-
    nesses of impending shut down.
            Enforcement Plan:
            1)   Organize  surveillance and
                monitoring  teams  to spot
                check  closed  areas during
                alert.
            2)   Violators to  be subject to
                citations and/or  fines.
                                    Total VMT Reduction:
                                      Applicable to LDMV's  and HDMV's
                                                Emission Reduction:

                                                       (Percent of Total}
                                                  RHC
                                                              CO
                                                                          NOX
                                                                      Implementation:   Significant.
                                    Direct  Economic  Cost:  High.
Considerable  effort required for
all  phases  of this measure --
determining which areas to close,
organizaing publicity campaign,and
establishing  communication
channels.
laxable retail  sales  for the air
basin in 1971 were  about $60
million per day,  a  large snare
of which conies  from the areas to
be affected.
 Administrative Cost:  Moderate.
 Involves  primarily determing and
 contacting all the affected busi-
 nesses  of concern; also, providing
 for  an  effective surveillance pro-
 gram.
                                                                                                                      Other Cost  Considerations:
High social  cost  of  attemping
to close an  indirect source while
leaving many alternative  option
open, e.g. neighborhood and local
shopping centers.  Could  create
traffic jams and  safety problems
at local stores.
                                                                       Enforcement  Procedure:  Moderate
                                                                       to Significant.
                                                                       Dependent on favorable  cooperation
                                                                       of many businesses  affected.   It
                                                                       would be difficult  to detect
                                                                       significant numbers of  violations
                                     Singles out indirect source, I.e.
                                     businesses in shopping centers
                                     and CBD's, for control, which
                                     will  likely force a considerable
                                     activity to other alternatives --
                                     neighborhood and local stores.

-------
           Table 5.4     EVALUATION  OF  TRAFFIC  ABATEMENT  EPISODE  CONTROL  MEASURES -  continued
  CONTROL MEASURC  NO.   20
  Close Recreational  Facilities
         EFFECTIVENESS
                                       INSTITUTIONAL  OBSTACLES
                                                SOCIO-ECONOMIC  IMPACTS
General  Description:

Prior to and during  an  episode,
recreational facilities would be
closed to discourage vehicle
travel.   This measure would be
used primarily in  conjunction
with other types of  work  holidays
to minimize substitute  travel.
Total  VMT Reduction:

  7%%

  Applicable to LDMV's.
Implementation  Procedure:

1)  Arrive at working  definition
    of facilities  to be  covered
    1n shut down.
2)  Notify all  affected  facilities
    of shut down procedures.
3)  Establish communication  chan-
    nels for contacting  facilities
4)  Ban all parking within set
    distance of beaches,  parks,
    etc.
Emission Reduction:
       (Percent  of Total)
   RHC
 Enforcement Plan:

1)  Organize  surveillance and en-
    forcement teams  to  monitor
    program effectiveness.
2}  Violators  to be  subject to
    citations  and/or fines.
               CO
NOV
         Implementation:   Moderate  to
         ignificant.

        Some problem at defining  facilities
         :o be covered,  as  well  as  trying to
         3revent access  to  or parking at
         >ublic facilities.
Direct Economic Cost:  Moderate.
Primarily the loss  of  income to
owners  of recreational  facilities.
                                                                      Administrative Cost:   Moderate.
                                                                                                         Other Cost Considerations:
                                  Significant effort required  to
                                  inform public and affected busi-
                                  nesses of episode plan.
                                            Social  cost of not  having re-
                                            creational  facilities available
                                            as an outlet for  extra time
                                            available;  important for younger
                                            age groups  who  may  resort to
                                            less constructive activity --
                                            vandalism,  pranks,  crime.
                                   Enforcement Procedure:  Moderate
                                   to  Significant.
                                   Due to  the large number of parks,
                                   beaches, and private recreational
                                   facilities, it would be difficult
                                   to  prevent violations in any
                                   significant numbers.  Parking bans
                                   especially would be difficult to
                                   enforce.
                                            Singles out recreational  facilities
                                            to bear bulk of economic  cost of
                                            this control measure.

-------
            Table  5.4   EVALUATION  OF  TRAFFIC  ABATEMENT EPISODE CONTROL MEASURES  - continued
   CONTROL  MEASURE NO.  21
 Work Holiday  for Major Firms
         EFFECTIVENESS
                                       INSTITUTIONAL OBSTACLES
                                                                        SOCIO-ECONOMIC IMPACTS
 General  Description:

 All  firms which employ more than
 a  specified number of workers (100)
 would be closed to reduce work-
 related  travel.
Total  VMT  Reduction:
  11%

  Applicable to LDMV's and HDMV's
           Implementation  Procedure:

           1)  Arrive at working  list
^J             of firms  covered 1n  this
Co             exclusion.
           2)  Notify firms  of abatement plan
               and assist  in firms  developing
               individual  contingency action.
           3)  Establish communication chan-
               nels  to be  used in case of an
               alert.
                                    Emission Reduction:
                                           (Percent of Total)
                                      RHC
Enforcement Plan:

1)  Organize surveillance  and en-
    forcement teams  to  spot check
    and monitor program effective-
    ness.
2)  Violators  to be  subject to
    citations  and/or fines.
              CO
                                                  10%
                                                                       NOV
                                  Implementation:  Significant.
                                                                                 Primarily a problem of enlisting
                                                                                 cooperation of all  firms,  publici-
                                                                                 zing the program, and coordinating
                                                                                 the shut down.  Important  elements
                                                                                 will be good communication  and
                                                                                 advance publicity.
Direct Economic  Cost:   High.

In addition  to wage and salaries
the productivity of approximately
one million  employees would be
affected.
                                                                       Administrative Cost:   High.

                                                                       Cost of notifying the companies
                                                                       involved will be formidable,  both
                                                                       initially in regard to the  program
                                                                       and prior to an impending alert.
                                                                     Other Cost Considerations:
                                                                     All  the  goods and services offered
                                                                     by major firms would be curtailed;
                                                                     as a secondary impact, all firms
                                                                     dependent on major firms would
                                                                     likewise be affected.
                                  Enforcement Procedure:  Substantia
                                  Due to the  large number of firms
                                  involved, effectiveness is depen-
                                  dent on good  faith cooperation of
                                  companies involved.  Widespread
                                  violations  would be difficult to
                                  detect or prevent.
                                                                                                                    Singles out major firms  to bear
                                                                                                                    bulk of the direct cost  of this
                                                                                                                    control.   Depending  on circum-
                                                                                                                    stances,  employees,  company, or
                                                                                                                    both will  share  cost.

-------
                         Table  5.4   EVALUATION  OF TRAFFIC ABATEMENT EPISODE  CONTROL MEASURES  -  continued
             CONTROL MEASURE NO.  22
              Rotating Work  Holiday For
               .     Major Firms
                                             EFFECTIVENESS
                                                                           INSTITUTIONAL  OBSTACLES
                                                                                                             SOCIO-ECONOMIC IMPACTS
ID
            General Description:
            All  firms which employ more than
            a  specified number of employees
            (100) would be closed in alert
            situations on a rotating schedule
            to reduce work-related travel.
Implementation Procedure:
1)  Arrive at working  list
    of firms  covered in  this
    exclusion.
2)  Notify firms  of abatement
    plans and assist in  firms
    developing individual  con-
    tingency  action.
3)  Establish communication
    channels  to be  used  1n  case
    of an alert.
 Enforcement  Plan:

1)  Organize  surveillance  and en-
    forcement teams  to  spot  check
    and monitor program effective-
    ness.
2)  Violtaors to be  subject  to
    citations and/or fines.
                                    Total  VMT  Reduction:

                                    2.2% - 1/5 of Major Firms
                                    4.4% - 2/5
                                    6.6% - 3/5

                                    8.8% - 4/5

                                    11%  - All


                                    Applicable to LDMV's and HDMV's.
                                               Emission Reduction:

                                                      (Percent of Total)
                                                 RHC
    Fract.
      of
    Firms

1W  1/5

3°i  2/5

4%%  3/5
6%   4/5
7'i%  Al 1
                                                             CO
                                                          2%

                                                          4%

                                                          6%

                                                          8%

                                                         10%
                                                   Fract
                                                     of
                                                   Firms

                                                    1/5

                                                    2/5

                                                    3/5

                                                    4/5
                                                    All
3%
Fract.
  of
Firms

 V5
 2/5

 3/5

 4/5

 All
             Implementation:  Significant.

            Primarily a problem of enlisting
            cooperation of all  firms,  publici-
            zing the program, and coordinating
            the shut down.  Important  elements
            will be good communication and
            advance publicity.
                                            Direct  Economic  Cost:  High.
                                            In  proportion to the number of
                                            firms involved.  Consist mainly
                                            of  wages, salaries, and lost
                                            productivity of persons not
                                            working.
             Administrative Cost:   High.

             Cost of notifying  the  companies
             involved will  be formidable, both
             initially in  regard to the program
             objectives and prior to an im-
             pending alert.  Added  cost of
             explaining rotation schedule to
             avoid excessive shut-down or non-
             compliance.
                                                                                                                     Other  Cost  Considerations:
                                           Decreases efficiency of normal
                                           business activities  of region.
                                           Since many firms  are closed while
                                           others are open,  many transaction^
                                           which normally take  place will be
                                           disrupted.
Enforcement  Procedure:Significant.

Large number of firms  still in-
volved, making monitoring a
difficult task with  limited
personnel.   Widespread  vio-
lations still  would  be  difficult
to detect or prevent.
Equity:

Still singles  out  major busi-
nesses to bear major  cost of
control  measure.

-------
                         Table  5.4    EVALUATION OF  TRAFFIC ABATEMENT EPISODE  CONTROL  MEASURES  - continued
00
o
                 CONTROL MEASURE NO.  23
                 Operate  on Sunday  Status
               General Description:

               A work holiday would be declared
               in which  all  persons would assume
               their normal  Sunday schedules.
               The intent  1s to reduce work
               related travel for the majority
               of business,  industrial, commer-
               cial,  and governmental employees.
Implementation  Procedure:

1)  Notify public at large  of
    proposed program,  especially
    encouraging no substitute
    travel during the  alert;
2)  Establish  an effective,  com-
    munications channel  with
    normal mass media  avenues
    for broadcasting impending
    episode.
              Enforcement Plan:

              1}   Organize  surveillance teams
                  to  spot check and monitor
                  compliance  to control regu-
                  lations.
              2)   Offenders would be subject to
                  citations and/or fines.
                                             EFFECTIVENESS
                                    Total VMT Reduction:

                                       35-40%

                                       Applicable to LDMV's and
                                       HDMV's
                                                  Emission Reduction:
                                                         (Percent of Total!
                                                   RHC
                                                  25-302
                                                                CO
                                                              33-38%
                                                                           NO,
                                                            25-30%
                                                                           INSTITUTIONAL OBSTACLES
 Implementation:    Moderate to
Significant.

Almost completely depender*    an
effective  mass media campa.yn and
favorable  community response.
 Administrative  Cost:  Minor to

 Moderate.

 Primarily  the cost of communi-
 cating  the program to the public;
 presumably much of the radio and
 TV  time  could be gotten free as a
 community  public service.
                                                                                                             SOCIO-ECONOMIC  IMPACTS
Direct Economic Cost:   Very  High.
Proportional  to firms  affected,
and includes  lost business,
salaries, wages,  and lowered
productivity.   Probably an in-
crease in some businesses, e.g.
recreational  and  shopping
facilities.   Total  lost wages
may be up to  $50  mi VI ion.
                                                                                                                        Other Cost Considerations:
                                                                                                                       Disruption cost of an unanticipated
                                                                                                                       work holiday for many businesses  in
                                                                                                                       community.
                                                                      Enforcement Procedure:  Significant
                                                                      to Substantial.
                                                                      Dependent largely  on  community
                                                                      cooperation.   Many businesses, com
                                                                      panies, etc.  would make monitoring
                                                                      difficult.   Violations  would be
                                                                      difficult to  detect and prevent.
                                    Equity:
                                    Singles  out businesses, companies
                                    and agencies which operate on
                                    Monday-Friday  (or Saturday)
                                    schedule.  Provides a  "bonus"
                                    to facilities, e.g. recreational,
                                    shopping, which  are normally
                                    open on  Sundays.

-------
                        Table  5.4    EVALUATION OF  TRAFFIC  ABATEMENT  EPISODE CONTROL MEASURES  -  continued
              CONTROL MEASURE NO.   24
              Complete  Work  Holiday With
              .	 Recreation Closed
                                             EFFECTIVENESS
                                                                           INSTITUTIONAL OBSTACLES
                                        SOCIO-ECONOMIC IMPACTS
00
            General Description:

            A complete  shut down of all normal
            businesses  and institutions, un-
            less  they are of an emergency or
            essential nature,  i.e. hospitals,
            fire,  police, etc.  An attempt to
            induce people to stay home because
            of no  work,  snooping, or recreate
            ion available.
Implementation Procedure:

1)  Notify public at  large of pro-
    posed control measure.
2)  Actively solicit  public  sup-
    port for program.
3)  Establish communication  chan-
    nels to be used for  impending
    emergency.
            Enforcement Plan:

            1)   Organize  surveillance teams,
                to monitor  program effective-
                ness.
            2)   Violators to be subject to
                citations and/or  fines.
                                    Total  VMT  Reduction:

                                     75%

                                     Applicable to LDMV's fcHDMV's,
                                                Emission  Reduction:

                                                       (Percent of Total)
                                                 RHC
                                                 55%
                                                              CO
                                                              70%
                                                             NOX
                                                                        55%
 Implementation:   Substantial.
Dependent largely  on  favorable
public response.   Adequate
aublicity and  communication are
critical  to  successful execution
of this measure.
•Direct  Economic Cost: Very High.
This measure would virtually
bring the economic activity of
the community to a halt.   Costs
would run 50-100
millions of dollars in  loss  wages,
business, productivity,  services,
etc.
 Administrative  Cost:  High.

 Substantial effort would be re-
 quired  to  inform everyone of this
 program, and  to convince them
 that the requested actions on
 their part are needed for effect!v(
 control of the pollution.
Other Cost Considerations:

Indirect socio-economic impacts
are staggering.  Would result
in almost a  complete disruption
of all  normal  community activities.
                                                                       Enforcement Procedure:  Substantial

                                                                       Completely dependent  on favorable
                                                                       community response.   Violations
                                                                       would be nearly impossible to
                                                                       stop and could be widely antici-
                                                                       pated.
                                    Equity:

                                    Activities of groups such as
                                    young  and elderly who do not
                                    contribute to the pollution might
                                    be  curtailed.  Loss wages felt
                                    harder by the poor.  Ultimately,
                                    however,  all in the conmunity will
                                    bear the  costs.

-------
     Of the three general groups of measures, Controls on Traffic  Support
 Facilities, appear to be the least desirable type of control.  "Controls on
 Parking" and  "Closing Freeways" rate poor in effectiveness; these controls
 may actually  lead to an increase in traffic emissions.  They also rank poorly
 as to  institutional  obstacles and socio-economic impact considerations.
 "Rationing Gasoline" can be somewhat effective, but it is extremely poor
 in the realm  of public acceptability and implementation and enforcement
 problems.  "Close Gas Stations", the most attractive of this group of controls,
 is nearly equivalent to gas rationing.   It involves fewer enforcement-
 implementation problems than gas rationing, but results in slightly more
 economic disruption.   However,  it too ranks extremely poor as far as social
 and political  acceptance.   Overall, it  seems that more attractive measures
 are available among  direct controls and incentive controls than under support
 facility controls; the latter group will  thus be eliminated from the present
 discussion.
     There are ten different control  measures listed under Direct Controls
 on Traffic.  The first of these, "Voluntary Abatement", essentially has no
 effect.  The  last control  "Commercial Abatement Schemes", unless  they involve
 massive shutdowns (in which case they come under "Incentive Controls"), are
 also of very  limited  effectiveness  due  to the lack of enforcement for car-
 pooling and public transit aspects  of these plans.   Review of the implementa-
 tion of federal  agency abatement plans  on July 26,  1973 reveals that negligible
 increases  in carpooling  and  public  transit ridership occurred among the
agency employees.  A  new scheme  would have to be developed for enforcing
commercial  abatement  plans if this  were to be considered  a viable alternative.
     "Banning  All  Non-Emergency  Traffic"  is rated as a less desirable alter-
native than a  similar measure which controls incentives to travel.   The most
                                    182

-------
severe control on travel incentives -- work holiday plus closing of shopping

and recreational facilities -- can yield VMT reductions which are equivalent

to a traffic ban.  Yet, the travel incentive control  has fewer enforcement-

implementation problems and is more palatable socially and politically than

a complete traffic ban.  Both cause severe economic disruption.   Other factors

being equal, the indirect approach, i.e. disincentives to driving, is con-

sidered more acceptable than direct regulation, i.e.  traffic bans.*

     "Banning Fleet Vehicles" is a somewhat acceptable alternative.  Although

this measure also presents certain implementation-enforcement problems and

causes some economic disruption, it is effective and  ranks relatively high

in public/private acceptability.  Per car controlled, this method has

considerable impact since fleet vehicles are generally driven more miles

per day than non-fleet vehicles.  The main problems associated with this

measure are compiling a list of exempt fleets, e.g.,  emergency vehicles,

and setting up an implementation and enforcement program.  These problems

appear to be workable and thus, this control will  be  part of the recommended

program.

     "Banning Fleet Vehicles and Vehicles for Commercial Use" is more effec-

tive than just "Banning Fleet Vehicles"; however,  it  presents much greater

institutional problems and has a greater economic  cost.  Because of the very

significant implementation and enforcement problems in attempting to single

out all (non-essential) commercial vehicles, it appears that the extra effec-

tiveness of banning all commercial vehicles is not worth the extra difficulties

Only a fleet vehicle ban will be recommended here.
*  Actually, the one advantage of a traffic ban is that it might be put
   into effect sooner.  However, this discussion assumes a predictive
   type alert system is available.


                                     183

-------
     "License-Plate Lottery", "Mandatory Carpooling", and "Windshield Sticker

Program", are the direct controls remaining to be discussed.   These controls

are similar in that only vehicles satisfying certain conditions (e.g. having

certain last license numbers, having certain numbers of passengers, or having

certain colored stickers) will  be allowed to operate during an episode.   These

controls present obviously difficult implementation and enforcement problems,

and all three rank very low in  public acceptability.  Each of the controls

could be effective if adequately enforced, and they all minimize economic

disruption by allowing businesses to remain open.   Of the three, the "Wind-

shield Sticker Program" is apparently the best.   Several  states, for example,

have instituted windshield sticker programs in connection with annual safety

inspections.  It is also likely to be less confusing than the "Lottery";  as

such, it should be more enforceable and effective.  "Windshield Stickers"  also
                      \
appear to be better than the two "Mandatory Carpooling" measures.   The latter

are less organized and can create an extremely confused situation.   Although

"Windshield Stickers" is apparently better than  either a  "Lottery"  or "Mandatory

Carpooling," the probability of difficult enforcement problems must be empha-

sized.   Even more, the outcry of many members of the public to not  being  able

to drive because they have a certain type of car should not be underestimated.


     The third general type of traffic abatement strategy, Controls on

Incentives to Travel. solves certain problems but creates one major problem.

Incentive controls involve shutting down reasons for traveling.  A plausible

order might be:   government offices, then government contractors, then

specific industries, shopping areas, and recreational facilities, and finally

a complete work holiday.  This  provides a program of flexible impact, ranging

from a  few percent VMT reduction to more than 70% VMT reduction for a complete

work holiday.   The implementation and enforcement problems, while still


                                     184

-------
considerable, are somewhat less than those associated with most other schemes



discussed.  For example, they are much less than corresponding problems for a



"Windshield Sticker Program."  Overall social acceptance to this type of con-



trol probably compares favorably to other schemes with considerable VMT



impact.  However, the problem introduced by incentive control  plans is major



economic disruption.  It is estimated that a complete work holiday for one



day would cost around $50 million.  The economic disruption per VMT reduction



gained is the principal  drawback to the "Shut-Down" approach.




5.2.1  Summary of Traffic Abatement Control Measures




     In summary, an examination of the alternatives available  for traffic



abatement reveals a number of problems.  Nearly all the control measures



would encounter considerable implementation and enforcement difficulties.



Most measures under consideration also result in a certain amount of economic



and/or social disruption.  The most formidable institutional  obstacles will



be dealing with private interests affected by different measures and the



political forces which exist at all levels of government -- local, regional,



state, and federal.




     Recognizing that any plan which is to be effective at controlling an



episode will have these disruptive elements in it, one can only hope to



examine the trade-offs in an attempt to minimize the negative  impact of a



control plan.  With these considerations in mind, certain traffic abatement



controls are recommended below for inclusion in the staged implementation



plan for emergency alert situations.  These controls are basically of the



"incentive to drive" type.  It is emphasized that the recommendations basically



come from a subjective value judgment weighting economic disruption,



(primarily from incentive controls), with problems of enforcement,
                                    185

-------
implementation, social, and institutional  acceptability.   Obviously, in an

area of such controversy,  it is likely that different individuals (or groups)

will evaluate the situation differently.


     Finally, it should be noted that this judgment is  a  function of the

frequency with which alerts are called.   If advanced alert stages are called

often -- five or ten times per year -- the economic disruption associated

with "incentive controls"  may be intolerable.   In  this  case it might be

preferable to try to implement another type of control  program, e.g.

"Windshield Stickers".   If complete alerts are called very infrequently, a

complex "Windshield Sticker" program looks less attractive and incentive

controls become more acceptable.   Thus,  if the .40 PPM  -  4 hour "harm level"

were to be modified, so that alerts were  called less frequently, incentive

type controls would become even more attractive.


5.2.2  Recommended Control  Measures for  Inclusion  in Control  Plan


     In an attempt to identify several  control  measures with  increasing

effectiveness in reducing  total  VMT (and  thus  RHC  emissions),  four measures

are recommended for inclusion in a staged  implementation  control plan.

Listed in order of increasing emission control  effectiveness  (and increasing

socio-economic impact), the control measures are:


     e  Ban (Private) Fleet Vehicles - (Control Measure No. 5).

     e  Close (Non-Essential) Governmental Agencies (Federal,
        State, and Local)  - (Control Measure No.  17).

     •  Operate on Sunday  Status - (Control  Measure No. 23).

     •  Complete Work Holiday with Recreation  Closed -  (Control
        Measure No.  24).


     The selection of these four controls  is in part dictated by attempts

to minimize administrative costs and procedures.   Thus, groups have been


                                    186

-------
selected for controls which are readily identifiable, and  few new operational
definitions have been added.  This avoids extensive public knowledge-publicity
campaigns.
     The California Department of Motor Vehicles (DMV) has extensive
information on private fleet vehicles  (i.e. ten or more vehicles registered
under one name).   It would be relatively straightforward to get a complete
listing of the fleets, their location, number of cars in each fleet, etc.
This information could then serve as a basis for notifying the affected
groups of the control measure requirements.  Due to the disproportionate
share of mileage which fleets account  for, prohibiting their travel  during
an episode should  have a perceptible impact on the problem.
     Closing governmental agencies is  another means of eliminating a
significant fraction of VMT.  While this measure invariably invokes  out-
cries of wasted taxpayer money, it presents far fewer problems than  attempt-
ing to close down  private enterprises.  It also circumvents some of  the
complaints voiced  during the July 26,  1973 shutdown of federal  facilities  --
i.e. giving a work holiday to federal  employees only and not all governmental
employees.  Lastly, this measure is attractive because it affects a  sizable
group of easily definable commuters in agencies where well established channels
of communication already exist and a spirit of cooperation can generally be
enlisted.
     The control  measure calling for Sunday status has several  desirable
features -- it is operationally simplistic and inherently incorporates large
portions of other,  more cumbersome control measures.  No new procedures or
working definitions have to be developed.   With adequate publicity,  few indivi-
duals would not know what was expected of them.  Without notifying many
                                     187

-------
agencies, schools, fleets,  etc.,  operating on Sunday status would automatically
curtail  almost all of their activities because of normal  work schedules.   '
     The severest control measure proposed,  a complete work holiday with
recreational  facilities  closed,  is  used as a last resort  primarily for its
technical effectiveness.  As previously noted,  very significant socio-
economic disruptions  are  associated with  this control.  However,  it also
appears  that  no other alternative will be able to sufficiently reduce RHC
emissions to  the levels  required  to protect  public  health  from significant
harm levels when particularly high  concentrations of  photochemical  oxidants
are predicted to occur.
                                   188

-------
                         REFERENCES -  CHAPTER  FIVE
1.   California State Office of Emergency  Services, Air  Pollution Emergency
    Traffic Control  Planning Committee, Traffic Abatement Plan for Air
    Pollution Episodes, 1973.

2.   Environmental  Protection Agency,  Region  IX, Preliminary Report of the
    EPA Region IX  Task Force on Traffic Abatemen~During Air Episodes,
    15 September 1972.

3.   Verleger, P. K., Counsel for Western  Oil  and  Gas Association, Statement
    before the California Air Resources Board, Los Angeles, 18 September 1973
                                    189

-------
                 6.0  A PROPOSED EPISODE CONTINGENCY PLAN

     Alternative formulations of the four basic elements of an episode
contingency plan for the Metropolitan Los Angeles AQCR have been described
and evaluated in Parts 4. and 5. of this report.  On the basis of this
discussion, a specific plan is recommended below.  Section 6.1 deals
with episode criteria.  Section 6.2 proposes a four stage control pro-
gram and describes the effectiveness, implementation-enforcement
procedures, institutional problems, and socio-economic impact associated
with the control strategies.  Recommended surveillance and operational
procedures are presented in Sections 6.3 and 6.4, respectively.
Finally, to put the uncertainess in the results into perspective,
Section 6.5 discusses the limitations of the analysis.

     Overall,  the episode criteria  and  control  strategies recommended
here are similar to those in a contingency  plan recently developed by
EPA- Region IX.   The EPA plan  is presented  in Appendix  B.  However,
certain differences do exist between the two plans,  particularly in
the intermediate control  strategy stage.  Part 2 of  Appendix B compares
the two plans  and discusses  the relative advantages  and disadvantages
of each.
                                   190

-------
6.1  RECOMMENDED EPISODE CRITERIA



     The need to use forecasting as opposed to feedback episode criteria



for a Los Angeles contingency plan was stressed in Section 4.1.2.   For



one, advanced notice is required to effectively implement traffic  abate-



ment measures.  Secondly, for controlling oxidant, the pollutant likely



to cause nearly all episodes, forecasting appears to be an absolute



necessity because oxidant levels can continue to rise during the day



even if emission sources are instantaneously shut down.



     Section 4.1.2 distinguished between two types of air quality  fore-



casts:  future-weather type forecasts for more than a day in advance and



shorter term, weather-change type forecasts.  For the Los Angeles  episode



plan, it is recommended that control strategy selection be completed



at the latest by 4:00 or 5:00 P.M. of the previous day so as to make the



evening news.  A 4:00 or 5:00 P.M. deadline would allow the forecast to



be made in the late afternoon of the previous day.  Since previous day



maximal oxidant is likely to be known at this time, a weather-change type



prediction technique would be possible.  The episode criteria recommended



below will be of the weather-change, forecasting type.  However, it should



be noted that to allow for more implementation time, it may prove  desirable



to use an earlier, (late morning),  future-weather  type  forecast.  The



trade-offs of operating time vs. forecast accuracy deserve further



investigation.





      Table 6.1 and 6.2 list the proposed episode criteria.  Table 6.1 is



 for oxidant.  Criteria are given for both the present  .40 PPM - 4 hour



 average harm level, (the binding constraint of the three existing harm



 levels; see Section 2.2.5), as well as for the recently proposed



 .60 PPM - 1 hour average substitute.  Table 6.2 gives criteria for CO




                                    191

-------
                                 Table 6.1

                       EPISODE CRITERIA FOR OXIDANT
 a)  For the Present .40 PPM (4-Hour Average) Harm Level
Previous Day
Max. Four Hour
Average Oxidant
(with no
control*)
(PPM)
.00 - .27
.28 - .33
.34 - .38
.39 - .44
.45 - .50
.51 +
Alert Stage to be Called
Forecasted Change in Max. 4-Hour Average Oxidant
Much
Lower
-
-
-
-
I
I
Lower
-
-
-
I
II
II
Same
-
-
I
II
III
IV
Higher
-
I
II
III
IV
IV
Much
Higher
I
II
III
IV
IV
IV
 b)  For the Proposed .60 PPM (1-Hour Average)  Harm Level
Previous Day
Max. One Hour
Average Oxidant
(with no
control*)
(PPM)
.00 - .39
.40 - .49
.50 - .57
.58 - .65
.66 - .75
.76+
Alert Stage to be Called
Forecasted Change in Max. 1-Hour Average Oxidant
Much
Lower
-
-
-
-
I
I
Lower
-
-
-
I
II
II
^
Same
-
-
..o
r^
fur)
;' iv ;
Higher
-
I
II
III
IV
IV
Much
Higher
I
II
III
IV
IV
IV
*  If controls were  in  effect  during  the  previous  day,  the  previous  day
maximal  oxidant level  should be  adjusted  upward  to reflect  a  "no  control"
value before using  these  tables.
                                    192

-------
                                 Table 6.2

                      EPISODE CRITERIA FOR CO AND N02

For the 50 PPM (8-Hour Average) CO and .50 PPM (24 Hour Average) N02 Harm
                                 Levels
Previous Day
Max. 8-Hour
Average CO
(with no
control*)
0 - 39
40 - 46
47 - 54
55 - 66
67+
Alert Stage to be Called
Forecasted Change
Previous Day
Max. 24-Hour
Average NO^
.00 - .39
.40 - .46
.47 - .53
.54 - .61
.62+
Much
Lower
-
-
-
I
I
Lower
-
-
I
II
II
Same
-
I
II
III
IV
Higher
-
II
III
IV
IV
Higher
I
III
IV
IV
IV
*  If controls were in effect during the previous  day,  the  previous  day
maximal  pollution levels should be adjusted upward to  reflect  a  "no  control1
value before using this table.
                                   193

-------
and NO .   These are based on the 8 hour CO harm level  and the 24 hour NO^

harm level which appear to be binding over the other CO and N02 significant

harm standards, (Section 2.2.5).


