SATELLITE VEHICLE SYSTEMS FOR  SOLID  WASTE  COLLECTION,  EVALUATION
AND  APPLICATION


Ronald  A.  Perkins


Environmental  Protection  Agency
Cincinnati,  Ohio


1971
                                                         Distributed , . .'to foster, serve and promote the
                                                                     nation's economic development
                                                                     and technological advancement.'
               NATIONAL TECHNICAL INFORMATION SERVICE
                                                                  U.S. DEPARTMENT OF COMMERCE
                            This document has been approved for public release and sale.

-------
                                                      IEPA-SW-82TS-71
SATELLITE VEHICLE SYSTEMS FOR SOLID WASTE COLLECTION

               Evaluation and Application
               This study vae made by tits

            Division of Technical Operations

 and its report (SU-82te) was written by RONALD A. PSRKISS
          U.S. ENVIRONMENTAL PROTECTION AGENCY
              Solid Waste Management Office
                          1971

-------
BIBLIOGRAPHIC DATA
SHEET
1. Report No.
     EPA-SW-82TS-71
                                                 I
4. Title and Subtitle
    Satellite Vehicle Systems for Solid Waste Collection;
    Evaluation and Application
7. Amhot(a)
   Ronald A.  Perkins
9. Performing Organization Name and Addresa
   U.S.  Environmental Protection Agency
   Solid Waste Management Office
   Cincinnati, Ohio  45213
12. Sponsoring Organization Name and Addreaa

   U.S.  Environmental Protection Agency
   Solid Waste Management Office
   Rockville, Maryland  20852
                                                                   3. Recipient'a Accesaioo No.
                                               5. Report Date
                                                 1971
                                               8. Performing Organization-Kept.
                                                 No.
                                                                   10. Project/Taslc/Tork Unit No.
                                               U. Contract/Grant No.
                                               13. Type of Report & Period
                                                 Covered
15. Supplementary Notes

   COLOR  ILLUSTRATIONS REPRODUCED IN BLACK  AND WHITE
16. Abstracts
The  productive efficiency and costs  associated with residential  satellite vehicle col-
lection systems are quantitatively evaluated and compared with conventional collection
systems for collecting residential solid waste.  Those factors that  affect efficiency
and  economy were quantitatively analyzed to develop models that  can  predict collection
cost and crew efficiency with a high degree of accuracy and provide  a practical and
reliable basis for designing future  collection systems using satellite vehicles.  Four
makes of satellite vehicles (Cushman, Trashmobile, Westcoaster,  and  Trash Taxi) opera-
ting in six communities with diverse terrain, type of collection agency, and collection
frequency were chosen for the study.  These small three- or four-wheeled vehicles
transport wastes from the dwelling unit storage point to a packer truck for ultimate
transfer to a disposal site.  The vehicles reportedly work best  in areas with single-
family homes where housing density is medium to low.  The costs, which could not be
compared because they occurred under distinctively different conditions, are measured
as "annual cost per dwelling'^ rather than as "cost per ton. '^  (  ' !      ~~	
17. Key Vords and Document Analysis. 17o» Descriptors

*Refuse disposal, *Collection, *Ground vehicles, Bulk transporter,  *Systems analysis,
Mathematical models, *Cost comparison
17°. Identifiers/Open-Ended Terras

*Solid waste disposal, *Waste  collectors, Cushman, Trashmobile,  Westcoaster, Trash Taxi
*Solid waste collection system,  *Satellite vehicle waste collection, Mini motor truck,
Resldneltal density, Atlanta (Georgia),  Columbia (South Carolina),  Knoxville (Tennessee)
Medford (Oregon), Pasadena  (California), Waukesha County (Wisconsin)
17e. COSATI Field/Group
                      13B
!«. Availability Statement

   Release to public
                                    Insecurity Class (This
                                      Report)
20. Security Class (This
   ^UNCLASSIFIED
                                                                             21 NoTof Pages
                                                                                253
                                                                             21 Price
                                                                             $3 or $0.95
                                                                                                                                         FOREWORD


                                                                                                              Current solid waste collection  practices do not differ  significantly

                                                                                                         from those used at the turn of  the century.   The horse-drawn cart  has

                                                                                                         merely been replaced by a motor-powered van that has a larger capacity

                                                                                                         and  a compaction capability.  This lag in the development of new tech-

                                                                                                         nologies has had a marked economic impact, because it costs  four times

                                                                                                         as much to collect residential  solid wastes as it does to dispose  of

                                                                                                         them.   Ways must be found, therefore,  to maximize efficiency while

                                                                                                         minimizing costs.

                                                                                                              Thus, an in-depth study was made  of a new collection system in

                                                                                                         which small ''satellite" vehicles are used.  The present report is  the

                                                                                                         result and presents operational details of the system and compares it

                                                                                                         with conventional walking pick  up.   The relative efficiencies achieved

                                                                                                         by private and municipal operators are also outlined.  Finally, the

                                                                                                         report postulates that a community can use the mathematical  models

                                                                                                         developed by the study to estimate the efficiency and cost of such a
                                                                                                                                                                             I
                                                                                                         system for its area, thus avoiding expensive field studies and trial

                                                                                                         implementation.
                                                                                                                          —RICHARD D. VAUGHAN
                                                                                                                            Assistant Surgeon General
                                                                                                                            Acting Commissioner
                                                                                                                            Solid Waste Management Office
                                                                                                                         ill

-------
                        CONTENTS.

                                                            Page

SUMMARY AND CONCLUSIONS                       "

   General . 		  1

   System Design and Application 	 	 ....  1

   System Capabilities .	5 °

   Crew Activity Analysis  ..............'.....  6

   Items Collected per Dwelling Unit .......	6

   System Costs	  6

   Comparison of Private and Municipal Systems	7

   Qualitative Evaluation of Satellite Vehicle Systems ...  7

   Operational Recommendations 	 .....  8

INTRODUCTION .............	 .,11

STUDY PROCEDURE		 18

 . -Phase one ~ Cost Study  .................19

   Phase Two - Field Evaluation	2O

       Collection Time Study Work Sheets .	2O .

       Driver Collection Time Data Sheet_. . ... . . . . . . 26

FIELD STUDY DATA ANALYSIS  .	.- . .	32

   C3tellite Vehicle Operational Analysis  . .... . . . . . 32

       Regression Model Analysis . . .  .	 36

            X , Dwelling Units per Satellite Vehicle
                Load . .	 36 ..

            X_, Items per Satellite Vehicle Load '. . . . . .41
                                                          Page

          X , Distance Satellite Vehicle Travels
              on Dwelling Unit Property	41

       •  X., Average! Distance from Satellite Vehicle
              to Storage	44

          X,., Satellite Vehicle Route Distance per
              Load		47

        *  Y , Productive Collection Time per Satellite
              Vehicle Load  ..........:....  49

     Analysis of Other Variables	 ...  49

          Items per Dwelling Unit . .	SI

          Distance from Street to Storage	  54

          Satellite Vehicle Unloading Time   	  54'

          Other Time	  61

     Effects of External Influences on
    'Collection Efficiency	61

          Collection Frequency	61

          Collection Agency 	  64

          Satellite Vehicle Type	64

          Collection Area Terrain	68

          Type of Item Collected  ............  71

          Weather Conditions  	  	  73

Packer Truck Driver Operational Analysis	74

     Regression Model Analysis  	  76

          b , Dwelling Units Serviced Between
              Truck Moves .	  .  76
          b2, Total Number of Items Collected
              Between Truck Moves	
          b ,  Average Distance from Packer Truck
              to Waste Storage Location ......
                                                           76
                                                           79

-------
       Packer Truck Driver Activity Analysis	79

SYSTEMS COST ANALYSIS	  84

   Daily Crew Costs	  .  .  84

       Labor	84

       Satellite Vehicle Operation and Depreciation   .  .  .  .  86

       Packer Truck Operation and Depreciation .Costs  ....  87

       Overhead	.88

   Annual Collection Cost per Dwelling Unit	88

       Collection Frequency  Effects on Cost	91

       Collection Agency Effects on Cost	93

   Collection Cost per Ton	93

   Development of Cost Estimation Model	94

       True Crew Cost per Hour	95

       Crew Efficiency  ...................  96

       Cost Estimation Model .'	...  98
SATELLITE VEHICLE WASTE COLLECTION VS. CONVENTIONAL
COLLECTION	
                                                              99
   Qualitative Evaluation	  .   1OO

   Quantitative Evaluation	1O2

       Collection Efficiency		1O2

       Collection Economy   ..'.	.  .   1O2

SATELLITE VEHICLE WASTE COLLECTION MODEL APPLICATION  .  .  .   1O4

REFERENCES	-  ...   HI

BIBLIOGRAPHY	'.  .  .   112

ACKNOWLEDGMENTS	114
                                vi
                                                           Page
APPENDICES

    A     Walking Collector Regression Model	115

    B     Atlanta, Georgia, Field Study -
          August 18-22, 1969	116

    C     Columbia, South Carolina, Field Study -
          June 24-27, 1969	141

    D     Knoxville, Tennessee, Field Study -
          July 14-18, 1969	16O

    B     Medford, Oregon, Field Study - August 4-8,  1969  .  183

    F     Pasadena, California, Field Study -
          July 28-August 1, 1969	2O4

    G     Waukesha County, Wisconsin, Field Study  -
          July 8-11, 1969	226
 TABLES
    1     Description of Satellite Vehicle Waste Collection
          System-- Study Sites	1'

    2     Average Values for Satellite Vehicle
          Collection Variables by Study Site	33

    3     Satellite Vehicle Collection Model  Coefficients
          by Study Site	35

    4     Correlation Coefficients of Satellite Vehicle
          Collection Model Variables	37

    5     Statistical Information - Regression Model
          Variables	39

    6     Accessibility to Waste Storage Location   ....   43

    7     Satellite Vehicle Round Trip Time Analysis  ...   SO

    8     Average Number of Items per Dwelling Unit at
          Bach Study Site	52

    9     Satellite Vehicle Unloading Time Factors  ....   56

-------
TABLES (Contd.)
                                                              Page
  10
        Satellite Vehicle Collection .Model..Coeff icients
11
, (
12
13
14

15

16

17

18
19
2O


Satellite Vehicle Collection Model Coefficients


Satellite Waste Collection Vehicle Specifications . .
Satellite Vehicle Collection Model Coefficients

Satellite Vehicle Collection Model Coefficients
for Hilly and Flat Terrain 	 	
Packer Truck Driver Collection Model Coefficients
All Study Sites 	 ............
Average Values for Packer Truck Driver



Estimated Cost of Residential Solid Waste Collection
Using Satellite Vehicles, Under Actual Community


65
66
67

69

7O

75

77
82
85


90
  21
  22
Cost Estimates for Satellite Vehicle Waste
Collection Service, Identical Conditions Being
Assumed 	  	

Residential Waste Collection Costs and Crew
Efficiencies - Satellite Vehicle Method-vs.
Conventional Method . . .	
FIGURES

  .1     Loading a Satellite Vehicle . . .  .  .  ...

  2     Unloading Satellite Vehicle at Packer Truck

  3     Cushman Waste Collection Vehicle   .  .  .

  4     Trashmobile Waste Collection Vehicle   .

  5     Westcoaster Waste.Collection Vehicle   .  .  .
 92



1O3




 13.

 13

 15 .

 15

 16
                               viii
                                                                                                                                                     Page
FIGURES
6
7
8
9
10
11
12

13

14

15

16

17
18
19
2O
21
22

23

• 24
(Contd.)
. Trash Taxi Waste Collection Vehicle . . . . . . ...
Satellite Vehicle Data Collection Procedure ....
Recording Data on Packer Truck Driver . 	
Collection Time Study Work Sheet .........


Distribution - Dwelling Units per .'.
Satellite Vehicle Load ...... 	 . .
Distribution - Satellite Vehicle Up

Distribution - Distance from Satellite

Housing Density vs. Satellite Vehicle Distance
per Dwelling Unit 	 	 	 	 .
Distribution of Items per Duelling Unit (Once per
Week and Twice per Week Collection) 	 	 . .
Distribution - Distance from Street to Storage . .


Unloading by Hand into. 'Rear-Loading Packer Truck
Unloading into Side-Loading Packer Truck 	
Distribution - Dwelling Units Serviced by

Distribution - Items Collected Between Truck
Moves by Packer Truck Driver 	 	
Distribution - Average Distance Truck to Storage .

16
. 21
21
22
27
28

4O

45

46

48

53
55
57
57
58
58

78

8O
81
                                                                                                                         ix

-------
     SATELLITE VEHICLE SYSTEMS FOR SOLID WASTE COLLECTION
                     SUMMARY AND CONCLUSIONS
     The use of satellite vehicles for collection of residential



solid wastes can be an efficient and economical alternative to



the conventional collection systems currently in use when the



wastes must be collected at the house.



     The efficiency and cost of a potential satellite vehicle



waste collection system can be accurately approximated without



costly field studies or experimental implementation by a commu-



nity investigating alternative collection systems, by using the



collection models developed in this report.




                  System Design and Application





     The efficiency, which this study defines as the total num-



ber of dwelling units a crew is able to service per unit of



time, can be determined by the satellite vehicle operator and



packer truck driver collection models developed in this study.



Usage of these models depends upon knowledge of the effects of



collection frequency per week, housing density and lot size



upon the variables which determine productive collection time



required to service a given number of dwelling units.  The



model which describes the time required to complete one round



trip with the satellite vehicle is:
                                                                                                                - 2  -
                                                                                     v  _ 0.80X. + 0.13X. * 0.007X. + O.086X. + 3.45X, - O.78 * U  .
                                                                                     YT - 	1	-2	3-g	4	5	
where Yj = the total time, in minutes, to complete one satellite



           vehicle round trip, servicing Xj dwelling units



      X  = the total number of dwelling units serviced in one



           round trip



      X  = the total number of containers and other distinct and



           separate items of waste material collected from X



           dwelling units



      X3 = the average distance,  in feet, which the satellite



           vehicle is able to drive up each dwelling unit driveway



    •  X4 = the average distance,  in feet, which the satellite



           vehicle operator must  walk from his satellite vehicle



           to the waste storage location at each dwelling unit



      Xg = the total route distance of the satellite vehicle



           round trip,  in miles



      U  = the unloading time of  the satellite vehicle at the



           packer truck, in minutes  .



      ES = the fraction of the satellite vehicle operator's total



           time on the collection route which is occupied by driving,



           collecting,  and unloading activities.
                              - 1 -

-------
                              - 3 -
     The packer truck driver collection model is:         - .





             . T = O.6Ob  + O.31b  + O.Olib. - O.6O





where T  = the total time, in minutes, to service b  dwelling



           units between truck moves    .



    ' . b  = total number of dwelling units serviced by the' packer



           truck driver between truck moves



      b  = total number of items collected from b  dwelling



           units



*. '     b,, = the average distance, in feet, from the packer



 .          truck to the waste storage location at each dwelling



           •unit
     The selection of proper values for each of the variables in



the collection models is explained in this report.



     The expected cost of a. satellite vehicle system, expressed



as annual collection cost per dwelling unit,-can be accurately



estimated using the crew efficiency and the expected crew costs.



The satellite vehicle collection cost estimation model is:
                      (z-1)c
6QX.(Z-1)
                                                           (N)
where C  .= annual collection cost per dwelling unit



      Z  = total number of men -in crew    '     :



   • •  C  = labor cost in dollars'per man per day
                                                                                                                - 4 -
      C  = packer truck operation, depreciation and maintenance



           costs, in dollars per day



      C  = satellite vehicle depreciation, operation and main-
       5


           tenance costs, in dollars per day



      C  = overhead cost of crew, in dollars per day



      H  = total number of hours per day which the crew spends



          •on the collection route:  This time does not include



           driving times to and from the route, to and from the



           disposal site, formal break times, or lunch time



      Y  = total time, in minutes, to complete one satellite



           vehicle round" trip



      X  = total number of dwelling units serviced in one



           satellite vehicle round trip



      T  = total time, in minutes, required by the packer truck



           driver to service b  dwelling units between truck



           moves



      b. = total number of dwelling units serviced by the packer



           truck driver between truck moves



      B- = the fraction of the packer driver's total time on the



           collection route, H, which he devotes to collection



      N  = the number of collections which the dwelling unit



           receives in 1 year





     The values for Qj. and Cg are discussed in the report,



while values for C  and C  are different for each particular



community.

-------
                              - s -
                       System Capabilities

     The satellite vehicles serviced an average of 5 dwelling
units per load in areas with once-weekly collection and 1O
dwelling units per load in areas with twice-weekly collection
(P 61)-
     Satellite vehicle operators collected an average of 15 items
per load for both once and. twice-weekly collection (p 41).
     The mean satellite vehicle route distance was O.45 miles
per load for once-weekly collection and O.75 miles per load for
twice-weekly collection (p 47).
     The average time required to unload a satellite vehicle was
1.5 min per load or 11 percent of the total trip time (Table 7).
     The time required to service the average dwelling unit
(Table 2) on once-weekly collection by satellite vehicle was
2.11 min.  The time required to service the average dwelling
unit on twice-weekly collection was 1.84 min per collection or
a total of 3.68 min per week (p 62).
     Collection in hilly areas required 15 percent more time
than collection of dwelling units with similar characteristics
in flat areas due to the reduced speed of the satellite vehicles
imposed by steep roads and driveways  (p 68).
     Satellite vehicle operators were able to service more
dwelling units per hour than hypothetical walking collection
crews under identical conditions, - in five of the six sites studied.
(Table 22).
                                                                                                                     - 6 -
                      Crew Activity Analysis

     Values for  E  for  individual satellite vehicle operators
ranged from 6.764 to  l.OOO with a mean of O.95O (Appendices).
     Every crew  observed worked under the incentive system and
the average value for H was 6.O hr per day (p 91).
     Packer truck drivers devoted only 16.1 percent of their
time to driving  and were able to participate in collection 3O.2
percent of their total  time on the route (Table 18).
     Packer truck drivers spent 16.O percent of their time assist-
ing the satellite vehicle unloading process (Table 18).

                Items  Collected per  Dwelling Unit

     Dwelling units receiving twice-weekly collection averaged
1.6 items when garden wastes were not included (Table 8).
     Dwelling units receiving once-weekly collection averaged
2.3 items per collection when garden wastes were not included.
Inclusion of garden wastes increased the items collected per
dwelling unit by.4O to  1OO percent (Table 8).

                          System Costs

     Satellite vehicle operation, maintenance and depreciation
costs accounted for 11 percent of the total cost of a 3-man
crew with two satellite vehicles (p 87).
     Labor wages and fringe benefits accounted for 6O percent
of the total cost of a 3-man crew with two satellite vehicles
(p 86).

-------
                                  - 7. -
         For  identical  conditions  (p.  89)  the average cost  for once-
  - weekly collection was'$19.00 per  dwelling unit  per year and $10.OO
   per ton collected (Table 21).                 '               .
        For identical conditions (p. 89) the  average  cost  for  twice-
   •weeltly 'collection was $28.SO per dwelling unit per year and $15.00
   per ton collected (Table 21).
                   rison of Private and Municipal Systems
        The average  private satellite vehicle operator was able to
   service  the standard dwelling unit  22  percent  faster than  the
   average  public satellite  vehicle operator  (Table  12).
        Packer truck drivers in private crews  deveoted an  average
  of 59.0-percent. of their time to collection.  Municipal packer
  truck drivers used only 1.4 percent of _ their ..time to collect
  (Table 18).

       For  identical conditions the cost  of collection to municipal
  agencies  averaged  53 percent more per dwelling unit  and per ton
  than  that to private collection, due to the higher efficiency of
  the private satellite vehicle operator  and  the participation  in
  collection by the private packer truck driver  (Table 21).   It  must
  be remembered that this does not refer to the cost to the resident
 but to the collection agency.   Cost to the resident would be higher
, because management  costs are hot  considered in this report.

        Qualitative  Evaluation of  Satellite Vehicle  Systems

     Satellite vehicle system users  reported increased morale,
 lower.absenteeism, and fewer injury claims due to the reduced
physical work required of waste collectors using satellite
vehicles.
                                                                                                                     - 8 -
    , Homeowners receiving satellite vehicle waste collection rer.C;
garded its use as a technological ,advance by the collection agent
and this improved the homeowners* image of the waste collector.
     Satellite vehicle usage resulted in littering of streets
in some of the areas in which they were observed.
     Satellite vehicles were excessively noisy in several in-  .
stances.
     Satellite vehicle users reported high maintenance costs and
short useful lives due to the lightweight construction of these
vehicles.
     Satellite vehicle users reported several instances of satel-
lite vehicles over turning • when subjected to sharp turns.

                   Operational Recommendations

     The selection of the appropriate packer truck to provide
maximum operating efficiency in conjunction with the satellite
vehicles is extremely important.  The packer should have a hop-
per of 2 cu yd or greater to accommodate the satellite vehicle
load in one cycle of the packer.  .Slight equipment modifications
may be necessary to provide maximum coordination between the
satellite vehicle hopper and the packer blade during unloading,
and to minimize spillage of wastes.
    . Packer trucks should be radio-equipped for communication
with the garage or central office to inform collectors of missed
dwelling units or. extra pick-ups.  This contact will also mini-
mize down time due to packer or satellite vehicle breakdowns.

-------
                              - 9 —
     Satellite vehicles should be equipped to facilitate waste




transfer to the packer with a minimum amount of spillage.




Rubber or canvas flaps attached to the rear lip of the satellite




vehicle hopper and welded metal wings on the sides of the hopper




at the rear will ease the flow of wastes from the satellite vehi-




cle hopper into the packer hopper.




     Satellite vehicles should not be loaded above the top of




the hopper.  Overloading results in waste spillage enroute to




the packer and impedes the unloading operation.  The vehicle




should be driven slowly enough to prevent paper from blowing out




of the hopper.




     Safety of the satellite vehicle operators should be a




prime consideration.  Visible turning signals, large rearview




mirrors, safety reflectors and extra taillights should be stan-




dard equipment on all satellite vehicles.




     Coordination between satellite vehicle operators and the




packer truck drivers working in one crew is necessary for opti-




mum crew efficiency.  The operators can avoid duplication of




service by establishing a set pattern to be followed each time




that a particular route is.covered.  The satellite vehicles




should work close enough to the packer truck to minimize haul




distances and avoid losing the packer truck.  The packer truck



driver should participate in collection, but should also be




available at the truck to assist in the satellite vehicle unload-




ing operation.
                                                                                                             - 10 -
     A means of transferring wastes from the dwelling unit stor-




age point to the satellite vehicle should be available.  Each




satellite vehicle operator should have a 4O- to 6O-gal light-




weight, manageable container with his vehicle to eliminate




carrying homeowners' containers back and forth to the vehicle.




     Satellite vehicles should be on strict preventive main-




tenance programs to extend their useful life and minimize




mechanical failures during their operation.  The lightweight




construction and small engines of the vehicles are subjected




.to rough usage, necessitating such a program.




     Establishment of good customer relations can be extremely




beneficial to the satellite vehicle operators.  Thoughtful




homeowners will remove obstacles which reduce satellite vehicle




accessibility to the waste storage point.




     Customer containerization of their wastes in paper or




plastic bags improves the handling of the wastes by the collec-




tor and prevents loose articles from falling from the satellite




vehicles when in motion.

-------
                            INTRO DUCTICN






     Since tfie inception-of the-Bureau of Solid Waste Management




 in 1965 much research time and monies have been devoted to the




 development of new and improved methods, for disposal of solid




 wastes, ' Despite the fact that collection and'transfer comprises .




 SO percent-of the total  cost of solid waste handling, there has




 been very li-ttle encouragement for new collection technology on




.the national level.  As  a result, current residential solid waste




 collection practices do  not differ significantly from .those used




 at the turn of the century.  The one-horse open cart has simply




 been replaced by a multiple horsepowered enclosed van with greater




.capacity.




     Residential solid wastes are collected from many different




 locations, depending in  large measure on the existence of suitable




 alleys and on the amount of money available for the collection




 service.  The collection point may be located at the curb, in an




 alley, on the back porch, in the backyard,, at the front or back




 house lot lines, in bassments and even inside houses.  When the




 collection point is anywhere except at the curb or in an alley,




 the collection system is said to be a "backyard" collection




 system.  Backyard collection is the most convenient and costly.




 system to the homeowner and is used in approximately one-third







                              — 11 -
                                                                                                                    12 -
of all communities in the United States.'1  To maintain this




convenience for the homeowner a means of maximizing efficiency




and minimizing costs must be sought.




     One new method currently being used in approximately SO




communities with backyard residential collection is satellite




collection vehicles.  These are small, manually operated, three




or four wheeled vehicles which transport wastes from the dwel-




ling unit storage point to a packer truck for ultimate transfer




to a disposal site (Figures 1 and 2).  These vehicles have capac-




ities of 1 to 3 cu yd and weigh 12OO to 26OO Ib.  The hoppers




are equipped with hydraulic lifts to transfer wastes from the




vehicle to the packer truck.  The vehicles reportedly work best




in areas with single-family homes, where housing density is




medium to low.




     With publicity focused on solid waste disposal research,




this new concept of residential solid waste collection has not




received warranted recognition and publicity.  Several munici-




palities and private collectors reported decreased costs and




increased collection efficiency using satellite vehicle systems.




Since residential collection costs are four times the cost of




disposal, any opportunity to decrease collection costs-cannot




be neglected.




     This study was initiated with the objective of quantita-




tively evaluating the productive efficiency and costs associated




with residential satellite vehicle collection systems.  Quanti-




tative analysis of those factors in the collection process

-------
        Figure 1.  Loading a satellite vehicle
                ^PttijSl&i ". -iSx '- ^-*'; -  v-i •' •: -Ti -5S5S*6


      £?   * '  -   *>Tv*<   •'ftfei-'' '    '   '  '->   ' • "   '•* • .
      '  W4 «ii? •-'rff's^—v5^.-"- '•-- '"•   ••••«*•-  '•",-•'.' '..»•••
         • v& -s ^sz^ffm.    ,:u..^  'irfS3»w:;*«si^
Figure  2.  Unloading satellite vehicle at packer truck


                                                1   NOT REPRODUCIBLE
                                                                                                              - 14 -
                                                                                 which effect the efficiency and economy of  the system would be


                                                                                 undertaken to provide a practical and reliable basis for design-


                                                                                 ing future collection systems using satellite  vehicles.   To


                                                                                 allow comprehensive investigation of these  factors,  four makes


                                                                                 of satellite vehicles operating in a wide range of conditions


                                                                                 were chosen for the study.  The four satellite vehicles  studied


                                                                                 were the Cushman, the Trashmobile, the Westcoaster,  and  the


                                                                                 Trash Taxi (Figures 3-6).  Six systems were chosen for the study


                                                                                 based on their diversity in vehicle type, terrain, collection


                                                                                 agency and collection frequency (Table 1).

-------
                      - 15  -
    Figure 3.  Custartan waste collection vehicle
Figuce 4.  Trashmobile waste collection vehicle
Figure S.  Westcoaster waste  collection  vehicle
Figure 6.  Trash Taxi waste collection vehicle

-------
:-  17  -







a
(«


ds
H tn

>5
igj
5- i?
§2
0
§h
S a
o

Q












•3
u
01
H

E
O >
•ri O
O O
•3!
(u
i-i
o
•rl
^ >
Of Q)

H S
0
x
B *ri

+> J3
0) a
r-l B
r-l O
38-
o
u




01
"• £
T3
11
 §
< <>

a


t
i
u
P




E
a
1
a
3



d
a
o
• •rl
E



E
•H
rH
S
3»
3 *
" OJ
•S ^
J3 01
§
"o I
O -0
^
•H
•H


*
0)
i
u
B
O


o
I-I
•ri
1
O
d
h
H




o
§
•ri
Kt
O.





0)
0)
m
Ss
S *"
H rt
f<8
•® rH
i-l ^
•rt r^
^
g rt
S^ "I
^ rt 13 *»
d rj B a
i-l *rl 10 H
 «U
1-5 1
•>s >
B? B
0 +* 0



0)
*»
I 1

• co co
3 s



•3
o) a
d ^
•ri B
W 3
0. B




a
d *
Irt"
Is S^
f - 5a
*3 «-i

*^ +J W M
O CD *g
0> •) d 3
Z< 0..?

d
H

                                                                             sryoy PROCEDURE




                                                         Prior to the initiation of the actual study,  two pilot


                                                    studies were conducted in the greater Cincinnati,  Ohio area.


                                                    These studies were conducted to test collection time study proce-


                                                    dures which had been previously developed by the Bureau for the

                                                                      2
                                                    Territory of Guam.   These time study procedures were altered


                                                    significantly to correspond with the operations peculiar to


                                                    satellite waste collection vehicle systems.   During these


                                                    studies it became obvious that there was little correlation be-


                                                    tween weight of solid waste collected and time required for col-


                                                    lection.  Collectors took longer to collect three containers


                                                    with 1O Ib of waste in each than they did to collect one con-


                                                    tainer with 3O lb of waste.


                                                         Based on this observation, it was decided that the number


                                                    of items collected and the number of dwelling units serviced


                                                    would provide higher correlations with collection time.  Corre-


                                                    spondingly, since the object of waste collection is to service


                                                    the most dwelling units in the least amount of time at the low-


                                                    est cost, it was decided that annual .cost per dwelling unit


                                                    would be more practical and meaningful than the traditional


                                                    cost per ton.  The author of this report strongly believes in


                                                    this concept and thus cost per dwelling unit is the ultimate


                                                    measure of cost used throughout the investigation.
                                                                                 - IB -

-------
                              - 19 -
     The. evaluation of the six satellite vehicle systems was .  •.. .
                    .•  . •         '_'•••'••      .'->   V          '
conducted in two distinct phases;   the first phase was the deter-

mination of system costs;  the second phase consisted of,1-week-

long studies of satellite vehicle operations at each site.


                      Phase One - Cost Study


     Meetings were held with the owners of the private collection

companies and with the superintendents of the municipal collec-

tion systems studies to obtain the costs associated  with their

satellite waste collection vehicle systems.  The associated

costs obtained were crew hourly.wages and fringe benefits; satel-

lite vehicle depreciation, maintenance and operating costs; packer

truck depreciation, maintenance and operating costs; and.the

overhead required for the system.  Overhead costs were not gener-

ally known by the collection agent and therefore a factor of 2O

percent of all other costs was assumed for the collection system

cost computations' and analyses.  This assumption is  based upon

known overhead percentages from other collection systems with  -

accurate accounting procedures.       -            •     .  .

     Adequate collection system cost data were readily available

in all of the study areas with the exception of Knoxville,

Tennessee.  Collection system cost calculations for  Knoxville

were based on cost estimates from the Trashmobile manufacturer's

literature.
                                                                                                                       2O
                   Phase:Two - Field Evaluation        :


 .   ' Comprehensive field evaluations of  satellite vehicle collec-

tion systems.were conducted to determine the effect of selected

factors on the efficiency of the system.  The field evaluation

at each, study site was 4 to 5 days in length.  At least one col-

lection crew using satellite vehicles was studied at each study

site.  The study team accompanied the collection crew during

their entire work day to account for any efficiency variation

due to fatigue at the end of the day.

     Three-man study teams were used on  all of the field evalua-

tions.  This allowed accompaniment and observation of two satel-

lite vehicles and the packer truck.  The activities of the

satellite vehicle operators and the packer truck driver were

timed and recorded in detail.

     Motorcycles were used by the study  team members to maintain

a close contact with the satellite vehicle operators at all

times.  These members were .equipped with stopwatches and clip

boards to facilitate data collection (Figure 7).  The third

member followed the packer truck in a car and was equipped with

several stopwatches to record all other  necessary times (Figure 8).

     Collection Time Study Work Sheets.  The "Collection Time

Study Work Sheet" was used by the study  team to record the neces-

sary data associated with .the activities of the satellite vehicle

operators (Figure 9).  Each sheet corresponded to one round trip

by a satellite vehicle to and from the packer truck.  A discussion

-------
Figure 7.  Satellite vehicle data collection procedure
   Figure 8.   Recording data on packer truck driver
                                                                                                            -  22  -
                                                                                                           Figure 9

                                                                                         COLLECTION  TIME  STUDY  WORK SHEET
                                                                         COLLECTOR-
                                                                         END ODOMETER READING-
                                                                         BEGIN ODOMETER READING.
                                                                         TOTAL DISTANCED	
DATE      ,
ELAPSED TIME_
UNLOADING TIME.
OTHER TIME	
SERVICE
.. 1
2
3
4
5
6
7
8
9
-- IO
II
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
3O
NUMBER
OF ITEMS































AVERAGES

DISTANCES (feet)
STREET TO
STORAGE






























,

SCOOTER UP
DRIVEWAY.

