-------
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
Ǥ
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
-------