     The episode criteria call  for control  stages I through IV to be

implemented under various conditions.  These control  strategies will be

presented in the next section.   The effectiveness of each control  stage

in abating episodes, (given later in Table 6.5),  is consistent with the

episode criteria as they appear in Tables  6.1  and 6.2


     It should be emphasized that the recommended episode criteria ar°

intended to be general  guides for control  strategy selection rather than

rigid regulations.   The director of the episode program should be  allowed

some flexibility.   This is particularly true in regard to the "Much Lower"

and "Much Higher"  predictions.   For instance,  in  Table 6.la, a previous day

4 hour oxidant level of .50 PPM along with  a forecasted change of  "Much

Lower" still calls  for  a Stage  I alert (voluntary abatement).   However, if

the "Much Lower" forecast is due.to the fact that a rainstorm is descending

on the basin, it would  not be appropriate  to call  any  alert at all. Con-

versely,  different  actions might be taken  with a  "Much Higher" forecast

depending on just  "how  much higher" is really  expected.  However,  since it
                                                       t
may prove administratively difficult to make a firm decision to call an

episode,  flexibility should be  allowed more in calling for over control

than in calling for under control.


6.2  RECOMMENDED CONTROL STRATEGIES


     On the basis of the discussions thus  far, this section recommends a

control plan which  has  the required technical  effectiveness to prevent sig-

nificant  harm levels with minimal  implementation  problems and socio-economic



                                     194

-------
disruptions.  A four stage alert plan is proposed with control ranging from



essentially zero emission reductions (Stage I) to approximately 64% reduction



in RHC emissions (Stage IV).  Two intermediate stages (Stages II and III)



would reduce RHC emissions by about 16% and 39% respectively.




     In many ways, the strategies recommended here are similar to those in



a contingency plan developed by EPA-Region IX; in other respects, distinct



differences exist, (see Appendix B for a detailed comparison).  The rationale



for the strategies recommended here is detailed forthwith, along with the



specifics of the plan.




     Tables 6.3 and 6.4 summarize the control  measures proposed for each



of the four stages.  Also presented are the VMT reductions and emission



reductions which can be expected for each measure and strategy.   Table 6.3



gives RHC emission reductions; Table 6.4 is for CO and NO
                                                         /^



     Stage I consists of a press release and a request for voluntary



abatement.  This "passive" control is included in all subsequent stages.



Stage II has four active (mandatory) controls, a ban on filling of under-



ground service station tanks, control of stationary RHC sources, a ban on



non-essential fleet vehicles, and closure of non-critical government



agencies.  Stages III and IV have only one active control each, operating



on Sunday status and a complete work holiday (including recreational shut



down) respectively.  This overview of the plan highlights one of the

     \.i


attractive features of the strategies -- their relative simplicity.  In



view of the massive communications, surveillance, administrative, and



enforcement problems associated with episode control, it is highly desirable,



if not essential, to formulate simple strategies which are operationally as



easy to implement as possible.  Complicated strategies are likely to fail



due to an inability to effectively implement the control measures.



                                    195

-------
                                                       Table 6.3

                                             PROPOSED CONTROL STRATEGIES:
                                       VMT REDUCTION AND RHC EMISSION REDUCTION
pggSSSSME
! Stage
I
rr
III
IV
*Add
m«Xi^VS^smamr^!i-Jmaa^^M'^m^ii^mmKw^w^^^^s^<:
Control Strategy
1. Media Announcement & Request
for Voluntary Reduction of
Vehicle Use & Emissions
1. Stage I Actions
2. Ban Filling of Underground
Service Station Tanks
3. Shut-down Major RHC Point
Sources & Other RHC Source
Categories Designated by the
APCD- (Exemptions Allowed)
4. Ban (Private) Non-Critical
Fleet Vehicles
5. Close (Non-Essential)
Government Agencies
1. Stage I Actions
2. Operate on Sunday Status
1. Stage I Actions
2. Complete Work Holiday
with Recreation Closed
Inc
VMT
negl .
negl .
—

5%-LDMV
30%-HDMV
5%*-LDMV
negl .
40%-LDMV
& HDMV
negl .
75%-LDMV
& HDMV.
:remental Rec
RHC Emi
Vehicular
negl .
negl .
—

6%
3%
negl .
29%
negl .
54%
lila&SSfaiiiSiSiiiiiSBSfiiEi
uctions [
ssions 1
Stationary i
[
negl.
negl .
4%
3%


negl .
10%
*wB»E!fciiiIX£iiJ.i -3
negl .
i
1

i

64%
.VH^W***-!-^
1% if public school closure is to be included.
cr>

-------
                                             Table  6.4


                                    PROPOSED  CONTROL STRATEGIES:

                           VMT  REDUCTION AND  CO AND NOV  EMISSION REDUCTIONS
                                                     A
Stage
I
II



in
,y
Control Strategy
1. Media Announcement & Request
for Voluntary Reduction of
Vehicle Use & Emissions
1 . Stage I Actions
2. Ban Filling of Underground
Service Station Tanks
3. Shut-down Major RHC Point
Sources & Other RHC Sources
Designated by the APCD-
( Exemptions Allowed)
4. Ban (Private) Non-Critical
Fleet Vehicles
5. Close (Non-Essential )
Government Agencies
1 . Stage I Actions
2. Operate on Sunday Status
1. Stage I Actions
2. Complete Work Holiday
with Recreation Closed
^umjmjatjtijiMLJAi^Arakma^
Increment
VMT
negl .
negl .
--
5%-LDMV
30%-HDMV
5%*-LDMV
negl .
40%-LDMV
&
HDMV
negl .
7 5%-LDMV
&
HDMV
;al Redu
Emiss
CO
negl .
negl .
N/A
7%
4%
negl .
37%
negl .
70%
ctions
ions
M°x
negl .
negl .
N/A
6*
3%
negl . |
I
negl. 1
52%
Cun
VMT
negl .
10%-LDMV
30%-HDMV
40%-LDMV
&
HDMV
75%-LDMV
HDMV
gS^SS^SSSMEL
nulative Re
CO
Emissions
negl.
11%
37%
70%
ductions 1
NO i
Emissions i
negl . 1
9% I
28% i
52% I
-n.-ffl-i-,i^>^iJLL^j*'l3Ji^.^flilIF;,v-i--i ^
*Add 1% if public school  closure is to be included,

-------
     Since this control plan is tied to forecasting episode criteria,
the more severe stages  involve  controls  with  more general  scope than previous
stages, but they are not necessarily derived from previous stages by
"tacking on" additional measures.   The latter approach, where each
stage calls for all the measures of the previous stages, is more suited
to feedback type criteria.   However, each successive stage in the
recommended plan does encompass nearly all the effects of the previous
stages.  For instance,  the work-holiday certainly includes all of the
Sunday status measure.   Operating  on Sunday status in turn covers the
filling of underground  service  station tanks  and government agency
closure and includes most of the effects of banning fleet vehicles and
shutting stationary RHC sources.
6.2.1  Control Strategy Applicability
     As discussed in section 2.2.2, the severest oxidant problem in the
Los Angeles Basin occurs predominantly in the Los Angeles, Orange,
Riverside and San Bernardino Counties.   Since air quality is never
expected to reach alert levels  in  either Santa Barbara or Ventura
Counties, the proposed  control  strategies for the Basin would usually
exclude them.  These two counties  are neither the sources of a large
share of precursors nor the receptors of high levels of pollutants.
Present and projected growth patterns in these areas for the near term
are not likely to change this situation significantly.  They should
be included in the episode controls only when very special conditions
occur.
     The control strategies have been developed for weekday abatement
actions.  It is not anticipated that alert levels of pollution will occur
during the weekend.  During the period since 1970, only one out of about

                                   198

-------
twenty alert incidents has occurred on a weekend (see Table 2.10), and



this occasion was under unusual circumstances.  It occurred  on a holiday



(Saturday, July 4, 1970) when one would expect an unusually heavy amount



of traffic to have taken place both during and prior to the holiday (i.e.



Friday evening).  Even then, though, the emergency level  only exceeded



the harm level by about 10 percent.




6.2.2  Control Strategy Effectiveness




    Of primary importance in any contingency plan formulation is the



technical effectiveness which is to be expected from the  control measures.



Depending on the severity of the impending alert, the programmed control



strategies allow for reductions of approximately 0, 16, 39, and 64 percent



of RHC emissions during Stages I-IV, respectively.   Reductions from indi-



vidual control measures were presented in Tables 6.3 and  6.4.




    Table 6.5 summarizes the effectiveness of the various proposed



strategies and presents the episode levels which can be abated by each



stage.  Analysis of air quality data [Section 2.2.5) suggest that Stages I,



II, and III are the only alerts which may have to be called in this



decade and then only because of oxidant pollution.   It is highly unlikely



that any CO or N02 harm level or a Stage IV oxidant level will ever be



reached.




6.2.3  Implementation and Enforcement




    The implementation and enforcement aspects of the proposed abatement



strategies will, in all likelihood, present some of the most difficult



problems of the program.  With the exception of the Stage I abatement



actions, which are voluntary, the subsequent control strategies all have



direct impacts or numerous public and private groups.  The magnitude of




                                    199

-------
             Table 6.5    EPISODE  LEVELS WHICH  CAN  BE  REDUCED
                         TO  BELOW SIGNIFICANT  HARM LEVELS  BY
                         THE LAPSED CONTROL STRATEGIES
Stage
I
II
III
IV
Emission Reductions
RHC CO NO.,
negl .
16>2%
39%
64%
negl .
12%
37%
70%
negl .
9h%
28%
52%
Episode Levels whic
Oxidant**
.40-4 hr. .60-1 hr,
.40
.435
.50
.59
.60
.65
.75
.88
h Can Be A
CO**
50-8 hrs.
50
54
66
94
Dated (PPM)
NO/*
.50-24^hrs
.50
.53
.61
.77
 *  These  results are based on the approximate air quality-emission level
    relationship developed in Section 2.2.7.  Oxidant concentrations are
    assumed  to be reduced by % of the amount of RHC emission reduction on
    the  day  of the episode.  CO and NO- concentrations are assumed to be
    reduced  by 2/3 of the-amount of CO and NO  emission reductions on the
    day  of the episode.

 **  As noted  in Section 2.2.5, the .40 PPM-4 hr. oxidant level, the 50 PPM-
    8 hr.  CO  level, and the .50 PPM-24 hr. N02 level appear to be the
   "binding  constraint" significant harm levels for each of the three
    pollutants.  (There are three oxidant and CO harm levels and two N0?
    harm levels.)  The .60 PPM-1 hr.  oxidant level has been proposed as
    an alternative to the 4-hr, oxidant harm level.

the implementation  and  enforcement problems  increase  with  each successive alert

stage.   These problems  are discussed  below  for each alert  stage individually.

     STAGE I - The  intent of the proposed voluntary abatement actions is

to create public awareness of the air pollution episode.   Basically,

the controls called for are similar  to the  actions solicited by the

local  APCD during  a first stage smog  alert.   Since no mandatory actions

are required and a  similar program  is  in  existence, no implementation

difficulties are anticipated.   In a  similar vein,  no  enforcement pro-

cedures  are  necessary.

     STAGE II -  Each  of the four control  measures  proposed for this

strategy will require rather detailed  implementation  planning.  The

stationary source controls  would  be most  readily  implemented through
                                   200

-------
the local APCDs, who have established working relationships with the

majority of primary emitters.  A potential difficulty may be enlisting

the cooperation of local APCDs, some of which have their own abatement

plans developed.

     Underground  filling  of  service  tanks would be most effectively

 curbed  at  the  bulk terminals,  a process which would require obtaining

 a  listing  of such stations and informing  them of  the pollution abate-

 ment schemes.   Each bulk  terminal  in turn, would  be responsible for

 informing  the  stations  it serviced.  While requiring advanced pub1 •'city,

 the  information aspects of implementing this control do not appear to be

 overly  cumbersome.

     The successful implementation of control for major point sources

 and  other  designated source  categories will be difficult for several

 reasons.   First,  a fairly large number of such firms exist, employing

 a  sizeable number of employees.  Many of  the firms are politically

 influential and can be  expected to lobby  extensively.  In other cases,

 plans for  shut down procedures, already submitted to the local APCD's,

 may  or  may not be acceptable to the  EPA enforcement personnel.  Some

 firms,  due to  the nature  of  their  product, would  face extreme

 difficulty in  trying to rapidly shut down.  Because of this, it is

 deemed  essential  that a review committee  be established to allow certain

 operations exemption from the  controls.   Enforcement of stationary

 source  compliance is likely  to be  hampered by firms willing to risk
                                                                    -.
 violating  the  regulations and  paying fines.

     The implementation of the Stage II traffic abatement measures

 require  an extensive effort  to inform all of the  groups involved and

will depend heavily on a  cooperative spirit from  the affected agencies.

                                   201

-------
Banning private fleets will most certainly result in negative economic



impacts.  Therefore, it is likely that many fleet vehicle owners will



either object to imposed constraints or not comply when they are in



effect.  In the latter case, enforcement will  pose problems because



of the many fleets which would need monitoring.   Identification of the



fleet vehicles may also pose enforcement problems.



     Notification of the fleets affected can probably be most readily



accomplished by utilizing the records of the Department of Motor



Vehicles.   One inequity present which cannot be  easily eliminated is



the out-of-state fleet vehicles which accumulate significant mileages



within the basin.



     Closing all governmental  agencies will  pose problems of a different



nature.  For these groups successful  implementation is dependent on the



individual plans drawn up by each agency.   Actual  implementation pro-



ceedings are highly dependent on an efficient  and widespread communication



system to all  the agencies involved.   Without  uniformly understood



instructions,  the situation can become somewhat  chaotic.



     STAGE III - The key to successful implementation of this  control.



strategy is adequate publicity of the program's  goals and requirements,



i.e., to operate everything on Sunday status.  Since no new working



definitions are introduced, most individuals should require no special



instructions.   Sunday status has an additional advantage in that



normally the largest share of the individuals  affected by Stage II con-



trols are similarly impacted by a Sunday status  measure.  For example,



all governmental agencies are closed on Sundays; also, service station



filling does not take place on Sundays.   As  before, successful imple-



mentation is largely dependent on voluntary compliance with the  proposed




                                   202

-------
regulations.   Due to the extensiveness of controls, monitoring activities
could,  at most, survey a fraction of  the total number of companies under
control.   Widespread violations, intentional or otherwise, would be
difficult to apprehend and prevent.
     From an administrative viewpoint, the Sunday status control measure
is very attractive.  The only question a company needs to ask is what
its normal Sunday routine is and to execute those activities; in most
cases,  it should be either nothing or very little.  It must be borne
in mind,  however, that the communications  necessary are directly pro-
portional to the number of individuals affected.   The latter stages of
alerts  affect so many individuals that effective  communication  is
crucial to the success of the program.
     STAGE IV - The formulation of a fourth stage alert or maximum con-
trol has  been dictated mainly by the technical effectiveness requirements.
Implementation and enforcment of a complete work holiday with recreation
closed  will  be extremely difficult to execute.  The requirements would
essentially call for shutting down nearly all  normal activities.  The
implementation of this stage would require wide public acceptance and
willingness  to  cooperate.  To achieve such cooperation would  require
a very  extensive mass media campaign.
     While the general public is very concerned about air pollution, it
has become accustomed through the years  to living with these conditions.
In view of the recently improving air quality  trends, the public might
object  strongly to taking abatement actions now for conditions which
used to occur relatively frequently and  consistently, years ago.
                                   203

-------
 6.2.4   Institutional Problems and Soclo-Economic Impact
     Any .attempts to implement and enforce strategies such as those
 proposed  will inevitably encounter a series of institutional problems
 and result in numerous socio-economic impacts.  Wherever possible,the
 formulation of specific strategies has attempted to avoid these
 difficulties.  Institutional problems can potentially arise in a
 number  of areas -- local, regional, state, and federal government or
 public  and private business enterprises.   Socio-economic impacts may
 be felt by institutions or individuals.   In either situation, the
 associated impacts increase with the higher stages  of alert action
 called  for.  To be more definitive, these problem areas are discussed
 in more detail according to stages or alert.
     STAGE I - Since the abatement actions called for are entirely
 voluntary, only minor impacts are expected from this stage of alert.
 Institutional obstacles, likewise, are anticipated  to be minor due to
 the existent emergency alert level program initiated by the local
 APCD's for Stage I alerts.   In essence,  the actions called for the pro-
 posed Stage I abatement,are the same as  the L.A.  County APCD plan for
 first stage alerts.   The primary emphasis at  this level  is individual
 abatement action, and socio-economic impacts are expected to be
minimal.
     STAGE II - In comparison to Stage I  actions, the controls called
 for in this stage affect numerous institutions and  affect tens of
 thousands of individuals.   Since certain  groups have been isolated for
 control, e.g. bulk terminal  operators, major  stationary sources, fleet
 vehicle owners, and  governmental  agencies, it is  certain that
 institutional objections will be raised.   Moreover, the loudest
                                   204

-------
objections are most likely to be voiced by those groups suffering the
largest impacts.  For governmental employees granted a work holiday,
charges of wasted governmental funds and tax dollars are likely to
be levied.  Many fleet vehicle owners will also complain of unreasonable
controls and/or severe economic impacts on private business.   Since
the larger businesses are most likely to own fleets, this group will
suffer more economically from Stage II controls.  On the other hand,
these businesses are also most likely to be able to absorb the costs
of the control actions.  Due to a lack of previous experience, it is
difficult to predict where and in what form, objections are likely to
be voiced.
     STAGE III - The incremental impact from the Stage II to Stage III
control measures are very substantial.  Operating on Sunday status
affects many more individuals and private enterprises.  Economically,
a great impact would definitely be felt regionally.  As a result,
institutional obstacles are also likely to be more significant.  The
importance of an effective mass media campaign cannot be overemphasized
for minimizing the institutional objections likely to be raised.
Because of the economic impacts which would accrue, especially to the
private sector, violations of the controls would be expected.   It
appears that these problems would be most efficiently dealt with well
in advance of any declared alerts, namely through soliciting public
cooperation to the proposed controls.
     STAGE IV - It is projected that only in an extreme situation
would it be necessary to declare a fourth stage alert.  The socio-
economic impacts of an effectively implemented fourth stage alert are
staggering.  The anticipated impacts are of such a magnitude that it
                                  • 205

-------
is questionable whether the public would tolerate them.   More probable



would be the public's tolerance to the severe air pollution -- levels



which the public has previously endured with little more than vocal



objections.   The principal obstacle facing any attempts  to implement



such harsh  actions  is the  lack  of  a precedent for it.  Never has  any area  ever



responded,  as  it is proposed, to air pollution levels  which have  been experi-



enced before without generally  apparent serious  effects  on health.   It remains



to be seen  whether  such an abatement action  can,  in fact,  be implemented.








6.3  RECOMMENDED SURVEILLANCE PROCEDURES



     The four  stage control program outlined above will  require two



different types of  surveillance procedures.   The first,  monitoring  of



air quality and meteorology, will  provide  the basic atmospheric data



for the plan.   The  second, control  strategy  surveillance,  will  provide



information on how  well the control program  is being implemented.



Recommended procedures for both kinds of surveillance  are  presented



below.



     As the description of air  monitoring  capabilities in  section 4.3.1



pointed out, existing air  quality  and meteorological measurements in



the Los Angeles basin are  adequate  for the kind  of episode control  plan



envisioned.  The main problem is that reliable methods for quickly



disseminating  this  data (especially to the EPA)  have not been imple-



mented.  Section 4.3.1 examined several ways, including  the installation



of a real-time telemetry network and the development of  a  regional  data



processing  center,  in which the situation  could  be improved.  It was



concluded,  however, that (at  least for the present) the  best way to
                                   206

-------
conduct air quality surveillance is to let the California Air Resources Board


act as the principal disseminator of information, based on real-time informa-


tion links with the APCDs in the Los Angeles Basin.



     Strategy implementation surveillance consists of two parts, agency


surveillance and source surveillance.  As discussed in 4.3.2, agency surveil-


lance can be conducted in several different ways - by field personnel,  by


telephone contact with all the agencies, or through telephone contact with


the ARB alone.   For the EPA, the last approach appears most practical  in terms


of communications simplicity and administrative cost.   The most extensive


stationary source surveillance program now existing is that of the APCDs.


The ARB also has some stationary source monitoring capabilities.  The


California Department of Transportation has a very useful  traffic monitoring


system.  At present, (if possible), it seems more practical  to rely on  the


existing procedures developed by these agencies,(described in section  4.3.2),
                                                                           y

than to institute a whole new program.   Programs like  these are very expensive,


and duplication of effort should be avoided.  Efforts  should be made,  however,


to organize well defined channels of communication to  EPA Region IX.



6.4  RECOMMENDED OPERATIONAL PROCEDURES



     The operational procedures of an episode contingency plan involve


three aspects.   They are organizational structure, communication channels,


and activity phases.  Each of these has been discussed in Section 4.4;  only


a brief synopsis will  be given below.



     The organizational  structure which will support the proposed plan  must


be capable of sustaining four different kinds of operations:  surveillance,


strategy selection, enforcement, and communication.  Section 4.4.2 has


reviewed three  different kinds of organizational  structure, including  the




                                    207

-------
EPA Region IX Regional  Emergency Operations Control  Center (REOCC), and has



found all of them adequate to support the  needed operations.   There seems



to be no reason why the proposed plan should not fit smoothly into the



existing operations of  the REOCC.




     In section 4.4.3 several  different types of communication are identified



as essential to episode control  operation.   These include communication with



the ARB, APCDs, field personnel, and  public.   EPA Region  IX  appears to have



provided adequate channels for handling all  types of essential communica-



tion.  As section 4.4.3 points out,  however,  it  may  be  advantageous to



incorporate some kind of local communications center into the REOCC structure.




     Section 4.4.4 suggests that episode control  operations  can be viewed



as consisting of three  phases:  an  initiation phase, a  control phase,  and



a termination phase.   For operations  in the L.A.  Basin, it is recommended



that the initiation phase be completed no  later  than 4:00 to  5:00 P.M.  of



the day before actual control  measures are  implemented.   The  control  phase



would last from 6:00  P.M.  of the previous  day until  4:00  or  5:00 P.M.  when



the operation would enter the termination  phase.   In this phase, officials



would decide whether  to continue control measures or to terminate the  episode.




6.5  LIMITATIONS OF THE ANALYSIS




     In developing an episode contingency  plan,  many simplifying assumptions



and approximations must be made.  Certain  simplifications will be rather



crude due to the very limited nature  of relevant data.   These errors in



approximations lead to  uncertainty in the  final  results.




     A considerable effort was made in the present study to eliminate as



much uncertainty as possible.  For instance, in  compiling the basic informa-



tion for traffic controls, a thorough review of  available transportation





                                    208

-------
data was conducted.   The latest additions to this data base, as well  as the



latest reported correction factors to previous data, were included in the



analysis.   Conflicting results in different transportation studies were



resolved by direct contact with the authors.  Air quality data was similarly



reviewed and scrutinized.  In two cases it was discovered that monitoring



results were not reliable.  In organizing an emissions inventory, checks



were made against several existing inventories to insure that no signifi-



cant and inexplicable discrepancies existed.  Although a very approximate



air quality-emission level relationship has been used, this relationship



was developed from an analysis of air monitoring and emission data rather



than just arbitrarily assumed.




     However, many uncertainties could not be resolved, either because the



relevant data were unavailable or because the effort involved would well



exceed the scope of the project.  Thus, many unknowns remain.  The results



of this study should be interpreted with these possible errors in mind.




     The main areas of uncertainty are briefly summarized below:




Hydrocarbon Reactivity Assumptions




     The issue of hydrocarbon reactivity is a very uncertain area in



air pollution control.  Several reactivity scales have been developed.



Although these scales generally agree on many hydrocarbon compounds,



very significant disagreement exists concerning others.  The present



study uses recent EPA reactivity results for mobile sources and APCD-



ARB reactivities for stationary sources (except for gasoline marketing



where EPA factors are used).



     Under nearly all  alternative reactivity assumptions, gasoline



powered motor vehicles are the dominant RHC source in Los Angeles.





                                     209

-------
Thus, the implications of traffic control  are much the same under
different reactivity assumptions.  However, alternative reactivity
indices can lead to major changes in the estimated contributions from
stationary sources, (from around 10% to 25%).  This,  in turn, leads to
considerable uncertainty in the overall effectiveness of stationary
source control.
Compliance with  Traffic Abatement Measures
     From the transportation data base used in this report, fairly
accurate estimates can be made  of the amount of VMT associated with
various activities (e.g.  recreation,  work,  shopping,  government  work,
fleet vehicle travel,  major firm work,  etc.).   The VMT reduction from
completely eliminating these activities,  (not counting substitute
driving), is thus fairly certain.  However, much uncertainty exists
in the amount of compliance to  be expected  from an attempt to eliminate
these activities.  Compliance will depend  on the degree of public
knowledge and acceptance as well as the impact of enforcement and
implementation procedures.  Much of this will not be  known until the
measures are actually tried.
Substitute Driving
     Related to  the previous qualification  is the Issue of substitute
driving.  When certain activities are shut  down, VMT  related to these
activities is cancelled, but extra VMT may  be created in other categories
For instance, in Section 2.2.6  it is estimated that eliminating work re-
lated VMT on weekends results in an increase in other types of travel
amounting to about 1/5 of the cancelled work VMT.  However, during an
air pollution alert, public consciousness  may well prevent people from
                                  210

-------
doing extra driving in non-cancelled activities.  This effect was, in-
fact, indicated by a poll of the federal employees on leave during the
July 26, 1973 alert.  Assuming the results of the poll are unbiased,
the federal employees exhibited no (and possibly negative) overall sub-
stitute driving.
     Whether positive, negative, or any substitute driving can be ex-
pected from the general public during an air pollution alert is not
known.  To be conservative, the normal  weekend substitute driving effect
was included in the VMT reduction calculations made in this report.

Air Quality Forecasting
     A basic dilemma encountered in this study concerned the necessity
of using forecasting episode criteria for a contingency plan to be
effective.  However, forecasts on episode type days are not very reliable
Ways of improving forecast accuracy are discussed in Section 4.1.3.  It
is not certain how much forecast reliability can be improved.   Further,
it is not certain how much forecast errors will  inhibit the overall
utility and consistent application of the contingency plan program.
The Air Quality-Emission Level Relationship
     An air quality-emission level relationship for episode control has
been developed here by comparing average weekend/weekday pollution
reductions and emission reductions.  This result accounts for residual
pollutants and non-linearities in the oxidant-precuser relation in a
very approximate way.   Two separate uncertainties are involved in
applying the relation to any given episode.  First, the relation is
not exact, even on an average basis.   Second, the effect of residual
pollutants depends on specific meteorological conditions, and thus it
                                   211

-------
changes somewhat from episode  to  episode.   Fortunately,  most oxidant
episodes occur in the rather consistent summer-fall  meteorological
pattern for Los  Angeles  (strong daily  sea  breeze  and weak  nocturnal
land breeze).   Although  the air quality relation  developed here is
a "first approximation," it should  provide a  very significant improve-
ment over arbitarily assumed relations, (e.g.  proportional  roll-back).
                                   212

-------
                                APPENDIX A
            EVALUATION OF INTERIM EMERGENCY ABATEMENT ACTIONS
                               (July 1973)


     This Appendix provides a brief evaluation of the Federal interim

emergency air episode program, and is based mainly on the alert action

of the 26th of July, the sole occasion the "previous day prediction"  plan

has been instituted.