•





















SCOOTER
TO STORAGE























i














;
ACCOMPANIED













1









I


1






-------
of each item on the work sheet will familiarize the reader with




the factors evaluated and the method of recording them; ,....•




     Collector - the name of the satellite vehicle operator




          being observed        .




     End odometer reading - the reading on the motorcycle        .




          odometer when the satellite vehicle returns to  the




          packer truck to unload                      .'"•*•'




     Begin odometer reading - the reading on the motorcycle




          odometer when the satellite vehicle leaves the  packer




          truck to begin a round trip




     Total distance - the difference between "end odometer




          reading" and "begin odometer reading", in miles




     Date - the month and day on which the~data~ were recorded




     Elapsed time - the total time which elapses from the time




          that the satellite vehicle leaves the packer truck




          until the time that it has returned to the truck and




          unloaded.  This is the time of the satellite vehicle




          round trip cycle.  Time was recorded to the nearest




          O.I min.  This time was recorded by the study team




          member following the packer truck."




     Uhloading time - the time required for the satellite vehicle




       :   to empty its entire load into the packer -truck.  The




          time starts when the satellite vehicle .arrives  at the




          packer truck.  The time ends when the satellite vehicle




          hopper is emptied and lowered to its original position.
                                                                                                                - 24 -
          Unloading time and other time are subtracted from




          elapsed time to yield productive time per load.




     Other time - the time spent on activities which do not con-




          tribute constructively to the process of waste collec-




          tion.  Some examples are:






          1)  talking to homeowner




          2)  getting a drink of water'




          3)  receiving instructions from the packer truck dr.iver




          4)  picking up roadside litter




          5)  waiting for homeowner to open garage door




          6)  lighting and smoking a cigarette




          7)  salvaging




          8)  other miscellaneous activities not part of the




              actual collection process






     Formal break times normally allowed to the crews are hot




considered as "other time", but are accounted for in the ensuing




analyses.




     Productive collection time - total elapsed time per satellite




          vehicle round trip minus unloading and other time




     Service - a solid waste storage location.  One service can




          include the wastes from any number of dwelling units.




          An apartment building with four families which has one.




          storage location is one service, but four dwelling




          units.  The number of dwelling units was recorded in




          parentheses when there was more than one per service.

-------
                          •25
Number of items - the total number of standard solid waste




     containers, other containers and distinct pieces of




     wastes collected by the satellite vehicle operator at




     each service.  This variable was sometimes broken down




     to measure the effect of garden wastes and particular




     types of containers.




Street to storage distance - the estimated perpendicular




     distance in feet, from the edge of the street to the




     point of waste storage for each service.  This distance




     was estimated to the nearest 1O ft by the study team.




Scooter up driveway distance - the estimated distance which




     the satellite vehicle traveled on each dwelling unit




     property.  This distance was estimated to the nearest




     1O ft by the study team.




Scooter to storage - the estimated distance which the oper-




     ator walked from his satellite vehicle to the point of




     waste storage at each service.  This distance was esti-




     mated to the nearest 1O ft by the study team.




Accompanied - this column was checked if the study team




     member on motorcycle accompanied the satellite vehicle




     from the edge of the street to the point where it stopped




     at the service.  If checked, the corresponding "scooter




     up driveway distance" was doubled and subtracted from




     "total distance" to yield the actual route distance




     that the satellite vehicle traveled along the road




     during each round trip.
                                                                                                             - 26 -
     Data by - the name or initials of the Bureau member who




          recorded the data






     In addition, each work sheet was numbered to record the




order in which they occurred.  The work sheet was given to the




member of the study team following the packer truck by the data




recorder on motorcycle, each time the satellite vehicle returned




to the packer truck (Figure 1O).  The elapsed time for the satel-




lite vehicle round trip was recorded when the satellite vehicle




had completely unloaded.






     Driver Collection Time Data Sheet.  The "Driver Collection




Time Data Sheet" was used to record the activities of the packer




truck driver (Figure 11).  Only one or two sheets were necessary




to record all of the driver*s activities each day.  In addition,




the sheet was used to record general collection crew information




required for the study.  Each item on this form will be discussed




to illustrate its purpose.






     Date - the month and day on which the data were recorded




     City - the name of the community where the study was being




          conducted




     Pickup point - the location of waste storage at collection




          time, for the crew being studied.  The pickup point




          was classified as backyard in all the situations




          encountered during the study.

-------
Figure 10.   Transferring data sheet at packer truck
                                                                                                                     /- 20 -
                                                                                                                   Figure 11


                                                                                                 DRIVER ^COLLECTION TIME DATA  SHEET
                                                                         DATE-
                                                                         CITY.
.VEHICLE NO..
_' frf-E	_
                                                                         PICKUP POINT-
                                                                                                         CAPACITY..
. CREW SIZE;	
. TOTAL SERVICES-
. TOTAL ITEMS	
 .DATA BY
                                                                         BEGIN.ODOMETER. READING;.
                                                                         END ODOMETER  READ!NG_L
                                                                         TOTAL DISTANCE _i______
                .ASSISTANCE TIMEj
                WAITING TIME__
                OTHER  TIME	
                                                                         TOTAL' COLLECTION
SERVICES

























1

L 	
NUMBER
OF ITEMS











..-
















DISTAKCE
St TO STOR.




•









	






•



'

\
ELAPSED.
TIME





















• ,






SERVICES




























.'•.'UMBER
OF ITEWS




























DISTANCE
ST. TO SfOK.



























.
ELAPSED
TII.-C





























-------
                         - 29 -











Vehicle no. - the identification number of the packer truck




     being observed




Type - the make of the packer truck body




Capacity - the size of the packer truck in cubic yards



Crew size - number of satellite vehicle operators, walking




     collectors and the packer truck driver




Total items - the total number of. containerized and* uncon-




     tainerized waste articles collected by the packer truck




     driver




Data by - the name or initials of the Bureau member follow-




     ing the packer truck




Begin odometer reading - the odometer reading of the car




     following the packer truck, recorded when the packer




     truck reached the first stop on the collection route.




     The odometer readings were taken to the nearest O.O5




     mile.



End odometer reading - the odometer reading of the car at




     the last stop before going to the disposal site or any




     other destination off route.




Total distance - the difference between begin and end odom-




     eter readings.  This is the route mileage of the packer




     truck.




Total collection time - the total number of minutes that the




     packer truck is on the collection route.  This is the




     time between the first and last stops which the packer
                                                                                                                  - 3O -
     truck makes on the collection route.  Formal break




     times are not included.




Assistance time - the time which the packer truck driver




     spends in assisting the satellite vehicle operators




     in unloading wastes into the packer truck and instruc-




     ting the operators in their collection activity.




Waiting time - the time in minutes which the packer truck




     driver spends standing or sitting at the packer truck,'




     waiting for the return of the satellite vehicles.  This




     is non-productive time for the driver.




Other time - the time in minutes which the packer truck




     driver spends on any activity which prevents him from




     being available to assist the satellite vehicle opera-




     tors upon their return.  A list of common activities




     is given in "other time" for satellite vehicle operators.




Services - the number of solid waste storage locations




     serviced by the packer truck driver between truck




     moves.




Number of items - the total number of distinct container-




     ized and uncontainerized articles of waste collected




     by the packer truck driver between truck moves.




Distance street to storage - the average distance in feet,




     which the packer truck driver walked from the truck to




     the waste storage point for each service collected




     between truck moves.

-------
                         - 31 -
Elapsed time - the total .collection time required by  the ; •..



     packer truck driver to collect the number  of services



     given on the same line, in minutes. .
                     FIELD STUDY DATA ANALYSIS






     A total of 1O50 satellite vehicle loads and 592 packer




driver trips were observed and recorded during the field studies.




The data obtained were subjected to standard statistical analyses




and extensive multiple regression analyses.  The primary objec-




tive of the data analyses was to determine the effects of selected




measurable factors upon satellite vehicle collection crew .effi-




ciency.  Accurate establishment of these- effects could then pro-




vide a~ sound basis for the design of future satellite vehicle




collection systems.




     The study results were based upon observations of twenty-




six satellite vehicle operators and twelve packer truck drivers




working in six distinct areas of the country.






              Satellite Vehicle Operational Analysis






     Individual stepwise regression analyses were performed on




the data from each satellite vehicle operator observed.  These




results are presented and analyzed in the individual Appendices




for each study site.  Analyses were also performed on the total




data obtained at each study site (Table 2).  These analyses




yielded equations which described the productive collection time.




required for each'satellite vehicle load.  The productive collec-




tion time, Y, was best described by the following variables:
                                                                                                            - 32 -

-------











«
m
a
ce
5

1

J
1 O.
9
Sg
PM
0)
Id >*
-J U Q
H H a
< w H
m >•
Hffl
K
£
en
1
td

1

























rH
£1
(0
•H
>




























. 01
.. 3 O-T)^
Q^ O (0 C
s^-
Du
01
O +• 0)
§01 Di-~
oi oaf
eo** Vi v vi HH
Q CO +> O -^
•H CO +>
Q co
01 T3
O eg ^~
O C O 01
« ^5 S S,
•o ti.^

01
U 01 01
^ CO O O CO "t>
X *> -rl V VI IH
CO £ O ^-"
•H 01 •»>
O > CO
01 >,
u o a
g |.o g.|^
•H 'S Tl *^
o > •o

•a-a
s a
CO -H O
x" § grt
M tH 0>
o a
u
o» "0*0
c o a
•H CD U O
x^S trt
« C H h
5 3 Q> Q»
Q a> a
Q> -a
+* 0) 0)
•H iH (0 >
,-1 O T) H
IH -H a 0)
01 j: o «
*> Gl i-H .D
<8 > O


m

oO
eneo min-eo inrn -o •n
OO rHOO OO OO




8OOOO OOOO
m m N « .H rH H



§o ooo ooo o
tCioBo eNSi--o>
rH rH rH rH




S2S32 22 S3




coeMrHOO enmmo
rH rH iH


oe>j 'jeoeo eno t^tn
rH S rHCMrHO

0) 10
en C •§
0) Vi 0)
'3 * S 1 I § 5 ^ 1

OHO SH8o qc°
0-C'c.c - s 5 S- 5 1 s
^ 11 s s | s " s- -
5 -p 8 1 « 1S 2
< O ^ X ft. 3 <















I





















;

i






                             . - 34 -
X., the total number of dwelling units serviced during one round




trip of the satellite vehicle; X , the total number of items col-




lected from X  dwelling units; X , the average distance, in feet,




which the satellite vehicle is able to drive on each dwelling




unit property; X  , the average distance, in feet, which the satel-




lite vehicle operator must walk from his satellite vehicle to the




storage point of each dwelling unit; and X , the total route




distance, in miles, which the satellite vehicle travels during




the round trip.  The values of these variables depend upon col-




lection frequency, housing density and lot size in the area to be




collected.  Thus a community can design a satellite vehicle col-




lection system without expensive field studies or trial implemen-




tation, simply by knowing the effects of these three factors




upon the five variables which determine productive collection




time required for a satellite vehicle load.




     A regression analysis of these same five variables against




the productive collection time for all 1,O5O satellite vehicle




loads observed during the study was performed.  The resultant



equation was:






  Yp = O.8OX  * O.13X2 + O.OO7X3 + 0.086X  + 3.4SX5,- O.78  (1)






     This model was able to explain 81.5 percent of the total




variation in the data, and the standard deviation of the residu-




als was 27.1 percent of the response mean, productive collection




time.  The model coefficient values of the five variables were




very similar to those for the individual study sites (Table 3),

-------
d

M
• ."












0)
rH

5
'.N~
a



























C "O
V O 0)
C-H C
8 V -rl
O  O
-•r*
o IE
X^ o) O
C +*
O
O
01 1)
U 10
o c o
in v a IH
X g v ^
• '' ? ? a
Hi
U O 0>
C rH OV
f a o o ffl
X «• -ri f • X
•H "§ S
. Q > . 0)
ffl >>
U 0 <0
X|« 1
Q > TJ
'S'D

. 2 "tj Q
X § O rH
•M i-l .
M rH rl
o ®
o a

? "2 "2
C (Pa
. -H a o~ C
X rH -rl >
O C h h
i 3 o w
a • « a





V


o

.£
V
tn


• ; t-
CO
'r*-.

--S
O)
1


rH
GO
^f


i
*
o


e»
O
O



- in

O




§
rH
*








•^ *H
W X
t? a
8- H
O '£
a
5 H
c :
0)
*»

•o
fH
^_ •

O
rH



O>
^
*

in
0
O


rH
N
O
O



^
rH
0




S
o







(V

ft
JD
^"
0
tf)
H


^
f!


O)
in



OJ .
O
en


i
i


i





i





O
I-l
rH
*



. *H
8
(0
U

JC
•*>
g
CO

a
•H
A
Col urn

„
2
•!•

m
O



in
rH


O
O


8
0



•o
•e
o




i- —
t- •
o




8
(A


C.
C
H
*".
0)
rH
«H
•H
Knoxv

,. -
g


S
m
i


s
en


i
o


o
s
o



Ol
^
o




§
o








c
o
v
o
i
1

in
CO


M
rH
1
^

0
m


rH
O


tn
0



o
CM
O





ot
*

0
o


rH
rH
0
O



tn
rH
O




S
O
c
to
c
8
m
*H
3


>,

c
3
O
O
t
CD
to
CO
3

in
rH
GO


CO
O
1 .


m
en


-o
8
o


8
O



en
rH
O




O
CO
*°





•a
Q)
c
•rl
*§
8:

01
4-*
•H
a
rH
rH
















.
8
+*.
<».
•H .
U
0
h
a
•
v
•H
^
•o
2
CD
rH
to
corn
>
j-
o>
•H
£
^
0)
0
E


rH
CO
•H
S
*

                                                                                                             - 36 -
despite the dissimilarities in operating conditions at each site.




The, correlation coefficients between variables illustrates their




independence (Table 4).




   •  The confidence to be placed upon the use of the regression




model presented above'is'dependent upon the quality, quantity and




distribution (Table 3) of -the variables which are the basis of




its development.  The fact that the model covered 81.5 percent




of the total variation in the input data and that the P value




was 922.1, .is a significant indication of its accuracy and qual-




ity.  The quantity and distribution of the input data.becomes




obvious in the analyses of the regression model variables and




other factors which effect satellite vehicle collection system




efficiency.




     Regression Model Analysis.  The regression model consists




of five variables which accurately describe_the productive col-




lection .time required per satellite vehicle load.  A discussion




of each of these variables will illustrate their significance




and will indicate the values to be used for design purposes.




     X , Dwelling Units Per Satellite Vehicle Load.  The number




of dwelling units serviced in a satellite vehicle load was the




most significant variable in describing Y , the productive time




required per satellite vehicle load.  Since the satellite vehi-




cle has a fixed capacity, Xj increases as the number of items per




dwelling unit•decreases.  Therefore, X  is highly dependent upon




collection frequency,-since items.per' dwelling unit per collec-




tion increase-as the-number of collections per week decreases.-

-------
                             - 37 -
                             TABLE 4

              CORRELATION COEFFICIENTS OP SATELLITE
               VEHICLE COLLECTION MODEL VARIABLES
Variable
                                   Variable
              l.OOO
                         O.518      0.489     -O.186     O.4O7
 0.518      1.000      0.393     -O.179     O.O56

 O.489      O.393      l.OOO     -O.OO8     O.135

-O.186     -O.179     -O.OO8      l.OOO    -O.112
              O.4O7      O.O56
                                    0.135     -0.112     1.000
                                                                                                                      - 38 -
                                                                           The satellite vehicle hopper  capacities  observed ranged from 1.25
                                                                           to l.SOcu yd.  This limited range did not allow any direct correla-
                                                                           tion between number of items per load and hopper capacity.
                                                                               There is a high correlation  between collection frequency
                                                                           and number of dwelling units  serviced per load.   For once  per
                                                                           week collection the satellite vehicles collected an average  of
                                                                           4.53 dwelling units and 14.42 items per  load.  Where twice per
                                                                           week collection was in effect, the satellite vehicles collected
                                                                           an average of 9.78  dwelling units and 16.1O items per load.
                                                                               For design purposes the  value of X-  is determined  by  the
                                                                                       formula:
                                                                                                    Xl =
                                                                                                                        15
                                                                                                         average items per dwelling unit
                                                                               The number of dwelling units per load is also dependent
                                                                          upon the types of wastes which'the collection agent is respon-
                                                                          sible for collecting.  Where crews are required to collect gar-
                                                                          den wastes with household wastes, the number of items per dwel-
                                                                          ling unit increases as much as 1OO percent, and thus the number
                                                                          of dwelling units per load is reduced  significantly.   Many
                                                                          communities designate garden wastes as "trash" and collect it
                                                                          separately.
                                                                               The average number of dwelling units serviced per satellite
                                                                          vehicle load was 5.71 and the coefficient of X  was °-80 (Table 5)-
                                                                          The wide distribution of observed values for X  lends statistical
                                                                          validity to the coefficient value determined by the regression
                                                                          analysis (Figure 12).

-------
          - 39
                                                                                                     - 4O -
          TABLE 5
 STATISTICAL  INFORMATION
REGRESSION MODEL VARIABLES
Variable
>1
X2
*3
X4
Xs

Correlation
coefficient
with Yp
O.818
0. 538
-O.O64
0.483
O.661

Coefficient and
95 percent , Mean _
confidence value «ange
interval '- •
O.8O ± O.O6 . 5.71 1-25
O.13 ± O.O27 14.8 1-43
O.OO67 ±O.OO22 92.7 ft O-3OOO ft
6.O86 ± O.O11 14.7 ft O-3OO ft
3.45 ± 0.3O O.5O O.O-4.7
. miles miles
                                                          t    ' -:.r;:nrr_:.i^:-: -::/:'  :.:.:-..(.

-------
                              - 41 -







     X^, Items Per Satellite Vehicle Load.  The number of items


collected per satellite vehicle load was fourth in order of


importance for describing productive collection time, Yp.  The


correlation coefficient between productive time,' Yp, and X- was


 0.538 for the 1.O5O loads observed (Table 5).


     X  is dependent upon the hopper capacity of the satellite


vehicle and the size of the items collected.  The satellite


vehicles observed had an average capacity of 15 items per load.


     The study observations included 2,31O dwelling units reev-


ing twice-weekly collection and 3,684 dwelling units receiving


once-weekly collection.  The mean number of items per satellite


vehicle load .was 14.8O and the coefficient of the variable, X ,


was O.13 (Table 5).


     X t Distance Satellite Vehicle Travels on Dwelling Unit


Property.  The average distance which the satellite vehicle


traveled on a dwelling unit property was the least important of


the five variables used to describe productive time required per


satellite vehicle load.  The correlation coefficient for X.


and the productive time per load, Y , was -O.O64 (Table 5).  X


is dependent primarily upon dwelling unit lot size and waste stor-


age location, and secondly upon homeowner habits and practices.


The relationship between lot size, waste storage location and X


can be expressed as follows:
D_- = O.83X  + O.76X  + 11.4
 SS        3        4
                                                          (2)
                                                                                                    - 42 -
                                                                      where D   = perpendicular distance from edge of street to waste


                                                                                  storage location, in feet


                                                                            X   = distance satellite vehicle travels on dwelling unit


                                                                                  property, in feet


                                                                            X   = distance from satellite vehicle to waste storage


                                                                                  location, in feet
                                                                           The relationship between homeowner habits and X  is called


                                                                      accessibility and is expressed as:
                                                                                                     Y  + V
                                                                                                     X3   X4
                                                                                                                                (3)
where A = accessibility,  expressed as a fraction




     D   and A may be estimated from a visual survey of the col-


lection area.  X  and X  can then be determined by simultaneous


solution of these two equations.


     Accessibility averages in the six study areas ranged from


O.727 to l.OOO (Table 6).  Low accessibilities are indicative


of obstructions in driveways, such as automobiles, or poor stor-


age locations, which force the operator to leave his vehicle and


walk to the storage area.  Collection efficiency is reduced when


wastes are stored in basements, at the rear of homes and at the


bottom of stairways.


     Five thousand, nine  hundred and ninety-four individual satel-


lite-vehicle-on-property  distances contributed to the 1.O5O aver-


age values per load for X .  The mean value of X  was 92.7 ft and


its coefficient was O.OO67 (Table 5).

-------
43
                TABLE 6
ACCESSIBILITY TO WASTE STORAGE LOCATION
Study site
Atlanta
Columbia
Knoxville
Medford
Pasadena
hilly
flat
Waukesha
All cities
Distance
. vehicle
up
driveway
(x3)
9O
8O
1OO
9O

:'12O
100
7O
9O
Distance
traveled
''-• to
storage
<*3 * X4>
120
110
120
110

130
100
80
110
Accessibility
tx^rV
1*3 *X4/
7 0->50
0.727
0.833
0.818

. 0.923
^.Jl.OOO
Or875
0.818
                                                                                     - 44.-
                                                             The observed values for X_ were well distributed from O to


                                                        ISO ft and ranged from O to 3.OOO ft (Figure 13).  The value of


                                                        X  was zero at 15.3 percent of the dwelling units serviced.


                                                        This meant-that -the waste was located at the street or that


                                                        some obstruction prevented the vehicle from entering upon the
                                                              >.,_.'.

                                                        dwelling unit property.  When either, case occurs, the purpose of


                                                        using satellite vehicles is defeated.           "                .


                                                             XA, Average Distance from Satellite Vehicle to Storage.


                                                        The average distance from the satellite vehicle to the waste


                                                        storage location for the dwelling units serviced in one load


                                                        was the third most important factor in describing the productive


                                                        time required.  The correlation coefficient for X^ and Yp, the


                                                        productive time required per load, was O.4S3 (Table 5).  X  is


                                                        dependent upon accessibility to the waste storage location and


                                                        may be determined by simultaneous solution of equations (2) and


                                                        (3) discussed previously.  For a given dwelling unit, as X  in-


                                                        creases accessibility approaches zero, and the satellite vehicle


                                                        operator assumes a role similar to the conventional walking col-


                                                        lector, since satellite vehicle usage, X , is minimized.


                                                             The values for X  ranged from O to 3OO ft with a median of


                                                        1O ft (Figure 14).  The optimum distance for greatest efficiency,


                                                        O ft, occurred at 47.0 percent of the dwelling units observed.


                                                             The average values for X  in the six study areas ranged
                                                                                     4

                                                        from O to 30 ft (Table 2), thus providing sufficient testing of


                                                        the regression model over this range.

-------
ff      *

-------
                             ... 47  -











     X  , Satellite Vehicle Route Distance Per Load.   Route dis-




 tance per .satellite  vehicle  Ipad was  the second most  important




 factor  in describing the productive time required.  The  correla-




 .tion coefficient -between X5  and Yp, the  productive  time  required




:per load, was 0.661  (Table 5).  The route distance  is a  function




 of housing  density in  the collection  area, total number  of dwel-




 ling units, serviced, and the proximity of the satellite  vehicle




 collection  area  to the packer  truck.                   ....



     X_ can be estimated best  by using the housing  density of the




 collection  area.  A  relationship between route distance  per dwel-.




 ling unit  and housing  density  was  developed from the  study data




 by plotting housing  density  versus X^ divided by X_ at each site




 (Figure 15). Housing  densities  in the study areas  ranged from




 4O to  14O dwelling units  per mile  of street.  The route  distance




 per dwelling unit increases  as housing density decreases.




     For a given housing  density the route distance logically




 increases as the number of dwelling  units serviced per load




 increases.   As  stated earlier, the number of dwelling units per




 load is a function  of  collection frequency.- The mean route dis-




 tance  for once-weekly collection areas was O.4S miles per load,




 while  the mean  for  twice-weekly collection was O.75 miles per,




 load.                                               .-.-..'



     The route distance can be kept  at a minimum by operating




'the'satellite vehicles as close to the packer truck as possible.




 .If the average  route distance of O.SO miles per load was increased
                                                                                                                 -  48  -

-------
                               49 -
SO percent by not working close to the packer truck,  the produc-.




tive time required per average load would be increased by O.86




min.  This would cause a decrease in collection efficiency of




approximately 1O percent.



     Y , Productive Collection Time Per Satellite Vehicle Load.




The productive collection time required for one satellite vehicle




round trip is dependent upon the five variables discussed and




the particular movements of each operator.   It has been illus-




trated that the five variables in the model  explain approximately




8O percent of the variation in productive time values, with a




standard deviation of 27 percent.  The remaining 2O percent of




the variation is unexplainable because of particular movements




peculiar to the individual operator.



     The productive time per  satellite vehicle load only  includes




that time which productively  contributes to  the collection pro-




cess.  The unloading time of  the satellite vehicle and "other




time" must be taken into account to  obtain the actual  total




elapsed  time per  satellite vehicle round trip.  These  two fac-




tors are discussed in  the Analysis of Other  Variables  section.




Productive collection  time accounted for  approximately 84




percent  of the  total trip time for the  average  satellite vehicle




operator (Table 7).



     Analysis of Other Variables.  In addition  to the five vari-




ables  included  in the regression model  there are other factors




which must be taken into account before a total system may be
                   - 50 -












                 TABLE 7




SATELLITE VEHICLE ROUND TRIP TIME ANALYSIS
Study Site
Atlanta, Georgia
Trash Taxi
Trashmobile
Columbia, South Carolina
Knoxville, Tennessee
Medford, Oregon
Pasadena, California
Central
West
Waukesha County, Wisconsin
Average for all sites
Percent of total time per round trip
Productive

87.9
92.2
82.3
87.1
82.2

SO. 2
82.3
77.5
83.9
Unloading

8.6
7.0
8.3
8.4
11.4

14.9
13.7
16.6
11.1
Other

3.5
0.8
9.4
4.5
6.4

4.8
4.0
5.9
5.0

-------
                                51 -






designed.  The number of items per dwelling unit, distance-from .




street to storage, unloading time and other "time'all.contribute




vital information and effect the five variables used in .the col-




l^jction model.  A discussion and analysis of each of these




factors will establish their significance, in the satellite vehicle.  .




system design process.




   .  Items Per Duelling Unit.  The number of items to be'collected"




at one dwelling unit is a function of the collection frequency and




the limitations on types of wastes which are to be collected by ..




the collection agent.  As collection frequency increases,.the




number of items per dwelling unit per collection decreases corres-




pondingly.  Where garden wastes are collected with household wastes




the number of items per dwelling unit will be increased significantly.




     The three study sites with once-weekly collection averaged 2.3




items per dwelling unit when garden wastes were not included and




3.9 when they were (Table 8).  The three study sites with twice-




weekly collection had separate collection of garden wastes and




averaged 1.6 items per dwelling unit (Table 8).  During the study,




2,31O dwelling units with twice-weekly collection and 3,684 dwell-,.




ing'units with once-weekly collection were observed.  The distribu-




tion of items per dwelling.unit illustrates significant differences




when plotted by collection frequency (Figure 16).  Two items per.




dwelling unit was the most common occurrence for once-weekly  .




collection and 1 item, per dwelling unit the most common occurrence




for twice-weekly collection..
- 52 -









J^
§
g
§
5
D U
s »
3 10
E.
<
g 1|
H H X
** S


B
3 <


Z
M
1











0)
s
** +
-H T
'o :
c
N t
at i.
S c
*:.
- T3
O
C U
•0 r-l
a o
Cl U
• 2 'S
Jl



§x
•••g-
9 '.!
* '
VI' 1
. s
0 >
•rt (
*> (
jT

6





- .«.

 m N en '« m v
• H ' rH ' CJ rH M M pj ,

i -. :- ; ..'. ...
i .
1 II 1 ' 1 * * •
• cn ^» • >.










« « rH C« rH rH n n
. - . ' •




•. i. '.;.-...• ....-•.
C :
•« :. .• • U >,
C C X rt
••H O . rH . ,v
:. rH 0 J< 0
-O '?". . 0 .rf
Ot+^-C O *H *» C S'
* 3 ID 0, rt c 0 5
0 0 H- * «g .
-0> U) -.M'U o ><' M^
W • .7 O o ,0 c o c
•». -» • Q) - » - Vi o Vt o
•cd rH -* « Id -H '-rl
01 -I* rH TIC i <»*• » f
*» £ M H 0 n OIUDIU
C E > O Tl O <3B « 41
0.13 x O >O
< y « s_ £s 
-------
               ib^
                              --£-•
   i  '        _•      	.    	"^i   •    „_   r	*  j
 ...	         -      .   i    	      ,  ,
~"t;zrzi^z^i~~ ri~^"inizii;!i'irr_n~ .^.iiiir.-". ...-i N

r.
                              - 54 -









      Communities which allow garden wastes to be combined with



 household wastes can expect as much as a 1OO percent increase in



 the number of items to be collected per dwelling unit (Table 8).



 Garden wastes accounted for 30 percent of the items  collected in



 Pasadena and SO percent of the items collected in Waukesha County.



      Distance from Street to Storage.   The perpendicular  dis-



 tance from the edge of the street  to the waste storage  location



 was recorded for each dwelling unit  observed during  the study.



 The distance from street to storage,  D  ,  is used to determine X
                                       oo                       3


 and X4 (p 41).   Dgs may be estimated from visual  observation of



 any collection  area.   It is dependent upon house  lot  size and



 customer practices.



      The mean street  to storage distance of  the 5,994 dwelling



 units observed was 99.4 ft.  The values for  D  ranged from O to



 3.OOO ft with a median  value of 8O ft  (Figure 17).



      Satellite Vehicle  Unloading Time.  The time required to



 transfer wastes from the satellite vehicle to the packer truck



 constituted approximately 11 percent of the total time that the



 satellite vehicle operators were observed  (Table 7).   Unloading



 time  is  dependent upon the satellite vehicle hopper loading prac-



 tice and  dumping mechanism, the hopper size and compaction cycle



 time of the packer truck, the amount of assistance provided by



 the packer truck driver and queuing at the packer truck  (Table 9).



     Satellite vehicles have several methods of unloading  and may



be used in conjunction with front,  rear, or side-loading packers



(Figures 18,  19,  2O,  21).   The  four makes  of  vehicles studied

-------
56' -  '











in
§
b.

M
H .-
S
H
:•§' '.
.O
' - C 'r'''"

B)
d
M
| ' ..'.-
w











• -(rt
00)
1 ^
o
a
S8§
*j'*^ ^j
ft. -0 «
0)
d
h
|1
~ H 0)
O N
3-3
d
D.
* " " ' : '
2 '
g S S
:;s.a3



tO r'...
+* O
•H iH
i-t U
i-l -H
",!*
•v.W'^'





V
•H
•H -
O
'"X
4*

tH
CM



b






|H





O




TJ
a
o
4*
d
VI
Cuahman {











Ov
O



!§
rH





iH





n






9 •
iH
Traahmobj




«
a
B
o
c
c
o
H
0
Knoxvill
CO
•H'



^
tH





rH





CM







Cuehman







c
o
Ol
o
o
Medford,
en
tH



$
^




o
en





O






H
O
Weatcoast




a
c
o

•H
tH
a
U
Pasadena
CD
O



15
in




o
CM





o






M
0
Westcoasi











r-
rH



§






f-l





0)
«






Cushman
c
•H
n
0
U
n
•H


^
C
J
Waukesha
in
.-5



$
tH




(s.
I-l





3











0)
Q)
•H
0)

•"*
in
M
Average











;•



>»
4*
• *rl
O
d
&
S
h
0
O.
1
O
^J
•H
. C
•H
4*
3
O
.c
4*
•H

O
if
a
a
c

*o
0
tH'
1
a
•o
•H
(A
* •

-------
                         - S-7-
Figure 18.  Unloading into front-loading packer truck
Figure W.   Unloading into rear-loading packer truck
                                                                                                                              -58-
Figure 20;  Unloading  by  hand  into rear-loading packer truck
                                                                                                    Figure 21.  Unloading into side-loading packer truck

-------
                             - 59 -




were equipped with hydraulic dumping mechanisms which had varying

dump lifting times.  Also'a flat-bed unloaded manually was ob-

served in Columbia, South Carolina.     '                     .      .

&- ,-.Overloading..the . satellite .vehicles with .too many items

resulted in an increased unloading time.  Overloading required

additional compaction cycles of the packer truck and a concerted

effort by the operator to minimize waste spillage during the

unloading process.   In most cases IS items were collected per

load ~to maintain low unloading times.  Columbia and Waukesha

County crews required higher unloading times  due to overloading

 (Table 9).            .         .                       .,'.'.'

     Unloading time is very much dependent upon the hopper capa-

city of the packer truck and  the length  of the compaction cycle.