A.I  The Interim Emergency Air Episode Plan

     Scheduled for adoption this fall, the air pollution emergency

contingency plan will be in operation too late to be applied during

the majority of the 1973 smog season experienced in the Los Angeles

Air Basin.  In the meantime, the EPA has been charged (under the Clean

Air Amendments of 1970) with implementing an interim emergency episode

program.  In establishing such a program the EPA has requested each of

the federal agencies located in the Los Angeles Air Basin to submit a

voluntary air pollution episode emergency abatement plan.   At the  time

of this writing (two months following the request), 45 of 90 federal  agencies

contacted had submitted a proposed interim plan.   The abatement actions  of

these  plans are to  be  voluntarily executed by the  respective agencies when

smog episodes are either forecast by the EPA (corresponding to an

oxidant-level prediction of .40 PPM), or when they occur without

previous day forecast.  The plan of action for a previous day pre-

diction" is usually quite different from the "same day" episode in that

more decisive action is generally planned when the episode is predicted

the day before its occurrence.  Table A-l is a summary of the partici-

pating federal agencies and the voluntary abatement plans they have

submitted for a previous day alert prediction.  The plans are comprised


                                   A-l

-------
Table A-l,  Federal Agencies Participating in July 26 Air Episode Abatement Action
AM'iln All!?
HAIOUACTEH'.,
LOr.A' IT!
,; •, 'WM ,,JdM
nT.rrr


417 3, Httl Str^t

? CrfSarl^nt Of *h*
'' ',. Secret Service)
jl? '. 3rrir.(; Street
Los Anqdei. Calif
4 U.S. Forest Service
lolela. Calif
o i. cosul Servlc*
P 0 Bo* 3399
6 Long be*c^ l*av*l
ifttfiyarC
Lorvi teach, Cal If.

7 U.S. District Court
Los Angeles. Calif.
8 U.S Department of
Justice (FBI }
} 1000 W1 lihlre Blvd.
Los Angles, Calif.
9. National Aeronautics
(NASA)
10 Deparbnent of The
Air Force
4727 Kllshlre Blvd.
LOS Angeles, Calif.
N ippfb (j
<•',['.<•'
• "KO'cr
1390





77

139


7000

350
750
appro* )
39

93
U.A;'' ILtHFNT', Of
MCCKIM ABflTEMl'IT
PlAfi 'JJfiwiTTtL TO EP/
- me ^'0/e^ to o*
tr.vtl curt^)^


hol idaj

,.p to 75t of tmD|oyefi
on *orli loed priority
Travel to work
encouraged to be via
bicyde, on foot, or

public transportation,
* Three different plani
(condition 1,213)
of var/lr.9 degree* of
activity curti1!»>ent
depending on severity
of smog anticipated
and priori ty of
«m.
Urge employees to use
car pools.
• Urge employees to me
c«r pools irnl public
transportation.
23 ejnp)oje«i to t*

5 *mploye« to b<
assigned on hoi 1day.
Reduction of 1875
VKT from curtfltl«nt
of department vehicle
uSBOe
[>l',DH Of EUCUTIO'.
filC1 PCL^^D
tOMHfNH
flan oil/ DArual I/
""!•';«" s"r un"rt
(rwde Initially
lat- Tjpsrh/, t^-P
^galn lale rfecinf'^^/',











Plan fully eMPcute^.



EPPLC'Lt', 0V
*C.1'IMr.*P/T!/E
LE*,'[ 0^1: ro
^C.' 'OH
•!».,«








None




None



V'tDuCMOr. OF V-r
0r CijPTAH."[',T
Oc OrnciH VCHJO-t
IjS^A
Cu--talUd aopro- JO" ,



minimal.




None. (Nature of
utll liatlon of
vrhlcle travel.;



Horn (Katun- of
work requires
department to b*
In ful 1 pfwretlon. )



wEL/JtT[Ofi Of ^T
£' CAR POOL 1 10.
p';6L 1C
TRAM5PO?~AT1DM
Neil^ible.








H^gl iqifle
ooo 1 s ignored.
Ke^Holble


Nenli^ible



^«!UE T°:P'
= [Du(.LL- £Y
'.ANCLLLA-:OJI -jf t'JP.UC
JlilT: 70 AuE'lCr

OD-.ervO







,»n, «n-., „„»„

'tort ••orr.a! na-iber
of visitation cr>ntarts


•tone. '11 scheduled
appolnti^ents and usgal
carried out .



-',:;< ' PlA'.TIfj'l
• , ; i - [ ? : '-. At
/WHMl ,i A:.'l /,-

n-.jnt-nen!. p^'.ud^OS )r0r n>;^a ar,d

-.^ ,e-,Pn,r«n, p^rts-.Kd ar-,rjr.-
™ -,ni
-------
Table'A-1.   Federal  Agencies  Participating  in July 26  Air Episode  Abatement Action (Continued)
                                                        or IMTE»EIIT «c:io; CXECLTUD J'JIY 26-
AGENCV AND
HEADQUARTERS
LOCATION
11 . Federal Aviation
Administration (FAA)
Los Angelas , Calif.

12. National Transportation
Safety Board
Los Angeles, Calif.
13. Bureau of Indian Affair!
300 H Los Angeles St.
Los Angeles, Calif,
14. Ofrice of Information
1016 N McCadden PI .
Los Angeles , Cal If.

15. U.S. Geological urvey
Los Angflei, Cal f,
16 Department of Th Army
Los Angeles, Cal f,

i7. U.S. Government Small
Sasinei* Administration
649 S Broad-ay
Los Angeles, Calif.
18 U.S. Marine Corps
El Tore (Santa Ana)
Calif.

19. U.S. Department of
Justice
Los Angeles, Cal!f.
20. U.S. Savings Bonds
Division
11000 Ul hhtre 8)vd,
Los Angeles, Ca] f.
?1 . Securities and Exchange
31? N Spring St.
Los Angeles, Cal f.
NUMBER Of
AGCNCV
EMPLOYES
1400

8

20
134


10

1790


160
D.OOO

80
15
40
MAIN CLEMENTS OF
INTERIM ABATEMENT
PLAN SUBMITTED TO EPA
• 700 employees to be
on leave.

Encouragement of car
pools and curtailment
usage.
18 employees to be on
holiday; remaining to
arrive at work by
car pool
Comander 1n charge
to be taken. Among
personnel will take
Approximately 15
employees to be placed
car pools.



146 wi01oy«s to bo
on hoi iday.
* Up to 4250 military
personnel on hoi Iday.
Most of 1500 civilian
personnel on hoi Iday.
curtailment of fedaral
vehicle usage.
Approximately 64
employees to be on
hoi id^y.
All IS employees to
be on holiday.
Al 1 40 e-nployeej to
be on holiday.
DECREE OF EXECUTION
AND RELATED
COMMENTS
Plan not Implemen ed.
Th s age cy li wi h-
ho d ng articlpa Ion
un 1 th CPA Inl iates
ma d tor and col active
pa t dp lion of 11
ag n les.









t had not yet been
pproved by comwndtng
nnounced (Plan was
ubmltteO to EPA 2$ July.)

Plan Implemented only
part ally. Short not ce
made execution di ff Ic 1 1
nd severity of e*pec e«J
eld sufficient by Oe t.
f Defenie to warrant
ull act on.



EI-PLOVECS OH
ADMINISTRATIVE
LEAVE DUE TO
ACTION














Token number.




REDUCTION OF VMT
6T CURTAILMENT
OF OFFICIAL VCHICLC
USAGE










None



0 fklal vehicle use
c rtatled slightly
b t quant 1 ff cat ton
o VMT is indeter-




REDUCTiON OF VMT
BY CAR POOLING,
PUBLIC
TRAhSPORTATlOS
None









None



negl fq'blt




NUHBER OF
VEHICLE TRIPS
REDUCED Br
CANCELLATION OF PUBLIC
VISITS TO AGENCY
'tone









'i^np



None. (Visitation travel
(although pot substantial)
»as carried out as usual




AGFNCY PEACTIO^
TO AIP EPISODE
AMTEuEtfT ACTION'
Feel EPP nas taken e fectlve a proach
by r questing unilate al volurt r-y abats-
nxtnt actions, and oil of the f oerjl
aqcn lei. Instead, I is felt hat EPA
fo^ these across '.He board.








favoraoU attitude rega-dlng concept
be f-vident in initial abdtc^nt efforts -


enforced within the agency




DISRUPTION OF
AGENO OUSJNG
ACTjOfl
Non . (However, 
-------
Table A-l.  Federal Agencies Participating in July 26 Air Episode Abatement Action (Continued)

—!?••

Board
" ^:r™°'
]'J, ', f.rry it.
,'J , S Civil jfrvlce
851 ' Broadway
Los Angeles . Call f



MOOO WiKMre BUJ
Los Ar-eles. Lal it.
2!. n. s General Actouit'ng
Office
iLXJ N, Los Angpli", Si
LCI Angeles, Call'
^H U S. Denartr>ent of
II 000 Wt'shlrc 0Kd.

29. Internal ct>veiui? Service
30C ,'i Los Anyeles St.
Los Anrslf, . CjHf.


* Wes—n Regional
300 N. Los Angeles St.
Los Angeles, CaHf
31 . U.S Forest Service
Pasadena. CaHf.
(Eur«au of Alcohol ,
Tobacco, and Ffreat-ws)
(Headquarters In

•!K,


37

101






'30
!26


2100


3

00
00

OF MAIN ELfO.'T' OF
IIITEP1K ACATEMEMT
ES PLA,'I 'jUUHJTTED TO EP1

be on holiday-
33^10,..^ to b, on
"^rrtr^ r'™tn~
Car -jools and public

hoi iday. Use of rotor
[tools tjrtal led.


Al 1 employee! (45} to
be nn hoi iday.
90 employeas lo be on
hoi Iday.
Appro, ImatMy 81
Other enployees urged
of transportation to
reduce V«T.
* All (£100) enoioyee-,
to be on hoi iday.


All (31 ) ernnloyees
to go on hoi iday.
Several nwasures t«1ng
lak?n as part of ongoing
program to reduce
pollution (I) conversto
of vehklei to dleiel an
LPG fuel ; (2) closing al
burnlnij dumps . will als
encourage c.ir pools.
Some employees to taVe
leave. Others to exer-
cise Individual discre-
tion (n curtal 1 ing
vehicle ntleag«
t
DEGPEE OF EXECUTION
AND RLI/-TF.D
COMMENT:






Plan only partially
could not justify the
execution and the
EPA directives. In
difficulties also





Plan fully executed.







E^PL^fCT j OH
ADMINISTRATIVE
LCPVE DUE TO
ACTION






Nore








2100







PEDUCT10N OF VHT
Br CURTf.uvrNT
OF OFFICIAL VEHICLE
r,/,r,E






•njtely fcO" reduc' Ion
0' motor pool
activity







^DO>~O» t""a tel ^ 25,000







REDUCTION OF V^T
6' CAR POOLI'IG,
PUBLIC
TRA'tS^ORTAI ION















Not Aprllcable






JULY ?6 	
VEHICLE TRIP'j
REDUCED ov
CANCELLATION OF PUULIC
VISITS TO AGCNCY















SCO







AGENCY REACT ION
TO AIR EPISODt
ABATrMFHT ACTION*







ty.







D1 a polntmtnt was expressed regard) g
r.o f slon about onrDttiMnt to eptsod
c 1 rr on preyed ng day (Wednesday) This
e poor tone, o^e concern also no ed
•3 n les (EPA, A CO, ARB}. Agency r -
nls^r -'" "" be'""!" "™C"' ""

*

-
t
DISRUPTION OF
AGLNCT DURING
ACM UN















Intern \ o-srup-
signif cance.
Backlo generate
--Mjft -o»l^.






-------
          Table A-l.   Federal  Agencies Participating in July 26 Air Episode Abatement Action  (Continued)

AGENCY AND
HEADQUARTERS
LOCATION
33. Department of The
Air Force
Los Angeles Office
of Information
11000 WlUhtre Blvd.
Los Angeles, Cal If.
34. Agricultural Research
Service
P.O. Bo« 2326
Riverside, Calif.
J5. Naval Regional
Procurement Office
312 N. Spring St.
Los Angeles. Calif.
36. Department of The Navy
(Navy flecruHIng UistMc
4727 Uthhlre Blvd.
Los Angeles. Calif.
37. General Service-.
Administration
Los Angeles. Calif.
38. Forest Service
San Bernardino. Calif.
Immigration and
Natural Izatlon Service
Terminal Island

40. Defense Contract Audit
Agency
1340 W, Sixth St.
Los Angeles. Calif.
Regional Office
Los flngeles, Cal If.

UMBER 0
AGENCY
MPLOYEE
e
66
109
162
)
542
113

527



iHAIN ELEMENTS OF
INTERIM ABATEMENT
PLAN SUBMITTED TO EPA
Car pools and publ 1c
organized.
5fl employees to be
placed on hoi Iday.
Commanding supervisors to
takp Initiative 1n forma-
tion of car pools,
110 emoloyees to bo
placed on hoi Iday , Use
of government vehicles
curtailed.
• 365 employees to go on
hoi Hay. Government
vehicle usage to be
be urged for employees
not on hoi Iday.
Encourage car pools,
eliminate vehicle
trips where possible.
car pools and use of
publ 1c transportation
whure possible.
Employees requested to
form car pools or use
puhltc transportation.

roent. Use of public
pools to be encouraged.
AD government vehicle


DEGREE OF EXECUTION
AND RELATED
COMMENTS




Plan partially
executed. Internal
approval of plan had not





agenc Instituted
plan ue to short
notic "previous
given late Wednesday.)
EMPLOYEES ON
ADMINISTRATIVE
LEAVE DUE TO
ACTION




(these were al 1
aytlme empl yees'
Note that 4 0 of
chedule. T Is
11 be alte ed In
ar future, hw-
er, when o ly
0 employee
11 work th
ght assignment)



None
REDUCTION OF VHT
BY CURTAILMENT
OF OFFICIAL VEHICLE
USAGE




Approximately 3000
miles.

Very mlno reduction.
of work ( aw enforce-
majority f travel

Approximately 1500
miles.
REDUCTION OF VM7
BY CAR POOLING,
PUBLIC
TRANSPORTATION




No e. (a pools
no emph s ed due to
ve y 1 1m t dflytlme
tr vel 1 rpoois).

Negl Iglble. Approxl -
or carpools - t Is
creased due to ngolnq

Negl 1 giiile
NUMBER OF
VEHICLE TRIPS
REDUCED 8V
CANCELLAT ION OF PUBLIC
VISITS TO AGENCY




None. Observed normal
number of walMn contacts.

None observed - normal


*•"• /'°™<' '«'»".,

AGENCY REACTION
TO AIR EPISODE
ABATEMENT ACTION*




Close proximity f this aqenc to
Federal Informal on Center (r cinl-
enl of numerous omplalnts an In-
enlsode alert ac Ions. Gpner 1 con-
Agency record 1) hanrony between
2) a local spotesnwn for fPA br availa-
ble during actions, and 3) jn emergency
publ Ic net'di,

Basically lyrical reaction tr "holiday'
holiday ipUure wuuld include vehicle
liglitncnt to abatement objectives Qe


wishes to participate 1n smog alert
.let ions
DISRUPTION OF
AGfHCY CORING
ACTION




No dls uplinn apparent.
(full xecutlon of pUn
not ei ected to cause
to nat re gf work -
hacklo Is quickly


None


slant la) backlog pro-
t' I Cm .
cn

-------
     Table  A-l.     Federal  Agencies  Participating  in  July  26  Air  Episode  Abatement  Action  (Continued)
                                                                                       - DESCRIPTION OF ABATEMENT ACTION EXECUTED JULY  26
A&ENCY AND
HLADQUARTERS
LOCATION
«?, Ve;erans Acjnl nl stra I ton
Supply Subdepot
*'47 5. Eastern A«.
Bell . Calif.
43 Veterans Administration
Reqlon*! Hrdlcal Dlstrlc
of Southern Col ( fomtt
Us Angrles, Calif,
44 Marine Corps Pub He
Affair! Office
11000 WUshlre Blvd.
Los Angeles. Cal If.

air episode abatement
plans.
All agencies of
predicted major (T
slgnlf 1carx:e (baled re
on submitted 93
abatement plan). ag
tmt
am
ag
su
ab
Pi
IUMBER 0
AGE NO
:WLOYEE
30
9050
6


9.345
h j
pre ents
t o all
enc
ployment
ong those
snc ei
bral ttng
ilefwnt
ns.)
MAIN ELEMENTS OF
INTERIM ABATEMENT
PLArt SUBMITTED TO EPA
ran 2B to 8 for
rrlval at work place.
ancel all use of
government vehicles.
' urtal) flpency vehicle
saqc to mlnlnum.
ncouragp Curtailment
f private vehicle use.
4 employed to be given
holiday. Official
vehicle usage to bo
cancelled.
14,540 erooloyeei to be
on holiday. Car pools
and public transporta-
tion to b* urged .
curtailed extensively.
13,650 employe*! to b«
on hoi 1day (thli
represents 941 of all
employees to be on
holiday during full
Implementation epfso by uch of agffnclei *bove.
      Thtie plsnj w-re obUlnrd for evaluation fro* the EPA Regional Office.
  Denotes agenclei xMch are of Mjor significance In overall air pollution epliod* f«^r*l abatownt plan.
  •Kajor jlgnlflcifKe' Is assigned on the basis of total matter «p'oy«i •«! ttsgre* of atatanent action planned.
  The igencles of major significance were Included In th« telephone Survey of this study.

-------
mainly of three types of abatement measures:
     1)  Administrative leave to employees.
     2)  Curtailment of official vehicle use, and
     3)  Encouragement to share vehicle travel by carpooling
         or use of public transportation.
As can be seen in Table A-1-, the measure of administrative leave varies
markedly among those agencies submitting plans.
     Total employment of the 45 agencies submitting smog episode abate-
ment plans was 42,407, or about 36% of the 118,000 federal employees
in the South Coast Air Basin [A-l].  The majority of the agency abate-
ment plans provide for implementing some degree of administrative leave.
Complete execution of each separate agency plan would result in a
"holiday" for 14,540 employees throughout the South Coast Basin.
A.1.1  Initiation of the Plan
     On Wednesday afternoon (July 25, 1973), the EPA contacted the
various agencies submitting plans, requesting they implement their
Emergency Air Episode Plan.  The EPA request was supported by the
announcements of Gordon Elliot, chairman of the Federal  Executive Board
in Los Angeles.  In addition, announcements of the intended air episode
abatement implementation were carried over television networks, the
radio, and newspapers.  This "previous day" advisory presumably
allowed the various agencies to initiate advance telephone notification
to all employees targeted for participation in the abatement action.
     In addition to the action planned for Thursday July 26, the EPA
had also attempted to initiate an action for the 25th.  The advance
notification provided by the EPA in this instance (late Tuesday evening)
was insufficient to allow a "previous day prediction" action to be taken

                                   A-7

-------
In fact, due to the lateness  of the notification,  the Federal  Executive

Board in Los Angeles  refused  to support a full  fledged abatement action

recommending instead  that a "present day" abatement action be imple-

mented on the 25th.  As  a result of the uncertain  format in directives,

only a partial  abatement action was voluntarily implemented by the

federal agencies on the  25th.   Because  the abatement effort of the

25th was minimal and  unrepresentative of the potential  mobilized "present

day" version of the federal air episode abatement plan, the scope of this

investigation did not include  an evaluation of  the actions taken by

agencies during this  day.

A.2  Survey of Agency Participation

     Presumably only  those agencies which had already submitted air-

episode abatement would  have  participated in the July 26th air pollution

abatement action.   The degree  of action executed by  these  agencies on

the 26th was determined  by telephone interviews with the air pollution

episode coordinators  (or other cognizant officials)  of the respective

agencies.  The participating  level  of those agencies which had not

submitted plans was also briefly evaluated in this telephone interview.

Additional information regarding individual employee participation was

obtained by means of  a questionnaire survey which  will  be discussed

in the next section.

     The telephone inquiry of each  agency was constructed to reveal:

     e  The level  of  commitment demonstrated by the agency
        in carrying out  the submitted plan.

     §  The number of employees on  leave.

     e  The reduction of VMT  due to curtailment of official
        vehicle usage, and by sharing of commuter  travel in
        carpools and  public transportation.
                                   A-8

-------
     •  The reduction in normal number of vehicle trips achieved
        by cancellation of public visits to the agency.

     •  The degree of disruption experienced by the agency during
        the action.

     •  The agency opinion and reaction to the overall emergency
        air episode action.

     •  The operational problems associated with implementation
        action.

     Due to time  constraints  imposed on the study, only 14 of the 45

agencies submitting emergency abatement plans were included in the

telephone survey.  This limited sampling was justified on the basis of

the significance  of the selected agencies in terms of:  1) their over-

all employment numbers, and 2) the apparent strength of their abatement

plan.  The 14 agencies represent a combined employment force of 39,345,

or 93% of the combined employment from all agencies submitting abatement

plans.  According to a tabulation of the proposed abatement strategies

(Table A-l),  94%  of all employees expected to be on leave during a

full implemented  emergency abatement action are employed by the 14

agencies in the interview survey.

A.2.1  Results of the Agency Survey

     Table A-l provides the summary of the inquiries made to the selected

major agencies which had not submitted air episode abatement plans

before July 26.   The results confirm clearly the assumption that

agencies in this  category did not participate significantly in the

abatement order.

     Table A-1 provides a summary of the inquiries made to the selected

agencies.  Several findings were apparent.  It can be  seen that the

majority of agencies surveyed did not participate in full accord with

their submitted plan.  The more disruptive the submitted plan, the  less


                                   A-9

-------
                Table A-2.  MAJOR FEDERAL AGENCIES WITHOUT PLANS - PARTICIPATION  IN  JULY  26  ABATEMENT ACTION
          Agency and
          Headquarters Location
Number
of Agency
Employees
In Basin
Notification Received
Requesting Action on 26th
Action Taken by Agency
          Defense Contract Administration
          Services Region
          11099 S. LaCienega Blvd.
          Los Angeles
  2448
Received telegram requesting
action by EPA on 26th.
Received no official  advance
notification except by
news media.
Essentially none.
Commander chose to continue
work routines as usual.
A plan is now being prepared
for jiext air episode
prediction.
>

CD
          Naval  Station
          Long Beach
  1000
No official notification received.
Only knowledge of proposed action
was by newspaper and TV
announcements.
No action was taken.
This agency will be phased
out in June, 1974, and
probably will not formulate
plan in interim
          Naval  Underseas Center
          3202 E.  Foothill  Blvd.
          Pasadena
   570
EPA telephoned on afternoon of 25th
to issue "previous day" episode
prediction.  Telegram from EPA
was received on 26th.
Agency encouraged employees
to use public transportation
and car pools.  Official
government use of vehicles
was rescheduled when possible
(40 vehicle trips to
San Diego cancelled).

-------
                                                                THE INTERIM FEDERAL EMERGENCY AIR EPISODE ACTION


              This  questionnaire 1s being circulated to determine the effect  of the emergency smog episode abatement actions 1n the Los  Anoeles Basin on July  26  1973-
         The action consisted of federal agencies participating In a voluntary plan aimed at reducing vehicle mileage In the basin.  Some  of the agencies were shut
         down, while others encouraged their employees to share rides to work.  Many agencies curtailed official  vehicle usage.

              Tne f1n^ air pollution  ewsrgency air episode plan 1s scheduled for adoption this  fall.  Until  it is  adopted, the Interim  version of the plan  (as
         implemented July 26) will  be  applied.  Because the Interim plan consists primarily of participation by only the federal agencies,  its Impact is minimal
         compared to the expected final  version.  However, the actions taken  during July 26, and specifically the overall behavior of federal employees with a  role
         in the plan, will provide  the Environmental Protection Agency (EPA)  with Important Indicators which will be helpful in formulated the final  emergency air
         episode plan.  Your caraful and anonymous responses to this questionnaire are an essential  ingredient in providing the EPA with some of the accurate
         Indicators needed.                                                                                                                    	
              The  questionnaire 1s being conducted for the  EPA by TRH.  It  has  been distributed  randomly among several employees of  your agency   '
         be entirely anonymous_. and all returns are to be maintained 1n TRW's possession.  Please  return your response by folding this  postaged qu<
         Indicated,  and mailing.                                                                                                      y

              Please dispose of any Inquiries you may wish  to make by phoning George Richard.  (213) 535-1592, or
         Environmental Operations Group.
                                                         The answers  are  to
                                                         iestionna1re  as
                          John Trljonis (213) 535-1571. of the TPW
              How  do you normally travel  to work.  (If your mode of travel  differs
              during week, please respond for Thursday.   Note that the emergency   *
              action was called on a  Thursday).             - -..   •      .       JB»
              If you normally 90 to work  In your own vehicle, estimate the  total
              miles of commute and work-related driving  (to work, at work,  return
              home) you would usually  accomplish on a Thursday.         i      rm ,
              How were you notified you should not report to work? •
              Please state briefly  your reaction to the air episode abatement
              action (I.e., do you  think the current plan has any value, should
              It be used again, what are Its weak points, how can 1t be
              Improved, etc.)?
1. a.  Alone (or  w1th fami1v member)  In a personal  motor vehicle
  b.  In personal  vehicle with family ntember who uses vehicle during  day
  c.  By carpool  (with one or more other persons travplina to work)
  d.  Public  transportation
  e.  Walk
  f.  Bicycle
  o.  Other (airplane, boat,  roller  skates, etc.)
              Estimate the miles of driving you and your family would normally do  •
              on  a Thursday 1n all  your  family vehicles  (this includes normal
              driving to work, work-eelated driving, shopping trips, escort trips, ,
              visits, etc.).                                                 jfr

              Estimate actual  miles of family driving 1n family vehicles  durina
              the day of the smog abatement action (Thursday, July 26, 1973). —s^-,

              Which of the following describe the purpose of trips you made 1n
              your vehicle during the July 26 smog episode abatement action. —H
                                                                                                miles
              miles
                les
5.a.  For shopolno
  b.  For recreation
  c.  For social visits
  d.  Alternate work
  e.  Other
  f.  Did not make any trips

6.a.  Telephone at home
  b.  Announcament at work the day before
  c.  Not officially notified, made decision  based on news media
  d.  Telegram at home
  e.  Other  (please describe)
Figure  A-l.     Survey  Questionnaire  for  Federal,  Employes  on  Administrative   Leave  July  26

-------
                                                    THE  INTERIM FEDERAL EMERGENCY  AIR EPISODE ACTION
     This questionnaire Is being circulated to determine the effect of  the emeraency smoa episode abatement actions  in the Los Anqeles Basin on July 26,  1973.
The action consisted of federal agencies participating In a voluntary plan aimed at reducing vehicle mileage 1n the  basin.  Some of the agencies were shut
down, while others encouraged their employees to share rides to work.   Many aqendes curtailed official  vehicle usage.

     The final  air pollution emergency air episode plan Is scheduled for adoption this fall.  Until 1t Is  adopted, the Interim version of the plan (as
Implemented July 26) will be applied.  Because the Interim plan consists primarily of participation by only the federal agencies, Us Impact 1s minimal
compared to the expected  final  version.  However, the actions  taken during July 26, and specifically the overall behavior of federal employees with a role
In the plan,  will provide the Environmental Protection Agency  (EPA) with Important Indicators which will be helpful  1n formulating the final emergency air
episode plan.   Your careful and a_nonymou_^ responses  to this questionnaire are an essential  Ingredient in providing the EPA with some of the accura_te_
1ndicators needed.                                                                                                             ~

     The questionnaire Is being conducted for the EPA by TRW.  It has been distributed randomly among several employees of your agency.  The answers are  to
be entire]^ anonymous, and all  returns are to be maintained in TRW's possession.  Please return your response by folding this postaaed questionnaire as
1ndicaTed, and  mai11ng.

     Please dispose of any Inquiries you may wish to make by phonlna George Richard (213) 535-1592, or John Trljonis  (213) 535-1571, of the TRW
Environmental Operations Group.
1
2.
3.
4.
5.
6.
7.
How do you normally travel to work. (If your mode of travel differs
during the week, please respond for Thursday, Note that the emergency

On the day of the air episode action, how did you travel to work? — **»-
If you normally go to work In your own vehicle, estimate the total

Estimate the miles of driving you and your family would normally do
on a Thursday In all your family vehicles (this includes normal drivinq
to work, work-related driving, shopping trips, escort trips, visits)-£a*
Estimate actual miles of family driving 1n family vehicles during the
day of the smog abatement action (Thursday, July 26, 1973). - s>

How were you (or were you?) requested to participate in sharing of

Please state briefly your reaction to the air episode abatement action
(I.e., do you think the current plan has any value, should it be used
again, what are its weak points, how can it be Improved, etc.)?-"**"
• i ..
• b.
• c.
• 1:
• f .
• q-
*2. «•
• b.
• c.
• I:
• f.
• g.
e
• 3.
•
•* 4.

t
"* 5

8 6. «
. b.
c.
•
»
. d.
• 7.
•
•
Alone (or with family member) in a personal motor vehlc
In personal vehicle with family member who uses vehicle
By carpool (with one or more other persons traveling to
Public transportation
Walk
Bicvcle

..... , .
In personal vehicle with family member who uses vehicle
By carpool (with one or more other persons traveling to
Public transportation
Walk
Bicycle
Other (alrnlane, boat, roller skates, etc.)
mi les
miles

miles

le
during day
work)

dunna day
work)



By company request the day before the action
I was not aware of any such requests
No particular request was made for the dav of the action, however
the company is promoting an ongoing program to encourage car-pooling
and publ ic transportation
By news media and previous understanding of my company's emergency
action plan






           Figure   A-2.
Survey  Questionnaire   for  Federal   Employees  of  Agencies
Encouraging  Transportation' Sharing

-------
reported receiving a large number of telephone inquiries regarding their
open/close status on the day of the abatement order.
     Few agencies (except for those that shut down) noted significant
changes in their contact activity with walk-in clients on the 26th, although
numerous telephone inquiries were made the same day to ascertain the open/
close status of some agencies.  At the two major agencies with shut down
abatement procedures, the cancelled number of walk-in visitations was
estimated at approximately 7500.  Agency disruption produced by abatement
action depends on the degree of execution of the abatement plan, the
abatement plan itself, and the nature of service provided by the agency.
Almost all agencies reported minimal disruption of their operations on the
26th.  However, several of these agencies (those with a proposed plan
including significant administrative leave) conceded a full execution of
their proposed plan would generate appreciable disruption and cause
irreconcilable problems in the short term.
A.3  Survey of Agency Employee Participation
     The successful implementation of the federal agency emergency air
pollution abatement plans depend on the conscious behavior of the individual
employee to voluntarily reduce their personal vehicular travel.  Hence, those
employees on administrative leave-the day of the action were expected to stay
home and minimize their travel in family vehicles, and the employees of
agencies advocating transportation sharing as an abatement measure were
expected to take public transportation or travel in a carpool to work.
     The actual specific actions which were executed voluntarily by
federal employees in the July 26 abatement action were determined by a
survey of individual  employees of the major participating federal agencies.
The selection of the five major agencies for this survey was based on
                                  A-13

-------
their demonstrated committment to implement their submitted abatement
plan.  These agencies,  their overall  employee working force, and their
abatement plan, are tabulated in Table A-3 below.