Trucks with hoppers -mailer than 2 cu  yd require two  or  more

 compaction cycles  to accommodate the wastes  from any  satellite

 vehicle.   The length of  the compaction.cycle becomes  important

 when multiple compaction cycles are required.  The packer trucks

 with 2 and 3 cu yd hoppers used in Pasadena, California were able

 to accept the wastes from the satellite-vehicles  in an average of

 1.1 nin,  while trucks with smaller hoppers averaged 1.6 min

 unloading time.      -

      The amount of time which the packer truck driver spends

 assisting the satellite vehicle operators during the unloading

 process is also important.  Assistance from  the packer driver

 allows the satellite vehicle  operator to remain in his vehicle.

 The packer truck  drivers remove waste stuck  in the satellite vehi-
                                                                                                                      - 60 -
cle hopper and, clean up the area around the truck after the satel-

lite vehicles .have left.  The drivers observed during the study

assisted the satellite vehicle operators from O to 6O percent of

their time.  Assistance time of 20 percent of total time on the .

route would be adequate for a crew with two satellite vehicles.

Queuing at the packer truck causes an increase in unloading time

as satellite vehicles wait to be unloaded.  Queuing could be

easily eliminated for crews with only two satellite vehicles by

proper coordination between the two operators.  For three or more

satellite vehicles per crew, some queuing is unavoidable but can

be minimized by proper operational coordination.  During the

study, queuing occurred more frequently than appeared necessary

under the circumstances.  Operators used this waiting time for a

brief rest between trips and an opportunity to talk with each

other..      •         .    .

     Following .the operating procedures outlined above, it is

estimated that all. of the vehicles observed could unload in approx-

imately 1 min if used in conjunction with a packer truck with a

hopper of 2 cu yd or greater.

     Adding unloading time to productive collecion tine yields

the total time required to complete one round trip, with the sat-

ellite vehicle, assuming there is no "other time" involved.


      Ym. = O.8OX  + O.13Xm + O.OO7X, + O.O86X. + 3.45X, - O.78 + U
      *M
                                           ,  4
where Y  = minimum round trip time required to service X.
  .         dwelling units

      U  = satellite vehicle unloading time

-------
                             - 61 -







     Other Time.  Inclusion of non-productive time or "other


time" in the equation will yield the total time which any opera-


tor will require to complete one round trip with the satellite


vehicle.  This  is  expressed as:


       O.8OX, + 0.13X_ + O.OO7X. + O.O86X  + 3.45X  - O.78 + U
...  _.	1	2	 3 '	4	5	
where E_ = the fraction of the satellite vehicle operator's total


           time which is occupied by collection and unloading


           activities



     The average values of B  for each study site ranged from


O.9O6 to O.992 with a mean value of O.95O.




    Effects of External Influences on Collection Efficiency



     In addit ion to the. variables which must be considered


within a satellite vehicle system, there are outside influences


which are not under the control of the satellite vehicle operator,


and affect collection efficiency.  The design of the investiga-


tion allowed effects of collection frequency, collection agency,


satellite vehicle type, collection area terrain, type of item col-


lected and weather conditions to be evaluated.


     Collection Frequency.  Increased collection frequency enables


more dwelling units to be serviced per satellite vehicle load, due


to the decreased number of items per dwelling unit.  During the


study, the satellite vehicles operating on once-weekly routes


averaged 4.53 dwelling units per load while those on twice-weekly
                                                                                                               - 62  -
routes averaged 9.78 dwelling units per load.  To service: the,-


average dwelling unit once-weekly requires 2.11 min.  To  service  ,


the average dwelling unit twice-weekly requires 1.84 min  per col-


lection or a total of 3.68 min per week.  Therefore twice-


weekly collection is not justified on reasons of economy  or


efficiency, but should be used when~ health considerations


warrant it.


     Separate regression analyses based on collection frequency


produced models with significantly different coefficients (Table


1O).  For once-weekly collection, X , the number of dwelling units


per load, is the most highly correlated variable to productive


time.  The small coefficient for number of items collected pro-


duces only a minor increase in productive time required to col-


lect an increased number of items per dwelling unit.  For twice-


weekly- collection, X., the number of items collected per  load,


is the most highly correlated with productive time.  Small


changes in items per dwelling unit have a significant effect on


the productive time, due to the large coefficient, 0.47, of this


variable.  Therefore, conmunities with more wastes per dwelling


unit would theoretically increase efficiency significantly by


using once-weekly collection.


     Use of the regression models based on collection frequency


is not necessary since the general model takes into account the


effects of different collection frequencies.   More frequent col-


lection increases variables X  and X_, decreases X , and vice-


versa for less frequent collection.   Tnc general model explains

-------
1-63 -
                                                                                 - 64 -


.
M
K
















O
i-l
0
d
14
d


























c -o .
•H O O
C -H C
01 -H -rl
Odd
*4 -H rl
o w p.
01 5 8
§
O +* H '
C +*
5
0> TJ
u d
e» c o

. o 
O 01 01
C rH 0>
• T4 4> H
 «



O X
o oi d
C rH $
en d o o. 5
X «* -H 3 >
. -S « °»4
D > t>

•a
• - 01 •O
a +* d
& IS2
•P rH
M rH V<
80)
a
O" T) T>
c oi d
•H O U O
X rH -H >

6. a> p.

C
>. o
0 -H
C - V
§08
5T *^
S "o <
b. o



"1
S

. .„
a>
en
d



•o .
m
en

	

S
o
•
0





m
8
d




«
o
• '
o



CM
d
*


8-

s

a
o
c
o


CO'
{;



en
en



t-
00
en



t-
o

d





r-
°
d





*




2
d


x
S


0
a

0

S


"J---
3



C1*
d
i


in .

en



•o
-CO
O

d





1
d




en
tH
. d



S



a
01
u


fa
' \4 -5
,0
rt O
< 0


















a
s
M

O
^
*J
a
3

1
a
c

t
•o
0)

d
tH
o
h
- M
8
rH
£
.:. ^Ot
. f.

S
O
,O
d
••H
1
*



                                                    a higher percentage-of the variation in the total data than each



                                                    of the models based on collection frequency (Table 1O).



                                                         Collection Agency.  Regression analyses were made on the



                                                    data separating private and municipal satellite vehicle opera-



                                                    tors, to determine any differences in efficiency (Table 11).  The



                                                    six  agencies observed operated under the incentive system, thus



                                                    eliminating efficiency differences due to this factor.  Substi-



                                                    tuting average values for variables X  through X  (Table 3) and



                                                    values for E  and U from Table 12 into the private and public
                                                                • S       -             • •               .     •


                                                    operator models shows the private operators to be 22 percent fas-



                                                    ter than the public operators.



                                                         Satellite Vehicle Type.  Of the four types of satellite



                                                    vehicles studied, three had the required capabilities to cope



                                                    with all types of route conditions. The  Trash Taxi was observed



                                                    during its first week of operation and it did not have adequate



                                                    power to climb hills and steep driveways at a reasonable speed,



                                                    thus significantly slowing down the collection process.



                                                         The expertise and collection habits of the individual



                                                    satellite vehicle operator far outweighed any slight advantage.



                                                    one particular vehicle might have.



                                                         The specifications for each of the four vehicles observed



                                                    are very similar (Table 13).  .The larger hopper size and faster



                                                    unloading mechanism of the Trashmobile would appear to give it



                                                    an advantage over the three-wheeled vehicles,  but this advantage



                                                    is eliminated by its lower mobility and maneuverability.  Sepa-

-------
- 65 -
                                                                               - 66 -
                                                                              TABUE 12
c-o












 -M -H
« o to rt
K M -rl iH
Q) H CU
a 5 «
P OP.








a

° C ||
J U
SQ CD

0 <0 h
K-rl «
•o o.
O
o o> o>
SrH Ot
O O O
X *» TH 

o> >>
U Q) 01
C rH 3
cna o a 01
X *• -H 3 >
0 £ -H
•H 01 rl


^
O TJ
o> *• a



M rH rl
8V
a
E "S "S
•H m U o
rHrH +> -rl rH
X rH *rl >
JB rl rl
3 01 0)
to a.

. B
o
II -H >>
II *" °
II ° C
Sg,
II "-i o
1- 3

CO

•<*





rH
O
d


o>
in
CM



CO
8
O


m
8
d





en
rH

d




•o

d
*





Q)

a

rl
0.

rH «

0 rH
O> CO




00 00
t^ t-
rH O
I 1

Tf 'in
H •**
^» en "



d *O
00 00
0 C)
d d


§ 8
• •
o o





TJ en
« rH

d d



m o
CO 00

0 0
* *






u
•H

1 2

COMPARISON OF PRIVATE AND PUBLIC AGENCIES





Y »
U T
Agency E- . . . . . .
S (mm) (mm)


Private O.944 . 1.5 1O.O

Public . O.947 1.6 12.9


All O.945 1.5 11.3
e
••H :
a
•H ' *Using average values for variables X through X^ (Table 3).
•0
.§
o
D.
*
O
**
d


M
M
0
U
f-t
•H


•M
0)
o
€
V
T4
h
(0

*




-------















-
H
s
H
ft
-a
id
S. :

1
8.
cd
• H

S
cd
P!
s












.0
.01
§'
-
"O
- *tn •

-'s



•rl
H

in
(0
(H




0)
rH •— *
•H C
JS 3
Trashmol
(Dats




i
0) -
3




CO -
. in X
in -CM
. . o
rH O .rH CO in
m t? . •- •
>.
,8

c
5S
H 1
in co
« « „
• H X
H. . m **
o in

m
in

^
in
. X
rH CO CM ^
" * 8

CO
x

in m o
CM t- t^
• * • '•
.00 , ,|
m

TJ * •
>, C C
- v -H -H
*>. rt
S £ £ 01
o o> 01 j:
g. "Si "oS ' * ' o
8£ £  E o>
. 0 > 3 £
. . j o z -. S
:-

„ 	 v




rH
CO"







^ .
CO






•
o>
o>





CM




__.
c
•H
0)
a
o

rH
0
O
-fi.
67 -

1

8 '


•<*>
•rt


C

Ci
«

•o" •>-
M O
•o u " ~~
.§ o , • • • -
*» \
'•" ' * - .
/; *-* \o co m '-
•o o m rH;-
O O rH
O CM
O, CM
ID *
1 H
CO
'•, ';:•


^« :tf* •
S " ^
fr.
1
CM_ "



•6
1
1
0 • o m CM o
•o o « « CM
§ * rt!
e- <*
•
•0 . ' •'
M
•n
C
a
» ... '«-
TJ CM m in m
0) 00 i-t Tf rH
0 M rH .
0 rH

B» .-. -. .'•.. . •
T T ^- '„•
ja c • .**
rH -H . C \^
*«•«''** o
. C v G i-<
C 01 0> *> 0
O -H C TJ K
-< 0) 5 -H -H
0 S r-l S O
o>-
•H O rH rH Ol
E . rH rH rH ' C
0) U (8  .O O H





.' r



































                              - 68 -







 rate regression analyses  for  each vehicle  did not  reveal any


 efficiency'differences, with  the exception of the  Trash Taxi


 (Table 14)..              .


      Since all the  models have coefficients of the same order


 of magnitude and the vehicle  type does  not affect  collection,


 there is no advantage to  the  use of  the separate models, with
           .*

""the exception of the Trash Taxi.


      Collection Area Terrain.  Separate regression analyses were ' •


 performed on the data from hilly collection routes and flat col-


 lection routes (Table 15). the  regression model coefficients for


 each are very similar, with the  model for  flat areas yielding


 slightly lower load times for standard  conditions.  The time


 required to complete one  load in flat areas was approximately 15


 percent less than the time that  would be required  to service the


 same houses in a hilly area.  The longer time required by the


 satellite vehicles  operating  in  hilly areas was partly due to


 the reduced speeds  of the vehicles when negotiating steep drive-


 ways and roads.


      Homes located  in hilly areas of the.study were usually high


 income types and tended to be farther from the street than houses


 on flat terrain. The average distance  traveled up the drive-


 ways of homes in hilly areas  was 1OO ft, but was only SO ft for


 homes located on flat terrain.   The  longer distances from the


 street to the storage point for  homes in hilly areas also resulted


 in longer walking distances for  the  satellite vehicle operators.


 Operators walked an average distance of 2O ft from their vehicle

-------
                                        TABLE 14



                     SATELLITE VEHICLE COLLECTION MODEL COEFFICIENTS

                                    BY VEHICLE TYPE
                                           Variable                          '



  Vehicle      DweUing       *     Distance   Distance     **         *°     vlr"^on

                units     JJS-  -hiol°    VehtCle    cSce   ««*«*   e^ained

               portal   p*ri°*d   d^u   .»£...    p~ioad        	.



Cushman   ^      »0.76       0.16      0.006      0.091"""    4.14      -2.45        82.3
(301 loads) .
                                                                                                  i

                                                                                                  s
Trashmobile      #074        76<2            ,
(308 loads)




Westcoaster       Q 63       Q 20      Q QQ.,      0>107      »5 Q4      _1>22        84tS

(337 loads)




T'"h,Ta''i.      »1.06       0.25      0.019      0.047       4.81      -2.69       , 87.7
(104 loads)                    -         •                  •                            •



Combined
(1,050 loads)    *0.80       0.13      O.O07.     0.086       3.45      -0.78        81.5.





     •Most highly correlated variable to productive collection time.
                                         TABLE 15



                     SATELLITE VEHICLE COLLECTION MODEL COEFFICIENTS

                              '  FOR HILLY AND FLAT TERRAIN
                                           Variable


                                                  V          V          *r          RI
  Terrain       Dwelling               Distance   Distance      f                 Percent
                 .,»<<-o       items     ..^h^!^    ,.«».4--i_    Route     Constant
	;;r;;s   perioad.   d.i^.^   .t,,^.   p°rioad     tern	


^506 loads)       *°*73        °*12     0>006      °'074      4*13      "°'9S       85'9
(544 loads)       *°.82     ,   0.16     0.005      0.071      2.76      -0.03       75.3




                                     t



(l?050eioads)     *°'80        °'13     0'°°7      °'086      3'45      -°'78       81'5
     •Most highly correlated variable  to productive collection time.

-------
                             - 71 -
to the storage location in hilly areas and only 10  ft .in flat




areas.-    .        -                      •             ...--,-




     Unless an area has exceedingly steep  hills and driveways,




similar to those.observed in Pasadena, California,  the regres-




sion model for all areas combined should be used.   The variables




X , X , and X, account, for the longer distances to  be expected




in hilly areas.  The special mo<3el for hilly areas  should, only




be used when reduced satellite vehicle speeds are expected, from




excessive gradients on collection routes,  streets,  and driveways.




     Type of Item Collected.  The type of  item which is collec-




ted by a satellite vehicle operator affects collection efficiency.




Containers without handles and having excessive container weight




require more time to transfer to the satellite vehicle from the




storage point and unload.  .




     The items collected in Knoxville, Tennessee were classified




into three groups to determine the effect  of the type of  item •




collected on collection efficiency..  The three groups were  55-




gal druns, standard waste containers, and miscellaneous-items.




Of all the items classified in this manner, 68 percent were




standard waste containers, 19 percent were miscellaneous, items,




and 13 percent were 55-gal drums.       •-.-...




     A regression analysis was made with the items  classified




into three groups.  The equation with the item classification




included is:                           ,                    .




Yp = 0.59 +-0.71X  + 0.22X2c +.O.30 *2b + 0.14X2m + O.O3  Xj +2.04X5
                               - 72 -






where Y    = productive time required per load




      X^   = number of dwelling units serviced per load




    .  X2c  = total  number of standard .containers collected




      X2b  = total  number of 55-gal  drums collected "




      X2m  = total  number of miscellaneous items collected •




      X4   = average distance,  in  feet,  from the satellite




            vehicle to the storage  point at each dwelling unit




      Xs   = route  distance, in miles, which the satellite vehicle



            travels per load                                .





     This  equation was able to explain  7O.4 percent of the  total




variation  in the data,  and the standard deviation of the  residuals




was 19.8 percent of the response mean,  productive time.   The coef-




ficient for 55-gal drums collected, O.3O,  is more than twice as




large as the coefficient for miscellaneous  items collected, and




35 percent greater than.the coefficient for standard containers



collected.  Thus,  whenever a 55-gal drum is used instead  of a




standard container,  the pickup time per item is  increased by 35



percent.           .




     Miscellaneous  items were  usually paper or plastic bags or




cardboard boxes which were tossed into the satellite vehicle




hopper, leaving no  container to be returned to the storage point.




The small coefficient for miscellaneous  items, O.14, illustrates




the lower collection time required for  such  items  as compared to




standard containers which-must be returned  to their original




storage position.

-------
                             - 73 .-
Fifty-five gallon drums are cumbersome, difficult to empty,




potentially dangerous to the health of the collector and the




homeowner, and should be restricted from usage as waste con-




tainers in all communities.  If a satellite vehicle operator




collects 15 55-gal drums per load instead of 15 standard waste




containers, the productive time required to service 6 dwelling




units would be increased 14 percent.




     Additional study is being conducted by the Bureau on the




effect of type of container on collection efficiency to deter-




mine the optimum container size and weight from the standpoint




of the collector.




     Weather Conditions.  All of the field studies were con-




ducted under ideal summer weather conditions.  The satellite




vehicle capabilities were not evaluated in the winter, but




users reported no difficulties until there was more than 2 in.




of snow on the ground.  It would appear that.ice or larger accu-




mulations of snow could seriously impede the efficiency of satel-




lite vehicles to the point where a conventional walking collector




could be of equal or greater efficiency.




     In areas of high temperature and humidity, the satellite




vehicles offer relief from fatigue and heat exhaustion by reduc-




ing the amount of walking that a conventional walking collector




normally does.
                                                                                                                - 74 -
            Packer Truck Driver Operational Analysis






     Twelve packer truck drivers, working in conjunction with




the satellite vehicle crews were observed and evaluated during




the study.  Five of these drivers collected wastes in addition




to their primary driving duties.  .Of the five drivers that




collected wastes, four acted as conventional walking collec-




tors and one used a satellite vehicle.




     Regression analyses were run on the four drivers serving




as walking collectors individually and together, to describe




the time required to service the dwelling units collected by




the driver between truck moves (Table 16).  The productive col-




lection time, T, can be accounted for by the variables: b , the




number of dwelling units serviced by the driver between truck




moves; b2> the total number of items collected from b  dwelling




units; and b3> the average distance from the truck to the dwel-




ling unit waste storage location.  The values for these variables




for a particular community are determined in the same way in




which the similar variables in the satellite vehicle collection




model are determined.




     The packer truck driver collection model resulting from the




regression analysis of 592 data points for the four drivers was:
                                                                                                  T = O.6Ob1 + O.3U>2  *
                                                                                                                                - O.6O

-------
                           - 75 -
                                                                                                                76 -.
8
g

$
i
6
c -o
+» O O ' -
„ si .5
M -,G). +* *rt
• -o a .a
rl iH iH
01 VI p.
0-5 S?
> 01

+;
o° 1 1
.O CD -W
C +»
Q

-O 0) O V
Ov 0 +* »
flj C cfl
^SSiS S
< *H 2 0)
*O -H


» 1 »
o 8 +>
/« «4j g

o
0 i
4. 5
•H S «
O 3 01 S
J3 O
•HU Ol E
O 0> C T>
,O M 0> Ji
E iH O O
9 iH -H :
•po**


01
'irl
Q
- ' .' -'
'O> CO ' rH ' . Ol Ol
o> co t^ *f m





O> CO rH t-- O
M Ol . l~ CO . 
cO
_j
cbrre!
rH
•H
0)
i
. *


     This equation was able to explain 75.9 percent of 'the

total variation in the data, and the' standard deviation of the

residuals was 31.8 percent of the response mean, productive

collection time. .The model coefficient values are similar to

those for the individual drivers (Table. .16), indicating the

validity of the model, despite different operating conditions*

     Regression Model Analysis.  The regression model consists of

three variables which accurately describe the productive collec-

tion time required to service b  dwelling units between truck

moves. ' ~A discussion of each of the variables will illustrate

'their significance and now they may be estimated for design

purposes.

     b_ , Dwell ing -Units Serviced .Between Truck Moves.  The number

of dwelling units serviced between truck moves by the truck

driver was the second most important variable in describing the

productive time required.  The correlation coefficient between'

b  and Y  was O.691 for 592 observations.

     The average number of dwelling units serviced between truck

moves was two (Table)}?).   The values for b.. ranged from one to

five, but the drivers rarely collected more  than 3 dwelling units

between truck moves (Figure 22).

     b2, Total Number of Items Collected Between Truck Moves.

The total number of i teas collected by the truck driver between

moves was the most significant variable in describing the

-------
                                                                                    - 78 -
- 77 -





I
«§
O M
b H
w M
g8
AVERAGE
!UCK DRIVEF
H
£








rH
a
M
a








0
o e -H
H •§ +1 -5.
o
'o, -
V 01 O O
O» u -r» C»
*8 c (tf ^~*
cn w o J< » +>
J3 0) +> o O  III 3 -r> —
< -H H 0)
•a *>
" 1 ?
CM $ " 0
il B -rl <0
3 H
o

•H O
^Ili .
M^O,*!
.0 0( C TJ
J3 -rt 0) JS
6 iH O 0
3 iH -r) 3
2 0 > S
01
a
•H *•
0 C
| 0.
£ (S

»
•rl

O ^» CO •O rH
CM CM CM iH CM

O O O O O
o* co co in co

m ^» en CM m


iH « ' CM CM CM


CO CO O « CM
f-t CO PI rH O*
•o
rH CM iH 0) -H
m m o o A
CM CM CM CM 6
Q U I Z O
2 S £ Z 0


-------
                               - 79
productive  time  required to  service.b.  dwellingunits.".The   . .   .


correlation coefficient between b  and'.T was 6.755.

     ,.ijje.,average number o.f items..cpllected between truck moves


was  three  (Table 17).  The values  for b  ranged from O  to  14  with

a median value of 2  (Figure  23).                    ..'.'•'..••••


     B  , Average Distance From Packer Truck to Waste Storage  '.'''•'•'


Location.   The average distance from  the packer truck to the

waste storage  area for b  dwelling units serviced between  truck


moves was the  least  important  of the three variables used.  The


correlation between  bn and X was 6.365.            .         •  . • .
'  -'    •           .-     3    \  -  V     .                .  •
     The average value for b  was  8O  ft and the range was -from O

to  160  ft  (Figure.24).. The  drivers usually serviced  dwelling

units with shorter street to storage  distances than those  serviced


by  the  satellite vehicle operators in the crew.

Packer  Truck Driver  Activity Analysis.    All of the 12  packer


drivers observed during the  field  study devoted less  than  2O


percent of their time to driving the-track" (Table.'18).   Therefore,


SO percent  of  the driver's time is available for  collection or

assistance to  the satellite  vehicle operators. To effectively


assist  a crew  of two satellite vehicle operators  should require

approximately  2O percent of  the driver's time.  Of the  remaining

6O.percent, 45 percent should  be devoted to collection, 1O per-

cent to waiting  at the truck and 5 percent.to other time.  Bach


additional  crew  member would require 1O  percent of the driver's

-------
                                                                                    - 82 -
-o-
    O  to














>.
K •
5
B
3

=.
FH

•
y
H
fj;
5
X
S*
&•
fit
o
tn
H
S

§
*





















•o
01
>
VI
to
.0
0

. e
•H
4^

rH
10
O
4-»
VI
o
f
B
S
U
VI
8.







T
1
i
(
1
i


















H HI
flj g
j=-H
S"



TO
C HI
*rl B
+* »H
•H *>
S
S


0}
U
B
to 0)
S-H
Tf *"
to
to
<

c
o
•rl
+> 0)
S-S
rH +•
fH
O
O

Ol
B 0
•H B
> -H
fi **
U
J
: **»
S 2-2
i O «
)







Of
H-»
•H
m
>i
TJ
3
•e
(0




r^- o> -
rH





cn in cn
O rn cn
f

1^ ^ rH
Tf Ch CO



cn m m




g
•rl
rH
O 01
H a
5(0
m
a a>
•H c q
0> *• E
n 3 o
o o H
• OJ C/l
o
o
<0 rH .
a .'r-f r-l
*J ^3 -H
C E >
.a a X
rH rH 8
•f* O B
< o 2
cn t» ^
t> in cn
rH




CM PI CO
CO « O
O)






co c^ in
cn o p»
rH ^





oo •<* o>
in d i->
•* tn

f 0 Oi
to t> in



cn in M


*
c
•H
to

U
a to
•H -H
C 3
VI
o
C ±i >,
&Ti +>
iH C
G» (3 3
Vi U O
O (J

• C3 10
•0 C £
Vi « n
o -a o
vi a X
*O 0) 3
o is a
S B. 3
r»
rH
rH




CD
T»







O
rH
rH





O
a
m

rH
^
rH




0)
O
•*4
U
S
at
«

a>
«•>
a
>
•H
U
a
V4
O
SH
&
a
VI
V
>
<
rH Ol
O CO




m co
CO CO
in M






rH O
rH 
+j
U -rl
-H 91
rH
^3 IH
i «
h n
O 0
Vt <^4
V 0)
at o>

•H
•o
q
•H
C
"O

U

3
d>
)-<
U
*




X
u
g
D»
H
Or
€
v
+* ".
3

•o
0)
M
•H
3
0*
0)
tH
01
V4
S
>
•H
fi .
+• a
91
.X
to
Vi
A

-------
                               83 -
time, correspondingly reducing the time available to;the driver




.for collection.                                           ,  •..



   ... All of the packer truck drivers in the private agency crews




did some collection.  These drivers collected an average of 59.O




percent of the available time, while the amount.of time devoted




to collection by the municipal packer truck drivers was negligi-




ble (Table 18).
                                                                                                               - 84 -
                      SYSTEMS .COST ANALYSIS






     The cost information obtained from the six collection agen-




cies studied was compiled and analyzed to determine the cost of




residential waste collection using satellite waste collection




vehicles.  The daily crew costs, annual collection costs per dwel-




ling unit and collection costs per ton were calculated for each  --




of the study sites to evaluate the economics of satellite vehicle




waste collection.






                        Daily Crew Costs






     The costs associated with satellite vehicle waste collection




systems are labor, satellite vehicle operation and depreciation,




packer truck operation and depreciation, and overhead.  The




costs reported in the six study areas varied widely due to dif-




ferent economic levels, price and wage indices, and accounting




methods used across the country (Table 19).  The average daily




crew cost was $130.82 per day, for a crew of two satellite vehi-




cles with operators and a packer truck with a driver.




     Labor. -: The wages, and fringe benefits of crew personnel




varied significantly from one section of the country to another




(Table 19).  Labor costs ranged from $79.97 to $195.62 per day.




The wages and fringe benefits paid  to the packer truck driver




were usually slightly higher than those of the satellite vehicle




operators.  The packer truck driver was frequently designated as

-------
- 85 -

















O
M


g


Id
J
M


rH

cd H
,_} M
Eg j
g d
S

IX
8
U)
te
8

*g
IX







































ffl
•3

&

U)

r-f
rH
S











H
5
o
(H
+
TJ
10

O

O


§
. 8





X
$

+*
M
O
.3
o
£





tn

o
•H

0)
•*•»
•H
rH
rH
O
•H
a
m
i
H

g
•H
flj
J-t
S.
o

g
•H

Hi
•H
U
0)

a


c

•H
(0
0)
Q,
a
c
o
-H
-H
flj
-H
^
Vo
O.










CJ
•H
01

t
U)




co « . r> eo o . ' -o* o» O «
CO CN) -O* tn ^ 00 *O f*l GO
. »O rH 0* rf \O • f> tfl 00 O
rHrH h* N - f*l " COO* CO PI
rHrH. rHrH rHiHiHrH


CO IA CO f> tS + . + O O)
O* 00 co O d ro »O
fH rH rH CM 0) W rH



. CO tO in IAN rHiHO tA
is is ' in " o " ^ r> is d co'
. rH rH



O *O (D O O rH CO
^ ^J r> O Is- • O CO

IA *A lA O **! O» O
fH r-t rH
* *
* *
O to
d  « r- in
ot m 
0
rt rH
O M O O CM ^* CO
oj *o o cj in m CM
m ^ m o* fi rH t^
f-H




c
fH

c c
fH O
rH U
O Q) ffl' 0) 0)

— ffl 0) C S P
O U 0) rl *••<
(0 fH rH O O * CO
•HX'^'.CCC'H ^
Oi0^PCO*n pi— t
>*HQ3OCflrH SM
oSoHOIffl — 3ffl
5c. c. u) K ej x— o
0)0) -O *-H-r»OVl
a a - a .i-ia o
• MM ffl rH ~ (0 fHrH ffl
nj (-, H fH rH V C Xb. £ 0
p x^< 	 J3 fH rl QJ 	 ^^* CO Ol
c e > o -a a a
a 3 x vi 0 •* «
rH rH O V 10 3 O
P O -C tt ffl ffl >
< S S S 0. 3 <
U O
O 0)
> rH
•H ffl :o
•OB ' c
O -H
ffl £
S C
•£ .• S
•HO %
: 9
G) C
V c - o
o 5 j2
3 O
4-> U) J3
MO P
OP O
X a c
U rl
ffl O 0)
a S. rl
O CO
ffl
O TJ
•as c
C P ffl
A
•H . 01
CO O 0) tO
H p Ol
O*O <0 rl
p 01 O ffl
a p u £
M m u
V *rl %
5. 01 0) rH
O C H ffl
O UP

P rl O
•H S A) M
S » £

to u o o P
0 to
rH O rH "O O
U £ rH o U
fH -P CO fO
£ ^ ffl g
> 8 o A
O O ffl
01 £ P H rH
PS C ffl
fH S "0
rH C U ffl C
rH -H KG ffl
0) tO O O
P C • O, T) P
a o u a G
to U 0> O 0) O •
to > o» « i
O -rl fH 0. Q.
S 3 rl 0 M
p 1} .0 X • 3
XI C cr
a >>^ o o a
ffl P U P 73 fH
C 3 .OP G
0 3 H •D P ffl fH
H Q P O rl fH
01 O POUT)
£ M ffl (X O O
ffl 01 B Q) rl "O
TJ £ JC -5 HO, 3
O 01 U P O rH
c oi co tn to t3 u
•H J< O. 0) P C
IH 3 to -a -H
«ffl O 0) O C
3 £ ffl U ffl 0)
PS -H
01 C PC
•H -rl E T3 CO T3
ffl ffl O TH ffl
S P 0) § p Q)
rl O O rl fH VI H
UUP 0) 3O O
x u > 0*0. >
< o a o m o o
* S t


                                                                                    -  86  -





                                                       the crew foreman  or  chief, and  was responsible  for  the crew's .




                                                       activities.




                                                            Labor costs  accounted for  an  average of  6O percent  of  the




                                                       total costs  associated with a three-man satellite vehicle col-




                                                       lection  crew.




                                                             Satellite Vehicle Operation  and  Depreciation.  The opera-




                                                       tional costs of satellite vehicles depend upon  the  operator




                                                       treatment and maintenance program  to which  it is subjected.  Oper-




                                                       ational  costs averaged $3.29 per satellite  vehicle  per day  for




                                                       the areas investigated.  Maintenance costs  comprised the major




                                                       portion  of the operating costs  as  the  satellite vehicle's light-




                                                       weight construction  and small engines  are exposed to much abuse.




                                                       The four-wheeled  vehicles observed had lower maintenance costs




                                                       than the three-wheeled vehicles due to their heavier duty con-




                                                       struction.   Fuel  costs are very low since the satellite  vehicles




                                                       average  approximately 8O miles  per gallon and seldom travel more




                                                       than 2O  miles per day.  Operating  costs accounted for an average




                                                       of  5 percent of the  total costs of a three-man  satellite vehicle




                                                       crew.




                                                             The depreciation costs of the satellite vehicles depend




                                                       upon their initial cost and assigned useful lives.  Initial costs




                                                       of  three-wheeled  satellite vehicles ranged  from $2,10O to $3,OOO




                                                       depending upon the number of vehicles  purchased and the  optional




                                                       equipment desired by the purchasing organization.  The four-




                                                       wheeled  vehicles  cost approximately $3,6OO.  The useful  lives

-------
                             - 87 -
of the vehicles ranged from 2 to 5 years depending upon the .  .




amount, of abuse the vehicles were subjected to and'the-preventive




maintenance programs of the owners.  The average depreciation
m.-.  'I----—  *»••>. r._ 2-—  •-*  -  -.•	--             =    r          •--• j.


cost was $3.64 per satellite vehicle per day.  The depreciation



cost accounted for an average of 6 percent of the total cost



of a three-man collection crew.



     The average satellite vehicle operation and depreciation



costs accounted for 11 percent of the total crew cost in the




six systems studied.                          •




    .Packer Truck Operation and Depreciation Costs.  Packer




truck operating costs depend' primarily upon the size of the




truck chassis and the abuse which it is given. ..The average oper-



ating cost for the packer trucks operated in the study areas



was $10.33 per day.




     Packer truck operating costs accounted for an average of 8




percent of the total costs of a three-man satellite vehicle col-.




lection crew.                                          .  .



     The depreciation costs of a packer truck depend on the ini-




tial cost and the estimated useful life.  The cost varies pri-




marily as a function of the capacity of the packer unit.  The.



most commonly encountered packer trucks: had 18 and 2O.cu yd



capacities and an approximate initial cost of $15,000.  Useful




lives assigned to packer trucks ranged from 4 to 1O years in•the  .




six areas studied.  Packer truck depreciation averaged $12.O2 per




day and accounted for 9 percent of the total crew.cost.
                                                                                                                            - 88 -
     The average packer truck operation and-depreciation costs




accounted for 17 percent of the total satellite vehicle waste




collection crew costs.




     Overhead.  Overhead costs were not available from five of




the six collection agencies studied and therefore it was esti-



mated .to be 2O percent of all other crew costs.  This figure is



based upon known overhead costs from communities with adequate



accounting records.  Based on this assumption, overhead cost



averaged 12 percent of the total satellite vehicle waste col-




lection crew costs.






            Annual Collection Cost per Dwelling Unit






     The homeowner judges the effectiveness of any residential



waste collection system by the cost which he must bear for the




particular level of service he receives.  As the level of ser-




vice increases, the cost must increase correspondingly.  Back-




yard collection costs more than curbside; twice-weekly collec-




tion more than once weekly.  Within a backyard collection system




a house 2OO ft from the road requires more time and therefore




more.money to service than.a similar house 1OO ft from the road.



     If one house has three items to be collected and another has



only two,, the former will cost more to service even if all other



things are equal.  Therefore, in order to make valid comparisons




between any two collection systems all of the variables must be




accounted for.

-------
                             - 89 -
     The actual annual collection cost per dwelling unit for the




average conditions observed in each study site ranged from $11.OO




to $46.00 per year, not taking into account any other influences




(Table 20) .  These costs cannot be compared as they occurred




under distinctively different sets of conditions.  There is no



such thing as an average cost of collection.  There is such a




thing as an average cost of collection for a particular dwelling




unit and a crew operating under given costs.  To change any




characteristics of the dwelling unit or one of the costs con-




tributing to 'the total daily crew costs changes the cost of ser-




vice to that dwelling unit.




     To compare the cost of service provided by any two satel-




lite vehicle collection crews requires that they service dwel-




ling units with identical characteristics and have identical




labor, equipment and overhead costs.  Evaluating a number of-




crews under the same basic operating conditions will then pro-




duce an average cost to service a dwelling unit taking into




account all influencing factors*'            , '




     Using the average values for the variables recorded during




the study, the average daily crew cost and the satellite vehicle



and packer driver collection regression models for each study




site, the annual collection cost per average dwelling unit for




once-weekly and twice-weekly collection was calculated.  The




average values used in the calculations were:






X  = five dwelling units for once-weekly collection
                                                                                                                    -  90 -








U)
2 |j
5 a
9 §
8 >.
B) ti
In S


a 8

"> S
° < <
g 1 «
S Q E
a .
Of (A
fc 2
b O
to 5
8 >
0)
s S
H
(0 (O
ffl
2
i













y,
it
• H
o
•o
'•H
8









'

1
1















*


C




^j
•H
§
0
C
-ri
rj
I
•6
n

CL





&
S
of ~4
h O
 M rt rt M rt

ooooo ooo
mininom omo
.ffi^ojcnrtw ^joji-i



S3 -o o 0 oo c- ,
>, — O
o»» -oiH*-o
- 22«rt .cd5
*Jt-£i2M1'c:i:6''c
B^'^B;>0'S^'^'5
s 2 K a « *
i-i rH o 13 (0 3
f O C « o a
H

01 °
01 ffi
S P. •
o
» .0 •u
S E
Q) (^ O
*• C

iH Q) 0)
0 S S

*" *> C
(0 01 £
3 0)
G. -H
TJ < 01
i-H
O • O
in >> n
"S 3 S
E 0?
§10 r-t
JC r-l
fi in -H
to V >
JI X
S3|
H25




-------
                              - 91 -
-X  = 1O dwelling units for twice-weekly 'collection   .   . :



 X2 = 15 items per -load for once or twice-weekly collection



 X  = 90 'ft                     .'..'••

 /••-..;'•>-

 X.. = 1O 'ft
  4           -                    -      ••             • ..


 X  = O.45 miles for once-weekly collection  -



 X  = O.75 miles for twice-weekly collection



 U  = l.O min



 B  = 95 percent                                      -



 H  = 6.O hr                                           ...



 Z  = three men                 .



 C, = $26.OS per day
  1-                                         "


 CT = $22.35 per day



 C  = $6.93 per day



 C  = $16.24 per day



 b  = two dwelling units      .



 b_"= six items for once-weekly collection  . •-'



 b  = three items for once-weekly collection



 b3 = 10O ft



 R  = 45 percent         ,              	      .



      Collection Frequency Effects on Cost.  For once-weekly col-



 lection the average annual cost per average dwelling unit for •



 standard conditions .was $19.OO per year and for twice-weekly col-



 lection the average cost was $28.SO per year  (Table  21).  There-



- fore, using standard conditions, once-weekly  collection  costs



 approximately 33 percent less  than twice-weekly collection.   .•
                                                                                                                  -92  -





Q
i
S
M **
§..!
'»B
*|
« dj
3H 2
< p

S K 1
£2
b
P
•H in
S z
w 11
85


o










'c
o

b
u
X -
f
7
O
U
1


1
0)
rH
0
U
X
I
t
ts
9
i

















• !+*
S -•
^
•H
§
o a
c 







a
•H
g
O
VI
•H
rH
a
u


a


0)
0)
0.
8
CM



8
en
CM



O
o
CO



8
m



B
-H
a -
c
o
o
o
•H.


1

a

o


a
. s
8
. m



o
tn
CO
• CM



8
o.
rH



8
0







Q
•H
(0

rH
rH
05
o

9

•H
S.
M
O

o
m
t-



8
•

-------
                              - 93 -






     Collection Agency Effects on Cost.  Cost to private collec- .




tion agencies averaged $15.OO per dwelling per year for once-




weekly collection, while public agencies averaged $23.OO per




dwelling unit per year, when all other influences were taken




into account.  Twice-weekly collection by private agencies aver-




aged $22.SO per dwelling unit per year, while identical service




by public agencies averaged $34.SO per year.  The 53 percent




lower collection costs achieved by private agencies was due




solely to higher crew efficiencies (dwelling units serviced




per hour).  The higher crew efficiencies were due to the fact




that the.private packer truck drivers also collected, thus con-




tributing to the total number of dwelling units serviced per




hour by the private agency vehicle operators were slightly




faster than public operators (Tables 12, 18).  This does not




refer to the costs charged to residents for service but is




related only to the time needed for collection by each type




of agency.





                     Collection Cost per Ton




     Collection costs per ton were calculated for the average




dwelling units observed at each study site  (Table 2O).  Also,




collection costs per ton were determined for each study site




using the average variable values used in the annual collection




cost per dwelling unit calculations (Table  21).  For these cal-




culations, the standard-dwelling'unit was assumed to have an




average of 3.5 occupants with a waste generation rate of 3.O




Ib per person per day.
                                                                                                                 - 94 -
      The same lower  costs  for once-weekly collection and col-




 lection by private agency  result from these calculations (Table




 21).   The actual  costs per ton achieved by the crews in the six




 study areas ranged from  $5.SO to $24.OO per ton, disregarding




 all  influencing factors  (Table 2O).  A recent national survey




 of national collection practices, disregarding type of frequency




 of collection,  gives an  average cost of $13.OS per ton for 1O




 cities providing  backyard  collection.






               Development  of Cost Estimation Model






      Ideally,  the designer of residential waste collection sys-




 tems would like to be able to estimate the collection cost per




 dwelling unit  serviced by  any type of collection system in order




 to make economic  comparisons between alternative systems with-




 out actual trial  implementation.  Such a cost estimation model




 can be developed  from the  satellite vehicle operator and packer




 truck  driver collection regression models that would take into




account  all of  the factors peculiar to each community that affect



 cost of  collection.




     The large  number of factors which affect the collection




cost of  a  dwelling unit  by any method make cost comparisons




from one community to another meaningless,  unless all influenc-




ing factors are accounted for.   Each community is a unique case




and must account for its own particular housing characteristics




and labor  situation to determine the most  economic collection

-------
                                95
 system.   A cost estimate per dwelling  unit  can be obtained.by



 multiplying the true crew cost  per hour by  the reciprocal  of  ,



 the. estimated crew efficiency.                   • -'      '      •



      The following are derivations of  true  crew cost  per hour



 and crew efficiency.                     .



      True-Crew Cost per Hour.   The costs associated with any



.satellite vehicle collection crew are:



      Cj, the wages and fringe benefits per  man per day



      C , packer truck operation and depreciation per  day



      C , satellite vehicle operation and depreciation per  day
       s                      '. •      . - •.	_..-—.


      Cj overhead cost of the system per day



      It must be emphasized that only those  hours devoted to



 the actual collection process are used in the calculation  of



 true hourly crew costs in order to prorate  the cost of the non-



 collection activities, such as  break times  and going  to-or grom



 the garage or disposal site, among the dwelling units which are



 serviced during the actual collection  process.  This  method allows



 the total costs to be included in those.hours which the collectors



 are collecting.  The total true cost. per. hour..6f any  satellite



 vehicle^collection crew may be expressed as
                                              c;"
                                             	o
•'where-C  = true crew cost per hour   .       '   -



       Z  = number of men in crew



       H  = number of hours devoted to the actual .collection process
                                 96. -.





      Crew Efficiency.  Crew efficiency is the  total  number  of



. dwelling units which a residential waste collection  crew col-



 'lects. per _unit -time.: _For this report on satellite vehicle



 collection the .crew efficiency is defined as the total dwelling



 units collected per hour by each satellite vehicle operator and -



 the packer truck driver.



      The total time required by a satellite vehicle  operator  to



 service X^ dwelling units in one load is expressed by the equation:
       = 0.803^ * 0.13X_ * 0.007X.J * 0.086X4 + 3.45X. - 6.78  + V




          '                              "
 where YT = total time required to service X  dwelling units. in



            one load, in minutes       '  ' .



       Xj = dwelling units per satellite vehicle load



       X^ = total items collected from X^ dwelling units



       X  = ' average distance satellite vehicle travels up



            dwelling unit driveways, in feet



       X4 = average distance from satellite vehicle to waste



            storage point, iri feet               .          -    .    .



       Xj = route distance of satellite vehicle per load,



            in miles                                              .



       U  = satellite vehicle unloading time, in minutes



       E  = fraction productive and unloading time of total



            time on collection route of satellite vehicle operator

-------
                              - 97, -






     Dividing YT by X  yields minutes per dwelling unit.   Mul-




tiplying Y_ by 1 hr per 6O min yields hours per dwelling  unit.




Inverting Y  produces the number .of dwelling units serviced per




hour by a satellite vehicle operator.  The resulting equation is:






     Dwelling units per hour =   .
     A crew of Z men has (Z-l) satellite vehicle operators, unless




the packer driver also operates a satellite vehicle as in Waukesha




County, Wisconsin.  The total number of dwelling units serviced




per hour by the satellite vehicle operators is then:






                               (Z-l)
     The total time which a packer driver requires to service b




dwelling units between truck moves is expressed by the equation:
               T = O.
                          + O.31b2 + O.Ollb3 - O.6O
where T  = total time in minutes to service b  dwelling units




           between truck moves




      b  = number of dwelling units serviced between truck moves
                                                                                                              - 98
     b  = total number of items collected from b  dwelling units



     b  = average distance from packer truck to waste storage



          location, in feet
Dividing T by b  yields the number of minutes per dwelling unit.




Since the packer driver only devotes a certain fraction of each




hour, E , to collection, the minutes per dwelling unit must




be divided by E .  Converting minutes to hours and inverting




the model produces the total number of dwelling units serviced




per hour by the packer truck driver is:
                                 . T




     The crew efficiency in total number of dwelling units per




hour for collection crew using satellite vehicles is:










                       6OXj (Z-l)    ^-i8*)
     Cost Estimation Model.  The annual collection cost per




dwelling unit, for N collections per year,  is obtained by mul-




tiplying the expected true crew cost per hour by the reciprocal




of crew efficiency, hours per dwelling unit.   The"cost-estima-




tion model is:

-------
                               - 99  -
                  MZ-I)C
                    H
                                                             (N)1
      To compute an estimate of the  annual collection  cost per  dwell-

 ing unit for a satellite vehicle crew,  the designer first determines

 hiswpar.ticular_ccsts,  estimates operational characteristics  from the

 previous analysis of variables based on physical route and housing  .

 characteristics, and then places them into the cost estimation

 model for calculation.   The cost estimate obtained is then compared

 to  the cost estimates for other methods of conventional  collection

 produced from similar collection regression models for backyard


 collection service (Appendix A).

      Quantitative and qualitative comparison of different waste
  •  • -          i.   •             '         .     .   •
 collection methods permits a rational and responsible decision

 for the type of waste collection system to be used in any community.



  •SATELLITE VEHICLE HASTE COLLECTION VS. CONVENTIONAL COLLECTION
              -..:'('.

      A residential solid waste collection system should  attempt

 to  provide the most conventional, aesthetic and sanitary service


possible to the customer  in the most efficient and-economical  •
                  i '.
manner in conformance with the health, safety and morale  of the

collection agency personnel.  Comparison of one collection  sys-

tem with another requires qualitative and quantitative evaluation

of each of these desirable features.
                                                                                                                     -  100 -
                     Qualitative Evaluation


     The following statements are the author's opinions, based

on observations and user reports, since they are not subject to

numerical analysis.

    The most convenient waste collection service which can

be^provided to'a homeowner is collection of wastes-from-the  -  •

point of storage..  This service may be provided by collectors

walking or riding satellite waste collection vehicles.  From

observation of both methods, collectors using satellite vehi-

cles provide a more sanitary service since walking collectors

tend to spill wastes around the storage area in their attempt

to minimize their number of trips to the packer truck by trans-

ferring and consolidating wastes from several containers into

their own carry container.  The satellite vehicle operator

has a 1- to 2-cu yd carrying capacity at his disposal and is

content to merely dump the wastes from the homeowner's container

into the satellite vehicle hopper.  In addition, the homeowner

views the satellite vehicle as a technological advance on the

part of the collection agent, thus eliminating a portion of his

natural psychological adversity to the waste collector.

     Use of satellite vehicles facilitates the work of the col-

lection personnel.: . For. the average dwelling unit located 1OO

ft from the street, the .satellite vehicle operator would walk

a total of 2O ft, but a walking collector would be required to

walk 3OO ft (Appendix A).  Thus the physical work required of

-------
                            - 101 -
the collector is reduced 93 percent.  As a result, the men do




not tire physically and are not as frequently exposed to injuries




from lifting and carrying excess weight.  The six agencies




studied reported significant increases in satellite vehicle oper-




ator morale and corresponding decreases in employee absenteeism.




Increased morale is partially attributable to the fact that




satellite vehicle operators assume an aura of status over their.




walking peers, as laborers operating machinery are usually more




highly regarded and more highly compensated than manual laborers.




In addition, satellite vehicles provide shelter for the collec-




tor during inclement weather.




     Satellite vehicle usage indirectly affects packer truck




operating and maintenance costs.  Packer truck movements and




aileage are reduced since the trucks remain on main streets




while the satellite vehicles service narrower side streets.




Thus operating and maintenance costs are reduced.  In addition




driveway and curb damage is reduced.  Also the hazards to chil-




dren playing in the collection area are significantly lowered,




with reduced reverse movements of the packer truck.




     Disadvantages of satellite vehicle usage are potential




littering of collection areas, excessive noise and damage to




dwelling unit property.  The open hoppers of the satellite vehi-




cles allow waste to be blown from them by high winds or high




speed of the vehicle.  Many vehicles made excessive noise which




could bother some customers.  Careless operators were observed




to run the vehicles over lawns and flowers, causing some damage.
                              - 1O2 -







                     Quantitative Evaluation




     The economy and efficiency of satellite vehicle collection




systems can be evaluated quantitatively.  .A comparison of this




method with the conventional backyard collection system using




walking collectors reveals the relative economies and efficiency




of one to the other.




     Collection Efficiency-  The efficiency of two collection




systems can be determined by' comparison of the total number of




dwelling units each collection method could service under the same




conditions.  The satellite vehicle crew efficiencies observed




in the six study areas were compared to the estimated efficien-




cies of conventional crews servicing the same areas (Table 22).




It can be seen that in five of the six study sites the satellite




vehicle crews were able to service more dwelling units per hour




than the corresponding conventional  crews.  The calculations for




the crew efficiencies are contained  in the Appendices for each




study site.  Satellite vehicle crew  efficiency ranged from 43




percent more efficient in Pasadena,  California to 28 percent less




efficient in Columbia, South Carolina.




     Collection Economy.  The ultimate comparison between alter-




native methods of residential waste collection is between their




relative costs to accomplish the same objective.   The annual




collection cost per average dwelling unit observed for each study




site was calculated for satellite vehicle collection and compared

-------






01
M

' w Q
rH O
° £
Eg
fe

§3
z
gs
Q Z
i|
U] Q
H O
« O •
M (J UJ
ra 8 8
* f 1
J
-eg
«a
K 5
^ ^
ca
S3
g 5

s S*
M (n
• •. ; a
K





' *
1 .,|T
: -5 •£ l-t
rH CI -H
'rH -> +J
»
" '„. '
+J ' +j o
0) - -H «-H
O rH O
° a! 3
rH -+J 0)
to to "> .',
. 3 in • \
c . •

^ .
*~\ *~*

3 S^
O - O tD
+- * -rl +>
U >4 C 6
C 3 0 -H
3 ° & «
O 0 -8-«r-
S^
1| Sa;
U i-H r-< -O
0 0 3
i <-> o
a to '> ;
-M tn •
J fr?

Si 1
* S r
- ' rH C^ ll
rH 9 O
OHO.
U V —


.


' ' ' ffl
-H
. 3 '

' .


o o boo
m o . ;in o in
CO ' ^f . CO t\ O
Cn M «H rH CO


^n in mo m
tn rH rH rH in
en oi en rH «





;



t> m t-» ca in

-". -.. ...


in  j: £ w) M
. o tn « ' • . « O
. 10 ffl - 01
. H H a rH
-OS H H -H rH •O
+> • — ' 	 ^ *H M
' ^ III
V •«-.-. 8 S..I


,m m O
•;-.& -3 S


8 . 8 8
S ^ r^









in  rH U
fO rH
HJ rH 'H tO
C "*• * •§
* 3
1 1


SO
5
en >o
tt CN


in O
rH 09
O C>)









rH ^
- in . «o




21^
t>







*> 3
Ji 0)
0) Q]
s * •
1 01
01 U
O -H
§ 3
4-) +>
's 's
§c

a V
§ -Iv
•H 0
+*
•H m
8 .5
*H
"S ?
£ 8
n o
0- <0
| |
Q) O
S 5
'£• £
4. *


                                  T-  104 -






   with the estimated cost by conventional walking collectors




   (Table 22).  Satellite vehicle collection costs were less than




   conventional collection costs  in four cases, even in two cases,




   and more in two cases.  Satellite vehicle collection costs ranged




   from '24 percent less to 7O percent, more than the estimated cost




   per dwelling unit by conventional walking collectors.  Therefore




   neither method of collection can be  said to be more economical




   in all cases.  In order to determine the most economical col-




   lection system for a. particular community several factors must.



  . be considered and used in the  cost estimation model previously




   outlined.





          SATELLITE VEHICLE WASTE COLLECTION MODEL APPLICATION





        The practical significance of the satellite vehicle waste




   collection model is based upon its capacity to accurately esti-




   .mate the collection time necessary to service any given number




   of dwelling units in any given community.  The following appli-




   cation is presented to demonstrate the validity of the model and




   the utility of the accompanying analyses in designing a satellite




. .  vehicle waste collection system.      ' ;   -  .               ,:




        The Village of Glendale, Ohio has a satellite vehicle waste




   collection crew for their residential waste collection, but was




   not one of the study sites used in the development of :this re-




   port.  .The-crew,- consisting of a packer truck driver and two




   satellite vehicle operators, spends 330 min per day or 1,65O

-------
                             - 105 -
                                                                                                                    - 106 -
 min per week on collection.  The crew services approximately

 854 dwelling units an average of 2% times per week,  for a

 total of 2,135 dwelling units per week.

      By evaluating the factors analyzed in this report, and

 using them in the waste collection models, the collection time

 required can be estimated and compared to the actual time

 required given above.  The satellite vehicle waste collection

 model is:
    v=
                                     O.O86X
                                               3.45X  - O.78 + U
. where Y_ — the total time, in minutes, to complete one satellite

            vehicle round trip

       X  = the total number of dwelling units serviced in one

            round trip.  As discussed earlier, X  is primarily

            a function of collection frequency, which in turn

            determines the number of items per dwelling unit

            Since garden wastes are combined with household

            wastes in Glendale it is estimated that the average

            dwelling unit has 3.9 items to be collected per week

            or approximately 1.5 per collection (Table 8).
       *1
                               15
            average number of items per dwelling unit

             15
            1.5
= 1O (p 38)
       X  = the total number of items collected from X  dwelling

            units.  The analysis of X  showed that the satellite

            vehicle averaged 15 items per load (p41).
                                                                    (1)
                                                                                  the average  distance,  in  feet,  which the satellite

                                                                                  vehicle  is able  to  drive  up each dwelling unit drive-

                                                                                  way.  This is a  function  of street to storage distance

                                                                                  and accessibility.   From  the average Village house lot

                                                                                  size  it  can  be determined that  the average street to

                                                                                  storage  distance is 9O ft.
                                                                               " DSS  ~
                                                                                                               * 0t76X
                                                                                                                          11-4  (p 41)
                                                                                                                                               (2)
                                                                                 9O   = 0.83X3 + O.76X4 + 11.4

                                                                                  The accessibility factor may be determined by a visual

                                                                                  survey of the Village on collection  day  or by assuming

                                                                                  the average value, O.88,  for A (Table  6).


                                                                                                                               (3)
                                                                                  O.88 =
                                                                                  Simultaneous  solution of equations (2) and  (3)

                                                                                  determines the values of X  and X .
                                                                                                   X  = 94 ft
                       X4 = 13 ft


X  = the average distance,  in feet,  from the satellite

     vehicle to the waste storage location.   This  was

     calculated above and the value  is  13 ft.

-------
                              - 107 -



     •X  = the total route distance^ in'^miles, of. the satellite

         .  vehicle^rpund trip.  Route distance is a function of

           the housing^density and the total number of~houses

           serviced per load (p 47).  Glendale has a housing den-  '

           sity of. 53 dwelling units per mile of street..' The .   :

           route distance per dwelling unit .served is 0.114 miles

           (Figure 15.);  For 12 dwelling units the total route

           distance would then be approximately 1.35 per load.


       U = the unloading" time, in minutes,.x>f the satellite

           vehicle at the packer truck.  Assuming average operat-

           ing practices by the crew in Glendale, using Cushman

           satellite vehicles, U is estimated to be 1.5 min (Table 9).

      B  = the fraction of the operator's total time which is

           occupied by collection and unloading activities.

           Since the operators are assumed to use good operating

           practices,.the. average value of E , O.95, is used (p 61).



     The determination-of values for the variables in the satellite

vehicle waste collection model was based on knowledge of three

factors; the collection frequency, the housing density, and the

average dwelling unit lot size.   Thus, it can be seen that the

design of the collection system does not require a- field study'

and can be easily determined from common community records.
                            -  108 -







      Substituting the values assigned into the model,


 0.80(10) + 0.13(15) + O.OO7(94) + 0.086(13) + 3.45(1.35) - 0.78 + 1.5O

                                  O.95


    .    >        =,18.O min/satellite vehicle load


     The efficiency of the satellite vehicle operator is io dwell-


ing units per 18.0 min or 33 dwelling units per hr.


      The number of dwelling units per hour which the packer truck


driver can service can be calculated using the packer truck driver


collection model.  These dwelling units have the same characteris-


tics as those serviced by the satellite vehicle operators.
               T = O.oObj + °-31b2 + O.Ollbg - O.6O    (p 74)



where T  = total time, in minutes,  to service b  dwelling units  -


           between truck moves

      b  = total number of dwelling units serviced by  the packer


           truck driver between truck moves.  The average value


           for b  observed during the study, 2, will be used here


           (Table 17).                     .   _


      t>2 = total number of items collected by the packer truck


           .driver from b  dwelling units.  The dwelling units in


           Glendale were estimated to have 1.5O items each.   There-


         .  fore t>2 = 3.0O.


      b  '= average distance,  in feet,  from the packer truck to

                                                     7
           the storage at each dwelling unit.  If the average

-------
                                109 -
           .street to storage distance is 90 ft, the value for




           b_ would be approximately HO ft.






      Substituting these values into the packer truck driver




collection model;






     T = O.6O(2) + O.31(3.O) + O.Oll(llO) - O.6O = 2.7 min






     A packer truck driver working efficiently can devote 3O




percent of his time to collection (Table 18).  Assuming this to




be the case in Glendale, the efficiency of the packer truck




dr iver is:





  2 dwelling units/2.7 ain x 6O min/hr x O.3O = 13 dwelling units/hr






     The crew efficiency can now be calculated:





     crew efficiency = satellite vehicle operator efficiency




                       + packer truck driver efficiency




                     = 2 (33 d.u./hr) + 13 d.u./hr




                     = 79 dwelling units per hr





     Since the total number of dwelling units which must be




serviced per week is known, then the total time required may be




determined.






        2,135 dwelling units/week x 1 hr/79 dwelling units






                         = 27.02 hr/week





                       = 1,620 min per week
                                                                                                                        - 110 -
     The actual collection time required is 1,6SO min per week




or approximately 2 percent above the estimated collection time.




Therefore, this application clearly illustrates the usefulness




and accuracy of the satellite vehicle operator and packer truck




driver collection models for designing satellite vehicle waste




collection systems.

-------
                             - Ill -
                            REFERENCES •
1.  American Public Works Association, Refuse collection practice,
    third edition; " Chicagoj'Public Administration Service, 1966.
    525 p.'  •         - '.  • .    •-,*.!--- ;, .-'                  .  . '

2.  Little; H. R.  Solid waste management  iri~ the Territory of
    Guam:  Cincinnati; U.S. Department of  Health, Education and
    Welfare, 1969.-  116 p.                               -

3.  Ralph Stone and Company, Inc.  A  Study of solid waste collec-
    tion systems comparing one-man with multi-man crews.- Public
    Health Service Publication No. 1892.   Washington,  U. S.
    Government Printing Office,  1969.   175 p.
                                                                                                             " - '112 -
                                                                                                           BIBLIOGRAPHY
 American Public Works Association,  Refuse collection practice,
 third edition.  Chicago, Public Administration Service, 1966.
-.525 p. -     ' •.-"••."•"  '- '•'  .  .';';' '•"'" --.••..  •        .     '  •  "  .-   .. :  .

 An analysis of refuse collection and sanitary landfill disposal.
 Technical Bulletin No. 8,  Series 37.   University of California;
 1952.  133 p.    '   :  . '    .-.= .."       :

 1968 Annual Report, City of Cincinnati Department of Public Works
 Waste Collection Division.                              -.'•''.•'

 Anon.  Backyard service improved by use of small vehicles^  Refuse
•Removal Journal.  7(1):24,  January, 1964.               ' ••!.  .'':•-

 Anon.  Three-wheel carts double number of daily collections.
 Refuse Removal Journal.  .7(5):34-36,  May,  1964.               .

 Anon.  Solid wastes collection.  Ohio Cities and Villages.  15(6):.
 June, 1967.                              .        '' ,'

 Black, R. J., A. J. Munich, A.  J. Klee,  H. L. Hickman, Jr.,  and
 R. D. Vaughan.  The national  solid  wastes survey; an interim
 report.  Cincinnati , U. * S. Department of Health, Education and
 Welfare,  1968 ...-,53 p.                            .

 Brambier, N.  Solid waste  collection  simulation:  a generalized
 approach.  Paper presented at  the Third Annual Simulation Symposium.
 Tampa, Florida,/January, 197O.     '•'•••                 ;'    ' • ••   .

 Clark, R. H., B. L. Grupenhoff and  G.  A. Garland.  Cost relations
 in solid waste collection.  Cincinnati^  U. S. Department  of Health,
 Education and Welfare, 1969.   11 p.

 1969 Commercial Atlas and  Marketing Guide, one-hundredth  edition.
 Chicago, Rand McNally and  Company,: 1969.  611 p.

 Draper, N. R. and H. Smith.  Applied  regression analysis.  New
 York, John Wiley and Sons,  Inc., 1968.  4O7 p.

 Henningson, Durham and Richardson,  Inc., Collection and disposal of
 solid waste for the DBS Moines  Metropolitan area.  Cincinnati,  "
 U. S. Department of Health, Education and.Welfare, 1968.   268 p.

 Little, H. R.  Solid waste management-• in the Territoty of Guam.
 Cincinnati, U. S. Department'of Health,' Education and Welfare,
 1969.  116 p.

-------
                               - 113 -
.Mu
tions
 unicipal collection of  refuse:   progress  report  with recommenda-
 ions.  City of Pasadena, 1965.   22 p.                         -

Ralph Stone and Company, Inc.  A  study of  solid waste collection
systems comparing one-man with multi-man crews.   Public Health
Service Publication No.  1892.  Washington,  U.  S.  Government  Print-
ing Off ice, 1969.  175 p.

Shuster, K. A.  Unpublished data, February,  197O.

Truitt, M. M., J. C. Liebman and  C. W. Kruse.   Simulation  model
of -urban refuse collection.  Journal of The Sanitary  Engineering
Division. ASCE.  95{SAZ):  289-297, April,  1969.
                                                                                                                  - 114 -
                                                                                                             ACKNOWLEDGMENTS


                                                                                         This study required the cooperation of many people and

                                                                                    organizations.   The Division of Technical Operations  extends

                                                                                    sincere appreciation to the personnel of the Division of Sani-

                                                                                    tation, City of Atlanta, Georgia;  Department of Sanitation,

                                                                                    City of Columbia, South Carolina;  Sanitary Disposal Company,

                                                                                    Knoxville, Tennessee; City Sanitation Service,  Medford,  Oregon;

                                                                                    Sanitation Division, City of Pasadena, California;  and Sanitary

                                                                                    Disposal Service, Inc., Waukesha County, Wisconsin.

                                                                                         Members of the Division of Technical Operations  participat-

                                                                                    ing in the field work were:  Gregory Cole, Stephan  Freedman,

                                                                                    Jeffrey L. Hahn, Harry R. Little,  Claude A. J.  Schleyer,

                                                                                    and Kenneth A.  Shuster.  Special thanks go to Betty L.

                                                                                    Grupenhoff for hex statistical services, which were a major

                                                                                    contribution to the success of" the study.

-------
                              -  115 -
                             APPENDIX A



                 WALKING COLLECTOR REGRESSION MODEL




      A regression model to describe productive collection time .


 for residential  waste  collection by walking  collectors has- been .f •


 developed by the Bureau in a manner similar  to that for the


 satellite vehicle regression model* Data on 16 walking collec-.


 tors were gathered in  five U.S.  cities,  by following the collec-


 tors on their collection routes*  Comprehensive analysis.of the


 data is currently in progress and only the preliminary results


'have been used in this report.                       -


      The regression model which  best describes the productive


 collection time required to service W  dwelling units is:




            C = 0.18 -  0.12W1 + 0.12VT  +  0.24W3 + O.OOW



 where C  = productive  collection time  required, in minutes, to


            service W^  dwelling units

       W  = number of dwelling units to be serviced


       W  = total number of items to be collected from W


            dwelling units


       W  = total number of round trips which the collectors make


            to the packer truck.   Preliminary analysis of the data


            indicat es . that an average of  three items may be coll-


            ected by the typical  walking  collector before he must


            return to the packer  truck  and empty his carry barrel.