         TABLE A-3.  FEDERAL AGENCIES OF THE EMPLOYEE SURVEY
     Federal  Agency
No. of Employees
Emergency Abatement Plan
     Internal  Revenue Service
     Department of Health,
     Education and Welfare
     Federal  Bureau of
     Investigation
     Veteran's Administration
     (Regional Medical  District)
   2100
   1140

    750


   9050
Employees on Leave
Employees on Leave
Transportation Sharing
(Car Pooling, Public
Transportation)
Transportation Sharing
     Of all the federal  agencies  employing  an  administrative leave policy
the day of the abatement action,  the  two  selected  in  the  employee survey
accounted for 86-1/2 percent of the entire  work  force on  leave.   Of all
the plans of encouragement to participate in sharing  of transportation,
the three selected represent agencies of  major committment  to the plan.
     The survey was conducted with the use  of  brief and anonymous
questionnaires distributed randomly among several  employees of each of
the selected agencies.   Two separate  questionnaires were  formulated.   For
those employees on leave, the questionnaire (see Figure A-l) was structured
to provide information  about the  following  specific items:
     ©  The employee's  normal mode of travel to  work
     e  The usual  vehicular travel the employee  and his family are
        responsible for on Thursdays
     e  The vehicular travel  (mileage and trip purpose) the employee
        and his family  were responsible for during the July 26 smog
        abatement  action
     e  The means  by which the employee was notified  to participate in
        the July 26 abatement action
     e  The employee's  personal  reaction  to the  air episode abatement
        action.
                                 A-14

-------
 likely was its full execution.  Of a total of 13,650 employees targeted
 for leave under full implementation of all federal agencies'  voluntary
 emergency episode plans, only 3,379 (25%) were absent from work on the
 26th of July.
      The reasons given for failure to fully implement proposed abate-
 ment actions were varied, but could generally be accounted to the
 following:
      »  The confusion and uncertainty which had been generated
         by the initial attempt of the EPA to direct a "previous
         day" action for the 25th (see previous sections).
         Additional uncertainties were generated by apparent
         disharmony among the various air pollution agencies
         (see section A.6!2) as conveyed by the media.   Amid the
         resulting uncertainty, several agencies reported they
         were unable to justify a complete execution of their
         plan.
      o  Internal administrative details remained to be resolved
         before full execution of the plan was possible.
      o  Action was withheld pending resolution of conflicts
         created by the unilateral and voluntary plan submittals.
      Almost all agencies reported a carry-through with a gesture  to
 encourage carpooling or other form of ride-sharing.  In general,
 however, spokesmen for the various agencies estimated the  effect  of
 encouragement to participate in carpooling and public transportation
 as negligible.
      It was assumed by most contacts that the well established travel
habits of employees could not be altered on the short-term notice  typical
of previous day abatement mobilization attempts.  An effort to confirm
these agency impressions by seeking public transit passenger data  only
revealed that daily passenger counts are not recorded in the Rapid Transit
District [A-2].  However, a spokesman for the Rapid Transit District was
                                   A-n

-------
quoted in the newspapers as saying there were no apparent significant
increases in bus ridership during the 26th.
     The majority of agencies reported activating a program which limited
official vehicle travel.  It appeared that this measure caused very minor
inconvenience since almost all  of the cancelled trips were reschedulable
in nature.   The quantification  of vehicle mileage reduction due to this
measure consisted of very rough estimates, since suitable data was lacking
in most cases.  Many agencies reported an intention to maintain more complete
information of their abatement  efforts during future episode alerts.
     All of the participating agencies contacted in the telephone survey
reported they had received the  request to implement the "previous day"
abatement plan from the EPA on  the afternoon of July 25.   This notification
was received by telephone in advance of agency closing hours, allowing each
of the agencies the opportunity to inform employees of the air episode order.
Nearly all  of those agencies contacted also  received a confirming telegram
from the EPA, in most cases on  the day following the July 25 telephone
announcement.  These direct communications were clearly understood by the
participating agencies, and all agencies contacted reported minimal  problems
in informing employees of the air episode abatement order.  By contrast, the
press releases were not as clearly understood, and many of the agencies
reported that numerous employees became uncertain regarding their proper role
in the action after being in touch with the  news media.  The confusion which
developed concerned the question of administrative leave.  When it was
reported by the news media on the eve of the action that federal  agencies
would be closed the following day, many employees expecting to report to work
began to wonder whether their agency would be participating in the leave
policy.  The general public was also confused, and many of the agencies
                                   A-12

-------
The other questionnaire, distributed among those employees of agencies
whose plan consisted of encouragement to share transportation, was
similarly constructed (see Figure A-2).
     All questionnaires contained a brief but informative preamble designed
to enlighten the respondent regarding the purpose and importance of the
survey.  The anonymity of the questionnaire returns was stressed in an
effort to solicit honest and abundant responses.  The responses were pre-
postaged for return directly to the survey headquarters to avoid possible
complications of agency involvement.
A.3.1  Results of the Employee Survey
     The survey questionnaire distribution and return rate were as shown
in Table A-4 below.  The highest return rate was achieved by employees of
the  Internal Revenue Service (86%).  The low return rate of the Veteran's
Administration was probably due largely to the belated distribution of
questionnaires within this agency.  Consequently it was expected that a
substantial number of returns would continue to be received from this
agency after the due date of this writeup.
     Table A-4.  FEDERAL EMPLOYEE QUESTIONNAIRE RESPONSE RATE
                         Employee           Number of
                         Abatement          Questionnaires % of Questionnaire
Federal Agency           Measure            Distributed    Returns
 Internal Revenue         Employees              100             86.0
 Service                  on Leave
 Department of Health,    Employees              110             62.7
 Education & Welfare      on Leave
 Federal Bureau of        Car Pooling,           100             68.0
 Investigation            Public Transportation
 Veteran's                Car Pooling,           200             20.0
 Administration           Public Transportation
 (Regional Medical
 District)
                                   A-17

-------
     Because the questionnaires were directed at two distinct employee
groups (those on leaves and those encouraged to share transportation), the
results were tabulated according to the same distinction.   These survey
findings are discussed in the next sections.
A.3.1.1  Usual Mode of Travel
     To determine the actions of the individual  employees  expected to
participate in the smog abatement plan, it was necessary to establish first
the usual mode and extent of vehicular travel  of these affected employees.
Background information regarding usual  mode of travel  was  tabulated from
responses to the initial  items of the questionnaire, and is shown, for
each of the response groupings, in Table A-5 and A-6.   Since a large
majority of the respondents drive alone to work in  their own personal
vehicle, it is clear that voluntary abatement actions  on July 26 could have
a significant effect on vehicle mileage traveled by those  in the affected
agencies.  This reduction could be achieved either  by  the  administrative
leave plan, or by the increased sharing of transportation.
     In those agencies in which employees were requested to share trans-
portation to work, minimal  action appears to have been taken by the
employees to effect an implementation of their agency  plan (see Table A-6).
On the day of the emergency smog abatement action,  practically all employees
responding to the questionnaire reported no change  in  their normal mode of
travel to work.  For those employees on administrative leave the day of the
abatement order, the expected change in vehicular travel was, of course,
substantial (see next section).
A.3.1.2  Reduction in Usual Vehicle Mileage Traveled (VMT)
     The amount of reduced vehicle travel which each of the employee
response groups accounted for on July 26 is tabulated in Table A-7.  The
                                   A-18

-------
           Table A-5.   USUAL MODE OF TRAVEL - FEDERAL EMPLOYEES ON ADMINISTRATIVE LEAVE JULY 26
Federal Agency
Internal Revenue Service
Department of Health
Weighted Average
Mode of Travel to Work (Frequency of Responses Expressed in Percent of Total
Returns Responding to Item)
a) Alone in
Personal
Motor Vehicle
84.9
82.3
83.8
b) In Personal
Vehicle with
Family Member
Who Uses Vehicle
During Day
1.1
4.4
2.3
c) By
Carpool
11.6
4.4
8.9
d) Public
Transportation
2.3
4.4
3.1
e) Walk or
Bicycle
0
2.9
1.1
Weighted averages are calculated based on the proportion of employees an agency contributes  to  the  overall
number of federal employees reportedly participating in the particular type of abatement plan.

-------
            Table A-6.  CHANGE IN MODE OF TRAVEL - FEDERAL EMPLOYEES WHICH WERE URGED TO SHARE TRANSPORTATION  JULY  26
Federal Agency
Federal Bureau of
Investigation
Veteran's
Administration
Regional Medical
District
Weighted
Average
Date
Usually
July 26
Usually
July 26
Usually
July 26
Mode of Travel to Work, Usual, and on July 26. (Frequency of Responses
Expressed in Percent of Total Returns Responding to Item)
a) Alone in
Personal
Motor Vehicle
64.2
61.2
57.5
60.0
57.9
60.1
b) In Personal
Vehicle with
Family Member
Who Uses Vehicle
During Day
2.9
2.9
0
0
.2
.2
c) By
Carpool
29.8
32.8
17.5
15.0
18.4
16.3
d) Public
Transportation
0
0
20.0
20.0
18.5
18.5
e) Walk or
Bicycle
3.0
3.0
5.0
5.0
4.8
4.8
I
[S3
O
     Weighted averages are calculated based on the proportion of employees an agency contributes to the overall
     number of federal employees reportedly participating in particular type of abatement plan.

-------
Table A-7.  CHANGE IN VEHICLE MILEAGE TRAVELED - EMPLOYEES ON LEAVE AND EMPLOYEES ENCOURAGED TO SHARE TRANSPORTATION





Federal Agency
Internal Revenue
Service
Department of
Health, Education
and Welfare
Weighted Average
(Employees on Leave)
Federal Bureau of
Investigation

Veteran's Adminis-
tration, Regional
Medical District
Weighted Average
(Employees in Ride
Sharing Program)




Main Element of
Abatement Plan
Employees on
Leave
Employees on
Leave

Employees on
Leave
Sharing of
Transportation
Encouraged
Sharing of
Transportation
Encouraged



A
Usual Miles of
Work-Related
Driving (Thurs-
days) in
Personal Vehicle
34.4

18.5


28.6

51.5


14.5


17.4


B
Usual Overall
Miles of Driv-
ing (Thursdays)
by All Vehicles
in Family
54.7

41.7


49.9

77.0


33.4


36.7


C
Total Miles of
Travel for All
Vehicles in
Family on
July 26
16.5

22.5


18.6

73.2


30.3


33.5





Reduction in
Normal VMT
(B-C)
38.2

19.2


31.3

3.8


3.1


3.2
|
J
1.  Mileage figures represent averages of return responses for the designated categories  above.

2.  Weighted averages are calculated based on the proportion of employees an agency contributes  to  the  overall
    number of federal employees reportedly participating in the particular type of abatement  plan.

-------
survey results show that employees  who were on leave reduced the usual
overall travel of vehicles  in their family by approximately the usual
amount of work related travel they  normally account for.   Hence, employees
who were on leave July 26 did not appear to generate any  substitute travel
in place of their normal work related driving.  (If anything, the results
indicate that they also drove less  other types of trips.)   This trend  was
in distinct contrast to the expectancy for holiday or weekend travel,  when
employees off from work usually do  more non-work driving  than they do  on  a
workday.  This substitute driving usually amounts to about 20 percent  of
their work related VMT.
     Table A-8 summarizes the nature of vehicle trips taken by the
employees on leave, and tends to explain why substitute driving (in place
of work related travel) was minimal.  Only 4.3% of all  trips were executed
for recreational  purposes,  and 42.1% of all  employee respondents made  no
trips at all.
     Employees working for agencies which encouraged transportation
sharing as the main element of their emergency abatement  plan, achieved
a very minor reduction in the normal travel  of their vehicles.  Table  A-7
demonstrates the  almost negligible  (9% reduction) effect  on VMT for
agencies attempting a carpooling and ride-sharing abatement plan on
July 26.  This reduction apparently was due to the agency's program
to curtail work-related vehicle travel during working hours, since the
survey clearly showed that there was no increase in car pooling efforts
July 26.  This confirms the findings achieved in the telephone survey  of
agencies discussed earlier in Section 1.2.  It is interesting to note,
however, that employees of those agencies featuring a ride-sharing plan
                                   A-22

-------
                  Table A-8.  NATURE OF TRIPS MADE .BY FEDERAL EMPLOYEES ON LEAVE JULY 26
CO

Federal Agency
Internal Revenue Service
Department of Health,
Education and Welfare
Weighted Average
Purpose of Vehicle Travel on July 26. (Frequency of Responses Expressed in
Percent of Total Returns Responding to Item)
a) For
Shopping
31.4
25.0
29.1
b) For
Recreation
3.5
5.9
4.3
c) For
Social Visits
1.2
1.5
1.3
d) Alternate
Work
3.5
7.4
4.9
e) Other
22.1
22.0
22.1
f) Did Not Make
Any Trips
41.9
42.6
42.1
Weighted averages are calculated based on the proportion of employees an agency contributes  to  the  overall
number of federal employees reportedly participating in the particular type of abatement plan.

-------
were normally in car pools much more often than those respondents of the
survey on an administrative leave plan (see Tables A-5 and A-6).  It
appears that many respondents of the survey were already somewhat instru-
mental in reducing VMT on a continuing basis,  by joining in car pools and
sharing commute mileage.
A.3.1.3  Notification of  Federal. Employees to  Take Action
     The manner in which  federal  employees were notified for
participation in the July 26, as  determined by responses to item #6  of
the survey questionnaire, is tabulated in  Table A-9.   The survey results
suggest that those employees who  participated  in the  administrative  leave
policy were far more aware of their expected role in  the abatement action
than were those employees of agencies  advocating transportation sharing.
In fact, it appears that  about one half of the latter employees were
completely unaware of their implied responsibility for the July 26 action
This data would help to explain the finding noted earlier that  increased
car pooling (and subsequent VMT reduction) was practically negligible
among those employees responding  in the survey.   Table A-9 also suggests
that agencies exercising  the administrative leave policy were  successful
in providing widespread advance awareness  of the upcoming "holiday"  to
its attentive (and responsive) audience.
A.3.1.4  Employee Reaction to Plan
     The final  item of the survey questionnaire provided an open ended
outlet for respondents to express their opinions with regard to various
aspects of the emergency  abatement action.  The responses received were
varied and abundant.  Less than six percent of the questionnaire returns
were without some type of written response on  this item.  Many  of the
comments were of a repetitious nature  and  were classifiable by  their
                                  A-24

-------
              Table A-9.   MEANS BY WHICH FEDERAL EMPLOYEES WERE NOTIFIED TO TAKE EMERGENCY  SMOG  ABATEMENT ACTIONS

Federal Agency
Internal Revenue
Service
Department of
Health, Educa-
tion, and
Welfare
Federal Bureau
of
Investigation
Veteran's
Administration,
Regional Medical
District
Main
Element
of Agency
Emergency
Action
Plan
Employee
Leave
Employee
Leave
Trans-
portation
Sharing
Trans-
portation
Sharing
Manner of Notification. (Frequency of Response Expressed in Percent of
Total Returns Responding to Item)
a) Announcement
at Work the Day
Before Action
80.0
82.3
19.1
2.5
b) Telephone
At Home
10.4
10.3
0
0
c) By News Media
and Previous
Understanding of
Agency's Emergency
Action Plan
1.1
4.4
7.4
25.0
d) No Particular
Notifications were
Made for the Day of
the Action; However
Agency is Promoting
Ongoing Program
Encouraging Car-
pooling and Public
Transportation
0
0
29.4
20.0
e) Was Not
Aware of any
Notification
0
0
41.2
50.0
f) Other
8.1
2.9
2.9
2.5
I
ro

-------
general content.   The response categories appearing to be most predominant
in theme were established from a tabulation of all  returns, and are shown,
with corresponding response frequencies, in Table A-10.
     It is clear that federal  employees who participated in the
"holiday" type action exhibit  a high regard for the value of the current
emergency abatement plan.  Whereas approximately 70 percent of these
"holiday" participants thought the current plan had positive or potential
value, only about 35 percent of those federal  employees participating under
the transportation sharing action thought the  plan  to be of value.   In other
distinctions between the response of the two employee groups,  it is seen
that employees exercising the  responsibility of administrative leave
July 26 were far more inclined to suggest improvements for the current
plan when they believed it had value.  Participants in the transportation
sharing campaign were greater  advocates of rapid transit, as ten percent
of these respondents (compared to three percent of  the employee groups on
leave July 26) recommended its development as  the most feasible and
effective smog abatement measure.
     The general reactions of  federal employees illustrated in Table A-10
supports to a certain extent the social consciousness demonstrated  by
each of the two employee groups in exercising  voluntary actions to  reduce
vehicle travel on July 26.  The employee group on leave appeared to have
accounted for substantially less vehicle travel than would have been
expected of employees in the usual  "holiday" situation (see Section
A.2.1.2 and Table A-7), indicating this employee group imposed deliberate
self-restraints to help achieve the goals of the abatement action.
Conversely, the employee groups which was urged to  share vehicle travel
demonstrated almost negligible change in their usual travel behavior in
                                   A-26

-------
                 Table A-10.   EMPLOYEE REACTION TO  EMERGENCY AIR  EPISODE PLAN  -  BASED  ON JULY 26 ACTION
                                    (Frequency of Responses Expressed In Percent of Total Returns Responding to Item)










Federal Agency
Bureau of
Internal Ravenue
Department of
Health, Education
and Welfare
Federal Bureau
of Investigation
Veteran's
Administration
(Regional
Medical District)









Main Element of
Abatement Plan
Employee
on Leave
Employee
on Leave

Transportation
Sharing
Transportation
Sharing











Current Plan
Has Little
or No Value
9.3

17.6


33.8

27.5











Current Plan
Has Positive
or Potential
Value
69.7

72.1


33.8

37.5







Plan Should be
Extended to
Include Greater
Participation
(Private
Businesses &
All Government
Aqencies
44.1

32.3


13.2

15.0



Current Plan
has Weakness
In that it
Depends
Heavily on
Voluntary
Action of
Individuals
to Curtail
Personal
Vehicle
Travel
8.1

10.2


7.3

7.5







Organization,
Forecasting,
and News Media
Treatment
Should be
Improved to
Make Plan
Effective
15.1

14.7


2.9

22.5









Rapid Transit
or Adequate
Public
Transportation
the
Only Answer
3.4

2.9


10.2

10.0












NO
Response
3.4

2.9


8.8

10.0













Other
24.4

14.7


22.1

15.0



 I
ro

-------
terms of mode of travel.   This  group  did  accomplish  a  nine percent
reduction in vehicle travel,  due  presumably to  the agency's active program
to curtail  all  work-related  vehicle travel  during  working hours.
A.4  State  and  Local Actions
A.4.1  Actions  by the California  Air  Resources  Board
    The California Air Resources  Board  (CARB) assumes  the responsibility
of supervisor and coordinator to  the  various air pollution control  districts.
The ARB receives information  from air monitoring stations  and  advises  the
districts regarding expected  pollution  levels.  When a smog siege is  fore-
casted in a given district,  the ARB coordinates the  activation  of emergency
district programs.  The advisors  are  prepared by 9:30  am for a  "same  day
prediction" and by 2 pm for  a "following  day prediction."
    Preceding the emergency  air episode action  of  July 26, the  ARB
received information indicating that  sub-emergency levels of pollution
were expected for the 26th.   The  ARB  advised the districts on the
afternoon of Wednesday (July  25)  that lesser values  were expected for -
Thursday, and that emergency  district programs  would probably not be
required as they were in some districts on  July 25.
A.4.2  Actions  by the APCD's
    Each of the Air Pollution Control Districts (APCD's) is responsible
for executing its own distinct  emergency  air episode abatement  program.
Emergency episode plans defining  alert  stages and  actions originated  in
Los Angeles County and have  been  followed since 1955.   First stage
actions in the  various districts  include  banning of  burning of  combustibles,
voluntary curtailment of vehicle  trips, and notification of industry through
the district's  emergency radio  network  of the alert  and need to prepare
                                   A-28

-------
for shutdown.  More severe second stage actions,  initiated  upon
increasing pollution levels, involve shutdown  of  industrial  operations
and,possible control of motor vehicle operations.
     Since air pollution levels on July 26 were considerably less than
the first stage levels required for activation of any  of  the separate
APCD emergency abatement programs, no actions  were taken  by the APCD's
of the basin.
                                   A-29

-------
A. 5   IMPACT OF ABATEMENT ACTION ON VEHICLE MILEAGE TRAVELED  (VMT)

      The reduction of VMT due to the emergency air episode action was

estimated in two ways:  1) based on information obtained from the

Agency and Employee surveys and 2) vehicle traffic count data obtained

from  the California Department of Transportation (DOT).

A.5.1  VMT Reduction as Determined from Agency Interviews and Employee
       Surveys

      Through a series of assumptions based on pertinent transportation

studies, and a utilization of the Agency interview findings  (see Table A-l),

the result of the various abatement actions executed on July 26 may be

may be expressed in terms of the reduction of VMT attributable to the

action.

      For the case of administrative leave, the VMT reduction due to

federal employees absent from work was calculated on the basis of

mileage figures tabulated from the employee survey results (see Section

1.3.1.2).  These survey findings indicate that each employee on leave

was responsible for a reduction in vehicle travel  (July 26) of about

31 miles.  Hence the percentage VMT reduction achieved by the 3379

employees absent from work is


               3379 x 31
       (Basin total weekday VMT)  x  10° = '055%

      Total weekday VMT in the Basin is 183,000,000 (see Table 2-12).


     Total  VMT reductions attributed to a new sharing of commuter travel

was also calculated on the basis of the employee survey results.  The

survey indicated that employees of agencies encouraging ride-sharing

accounted for a mileage reduction of about 3 miles each.  Hence the

percentage  VMT reduction  achieved by the approximate 22000 employees

                                    A-30

-------
urged by their agency to participate in this plan is


                    22.000 x 3
                   183,000,000
                                        _
                                  x  IUU ~  '
     The total estimated decrease in Basin daily weekday VMT due to

curtailment of official vehicle usage was determined from the agency

survey to be approximately 37,000.  This may be expressed in percentage

reduction of overall weekday Basin VMT  as
                    37>OQO	x 100 = .020%
                  183,000,000


     The percentage decrease in overall daily weekday Basin VMT

attributed to cancellation of some 7,500 customer visits to those

agencies which were shut down during the abatement action, was esti-

mated as follows:


% VMT      _  [no. of cancelled )     (average trip]     [partial  cancellation]     ,,
Reduction  ~  \    visits	/  x  \   length   /     \  adjust,  factor   /  x  '
                               (overall weekday Basin VMT)

           -   7,500 x 6.5 x %     ,nn  _  nny
           "     183,000,000    x  10°  ' '013%


In the above calculation, 6.5 (Table 2.13) is the average trip length

in the "Home to Other" category for driving patterns studied in Los

Angeles.  An adjustment factof or % was assumed in recognition of the

fact that only a portion of the cancelled agency visits account for

full cancellation of vehicle trips.

     The overall estimated percentage by which Basin daily weekday VMT

was reduced due to abatement actions taken on July 26 is
                                   A-31

-------
   .055     +   .036   +    .020     +      .013        =   .12%

'employees^ /car    \ /official   ^  /cancellation^
                                  /customer
                                  wisits
:employees\  /car   \ /official   \
on leave j  (pooling)! vehicle    J
         '  \      / \curtailment/
 Most  of  this  reduction was  achieved by  the  shut  down  actions  of the

 Department  of Health, Education and Welfare and  the  Internal  Revenue

 Service  (from both  employees  and customers  staying..home).   Had  all

 other agencies  carried out  their emergency  air episode  plans  as

 submitted,  the  expected  reduction  in Basin  VMT would  have been  in-

 creased  four  times  (based on  a linear proportioning of  number of

 employee leaves  versus VMT  reduction).  Carrying the  hypothetical

 projection  further  by conjecturing an air episode  action  in which all

 federal  agencies  participate  according  to plans  along the lines  of

 those now submitted, one would expect a .7% VMT  reduction.

 A.5.2 VMT  Decrease  as Determined  from  Traffic Count  Data

      Typical  Thursday traffic patterns  were considered  by evaluating

 vehicle  counts  (obtained from DOT) from the "42  mile" freeway loop

 (Figure  A-3).   Vehicle-counts for more  extensive coverage of  the

 Basin transportation facilities were not yet available  at the time of

 this  study.   Traffic patterns on the 42-mile loop  were  presumed  to be

 somewhat indicative  of overall basin VMT, although the  limitations of

 this  assumption  are  easily  recognized.   Table A-ll documents the

 vehicle  count of  the loop at  various selected locations for a number of

 Thursdays in  nearby  temporal  proximity  to the July 26 air pollution

 episode  action.   The data show relatively small  fluctuations  in  the

 average  of  the  aggregated loop traffic  count from  one Thursday  to  the

 next.  For  example,  corresponding  to the four dates of  Table  A-ll,  the
                                  A-32

-------
                                    SANTA MONICA FREEWAY
Note:  Arrows  indicate the
      location of monitoring
      stations
             Figure A-3, "42 Mile Loop"

             Source:   California  State Department of Transportation
                                   A-33

-------
                                                           Table A-ll

                                                  TOTAL DAILY  VEHICLE  COUNT
                                        SELECTED STATIONS  IN  "1»2 MILE FREEWAY LOOP"
DATE
June 28
July 12
July 19
Aug 2


July 26

15E
102000
1023003
100200
100400


101000

SANTA
24E
130200
130900
128700
129200


129000

MONICA
15W
106200
105400
104500
105900


107200

24W
138100
137500
135900
136000


137600
FREEWAY
13N
97200
100300
98000
97800


97500
& FREEWAY STATION
HARBOR
24N 13S
114200 88*400
115400 88700
115000 88100
111900 87200


1 12200 88000

24S
1 10500
109300
109500
105100


107200

4N
98100
99100
97500
98200


97600

SAN
27N
88300
84200
86700
86400
OVERALL
STANDARD
84100

DIEGO
4S
90600
90400
91200
89900
AVERAGE
DEVIATION
91400

275
97900
100000
100500
100800


99600

AVERAGE
ALL FWY'S
105100
105300
] 01*700
104100
104700
540
104400
3>
co
           a.   Figures for July 12 were adjusted  to  represent all lanes of  fwy when available data contained  traffic counts
               for all but one lane of traffic.


               Data for July 5 was not included due  to its temporal  proximity  to a national holiday, which,  it  was  felt, would
               cause  it to be unrepresentative of normal Thursday traffic.
              Source:  State of California,  Department of Transportation,  500  South Vermont, Los Angeles,  California.

-------
standard deviation for the average vehicle count in the loop is only
540 vehicles/day, or just .5% of all vehicles counted daily in the
loop.  In comparing the vehicle count data for July 26th to the average
Thursday traffic count, it is seen that the July 26 count value is well
within the expected normal variation for a Thursday vehicle count.
Hence the transportation data indicate that a significant reduction
in vehicle count did not occur in the 42 mile loop on July 26, and that
any reductions in VMT due to the July 26 abatement actions would not
be discerned in the data.  Actually, the fact that typical Thursday
vehicle counts fluctuate relatively little  (as evidenced by a .5%
stand deviation), would suggest that measurements of the 42 mile loop
could serve as a sensitive indicator to traffic control measures to
reduce VMT.  However, this sensitivity is not tuned fine enough to
discern the .12% VMT reduction estimated by the results of the agency
interview survey.
     In another tabulation of transportation traffic counts (Table A-12)
on the freeway loop morning rush-hour vehicle counts were examined for
typical Thursday traffic preceding the July 26 air episode action.
The expected fluctuation in morning rush hour vehicle count is sub-
stantially greater than that observed for the daily variation (Table A-ll),
In comparing the vehicle count data for July 26 to the average Thursday
traffic count, it is seen that the July 26 count value is well within
the expected normal variation for a Thursday morning traffic count on
both the Harbor and San Diego Freeway portions of the loop.  However,
it can be seen that morning traffic counts for July 26 on the Santa
Monica Freeway section were significantly less than usual.  This decrease
                                   A-35

-------
                                Table  A-12.
                    MORNING  RUSH HOUR  (6-9  AM)  VEHICLE
                        COUNT,  SELECTED  STATIONS
                        IN "A2  MILE  FREEWAY LOOP"
DATE

June 28
July 12
July 19
Overal 1 Average
Standard Devia
July 26
HUNDREDS OF VEHICLES
HARBOR
(Northbound)
195
202
197
198
2.9
197
SAN DIEGO
(Northbound)
212
210
209
210
1.3
209
SANTA MONICA
(Eastboard)
213
212
213
213
.6
200
Values in table are average  of  5  stations  for  each of  freeways.   The
apparent low value for traffic  volume  on  July  26 for the Santa Monica
Freeway is due, at least  mainly,  to  a  traffic  accident,  which caused
appreciable conjestion.
Source:     State of California
           Department  of Transportation
           550 So.  Vermont
           Los Angeles,  California
                                  A-36

-------
was due in all likelihood to a traffic accident and the subsequent

traffic congestion.[A-3]  Discounting the vehicle counts on the

Santa Monica Freeway, it is clear that any reductions in morning rush

hour VMT due to the July 26 abatement actions are not discernible in

the data.