            This  relationship was assumed for the calculation of  .
                                                                                                                     - lisa -
           time estimates for the six areas where satellite


           vehicles were studied.


      W  = the total distance, in feet,  walked by the collector
       4 '  .

           while servicing W  dwelling units.  Preliminary


           analysis of the data indicates that the walking dis-


           tance is a function of the street to storage distance


           for each dwelling unit.  A regression analysis of


           street to storage distance against walking distance


          -.. produced the equation:
     W4 =  £5 + 2.
istance)!   W
55 (average street to storage dista:
          •This equation was used in the time estimations for


   ."'.-•' -   'comparison of collection methods in the six satellite


           vehicle study sites.




     The walking collector regression model was able to explain


92.6 percent of the total variation in the data, and the standard


deviation of the residuals was 19.0 percent of the response mean,


productive collection time.


     The regression model for productive time does not include the


non-productive or "other" time which normally occurs during the


work day.  Productive time accounts for approximately 85 percent


of the average walking collector's time.  When the walking collec-


tor also drives the packer truck his productive time is reduced to


approximately 8O percent of his total time on the route.

-------
                             - 116 -
                            APPENDIX B

         FIELD STUDY, AUGUST 18-22, 1969, ATLANTA, GEORGIA


     The Sanitary Division of the City of Atlanta Public Works

Department is responsible for the collection and disposal of

all residential wastes generated by the 502,500 inhabitants of

Atlanta (Table B-l).  The Division provides backyard collection

of refuse twice weekly and curbside collection of garden wastes

once weekly.  The refuse collection department operates 127

packers and employs 1,O25 men on 152 routes in 15 districts.

     A shortage of manpower was the initial reason for Atlanta's

purchase of 41 Cushman scooters in 1966.  The Division  suffered

from an absentee rate which . ran as high as  4O percent of the

total department.  The City found that two collectors with

Cushman scooters could replace three walking men collecting

with tubs.  The Cushman scooters operated well and efficiently

at first, but developed mechanical troubles 3 to 6 months after

introduction into the system.  Rugged usage over the hilly terrain

in northern Atlanta caused the small 8 in tires to fail after 3O

days.  All of the Cushman vehicles became obsolete within 3 years.

     The City then purchased a number of four-wheeled Trashmobiles

to use in this area.  The Trashmobiles worked well and are still

in use, but they lack the maneuverability of the three-wheeled
                                                                                                                -  117 -
                              TABLE B-l
                       CITY OF ATLANTA, GEORGIA
                  (from Atlanta Chamber of Commerce)
Population

Dwelling units

Persons per dwelling unit

Land area

Population density

Dwelling unit density

Miles of roads

Dwelling units per street mile
5O2,5OO

157,900

3.18

136 sq miles

3,7OO persons per sq mile

1,16O dwelling units per sq mile

2,057 miles

8O dwelling units per mile

-------
                             -  118 -







 satellite collection vehicle as they must  back out  of driveways,




• thus causing a safety hazard.   Recognizing the need for a heavy-




»duty-'vehicl«> -the-Gity-had the-T-ra-sh Taxi  designed  to meet  the




•requirements imposed by the varied terrain in northern Atlanta.




 The-1<| cu yd hoppers from the obsolete Cushman scooters were     •




 used on the Trash Taxis.  The Trash Taxi features a 13 in.  wheel,




 a heavier brake,  larger engine  and hydraulic transmission to




 combat the problems  previously  encountered.   Eighteen Trash Taxis




 were put into operation in August 1969 to  supplement the 26 Trash-




 mobiles currently in operation.                  .  •'. ,   •






                        Field Study Analysis--  -.






      The field study on Trash Taxi and Trashmobile  operations




 in Atlanta was conducted August 18-22, 1969.  Two crews using




 Trash Taxis were observed during the first 4 days of the study.




 Data were recorded on four Trash Taxi  operators and two packer




 truck drivers.  The final day was spent with a crew consisting




 of two Trashmobiles and a packer.  Both Trashmobile operators   .




 and the packer driver were observed for 1  day.




      Due to the distinct characteristics of these two types-of-




 satellite vehicles,  their operations were  analyzed and will be




 discussed individually.
                             - 119 -
     Trash Taxi Collection Operational Analysis.  One hundred




and four data points were collected on Trash Taxi operations in




Atlanta.  Step wise regression techniques were applied to two




individual operators and also on all the operators together.




The -best . regression model for describing productive collection




time for the four Trash Taxi operators in Atlanta is:  (Table




B-2);
    Y  = - 2.69. + 1.O6X, + O.25X  + O.O19X  •«• O.O47X^ + 4. SIX






where Y  = productive collection time per satellite vehicle




           load, in minutes




      Y  — number of dwelling* units serviced per load




      X  = number of items collected per load




      X, = average distance that the satellite vehicle goes




           up each dwelling unit driveway, in feet




      X  = distance from satellite vehicle to storage point,




           in feet




      X  = route distance of satellite collection vehicle per




-.          load, in miles





     This model was able to explain 87.7 percent of the total




variation in the data with a standard deviation of the response




mean, productive time/>f 14.0 percent.  An overall F value of




139.6 with 5. degrees of freedom in the numerator and 98 degrees

-------
- 120 -











ti
1
u o
CL. O>
fe rH
O •*
H rH
> o>
•p
1? *J I
c +>

o
0) TJ
u d
in -H a rH
o to n
rx -H o
•o o.

• u of 0)
SiH Ol
o o d
^ +J ^1 « H
to js o
•HO) +>
a > «



go, |
C rH S
cod u D. o
X 01 !c -H
Q > TJ

?«
0} +* to

X V 0 rH
M rH h
8pL

? ?•§
^H 0) O O
iHrH +* ^1 rH
"3 C S K
a 3 01 0<
o> o.






o
(0
0)

o*



00
^
»

in
•o
rH
1

i? 5"
01 0>
rH rH (7.
1 & *

I 1
tn o
•rt -rt 00
>> >. 0
d d O
C c
d d

C c
S S
(D ffl ^*
« 01 rH
SO O
M »
O>QiO
O V
0- CC

(0
rH
O



§
O
*











rH M CO
fn H H
i s s

0)
in
00

in
O
CO




rH
m



o»
O
o




o>
3
*
o



o»
0



s
rH
*












Hi

r» o> •<»
t^ OJ t~
oo a o.

O CO rH
MOO
1 1 H
1


oo in o
Tf CO T»



t^ in  0
iH O»
O • 1 O
0 0


m t> w
« CO rH
O O O
*


0 ° •O
rH 0 0



01
•H
X
d
H
£
tn
d

(H
rH M
rH £ Z
3 S §

O
rH
0*

§
rH




O»
?



tn
0
o




rH
0
O
0



rH
0



CO
O


in
01
rH
•H

Q
tn
d
M
H

rH
rH
















0
E
T*

0

•H
roduct
a
£
+*
•o
0

rH
0
m
8
x
rH
-H
+*
0)
O


(U
i— t

(0
M
(0

*



                              - 121 -





of freedom in the denominator was obtained.  Therefore, the five




variables included in the model were significant in explaining




productive time.  The values of the coefficients for the regres-




sion model variables for operators ATT , ATT  and all Trash Taxis




were very similar, with X , number of dwelling units serviced per




load, being the most highly correlated variable to productive




time (Table B-2).  Due to this similarity, more confidence can




be attributed to the variables used in the regression model.




The utility and degree of accuracy of this model can be illustrated




by comparing its result to the values observed in the field investi-




gation.





     The model can be used to predict the productive time required




to make a complete load with the Trash Taxi using the five regres-




sion model variables.  Productive time is converted to total




elapsed time for the load by dividing by the percent productive




time (Table B-4).  The average values of the variables observed




during the study were used:  (Table B-3).







       Fraction productive time = 87.9 percent





       X  = number of dwelling units per load = 8




       X  = number of items per load = 16




       X  = average distance of vehicle up driveway = 1OO ft




       X  = average distance from vehicle to storage = 3O ft




       X,  = route distance of the satellite vehicle = 0.35 miles

-------
                              -122 -
 a
•s
 iS
 I O>
 < o
 U)

 II
 in 5


¥
 a








Variable .










0) •
> M
•H 01
V {I'D •— •
>. ' jj 0> O -S
K V> ""
a.
i"
coi ci^~
u) co o) o (fl +*
0) +> ' h +J h *W
Q 01 *> O »••
•H 01 ^
0 01
01 -O
m ° § 1 j»
X 3 *• -~
x*5^S!2«
at iT
Q > w
8 w |
C H .3 '-^
to aj o o* <*> +>
XSi3-S-
0 > . TJ
•O T3
. 5 a
0) V O
4* rH r4
H o o.
u
Oi • -O 13
C a> & a
•*> *> o o
a"-1 -riTj rt
*^ 5 ij
P , 01 O.
*• 01 01
^i 1-1 m ?
S -H § »

« V
a
Tf en ' rH (0 H - .N . O« O
« • o> ^> •« —m - • ^ • t» -o

§ § S § . § S 8 S
•H . H H . rH


omoo *°P2 S
m.*o ^ co co..r* o» OD
ooooo op o
O-OOO O OQ^9
Tt in eo co co « ^ co
O O O O - --O O O O
•^ m o> - cj " o . .t*- o - I*-

tH i-t rH ' rt rrH iH.d rH

O O 00 00 00 O"> OI
rH rH rH r-l . rH

CO O • t*» 
-------
                             - 124 -





    Substituting these values into the regression model:





      Productive time =  - 2.69 + l.O6(8) + O.25(16) + O.O19(1OO)



                         + O.047(3O) + 4.81(0.35)



      Productive time =  14.8 min



      Elapsed time =—rrr = 16.8 miti.
                      • o / y
                                             *

      Actual average productive time per load   =  15.1 min



      Actual average  total  time per load =  17.2 min




     The predicted productive time  required per  load is  2 percent



below the actual time recorded.  Two Trash Taxis working together



 collect from  16 dwelling units in 17.2 min.-



     Trashmobile Collection Operational Analysis.   Twenty-one data



points on Trashmobile operations were  obtained during the field



 study in Atlanta.  Stepwise regression techniques were applied to



 the data for  each operator individually as well as to both opera-



 tors together.  The best mathematical model for the two Trashmobile



 operators in Atlanta is  (Table B-2)





     Yp = 1.40 + 0.39^  + O.14X2 + O.O21X3 + O.O59X4 + 4.993^





 where Y  = productive collection per satellite  vehicle load, in  minutes



      X_ = number  of dwelling units serviced per  load



      X = number  of  items  collected per load



      X = average distance that the satellite  vehicle goes up each



           dwelling unit driveway,  in  feet
                              - 125 -







      X  = average distance from satellite vehicle to storage



           point, in feet



      X  = route distance of satellite collection vehicle per



           load, in miles




     This equation was able to explain 91.6 percent of the total



variation in the data, and the standard deviation of the residuals



was 9.4 percent of the response mean,  Y .  An overall F value of



32.9 with 5 degrees of freedom in the numerator and 15 degrees of



freedom in the denominator was obtained.  These five variables



were significant in explaining the productive time necessary for



collection by a Trashmobile operation.   The coefficients of the



variables in the models for the individual operators were all of



the same order of magnitude (Table B-2) .  The most highly corre-



lated variable to productive time for  the two operators together



is X , route distance per load.   The number of items per load,  X ,:



was the most significant variable for  operator ATM .



     The Trashmobile regression model can be tested for. accuracy. -



similarly to the Trash Taxi by using average values for the vari-



ables in the regression model (Table B-3).



     X  = 12 dwelling units



     X  = 19 items
      2


     X  = 7O ft



     X4 = 30 f t



     Xr = O.8O miles
  Table B -3 •

-------
                             - 126 -
                                                                                                                                         -  127  -
     Percent productive time is 92.2 (Taoxe c-aj.




     Substituting-the-above .values into the regression model:



        Productive time per load = 1.4O + p.39(12) + O.14(19)     .  .



                                   + O.O21(70) + O.O59(3O) + 4.99(0.80)



                .              •  s 16.O min

                                                »
        Actual productive time per load observed  = 16.0 min




     The predicted productive time required per average load



agrees perfectly with the actual productive time observed.  To



account for unloading- and other time- inherent to the operation,


                                          16.O
the actual total elapsed time per load is  „„- or 17.4 min per



load.  Therefore, two Trashmobiles working together collect 24



duelling units per 17.4 min.



Packer Truck Drivers Operational Analysis.



     The packer truck drivers did not do enough collection to



apply regression techniques to.their activity.  Two of the three



drivers observed did not collect at all.  The third only collected



4.7 percent (Table 8-5) of the time he was observed.  The three



^drivers spent over three-quarters of their time waiting in their



trucks while the satellite vehicles collected.  Only 23 percent



of the driver's time was utilized in contributing to collection



route completion.
 Table B-3

-









' ' o>
S
S3-
. . S2 N"
Is
« S|
i > o
eq IH 3 '
ta C <
J < i
H p S
rH B
B Q
O S
«°
g< .

CL <
p
• '• .- - ' J ' ' < '






-









•H
r-|
d
V
o
+J
0
•M
8
o
M
0}
 O
0 -H £
•rt -O
£
o>


Q

M
0
•H

PI CO O rH ^ r-( f*.
•H O O O O «O rH
rH

m IH • tn . »o m N (f,
2 ? 2 ° N ^ *o
CO - r- co (*• t^ io t^


rH ^ Ch . ^ CO CO ^
iO O O CO Of* ^f
W rH rH PI M





O O O O O t* CO
o o o o o «r o




rH «O t*. O O . O) O
o> <* ^» 'f (^ i^ r>
. rH




O CM ^f tx O O rH
• • • » • » •
r^ co ^j (*«. o 10 ~co
*"^

CO O O i-1 CJ O W
.rH rH CM Ct Ol N 1
"X \ \ \ \ \ QO
\
m
tj
•H
iH CM ' • m "O .
• IT ' ' i i 5

-------
                             - 128 -







                       System Cost Analysis





     The daily costs of the Trash Taxi and Trashmobile crews




were calculated separately due to the distinct  differences




in their operation and efficiency.  All costs,  with the excep-




tion of the Trash Taxi operating costs, were provided by the




Sanitary Division of the City of Atlanta.




     Trash Taxi Crew Daily Costs.  The costs associated with




any satellite vehicle crew are labor wages and  fringe benefits,




equipment operation and depreciation, and overhead.  Equipment




operation includes repairs and maintenance in addition to




direct operating costs.  Operating costs were not available




for the Trash Taxi since it was studied during  its first week .




of operation.  Operating costs are therefore based on estimates




by the manufacturer.  Employee fringe benefits  were estimated




to be 15 percent of wages and the overhead was  estimated to be




one-sixth of the total crew cost.  The costs are expressed in




dollars per day.





Labor




     2 - satellite vehicle operators at $19.1O per day +




         15 percent fringe benefits                          = $43.92




     1 - packer truck driver at $20.75 per day + 15 percent




         fringe benefits                                     = $23.86




                                                Total'labor  = $67.78
                             - 129 -







Satellite vehicle operation and depreciation




     Operation




        2 - Trash Taxis at $1,200 per year               = $ 9.22




     Depreciation




        2 - Trash Taxis, $2,711 new, 4 year life         = $ 5.2O




                      Total operation and depreciation   = $14.42





Packer truck operation and depreciation




     Operation




        1 - 2O cu yd Loadmaster at $1,42O per year       = $ 5.46




     Depreciation




        1 - 2O cu yd Loadmaster, $12,O59 new, 5




            year life                                    — $ 9.28




                      Total operation and depreciation   = $14.74




Overhead




     Overhead costs were not available and therefore




     were estimated to be 2O percent of all other costs.




     Overhead cost = 0.20(67.78 + 14. 42 + 14.74)         = $19.39





-Total crew cost




     The total daily cost for a Trash Taxi crew in




     Atlanta is                                           $116.33






	Annual Collection Cost Per Dwelling Unit.  The annual cost




for twice weekly collection for the average dwelling units

-------
                                  -  130  -
                                                                                                                       - 131 -
    observed  in Atlanta can be determined by multiplying crew cost

    by  crew efficiency.  The crews observed during the study spent

    •an  average of 6.5 hr per day in the actual-process of collec-

    tion.  .Therefore the true cost of ^he  Trash-Taxi  crew is $116.33 per

    6.5 hr or approximately $18.OO per hr.  The crew efficiency is  •

    defined as the number of dwelling units (d.u.) the crew is able

    to  service in 1 hr.  The Trash Taxi crew serviced an average of

    16  dwelling units per 17.2 min during the study, or 56 dwelling

    units  per hour.  The annual collection cost per dwelling unit

    is  then:



      $18.0O/hr x 1 hr/56 d.u. x 1O4  collections/d.u./yr = $33.5O/yr



         Collection Cost Per Ton.  The residential waste generation -

    rate for  the study area was determined by weighing all of the

    wastes collected during the field study (Table B-6).

         The  collection cost per ton was calculated by multiply-

    ing the crew cost per hour by  the number of hours required per

    ton collected by the crew.  The amount-of waste collected per

    hour by the Trash Taxi crew was:



56 d.u./hr  x 3.18 persons/d.u. x 2.4 Ib/person/day x  3.5 days = 1,5OO Ib/hr


         The  collection cost per  ton  was:



           $18.0O/hr x  1 hr/l,5OO  Ib x 2.0OO Ib/ton = $24.0O/ton   '
             TABLE B-6


 PRESIDENTIAL SOLED WASTE GENERATION
ATLANTA, GEORGIA, AUGUST 18-22,  1959
Date
8/18
8/21
Total
Weight
(Ibs)
5.9OO .
7,360
13,260
Services
174
347
521
Lb/capita/day
2.7
2.2
2.4

-------
                             - 132 -
                                                                                                                   -  133 -
     Trashmobile Crew Daily Costs.  The labor and packer


truck costs for the Trashmobile crews are the same as those


for the Trash Taxi crews.  Satellite vehicle operation and


depreciation costs and overhead costs must be calculated for


the Trashmobile crews.


Satellite vehicle operation and depreciation


     Operation


        2 - Trashmobiles each at $727 per year            = $ 5.58


     Depreciation


        2 - Trashmobiles, $2,4OO ncw»4 year life


                     Total operation and •depreciation



Overhead


     Overhead .cost is assumed to be 2O percent of all


     other costs.


     Overhead cost = O.2O (67.78 + 1O.20 + 14.74)         = $18.54


Total crew cost


     The total daily cost of the Trashmobile crew


     described is                                         = $111.26



     Annual Collection Cost Per Dwelling Unit.  The Trashmobile true


crew cost was $111.26 per 6.5 hrs or approximately $17.OO per hour.


Two Trashmobile operators were each able to service an average


of 12 dwelling units per 17.4.min.  The total crew efficiency was
     24 dwelling units per 17.4 min or 83 dwelling  units per hour.


     The annual collection cost per dwelling  unit is:



        $17.OO/hr x 1 hr/83 d.u. x 1O4 collections/d.u./yr = $21.5O/yr



          Collection Cost Per Ton.   The amount of waste collected per


     hour by the Trashmobile crew observed was:



83 d.u./hr x 3.18 persons/d.u.  x 24 Ib/person/day x 3.5 days = 2.2OO Ib/hr



          The collection cost per ton  becomes:



           $17.OO/hr x 1 hr/2,2OO Ib x 2,OOO  Ib/ton = $15.5O/ton




          Satellite Vehicle Collection vs. Conventional Collection


          It is important to determine if the satellite vehicle collec-


     tion systeas in Atlanta are as efficient and economical as conven-


     tional walking collectors  could be in their place.


          A regression model which estimates  the productive time required


     for a walking collector to service any number  of dwelling units with


     some particular characteristics can be used to calculate their


     efficiency.  This regression model is:
               »'C= O.18 -  O.12W, + 0.12VJ  + O.24W  + O.OO5W
                               123         4
                                                                                      •Appendix A

-------
                             - 134 -
                                                                                                                   - 135 -
 where C   =  productive time," in  minutes,  to  service W  dwelling  units .   ,



      W   =  the number of  dwelling units  to  be  serviced



      WL  =  total number of  items to be collected from W  dwell-



            ing units            "                                     •



      W   =  total number of  round trips to the  truck'made by the         ,



.'  •     .    walking collector while servicing W dwelling units



      W   =  total distance,  in feet, walked  by  the collector while



            servicing  W dwelling units                                 .




      Separate comparisons  were-made of the walking collectors



 with the  Trash Taxi crews  and the Trashmobile crews.



      First,the time required by  a walking collector to service



 eight dwelling units with  characteristics identical to the  average



 dwelling  units serviced by the Trash Taxis was  estimated



 (Table B-3).




      W   =  eight dwelling  units



      W   =  two items per dwelling unit x eight  dwelling units =  16 items

                                   »
      W   =  one trip per three items  x 16 items = 5.33 trips



      W   =  J45 + 2.55  (average street  to storage)] W *     ' :'         •



          =  1*5 + 2.55(130)]  8 =  3,010  ft
          Substituting into the walking collector regression model :





       C= 0.18 - 0.12(8) + 0.12(16) + 0.24(5.33)  + O.OO5(3,01O)



         = 17.47 min of productive time





          Since the average*. walking collector is productive 85 percent



     of the total time on the collection route, the total time required



   -  for him to service these eight .dwelling units would be • •-.'.,,, or
                                                             O.85


     2O. 5 min.                    '



          The average total time required by a Trash Taxi operator to



     service these same dwelling units was 17.2 min. " Thus, the use



     of Trash Taxis, in ttiis area is theoretically 19.1 percent more



     efficient than the use of walking collectors.



          The total cost of a crew with two walking collectors, a



     packer truck and a packer driver in Atlanta is $99. 02 per day.



     Since the crews are on the collection route an average of 6.5 hr



     per day, the true hourly cost per collection hour -are approximately



     $15.OO per hr.  The walking crew services a total of 16 dwelling units



     each 20.5 min.  The annual collection cost per dwelling unit is



     ihen:   .
$15.00/hr x
                                                                                                                  hr/6O min x  1O4 collect ions/yr = $33.5O/yr
  Appendix A
          The cost for equivalent service by the Trash Taxis is also .


     $33.50 per dwelling unit per year.  The walking collection system



     is theoretically just as economical as the Trash Taxi system.

-------
                             -  136 -




     The time required by a walking collector to service 12

dwelling units with characteristics identical to the average

dwelling unit serviced by the Trashmobiles in Atlanta also

was estimated*


      W1 = 12 dwelling units

      W  =1.6 items per dwelling unit x 12 dwelling units =

           19 items

      W  = one trip per three  items  -x 19 items = 6.33 trips
                                               +     »
      W  = 45 + 2.55 (average  street to storage )  W
       4                                            x

         = 45 + 2.55(90)  12 = 3.29O ft


      Substituting into the  walking collector regression model:

    C =  0.18  - 0.12(12) + 0.12(19) +  0.24(6.33)  +  O.OO5(3,29O)


      =  18.99 min  of productive time


      Since the average walking collector is only 85 percent pro-
                                           18 QQ
 ductive,  the total time required would be  Q|85t  or 22.3 min.

      The Trashmobile operators averaged 17.4 min to service 12

 dwelling units with identical characteristics.  Therefore, the

 'average Trashmobile operator  is theoretically 28 percent more

 efficient than the average walking collector when servicing these


 dwelling units.
                                  - 137 -

          The walking collector crew costs  $15.OO per hr.   This  crew

     theoretically services a total of 24 dwelling units  each 22.3

     min.   The annual collection cost per dwelling unit would then be:


$15.00/hr  x 24 dweiiing'°Units x x hr/6O min x 1O4 collections/yr = $24.OO/yr


          The annual collection cost using  Trashmobiles was $21.SO per

     dwelling unit in this area.   Theoretically,  waste collection  by

     Trashmobile is 12 percent more economical than the conventional

     waste collection method.


                      General Comments and  Conclusions


          The Trash Taxis  and Trashmobiles  observed were operated  in

     the north-east section of Atlanta.  " This  area  has a varied  terrain

     with  many hills and steep driveways.   The homes  serviced were old

     large high-income family homes,  unlike housing development  areas

     of today.  The houses were located on  large  lots with  very  long

     driveways.  The average street to storage distance for dwelling

     units served by the Trash Taxis was 14O ft as  opposed  to only 9O

     ft for residences served by  the Trashaobiles observed.

          the satellite vehicle operators used SO gal rubber  barrels

     to transfer the waste from the dwelling unit storage point  to

     their vehicles.   The  average distance  for operators of both types

     of vehicles was about 3O ft.   The average distance for all  cities

     studied was only 15 ft.   Easier accessibility  to the waste  would

     result in an increased crew  efficiency for the collectors.
  Table B-3

  Appendix A

-------
                                - 138 -



      The  low number of items per residence, approximately .1.7,

 waif due to the.separate collection of all garden wastes.  Assum-

  ing garden wastes to be 30 percent of total wastes during the

  summerf" combined collection would decrease crew efficiency by

  approximately 10 percent.  Eighty-seven percent of the items

  collected during the study were stored in acceptable standard

  containers (Table B-7).  The percentage of miscellaneous items

  was lower than in other study areas.

      .The  regression models illustrate the large difference in

  efficiency between the Trashmobiles and the Trash Taxis in

  Atlanta.  Under identical conditions, the.Trashmobile was able

  to service 5O percent more dwelling units per elapsed hour than

  the Trash Taxi.  Since this was the first week of operation for

  the Trash Taxi and the Trashmobile has been in operation for 3

  yr, some  discrepancy is naturally expected.  However, this fact

  cannot account for the entire difference.  The lower crew

  efficiency of the Trash Taxi crew was also caused by particular

  deficiencies in the satellite vehicle.

      The  major problem which the Trash Taxi operators encountered

  during the vehicle's first week of operation was its. inadequate

  power to  climb steep driveways.  Many times the packer truck

  driver had to help push the vehicle out of a driveway in which
i                 ' .            •                        •
  it had become stuck.  This lack of power is partly due to the
                              - 139 -
                            TABLE B-7

                       ITEM CLASSIFICATION
               ATLANTA, GEORGIA, AUGUST'18-22, 1969
Date
Total items
 collected
                                      Percent of total
 Standard
containers
                                                Miscellaneous
8/18 .

8/19

8/2O

8/21

8/22


Week
    412

    38O

    265

    619

    389


  2,065
   83.1

   89.O

   77. O

   90.5

   90.5


   87.O
16.9

11.0

23.O

 9.5

 9.5


13.0

-------
                             - 14O -


provision of a hydraulic power inadequate for the size of the

vehicle and, the restriction of the vehicle's maximum speed to

28 mph.  In addition to lack of power, the vehicles suffered

steering column failures, frequent stalling, hydraulic fluid

leakage, gasoline leakage and emergency brake failure during

the one week study.  The leakage of hydraulic fluid damaged

asphalt driveways of customers in addition to being a mechanical

problem.  Unless several corrections are made and the vehicle can

produce more power, it cannot compete with the more efficient

Trashmobile in the City of Atlanta.

     The average unloading times for the Trash Taxi and Trash-

mobile were 1.5 and 1.2 min respectively-  Both of these times

were less than the average unloading time for Cushman scooters.

Twenty yd,  Loadmaster packer  trucks were used on all routes

observed.  The Loadmaster has a hopper  capacity of 1.5 cu yd and

two compaction cycles were normally needed to accommodate the

wastes  from the satellite vehicles, thus creating high unloading

times.
                               - 141 -

                             APPENDIX C

                      COLUMBIA, SOUTH CAROLINA
                   FIELD STUDY - JUNE 24-27, 1969


      Columbia, the capital of South Carolina,  is a. city of 98.OOO

 persons (Table C-l) located in the south central area of the state.

 The city lies on flat terrain,  but is  surrounded by hills.  The

 summer climate of Columbia is very hot  and humid and temperature

 and humidity readings of 10O are common during this season.

      The Department of Sanitation has had the  responsibility of ...'

 collecting and disposing of all  residential solid wastes in  the

 City since November,  1967.   Prior to this date waste collection

 had been the responsibility of a private company on contract with

 the city.   In the transition, the City  inherited poor equipment

 and the problems  of the  contractor who was  forced out of business

 due to  unsatisfactory service.

     The Department of Sanitation  presently provides  once  weekly

 curbside collection of "trash" (tree trimmings  and grass)  and

 twice weekly backyard  collection of "garbage"  (non-garden  wastes)

 for  the  26.OOO dwelling units in the City (Table C-l).

     There are 24 residential "garbage" collection  routes.  Cush-

man satellite vehicles are presently used on four of  these routes

 in the southeastern section of the City.  The City also uses con-

 tainer trains (a Scout and three container trailers) on six

 routes.  The Department inherited~*19 Cushman- scooters from the

private contractor, but only 13 are operational and only eight

-------
                                 -142

                                TABLE  C-l.
                   DESCRIPTION OF COLUMBIA, SOUTH CAROLINA
    .Population
    Dwelling  units
    Persons per dwelling unit
    Land area
    Population density

    Dwelling  density

    Hiles of streets
    Dwelling units  per street mile
(City)
(Study area)
(City)
98,000 (Est.  1/1/69)*
26,600 (est. 1/1/69)*
3.68       .        '
20.6 sq mi*
4,800 persons/sq mi
3,600 persons/sq ml
1,300 dwelling units/sq nri
(Study  area)   970 dwelling unlts/sq ml
        ^     560+
              50 dwelling units/mile
(Study  area)   70 dwelling units/mile
*From 1969 Rand-McNally Commercial Atlas and Marketing Guide.
^Engineering Department, City of Columbia, South Carolina. .
                              -  143 -

'are actually in.use..  Most of the  satellite vehicles are in
 various states of disrepair (Figure C-l),  due to the long lead
 time required on major part orders from the manufacturer.  The
 Department suffers from high absenteeism among the employees
 and'is unable to find'ehough employees  with driver's licenses
 to operate .the satellite collection vehicles.
      The four crews with Cushman satellite vehicles each consist
 of two men with satellite vehicles, two walking collectors with
 plastic barrels on wheels, and  a packer truck with a driver.

                        FIELD STUDY ANALYSIS  "

      The field study was conducted June 24-27, 1969.  During the
 four day study two satellite collection vehicle crews were .observed
 in the eastern section of the City.  This area was characterized
 by medium to high income houses on flat terrain with long drive-
 ways.  One crew used Cushman satellite  vehicles with extended  /
    .....           .                      H',
 hydraulic pumps to be used in conjunction with a side loading
 packer truck.  The other crew operated  Cushman satellite vehicles
 with flat beds carrying four 5O-gal plastic barrels.  A total of
 46 data points were obtained on five individual satellite vehicle
 operators.                   .

 Satellite Vehicle Collection Operational Analysis
      Stepwise regression techniques were applied to the satellite
 vehicle data in an attempt'to explain the productive time per

-------
                           -  144  -
Figure C-l.  Cushman satellite waste collection vehicle with
            flat-bed used in Columbia, S.  C.
                             - 145 -


satellite vehicle trip using the five variables  recorded  during


the study.  Individual regression analyses were  made for  two


operators, one using a flatbed vehicle, and one  using a dump


vehicle, and also on all of the data combined.   The regression

model which best described the productive time for all the data


combined was (Table C-2):


     where Y  =5.42 + 1.1O XJ + 3.O2 Xg


           Y  = productive time per satellite vehicle load


           X  = number of dwelling units serviced per load


           X  = route distance of the satellite  vehicle per

                load, in miles


     This equation was able to explain 71.7 percent of the total


variation in the data and the standard deviation of the residuals


was 20.2 percent of the response mean, productive time.   The


variable most highly correlated to productive time was 5C , the


number of dwelling units serviced per satellite  vehicle load.


Three of the variables recorded during the study were not  signifi-


cant in explaining productive tiae in this model.  The models for

the two individual operators included as many as five variables


and their coefficients were of the same order of magnitude as


the combined model.

     The regression model was tested for prediction accuracy by


comparing it to the actual field results for a particular situa-


tion.  The average variable values observed in the study  (Table C-3)

-------
. - 146











B
(J u>
0 -H
M
CM -
CM f-
W CM
O 1'
CJ 3"
CM '
.-3
U M
Q Z
CM Z 1
I O
CJ M <
E-i Z
W On
, i M »J
s ii
M g
O O
"«

ul
«*! *~*
CO












a




.,











Q)
Xi
(It

fl)
















• 4J C "O
c o a>
C4 Q) -H C
0 ^ -H .

a, ^' a.
to x
> 0)
" . -N
C
0 $ g
X CO Q)
C 4->
O

O ro
in *" C ^
x a v
0 2 **
•0-0.