     The data from the freeway loop suggest there is little need to

examine further data or invest more time in quantifying the July 26

abatement actions in terms of its effect on overall  vehicle usage in

the Basin.  It is improbable that the effects of the action can be

measured since the resulting changes are within the usual  daily vari-

ations.  Suffice it to say that the impact of the actions  were un-

doubtedly insignificant in the sense of air pollution abatement.

A.6  PUBLIC REACTION AND MEDIA INVOLVEMENT

     The significance of the July 26th air episode action  in terms of

public reaction and attention received from the media was  appreciable.

The following sections provide a summary of this activity.

A.6.1  Agency Reaction

     Table A-l includes a summary of the attitudes and recommendations

voiced by respondents of the agency survey interview.  These reactions

may be broadly summarized as follows:

     •  Skepticism is prevalent regarding the effectiveness of
        limited administrative leave as a pollution control
        measure to decrease Basin VMT.  Recommendations are that
        employee leave be enforced throughout all  agencies, along
        with accompanying controls to ensure substitute motor
        travel does not occur.

     0  Most agencies appear to feel the overall emergency air
        episode implementation is poorly administrated.  Recom-
        mendations for improvements include educating for
        increased awareness, more definitive abatement alert
        action announcements, better harmony between the air
        pollution agencies, and stronger committment and
        leadership by the EPA.

                                  A-37

-------
     In general it appears that much of the discontent over the current



emergency episode program is a derivative of its  interim status.   The



interim approach, calling for a voluntary submittal  of abatement  plans



from the federal agencies, is a temporary stopgap until  a more complete



emergency air episode program is adopted later this  fall.   This interim



perspective is evidently absent as  common public  knowledge.   Consequently,



it was found that many of the criticisms of the July 26  action were



directed at the failure of the, plan to  be comprehensive, definitive,



all inclusive, and mainly, authoritative.   During the interview survey,



several respondents (usually the agency air episode  coordinator)



indicated that the July 26th action demonstrated  the futility of  de-



pending on voluntary individual agency  committment in affecting



positive encompassing action.  These respondents  called  on the EPA to



define and direct appropriate abatement control measures,  and to  enforce



these collectively among all agencies.   In one case, an  agency (the



FAA) had withheld its participation on  July 26th  and future  abatement



actions pending a decision by the EPA to issue more  definitive direc-



tives.   It is evident, based on the nature of these  complaints, that



the inherent weaknesses of an interim stopgap have been  responsible



for many of the inadequacies so apparent in the program  at this time.



The future adoption of the State's  plan later this year  should



alleviate some of the problems now  being experienced.   In the meantime



it would seem that greater effort towards shaping awareness  of the



interim nature of the present plan  could result in more  positive public



opinion of the entire emergency air episode abatement program.
                                  A-38

-------
A.6.2  Reaction from the Media
     The implementation of the Emergency Episode Plan by the federal
agencies drew much attention in the local newspapers.  Figures A-4 and
A-5 are representative articles demonstrating local newspaper reporting
on the action.  In general the reporting was consistent with information
obtained in this study.  The major contradiction of fact concerned the
number of federal employees absent on leave during the 26th.  Based on
interviews with the Federal Executive Board during the action, most
reports carried information that an estimated 25000 federal  employees
were-out of work due to the action.  The survey of agencies, however,
in this study, indicated less than 4000 persons out of work.
     In addition to the comprehensive reporting of the July  26th action,
a number of editorials appeared in the media.  Figures A-6,  A-7, A-8
demonstrate the tone conveyed by these editorials.  As can be seen, the
emergency action generated substantial controversy, with both affirmative
and negative positions being expressed throughout the media.
     The evening outlook (Figure A-6), in an editorial July  28, issued
the criticism that the EPA abatement action was irresponsible, and
accomplished little in the way of air pollution abatement while at the
same time costing taxpayers millions of dollars.  (It was noted that
this cost occurred on a day in which the predicted conditions changed,
such that heavy smog failed to materialize.)  The editorial  referred
to "an estimated 100,000 federal employees out of work," and noted it
was unrealistic they would stay home according to the intent of the
abatement plan.
     In a Los Angeles Times editorial (Figure A-7), the action was
said to have performed a valuable service by demonstrating what will
                                   A-39

-------
                                 V.,  v
                                 Vv
                             K '**   i *> 0 • ' |C\
                             m   li  i-vG
                          BY DICK MAIS'
                      __    THta Hid Writer

  Federal arencies in five Southern California counties \Yedr.esday \vere
Ejked to c'.osc ih^r cfi'jces or curb  activities today because of extremely
 ;>h daytime smog level* expected a', many in": and citiei.
 I: \viilte ;he first time the recently
.f she K:-.\i
r.c.1" i.i S
 T.-c  p^
             fc^cr:!  Err.errcncy Air
             n is 10 be pui into cffccL
              ;:-:'. Hi. :n--._ nf fhe


              ^t of ihr rrr:~.~-Tl rfiice
             irr-.cr;al Pro'.cciion A;-
               Frcr.tlico.
              actually c;!l- only fnr
  rjr'-.iij all xmncrc.-.'.'.r-- SLio^.obilc
  ir.i\^J. u.-e cf »xc!jlnn  o.'.':cc forcci,
  pr-i ur^rcs c--cr.iiii ncr.v.r.-.c-I to MM
  pc-V:c u-j-i-porutior; cr c^r pools (n-
  x'.ccc r.f ihcir own vehicles to  cora-
  .r.i;;e Troir. home to work.

  fr. .-:<:! C;:.L.-S will be clo/ca today.
    r.".V,;i  .'-.ii,i t!:ci~c  arc mo:t  thjn
  K- '/ 3-J  f^ric-rs!  cr-.jTlr.yci  ;;i   t"ic
  S«..'h Co;jt  Air K-^:.i  \\ hich in.-
  rir ,c5  Lo.> An^cir5. \'i*rtura.  Or-
  rrpc. Tijvcriidc aua San Ecrnirdino
  cojr.ifci

     XumUcr Affected Unknown
   ]--;!:o:t i^:'! he  could re: estimate
 hf.iV jr.ir.y ..xcrai cmp:o*--cs \vouM
 re.r.oln ai r.xr.e, cxi.^L-.ln^ that of-
 /.cr. s'.a.'fir.^ requiremenis for to'lay
 v/orc Jc.rl to the ciscreucn  o;" ageaxry
 hecc'.s.
   Many ac^rxie* sub*er,uen'ly an-
 Eo-inc^d oi":"iccs wilJ CJOSG  toJay,
  The Jrtprr.i.1  Rever.ue Service paid
 *H  ;:.> r.j!~. .*. ;r four ;c'jTj":^5 — Los
 : '. , c!?s. £: n ?,irr.rrd:no. ^trcrs:dc
     C1: :~^L — •••I.i tc c^o*c.i io-':iy.
     • JiiS r..V,c-.s  in Southern. CaU-
    . . M .li rrrric.n  oncn,
  . - ..-o.   K.r.r.cl],  Social  Secuntv
            '. •*..•;!; be ci^-C'j.
           ."..ce t--oxc=r.;i.: in Lcs
                          .
  7.-j .-,.:~c!ct  city sr.d cojr.iy  of-
,.c:~-  v/::i  ;-c,r.;.n  cpcn,  hov.-cver,
 ;^l:cjnie:i for tr.ofe entires fiiid.
  ?":s;e  offices  *I.so viii  V-e  open.
  owever,  Cov.  Rta^an  ordered  »r.
                                     Smog  Emergency


                                     to  Close  Fedora


                                     Agsncies   Today

                                     Continued from Fint Pe;«
                                     LT..t!ciflate hah to use of state vehi-
                                     cles except for enerjcr.cies in ihe
                                     sraos-plajjcd pans of Southern Cal-
                                     ifornia,
                                       The  governor's order covers
                                     dou-mou-n Los An^cics. eastern Lit*
                                     An^c'.us Coun'.y  and Rivcitice ino
                                     Sen Bernardino Coar.uc5.
                                       Herb Camribcll, circttnr nf the
                                     zUtc Orfice of Ecu^cncy  Son-ices.
                                     aljsn  vrgcd  mo'.crUu in the Scuih-
                                     !s.nd to  cut  back on  nane^s-cr.uJi 1
                                     drivL":^,
                                       The  federal air epLsxie order was
                                     isr-ed  on the second cay  of or.c o'
                                     the wui-st sno-2 sjejcs of the year.
                                       Jn Lo5 Ar.gcle* Cour.'.y,  the ozone
                                     ctMST snared  u> .49 pans per nuUion.
                                     jurt short of  the ^0 level for a first
                                     tU^c alert,
                                       The  .40 rciriing was  recorded at
                                     the A r Po:!uiio3 Conirol  D'.stnci'a
                                     'Ez.'-l Son Gabriel Va'.lcy station.
                                       It  v.-ss the  i^cccr.J highest oi;cne
                                     reiriir^ fn  Los .nJi^clc-s COT my fo
                                     far this year. Tne piuk r.zone c^unt
                                     50 far v/as .0? pprn recorded in Ccn-
                                     iral Los Ange'.e» on June 21.
                                              Fii»l-SU*e Alerts
                                       First - !>ta;e  f^r.c^   alerts  were
                                     issue:]  in tc:h >±n  E.'rnsrciino and
                                     pjver>:.'> ccur.ues during  tne  day.
                                       Top  oxidant counts  included:
                                     UpL^d; .-VS pn.T.; Konuna. ..rj5; San
                                     Bcrnjrcino, .."iC,  snd Tdverside, .32.
                                       At  the San Bernardino County
                                     riiic.- of Upland, Forruna,  Ki^im
                                     and SOT  Bernardino, shcrifrj<
   Per=or.s  v/ho  hao  ^Jiherca  esn
 .-iricv, o!".:5 and  in  ph^ppm^ areas
 \vcre ur^cd n s'.ay 5n.-;cc a buiidinj
"or ih.c-fr hor.ics and  r.oL to use ihtir
 autoranbi'.cs.
   "I; v.-c.^ £ove:h:E5  like a \V~orIi
 War II £ir raid \varr.ir.5,"  a startlei
 resident o.r L'p^r.d E^id.
   The ^r.io; c:.ur/.sjn  San  BornaitJf-
 r.o County a!-o \vc-rc the second hi;h-
 e.-t  reiis'.cred ;his  yen.-. Th*  h:gh-
 oil  :o cjis \v5s .t6 p^m at  Up!snd
 on J'jr.e 6.
   At Riverside, v/hc-re srr.oz, alerts
 are  ca'.lcd v/hcn the oxiJan: co-ni
 reaches .27, thj C4:h elm of the >ter
 v/£s i-"jcd. Tht prcv;n;i«  day thre«
 »ep4ra:c alms \.xrc c^'^ed there.
           5-ciiofiI Warning
   In Lns A^eJcs. the A:r Po!!-tt»o
 Cor. 'ml Dl^ir.Cv i-;--Jcd the  itird c®i-
  •?cui:\'e school ind hcal'.h v.-arr.tB^
 for  ;!~e S.~r. >*crrirnn. S^n  Gabriel
 zr..1 Porr.oro-V.'alnui. valleys.    •
   Mrx:r.TJ-n r-iG^.e !e\v's io.'iav will
 rarj.-? oci^ccp. ,-i ) ',o .i'i p;i^*. in  ihfi
 Sen GLLivid nr.t^ Pomona vaP.cys,  the
 APCD r :d. Oz. r.c Ic-.c:? of between
 .10  to  .'J> sri- cv>'-c.tc'd in '.he ian
 >'crr.sr/J3 Vfll'.cy and a hi^h  reading
 nf ^!5 wis 'nrcr^rL for hem the otft-
 Lr?,l ar.ri s',i:,h.-.~>.t rficnori^.
   Roocrt i',. L-n-rie, ar'jng  a:r po!-
 lu.i^T  t.ir.Tji o-fjccr,  >ai^ lhal  4c^
                                                                            s^.og- " i-vp'.s j'rrn V.*cyr.c>c:iy's  r.f^h
                                                                            iht APCD  nifice  *-LS urcin^  lik*l
                                                                            aj.l lir.rcco.-saL-y  moLor vehicle Lrav-
                                                                            ti b« cjru:"cc today.
                                      Figure   A-4
Los  Angeles   Times  Report   on  Abatement  Action,  Appearing
                                    July   26,  1973
                                            A-40

-------
-So  Do  25,
   Federal Workers Get Day Off in Move to Cut Traffic  but
   Freeways Are Still Clogged; Supervisors Protest Holiday
                         BY FICHARD WEST
                             Tltnei StaH Wnltr
  An estimated  ?\000  federal em-
ployes in five  Southern California
counties were  given Thursday  off
because of an expeclPd sipgp of hea-
vy smog that never materialized.
  Thp undprstanrtinp was  that the
workprs \VPVP tr> stay home and kp*>p
 their cars off the road in an effort  to
rut  down the amount of pollutant-
in the air.
  Put  police  and  the California
 Highway Patrol said that despite the
fpdpral action, streets and freewa\3
were as crowded  as ever.  In fart,
they said, traffic was a little heavier
than usual in some areas.
  The Los Angelea County Boaid of
Snpenisnrs protested  to  President
 Nixon that the federal action was a
shocking  example of bureaucratic
 misuse of public powers.
  And  Los  Angeles  Mayor Tom
 Bradley said the rity has no  inten-
 tion of automatically following  the
lead of the federal government in or-
 dering  municipal  workers  to stay
 home during smog crises.
         First  of  Its Kind
  The action, the first of ils kind
 taken under the new federal Emer-
 ppnrv Air  Kpisode Plan,  involved
 HS.nOO federal workers in I he South
 Coa=t  Basin, which is  made  up  of
 Los Angcfps,  Ventura.  Orangp,
 Riverside and San Bernardino conn-
 tip.'.
   It closed down  all Inlprnal Rev-
 enue Service offices in this area PX-
 cept in Ventura County and all  So-
| cial Security Administration offices
Jin the Los Angeles Basin.
   In addition, it drastically curtailed
 operations  of  the  Department  of
 Hpslth. Education and  Welfare, the
 Department r-f  LaTor.  the  Sm-ii
i
Ru^tnPss Administration, 'he Feder-
al  Trade Commission and the Equal
Employment rommis=inn.
  Thp:e agencies opetal»d will- «nl"
s'kflpfon suffs:
  Goidnn  Kllmlt, chaitrrun  nf  thr
Federal F.vpru'ivp Board here, an-
nounced  late  Thursday  that   im-
pro\ ing air  pollution  cnndiiion-"
will permil  "all fedn-al agcnnF-; and
m-lallalions in the South Coa-i  R.i
sin to  rplurn to  their normal \\ork
schedules"  today.
  It  \\a-   Klliotl  \\hn  announced
Wednesday afternoon lhai.  ih« ppi-

 sode plan  would go into effect here
 Thursday  at the request of the re-
 gional office of the federal  Environ-
 mental Protection Agency in San
 Francisco.
   This ujs after  fir^t-stdgc  tiling
 alerts had been called  in ltivei.-.iiie
 rind  San   Bernardino counties and
 the smog level had reached .-10 parts
 of ozone per million pans  of air in
 the east San Gabriel Valley.
   The Air Pollution Control District
 lud  forecu.it heavy smog  ior >ome
 areas of  the  l.os  Angeles  l-Ja^iii
 Thursday.
   During  the night, however, hri.-k
 ocean breezes swept in and did away
 •,vith  the  meteorological conditions
 conducive to heavy smog.-The on-
 shore flow of marine air  also held
 down  temperature?,  resulting  in
 moderate, pleasant  weather  Thurs-
 day.
           Light Readings
   Peak smog readings were light to
 moderate  in  the  ba^in Thursday.
 They included .111 in central IMS An-
 geles. .20 in east San Fernando Val-
 ley, .17 in the Pomona-Walnut Val-
 ley and .23 in the east and  west sec-
 tions  of San Gabriel Valley.
   1-evels  forecast for today include
 2.~p in the inland  valleys and .15 in
 .p'ilral l^is Angeles.
   The pleasant weather lured
 f-oivus totaling an estimated -12J.OOO
 10 the beaches  of  Los  Angeles and
 Gr:ni-p counties.
   Aim'.-P'nerit parks such as DI-UPV-
 I.IMI:'. lion  Country i-afari.  Magic
 ^.•"i.nt.-iin aiul Knotl's  Berrv I'otiii
 ;r!K"»ned  no >igniiicant o>op in  ^t-
 -••:-.'j,-i e bee.iuie of ihc .-mug «:irn
men!-.
snd 1m-
putilic
Kenneth  llahn.   u ho  wrote  thf
protect  mouon  whu-h the board  rii-
ivi-ted to Mr. NJMin  and Russell K.
Train, ihjinnnn of the Council  on
Knvironmental  Quality, comment,-,!
  "Probably i,me out of 10 of the-o
federal employes work  m  air-mnd,
tioned offices. l'Jvcrvbod\ in  pri\;,ie
business has to work and pay taxes
to pay the  salaries of the-e  feileral
employes,"
  Ilahn's  motion  said.  ''A  nim-h
greater contribution toward  soli MI;
the problems of air pollution an Id
be made by the federal goiernmcr.t

  by vigorouslv pursuing ef-     •  I
  forts  toivanl control  of   ]i",i|
  »;tlo exhaust emissions."
     Bradley  said  tha  cily
  "xvould have to confer with
  the  APOD  first  on the
  health aspect before  tak-
  ing any action" similar to
  tha federal  government's
  episode  plan.
     But, he added, "1  don't
  know what the future will
  hold. It may get so bad ve
  will have to take drastic
  action."
     Elliott, ho\ve\er. defend-
  ed the federal action, say-
  ing the  "federal  family"
  use of  parking  lots  and
  cars decreased  considera-
  bly Thursday as workers
  took to car pools and pub-
  lic transportation.
     V spokesman for the Ra-
  pid Transit.  District  said
  Thursday afternoon It v.,is
  too early to come up i\uh
  ;be number of persons rid-
  ing buses during the da;-',
  but that there was no Indi-
  cation of any significant
  increase.
    At the downtown  Los
  Angeles  Federal  Office
  Building,  330 N. IMS An-
  geles St. about  2,800  of
  the 5.000  persons  who
  work there look the dav
  off. Almost all of the 2.300
  employes  at  the  U.S.
  Courthonip. :'.12 N.  Spring
  bl, reported to woik.
      Kmcrqcnry I'se
   Ciiy.  (uillllv and st,,!e
  ofhces were all  open, but
  (Jov.   fJeagau  ordered  a
  halt to the u^e of slate \P-
  hicles  except in emergen-
  cies.
   Elliott  said reaction  to  i
  future smog alerts could  j
  be  made more effective if
  all  levels of  government  |
  developed  mutual  goals  j
  and r,kir,~
      •e '-an pet all of our
       in  federal, S'^te,
       and  local  gnvprn-
          -s to strut u-in?
          -Mutilation .;nd
          ,-'\ cry  woi km1.; ,
day. ii"i. m-t riming MTIOC;
slens. v.e  will have t.ikeii
a grot ftnda lowrud abat- j
tus; the Mini'* problem,"  he
5.ini                      ;
  Klliotl pr.,i-cd a'l fcdor.il :
fluencies tor '*  'ving '-land- i
hy plans re.nly ami ljkin.2 !
'quick, posit u « action" io '
meet the smog alert.      i
  "Thou-ands  of  feileral i
f m p 1 o v e s u^ed  ]niblic :
1 rail-port.1110:1 instead  of j
their "wn automobiles and
lode in car pools," he said. .
  In  San    Bernard inn, I
API. D  meteorologist Koi>
Mankcr said  tha  use  o( '
sneakcr-eqinpiied helicon- ]
tcrs to warn j>eople of the
smog danger had resulted
in "a lot of confusion and
panic "
  ''We are trying to get t^e
radio and newspapers  to .
explain  that the world is '
not coming  to an end  or t
that we are in danger of
annihilation," he said.
  "It's just  a  warning to
the  elderly   people,  the
children  and  those with
le-piratory and heart con-
ditions to take it easy and
fclay in the house."
  The helicopter warnings
were broadcast  Tuesdai-
and AVennesday  but  not
Thursday.
  Inland  smog   readings
Thursday included .iO at
Kontana. .30 at Upland. .21
st San Bernardino, .2S at
Riverside  and .30 at Per-
lis.
  Comparatively moderate
temp eratures  prevailed
Thuadav, with highs ol S!)
degrees at I^os Anircles (,'h -
i-: 'Center, 9:i  at l;i\er~ide
and 08 at  San I'ernavdm".
  The cooling trend  will
ointinue,   the  forecaster
ta'd, wiih high  tempera-
 tures at Los Angeles Civjo
Center  today and Satur-
 cav in the low SOa.
                                                       Figure  A-5
                                          Los  Angeles  Times  Report  on
                              Abatement  Action,  Appearing  July  27,
                                                         1973
                                                             A-41

-------
        0
                                   o
                     r -J

  It was both wasteful and  ir-
responsible for  the federal
government to order the closing
of  most federal offices  in
Southern California  Thursday
because of predicted heavy smog"
that day.

  It was a public relations stunt,
of  course, to dramatize  the
seriousness of the smog problem
—  as though Southern Califor-
nians aren't aware of  that alrea-
dy. But it cost the taxpayers a lot
of  money  needlessly, and was
violative of the obligation of
government agencies to serve the
people.     .          -•. t/,-

  The bureaucrats in the  San
Francisco office of the Environ-
mental Protection Agency, where
the idea was dreamed up,  said
"closing the offices would mean
about ICO.CGO federal employes
wouldn't  have to drive  to  and
from work,— and thus wouldn't
contribute to" the predicted  high
smog levels.

 . • But anyone with any
 sense would kssw t&at
                                    vs, ^ n
                  pie, given an unexpected day off
                  with pay on a hot, smoggy day,
                  wouldn't stay home'if they could
                  help it.  They'd go to  the beaches
                  or  mountains.  And JJ they
                  couldn't escape  El  Monte, or
                  West Los Angeles, or  wherever
                  they might live, they  would likely
                  make a lunch date with someone
                  in an air-conditioned restaurant,
                  or go to the movies,  or run down
                  to the hardware store to get the
                  washer to fix the leaky faucet.

                    So it's highly  unlikely there
                  was any substantial  reduction in
                  federal" employe  car use Thurs-
                  day.
                            >

                    But while the estimated 100,000
                   affected federal  employes were
                   having their day off, they were
                  'still earning their $35 or $-10 a day.
                   So it probably cost the taxpayers
                   about $4 million in lost services of
                   the federal agencies  involved.

                     The punchline to all this is that
                   the predicted heavy  smog Thurs-
                   day didn't materialize.  Weather
                       £&$$& changed 'sufficiently to
                              Angeles basin.
            Figure  A-6
 Editorial  from Evening Outlook,
Appearing Saturday,  July 28, 1973
               A-42

-------
                     Clear  Warnings  on  Smog
  Five Southern California, comities have now ex-
perienced their first Federal Emergency Air Epi- -
socle Plan. Because nearly one-fourth of the 100,-
000 federal employes in the five counties were not
required to report for  work, there were several
1'nousund fewer autos on the highways. And there
'was  less smog than the day before, whep first-
stage smog alerts were called in some areas.
  Whether the Environmental Protection Agency's
plan to reduce smog by closing offices and curbing
activities was responsible for the drop in the ozone '
content is debatable. Robert  G. Lur.che, Los An- _
geles County air pollution control officer, says the,
Air Pollution Control .District had  informed EPA-
that there would be no emergency smog conditions
for two days in a  row. But the  drastic federal
action did perform a valuable service. It provided
a dramatic demonstration of what  the future
holds  if a continuing effort at  upgrading  our
air st-ancVdi'ds is not successful.
   Tho. federal  emergency  plan  was  activated
.Thursday when the ozone count  the day before
reached .49 pail per million in section-of  Los
Angeles, just short of the .50 count required for
a first-stage alert. But first-stage alerts were is-
sued in  Sa-n Bernardino, and Riverside counties,
where first-alert requirements  arc  lower. Even
Los  Angeles  County issued school  and health
warnings for  some areas. And  Gov. Reagan or- •
dered a halt to the use  of  slate vehicles in" the
South Coast Air Basin except for emergencies.
  Although Los Angeles  city and county govern-
ment offices remained open  and private industry
conducted business as  usual, the partial federal
shutdown should not be taken as a stunt. Eventual-..
ly, unless the dangerous air pollution is 'conquered,.>
all government agencies and  private industry may.,
be forced to shut when the  ozone count hits un-.
healthy heights.
  The federal response to our polluted air should
spell out yet another warning. With motor vehicles
responsible for 909o of contaminants in the air, we
must place harsh controls on the private auto and
develop alternative modes of transportation. Right
now, the only possible alternative lies in the-devel-
opment  of an. adequate mass transportation  sys-
tem.  '.•'•-'
                                     Figure  A-7
                        Editorial  from Los  Angeles Times,
                         Appearing  Friday,  July  27, 1973
                                          A-43

-------
be required as future abatement actions if air quality cannot be im-
proved sufficiently.   The editorial  explained that the action of the
24th represented a necessary and rational  response to the degradation
of air quality, and it pointed out the need to place more stringent
pollution controls on the private auto, and to develop alternative
menas of transportation.   The only possible alternative, the editorial
concludes, lies in the development of an adequate mass transportation
system.
     Based on contacts with  the major television networks,  it was
determined only KNXT issued  an editorial in response to the action of
the 26th.  The KNXT statement (Figure A-8) was sharply critical  of the
EPA emergency air episode plan, and  also warned that it is  only  a pre-
view of the ultimate smog plan of the EPA which calls for a shutdown
of the Basin by 1977.  The editorial  added further caution  that  the
EPA has the power to enforce their plans,  and that this agency cannot
be trusted with this authority.  It  was suggested that decisions for
shut down action during emergency air pollution episodes should  be made
by local and state authorities.
     The critical tone voiced by the media and the federal  agencies was
shared by the Los Angeles County Board of Supervisors (See  Figure A-9).
Supervisor Hahn objected to  the "holiday" given to a select number of
employees in the federal  government, while the taxpayers were required
to work.  Mayor Bradley voiced doubts of the plan, saying the City
would not arbitrarily follow the action of the federal government, and
that the decision to shut down city  offices would depend on consultations
with the County Air Pollution Control District regarding health  hazards.
                                   A-44

-------
Subject:  The EPA Smog Holiday

Broadcast:  August 1  and 2, 1973

The free holiday for Fedora! employees last week was not just a one-time publicity stunt
for the Environmental Protection Agency. It was just the beginning of a plan by EPA to
cail a holiday every time they decide smog is bad.

As you know, smog was heavy in a few areas of the basin last Wednesday. For a while,
weather men said it would be worse on Thursday, so the EPA requested that Federal
agencies close.

The fact  that the anticipated first-stage alert failed to occur on Thursday is unimportant.
The really critical thing is that under the Clean Air Act of 1970, known as the Muskie
Bill, the EPA aces have authority to c~ii for a shutdown of Federal agencies.  They can
do so whenever ozone levels reach that arbitrary IcveJ set by the EPA, which happens to
be 20% below the first alert stagejn Los Angeles County.

What happened Thursday v/as nothing compared  to the ultimate plan of EPA for the basin.
You tl.ir.k you're having trouble finding gasoline? Just wait until  the EPA tries to clo.se all
gasoline stations, and block freeways, and perhaps even to close some industry.

It can come to that because bureaucrats of the EPA have decided  on a smog plan which we
• tr.ink couid shut down the LOJ Angeles basin by  1977.  Their own director, the head of
EI'A, said the plan could ban all cars from the streets in Los Angeles by 1977.