4)
O C V
C -H bO
-* «0 U-O Hi
X 4-» -H 4-< t*
0) .C O
•H 0) V
O > V\

O. V «J
to- cfl O O. W
X *J -H 3 >
10 £ • *H
3 > £
•b
9> *O
en 4-j n)
CN e o o
X (U 0) t-»
"'5 &
o o
u a.
DC T) -O
C 0) Q
•H (0 O O
X •-! -H >
a> c £* b
g 3 « a>
« o
•| :
2
0)
s-




. . «•
in in H .
en 01 r*^




CN CJ CM
CO. d" rf
CM CM- in
• i i in

O tO CM
in . j- .0
co tn to
•- -•- •-.-




m
1 CO
fr H — 1
i • i
1- O 1
i i


i- 1-1 i
t 2 1


j- r* i
<"1 W 1
i^
O O 1


CO CO 0
en m r-H
O O -tH '

•o
o
B
*H
b § 1
                               -  147  -




 were used  in this case.   Productive time was  converted to  the



 total elapsed time per  load using the fraction of productive



 time for all satellite  vehicles observed in Columbia  (Table C-4).



      The average variable values were:  (Table C-3) -

                                 I

            JC = number of  dwelling units serviced per  load  = 11



            3C = route distance .of the satellite vehicle per load



  •              = 1.55 miles                 .



            Percent productive  time = 82.3 percent



      Substituting these values^ into the regression model:



            Productive time = 5.42 +  1.1O (11)  + 3.O2 (1.55)



                            = 22.2 rain per load



      The actual average productive  time observed during  the



 field study was 22.6 min  per  load.



      The predicted productive time  per  load is therefore 1.8 per-



 cent lower than the .actual observed time.  The actual total elapsed


                   22 6
 time per load was       or 27.5  min  due to  the normal  unloading time



 and associated other time. Two satellite  vehicles working together



 were able  to service 22 dwelling units  in  27.5 min for the average



 conditions observed in  Columbia.




 Packer Truck Driver Operational Analysis



      The packer truck drivers were  required to remain in their



 trucks due to faulty emergency brakes, which  are presently being



 replaced by air lock brakes.   The amount of collection done by



- the packer truck drivers  was  insignificant.   The drivers spent

-------
- 148 -



ION VARIABLES
1969
C-.
t* CM
CM
u
TABLE C-3
rELLITE VEHICLI
CAROLINA - JUt
;3
OH O
£w
ESS
3g
> 3







/ariable







' o
> c<
' 'S8-?^
>• 3 O O "•
O -M O
CO bO
CO (0 0) O <0 *"*
01 4-* Cj.LJCj.i->
Q w £ 5 >n
•H 0) +J %_/
Q »
 •'
TJ O.
So «^
J- flJ O O CO *£
W ^3 O -^
2 ? a
s „ &
§I-H) J X^.
O O. O -P
X 4-» -H 3 > M*
W J= >H ^^
as •*
•§•8
§+* o
O «H
0
V «H t,
M M 0
00.
00 -O T)
C O HJ
•^ ra o o
0
•HO XI
•riVH a 0
H O "O >
3SS 1
H O -H CO
Q > A
|


co m to
in in CM
«H CM CM
o o o
en co co
If) 1A If)
•H r- m
•-t H .-H
O O O
CO CO (O
o o o
en CM H
•H CM CM


co in •-!
• ID iH 10
1
O CJ O
O CJ O
                                                                                               - 149  -
                                                                                             TABLE C-4
                                                                            SATELLITE VEHICLE OPERATOR ACTIVITY ANALYSIS
                                                                             COLUMBIA, SOUTH CAROLINA - JUNE 24-27, 1969
Percent of Total Time
Operator
CCF,
CCF2
CCD1
CCD2
CCD3
Combined
Total
Minutes
Observed
293.8
225.9
343.6
279.0
122.8
1295.1
Productive
77.1
81.6
76.7
86.0
88.0
82.3
Unloading
11.5
7.5
5.7
10.5
8.1
8.3
Other
11.4
10.9
17.6
3.5
3.9
9.4

-------
                              -  150 -





over 60 percent of their total  time waiting at or  in  the packer




truck  (Table C-5).   Based on  driver activity in other cities,




all of this time could be spent more constructively either




collecting or assisting the satellite vehicles in  the unloading




process.  A recently adopted  department policy has given the




drivers the responsibility of directing and coordinating the




satellite vehicle efforts with  those of the walking collectors




to achieve better waste collection  efficiency.






                      System  Cost Analysis





     The following daily costs for  satellite vehicle  collection




crews were based on figures obtained from the Superintendent of




Sanitation in the City of Columbia  for the  period  of  March 1968




•through February 1969.  For study comparison purposes, the two




walking collectors in Columbia's crews were ignored and  the crew




was divided as two satellite  vehicle operators and the packer



truck driver.





Daily Crew Costs                        .




     The costs of a satellite vehicle waste collection crew.are  .




labor,  equipment operation and depreciation,  and overhead.  Equip-




ment operation includes repairs and maintenance in addition to




direct operating costs.  The  following costs  are expressed in




dollars per day.                                      ,•    .




     Labor                 . • .                 .            ,„




       2 satellite vehicle operators at $12.3O/day




         + .15 percent fringe benefits                   = $28.3O.
                  -"151"-



                TABLE  C-5



 --•  PACKER TRUCK DRIVER ACTIVITY ANALYSIS



COLUMBIA, SOUTH CAROLINA - OUNE 24-27, 1969
Driver
CG20
CY30
Combined
Date
6/24
6/25
6/26
6/27
6/24-27
Miles per
driving
hour
8.3
9.4
6.1
4.1
6:2

Assisting
0.0
0.8
10.8
10C6
6.5
Percent
Collecting
0.0
. 0.0
0.2
6.8
1.5
of Total Time -
Driving
14.9
12.0
20.2
27.8
19.4
Waiting
72^1
83.6
57.1
'46.3
61.8
Other
13.0
3.6
11.7
8.5
10.8

-------
                        - 152 -






  1 packer truck driver at $14.9O/day + 15 percent




    fringe benefits                                  = $17.15




                                  Total labor cost     $45.45




Satellite Vehicle Operation and Depreciation




  Operation




  2 satellite vehicles each at $41O/vehicle/yr       = $ 3.2O




  Depreciation




  2 satellite vehicles, $2,6OO new, 4-yr life        = $ 5.OO




                  Total operation and depreciation     $ 8.2O





Packer Truck Operation and Depreciation




  Operation



  1 - 2O cu yd Garwood "7OO" at $l,5OO/yr            = $ 5.78




  Depreciation




  1 - 2O cu yd Garwood, $15,OOO new, 8-yr life       = $ 7.21




                  Total operation and depreciation     $12.99




Overhead




     Overhead costs were unable to be obtained and




were therefore estimated to be 2O percent of all other




crew costs.




     Overhead Costs =  -2O (45.45 + 8.2O + 12.99)     = $13.33




Total Crew Costs




     The total cost per day for this average satellite




vehicle crew is $79.97
                              - 153 -






Annual Collection Cost per Dwelling Unit





     The annual cost to the average dwelling unit observed in




Columbia for twice weekly collection was determined by multiply-




ing the crew efficiency by the crew cost rate.  During the field




investigation, the crews observed averaged 5.5 hr per day in the




actual process of collection.  The true crew cost rate is then




$79.97/day x 1 day/5.5 hr or approximately $14.5O/hr.  The crew




efficiency was 22 dwelling units per hr.  The annual cost per




dwelling unit is then:






                 81d"^  x 104 collections/d.u./yr ~ $31.SO







Collection Cost Per Ton





     The residential waste generation rate for Columbia was not .




able to be determined due to the unavailability of truck scales




at the time of the study.  The national residential waste genera-




tion rate of 3.O Ib/capita/day was assumed, and the collection




cost per ton was calculated.  The amount of waste collected per




hour by the crew would be:




 48 d.u./hr x 3.68 persons/d.u. x 3.O Ib/person/day x 3.5 days between




                     collecting = 1.85O Ib/hr





The collection cost per ton is then:





      $14.5O/hr x 1 hr/l,85O Ib x 2.OOO Ib/ton = $15.5O/ton

-------
                             - 154 -
     Satellite Vehicle Collection vs^ Conventional Collection





     The time and costs required to service the dwelling units




observed in Columbia by walking collectors was estimated and




compared with the time and costs using satellite vehicles.  To




facilitate the calculation, the time required to service 11




dwelling-units, the average.number per satellite vehicle load




was used in the calculations.  Using average values (Table C-3):






   '  Wj = 11 dwelling units




     W2 = 1.9 items/dwelling unit  x 11 dwelling units = 21 items




     V*3 = 1 trip/3 items  x 21 items = 7 trips




     W4*= J45 + 2.55 (average street to storage)! -Wj = 2ff4O ft






     Substituting into the regression model for conventional




collection (Appendix A):                          .      .'..•'•





     C = 0.18 - 0.12(11) + 0.12(21) + 0.24(7) + O.OO5(274O)




     C = 16.7 minutes of productive time





     Walking collectors are productive approximately 85 percent




of the tine which they spend on the route (Appendix A).  The total




time for one walking collector to service 11 dwelling units then




becomes :.



                       16.7
                        .85
                            or 19.6 minutes
 Appendix A
                               - 155 -






     The average satellite vehicle operator in Columbia required .




27.5 min to service the same number of dwelling units.  Use of




average walking collectors in Columbia would theoretically increase



collection efficiency by 4O percent.




   -  The cost of a crew with two walking collectors and a packer




truck with driver would be equal to that of a satellite vehicle




crew minus satellite vehicle depreciation, operation and overhead




costs.  This cost is $79.97 - 1.2O(8.2O) = $70.13 per day.  Since




the crews actually collect for 5.5 hr on the route per day, the




true cost is approximately $12.OO per collection hour,  the two




walking collectors together collect 22 dwelling units per 19.6




min or 67 dwelling units per hr.  The packer truck driver does not




participate in collection.  The annual collection cost per dwell-




ing unit is then:





     $12.OO/hr x 1 hr/67 d.u. x 1O4 collections/d.u./yr ^ $18.SO





     The collection service cost using walking collectors is




theoretically only 58 percent of the same service cost using




satellite vehicle collection.

-------
                             - 1S6 -
                       OPERATIONAL COMMITS






     There were several constraints which severely limited the




collection efficiency of the satellite vehicle operators in




Columbia.  The primary hindrance was the hot weather during the




study period.  The temperature and humidity readings approached




1OO on all four days of the study.  The operators had to pace




themselves to avoid heat exhaustion, and dehydration, which could




result in severe sickness or death.




     The poor mechanical condition of the satellite vehicles was




another factor adversely affecting the efficiency of waste collec-




tion.  Frequent engine stalls and tire punctures were observed




during the study.  Frequently the vehicles did not start in the




morning and delayed the operators from getting to their routes.




     General thoughtlessness on the part of the residents contributed




to the inefficiency of the system.  Accessability to the waste




storage area was imbedded by cars blocking driveways, closed fence




gates and menacing dogs.  This resulted in an increased walking




distance for the operators from their vehicles to the waste storage




point.  The average distance observed was 3O ft, twice the average




walking distance observed in the other five areas studied.  The




satellite vehicle operators used SO and 6O gal plastic barrels to




carry the wastes from the storage point to the satellite vehicle.
     The satellite vehicle unloading times observed were the




longest of the six areas studied.  The average time required for




unloading the extended hydraulic dumps into a 32 yd side-loading




packer was 2.5 min (Figure C-2).  The flatbed vehicles had an




average unloading time of 2_. 1 min.when unloading into a 20 cu yd




Garwood TOO rear-loading packer with a 1-1/2 cu yd hopper.  Two




packing cycles of ttte Garwood were required to accommodate each




satellite vehicle waste load of 1-1/4 cu yd.  Use of a packer




with a larger hopper could reduce unloading time by 50 percent




or more.  Packer trucks with rear loaders were currently on




order to replace all side-loading packers in Columbia.




     Two significant improvements could be made in the satellite




vehicle collection systems observed in Columbia.  Better coordina-




tion between the satellite vehicle operators would eliminate the




frequent duplication of services.  The duplication of effort




observed during the study occurred for 34 dwelling units or 6.5




percent of the total number of dwelling units serviced.  This




amounts to 3/4 of a dwelling unit extra per satellite vehicle load.




     The second improvement would be to reduce the route distance




traveled by the satellite vehicles during each load.  The average




distance observed during the study was 1.55 miles, over three times




the average for .the other five study areas.  By working closer




to the packer truck this distance could be reduced to approximately




O.5O miles and increase collection efficiency significantly.

-------
                          - 158 -
                                                                                                               -  159  -
Figure c-2.  Unloading satellite vehicle with extended hopper
            into side-loading packer.
                                                                                      The prediction model was used  to quantify the  amount of time


                                                                                 wasted by the two: system deficiencies discussed above.   The time


                                                                                 required .to . service the O. 75 .duplicate  dwelling unit  and to travel


                                                                                 the extra -1.O5 route 'miles was:    •..''"-•
     Wasted time per  load =  5.42
                                                                                                                          +  l.'lO  (O.75  duplicate
     dwelling units)  + 3.O2  (1.O5 extra route. miles) = 4.4 min



     This wasted  time was  approximately 2O percent of the average


productive time required per load during the study.  Thus, system


.improvements in these two  areas could', improve the efficiency of the


total system by as  much as 2O percent.

-------
                             - 160 -





                            APPENDIX D


                        KNOXVILLE, TENNESSEE


                   FIELD STUDY, JULY 14-18, 1969



     Knoxville is a city of 187,OOO people located 40 miles north-


east of the Great Smoky Mountains in Tennessee (Table D-l).  The


terrain of 1 the City is varied with steep hills common in many areas.


     Sanitary Disposal Company provides residential waste collection


on contract with the City to approximately SO percent of the popu-


lation.  The Company services the entire area surrounding the


central business district on a once per week collection basis.


The collection charge is $2.SO per dwelling unit per month.  The


City provides collection of garden wastes to these same areas once


weekly.


     Sanitary Disposal Company created  the Trashmobile, an  Interna-


tional Scout chassis with a 1-1/2 cu yd dump, to solve  labor problems,


reduce costs and increase efficiency in the hilly areas of  Knoxville.


Scout trucks are now being replaced with  Datsuns as  they appear  to


require less maintenance than  the Scout.   Sanitary Disposal operates


2 to 3 Trashmobiles on each of  12 collection routes  in  the  City  of


Knoxville.  The Company also has additional private  routes  in


annexed areas of Knox County at a cost  of $2.SO per  month  for  once


per week collection.
                                                                                                                       - 161  -
                            TABLE D-l


                       KNOXVILLE, TENNESSEE
                            STATISTICS
Population

              *
Dwelling units


Persons per dwelling unit

         *
Land area


Population density


Housing density


Miles of street-


Dwelling units per mile of street
187,OOO (1/1/69)


52,33O


3.57


77 sq miles


2,43O persons/sq mile


68O houses/sq mile


998


SO
 From 1969 Rand HcNally Commercial and Marketing Atlas

*
 From Advanced Planning Commission, City of Knoxville, Tennessee


-From Engineering Department,  City of Knoxville, Tennessee

-------
                              - 162 -





                       Field Study Analysis





    • The field study was-conducted July 14 thrpugh 18 in the'very




hilly eastern section of the City of .Khoxville.  Four. Trashmobile




operators in two different-crews.were observed during the week.




Two collectors used Datsuns and two collectors operated Scouts.




The activities of two packer truck drivers'"were also recorded.





Satellite Vehicle Collection Operational Analysis      ;




     Stepwise regression analyses were conducted on the four




satellite vehicle operators individually and together (Table D-2).




In addition to the regression analyses using five variables,




supplemental regressions were made using seven variables.  The




two extra variables resulted from classifying items as standard




containers, 55-gal drums, and miscellaneous items.  This analysis




was conducted to determine the adverse affects of SSrgal drums




on the operator's collection efficiency.




     The regression model which best describes the 287 satellite




vehicle loads observed in Khoxville, using five variables, was




(Table D-2): .
     P = + 0.57 + O.77XJ + 0.16X2 + O.OO2X3 + O.O39X4 +




where Y  = Productive collection time per satellite vehicle load,




           in minutes




      X  = Number of dwelling units serviced per load




   '.-:' X ' = Number of items collected per load




      X  = Average distance satellite vehicle goes up each dwelling




           unit driveway, in feet-
- 163 -












in
§••
o S
Cz. rH


82
i

B rH

8 S
? 5V
D rH 1
05 Cm


^ g|
§V
S 3*
>ri
El.
s§
a













• ' B




















rH
a
45'
n
a


















c TJ .
4J O 0)
C -H C
o  o
c
a B
9? "S 8
C *•
5

an
«' e 8
own
a -H m
•o a.


o •
u a «
C rH CA
J CO O O CO
a! .2
0) >.
U O CO
B rH 3
en a u Q. 0)
X +> -H a >
a £ -H
as «
•o
CO V <8
N e o o
X 01 0 rH
• +* rH
M rH r4
0 9
u a
01 TJ TJ
c on
•H 0) O O
rHrH V *H rH
X rH *H >
01 C M H
a a a a
a no.

• o

(Q
01
*


m .«
in in
N 00


'
rH en
1^ O
o o
1


O rH
^ c»






CM t»
S 8
d d


n 
CM
d
*

8-
o






cf
fe


^*
o
t^



Is-
m
O



5
rH







S
d


8
d



-------
          .                  •  - 164 -



      X. = Average distance from satellite vehicle to storage



           point, in feet                                  .



      X  = Route distance of satellite vehicle per load, in miles




     This equation was able to explain 70.4 percent of the total



variation in the data, and the standard deviation of the residuals



was 17.4 percent of the response mean, productive collection time.



The most highly correlated variable to productive collection time



was the number of dwelling units serviced.  The coefficients for



each operator are of the same order of magnitude for corresponding



variables, with the number of dwelling units being the most highly



correlated for three of the four operators observed (Table D-2).



     To illustrate the utility and accuracy of the regression model,



its results may be compared to the actual field observations.  Using



the average values of the variables observed in Knoxville (Table D-3),



the time required to make one round trip with the satellite vehicle



was predicted.  To account for unloading time and other time, the



productive time must be divided by the fraction of productive time



(Table D-4).  The average values of the variables for this example



were:



     >(  = 6 dwelling units per load



     X2 = 15 items per load



     X  = 1OO ft up each dwelling unit driveway



     X  = 2O ft from satellite vehicle to storage



     X_ = O.5S miles route distance



     Percent productive time = 87.1 percent
                   - 165 -
                  •o   «-

                  O   00
             t  ^
             g  o
     .S
        a o a f
   i *   ^ ^ M *M
   I O a> *>   SC..
   1   -  •   *»
          O T)
          O «-^
        O C O ffi
        *; 3 rt rt
 S   §  2  S   S
        rt  rt   rt






 !  3  8   S' S
      u v

.Sl^sups-v
h

5
                 o   o   o
o   o
«   o»
                            O   o
O   o
M   K
                2  S   8  S  8
                        •-1  -I  rt
                S   S  *   «>
    01  -a TJ
    c  mo
    •H 0} O O
 *H y "Q
               
-------
                   - 166 -
                 TABLE D-4




SATELLITE VEHICLE OPERATOR ACTIVITY ANALYSIS

   KNOXVILLE, TENNESSEE - JULY 14-18,  1969
Operator
KTS
FS2
KTD1
KID,,
Combined
• Total
minutes
observed
517.2
764.1 .
925. O
981.8
3,188.1
Percent
Product ive
8O.2
89.1 --..- .
88.8
87.5
87.1
of total time
Unloading
8.5
.7.3
8.4
9.2
8.4

Other
11.3
3.6
2.8
3.3
-!4.5
                              - 167 -





For one satellite vehicle load:




     Productive time = + O.S7 + O.77(6) +0.16(15) + O.OO2(10O)




                       + 0.039(20) + 1.75(0.55)




     Productive time =9.7 min per load
                                                                               Total elapsed time
                         9.7
                                                                                                       - 11.1 min per load
                                                                               Actual average productive time per load = 9.7 min per load
                                                                               Actual total elapsed time =
                                                                                                            9.7 _
                                                                                                           .871
                                                                                                                = 11.1 min per. load
                                                                               The predicted productive time required to collect this average




                                                                          load is  identical to the actual time observed.   Two Trashmobile




                                                                          operators in a crew were able to service 12 dwelling units in 11.1




                                                                          min.     .





                                                                               The regression model for the Trashmobile operation in Knoxville




                                                                          when the items collected were classified into three categories was




                                                                          (Table I>-5):





                                                                            Y = 0.59 + 0.71XJ + 0.22X2c + 0.3OX2fe + O.HU^^ + O.O28X4 + 2.O4X-






                                                                          where X    = number of welling units  per load




                                                                                \2c = number of standard containers collected




                                                                                X_,  = number of 55-gal  drums collected




                                                                                X    = number of miscellaneous  items collected
                                                                                • ^m             '



                                                                                X4  = average distance  from the satellite  vehicle to the




                                                                                      storage point




                                                                                X5   = route  distance of satellite vehicle,  in miles





                                                                              This  equation was able to explain  7O.4  percent of  the total




                                                                          variation  in the data, and the standard deviation  of the residuals




                                                                          was 19.8 percent of  the  response  mean.  The  model  coefficients were

-------
- 168 -
















(fl
' M
rj
b.
g g

H
W cd

I ° is
m z O %
1 O ri S
Q »HH
§Gg -

< 3fcH
H O M M
^ X*
3 3 §
0 H
M I
5 H
> 3
cc
tt
d
H
S!


















tt























0)
rH
O
a
•H
H
>



























§•0
01
§ 'H -S
O fl] ffl •
M *H rH
01 H p.
0. a 'X
• >  0 rH

Q  a]
01 >.
— o e «
X CD O P* 5
^ 4J -H 3 >
0) £ -H
•HO) K
Q > TJ

m
-~ C <0 01
e» « e v
X ^1 rH +> 01
3 U rH
Z n o
•H O
e

 I I

o o



O i-* rH CO iH
o o o o o






o N m tn
o o ' o d





CO CM tH 0) i
N en ^ oj
'dodo ;
1



IO rH ^ n iH
rH «O *O t^ t^
rH O O O O
•o
01
c
iH CM iH CM tt
tn to o Q 8
b b b b ,9
                                                                                   - 169 -





                                                     similar to values in the unclassified container model discussed




                                                     above.  The coefficient for variable X2b' numoer °^ 55-gal drums,




                                                     was larger than the coefficients for standard containers and




                                                     miscellaneous items, indicating the increased time required to




                                                     collect waste from 55-gal drums.  Container classification was




                                                     extremely important in describing productive time for operators




                                                     KTO2 and KTS2 (Table D-5).




                                                          Using the variable values from the first example with con-




                                                     tainer classification, the classified container model was compared




                                                     to the unclassified container model.  The 15 containers serviced




                                                     per average load included nine standard containers, two 55-gal




                                                     drums and four miscellaneous items (Table D-6).  Substituting




                                                     into the model:




                                                          Productive time = O.59 + O.71(6) + O.32(9) + O.3O(2) + 0.14(4}




                                                                            + O.O3(2O) -r 2.04(0.55)




                                                          Productive time = 9.8 min per load




                                                          Actual average productive time = 9.7 min per load





                                                          Comparing the results of the two models, it can be seen that




                                                     the regression model without container classification is able to




                                                     better describe productive time per load than the predictive model




                                                     with container classification, for.this case.





                                                     Packer Truck Driver Operational Analysis




                                                          Regression analyses were made on the data recorded for the




                                                     collection activity of the two packer truck drivers observed




                                                     (Table D-7).  Based on 356 data points, the best regression model

-------
                  - 170 -
                                                                                                           - .171 -
               TABLE.  D-6

        CONTAINER  CLASSIFICATION
KNOXVILLE, TENNESSEE  -  JULY 14-18,  1969
Operator
Krs1
KTS2
ICTD1
KTD2
Combined
Average
number
of items
per load
16
14
14
18
15
Classification
Standard
containers
1O
- 8~ "
8
12 •
v9
55-gal
drums
2
-.- 2
2 '
... . 2 - • .
2
Miscellaneous
items
4
4
4
4
4










in
Is 8
5 <" •
M fH
a .
b, CO

8 ^
^ 2
r> S "*
(4 S: '


J f-t t.0
CO U JO
£ 3|

O H
5 d
Q >
K o
§ i
g














c
0














0
i-f
(0
-H
h

>



















~ -*~ ,
§~ T)
ffi
C -r< C
M 0) -p -H
: o « a
fc -H fH
VHP,
o. S S<
> 8
c .
« me
O +J h
C f
3


OJ O

en o 10 j^ to
n a 3 o
> -H H f
< TJ *> 0


N h e u
_fi Q) O o
. ja +> fH
£ *H fH
3 O
z - o

«g
«s.
0 §0 §
fH h Ol i
.O *H OJ J^
1 |H 0 0



•6 «
0)



M
V

•H
H
"

o*

Q\
to




o*
CN»
O
1




8
d



o
en

o
*





S
d








rH
m
CM
y

00

m
03




00
o>
d
i




O1
d




in

°





-o
in
d








CM
in

$

m

-,_!





CM
in
d
i



o
0
d



rH


O





m
d






•o
a
c
-H
.a
§
B







§
•H
•H
u
rH

O
U


-------
                              - 172 -






for explaining productive collection time for the two'packer truck




drivers observed in Knoxville was (Table D-7):





              T = - O.52 t O.57b  + °-31b2 * O.OlOb3





where T  = Productive 'collection time required to service b_




           dwelling units between truck moves




      b- = Number of dwelling units serviced'between truck moves




      b  = Number of items collected from b  dwelling units




      b  - Average distance.which the driver walked from truck




           to storage point,  in feet





     This model was able to explain 71.5 percent of the variation




in the data,  and the standard deviation of the residuals was 34.2




percent of the response mean.  The model coefficients for the two




individual drivers were very similar, with variable b , the number




of items collected, being the most highly correlated to productive




time.   The significance of variable b  may be attributed to the




diversity in the types of containers used in Knoxville, and the




corresponding diversity in collection time required.




     The packer truck driver collection model may be tested




similarly to the satellite vehicle operator model.  For this




example, average variable values were used (Table D-8).  The




time devoted to collection by the packer truck drivers in Knox-




ville was 44 percent of total time (Table D-9).
- 173 -




i
5
«!
>
o *o
g s

3 CO
»j rH
84
i-t
§ >.
•>
& ° ,
« B S

g HS
g i


0. -
a 3
e|
il
§
u
2

§













' 00
o
rH

(0
•H
a












01
•H «
•H' O C
a. o 6 (S
>< 3 -H B
•6 *> ^
0
0) O
So +> 01
B O>
m  1-1
3 O
z u

n a)
•H ^ m

IH K o10 i
J « C T)
J3 -H 01 J<
B rH O O
3 *-* *H 3
Z g > H
I >J *»
•O Q)
ai

m
•H
a R
3 -H
^ (0
•o
V

•H
*


0 •« (\)
0) M N



§ ' g 'g





cn f tn







•-t O r-l






OQ CD 
-------
- 174 -












UJ.S
U) rH

< CO
3 *"*

^ "
. &i

A •§ ,
td OJ M
~j W W
3 512
H g|

u *•*
H aT
K H


u R
0, 'Z















1
rH


O
•8

+»
c
Q)
J^ .
0)
CL








m
r-l
•H
Q

g.
CO
M





h
O
'f-
' C
•. -P
"rl
(0
3

c
>
•H
fi

Ch
C
•H

U
0)
iH .
_rH


c
•H
to
•H
to
5


3


14
9 Ol
a c
•H
>
a
8
V
•H
fi-


(n to ^ co co co
Tf rH O\ CO CO W
.. r-l d *~*
CM i^ . CM m ** cy
C4 h- "* CM CO* 0»
' r-t . • " - H



m. ^ w * e^ r* -,,r
O o\ <* o» <)•••; rsW S

• .^'SB?
—



o» o t^ ' m en «>
O **" c> CM o> <*> -
^- «* rf (fj Tf ^.



rH O CO rH CO r-
f*J f«. *O 1** 'O CO
N rH N - iH r* '






O ** " CO in rH \O
(^ , t*. o m *o *o


*Q m co *o N
CJ
c
rH '• • (M *H - '
tn ' ' m ja
S3 S o
2 I . U
                               -  175  -


     b  =  1 dwelling unit collected  between  each  truck move


     b  =  3 items      .


     b3 =  80 ft walking distance .


     Productive time = - O.52  +  O.57(l)  +  O.31(3)  +  O.O1O(8O)


     Productive time = 1.8 min


     Actual average productive time  =  2.2  min


     The 'predictive time required is 18  percent below  the  actual


productive time observed.


                       System  Cost Analysis       .


      Complete cost records were not made  available  by Sanitary,

                                                    ."V  '
Disposal Company.  Labor costs were  obtained from employees and


satellite  vehicle operation costs were obtained from manufacturers


literature.


     Daily Crew Costs.  The costs associated with a  satellite


vehicle waste collection crew  are labor  wages and fringe benefits,


equipment  operation and depreciation,  and  overhead.  Equipment


operation  includes repairs and maintenance in addition to  direct


.operating  costs.  The  costs below are  expressed in dollars per  day.


Labor      •                          -


     2 - satellite vehicle operators at  $85/week  +


         15 percent fringe benefits                        = $39.1O


     1 - packer truck  driver at  $95/week + fringe


         benefits                                          = 2-1.85


                                             Total labor   = $60.95

-------
                             - 176 -



Satellite vehicle operation and depreciation



  Operation                    -        •     •             .


     2 - satellite vehicles each at $65 per month         = $ 5.98



  Depreciation                                                   .


     2 - satellite vehicles, $359O new, 3 yr life         =   9.20


                                                            $15.18



Packer truck operation and depreciation



  Operation


     1 - 25 cu yd Garwood 7OO at $2,6OO/yr                = $1O.OO



  Depreciation


     1 - 25 cu yd Garwood TOO, $23,OOO new, 5 yr life     =  17.69


                                                            $27.69



  Overhead


     Overhead cost was assumed to be 2O percent of all


     other costs.  Overhead cost = O.2O(6O.95 + 15.18 + 27.69) = 2O.76



Total crew cost


     The total daily satellite vehicle crew cost          = $124.58




     Annual Collection Cost Per Duelling Unit.  The annual collection


cost for the average dwelling unit observed during the field investi-



gations in Knoxville for once weekly collection was determined by



multiplying the crew efficiency by the crew cost rate.  The crews


observed during the study averaged 7.5 hr per day in the actual



process of collection.  Therefore the true crew cost rate was


$124.58 per 7.5 hr or approximately $16.SO per hr.
                                  - 177 -




         The efficiency of the crew is the total number of dwelling



    units per hour serviced by the packer driver and the two satellite



    vehicle operators.  The satellite vehicle operators averaged 12



    dwelling units per 11.1 min or 65 dwelling units per hr.  The



    packer driver services one dwelling unit per 2.2 rain and collects



    44 percent of his time for a total of 12 dwelling units per hr.



    The crew collects a total of 77 dwelling units per hr.   The annual



    collection cost per dwelling unit is then:



      $16.5O/hr x 1 hi/77 d.u. x 52 collections/diu./hr = $ll.OO/yr




         Collection Cost Per Ton.  The residential waste generation rate



    for Knoxville was not able to be determined due to the unavailability



    of truck scales at the time of the study.   Assuming the national



    residential waste generation rate of 3.O Ib/capita/day, the coll-



    ection cost per ton was calculated.  The amount of waste collected



    per hour by the average crew would be:



77 d.u./hr x 3.57 persons/d.u. x 3.O Ib/person/day x 7 days = 5,75O Ib/hr



    The collection cost per ton is then:



           $16.5O/hr x 1 hr/5,75O Ib x 2.OOO Ib/ton = $5.5O/ton




         Satellite Vehicle Collection vs. Conventional Collection.   It



    is important to determine if the present satellite vehicle system



    used in Knoxville is more efficient and economical than a system



    using walking crews might be.  An estimate of the efficiency and



    economy of a walking system was calculated using a regression  model



    similar to the one developed for satellite collection vehicles.*



    The walking collector regression model  is:




              C = 0.18 - 0.12W.  + O.USW,  +  0.24W, + O.OOSW_
                              •*-        Z         3        4




         Appendix A

-------
                                -  178 -  ' •  •





-'•where C  = productive, time,  in minutes,  required to service V



            dwelling  units              ..--  ' ...  _            ;



                                          to be collected



     .. W- = total number of  items to be collected from W,
W  = the number of dwellingjunits
 .1   ._-.-.-.•   ;    ...  •;.,•
                dwelling units.             .



           W  = total number of trips to the truck while servicing



 ~ •  '    -      W  dwelling units  '        .  .                     '    .--.--•



           W  = total distance, in feet, walked by collector while



     •       .    servicing W  dwelling units                    .





          The productive time required to service six dwelling units .



     with the average characteristics observed during the field study



     was compared to the required time using satellite vehicles (Table D-3).




           W  = 6 dwelling units                   •



           W  = 2.5 items/dwelling unit  x 6 dwelling units = 15 items



           W3 = 1 trip/3 items* x 15 items = 5 trips



           W4 =  &5 +2.55 (average street to storage)]  X^



              =  fas + 2.55(110)]  6 = 1,950 ft




     Substituting into the walking collector regression model:



C = O.18 = O.12(6) + O.12.(15) + 6.24(5) + O.OO5(1,950) = 12.2 productive min




     Since the average walking collector is only 85 percent productive,



     the total time required becomes   ^ or 14.4 min.



          The same six dwelling units were collected in a total time of     .



     11.1 min by satellite vehicle.  Therefore the average satellite
      -JTable D-3

        Appendix A
•'.'•'                          - 179 -.                      ,




  vehicle operator is 30 percent more efficient than the -average



  wa Iking" collector under the se conditi ons .       -



       The cost of a crew with two walking collectors and a packer



  truck driver in Ktaoxville amounts to $1O8.O7 per day.  Converting



  to an actual, workday of 7-1/2 hr increased the true crew cost to



  approximately $14. 5O per hr.  Assuming that the packer driver .would



  collect at the same rate as he presently does with the satellite



  vehicle crew,  the total production of the crew would' be:
                                                                                      ;,.,6 dwelling units   6O min.   12 dwelling units _

                                                                                      2( -        - ~ x       ^ * --     - =
                                                                                                   14.4 min

                                                                                                                                                 -_

                                                                                                                                                 62
                                                                                               The cost per dwelling unit  is then:



                                                                                             $14.5O/hr x .1 hr/62 dwelling  units x 52  collections/yr  =  $12.OO/yr




                                                                                                    The collection service cost per dwelling unit for  the satellite



                                                                                               vehicle operation was $11. OO per dwelling unit per year.   Thus  the



                                                                                               conventional walking system cost would theoretically  be 9  percent



                                                                                               greater than the present satellite vehicle system.




                                                                                                                      Operational Comments



                                                                                                    The collection routes  observed during the field  study were in



                                                                                               very hilly areas with narrow, winding  roads.  The dwelling units



                                                                                               were medium' and high- income single family structures  located on



                                                                                               large lots with long, sloping driveways.   '                .



                                                                                                    The customers of Sanitary  Disposal Company were  not considerate



                                                                                               of the collectors in waste  container type and storage point.  Heavy-



                                                                                               duty military style waste containers and 55-gal drums were quite



                                                                                               common.  Of the total items collected  by the Trashmobile operators,.

-------
                              - 18O


9.5 percent were 55-gal drums.  The homeowners set the containers

back -from the road on collection day, instead of moving the con-

tainers forward to facilitate collection.

     The wastes collected had a high percentage of food wastes

causing maggots and raalodors due to the once weekly collection

during the hot and humid summer season.  The high moisture content

of the waste resulted in the production of a large amount of drain-

age with each compaction cycle of the packer.  A drain in the packer

truck hopper allowed moisture and maggots to drop onto the street

on several occassions.  Paper and plastic bags were not used to

wrap the wastes, resulting in a loose, uncontainerized waste for

collection.  Sanitary Disposal Company is promoting twice weekly

collection and use of plastic bags to combat these problems.

     The average unloading time for the Trashmobiles, O.9 min, was

the lowest of all six areas studied.  The Trashmobiles used in con-

junction with 18 and 25 cu yd Garwood Load-Packer 7OO's with 1-1/2

cu yd hoppers required only one compaction cycle to accommodate an

entire load.  The sweeper blade of the packer scrapes the wastes

from the hopper while it is raised in a vertical position.  Spil-

lage occurs when the Trashmobiles lack a rubber flap on the rear

of the hopper, or when the packer truck hopper does not have a

vertical metal extension welded to the lip of the hopper.  Spillage

was cleaned up from the pavement by the packer truck driver, using

a broom and a dustpan after each unloading.
                       - 181 -
Figure D-l.  Packer truck driver assisting unloading
             of satellite vehicle.

-------
                              - 182 -                       .   '   .


     the packer truck drivers spent 43.3 percent of their, time .

collecting.  The. drivers used SO-gal light-weight cardboard barrels

to carry was'tes from the dwelling unit storage point to the packer

fruble to prevent having to retiirh homeowners' waste containers.     .

The drivers were able to collect approximately 12 dwelling units

during each hour on the route.  The efficiency of the driver's

collection was decreased, as 13.0 percent of'the items collected ™« •"—-

were 55-gal drums.  Approximately 20 percent of the driver's time

was used to assist the Trashmobile operators in unloading and to

direct them in their operation (Figure D-l).

     The collection operations observed in Knoxville were very

efficient.  The Trashmobile operators provided the customers with

responsible and careful service.  Operators were especially care-

ful to return waste containers to their original location and

replace the covers.'  In addition, the operators did not overload

their vehicles, and kept littering at a minimum.  Wastes which

did fall from the vehicles were always picked up by the operators.

A  little thoughtfulness on the part of the homeowners would result

in an even more sanitary and efficient system than the operators

are now providing.            -               •        '
                                -  183  -
              :             .  APPENDIX E

                          MEDFORD, OREGON   .
       ••.;•'     .     FIELD STUDY, AUGUST 4-8,  1969


     The City of Medford, Oregon  is  located  in the flat agricul-

tural  Rogue Valley of  southwestern Oregon.   Medford has a moderate

climate with marked seasonal characteristics.  High temperatures

in the summer average  slightly below 9O degrees and the air  is

very dry.  The lumber  and wood products industry is the largest

employer in the city as  timber is brought into Medford for pro-

cessing from the surrounding mountains.  The total population/of

the city is estimated  to be 3O.5OO (Table B-l).

'  .   City Sanitation Service Incorporated is franchised to collect

residential solid wastes in Medford  and surrounding areas of

Jackson County.  The company provides  twice  weekly collection to

approximately 5,SOO dwelling units in  Medford.  The collection

charge is $33.OO per year for one 3O gal container per collection.

For the 2,OOO dwelling units in the  outlying areas on once weekly

collection the charge ranges from $18.OO to  $3O.OO per year.

     The company began operating Cushman satellite vehicles  in

1966 and has been very satisfied with their  performance ever since.

City Sanitation currently operates one crew with two satellite

vehicles and four crews with one satellite vehicle and one walking

collector.   In addition, each crew has a packer truck and driver.

-------
                               - 184 -







                             TABLE E-l



        VITAL  STATISTICS OF THE CITY OF MEDFOHD, OREGON
Population


Number of dwelling units


Persons per dwelling unit

         2
Land area


Population density


Dwelling unit density


Miles of street2


Dwelling units per street mile
30,500 (1/1/69)


1O.4OO (1/1/69)


2.93


11.3 sq mile


2,TOO per sq mile


92O per sq mile


119.2


9O
      1969 Rand-McNally Commercial Atlas and Marketing Guide
      Engineering Department, City of Medford
                                                                           - 185 -
                         Field  Study Analysis



      During  the 5  day  study three  satellite vehicle operators and


 two packer drivers were  observed.   The collection routes  followed


 were  on very flat  terrain and  the  homes were predominantly medium


 and low income family  dwelling units.  Four days were spent on


 twice weekly collection  routes and 1 day on a once weekly collec-


 tion  route.                                              '


      Satellite Vehicle Collection  Operational Analysis.   Stepwise


 regression techniques were used on the data from each individual


 operator and  from  the three operators together (Table E-2).  The


 regression model which best described the productive time for coll-


ection for the three satellite vehicle operators in Medford was:




              Yp = - 3.02 + 0.40^ + 0.42X2 + 0.01X3




                   + O.O9X  + 3.5OX^



where Yp = productive collection time per satellite vehicle


           load,  in minutes


      Xj = number of dwelling units serviced per load


      X2 = number of items  collected per  load


      X3 = average distance the satellite vehicle goes up


           each dwelling  unit  driveway, in feet


      X4 = average distance  from the satellite vehicle to


           storage point, in feet


      X^  = route  distance of satellite vehicle, 'in miles

-------
- 186 -









in
§
M.
g§ '.
m (H.
00
(d ^f
la

o 5
H <
5 •

O u
°|
H O
X a
H o
> &

1
(H
Si













C "O
•H O 01
• C -H C • .
CMO V -H

"l||













i-f
XI
5
14
(0
















c
to e
x3 » a
§ **

U (0
0 C 0

, u •o a.

U 0} O
C rH C1
7  -rt *" W
ox: o
•H 01 ii
D > v
o o ig
§•-" 3
o o. 5
XSS35
"S5 «


•o

0) *> 0>
eg E o o
X 01 0 -H
+» •-! •
M "o o
u ex


Ol TJ T3
C 01 A
•H 0) U O
i-IH V -H rH
X rH -ri >
O C W h
s 3 « 
-------
- 188 -





s
eg
$
Is
H O>
Q rH
3-<0
84
ITS VEHICLE
CN - AUGUST
H (3
EJI
5°
"o-
K o:
2P
a
§ S
f>
CD
s










01
rH
JD
a
u
5








0
> h
v S.*a «
o o c
*• •§ I s -g
0 -H C-
n +*
CL
. QJ
o ** o
C & O»'-^
m  C O 0)
m +j rt »H in
X § « H B
K3£~
01
O 01 01
C iH O>-»
^fnj o o a *^
X ^ *H +* n V(
in J3 0^
•dO +J
D > 0)
O >>
g5 §-
^s-sg-gc
n £ -H •— '
25 «
TJ
O T>
0) *J (0
x" §gS
+j M
M IH 14
O O
0 S.
Ol T3 T>
c o a
•rl • 0 0
iH rH +• *H ^H
X rt-rt >
* e n H
I 3 o> a>
a a a
<» *o
+J (Jj Q)
•H I-H m >
- tH O T) W
iH -H flj Q)
fli J5 O «
+* O iH £
5=" °
h
0
*»
a
M
*

o » m «
O rH M iH
rH rH rH rH

§0 O O
O* • W rH
rH rH rH
in in m o
•O 0 Ol ID
0 O O O
O .0 0 O

o o o o
CO 1^ rH O>
rH
rH m rH M
rH rH rH rH


o oj o^ o
rH rH rH

m o «^ PI
01 04 
01
c
rH M m t>
Q O Q B
g y 8 3



Operator

MCD1
HCD2
MCD3
Combined
- 189 -
TABLE E-4
SATELLITE VEHICLE OPERATOR ACTIVITY ANALYSIS
MEDFORD, OREGON - AUGUST 4-8, 1969
Total Percent of total time
observed Productive Unloading
385.6 82.1 11.5
357.6 83.3 •; > •' 1O.6
570.3 81.5 ' 12.O
1,313.5 . 82.2 11.4




Other
6.4
6.1
6.5
6.4

-------
                            r 190 -    .    .      •..'-,





              •'  '.'  .  '      11.7    ;-.  ..      '

     Total elapsed time =  ,   = 14.2 rain.'  '
                          . ti££       ,'    •                .;''••     •


     Actual average productive time  per  load, =  11.6 min   ...

     .„-.„_;<  a!-*;	«.-,.-- ^       .-       .-.  '


     Actual elapsed time per load =  11.6 min =  14.1 min   '...  .. •' ..

          '      •'"        '      '     .822  --    .'      ' '    "'••'.."  '



     The predicted productive time for the average load was within



one percent ox* the actual productive time .per load.  This negli-  •



gible difference demonstrates the accuracy of the  satellite vehicle



collection model for Medford/   .           •  -    -    .



     Packer Truck Driver Operational Analysis.   Regression analyses.



were made on the data recorded for the collection  activity of the



two packer truck drivers observed.   Based  on 236 data points the



best equation for describing productive  time foi the two packer  ,



truck deivers in Medford was:  (Table E-5)
             T f- O.69 +
                                    O.311
0.01*3
where T = productive collection time  required to service b • dwelling



          units between truck moves. •            -



     b  = number of dwelling units  serviced between truck moves



     b  = number of items collected from b  dwelling units



     b  = average distance from truck to. storage point, in feet





     This equation explained 83.3 percent of the variation in the ,



data and the standard deviation of  the residuals was 26.9 percent



of the response mean, productive time.  The F value and the
                                                                               - 191 -







I
n
w
a ;
h, O>
&« .
o>
8".

. • J oo
r-S4

2 D
'.- O O -
£* <
,'^T/
i'B'a-''
M
. ' :{t OS
";'">>-•

'i 2
B §

(H
B


^









q -o . ..
•HOC)
C -H -C' : .
O) Q> +*' -H '
ae o a a
» -H rH
- . o> w a . -
•:••"• >«










o
Or
tH
jO
a
•H
N
CB






•- ;









0 ™ W
X) 0 01


p



O O O 01
D> O +> Ol
10 C (0
en ri a X u
XI O <-> 0 O
> 0 3 *>
< -H M: 0)



0> T)
K B 01
01 01 V
(M XI *> O
JQ , E .-H 01
2 Vi rH
00
o

i'S c "
V O
•H S *
•H C f 01

rHrl O E
Z> 01 C T)
XI -H 01 Jt
E rH 0 0
3 rH -H 3
2 0) > k.
a rl *J
•6 o
0)




•-*"- .0
fi ;• '



rH O* fl


. 00- -00 .00
r-4 t* o*
i^ tn «
d d d
lit





§O rH
.' rH rH
O O
• " d d d




3: -O rH
rH en
odd







SOv rH
in >o
o o o





•o
01
• • c
rH OJ -H
o o ,a
,1 1 . I'- '








.
Q)
e


o


o
o
M
a
j-
+*
•o
Cr
(0

(D
M
8
r-l
. -C"
* 
-------
                 ...            - 192 -                       .






similarity of the coefficients for each variable in each regression




model suggest that the three variables chosen are highly signifi-




cant *in describing productive time for collection.  The most highly




correlated variable to productive time was 5C, the number of dwell-




ing units serviced between truck moves, in each of three regressions.




     The regression model was tested to establish its accuracy in




predicting productive collection time.  The average values for the




variables observed during the study were used in the example




(Table B-6).  The time spent on collection by the packer truck




drivers was 45.8 percent of the total elapsed time (Table E-7).





     b  = two dwelling units collected between truck moves




     b^ = two items




     b  = 6O f t walking distance




     Productive time = - O.69 - O.61(2) + O.31(2) + O.O1 (6O)




     Productive time = 1.8 min




     Actual productive time observed = 2.O min





     The predicted productive time required is 1O percent below




the actual time.





                       System Cost Analysis




     The daily costs associated with the satellite vehicle waste




collection crews operated in Medford were obtained from City
- 193 -





U)
3
n
2
H
I
•z
S
H O
u *o
a o
•-] rH
8 «T
I

M f/)
O Q O

CD X <
3 § '


.2
41
•rt
(0












at
.•H *-
+j (]) (
>«H 3 -H 'i
*O 4-* ^w
O
£


o> 01 o 01
D> O *> Ol
ffl C ffl "
iD 01 4-* U O ^



in -o

JO +J O
OJ E -rt O
J3 3 rH
2 *N IH
°8


C
0) Q|
•H S O
°§|I
r-IH O> E
XI 01 C -O
6 >H o 0
2 | > H
•D 01
0)

n

-------
                              -  194 -

i C
1 >
I K
1 "
> 14
= 2.
V

s

oS
•H

en " N t^ *o N o»
M >H H- N N . rl
>* to m co' o M
I-H 01 co co m co


o> m o *o co ^
Ov CO O> O 9> CO




PI f- ^f O ** CO
O> O CO Q *fl "J
co - in ^ ^ co . ^



o O o - •-*• «H co
in PJ rH rH 1^ CO





o m N "~* in *n
O O* O O1 CO O*
HH rH C^




^ r^ co m o
CO CO' CO CO CO .
. • . "°
' r C
rt OJ -H
o ' "e
         .                     - 195 -  .




Sanitation Service Inc. for the year 1968.  The costs were for a


three-nan crew consisting of two satellite vehicles with opera-


tors, and a.packer truck with a driver.


     Daily Crew Costs.  The costs associated with a satellite


vehicle waste collection crew are labor wages and fringe benefits,


equipment operation and depreciation, and overhead.  Equipment


operation costs.include repairs and'maintenance, in addition to


direct operating costs.  The costs given below are expressed in


dollars per day.




Labor                             '


     2 - satellite vehicle operators at $24.25 per day


         +15 percent fringe benefits                     = $55.78


     1 - packer truck driver at $25.25 per day


         + 15 percent fringe benefits                     =  29.04


                                            Total labor   = $84.82



Satellite vehicle operation and depreciation


  Operation


     2 - satellite vehicles each at $725 per year         = $ 5.6O


  Depreciation       •                            -      •


     2 - satellite vehicles, $2,30O new, 5 yr life..       =   3.52


                       Total operation and depreciation   = $ 9.12

-------
                              - 196 -






Packer truck operation and depreciation




  Operation




     1 - 2O cu yd Heil Collectomatic at $3,580 per year   = $13.76  •




  Depreciation




     1 - 2O cu yd Heil Collectomatic, $16,OOO new, 10




         yr life                       '                   =   6.15




                        Total operation and depreciation  = $19.91




  Overhead




     Overhead cost was assumed to be 2O percent of




     all other costs




     Overhead cost = O.2O (84.82 +9.12 + 19.91)          = $22.77




Total crew cost




     The total daily cost of the satellite vehicle




     crew described above                                 = $136.62





     Annual Collection Cost Per Duelling Unit.  The annual cost




to service the average dwelling unit observed in Medford twice




weekly was determined by multiplying the crew efficiency by the




crew cost.  The crews observed during the field investigation




averaged 5.0 hr per day in the actual process of collection.




Therefore the true crew cost was $136.62 per 5.O hr or approxi-




mately $27.50 per hr.  The efficiency of the crew is equal to




the total number of dwelling units it can service per hour.  The




satellite vehicle operators each serviced 1O dwelling units per




14.1 min for a total of 86 dwelling units per hour.  The packer
                                   - 197 -






  .  . drivers serviced two dwelling units per 2 min,  but only collected




     45.8 percent of each hour,  or 27 dwelling units per hour.   The '




     total crew efficiency was 113 dwelling units per hour.  The annual




     collection cost per dwelling unit for twice weekly collection is




     then:




       $27.5O/hr x 1 hr/113-d.u.  x'lO4 collections/d.u./yr = $25.50/yr





          Collection Cost Per Ton.  The residential  waste generation




     rate in Medford was determined by weighing the  trucks which were




     followed during the study (Table E-8).  The collection cost per




     ton was calculated by multiplying the crew cost per hour by the




     number of hours per ton. The amount of waste collected per hour




     by the average crew observed was:




113 d.u./hr x 3.1 Ib/person x 2.93 persons/d.u. x 3.5 days = 3.6OO Ib/hr





          The collection cost per ton is then:




           $27.5O/hr x 1 hr/3,6OO Ib x 2.OOO Ib/ton  = $15.5O/ton






          Satellite Vehicle Collection vs. Conventional Collection




          It is important to determine if the satellite vehicle waste




     collection system in Medford is more efficient  and economical than




     a conventional walking collector system.   An estimate of the effi-




     ciency and economy of a potential walking collector system can be




     calculated using the data obtained in the field study in a regres-




     sion model for walking collectors (Appendix A).  The regression




     model is:

-------
                              - 198 -
                            TABLE -E-8


                   RESIDENTIAL WASTE GENERATION
                MEDFORD,' OREGON - AUGUST 4-8, 1969
 Date-
Weight
 (Ib)
                           Dwelling units
                  Lb/capita/day
8/5

8/7

Total
18.O9O

15,310

33.4OO
  58O

  466

1,046
2.7

3.7

3.1
                                                                                                                      - 199 -
                                                                                                  C = O.18 - O.12W  + 0.12W  + O.24W  + O.OO5W-
                                                                                                                  1       .23         4
where C  = productive time, in minutes, required to service

       '    W  dwelling units

      W  = the number of dwelling units to be collected

      VI  = total number of items to be collected from'W

           dwelling units

      W  = total number of trips to the truck while servicing

           W  dwelling units

      W  = total distance, in feet, walked by collector while

           servicing W  dwelling units


     The productive time required to service 1O dwelling units

with the average characteristics used in the satellite vehicle

regression model calculations was calculated for walking collec-

tors.  A comparison of the two collection methods was made to

reveal the more efficient and .economic system.
                                                                                             W  = 1O dwelling units

                                                                                             W  =1.2 items/dwelling unit  x 1O dwelling units = 12 items

                                                                                             W  = 1 trip/3 items  x 12 items = 4 trips

                                                                                             W  "=.  J45 + 2.55 (average.street to storage)] W

                                                                                                =  £l5 + 2.55 (11O)] 1O = 3,250 ft
                                                                                             +Table E-3
                                                                                              Appendix A

-------
                              - 2OO -


     Substituting into the 'walking collector regression model:



      C = 0.18 - 0.12(10) +0.12(12) + 0.24(4) + O.OO5(3,25O)


                    = 17.6 productive minutes
                                                     t
                                    •    •           -3
     Since the average walking collector is productive 85 percent

of the time on the collection route, the total time required would

be:                    .

                         17.6
                         O.85
                              or 2O.7 minutes
     The same 1O dwelling units can be serviced by a satellite

vehicle in a total time of 14.1 min.  Therefore the satellite

vehicle operator is 46 percent more efficient than the walking

collector under these average conditions.

     The cost of a crew with two walking collectors and a packer

driver in Nedford is equal to the cost of a satellite vehicle

crew ninus the costs associated with the satellite vehicles.


                 136.62 = 1.2(9.12) = $125.68/day


     The Nedford crews spend 5 hr per day on the collection route.

The true crew cost is then approximately $25.OO per hr.  Assuming

that the packer truck driver would collect at the same rate as he

does when working on a satellite vehicle crew, the crew efficiency

would be:
                                                                                                                       - 201 -
,1O dwell
1      20
       ling units  6O min + .27 dwelling units. _ 85 dwelling units
       .7 min        hr  '   l        hr        '          hr
     The cost per dwelling unit is then:


$25.0O/hr x 1 hr/85 dwelling units x 1O4 collections/yr = $3O.SO/yr


   .  The collection service cost per dwelling unit for the


satellite vehicle operation was $25. SO per year.  The theoretical

savings to be attributed to the use of the satellite vehicle

system is $5.OO per dwelling unit per year.  Use of walking

collectors would theoretically increase collection costs 2O

percent .


                       Operational Comments


     City Sanitation Service, Inc. operates an efficient, sanitary

and safe collection service in and around the City of Medford.  The

collectors have established an excellent relationship with the

customers resulting in mutual benefaction.  The customers maintain

neat and convenient storage of wastes and the collectors reward

them with clean and efficient collection.

-------
                            - 2O2 -
     The, company! s concern with' safety was very impressive.  The  . •




satellite, vehicle operators were required to wear safety helmets




when operating on heavily traveled, streets and the vehicles were.




equipped with safety reflectors, extra turning signals'and rear ..




view mirrors.  The operators made a practice .of entering into




traffic very cautiously when pulling away from the packer.




     The satellite vehicle operators were very careful not to




overload the vehicle hoppers.  This practice prevented refuse




from falling from the vehicles and littering the collection area.




As an extra precaution, metal wing extensions were welded, onto the




sides of the vehicle hoppers to prevent refuse from blowing out




while enroute to the packer or while the vehicle was unloading. .




   .  The packer trucks were radio-equipped for relaying informa-




tion on extras and misses to the crews from the company office.




Customer requests and complaints are thus rectified in minimum




time.  ...'"•'




     All of the satellite vehicle operators and packer truck




drivers used aluminum tote barrels for transferring wastes; from




the storage point to thei r respective vehicles.  These cans have*




a rounded shoulder rest, a one side handle, and a capacity of




approximately .40 gal.  The collectors were very, satisfied with




this type of tote barrel.      ,               '          .     .
                                                                                                                    -  203 -
      The packer trucks used in conjunction with  the  satellite  -




 vehicles were Heil Collectomatic Mark Ill's with a capacity of    .




 2O yd .  The 1% cu j>d rear hoppers were not adequate :for  accom-




- modating the 1^ cu yd ;of wa's'te in the satellite  vehicle hoppers..




 .Therefore, two compaction'cycles of  12 sec each  were required to "(-




 empty each satellite vehicle load of waste.  The.satellite vehicles




 partially unloaded, lowered their hoppers, pulled away for the




 compaction cycle, backed up to the packer hopper and repeated the




 process of unloading.  .The* average unloading time was 1.8  min.  A




 packer with a 2 or 3 cu yd rear hopper could reduce  the unloading    :




 time to approximately 1 min per satellite vehicle load.




      The primary advantage which the collectors  in Medford enjoy




 is the low number of items per dwelling unit.  Almost all of the




 dwelling units served had only one item to be  collected because




 of the company service charge policy.  Also, 87  percent of the




 items collected during the field study were standard containers.




 If the items per dwelling unit were  increased  to the study aver-




 age of 2.6, the regression model predicts that a 37  percent decrease




 in collection efficiency would result.  This illustrates  the effec-




 tiveness of the company' policy which minimizes items by charging




 for extras.

-------
                                - -204 -
                              APPENDIX F

                         PASADENA, CALIFORNIA
                FIELD STUDY, JULY 28 - AUGUST 1, 1969
     The City of Pasadena is 'located in the center of Los Angeles

County, separated from the City of Los  Angeles by the San'Rafael

Hills*  These hills in the northwest section of Pasadena have

been developed into modern high income housing areas.

The major portion of the city has flat terrain with older low to

medium income housing.  The population of 125,OOO (Table F-l) has

a large percentage of people over age 65 resulting in a low growth

rate in recent years.

     Residential solid waste collection and disposal is the res-

ponsibility of the Engineer ing-Street Department of the City of=

Pasadena.  Prior to 1963 the City provided conventional collection

using five man crews with dolly carts and 7O gallon fiberglass con-

tainers in conjunction with 16 yd  packer trucks.  At that time the

City operated 16 routes for combustible refuse and four routes for

non-combustible refuse.  Heavy accumulation of refuse due to once

per week collection and increasing incidence of back injuries caused

the City to try using the Westcoaster, a three-wheeled satellite

collection vehicle of 1 1/3 cu yd capacity in June, 1963.  The success

of these satellite vehicles in reducing costs and increasing efficiency

resulted in subsequent purchases until the entire system had been

converted.  The Department reduced the number of routes from IS in

1963 to 13 in 1969 with a corresponding reduction in men from 125

to 8O.  The 16 yd  packers have now been replaced by specially built
                              - 2O5 -




                            TABLE F-l

               DESCRIPTION OF PASADENA, CALIFORNIA
Population

Dwelling units

Persons per dwelling unit

Land area

Population density

Dwelling density

Miles of roads

Dwelling units per street mile

     (estimated for hilly area)
125,000 (1/1/69)*

47.2OO (1/1/69)*

2.65

22.7 miles2*

5,51O persons/miles

2.O8O dwellings/mile2

33O miles*

14O/mile

5O/mile
      Fran 1969 Rand-McNally Commercial Atlas and Marketing Guide

      From Director of Sanitary Service, City of Pasadena

-------
                              - 206 -              .       '










SO yd  packers requiring only two trips to the Scholl Canyon




Landfill per day.  The one route in the hilly,.western sector




uses a 25 yd  packer for the required maneuverability.  The    .




City now has 4O scooters with 32 to 36 operating on 13 routes




daily.  The crews have a packer driver, two or three collectors




with satellite vehicles, plus one or two men with containers on




.dolly carts.                                           v
                       Field Study Analysis






     The  field  study was conducted from July 28 through 31.  A




crew of one packer driver, three satellite vehicle operators and




one walking man was accompanied .in the hilly, western sector of




Pasadena  for the first 2 days.  The last two days were spent in




the flatter central area with a crew, of one packer driver, two




satellite vehicle operators and two walking collectors with con-




tainers on dolly carts.  Only two satellite vehicle operators




and the packer  driver were observed in each crew during the




study.  Due to  the distinct characteristics of the two areas




and the two types of packer trucks used, the crews




were analyzed individually.   A total of 337 data points
                              -. 207 -
were obtained on six different satellite vehicle operators in




Pasadena.  The two packer drivers observed did not do enough  .




collection to develop a, productive collection time regression




model. ,'••••




     Satellite Vehicle Collection Operational Analysis.  Step-




wise regression analyses were conducted on the data for each




individual satellite vehicle operator and ultimately on the-




data from all the operators combined.  The regression model which




best described productive collection time for the six satellite




vehicle operators observed in Pasadena was:  (Table F-2)
    Yp = - 1.22 + 0.63^ + 0.20X2 + 0.007X3 + 0.107X4
where Y  = productive time required per satellite vehicle load




      X  = 'number of dwelling units serviced per load




      X  * = number of items collected per load




      X  = average distance satellite vehicle up driveway, in feet




      X  = average distance from satellite vehicle to storage,




          .in feet




      X_ = route distance of satellite vehicle, in miles





     This equation was able to explain 84.5 percent of the total




variation in the data and the standard deviation of the residuals




was 25.9 percent of the response mean, productive time.  As the F




value, 271.1 was very high, it can be said that the model, which

-------
                              -  2O8  -
H O
(J 
U CM




!s

s ,
8S

c

















0)
JO
id
>H

a

















^ v Jj
X O nj a
h -H rH
0) K p.
CL «J ST
> 0)
Sa
IP s a
9% w w
c *•
o
u
0) "O
U (0
o c o
5? £ 2^
x s» *
DC -ri 0)
•o S.
01
r) Q) o
C rH Ol
•» OJ O O O
X +> -H *• W
n £ O
•HO) *•
Q > n

0 >.
U O <9
C rH §
m a t> a o
SO) H
> •o
•O
0> 73
0) *» «


*> rH
O 0)
u a
Ol 73 T3
C Q) <0
•H 0) U O
rHrH V -H rH
X ** -H >
P C H H
3300)
D 01 &
5
U
<0
rH
H
6
0 rH
•O rH
00 O



O* CO
rH 0
1 1



tn *o
t^- tn
^ en





O en
rH i-H
d d



m -o
8 8
d d


CO O
CM CM

d d



3^
r5
O rH
*


rH fN
i i
-o
S



0
o
i



S;
m




m
S
d



i



CO
rH .

O



rH
in
d
*


rH
1!
o
in
00



rH
•O
O
1



5
CM
rH
*



1




S
o
d




t






i


CM

in
j;



rH
rH
t



m
CO
CO





3
d



8
d


•o
rH

O



m
^
d
*


en
i!
in
§



rH
rH











,




CO
O
o
d




i




rH
rH
rH
*


^
CL.
m
CO



CM
rH




O
in
*




rH
d



8
d


0


o



m
*o
d
•o
0)
c
3

•H
+>

U




O


o.


£
                              - 2O9 -





was developed using five inde.pendent variables, is significant.



The variable most highly correlated to productive time in this



model was the route distance of the satellite vehicle.  There



was a large difference in route distances between operators in



the west and the central sections (Table F-3).  Western area



operators traveled O.SO miles per load versus O.15 miles by the



central area operators.



     The utility and accuracy of the productive time regression



model was established by comparing it to the productive time



actually observed during the field study.  The flat and hilly



areas were discussed individually due to their distinct char-



acteristics.  To predict the time required to complete one



satellite vehicle load the average variable values were sub-



stituted into the regression model (Table F-3).  The fraction



productive time was used to convert the productive time to



elapsed time required (Table F-4).  The following values were



used in the model:
                                                                                                                                     Hilly
                                                                                         X  - dwelling units per load
                                                                                                                                   =   3



                                                                                                                                   =  IO



                                                                                                                                   = 12O ft



                                                                                                                                   =  IO ft



                                                                                         X  = route distance of satellite vehicle  = O.5O
                                     X  = items per load



                                     X  = distance vehicle up driveway



                                     X  = distance vehicle to storage
                                                                                         Percent productive tine
                                                                                                                                   = 82.3*
                                                           Flat




                                                             3



                                                            IO



                                                           1OO ft



                                                             O



                                                           O.15



                                                           8O.2*

-------







1
. M .
I
696T 'le-ee
Noiioimoo •
i > i
S'|l
K
"'•ll'
>£
«
.,1.