Apparently a lot of people think the plan is just talk, but it's not. The EPA has the power
 anclit is  tneir pian. V.'hat happened  Thursday was just an example of their power, and it
should be a clear signal to Congress - and to Senator Muskie - that the EPA can't be trusted
WJth •nuijonry.  Y/iien ti^ci-JOi"5: !:f? n'.-jiK' 19 ^ill:P CiO^T: fQf vIYiCi'tiTick^ iH^e decision^
shouia be maue oy local uad ataio autnorities.
                                 Figure  A-8
                               KNXT Editorial
                      Broadcast August  1  and  2,  1973
                                      A-45

-------
   Los  Angeles County
Board of Supervisors yes-
terday objected to the day
off given to most local fed-
eral government  employes
because of high smog lev-
els.
  One   supervisor   called
the day off  shocking and
said  p u blic  employes
should not have a "Roman
holiday"  v/hen  the tax-
payers have to work.
  On a  motion by Super-
visor Kenneth Hahn (Sec-
ond  district),  the   Board
expressed its  disapproval
of the  s hutdown and
urged a day  off to  solve
the smog problem.
   Requested by EPA
   Hahn was  the nx>st vo-
cal critic of the shutdown
and held a  news confer-
ence after the Board meet-
ing to further expouse his
objectiqns.
  The shutdown was tak-
en at the request of the
regional office of  the Envi-
r o n m e n tal  Protection
Agency in  San  Francisco
to help  alleviate the high
smog  lefels expected yes-
terday.
  Included in the  shut-
down area were Los Ange-'.
les, Orange, Riverside and
San  Bernardino  Counties. :
  Hahn said that federal
employes with the day off
probably  went  to  the i
beach and used their cars'
more than  if  they  had'
gone to work.
  Asked if he would favor'
such  a  shutdown if the-'
smog situation  got worse'
Hahn replied, "Let  them
suffer with the rest of us.'1
    Bradley Comments   ;
  Later, Mayor  Torn  Brad--
ley was asked at an infor-
mal news conference if he-
thought city offices  would'
ever  be  shut  down be-
cause of smog.
  "We don't know  what
the future will hold," said
Bradley, adding that  pol-
lution might "get so bad"
that  this  could become
necessary..
  "But we-would not arbi-:'
trarily follow the action of/
the  federal  government,"
the Mayor said.
  Bradley said  he  woul..
not close city offices will;
out  consulting  with th.
County Air Pollution  Cor!
            abcut  healti'
                   Figure  A-9
   Report, on City Officials'  Reaction,
        Valley News  and  Green  Sheet,
       Appearing  Friday,  July  27,  1973
                     A-46

-------
     The tone of reaction demonstrated in national publications was
markedly different than that exemplified in the local press.  During
a communication with the Washington Post [A-4], it was conveyed that
national publications are somewhat removed from the  local problems of
the Los Angeles Basin, and consequently do not reflect an evaluative
tone.  The column in the Washington Post, shown as Figure A-10,
reflects this minimal degree of attitude presumption.  Several of the
national publications (such as the New York Times)were even less
committal by omitting coverage entirely.
                                 A-47

-------
                                                          THE WASHINGTON POST
   The  Federal  Diary
  While   Washington's   most
powerful  unregistered  lobby
                   ^ ^     ling  mixed  results in a  five-
                     '   'CT/d county area around Los Ange-
                   g         les.  which  has nearly 97.000
                             federal employees. There are
                             3-50,000 civil  servants in the
                             D.C. area.
                               Since Wednesday when L.A.
                             air pollution hit danger levels.
                             federal  agencies  have  been
                             trying to coordinate  an EPA
                                                         I   If  air  pollution  problenisj.tJiat1e:de^aT.^opet^tions
                                                         I continue, as seems likely,  the | hampered  By-reduced mongc,
                                                         1 EPA will  attempt  to  make; and problems of coordination
                                                          Washington  a  model  compli-: between inter-dependent agjn-
                                                         ' ance area, since government is! ties working different hours
                                                         ; the major industry here and a I   The Los Angeles  pollution
                                                         ! federal no-driving edict would ; alert plan, which may  be tie
                                                                                                  of
                                                          have more clout, and effect, in forerunner
                                                                            area
                             inspired plan of last May.  It
                             calls for voluntarv curtailment
 - the federal  carpool commu- .5rgovernment .induced venicle
 nity —  struggles for Parklng.^raffic
 spaces with the Environmen-,  ' As of  late  yesterday on]y
 tal  Proleclion  Agency,  Uncle . ,wo California agencies. Social
 Sam already has  been forced ! Security  Administration  and
 by  smog to shut down most ' ttle Internal  Revenue Service,
 operations in southern Califor-;had completely shut down.
 nja                          I   Post  offices, federal courts,
  _„.           ...         the U.S.  marshal's office  and
  EPA wants  to eliminate  or Los Ans,eies citv  and
 reduce cut-rate parking— 57.50 . offjces  \vere
 to S15 a  monlii-in or near • ske)elon crews working. ..]t is
 tederal agencies  here. Idea is bcin£;  run something like a
 lo force  more one-passenger snow7  emergency  here -  one
drivers inlo  c.rpools or  buses. \Vashington-based  official
Rep. Joel  T. Broyhill (R-\a). plained.
who  represents a major bed
                                                  county
                                          open, but  had
                                            a similar pfo-
jthe  metropolitan  area  than | gram  for  Washington,  v&s
i anywhere else.               • presented to the Federal Exec-
!   In trying  to  tackle federal' utive Board there  on May" 22.
 commuters, EPA will have Hs'FEBs  include   top   officials
 work cut out here.            ' from all federal offices within
   Despite  attempts by* Presi- given geographic areas.   <|
 dents  and the Budget Bureau   Working  on a  crash  ba|s,
'to  stagger  working hours  initheFEBgot agencies to  adcjpt
!Washington,  this  city,  morel air-pollution closing  plans.  «y
1 than most, retains its 9 a.m. to-June 25, less than a month  §e-
j5 p.m.  patterns.  Dependence!fore they were actually  calted
, on  the  automobile has  been'into play. The plan also  calls
 fostered  by  suburban  living for  the  shutdown  of fedejal
 patterns  and  inadequacy  01 parks  and beaches, and  elimi-
 mass transportation.           nation  of  all  but  essentjal
   In  private-industry   domi- travel  in  government-owned
' nated  towns, staggered  hours: cars. Such action in the Wain.
 are  standard.  If  a  factory ington   area   could   damage
 owner says everyone will be in,tourist-generated businesses?
                                                      ex-
                                     Employees who don't  at 7 a m. then that is the start-    Jobs:  Commerce   needs«i a
                             have to come in, or those with :ing time.
room commuter community, is  heallh problems are urged lo |
fighting the plan on behalf of :stay home» He said a ]iberjd,
 is constituents.              'annual leave policy, meaning ! with sizable protests from fed-! or 9  registered nurses,  GS? 9
  In  what could be  a  hand-:it could be granted  on-the-spot1
wriiing-on-the-wall    exercise,'rather than in advance, is in
the federal government is hav-! effect there.
                                                                                       Grade 8 secretary and  GSJ 5
                                                           Attempts to stagger working: clerk-typist. Call 967-3615. ^
                                                         'hours  here  always are  met I  D.C. Village  wants GS 5« 7
                                                         eral workers who,  as citizen-!physical-occupational   thcfa-
                                                         voters, call their congressmen. I pists a,nd  GS 12-14 medical #>£-
                                                         The congressmen in turn warnlficers. Call 629-8466.        j
                                                   Figure  A-10
                                          Column  from Washington
                                                Post,  July  1973
                                                      A-48

-------
REFERENCES - APPENDIX A
A-l  Federal employment total (118,000) from official  statements  of
     Gordon Elliot, Federal Executive Board, quoted in numerous  local
     newspapers July 26, 1973.

A-2  Private communication with Rapid Transit District, September 1973.

A-3  Private communication with Los Angeles Transportation  Department,
     August 1973.

A-4  Private communication with Washington Post staff, August 12, 1973.
                                   A-49

-------
                                 APPENDIX B

              THE PRELIMINARY EPA EPISODE CONTINGENCY PLAN FOR
                     THE METROPOLITAN LOS ANGELES AQCR


     EPA Region IX is presently in the process of constructing an episode con-

tingency plan for the Metropolitan Los Angeles AQCR.  Preliminary versions of

episode criteria and control strategies have been formulated; these are

summarized in Tables B-l and B-2, respectively.

     This appendix performs two separate analyses of the preliminary EPA plan.

First, in section B.I, the effectiveness of the proposed EPA control strategies

is computed.   VMT reductions are determined for the traffic abatement measures,

both without considering substitute driving and with accounting for substitute

driving and overlapping reductions from the various measures.  Using the

results of section 2.2.4, the VMT reductions are translated into percentage

emission reductions.  Expected emission reductions are also estimated for

stationary source controls.

     Second,  in section B.2, the EPA plan is evaluated in comparison to the

contingency plan proposed in this study, (Chapter 6).  It is noted that the

two plans are very similar overall.  The only major differences are that

the EPA plan  contains three episode stages (as opposed to four in the plan

proposed here) and that two control measures in the intermediate control

stage do not  coincide.  The relative advantages and disadvantages of each

plan are discussed, leading to the conclusion that neither plan can be

unequivocably recommended over the other.  The choice of which plan is

preferable depends on how the relative advantages of each are weighted.
                                    B-l

-------
     Table B-l    PRELIMINARY  EPA CONTINGENCY  PLAN  FOR
                   LOS  ANGELES  - EPISODE  CRITERIA
a For Present Significant Harm Lev^
Recorded Oxidant Level
.30 PPM-1 hour Avg.
or
.35 PPM-instantaneous
.35 PPM-1 hour Avg.
.40 PPM-1 hour Avg.
or
.50 PPM-instantaneous
.40 PPM-1 hour Avg.
.43 PPM-1 hour Avg.
or
1.00 PPM-instantaneous
Meteorological
Forecast for Next Day
Worsening
Worsening
Same
Worsening
Same
Is
Control Stage to be
Called on Next Day
I
II
III
b.    For Significant Harm  Level  of  .60  PPM-1  Hour  Average
Recorded Oxidant
Level (1 Hour Average)
.45 PPM
.50 PPM
.55 PPM
.55 PPM
.60 PPM
Meteorological
Forecast for Next Day
Worsening
Worsening
Same
Worsening
Same
Control Stage to be
Called on Next Day
I
II
III
                          B-2

-------
TABLE  B-2    PRELIMINARY  EPA  CONTINGENCY  PLAN  FOR  LOS ANGELES-CONTROL  STRATEGIES
          1)   Press release is  issued  by  EPA,  Region  IX, containing:

               a)  Description  of  the  air quality situation
               b)  Health warning
               c)  Request for  car-pooling
               d)  Request for  the elimination  of all unnecessary use of gasoline powered  devices
                   (e.g.  lawnmowers)
               e)  Request for  the elimination  of all unnecessary use of motor vehicles  by  businesses and
                   by the general  public
               f)  Request for  the use of public transportation wherever possible
               g)  Request for  elimination of  open burning
               h)  Statement that  cooperation will help prevent the necessity of EPA  taking more stringent
                   control measures to abate  the episode

          2)   Voluntary Co-operative elimination of non-critical  business  travel  of Federal, State and Local
              government  operations and commercial  businesses.
     Stage II

          1)   All Stage  I actions

          2)* Mandatory  closing of admission  to all public and private recreational areas  (e.g. beaches,
              parks, amusement  parks,  theaters, clubs, and bowling alleys).
                Those parks which  allow camping will  not admit additional  campers,  although those which
              were previously established at  camping  sites will be allowed free access.

          3)* Mandatory  cancellation  of all  recreational events (e.g. ball games, boxing and wrestling matches,
              track meets, automobile  and boat  races, other spectator sports, concerts and  plays.)

          4)* Mandatory  termination of use of  non-critical fleet vehicles.  A "fleet"  is ten or more vehicles
              with commercial plates  and  under  common ownership.   Ban on the renting  of  cars and trucks.
              Certain critical  activities (and  one innocuous activity) will  be exempted.   They will include
              those listed below among others:

               a)  Use of emergency vehicles  (e.g. ambulances, fire and police vehicles, and any vehicle
                   used  in a medical,  fire or  police  emergency)
               b)  Normal operation of limousines and mass transit vehicles  (except when used for sightseeing
                   purposes, which use is prohibited)
               c)  Deliveries of gasoline to maintain essential services (e.g. police  stations, fire stations
                   and ambulance services)
               d)  Deliveries of perishable goods and foodstuffs and essential drugs  and medical supplies
               e)  Use of vehicles by  public  utilities for emergency repairs and services  (e.g. gas leaks
                   and power failures)
               f)  Use of vehicles for purposes involving the national security, where such use is necessary
                   and cannot in good  prudence  be delayed
               g)  Operation of vehicles  powered by gaseous fuels (e.g. liquified petroleum gas)

          5)* Mandatory  closure of all non-critical activities of Federal, State, and  local government
              operation,  including administrative leave for non-essential  personnel.

          6)* AIT businesses with  more than 100 employees are to reduce their operations drastically so as to
              reduce substantially the emissions of air pollutants which they generate  (directly and indirectly).

               a)  Examples of  drastic reductions in  operations include:
                   1)  elimination of  all production  activities (whether of  a manufacturing or of a service
                       nature)  that result directly in emissions of "reactive hydrocarbons" (as defined by
                       EPA).  Examples: petroleum, gasoline refining and transfer (except  by retailer to
                       customer),  dry  cleaning, painting,
                  11)  other actual production  activities may continue but all peripheral  activities should
                       cease: such "peripheral  activities" would include:
                       (1) routine clerical  activities
                      (ii) marketing  activities
                     (1ii) loading, transfer  and transport except of "critical" or highly  perishable commodities
                           such as:  foodstuffs, medicine, fuels for domestic consumption  or for power  generation,
                           etc.
               b)  Exemptions from the above  requirements will be granted to certain  businesses and agencies
                   whose  activities are of "critical" nature such as:
                   1)  normal police  and  fire  services
                  ii)  normal operation of ambulance  services
                 ill)  normal sales of drugs  and medical aids by pharmacies
                  iv)  normal operations  of medical aid facilties (hospitals, clinics,  etc., including  such
                       facilities  for  animals)
                   v)  normal sales of foodstuffs by  foodstores
                  vi)  minimal  necessary  operating, maintenance and repair activities of public utilities  (power,
                       water, communications)
                 vii)  normal operations  of mass media  (newspapers, television, radio,  magazine)
                viii)  critical operations of  those businesses or portions thereof directly involved  in vehicle
                       use exempted under item 4 and/or with responsibility for deliveries necessary  for  such  use.

     Stage III.
          1)  All Stage  I actions

          2)* All Stage  II actions

          3)* Cjose all   non-critical  businesses.
      *  Mandatory action
                                                       B-3

-------
B.I.  EFFECTIVENESS EVALUATION OF THE PRELIMINARY EPA CONTINGENCY PLAN



      This section determines the effectiveness of the contingency plan con-



trol strategies proposed by EPA Region IX.  The analysis draws heavily on



basic data presented in the main text of this report.  Sections 2.2.4 and



2.2.5 provide the emission inventory and transportation data base used in



the effectiveness calculations.   Section 2.2.1 gives some of the basic



employment information.  Tables  and results from these sections will be



referred to throughout the following analysis.



     The effectiveness of some traffic control measures in the EPA plan



has been determined elsewhere in this report (Chapter 5 and Appendix).



At the risk of repetition, these calculations are repeated here to keep



the discussion of the EPA plan intact.



     Table B-3 presents the results of the effectiveness evaluation for



the EPA plan.   Table B-3a is  for RHC control; Table B-3b applies  to CO and



NO^.  The first column lists  the control  measures,  organized into the



appropriate episode stages.   More comprehensive descriptions of the control



measures are given in Table B-2.  The second column of Table B-3  gives



incremental  VMT reduction including the effect of substitute driving and



overlapping reductions  from other control  measures.   The number in paren-



theses  does not include the substitute driving and  overlap effects.   The



third column gives the incremental  emission reduction for each control



measure, broken down according to stationary and vehicular sources.   Emission



reductions from auto tank filling at service stations are included in the



vehicular category.   Finally,  the last two columns  give cumulative VMT



and emission reductions for each stage of the control plan.
                                    B-4

-------
                                          Table  B-3a

                            EFFECTIVENESS  OF THE PROPOSED  EPA  PLAN:
                          VMT REDUCTIONS AND RHC EMISSION  REDUCTIONS
KHWUfCmOH
Stage
I
BB^BIGE^^BES
II
III
Control Strategy
1. Press Release
2. Voluntary Reduction of
Vehicle Use & Emissions
1 . Stage I Actions
2. Close Recreational Areas
3. Close Recreational Events
4. Ban (Private) Non-Critical
Fleet Vehicles
5. Close Non-Critical
Government Agencies
6. Reduce Direct & Indirect
Emissions from Private
Businesses with more than
100 Employees
1 . Stage I Actions
2. Stage II Actions
3. Close All Non-Crittcal
Businesses
Incremental VMT
Reduction
negl .
negl .
7%%-LDMV
5%-LDMV
30%-HDMV
5%-LDMV***
(6%*)
3%-LDMV
(4%%-LDMV*)
(9%-HDMV*)
negl .
34%%- HDMV
54^%-LDMV**
40%%-HDMV**
Incremental RHC i Cumulative
Emission Reduction 1 VMT
Veh. I Stat.i Total i
negl .
negl .
4%%
6%
3%
2%%
negl .
16%
38%**
1
1
negl . negl .
negl . negl .
i
r • —
-- ' 4%%
1
4% ' 10%
I
-- | 3%
1
2% ' 4%%
1
1
I
1 £
negl . negl .
i P
i |
6% 1 22%
i
negl .
20^-LDMV
34%%-HDMV
i
1 I 75%-LDMV
4%** 1 42%** i &
i 1 HDMV
aaiaga'BcssasBaEaiBiaa
Reductions
RHC
Emissions
negl .
22%
64%
^?!^^^S3^a^^3L^^S^ill
*   Not including substitute driving and overlap with other control  measures.

**  In this case, cumulative effect was calculated and then incremental  effect
    was determined by subtracting Stage II results.
*** Add 1% if public school  closure is'to'be included.

-------
                                              Table  B-3b


                                EFFECTIVENESS OF THE PROPOSED EPA PLAN:

                           VMT  REDUCTIONS AND CO AND NO  EMISSION REDUCTIONS
                                                      X
r»~i rffl <'M- fla«h»ji.
Stage
I
II
^BiBJBHEi^S^
Ill

Control Strategy
1. Press Release
2. Voluntary Reduction of
Vehicle Use & Emissions
1. Stage I Actions
2. Close Recreational Areas
3. Close Recreational Events
4. Ban (Private) Non-Critical
Fleet Vehicles
5. Close Non-Critical
Government Agencies
6. Reduce Direct &
Indirect Emissions from Priva
Businesses w/more than 100
Employees
1. Stage I Actions
2. Stage II Actions
3. Close All Non-Critical
Businesses
^^a.,Wl^; LXJt.M J s .it.^ dkmLaiJaU'AflA.g
Incremental VMT
Reduction
negl .
negl .
negl .
7%-LDMV
5%-LDMV
30%-HDMV
5%-LDMV***
(7%%*)
3%-LDMV
te 4%%-HDMV
(4%°/-LDMV*)
(9%-HDMV*)
negl .
20%%-LDHV
34%%-HDMV
54ia%-LDMV**
40%%-HDMV**
li5S,&ma£Eakis&Z}iC
Increment
Redu
CO
negl .
negl .
negl.
6%%
7%
4%
3%
negl .
2Qh%
49^%**
'ife JYl^ftSSfe-', ;*;?tej«3SiaaaB£
al Emission
:tion 1
NO
E
negl.
negl.
\
negl .
4%%
6%
3%
2%%
negl .
16%
36*2% **
Cum
VMT
negl.
negl.

20%%- L
34%% H
75% LD
&
HD
^52£^£2S
jlative Redi
CO
Emissions
negl .
negl .
DMV
20^%
DMV
MV 70%
MV
s-ssssEisssRSsssarasra
^y^"l?."^^Ta.- TF^^h^SlJl
ctions
NOX
Emissions
negl .
negl .
m at!Aia.iai^!*aaii«n!j«.
16%
52%
^uaaisagm^ffjunf
**
***
Not including substitute driving and overlap with other control measures.

In this case, cumulative effect was calculated  and  then incremental effect

was determined  by  subtracting  Stage II  results.

Add 1% if public school closure is to HP -in^i.iri^

-------
     Table B-4 indicates the severity of the episodes that can be handled
by the various stages of the EPA plan.  For oxidant, the pollutant of most
interest, it is seen that a .45 PPM (4 hr. avg.) oxidant episode can be
reduced to below the significant harm level by Stage II.  Stage III can handle
episodes which would produce oxidant as high as .59 PPM (4 hr. avg.) with
no control.  If the oxidant harm level is changed to .60 PPM (1 hr. avg.),
Stage II can control episodes up to .68 PPM (1 hr.), while Stage III can
abate episodes up to .88 PPM (1 hr.).   Rather severe potential excesses of the CO and
N02 harm levels can also be prevented from reaching harm levels.  Of course,
as noted in Section 2.2.5, CO and N0? episodes are extremely unlikely to occur
ever in this decade.
     The effectiveness analysis for each of the proposed EPA control measures
 follows:
 STAGE I
 1.  Press Release
 2.  Voluntary Reduction of Vehicle Use and Emissions
       Controls based solely on voluntary cooperation are  likely to have
     negligible effect on VMT and on stationary source emissions.   The
     purpose of Stage I is basically to provide certain prerequisites for
     the later control  stages,  (health warnings, media announcements, pub-
     lic relations,  etc.)
       Past experience  in Los  Angeles  and San Francisco indicates that
     essentially no  discernable VMT reductions results from request for
     voluntary traffic  abatement, [B-3], [B-5].   However, in the future,
     with a more extensive and  better  organized media campaign, some VMT
     changes might be possible.   It is difficult to estimate just how much
                                      B-7

-------
                                                       Table B-4
                                     EPISODE LEVELS  WHICH  CAN  BE  REDUCED TO OR BELOW
                                   SIGNIFICANT HARM  LEVELS BY THE EPA CONTROL STRATEGIES

Stage
I
II
III
eaaoKnaatuL aai
1
EMISSION REDUCTION 1
E
RHC
negl .
22%%
64%
CO
negl .
21%%
70%
NO
x I
1
negl . I
1
15%% 1
i
52% 1
EPISODE LEVELS WHICH CAN BE ABATED* (PPM)
Ox id
.40-4 hr.
.40
.45
.59
tevatssia&u&imWiSigaiffw*
ant**
.60-1 hr.
.60
.68
.88
auaagiMfcaiiiB^ toUsfeaxsKSBSiisats
CO**
.50-8 hrs.
50
58
94
:«^:!TO!K5*^liV,.;%!t'At«iS*Jui;.'-J!k>ia!i,
NO?.**
.50-24 hrs.
.50
.56
.77
;jA'J»!s»s,«MS*.i*fe«'«S!iiWSESSJra!
CO
I
03
      **
These results are based on the approximate air quality-emission level  relationship developed in
Section 2.2.7.  Oxidant concentrations are assumed to be reduced by 1/2 of the amount of RHC
emission reduction on the day of the episode.  CO and N0? concentrations are assumed to be
reduced by 2/3 of the amount of CO  and NO  emission reductions on the day of the episode.
                                          y\
As noted in Section 2.2.5, the .40 PPM - 4 hr. oxidant level, the 50 PPM - 8 hr.  CO level, and
the.50 PPM - 24 hr. .NO^ level appear to be the "binding constraint" significant harm levels
for each of the three pollutants.  (There are three oxidant and CO harm levels and two NO-
harm levels.)  The .60 PPM - 1 hr. oxidant level  has been proposed as an alternative to tne
.40 PPM - 4 hr. harm level.

-------
    traffic reduction might be attained.  It is conjectured here that the

    maximum possible effectiveness of voluntary control would be 5%.

      Since VMT reductions from voluntary cooperation and very doubtful,

    no actual  reductions will be claimed for Stage I  in the effectiveness

    evaluation for the whole plan.
      STAGE I:

      Total VMT Reduction:                   Negligible to 5%.
                                             Most likely negligible.
                                             Applicable to LDMV's.

      Emission  Reduction:                    RHC      CO       NO
                                                                 /^
                Vehicular Sources           negl.    negl.    negl.
                Stationary Sources          negl.    negl.    negl.
STAGE II

1.   Stage I Actions

2.   Close Public and Private Recreation Areas

3.   Cancel  All  Recreational  Events

      Control  measures 2.  and 3.  of Stage II  will  be combined  in  the  effective-

    ness evaluation; total VMT reduction will be estimated for closing all

    recreational facilities.  It is much more difficult Jto estimate the

    effectiveness of closing "recreational  areas"  and "recreational events"

    separately.

      According  to the LARTS survey data described in Section  2.2.6,  (modified

    according  to recent corrections), [B-7],  recreation accounts for 7.2% of

    total weekday vehicle  trips in the Los  Angeles Basin, [B-3].   The indepen-

    dent results of an SDC survey, [B-9], indicate that around 11% of total

    weekday trips are for  recreational purposes.  The basic reason for this
                                     B-9

-------
discrepancy between the two surveys involves the definition of "recreation".
LARTS data distinguished between "recreation" and "social-entertainment"
trips, while the SDC study includes both under "recreation".   Thus,
the difference between SDC and LARTS estimates reflects "social-entertain-
ment" trips.
  For the purpose of evaluating measures 2.  and 3.  of Stage II, "recreation"
should include "entertainment" since theaters, playhouses,  clubs,  etc.
are to be closed.  However, most "social"  trips should not  be included.
Assuming around one-half of "social-entertainment"  trips  are  for  enter-
tainment," 9% of total weekday trips are accounted  for by the appropriate
definition of "recreation".
  Recreational trips basically occur in  the  "Home-Other"  and  "Other-Other"
LARTS categories, [B-3].  The average trip length in these  categories is
about 6-6% miles, (See Table 2.13).   However, it is expected  that re-
creational trips would be longer than average trips in the  "Home-Other"
and "Other-Other" categories.   Recreational  trips to the  beach, sporting
events, amusement parks, fairs, etc. would probably average at least as
long as "Home-Work" type trips, (10 miles).   This effect would only be
partially cancelled by recreational  trips  to clubs, bowling alleys,
theaters, etc., which would probably average about the same as "Home-Shop"
type trips, (3 miles).  It is arbitrarily  assumed here that the average
length of a recreational trip is 7 miles.
  With total recreational trips representing 9% of average  weekday trips
and having an average length of 7 miles, the maximum possible effect of
control measures 2.  and 3.  would be
                                   B-10

-------
      100% X 100% x 6.71miles (average length of all  trips  =  ™  of weekday  VMT
                                        in basin)
    However, it is extremely unlikely that all  recreational trips  will  be
    cancelled by the regulations.   Certain facilities  may  not be  covered,
    and  "pleasure driving" may still  lead to some recreational  traffic.
    It is  assumed here that adequate  implementation and  enforcement will
    eliminate 80% of total recreational  VMT.   This  means that
    .80  x  9%% - 7%% of total weekday  VMT will be affected.
    Total  VMT Reduction:      7%%
                             Applicable to LDMV's.
    Emission Reduction:             RHC      CO      NO
                                                     /\
             Vehicular Sources     4%       6%%     4%%
             Stationary Sources    —      —     —
4.   Ban (Private) Non-Critical  Fleet Vehicles
      For the purpose of making effectiveness  calculations,  the  fleet  vehicle
    ban is assumed to apply to  private (as  opposed  to governmental)  fleets
    only.   Elimination of non-essential  public fleet  traffic occurs  as  part
    of control  measure 5., closure of non-critical  governmental  agencies.
      Presently,  430,000 vehicles  are contained in  private  fleets  of ten or
    more operating in the Metropolitan Los  Angeles  AQCR,  [B-l],  [B-ll].  It  is
    not known how many of these are in categories  listed  as  exempt by  the EPA,
    (emergency  vehicles, mass transit vehicles, essential and emergency
    services, perishable deliveries, and gaseous fueled vehicles).   Here,
                                      B-ll

-------
    it  is  arbitrarily assumed that 1/3 of the vehicles are exempt.  Thus,
    290,000  vehicles are affected by control measure 4.
     Assuming  that the HDMV-LDMV ratio is 20%-80%  in  these  fleets,  [B-10], then  58,000
    HDMV's and  232,000 LDMV's are contained in non-exempt fleets.  These
    represent 25%  and 4% of  total basinwide HDMV's  and LDMV's, respectively.
    Since  fleet vehicles tend to be used more intensively than non-fleet
    vehicles,  (about 25% more,  [B-2],  [B-10]), the  affected  vehicles  will  account
    for around  30% of HDMV VMT  and around 5% of  LDMV VMT.
      It is  assumed that the fleet vehicle ban will also eliminate the
    stationary  source RHC emissions resulting from  the filling of
    underground service station tanks.  This service presently accounts for
    4%  of  total  RHC emissions in the Los Angeles Basin,  (see Table 2.5).
    It  should be noted that  over the next few years, vapor recycle systems
    will be  installed to control emissions from  underground  tanks.  These
    systems  will eliminate most emissions from this source.   Shut-down
    during episode will then  produce practically no further  effect.
Total VMT

Emission



Reduction:

Reduction:

Vehicular Sources
Stationary Sources


RHC

6%
4%
5% of
30% of
CO

7%

LDMV
HDMV




VMT
VMT
NO

6%




X


5.   Closure  of Non-Critical  Government Agencies
      Government agencies  account  for  13%  of  total  employment in  the Los
    Angeles  Basin,  (see Table 2.3).    Assuming  that  90% of these employees
    will  be  classified  as  non-essential  and noting  that work related VMT
                                          B-12

-------
accounts for 50% of total weekday VMT,  (Section 2.2.6), around
.13 x .9 x 50% = 6% of total VMT will be eliminated by government
agency shutdown.
    As noted in Section 2.6, a decrease in total VMT by "X" (by less
work related driving) normally brings about a "1/5X" increase in other
types of driving so that the total VMT  reduction is only "4/5X".  Thus,
allowing for substitute driving, the total VMT reduction from closing
government agencies would be 4/5 x 6 =  5_% of total  VMT.
    It may be that public schools would also be closed as a result of
control measure 5.  Estimates of trips  related to all schools vary from
around 1% to as high as 3% of total, basinwide trips.  The 1% figure
results from an SDC survey,  [B-9], but  it may be low since the SDC
survey is biased toward working members of the population.  The 3%
result comes from a LARTS report, [B-3].  Here it is assumed that 2%
of total trips are related to attendance at public schools.  Since
many school related trips may be very short, (e.g.  trips to local
elementary and secondary schools), school type trips will probably
average less in length than total basinwide trips.   Thus, it is
assumed that only ~\%% of total basinwide VMT results from travel
to public schools.
    Adding the VMT reductions from the  closure of government agencies
and schools, the total VMT reduction from control measure 5. is 7h%_
without considering substitute driving.  Using the 4/5 factor to
account for substitute driving would reduce this to 6%.  It is
interesting to note that this estimate  agrees almost exactly with
traffic reductions detected by the Los Angeles Freeway Surveillance
and Control Project during the Truman and Johnson Memorials, [B-5].
                                    B-13

-------
     Total VMT  Reduction:              5%*
                                      Applicable to LDMV s

                                      (6%, not considering
                                       substitute driving)


     Emission Reduction:               RHC      CO      NOX

             Vehicular  Sources         3%      4%      3%

             Stationary Sources       ---      --      -"
     * Add 1% if  public  school closure is to be included
6.   Reduce Direct  and  Indirect  Emissions from Private  Businesses with More

    Than 100 Employees


      It is very difficult  to perform  a definitive  quantification  of  the


    effectiveness  of control measure 6.  Certain  phrases  in  the  regulation,


    such as  "reduce  substantially" and "peripheral activities", may be


    subject to somewhat wide interpretation.  No detailed information is


    available on which specific employees and which specific operations


    of the firms will be affected.  Below, estimates of vehicular  and


    stationary source emission  reductions will be derived by using certain


    simplifying  assumptions.