Variable










1. s- - ~
i.'S:*?!5
1-^- . ^' Or O B
O--H; ^*
0,
S +* o' ^
U) (0 O O flJ +»
Q m -f* o *— •
.a-" ^
-S;1?^
tl C o O


(P
o ai o
SrH ' Ol*--
BOB'*1
X *> -rl -H H tH
CO £ O —
•HO V
Q- > 0)
01 '. >.
, O 0) g ^^
X"!^ 3.>^
•rl O 'n ^^
-Q > -0 .
•O T)
'•«•«•
!,. - 0 +• O
« e O rH
X « »
W rH O
oi -o -o
C •' CD B -
i a o o
rtrt +* "H* rt
X rt -H > '
I § S s
Q 01 a
. Satellite
vehicle
loads
observed <
o
a
. ° '
- 210 - -.",.'."
CD- ' IT!' ' .' O- • • rt 00 H *f rt
^ o» - f* . . ^" « ^ .in ^
• ' • -

o o o o o o o o
o* en »-i co o» •-( o o
rt rt -rt.rt.'-rt
in o o o m in o in
^-O" if| (M-rtrtN rt
oo o oooo o
o o o o oooo
rt rt rt
8 3 88 8 S 88
JH rH rH rH rH rH
0 rH O O> CO rH rH O
rH rH iH rH rH I-H

:.- • . ..... : ' ,.--.' •
en " en en cncj tn en- en-
r*.' . m « rt rt o en, m
km cn«ortOi-< O
rt -. • rt «
2 s
•£• • . Z
rt . _ rt
_< CM or rtotcn^ a
II ° £ £ £ ? °
E£:H KKS.-K H












'


•. - '

-


.•'•''. - 211 -
. TABLE F-4
SATELLITE VEHICLE OPERATOR ACTIVITY ANALYSIS
PASADENA, CALIFORNIA - JULY. 28-31, 1969


' ^ Total Percent of total time
scooter . .....
collector observed Productive Unloading Other

PWWj^ ' 65O;O 8O.7 14.6 4.7
PW»2 624.4 . 83.7 13.1 3.2
Total hilly 1,274.5 , 82.3 : 13.7 " , 4.O
PWCj^ 411.9 81.2 15.9 2.8
PWC2. 47.1 66.5 18.3 15.2
PWC_ ' 512.7 80.8 13.7 5.5
1 • 3 "
PWC,. 86.2 81.8 14.O 4.2
• 4 .-.''.
Total flat 1,O57.9 8O.2 14.9 4.8


' '. '.
.



-------
                              - 212 -


     Hilly.  Substituting the values into the regression model

yields:

   Productive time = - 1.22 + O.63(3) + O.2O(1O) + O.OO7(12O)

               + O.1O7(1O) + S.04(O.5O) =7.1 min
                                                                             - 213 -
                 Elapsed time =
 7.1
.823
                                     =8.6 min
     The actual average productive time per load observed during

the study period was 7.9 min (Table F-3).  The regression model

estimate is 1O percent low.

     Flat.  Using the data for the central area the model

becomes:

       Productive time = - 1.22 + O.63(3) + O.2O(1O)

      + O.OO7(1OO) H- O.1O7(O) + 5.O4(O.15) = 4.1 min
               Elapsed time =
                                   = 5.1 min
     The productive time per load during the study averaged 4.1

min (Table F-3).  The regress ion. model estimate agrees with the

actual productive time observed perfectly.

     Packer Drivers.  The time spent on collection by the two packer

truck drivers observed was negligible (Table F-5).  Most of the

driver's time was spent assisting the satellite vehicle operators

and walking collectors when they returned to the truck to unload

(Table F-5).












CO  S
m HI

b. < 1
9 M 2l
8" SI
(4 DC O

•H
j?^
« *"
£1
P.

















§
•H


Q
O

O
^
rcen
CO
tt








k
01

O
Dl
C
'^
'H

01
c
'•>
a


o>
c
•H
u
o

8


o>
5
00
•H

S8L5
8 -H
OJ h
> -o
O
S
IH
>
•H
£
^ « (x 10

d o w


O> N CO rH
r* o <> c>»
CJ M iH C4


« o w o
co cr TJ -o




m «o m oo

O Tf 0) rH




O) H CO Ok
d CQ 1** *O
^ ^ m if)


m ^ en O

rH CO -0 -O



CO O' O •"*
N oj en (n
\ \ \ \

in o
C4 W)
fc £

-------
                •  '           •  - 214 -•"•..     .'       '    .. .  •   '


                       System'Cost Analysis

     The costs associated with satellite vehicle waste collection were
obtained from tbe Director of Sanitation of the City of Pasadena.
In order to allow meaningful comparison with other satellite .vehicle.
crews, the walking collectors were not included in the time and
oost = analyses.                                          .   -
     Daily Crew Costs.   The costs of a satellite vehicle waste
collection crew are labor, equipment operation and depreciation,
and overhead.  Equipment costs in Pasadena are determined by the City's
Transportation Division, which rents the collection vehicles to the
Engineering Street Department.  The rental charges include operation,
maintenance, depreciation and overhead.  The following costs are
expressed in dollars per day.
     Labor.
     2 - satellite vehicle operators each @$562/mo. + 44 percent
         for fringe benefits and overhead = $89.6O

     1 - sanitation truck driver @$606/mo. + 44 percent for          :
         fringe benefits and overhead     = $48.31
                              Total Labor  $137.91
     Satellite Vehicles.
     2 - satellite vehicles each @$21O/mo. = $19.O8
         (Rental includes operation, maintenance depreciation
          and overhead)
    '.-'.'-              '     - 215 -


      Packer Truck.  A 25 cu.yd Pak-Mor packer truck was used in
 the hilly collection.area and Gaskin-Built SO cu yd packer trucks
 were used in the rest of the city.                      ;,

;      25 cu yd packer @$68O rental per month   .       = ,$30.9O,.
      (Rental includes operation, maintenance,
       depreciation and overhead)                            -  .

      SO.cu yd packer @$85O rental per month      .    = $38.63
   .  .(Rental includes operation, maintenance,
      depreciation and overhead)

  "=" Total daily crew cost
  •        Crew working in. hills                       =$187.89
      '    Crew working in flat area                    *$195.62

-------
                              - 216 -

     Annual Collection Cost per Dwelling Unit.  The annual coll-

ection cost for the average dwelling units observed in the hilly

and flat sections of Pasadena can be determined by multiplying

the crew efficiency by crew cost rate.

     Hilly Area.  The crew observed in the hilly area spent an

average of 5.5 hr per day in the actual process of collection.

The true crew cost would be $187.89 per 5.5 hr or approximately

$34.OO per hr.  The efficiency of the crew is the total number

of dwelling units which it can service in an hour.  This crew

averaged six dwelling units per 9.5 min during the study for a

crew efficiency of 38 dwelling units per hour.  The annual coll-

ection cost per dwelling unit is then:

  $34.0O/hr x 1 hr/38 d.u. x 52 collections/d.u./yr = $46.OO/yr

     Flat Area.  In the flat area the crew observed spent 5.O hr

per day in the actual process of collection.  The true cost for

this crew is then $195.62 per 5.O hr or approximately $39.OO per

hr.  The crew efficiency was six dwelling units per 5.1 min or

71 dwelling units per hr.  The annual collection cost per dwell-

ing unit in this area is:

  $39.OO/hr x 1 hr/71 d.u. x 52 collections/d.u./yr = $28.SO


     Collection Cost per Ton.  The residential waste generation

rate for both sections of Pasadena was determined by weighing the

trucks followed during the study (Table F-6).  Prom the waste
                      - 217 -
                    TABLE F-6
RESIDENTIAL WASTE GENERATION,  PASADENA,  CALIFORNIA
                 JULY 28-31,  1969
Date
7/28
7/29
Total
7/3O
7/31
Total
Section
West
West
West
Central
Central
Central
Weight

-------
                        '.-.•'  . - 218 -






    generation rate and the crew cost and efficiency, the collection     :


    cost per ton for both areas of Pasadena may be established.


         Hilly Area. 'The amount of-waste per hour collected by the



    crew observed in this area -was:          .              •--.




38 d.u./hr x 2.65 persons/d.u. x 4.4 Ib/person/day x ,7 days ,= 3.1OO Ib/hr




         The collection cost per ton is:




          $34.OO/hr x 1 hr/3,lOO Ib x 2.OOO Ib/ton =  $22.OQ/ton




         Flat Area.  The amount of waste per hour collected by the



    crew observed in the flat area was:                        .    .




71 d.u./hr x 2.65 persons/d.u. x 3.O Ib/person/day x 7 days = 3,95O Ib/hr




         The collection cost per ton is:          .         .             • "•



          $39.
-------
    •                             r 22O -                   .




        Substituting into the walking collector regression model:



Yp = O.18 -:O.12{3)  +O.12(1O) + O.24(3) + O.OO5(1,38O) = 8.6 productive min




        Since the average walking collector is productive 85 percent


   of the total time on the route, the total time required to service


   three dwelling units would be   '  , or 1O.1 min.
                                 O • OD


        The satellite•vehicle operators .observed in Pasadena used


   9.5 min to service three dwelling units with identical character-


   istics.  Therefore, satellite vehicles servicing this area would


   theoretically be 6 percent more efficient.


        The cost of a crew with two walking collectors, a packer


   driver and a packer truck operating in West Pasadena is $168.81


   per day.  Since the crew spends 5.5 hr on the collection route


   each day, the true crew cost is approximately $30.50 per hr.  The


   two walking collectors service a total of six dwelling units each


   1O.1 nin.  The annual collection cost per dwelling unit is then:




$3O.5O/hr x , > . *?: * ""•" -..   x 1 hr/6O min x 52 collections/yr = $44. SO
            o owe.Ll.ing units



        The annual cost per dwelling unit for satellite vehicle



   collection in this area was $46.50.


        Theoretically, use of walking collectors in this area would


   reduce  the present cost of collection by 4 percent.
                                      T  221  -
            W  =  three  dwelling units



            V*2 =  3.3  items per dwelling unit  x 3 dwelling units  =  1O  items


            W  =  1 trip per dwelling unit



            1*^ =  45 + 2.55 (average distance street to storage) X


               =  45 + 2.55 (10O) 3 = 9OO




            Substituting into the walking collector regression model:




Y = O.18 - O.12(3) + O.12(1O) + O.24(3.O) + O.OO5(9OO) = 6.24 productive min




            Since the average walking collector is productive 85 percent


       of the total time on the route,  the total time required to service


       three dwelling units would be  '    or 7.3 min.
                                     U. oj


            The satellite vehicle operators in  Pasadena required an


       average  of 5.1 min to service three dwelling units with identical


       characteristics.   Therefore,  the  satellite vehicle collection


       method is  theoretically 43 percent  more  efficient  than walking


       collectors  in  this area.



            The cost  of  a crew consisting  of two walking  collectors, a



       packer truck driver and a  packer  truck,  operating  in the flat



       sections of Pasadena  is $176.54 per day.   Since the crews spend



       5.O hr on the  collection route per  day,  the  true cost  per hour is
                                                                                               +Table D-3

                                                                                                Appendix A

-------
                                - 222 -
                                                                                                                 - 223 -
  approximately $35. SO per hr.   The two .walking collectors  service



  a total of six dwelling units  each 7.3  ihin" and the  driver does '



  not  collect^   The annual collection cost  per dwelling .unit is- .



  {then:     ' •        •    •-       . •   .          '  •
$35.5O/hr x ,
       x ,..        - ....
         6 dwelling units
                          x 1  hr/6O min x 52 collections/yr  =  $37.50/yr
                                                         '
       The collection annual  cost per  dwelling unit  for satellite -  ••




.  vehicle service in. this area  is $28. SO per year.   Theoretically,




  use of satellite vehicles in  this area of Pasadena provides



  collection at a rate 24 percent below that for conventional walk-




  ing collection.




       The cost estimates in  Pasadena  for the two distinct  types of .




  collection areas illustrate the fact  that there is no set answer




  as to which type of system  is the most economical, until  all




  factors have been evaluated and used in the regression models.





                     General  Comments  and Conclusions





       Hilly Area.  No major  inefficiencies in the satellite vehicle




  collection operations were  observed  during the study of the five-




  man crew in the hilly, western section of Pasadena.  However, it




  may be possible to reduce the crew to four men, as the one walking




  collector was able to collect from only 91 dwelling units in  two




  days while the scooters collected an average of 168 dwelling  units




  a piece. .           .              .                 ...:•'
      The-amount of waste and number of items collected at each




 residence was very high due to amount of garden wastes collected.



 From the container Characterization study conducted during the



 field investigation, the amount of garden wastes was estimated at




 3O percent (Table F-7) .  The low number of items per satellite



 vehicle load is due to-the size of the standard container used




 in Pasadena*  Homeowners used 3O-gal cans instead of the customary




 2O gal container ..used in most communities*  Therefore satellite.



 vehicles were only able to carry 1O items per load instead of



 15,  the average for other cities studied.




     The operators were particularly careful not  to overload




 their  vehicles.  Speeds were kept  at a minimum when the vehicles



 were full  and blowing paper and other refuse was  always  retrieved




 by the operator when dropped from  a load.  The satellite vehicles



 and  packer truck had to contend with very heavy traffic in the




 mornings,-but  managed to maneuver  safely and deliberately.  The




 packer truck had a train bell as a  safety  feature which worked



 automatically when backing-up.'




     The satellite vehicles, required an average of 1.3 min to



unload  into a'25 yd  Pak-Mbr packer with a 3 yd hopper.  The




 large hopper allowed the satellite vehicles to dump an entire



load with only one compaction cycle. -

-------
                              - 224 - .






     Flat Area.  The.major difference between this five-man crew.




and the crew observed in the hilly section was in the care exer-




cised during collection.  These operators overloaded their vehicles,




neglected some items at certain services and failed to retrieve




wastes which fell, from the vehicles.  In addition, the vehicles




were driven too fast when loaded, resulting in heavy littering of




yards and roads.  The crew was not as well coordinated as the




crew in the hilly area and operators duplicated dwelling units




frequently.




    : The number of items at each dwelling unit was similar to




the western area but the amount of waste at each residence was




less.  Garden wastes constituted 3O percent of the wastes coll-




ected (Table F-7).




     The average unloading time for the satellite vehicles was




O.8 in this crew.  This short unloading time was accomplished by




use of a front overhead loading Gaskin-built 50 yd  packer.  The




satellite vehicles dump into an unobstructed large bin at the




front of the packer which was more accessible than a packer with




a rear hopper.  The bin was then hoisted over the cab and emptied




into the packer after the satellite vehicle has left the packer




truck.
i












• • . 5


•

8 **

1
§
O.































(0
|
•H


•O 1
•3 c
c
CO
v>
rH 0)

•H 8
o *>
H -H
O


«
- 2

CO 
-------
                          •    - 226 -  .   •       .   '       -..'




                       • :    APPENDIX G  .                     '




                    WAUKESHA COUNTY, WISCONSIN   ._         .




                   FIELD STUDY, JULY 8-11, 1969






     Wauke-sha County.is located in the flat southeastern section  .




of Wisconsin directly west of Milwaukee County.  The County has




benefited from westward expansion of the City of Milwaukee, pro-




ducing one of the highest county population growth rates in the




United States in the 196O's.  The population has increased from




158,249 in I960 to an estimated 212,OOO in 1969.  This. suburb of




Milwaukee is characterized by upper middle class homes in large




housing developments built within the last ten years to meet the  .




needs of coamuters not wishing to live. in the city.  These homes




are located on small lots and are close to the street.




     Sanitary Disposal Service, Inc. provides once weekly waste-




collection service to approximately SO to 6O percent of the popula-




tion of Waukesha County on a private or contract basis.  The company




began using Cushman scooters in October 1966 in  an attempt to




increase collection efficiency.  Sanitary Disposal Service currently




operates 22 Cushman scooters in two man-crews-with two scooters per




crew.  The scooters are transported to the routes in pairs on




trailers hauled behind the packer trucks (Figuxe G-l).  The packer




driver operates a scooter in addition to the packer truck during




collection by attaching his satellite vehicle to the packer truck




when moving it.
                           -  227 -
Figure 0-1. • Satellite vehicles with hauling trailer.

-------
                              - 228 -



     The use of plastic bags for waste.storage has been recently


initiated by Sanitary Disposal to further reduce collection time.


The cost of these bags to the customer is $.O5 per bag and approxi-


mately 15 percent of the customers were using plastic bags at the


time of the study.



                    Field Study Analysis


     The field investigation was conducted July 8 through 11 in


the City of Brookf ield and the Village of Elm Grove in Waukesha


County.  One crew was observed for the four day period.  The area


in which the crew operated was very flat with modern upper-middle


class homes located on small lots with short driveways.  The hous-


ing density of this area was about 4O nouses per mile (Table G-l),


but since Sanitary Disposal only has about 75 percent of the


residences as customers, the effective density was reduced to 3O


houses per mile.  The company requires that the storage area be


easily accessible to the scooters in order to reduce walking


distance to an average of 1O ft at each dwelling unit.


     Stepwise regression analyses were made on the data for both


collectors individually and combined.  The average variable values


and the mathematical models for both drivers were very similar


(Tables G-2, G-3).  From this and the percent of the variation


covered it can be concluded that the variables recorded are very


significant and  therefore reliable in predicting productive time


per satellite vehicle load.  The regression model which best des-


cribed productive collection time for  the 173 satellite vehicle


loads observed in Waukesha County was:
                              - 229 -
                            TABLE G-l


                       COMBINED STATISTICS
           CITY OF BROOKFIELD AND VILLAGE OF ELM GROVB
                    WAUKESHA COUNTY,  WISCONSIN
Population
             *

Dwelling unit


Persons per dwelling unit

         *
Land area


Population density


Housing density


Miles of street


Houses per street mile
39,800 (est. 1969


9.5OO (est. 1969)


4.18


29 sq miles

                      2
1,370 persons per mile

                  2
33O homes per mile


233 miles


4O
     •Fro* Village Manager of Elm Grove and City of Brookf ield
      Department of Public Works.

-------
 Yp = - O;89 + O.54
       - 230 -






+ 0.15 X0 H- 0.011
                                            + O.O47
                                                         2.93
where Y  = Productive collection time per satellite vehicle




           loadt in minutes




      X  = Number of dwelling units serviced per load




    .  X  = Number of items collected per load




      X  = Average distance satellite vehicle travels up"     »  - •-




           driveway of each dwelling unit, in feet




      X  = Average distance from satellite vehicle to




           storage at each dwelling unit, in feet




      X_ = Route distance of satellite vehicle per load,




           in miles






     This equation was able to explain 83.0 percent of the. total




variation in the data and the standard deviation of the residuals




was 2O.O percent of the response mean, productive time.  The F




value of 163.5 indicates that these five variables are significant




in explaining productive time.  The route distance per load of the




satellite vehicle was the most highly correlated variable to pro-




ductive time in this equation.  This means that the variations in




productive collection time per satellite collection vehicle .load




were best accounted for by the corresponding variation in route




distance traveled by the vehicle.




     To illustrate the utility and accuracy of the regression




model developed, it was compared to actual field observation




values.  Using the average variable values, the productive time
                                                                                                                        - 231 -





m
" . . . EH-
' ' 2 CT\
WVD
O rH
M .
W rH
O 1
O CO

W ij
Q tD
0 ^
£
CM 2
1 O 2
C3 MM
EH CO
. -W . 02
J WO
CQ JO
, < j ra ;
; . EH o M •-. •

W « .
CJ EH
M 2.
W O
> 0
W ^1
' J W '
£§
EH <
< S





























|
m
5





















ip.SI.1 •
¥ U V -H

V -H fH
ft* b Qi
(0 X
> 0)
• c,
|Q E
0 +3 £
X w £
g^
CJ
O T)
O (If
-> 5 §S
X 3V
5(0 h
•H 01
•0 0.

0)
U O 0)
J § •i~l 1
si "I

8» S

xro B 2 §• s
w j: -
-HO L
Q > -3

O T3
 rr*
M.n

bO -0*0
C o m
•H O O O
X 'H "^ V ^
2* § V «
0) O



&

2
g.
~ in '
0 O O
*



•o
rH' CM C
Q O TH
0:0 a
336
0
/- 0
'













V
•H
0>
2
•o
0

0.
o

•o
id
rH
»
rH.
to
-H
x:
^
03
0
E
(U
rH
01
•H

0)
*

1

-------
- 232 --•
                                                                                        - 233 -



















ro
1
O
Ed
J
cB

EH































«>
M
O -
H rH
EH rH
0 1
64 
J X
0 J
O CD •

3 •
o z
M M
5G CO
M Z
> 0
O
td K>
EH M
M 3
J -
Id :*t
EH EH

5i .
PC, O
O
M X
[d CO
3 W
J X
> «<

Id
O •>
< CO
a: w
ta J
> m
M
g



























V

(d
-H
tj
rd
>





























? f.
•H 0>

>, 3 fl) O -H

H?
O -H (1)
co> bo
W dJ 0) O "0 *~»
(n -4J £4 +J t( 4-1
Q W +•> O 4-t
O 01


U Oj '~»
O C O 0)
X D +-» -H
o to f* e
a; -H o *-'

Q>
O « O ^^
X 4-> -H 4J fc 4-4

a > n
>,
o a>  »*H
W -C -+ '-'
•H V h
Q > -0


•o -o
V <0
m H b
~ 8 *

HO T3 T)
C O <0
-^ « O O
X .-1 -H >

-H --<  Q)  .a
to o


>
-H




o -=r t*-

GO t— t—



0 0.0
i- t- r-



o o o
YD VO ^O
O O 0


0 O O .





O O 0
r— t— t—




f\J C\J CM
CM CM CM






ir\ in in


C\J rH CO
O f- t—
fH rH



•a
0)
^ CM C
Q Q -rl
0.0 U3
s s e
o
o
                                                          required to make one trip with the satellite vehicle was* predicted




                                                          (Table G-3).




                                                               Unloading and other time were accounted for by dividing the




                                                          productive time by the fraction of productive time (Table G-4) to




                                                          yield total elapsed time required.





                                                                                                                =  5




                                                               X  = Items per load                              = 22
X  = Dwelling units per load
                                                               X  = Average distance vehicle goes up driveway   = 70 ft




                                                               X  = Average distance from vehicle to storage    = 1O ft




                                                               X^ = Route distance of vehicle per load          = O.6O miles




                                                               Percent productive time substituting into




                                                               the model;                                       = 77.5 percent






                                                               Productive time = - O.89 + O.54(5) + 0.15(22) + O.O11(7O)




                                                                                 + O.O47(1O) + 2.93(O.6O)






                                                                       Productive time = 8.1-min per load
                                                        Total elapsed time = Productive time = _8^1 = 1Q 5
                                                               The predicted productive time of 8.1 min per load is 5 percent




                                                          above the average time observed,  7.7 min per load,  during the field




                                                          study (Table G-2).  Therefore,  the regression model is very accurate




                                                          in describing productive time required per satellite vehicle load.

-------
                              - 234 -
                            TABLE Gr4

     SATELLITE COLLECTION VEHICLE OPERATOR ACTIVITY ANALYSIS
            WAUKESHA COUNTY, WISCONSIN, JULY 8-11, 1969
Operator
             Total
            minutes
                    Percent: of total tin
           observed   Productive   Unloading
                                                         Other
WCD,
WCD,,
1,117.4

  7O6.5
Combined   1,823.9
8O.7.

72.7

77.5
17.6       O.O

15.2       5.9

16.6       2.2
1.7

6.2

3.7
                   ...'..•     -  235 -


                        System Cost  Analysis


     The costs for  the  average satellite  vehicle collection crew

in Waukesha County; Wisconsin were  obtained from the owner  of

Sanitary Disposal Service,  Inc.  Two man  crews were  used in

Waukesha County,.but were equivalent to three man• crews in  other

areas.  The cost calculations and comparisons are based on  the

assumption of crew  equivalency.                    '      •       \.


     Daily Crew-Costs.  The daily cost of a satellite vehicle

collection crew consists of labor,  satellite vehicle operation

and depreciation, packer truck operation  and depreciation,  and

overhead.  The costs are given in dollars per day.

Labor         .    .

     2 - satellite vehicle operators at $18O/wk      =  $ 72.OO

Satellite vehicle operation and depreciation

     Operation   .                ..

       2 - satellite vehicles  each at $l,OOO/yr       =     7.7O

     Depreciation

       2 .- satellite vehicles, $3,OOO new>  2  yr  life  =  . 11.54

                    Total operation and depreciation  =  $19.24


Packer truck operation and depreciation

     Operation

       1 - 25 cu yd Heil at $2,5OO/yr                 =  $ 9.61

    • Depreciation

       1 - 25 cu  yd Heil,  $15,OOO new,  4 yr life      = ' «• 14.4O •

                    Total  operation and depreciation  =  $24.Ol

-------
                       •       - 236 -






Overhead        .              .



    . Overhead ccs t was unable to be obtained from Sanitary




     Disposal Service, Inc. and therefore was estimated at




     2O percent of all other costs.






           Overhead cost = 0.20(72.OO + 19.24 + 24.Ol) = $ 23.OS




Total crew cost




     The total daily crew cost is
$138.30
     Annual Collection Cost per Dwelling  Unit.  The annual coll-




ection cost for  the average dwelling unit observed in Waukesha




County for once weekly collection was determined by multiplying .•




the crew efficiency by the crew cost rate.  The crew observed




during the field investigation worked approximately 7.O hr per




day in the actual process of  collection.   The  true crew cost




rate  is then  $138.3O/day x  1  day/7.0 hr or approximately $20.OO




per collection hour.  The crew efficiency was  1O  dwelling units




per 9.9 min or 61 dwelling  units per hr.   The  annual collection




cost  per  dwelling unit  for  once weekly collection is then:






   $2O.OO/hr x 1  hr/61 d.u.  x 52  collections/d.u./yr =  $17.OO






      Collection  Cost  per Ton.  During the field investigation,




 five truck loads of waste  were weighed to determine the residen-




 tial solid waste generation rate in Waukesha County (Table G-S).




 The collection cost per ton was  calculated using this generation




 rate.  The amount of waste collected per hour by the crew would




 be:
          . •    '  .   '       '      .--.-237 - •






61 d.u./hr x  4.18 persons/d.u.  x 2.6 Ib/capita/day x 7 days = 4.6OO Ib/hr






    Multiplying by the crew cost per hour produces the collection cost




    per ton.




           $20.OO/hr x 1 hr/4,6OO Ib x 2.OOO Ib/ton = $8.SO/ton






         Satellite Vehicle Collection vs. Conventional Collection.




    Productive time requirements and costs for conventional walking




    collection were estimated and compared to satellite vehicle coll-




    ection using a regression model similar to the one developed for




    satellite vehicles.   The regression model for walking collection




    is:





           C  = O.18 - O.lZWj + O. 12»2 + O.24W3 + O.OO5W4





    where C  = Productive tine in minutes to service W  dwelling




               units




          W  = The number of dwelling units to be collected




          W  = Total number of items to be collected from W




               dwelling units




          W  = Total number of 'trips to truck while servicing




               W  dwelling units




          W  = Total distance walked by collector while servic-




               ing Wj dwelling units, in feet






         The  average housing characteristics used in the satellite




    vehicle calculations were used to calculate the productive time




    required  for one walking collector to service five dwelling units.

-------
              - 238 -
                                                                                                   - 239. -
            TABLE 10-5
RESIDENTIAL SOLID WASTE GENERATION
    WAUKESHA COUNTY, WISCONSIN
          JULY 8-1O, 1969
Date
7/8

7/9

7/1O
Total
Weight
(lb)
• 12,35O
1O.35O
14,250
6,100
lO.'SSO
53,600
•Dwelling units
serviced
154
164
161
88
143
71O
Ib-'capita/day
2.7
2.2
3.O
2.4
2.5
2.6
The results were then compared with satellite vehicle collection

to determine the most efficient method for Waukesha County.  The

average values for Waukesha County were: .                .'

     W  = five dwelling units                         .

     W  .= 4.4 items/dwelling unit  x 5 dwelling units = 22 items

     W  = 1 trip/3 items  x 22 items = 7.33 trips

     W-  = /45 +2.55 (average street to storage distance)! W *

        =  J45 + 2.55 (70)J  5 = 1,120 ft


Substituting these values into the regression model,
                                                                        Yp = 0.18 - 0.12(5').+ 0.12(22) '+ 0.24(7.33) * O.OO5(1,12O)

                                                                           = 9.6 lain of productive time           ,
                                                                         Since the average walking collector who also drives the crew

                                                                    truck is productive 8O percent of the time while on the collection

                                                                    route (Appendix A), the total time required to service five dwell-
                                                                    ing units becomes
                                                                                       9.6
                       or 12.0 min.  The satellite vehicle opera-
                                                                                      0.80
                                                                    tor serviced an equivalent number of dwelling units in a total

                                                                    tine of 9.9 min.  Therefore, satellite .vehicle waste collection

                                                                    in Waukesha County is theoretically 21 percent more efficient

                                                                    than the alternate method of waste collection by walking collec-

                                                                    tors.

                                                                         The cost of a walking collection crew in Waukesha County

                                                                    would be $115.2O per day.  Since the crew spends approximately

                                                                    7 hr per day on the collection route the true CCB t per hour  is
                                                                         ^Appendix A
                                                                          Table G

-------
                    .-'•.-- 2 - .





approximately $16.5O;  The two man  crew can collect a .total of




1O dwelling units per  12 min or an  equivalent of  SO dwelling




units per hr.  The annual collection cost  per dwelling unit is




then:               .





. $16.SO/hr x 1 hr/5O dwelling units x 52 collections/yr  =  $17.OO






     The annual cost per dwelling unit for satellite vehicle




collection was also $17.OO.  Thus the two  methods are theoreti-




cally equal on the basis of economics alone.






                       Operational  Comments





     The unique concept used by Sanitary Disposal Service,  Inc.




of having the packer driver also operate a satellite vehicle  is




very efficient.   Driving the packer truck  occupied a maximum  of




only 6 percent of one  man's time,   this leaves  94 percent  of  the




operator's time to be  available for productive  work instead of




waiting at Hie truck.   The time required  to attach or detach




the satellite vehicle  from the packer was  only  O.3 min.  The




only disadvantage to leaving the packer truck unmanned and run-




ning is the danger of  children innocently  or intentionally




taupe ring with it.




     The operators in  the area observed were benefited  in  their




operation by the  accessability to  the waste  storage point. Very




short driveways averaging 7O ft  in  length  with  little or no walk-




ing distance from the  vehicle  to  the storage point increased  crew




efficiency significantly.
                              -  241 -                     '




     Customers were very thoughtful in their choice of a storage




point and the collectors were seldom required to ring doorbells




to ask people to unlock garages, the most common storage area.




     The average number of items collected per dwelling unit,




4.6, was extremely high.  This value was nearly twice the average




value for the other five areas studied, due to garden wastes which




constituted approximately SO percent of the total wastes.  The




amount of waste collected in the winter would thus be reduced




significantly.  Approximately 4O percent of the items were not




in standard containers. . Most of the wastes collected were con-




tainerized in paper and plastic bags, increasing transportability




by the satellite vehicles.  Approximately 15 percent of the items




collected were plastic bags.   The operators carried a supply of




plastic bags in their vehicles and delivered them to customers




upon request.




     The satellite vehicle operators observed used an excessive




amount of time unloading the wastes into the packer truck.  The




high average unloading time of 1*7 min was due ID several reasons*




The hopper on the 25 cu yd Heil Collectooatic Nark III has only a




1.5 cu yd capacity requiring 2 to 3 compaction cycles to accommo-




date the waste from the satellite vehicles.  This problem was




enhanced by the consistent overloading of the satellite vehicles




by their operators (Figure G-2).  The 1-1/4 cu yd satellite




vehicle hoppers were usually heaped up to about 2 cu yd before




returning to the packer truck.  The operators averaged 22 items

-------
per load compared to an average of 15 items per load for the six




study areas.  Operators had to unload the satellite vehicles care-




fully and slowly to avoid spilling wastes onto the pavement from




the overloaded vehicles.    •  .   • •              •




  •'   Reducing the number of items per load and use of a packer     ~~ •




with a 3; cu yd -hopper. coU'ld decrease unloading time significantly.




In addition, the installation of a rubber flap on the back 'of the




satellite vehicle hopper could reduce the amount of waste spillage




and attendant cleaning up time.




     The Company reported very little trouble due to winter snow-




fall.  No adverse effects on normal operations were experienced




until there was more thin 2 in. of snow on the ground.
                                                                                                   Tigure G-2.   Overloaded satellite vehicle

-------