    Vehicular Sources:


      Private firms  with more than 100 employees account  for around 2B%


    of total employment in  the  Los Angeles Basin,  [B-5].  As noted in


    Section  2.6, around 50% of  average weekday VMT  is work related.


    Based on an  examination of  LARTS and SDC data,  it appears  that


    around 38% is  "commuter" type travel (home-work, work-home, work-shop,


    etc.) while  about 12% is "on the job" or "business" travel,  [B-3],  [B-9]


    Thus, commuter type travel  to firms of more than 100  employees should


    account for  around .29  x 38% = 11% of total VMT, while business related
                                      B-14

-------
travel in these firms should be around .29 x 12% = 3%% of total  VMT.



  The amount of commuter travel eliminated by control measure 6.



will depend basically on the number of employees who stay home due to the



measure.   Carpooling would also decrease commuter travel, but mandatory



carpooling is difficult to implement and voluntary carpooling is  not  reliable



The number of employees who will not report to work as a result  of



the control measure is not known; it will depend on the working  definitions



of "production" and "peripheral activities".  It is expected that most



firms would attempt to minimize the number of employees scheduled for



leave, so that the number who do stay home will  depend on how strictly



 the  stipulations  of  the measure are  enforced.   It  is  assumed here that



 a  program  can  be  implemented wherein % of  the employees  in  the affected



 firms will  be  on  leave.



   Control  measure 6.  is intended to  affect  "business  related" travel



 more  than  "commuter"  travel.   Marketing  and most transport  activities



 will  supposedly be cancelled.   The amount  of business  related travel



 that  is  actually  eliminated will depend  on  the working definition of



 "critical  transport";  it also  will be sensitive to the amount of



 enforcement attached  to the measure.  It is assumed  here that % of



 the business travel  in firms of more than  100 employees will be



 eliminated.



   Without  accounting  for substitute  driving and overlapping reductions



 of other control  measures measure 6 will produce the  following VMT



 reduction:
                                   B-15

-------
  % of the employees in         % x 11% = 2-3/4% of LDMV VMT
  affected firms on leave


  % of business related         1%% of LDMV VMT*
  travel eliminated             9% of HDMV VMT*


  However, as noted in Section 2.6, substitute driving should decrease


the "employee leave" VMT reduction by about 1/5, so that it would be


only 4/5 x 2-3/4% = 2%% of total  VMT.  Also, the fleet vehicle ban would


overlap with the business related travel  restrictions of measure 6.


The fleet vehicle ban apparently  eliminates about % of the total

business related travel in the basin.  Thus, about %  of  the  reductions


in business travel  claimed for measure 6. would already be achieved by

measure 4.   This reduces the effectiveness of measure 6.  to


h x 1%% = 3/4% of LDMV VMT and %  x 9% = 4%% of HDMV VMT.   Thus,


accounting for substitute driving and overlapping effects from control


measure 4,  the VMT  reductions expected from measure 6. are


  e  3% of LDMV VMT and


  •  4%% of HDMV VMT.


  The VMT reductions can be translated into percentage emission


reductions  by using Table 2.20, (note that the effect on emissions


from auto tank filling at service stations is to be included in the


vehicular source category).  The  results  are


  •  RHC Emission Reduction:   2%%


  e  CO Emission Reduction:    3%


  •  NO  Emission Reduction:   2%%
       f\
* The VMT reduction for "business related" travel is broken down into
  LDMV's and HDMV's as in control measure 4.  for fleet vehicles.
                                   B-16

-------
Stationary Sources:



  The portion of measure 6. pertaining to stationary source control



is rather ambiguous.  In fact, a self-contradiction appears to exist



since the measure applies only to firms with more than 100 employees,



yet "dry cleaning"  is included as one  of the industries to be



controlled.  Even the largest dry cleaning companies tend to employ



well less than 100 employees, [B-8].



  Another problem may result from the inclusion of petroleum



refineries in the control measure.   Refinery operations are difficult



to shut-down, and the shut-down process itself may actually lead to



extra emissions.  Because of such problems, a review committee, with



power to grant variances, should be established before a,definitive



Itst of affected industries is promulgated.



   Control measure  6., through control of marketing transport, will  supposedly



eliminate the transfer of gasoline  to  service stations.  This would  amount to



a 4% reduction in RHC emissions, (see Table 2.5).  The other sources to be



controlled have not been well defined  so that expected emission reductions



can only be estimated.   It is assumed here that the impact on other



sources will be the same as shutting down all "major point sources"  of reac-



tive hydrocarbons.  The effectiveness of this measure, evaluated in  Section 5.2,



is estimated to be 2% of total RHC  emissions.  Thus, the total reduction in



RHC emissions associated with measure  6. is 6%.



   It should be noted that the 4% reduction in RHC emissions from



eliminating the filling of underground service station tanks is



also attained by control  measure 4., ("Ban Private Fleet Vehicles").
                                    B-17

-------
    Thus,  allowing  for overlap  of other control  measures,  the stationary

    source effectiveness  of control  measure  6.  only  amounts  to 2% of

    total  RHC emissions.
    Total VMT Reduction:
3% of LDMV VMT
4%% of HDMV VMT
(4%% LDMV and 9% HDMV,
 not counting substitute
 driving and overlap of
 control measure 4.)
    Emission Reduction:
             Vehicular Sources
             Stationary Sources
RHC

2%%

2%
CO

3%
                                                   N0
STAGE III

1.  All  Stage I  Actions  (denoted  as  III.l below)

2.  All  Stage II  Actions (denoted as III.2 below)

3.  Close All  Non-Critical  Business  (denoted as  III.3 below)

      Considerable  overlap  exists between Stage  II, (measure 2. of Stage  III),

    and  measure  3.  of  Stage III.  In fact, measures 4., 5., and 6. of  Stage  II

    are  somewhat  redundant  when included as part of Stage  III.  Much of

    measure  II.4.,  banning  fleet  vehicles, would be a result of measure

    III.3.   Measure II.5.,  closing government agencies, also seems to  be

    a natural  part  of  measure  III.3.  Measure II.6., commercial abatement

    plans,  is  superfluous if measure III.3. is instituted.*  The  only  con-

    trol  measure  of Stage II that is not part of measure  III.3. is the
    * A possible exception is  that  measure  II.6.  may  shut-down certain
      stationary RHC sources  that might not be closed under measure
      III.3.
                                     B-18

-------
closure of recreational facilities.  In order to eliminate this re-

dundancy, Stage III could alternately be defined as

  1.  All Stage I Actions  (denoted  as 11 I.I  below)
   /
  2.  Close All Non-Critical Business (Work Holiday), Close
      Recreational Facilities, and  Shut-Down Major Stationary
      RHC Sources  (denoted  as  II\ .?' hP1nW)

This definition would have  a nearly identical effect as the original

Stage III, but it would be  operationally simpler since most of

State II measures would not have to be implemented separately.

  The new definition also happens to correspond with an easier way

of calculating the effectiveness of Stage III.   Instead of computing

the incremental effect of measure III.3.  and adding this  to the effect

of measure III.2,  (Stage II), the total,  cumulative effect of Stage III

(as embodied in measure III.2 ) can be  calculated at once.   This effective

ness calculation is performed below.

  Stage III  would drastically reduce incentives  to drive  by closing

all government agencies and industry except for "essential  employees",

all trade and other commercial  businesses  except for "essential trade

and services", and all recreational facilities.   The only traffic re-

maining would be from operating emergency vehicles, commuting of

essential employees, travelling for "essential  trade or services,"

and driving for recreational and social  purposes other than those

related to actual  recreational  facilities.  It is estimated that

this strategy would achieve around a 75% reduction in total weekday

VMT, [B-5],  [B-6].
                                    B-19

-------
    Stage III would also significantly affect stationary source  emissions.
  A 4% reduction in total RHC emissions would result from cancelling  under-
  ground service station tank filling.  Eliminating dry cleaning,  in-
  dustrial and commercial painting, and certain other business uses of
  organic solvents would reduce RHC emissions by around 6%,  (see
  Table 2.5).
    The cumulative effect of Stage III would thus be as follows:
     STAGE  III
     Total  VMT Reduction:              75%
                                      Applicable to LDMV's and HDMV's
     Emission Reduction                RHC      CO      NOY
                                                         /\
             Vehicular Sources        54%      70%     52%
             Stationary Sources       10%      —     —
B.2. COMPARISON OF THE EPA PLAN TO THE PLAN PROPOSED BY THIS STUDY
     This section compares the episode criteria and control  strategies
proposed by EPA Region IX with those recommended by this study and
discusses the relative advantages and disadvantages of each  plan.   The
episode criteria and control  strategies for the EPA plan were given in
Tables B-l  and B-2, respectively.  For the plan recommended  here,  they
were presented in Sections 6.1 and 6.2.
     Examining Table B-l  reveals that the EPA episode criteria are of the
weather-change forecasting variety.   That is, they are based on the present
day value along with a prediction of better, same, or worsening weather
conditions  for the next day.   This is the same type of episode criteria
                                     B-20

-------
recommended in this study, (Section 6.1).*  The criteria recommended here
differ in that a finer grid of present day concentration ranges and fore-
casted weather changes (e.g. much better, better, same, worse, much worse)
is used.   However, this difference just reflects the fact that the control
strategy recommended here has more stages.  Overall, the episode criteria
for the two plans are very similar.  They are both set to call controls at
about the same level, and they are both of the same (forecasting) variety.
     Table B-5. compares the control strategies of the two plans.  The
three EPA stages correspond to voluntary reductions, certain mandatory
controls, and a complete work holiday, respectively.  It can be seen that
EPA stages I and III are essentially equivalent to Stages I and IV of the
control plan recommended by this study.  Thus, for both plans, the first and
last stages attain negligible and 64% RHC reductions,  respectively.   The
two control plans differ only in that the recommended  plan has tw_g_
intermediate stages (between voluntary abatement and the work holiday)
and in that some of the control measures in the intermediate stages are
not the same.  Considering that 27 alternative control measures were
considered prior to recommending a plan in this study, the two plans are
actually quite similar.
     EPA Stage II essentially covers all of the measures in "recommended"
Stage  II.  Both contain "ban private fleets" and "close government agencies."
* The reader is also referred to Section 4.1 for a discussion of four
  alternative varieties of episode criteria, (feedback type, future-weather
  forecast type, weather-change forecast type, and combination feedback-
  forecast type).
                                      B-21

-------
                                    TABLE B-5  COMPARISON OF THE EPA CONTROL STRATEGIES AND THE
                                               CONTROL STRATEGIES  RECOMMENDED BY THIS  STUDY
                               EPA PLAN
                                                                                     PLAN PROPOSED IN CHAPTER 6

Stage


Control Strategy

Incremental
RHC
Reduction
Cumulative f
RHC I
Reduction 1
CD
t
ro
IND

Stage

iJElUllBauM

Control Strategy


Incremental
RHC
Reduction

Cumulative !
RHC i
Reduction ]
t
1!


1.
2.

Press Release
Voluntary Reduction of
Vehicle Use & Emissions


negl .


negl .
aJDLJJJUUBJJ

II



1. Stage I Actions
2. Close Recreational Areas
3. Close Recreational Events
4. Ban (Private) Non-Critical
Fleet Vehicles
5. Close Non-Critical
Government Agencies
6. Reduce Direct & Indirect
Emissions from Private
Businesses w/More than
100 Employees

negl .
4%%
10%
3%
4%%

maninfBiaiwttrKw -a



22%

III
1. Stage I Actions
2. Stage II Actions
3. Close All Non-Critical
Business
negl .
221
kn
64%
1'
1. Media Announcement &
quest for Voluntary
Reduction of Vehicle
& Emissions
Re-
Use
negl .
|
j
j
negl. \
v.
II
iiJJKLMJWIHJl
n**Af9i.t\.K<f.™*™e™>rf-'-W'^w*
1 . Stage I Actions
2. Ban Filling of Underground
Service Station Tanks
3. Shut-down Major RHC Point
Sources & Other RHC Source
Categories Designated by the
APCD - (Exemptions Allowed)
4. Ban (Private) Non-Critical
Fleet Vehicles
5. Close (Non-Essential )
Government Agencies
.uMun.ia .am,iimM-mjjH-
negl .
4%
3%
6%
3%
BaixsaiaiaHBXziaauj'.&MB*:'^
1
16%
ITT T

i.
2.
Stage I Actions
Operate on Sunday Status
negl.
39%
1
39% 1



1 . Stage I Actions
2. Complete Work Holiday with
Recreation Closed
)jfl«rf.aaam>BHijau»<^.
negl .
64%

S5&&S&&3U33&&232S5

64%

-------
 "Ban  filling  of underground  service  station  tanks,"  (recommended measure
 H-2),  is  covered  by measure II.4. of  the  EPA  plan.   "Shut down stationary
 RHC sources," (recommended measure 11.3),  is mostly  covered by EPA measure
 II.6.   However, EPA  Stage  II adds  "close  recreation" and  "traffic abatement
 plans for  firms with more  than  100 employees."  This  makes the Stage II
 RHC reduction 22% for the EPA plan,  while it is  only 16%  for  "recommended"
 stage II.
     Although it has a  slightly less stringent second stage, the recommended
 plan adds  another  intermediate  stage which consists  of "operating on Sunday
 status."   This  attains  most  of  the reductions  in "recommended" Stage II,
 (even without implementing the  specific measures of  that stage), and adds
 further vehicular  and stationary source abatement to  bring the total RHC
 emissions  reduction  to  39%.
     As indicated  in Section 6.0, neither  the  EPA nor the recommended plan
appears  to be unambiguously  better.  Each plan has  certain advantages,  and
 each has certain disadvantages.  The choice of which  plan to prefer depends
 on how the relative  benefits of each are weighted.    Some of these advantages
 are noted  and discussed  below.
Advantages of the EPA Plan
     •   One  point that  may  favor the  EPA  plan is that it contains
         three control stages rather  than four.  This may make the plan
         operationally simpler.   It also seems to be more consistent
         with the rather large forecasting error that exists for episode
         type days,  (see Section 4.1.).  Having four stages adds flexibility.
         However, the benefits of measured flexibility cannot really be
                                   B-23

-------
         reaped if forecasts are not reliable.   Forecast inaccuracy may



         make it difficult to know which of three stages is really appropriate



         for a given day.  In such cases having four stages to select from



         may be a useless complexity.





     •   The second major advantage of the EPA  plan is that it attains



         siightly more emission reduction per unit economic cost than does



         the recommended plan in the intermediate control  stage.  As noted



         above, the first 16% RHC reduction for both plans essentially



         comes from the same control measures.   The EPA plan then attains



         an extra 6% RHC reduction in  the intermediate stage by closing



         recreational  facilities and by instituting a commercial traffic



         abatement plan for major firms.  The next intermediate stage in the



         recommended plan attains an extra 23%  reduction (Bringing the



         total to 39%) by "operating on Sunday  status."  The economic cost



         per unit emission reduction is less if recreation is closed and



         major firms follow traffic abatement plans (in which they remain



         open and producing) than it is if firms are required to close as



         on a Sunday.   However, the total cost  effectiveness difference



         between the plans is not very large since the first 16% reductions



         for each plan have similar cost-effectiveness and since recreation



         shut-downs and commercial traffic abatement schemes do involve



         considerable  economic costs themselves.




Advantages  of the Recommended Plan




     •   The principle advantage of the recommended plan is that the control



         measures seem to be organized into stages that are operationally
                                  B-24

-------
simpler and  easier  to  implement.   The  two  "extra" control measures
in the EPA intermediate stage, close recreation and commercial
traffic abatement schemes, appear  to create more difficult enforce-
ment and implementation problems than  do the other control measures
in the EPA Stage II.   It will take quite an administrative program
to effectively  institute commercial traffic abatement schemes,
and closing  all recreational events and areas may present even greater
problems.  The  "extra" intermediate control stage in the recommended
plan, "operate  on Sunday status", may  be very severe as to economic
impact; however, it appears  to be  relatively easy to administer
(assuming the  legal authority to call  a control in this specific
form exists).
The proposed plan also contains operational simplicity in that
Stages III and  IV supplant all previous mandatory controls with
just a single,  more comprehensive  control.  Thus, in supplementing
Stage III, there is no need  to also implement the measures of
Stage II.  In  implementing Stage IV, Stage  II and Stage III measures
need not be  instituted.  This organization  of control strategies
fits well with  the  forecast  type episode criteria which are used on
the plan, (having continuous, sequential staging of controls is
most appropriate for feedback type criteria).

As noted in  the effectiveness calculation of Section B.I, control
measures 2.   and 3.  of EPA Stage III overlap considerably.  In
fact it does not make sense to implement certain aspects of
measure II.2. (such as commercial  traffic abatement plans) when
                             B-25

-------
measure 111.3. will be in effect (a work holiday).  As shown in



Section B.I., this redundancy can be eliminated and operational



simp!icites can be increased by eliminating measure III.2. and



modifying the definition of measure III.3.  If this is don.e., EPA



Stage III corresponds more closely to Stage IV of the recommended plan





Another possible advantage of the recommended plan involves the



way recreational closure is used as a traffic abatement measure.



In the recommended plan, recreation remains open until thr last stage,



when a complete work-holiday is instituted.  The recreational shut-



down then has the very significant effect of reducing substitute



driving which would occur as the result of the work holiday.





In the EPA plan, recreation is closed in Stage II, (before the work



holiday), as well as in Stage III, (with a work holiday).  In Stage



II, recreational shut-downs achieve around a lh% VMT reduction and



a 4%% RHC reduction.  This total reduction is considerable.  However,



it should be noted that most recreational driving (about 2/3) occurs



in the afternoon and evening on weekdays,  (see Figure B.I).  Since



the morning  (and possibly early afternoon) traffic is most important



to oxidant production, the effect of recreational closure will be



considered less significant than indicated by the total emission



reduction.  With this consideration in mind, recreational shut-



downs become much less attractive as an initial control measure.



It seems more appropriate to close recreation in the final stage



as a deterrent to substitute driving during the work holiday.
                           B-26

-------
                               f—I—I—I—I—*•—I—I—I
 12
12
                         HOUR
      Figure B.I  Hourly Distribution of Recreational
                    Trips In The LARTS Study Area
Source:  Reference [B-3]
                          B-27

-------
         It was previously noted  that the -simplicity of three stages (as

         oppossed to four) seemed to  favor  the  EPA plan and  that a three


         stage plan was  more  consistent with  the  rather large forecasting


         errors which now exist on episode  type days.   However,  if forecasting

         on episode type days  could be improved considerably, so that  the

         flexibility of  a four stage  plan could be better  put to use,  then


         a four stage approach might  be preferable.   Whether  or  not fore-
                                                          *
         casting can be  sufficiently  improved is  not known.
* Three possible methods  of  measures  forecast  accuracy are described in
  Section 4.1.3.
                                      B-28

-------
                         REFERENCES - APPENDIX B
B-l.    California Air Resources Board, Personal  Communication  of  Data
       Received from the Department of Motor Vehicles  on  Fleet Vehicle
       Population, October 1972.

B-2.    California Air Resources Board, Personal  Communication  of  Data
       Received from the Highway Patrol  on Fleet Vehicle  Use,  October 1972.

B-3.    California Department of Transporation,  LARTS Base Year Report,
       1967 Origin-Destination Survey, December 1971.

B-4.    California Department of Transportation,  Los  Angeles  Freeway
       Surveillance Control  Project Data, August 1973.

B-5.    California Office of Emergency Services,  Air  Pollution  Emergency
       Traffic Control Planning Committee, Traffic Abatement Plan for
       Air Pollution Episodes, 1973.

B-6.    Environmental Protection Agency,  Region  IX, Preliminary Report of
       the EPA Region IX Task Force on Traffic  Abatement  During Air
       Episodes, 15 September 1972.

B-7.    Los Angeles City Department of Traffic,  Transportation  Planning
       Division, Staff Report:  LARTS Trip Type Distribution,  Supplemental
       Report, July 6 1973.

B-8.    Kandlar, Howard, California Dry Cleaning  Company,  Personal
       Communication, Los Angeles, October 1973.

B-9.    Kearin, D. H. and Lamoureaux,  R.  L., A Survey of Average Driving   *
       Patterns in the Los Angeles Urban Area,  System  Development Corpora-
       tion Technical Memo (TM(L) - 4119/000/01), February 1969.

B-10.   Motor Vehicle Manufacturers Association,  1973 Motor Truck  Facts,
       Detroit, Michigan, 1973.

B-ll.   TRW, Transportation and Environmental Operations,  Transportation
       Control Strategy Development for  the Metropolitan  Los Angeles
       Region, EPA Contract No. 68-02-0048, January  1973.
                                    B-29

-------
                                APPENDIX C



     This appendix presents data on recent oxidant episodes in the



Metropolitan Los Angeles AQCR.  Table C.I lists the date and location of



all occurences of oxidant greater than .40 PPM - one hour average from



January 1970 to June 1973.  The maximum one, two, and four hour average



oxidant are given for each date and location.  Violation of the significant



harm levels are indicated by an asterisk, (these days are summarized in



Table 2.10 of the main text).  During the period under consideration,



only the .40 PPM - four hour average harm level was exceeded.  Such episodes



occurred fourteen times; the greatest excess of the harm level was .53 PPM -



four hour average at Riverside on 6 August 1970.  On this date, at Riverside,



the maximum one and two hour oxidant averages in the period under considera-



tion also occurred, .62 PPM and .59 PPM, respectively.
                                     C-l

-------
Table C.I
OXIDANT EPISODES IN THE
METROPOLITAN LOS ANGELES
AIR QUALITY CONTROL REGION
(1-70 to 6-73)
Federal Significant Harm Levels
Date Station
Saturday, 16 May 1970 Pasadena
Azusa
Monday, 1 June 1970 Pasadena
Azusa
Pomona
Tuesday, 2 June 1970* Azusa
Pomona
Pasadena
Saturday, 16 May 1970 Pasadena
Azusa
Monday, 1 June 1970 Pasadena
Azusa
Pomona
Tuesday, 2 June 1970* Azusa
Pomona
Pasadena
Saturday, 20 June 1970 Redlands
Wednesday, 24 June 1970* Redlands
Riverside
Thursday, 25 June 1970 Pomona
Azusa
Redlands
Riverside
Maximum
One Hour
Average
(PPM)
.70

.46
.42
.44
.43
.41
.47
.47
.46
.46
.42
.44
.43
.41
.47
.47
.46
.43
.52
.45
.48
.47
.45
.41
San Bernardino .40
Thursday, 2 July 1970 Azusa
Saturday, 4 July 1970* Pasadena
Azusa
Pomona
Thursday, 16 July 1970 Azusa
Newhall
Monday, 27 July 1970 Riverside
Thursday, 6 August 1970* Riverside
Azusa
*Significant Harm Level Exceeded
.48
.45
.42
.40
.42
.41
.46
.62
.49

Maximum
Two Hour
Average
(PPM)
.60

.44
.40
.41
.42
.37
.46
.45
.43
.44
.40
.41
.42
.37
.46
.45
.43
.40
.47
.43
.46
.38
.45
.39
.38
.37
.44
.39
.39
.41
.40
.43
.59
.49

Maximum
Four Hour
Average
(PPM)
.40

.38
.36
.35
.36
.31
.42*
.42*
.395
.38
.36
.35
.36
.31
.42*
.42*
.395
.39
.43*
.36
.34
.30
.37
.36
.32
.34
.41*
.38
.35
.35
.38
.36
.53*
.44*

                                     C-2

-------
Table  C.I  (continued)
OX1DANT EPISODES IN THE
METROPOLITAN LOS ANGELES
AIR QUALITY CONTROL REGION
(1-70 to 6-73)
Federal Significant Harm Levels
Date Station
Friday, 7 August 1970* Azusa
Riverside
Monday, 24 August 1970 Azusa
Wednesday, 9 Sept. 1970 Redlands
San Bernarc
Thursday, 10 Sept. 1970* San Bernar.
Azusa
Pomona
Friday, 11 Sept. 1970* Azusa
San Bernar
Wednesday, 16 Sept. 1970 Azusa
Pasadena
Thursday, 17 Sept. 1970 Pasadena
Friday, 18 Sept. 1970 Pasadena
Tuesday, 22 Sept. 1970 Pomona
Wednesday, 8 Sept. 1971 Riverside
Saturday, 11 Sept. 1971 La Habre
Riverside
Monday, 13, Sept. 1971* Riverside
Pasadena
Azusa
Anaheim
Thursday, 29 June 1972* Riverside
Azusa
Tuesday, 11 July 1972* Azusa
Riverside
Thursday, 13 July 1972 Azusa
Friday, 28 July 1972 Riverside
Fontana
San Bernar
La Habre
Significant Harm Level Exceeded
Maximum
One Hour
Average
(PPM)
.70

.58
.48
.48
.44
lino .44
Jino .42
.44
.42
.47
dino .40
.50
.45
.40
.41
.40
.42
.42
.40
.56
.53
.48
.42
.45
.40
.47
.41
.49
.50
.42
dino .42
.40
Maximum
Two Hour
Average
(PPM)
.60

.54
.45
.44
.42
.40
.42
.43
.40
.47
.39
.45
.40
.40
.40
.35
.40
.39
.35
.47
.46
.46
.33
.44
.39
.44
.39
.46
.44
.37
.31
.37
Maximum
Four Hour
Average
(PPM)
.40

.46*
.42*
.38
.36
.34
.44*
.395
.36
.42*
.36
.39
.34
.34
.34
.28
.37
.36
.29
.40*
.35
.38
.25
.40*
.37
.40*
.33
.38
.34
.29
.22
.31
                                    C-3

-------
Table C.I
       (continued)
OXIDANT EPISODES IN THE
METROPOLITAN LOS ANGELES
AIR QUALITY CONTROL REGION
(1-70 to 6-73)
Federal Significant Harm Levels
Date Station
Tuesday, 22 August 1972 Riverside
Thursday, 31 August 1972 Riverside
Saturday, 30 Sept. 1972 Riverside
Tuesday, 5 June 1972 Upland
Wednesday, 6 June 1973* Upland
Fontana
Friday, 8 June 1973* Fontana
Monday, 18 June 1973 Upland
Saturday, 23 June 1973 Fontana
Upland
Thursday, 21 June 1973* Downtown
Los Angeles
* Significant Harm Level Exceeded
Maximum
One Hour
Average
(PPM)
.70

.45
.40
.42
.43
.60
.45
.44
.45
.45
.42
.52
Maximum
Two Hour
Average
(PPM)
.60

.39
.36
.40
.42
.53
.44
.44
.43
.41
.37
.52
Maximum
Four Hour
Average
(PPM)
.40

.36
.32
.36
.38
.46*
.37
.41*
.38
• .36
.32
.47*
Data Sources:
NOTE:
        California Air Resources  Board,  California  Air  Quality  Data,
        Volumes  II, HI  and  IV, January  1970  through  December 1972.

        Air Monitoring Logs  of the  Los Angeles  County APCD
                        11    "    "  San Bernardino   "    "
                        11    "    "  Riverside        "    "
                        	'  Orange           "    "

This table includes all  days and  stations  with  max  one  hour oxidant
greater than or  equal  to .40 PPM.  Data  in 1970 and 1971  from the UC
Riverside station at Upland  have  not been  included.   That station
frequently experienced the  highest oxidant concentrations in the
basin; however,  a data analysis by the ARB has  lead to  the conclusion
that the UCR Upland data is  unreliable.
                                    C-4

-------
                                APPENDIX  D

               CALCULATION OF VMT REDUCTIONS ASSOCIATED WITH
                    VARIOUS TRAFFIC ABATEMENT MEASURES


     Chapter 5 of the text evaluates alternative control measures for air

pollution episodes in the Metropolitan Los Angeles AQCR.  Table 5.4

summarizes the effectiveness, institutional obstacles, and socio-economic

impact associated with the implementation of 24 different traffic abate-

ment schemes.  This appendix documents the VMT reductions expected from

each measure (effectiveness).

     Below, the various traffic abatement measures are evaluated in the

order of their appearance in Table 5.4.  Of course, for each measure, the

expected VMT reduction depends on the specific implementation and enforce-

ment activities associated with it.  Detailed descriptions of each traffic

control measure, including implementation and enforcement procedures,

can be found in Table 5.4.

DIRECT CONTROLS ON TRAFFIC

1.  Voluntary Traffic Abatement

     This "non-active" control is likely to have negligible effect on VMT.

Its purpose in an entire control strategy should be to provide certain

prerequisites for direct control, (media campaign, special announcements,

etc.)

     Past experience in Los Angeles and San Francisco indicates essentially

no change in total VMT resulting from such a measure [D-4], [D-5].  However, it

is possible that a more extensive and better organized public relations

effort on the media might bring about some discernable reductions.  Without

making provisions for workers to be paid if they choose to stay home (in

which case this comes under "Work Holiday"), the maximal effectiveness

should be less than 5%.

                                    D-l

-------
     It is assumed that voluntary abatement would be effective only for

LDMV's and not for HDMV's  or diesels.
     Total  VMT Reduction:    Negligible to 5%.

                            Most likely negligible.

     Applicable to:          LDMV's
2.  Windshield Sticker Program

     The effectiveness of the windshield sticker program varies

depending on the number of sticker colors that are  banned from travel.   The

theoretical maximal  potential reduction, with full  public knowledge, full

acceptance, and adequate enforcement,  is  85-90% of total  VMT (all  but emer-

gency vehicles), [D-5], [D-6].  The actual maximal  reduction would depend on

the actual degree of knowledge,  acceptance, and enforcement.  Considering these

problems, the State  Office of Emergency  Services Committee estimates around

a 50% maximum, [D-5].

     The strategy can  be applied to LDMV's, HDMV's  and diesels.
     Total  VMT Reduction:    Flexible.   Maximal  is  50-85% depending on
                            implementation and  enforcement procedures.
     Applicable to:          LDMV's,  HDMV's, and diesels.
3.  License Plate Lottery

     Like the Windshield Sticker Program,  the License Plate Lottery provides

for variable effectiveness.   In  this  case, traffic reductions depend on how

many "last license plate numbers" are banned.  The EPA Region IX Task Force

on Traffic Abatement,  [D-6], reports  the following maximal traffic reductions

per quantity of "last  numbers"  banned:  (The EPA results have been modified
                                    D-2

-------
to reflect an overall maximum VMT reduction of 85%, corresponding to all

non-emergency vehicles.)
                     "Last Numbers'
                  Maximal
Banned
1
2
3
4
5
6
7
8
9
10
VMT Reduction
negligible
4%
8%
13%
21%
30%
42%
55%
68%
85%
     Of course, as with "Windshield Stickers", the maximum reduction is

likely to be less than 85% due to problems of public knowledge and acceptance,

and to considerable enforcement difficulties.  Public compliance with "license

plate lottery" might be less than with "Windshield Stickers" because of the

apparently more arbitrary nature of the former.  Enforcement also might be

more difficult, (it would be harder to discriminate between vehicles in

general  and emergency auto use, e.g., doctor's cars, in particular).
     Total  VMT Reduction:
     Applicable to:
Flexible.   Maximal  is 50-85% depending on
implementation and  enforcement procedures.

LDMV's, HDMV's and  diesels.
                                    D-3

-------
4.   Ban Non-Essential  Government Vehicles
     In 1973, there are around 100,000 government vehicles operating in the
Metropolitan Los Angeles AQCR, [D-10].  Assuming 75% of these would be
classified as non-essential, 75,000 would be banned from travel.  This
would constitute 1.2%  of the total   6.2 million LDMV's and HDMV's in the
basin, [D-10].   Since  public and commercial  type vehicles  tend to travel more
miles per day than private vehicles, [D-2], [D-10], it is assumed that the total VMT
reduction would be around  1%%.    It is assumed that LDMV's and HDMV's are
equally affected by this measure.
     Total VMT Reduction:    1%%
     Applicable to:         LDMV's and HDMV's
5.  Ban (Private) Fleet Vehicles
     There are presently around 430,000 vehicles contained in private fleets
of ten or more which operate in the Metropolitan Los Angeles AQCR, [D-l], [D-10].
It is not known how many vehicles would fall in exempt categories (taxis and

transit vehicles, delivery of perishables, essential deliveries, and
emergency utility service).  It is arbitrarily assumed that 1/3 of the
vehicles would be exempt.   This implies that 290,000 vehicles would be
included.
     Assuming 20% of these are HDMV's [D-9], then 25% of the total 240,000
HDMV's in the basin and 4% of the total 5,900,000 LDMV's in the basin are
contained in non-exempt private fleets.  Since fleet vehicles tend to
be used more intensively than non-fleet vehicles, it is assumed that a 30%
reduction  in HDMV VMT and  a 5% reduction in LDMV VMT is attainable through
                                    D-4

-------
this strategy.   This  result  is  very  close  to  an  independent  estimate  of  5%


total VMT  reduction  by  the OES  Report,  [D-5].
     Total  VMT  Reduction:     5%  of LDMV  VMT

                             30%  of HDMV  VMT
 6.    Ban  (Private)  Fleet Vehicles  and Use of Vehicles  for  Commercial  Purposes


      As with  the  fleet vehicle ban,  insufficient data  are  available  to


 accurately  assess the effect of banning vehicles for commercial  use.  An


 estimate  will  be  made here by using  data on numbers  of trucks  (HDMV's)  in


 various use categories compared to the number in fleets, [D-9].  These data


 will  provide  a ratio of VMT by trucks (both fleet and  non-fleet)  in  com-


 mercial use as compared to VMT by  trucks in fleets only.   A similar  ratio


 will  then be  assumed for LDMV's and the effectiveness  of Strategy 6.  will


 be obtained by appropriately factoring the effectiveness of Strategy 5.,


 (which was  for fleet vehicles only).


      In commercial and  industrial  categories,  (other than  personal trans-


 portation and  agriculature),  around  60%  of  total  trucks are not contained


 in fleets of ten  or more,  [D-9].   Assuming  that  the same percentage of


 non-fleet commercial  trucks will be  exempt  as  are  fleet trucks, this  implies

                           flr\°i+A.r\°/
 that  strategy  6.  is 5/2,  (  40%  ).  as  effective  as strategy 5. for trucks


 (HDMV's).    It  is  possible  that  the ratio  of non-fleet  to fleet LDMV's in


 commercial use is less  than the ratio for trucks.  To  be somewhat conservative,


 it will  be assumed that  strategy 6.  is  only  twice  as effective as strategy 5.


 for LDMV's.


     Thus, with the approximate factoring of strategy  5.,  the effectiveness


of the present strategy  is 5/2 x 30%  =  75%  for HDMV and 2x5%=  10%  for LDMV.
                                     D-5

-------
This result agrees rather well  with a totally independent OES Report result,
 [D-5].  Combining the OES fleet and  commercial vehicle strategies  in the
appropriate way,  one obtains an estimated 12% VMT reduction.
     Strategy 6.  will  present much greater enforcement problems than the
more simple fleet-vehicle ban.   The actual effectiveness  of this strategy
may be significantly less than calculated.  Thus, the VMT reductions below
contain ranges indicating uncertainty due to possible enforcement difficulties.
     Total VMT Reduction   7 - 10% of LDMV VMT :
                                               i
                          50 - 75% of HDMV VMT |
                          Depending on enforce-
                          ment procedures.
7.  Ban All Non-Emergency Traffic
     Banning all traffic except for essential  vehicles and emergencies has a
maximum potential effectiveness of 85-90%, corresponding to all non-emergency
traffic.  The actual reduction will depend on  public knowledge and acceptance
as well as degree of enforcement.  The OES Report anticipated around a 50%
reduction, [D-5].
     Total VMT Reduction:   50-85%, depending on implementation and
                            enforcement procedures.
     Applicable to:         LDMV's, HDMV's and diesels.
8.  Mandatory Car-Pooling on Freeways
     From Table 2.16, average auto-occupancy in the Los Angeles Basin is
around 1.4 persons per vehicle on weekdays.  From Table 2.12, freeways
account for 48% of total  VMT.  Assuming carpooling causes an extra 10%
                                    D-6

-------
driving in picking up passengers) and that the same drivers use the freeway,

the total % VMT reduction would be


     A VMT(%) = 48% (1 -
where NFOR is the new freeway occupancy  rate,  [D-10J.   If NFOR could be increased

to 2, the total VMT reduction would be 11%.  If NFOR were increased to 3,  the

total VMT reduction would be 23%.  Reductions  in number of passenger trips

might yield even greater VMT reductions.

     However, it is likely that many people would use surface streets

rather than carpool and use the freeway.  This would negate some of the

above effect.  These drivers would be likely to travel  longer and produce

extra emissions.  These effects are likely to cancel at least 50% of the

above VMT reductions to around 5% (2 people per car) and 10% (3 people per

car), respectively.

     Depending on public awareness and acceptance and the level of enforce-

ment, actual VMT changes could be less than half the above estimates.
     Total VMT Reduction:   2-5%  (two people per vehicle),
                            4-10%  (three people per vehicle),
                            Depending on implementation and
                            enforcement procedures.

     Applicable to:      LDMV's
9.  Mandatory Carpooling on Freeways and Major Thoroughfares

     It will be assumed here that 90% of all vehicles use freeways and

"major thoroughfares", (this will actually depend on how the later is defined).

Average auto occupancy is 1.4 persons per vehicle.  Assuming an average increase

of 10% in trip length is required to accomodate carpoolers, then the total
                                    D-7

-------
change in VMT due  to  changing  the  average occupancy rate,  AOR,  is

     AVMT(%)  = 90% (1  -  K1  X  1>4) .
                          AOR

Increasing AOR to  2 leads to a 20% reduction.   An average  of 3  persons per
vehicle leads to a 44% reduction.   If passenger trips also decrease, even
greater VMT reductions might occur.
     In contrast to the  freeway carpooling increase above, there will be
much less possibility for drivers  to  escape the intent of  the regulation
by taking  alternative routes.  However, drivers who do attempt this would
likely travel much longer and cause  traffic jams; both of  these effects
would increase emissions from those  vehicles.   Both effects would depend
on the definition  of "major thoroughfare".  It will be assumed  that with
full enforcement and public awareness, maximal VMT reductions will be around
15% and 35% respectively for 2 and 3  persons per vehicle.
     However, enforcement cannot be completely thorough, and there will
certainly be public non-awareness  and non-cooperation.  This may reduce the
VMT changes by one-half or more.
     Total VMT Reduction:  7-15% (2 persons per vehicle),
                          15-35% (3 persons per vehicle),
                          depending on implementation and enforce-
                          ment procedures.
     Applicable to:       LDMV's
                                    D-8

-------
10.   Commercial Abatement Schemes
     The traffic reduction obtainable from commercial  abatement schemes  is
variable according to the number of schemes that are called upon for imple-
mentation.   Weekday, work related VMT, (see Table 2.13),  accounts  for around
50% of total VMT, (32% Home-Work , 10% Work-Other, and around 8% Other-Other).
To eliminate nearly all work related travel,would basically require  that
all  non-essential employees stay home.  This would then be basically a
"Work-Holiday", resulting in around a 35% VMT reduction.   To keep  businesses
operational, the VMT reduction would have to come basically from carpooling,
increased transit ridership, and elimination of certain postponable  business
trips.  Around  one-half  of work related  VMT,  25%  of total  VMT,  might  be
eliminated in this manner.
     Unfortunately, very difficult enforcement problems are inherent in
using commercial abatement schemes for increasing carpooling and transit
ridership, and for decreasing business trips.  As indicated in Appendix A,
experience with federal employees indicates that lacking  enforcement,
the  effectiveness of such schemes (without giving work holidays) is  minimal.
Enforcement and implementation difficulties are likely to severly  limit
the reductions that can actually be obtained.   A 15-20% maximum seems plausible
considering these difficulties.
     Total  VMT Reduction:   Flexible.
                            Maximum: 35%, calling "Work Holidays"
                                     0-20%*,  keeping business  operational
                            * depending on enforcement and implementation
                              procedures.
                                    D-9

-------
CONTROLS ON TRAFFIC SUPPORT FACILITIES
11.   Close Freeways (or Certain Freeway Lanes)
     Closing freeways  (or certain freeway lanes) to all  but emergency and
transit vehicles would achieve some positive effects by  discouraging person
trips and by encouraging a shift toward transit ridership.   However, this
effect would be counterbalanced by an  increase  in VMT due to longer trips
per auto and by an increase in emissions  per mile due to traffic congestion.
Overall, the strategy  is likely to cause  an  increase in  emissions.
     Total  VMT Reduction:    Doubtful,  very possibly an increase.
12.  Limited Availability of Parking
     Limiting parking spaces could theoretically reduce VMT by discouraging
some person trips and by encouraging the use of carpools and mass transit.
In practice, drivers may just "take the chance" of finding parking or
park illegally.   Actual  increases  in VMT may be produced by extra driving
in search of parking and by some commuters  who may have another family
member to drive  them to  work and later pick them up (thus doubling their
normal trip distance).
     Total  VMT Reduction:    Doubtful ,very possibly an increase.
13.  Special Parking Tax
     A "special  parking tax"  scheme may be more effective than "limited
parking availability" since it can be  conducted in such a way as to provide
the driver with  more certain  implications  about the outcome of his decisions.
That is, it could be made apparent to  the driver that he will have to pay an
                                   D-10

-------
extra $20 to park Downtown  (whereas "limited parking availability" just

means he may or may not  find a place to park).  This greater certainty

of parking penalty should provide a stronger incentive for carpooling

and public transit usage.


     However, there still may be a counterproductive tendency due to in-

creased driving in search of places not covered by the tax and due to some

persons being driven  to  work by non-working family members.  Many people,

of course, will just  decide to pay the burden of the tax (and complain

later).

     The OES report estimates about a 20% VMT reduction for this strategy,

(with a 50% possible  maximum).  In light of the above, the OES estimate

appears optimistic.
     Total VMT Reduction:   Very Uncertain.
                            Possibly as much as 20%
                            but possibly negligible.
14.  Gasoline Rationing

     A gasoline rationing scheme, based on a "coupon number system", provides

a flexible means of reducing traffic by controlling fuel use.  Estimates

of maximum effectiveness for this scheme vary widely.  The EPA Region IX

report assumes a 20-50% VMT reduction; the OES report gives 5%, 10% and

25% for the 1st, 2nd, and 3rd days respectively.  The maximum effectiveness

is estimated below as follows:

     The average vehicle miles travelled per day per car is 30 miles, [D-10].

Assume that an average vehicle can travel 180 miles per full tank of gas,.
                                   D-ll

-------
(12 gallons,  15  miles  per gallon).   Further assume  that on the average,

for every full  tank  of gas  used,  a  driver stops  for gas 1% times

(i.e.  drivers do not use  one  full tank  per gas  stop).   Then,  the

average driver stops once every 4 days  for gas.


     On the first day  of  gas  rationing,  a  maximum of 25% effectiveness  is

thus plausible (not counting  the  psychological  effect of the  danger of

running out of gas).  However,  of the people who stop for gas each day,

not all need to.  Much of this  25%  probably consists of drivers who

have enough gas to continue that  day.  Also, some  people have access to

two cars, and will use the one with  more gas.  A 10% maximum effective--

ness is thus assumed here for the  first day.  This  should increase to

around 20% to 50% by the  second and third days.
     Total VMT Reduction:    Flexible
                            Maximum:  10% 1st day
                                      20% 2nd day
                                      50% 3rd day

     Applicable to:          LDMV's  and HDMV's
15.  Close Gasoline Stations

     The effectiveness of closing gasoline stations  is  equivalent to the

maximum effectiveness of a gas  rationing scheme.   As estimated in 14, above,

this is a 10%, 20%, and 50% VMT reduction the 1st, 2nd, and 3rd days,

respectively.
     Total VMT Reduction:    10% 1st day
                            20% 2nd day
                            50% 3rd day
     Applicable to:         LDMV's and HDMV's
                                   D-12

-------
     The effect of  this  strategy will  be  found by adding together estimates

for the reduction in  shopping,  sales-service, and work trips.

Shopping:  The LARTS  and SDC  surveys,  [D-3],  [D-7]  indicate  that around 20%

of total vehicle trips  in the LARTS  region  are shopping related.  Since

only around 15% of  all  shopping trips  are to  RSC's  and CBD's, only

20% x  .15 = 3% of total  vehicle trips  will  be affected by this strategy.

Assuming an average trip length of 4.6 miles  to or  from a CBD or RSC,

[D-5], the total VMT  change from reduced  shopping trips would be



  10W * 10%0% xl.fmiles  (average length of  all '  a of total VMT
                           trips in  the basin)


Sales-Service:  Based on Los  Angeles CBD  data, [D-5], [D-ll], it is assumed

that for typical retail  establishments, sales and service trips are around

20% of shopping trips to these  establishments.  Assuming that a sales-

service trip is slightly longer than a shopping trip, around %%_ of total

VMT,(2% x.20 x  1 ),  is  accounted for  by  sales-service trips to retail

establishments in the RSC's and CBD's.

Employment:  Retail trade  accounts for around 14% of total employment in the

Los Angeles basin.  Since  only  around  15% of  such employees  are in RSC's

and CBD's, only .15 x 14%  = 2%  of total employment  in the basin will  be


affected by this  strategy.  Work-related VMT is  around 50% of total  basfn-

wide VMT.  Thus, .02 x  50% =  ]%_ of total VMT will be eliminated by

employees staying home.  This reduction actually will be slightly less

due to substitute driving.
                                    D-15

-------
     Adding together the traffic reductions from shopping, sales-service,



and employment, the total  effectiveness of closing retail outlets in



RSC's and CBD's is 2% + %% + 1% = 3%% of total VMT.
     Total VMT Reduction:     3h%                    \
                                                    t



     Applicable to:          LDMV's and HDMV's
20.  Close Recreational  Facilities



     LARTS survey data,  (modified according to recent corrections, see



Section 2.2.6) indicates that recreation accounts for 7.2% of total



weekday vehicle trips in the Los  Angeles basin,  [D-3.].   The results



of an SDC (System Development Corporation)  survey,  [D-7], indicate that



around 11% of total  weekday trips are for recreational  purposes.   The



basic reason for the discrepancy  between the two surveys involves the



definition of "recreation".  In both studies the basic  trip destinations



and origins can be classified as  "Work", "Home", "Shop", and "Other",



and in both studies  recreation is part of the "Other" category.   However,



within the "Other" category, the  LARTS study separates  out "recreation"



and "social-entertainment", while the SDC study  includes both of these



under "recreation".   Thus, the difference between SDC and LARTS



estimates basically  reflects "social-entertainment" trips.




     For the purpose of  the present study,  "recreation"  should include



"entertainment" since the control "Close Recreational Facilities" includes



theaters, playhouses, clubs, etc.  It will  be assumed that 9% of total



weekday trips are accounted for by this definition  of recreation.



     Recreational trips  basically occur in  the "Home-Other" and "Other-



Other" categories.   The  average trip length in these categories is




                                   D-16

-------
CONTROLS ON  INCENTIVES TO  DRIVE


16.  Close  (Non-Critical)   Federal Agencies


     As noted  in Table 2.3,   federal  agencies accout for   2%%  of the total


4.2 million  people employed  in the Metropolitan Los Angeles AQCR.  It will


be arbitrarily assumed that  95%  of federal employees can be classified as


"non-critical"    As  noted in Table  2.13,  50% of weekday VMT  is work related:


(32% Home-Work, 10% Work-Other,  and  8% Other-Other).  Thus, the decrease


in work traffic due to federal agency shut-down should be .025 x .95 x 50% = 1.2%


of total VMT.


     Section 2.6 noted that  a decrease in work related VMT by "X" brings


about a "1/5X" increase  in other types of driving so that the total reduction


in VMT is only "4/5X".   Thus, allowing for substitute driving, the total VMT


reduction from closing federal agencies would be 1%.
     Total VMT Reduction:    1%

     Applicable to:          LDMV's
17.  Close  (Non-Critical)   Government Agencies (Federal, State & Local)


     As noted in Table  2.3 , government agencies  account for 13% of the total


employment  in the Los Angeles Basin.  It is assumed that 90% of these employees


will be classified as non-critical   (more state and local employees will


be "essential" than federal employees because of police, fire, ambulance,


and other services).  Noting that work related VMT accounts for 50% of total
               A

VMT, (Table 2.13), around .13 x .9 x 50% = 6% of total VMT will be eliminated


by government agency shut-down.   Correcting for substitute driving, the


actual  VMT reduction will  be around  5%.
                                    D-13

-------
     If public school closure is included in this strategy, 1% to

extra VMT reduction will  be obtained, (see Appendix B).
     Total VMT Reduction:   5%*
     Applicable to:         LDMV's

     * Add 1% if public school  closure
       is to be included.
18.  Close Government Contractors

     It is difficult to obtain estimates of how many people are employed by

government contractors.   Based on  indications in the OES and Region IX

reports, [D-5], [D-6], it will be  arbitrarily assumed that 20% of the total

basinwide employees work for government contractors.   Assuming 95% of these

employees are "non-essential" and  noting that 50% of total  VMT is work

related, the effectiveness of this strategy would be around

.20 x .95 x 50% = 9%% of total VMT.   Correcting for substitute driving

(as in 16. above), the actual reduction would be around 7 %.
     Total VMT Reduction:     7%%

     Applicable to:           LDMV's
19.  Close Shopping Facilities  at Regional  Shopping Centers (RSC's) and
     Central  Business  Districts (CBD's)

     This strategy will  be assumed to apply to 48 regional  shopping centers

and 6 central  business  districts  which have been identified in the Los Angeles

basin, [D-5],  [D-8].   These RSC's and CBDls respectively account for $2.2

and $.74 billion of the total  $20 billion annual retail  sales volume in the

Los Angeles basin, [D-5].  Together, they thus account for (2.94/20) x 100% = 15%

of total  retail  sales  volume.
                                   D-14

-------
about 6-6% miles  (see Table 2.]3).  However, it is expected that
recreational trips would be longer than non-recreational trips in the
Home-Other and Other-Other categories.  Recreational trips to the beach,
sporting events,  amusement parks, fairs, etc. would probably average at
least as long as  Home-Work trips  (10 miles).  This effect would only
partially be cancelled by recreational trips to clubs, bowling alleys,
theaters, etc., which would probably average about the same as Home-Shop
trips (3 miles).  It is assumed here that the average length of a recreational
trip is 7 miles.
     With total recreational trips representing 9% of average weekday
trips and having  an average length of 7 miles, the maximal possible effect
of "Close Recreational Facilities" would be

  100% x 100^ x Smiles (average length of all = *& of weekda* VMT'
                           trips  in basin)
However, it is extremely unlikely that all recreational  facilities will
be covered by the regulations  (i.e. certain facilities may be neglected).
Also some recreation . may just involve "pleasure driving" and this
would not be covered.  With adequate implementation and enforcement
procedures it is  assumed that 80% of recreational driving can be
eliminated.  This would mean that 7%% of total weekday VMT would be
effected.
     Total VMT Reduction:
     Applicable to:         LDMV's
                                   D-17

-------
21.   Work Holiday for Major Firms
     Private firms  with  100 or  more  employees  account for around 29% of
total  employment in the  Los Angeles  basin  [D-5].  Assuming  95%  effectiveness
among these  employees,  (some workers will  be  classified  as  "essential"  and
there may be certain notification  problems with  others),  around 27%% of
total  employment in the  basin will be  affected by  the closure  of these  firms,
Since work related  traffic  accounts  for around 50%  of total  VMT on  an
average weekday, .275 x  50% = 13.8%  of total  VMT will  be  eliminated.
Substitute driving  will  negate  1/5 of  this decrease (see  section 2.6).
Thus, the expected  VMT decrease, allowing  for  substitute  driving,  should
be 11%.
     Total  VMT Reduction:    11%
     Applicable to:          LDMV's  and  HDMV's
22.  Rotating Work Holiday for Major Firms
     This control  measure is  essentially  the  same  as  control  measure #21
except flexibility has  been added  in that anywhere from 1/5 to all  of the
firms with 100 or  more  employees may be closed.  The  range of possible
VMT reductions thus is  2.2% to 11%
Total VMT Reduction:




Applicable to:
2.2%
4.4%
6.6%
8.8%
11%
LDMV
1/5 of Major Firms
- 2/5 	
- 3/5 "
- 4/5 	
- All 	
s and HDMV's
                                   D-18

-------
23.   Operate on Sunday Status
     According to this strategy, a "Sunday" type holiday would be called in
order to reduce incentives to drive.  Most industry and commerce would be
off work, and retail outlets would be closed or would operate on a Sunday
schedule, but recreational facilities would be open.  The VMT reduction
normally achieved on a Sunday is 36% of weekday VMT.  This is less than
would be expected from the cancelling most weekday work and shopping trips
due to substitute driving, particularly for recreational purposes.
     By calling a "Sunday" on a weekday, less than a 36% VMT reduction might
be expected due to incomplete public knowledge and other reasons for
non-compliance.  However, this effect would be counterbalanced by extra VMT
reductions due to the lack of time to plan the substitute driving trips
normally taken on a Sunday.  Also, as a show of good public concern, many
retail and recreational facilities might close even though they do not
normally close on Sundays.  It is estimated that calling a "Sunday holiday"
with appropriate penalties for non-complying industrial and commercial
establishments, should achieve a 35-40% reduction from normal weekday VMT
levels.
     Total VMT Reduction:   35-40%
     Applicable to:         LDMV's and HDMV's
24.  Complete Work Holiday with Recreation Closed
     A complete work holiday would entail closing all industry except for
"essential employees", closing all retail and other commercial businesses
                                   D-19

-------
except for "essential  services",  and closing recreational  facilities (as in
control  measure 20).   The only  traffic remaining would be  from operating
emergency vehicles,  commuting of  essential  employees,  driving for "essential
services" and driving  for recreational and  social  purposes other than those
related to actual  recreational  facilities.   It is  estimated that this
strategy would achieve around a 75%  reduction  in total  weekday VMT,  [D-5],
[D-6].
     Total  VMT Reduction:    75%
     Applicable to:          LDMV's  and HDMV's
                                  D-20

-------
                          REFERENCES  - APPENDIX D
D-l.   California Air  Resources Board, data concerning fleet vehicle
       population received from- the Department of Motor Vehicles,
       personal communication, October 1972.

D-2.   California Air  Resources Board, data concerning fleet vehicle
       use received from the Highway Patrol, personal communication,
       October 1972.

D-3.   California Department of Transportation, LARTS Base Year Report,
       1967 Origin-Destination Survey, December 1971.

D-4.   California Department of Transportation, Data Processing Depart-
       ment, Los Angeles Freeway Surveillance and Control  Project,
       personal communication, September 1973.

D-5.   California State Office of Emergency Services, Air Pollution
       Emergency Traffic Control Planning Committee, Traffic Abate-
       ment Plan for Air Pollution Episodes., 1973.

D-6.   Environmental Protection Agency, RegionIX, Preliminary Report
       of the EPA Region IX Task Force on Traffic Abatement During
       Air Episodes, 15 September 1972.

D-7.   Kearin, D. H. and Lamoureaux, R. L. , A Survey of Average Driving
       Patterns in the Los Angeles Urban Area,  System Development
       Corporation Technical  Memo (TM(L) -  4119/000/01), February 1969.

D-8.   Los Angeles Times, Marketing Research Department, Los Angeles
       Shopping Centers 1971-1972.

D-9.   Motor Vehicle Manufacturers Association, 1973 Motor Truck Facts,
       Detroit, Michigan, 1973.

D-10.  TRW Transportation and Environmental  Operations, Transportation
       Control  Strategy Development for the Metropolitan Los Angeles
       Region,  EPA Contract No. 68-02-0048, January 1973.

D-ll.  Wilber Smith and Associates, Los Angeles Central Business District
       Parking Study, 1967.
                                   D-21

-------