-------
Si
c "55
1 |
c |c »
8-B 12 i
£|o°
sml
£ |
o> £
(O
in
N
55
>r 0
® 0
n °-
O o
o
10
T- °
§"8"
° IO
1-8
Boo
co o o
O ^? O
CIO T-
0
Is I
n - O
"• o m
o
'E
0>
to
J1 0
T- O
0 0
in <*> o
£""-
• 0
T- 0
O CO
•» co
1 o
T- O
0 U>
o n
^
O
wnership Type
O
°> 0)
^° CO
CM _!_
O 10
OO 01
*~_ °o
r — CM
~
CO -^
O> co
--_ co
•v~
CM o
O> S
co" CM"
-^~
0 (^
CD o
CD" _i_
CD o
CO CM
^— 1
~
CO CO
^
Ogj
^_ ^
-(.
0) T-
T"
ublic Systems
Mean Residential Connections
Confidence Interval
o.
CM CJ) CD
•* CM 'f
~
CM C> CM
CO CD ij
•* T. °o
CD r— co"
^ ~
O CO CO
§ ?s
o co" i
CO +"
co in TJ.
•~
T— 1
+
° So °>
0 CD CM
•^-
0 CMco
+
2
to .Q
Median Residential Connection:
Mean Non-Residential Connect
Confidence Interval
CM 0
*- CD
CO
t— r-
CM in
CJ>"
^- in
CO T-
CM"
O CO
^" CJ>
CO
T-"
0 0
O CO
O IO
O T
m ^J•
CM -q-
O CO
CO
c
0
'•5
Median Non-Residential Conne
Observations
CM ^
^~
~
sis
CM" v-~
CM T}.
CM ,
~
CM o
O> f^
^ O)
5
CM"
OO
CD
o>
CO
o>
^
If)
T—
g
CO
§
CM
M
Median Residential Connection:
O T-
co ».
(
+~
CD CO
O> CO
f*~" •«-"
•£-
0$?
CM CO
•* -L
CM 0)
m 10
,— 1
co ?
1
~+~
CO CO
in oo
~+~
O> CO
"- CM
•^-
-7
+
10
c:
o
Mean Non-Residential Connect
Confidence Interval
O CD
Tfr
CO
O CD
CM
Y-^
CM
CO 0>
in T-
co"
T- CD
•
O CO
CO
m
c.
o
t5
Median Non-Residential Conne
Observations
tO (*}
CNj CM
^^
•<~ 10
^" ^J"
CD" o"
^~ ,
+
0 CO
S«"
•* o
TT §5
Is-" i-"
-j—
m (o
O CO
CM 10
CD" i
+
O CM
CO CM
T— 1
T"
CM o
~
0 "*
*~ ,
^.
a«
+
II Systems
Mean Residential Connections
Confidence Interval
<
co r- o
•
co m K
CM" co" i
CO +~
o co r\
CM O CO
CD CD CM
h-" V-" _i_
T~ +
CM CM S
»- CD -r-.
in" i
+
CO •»• 0
CD CD co
CD CM W
T— 1
+
CO CO
CO " ^
•£•
& -^
+
m
c
m o
Median Residential Connection:
Mean Non-Residential Connect
Confidence Interval
0 g
CM
"
o> in
m
co"
co ••$•
00 O)
r- T-
CM"
T- 0)
o o
T—
CD CM
co m
CO T-
0 Is-
O CD
CM m
T- O
CM
O CD
CD
0 8
^
CM
d
in
c.
o
t>
Median Non-Residential Conne
Observations
ata:
o
0>
CD
-------
§
i
K
c
0
'•ro
3
a
f
a>
0
•1
II
1?
^|
1
|
o
O)
O
c
o
1
a.
0
Q_
0)
0
'5
0)
(0
E
(0
co
g
N
in
^
**
*- o
O o
o
IO
•r- °
|i
,:§
0 0
SI
1 o
58
0 U>
^
1 o
T- 0
0 O
« 0
<"> *~
• 0
T- 0
0 CO^
1 o
55
3 10
o
wnership Type
o
o>
O
CM"
^~
CO
o
CM"
oo
to
o>
CM
CM
s
o
T—
i--
10
CM
CM
r-
co
0
5
o
10
CM
CO
I
£
CO
V)
ublic Systems
Mean Populatiol
CL
<° 0?
t oo
CM"
i
T"
00 CM
CM «>
O"
oo
T"
CO CO
OO CO
o"
CM
§5
05 '-
erf
+
«"> S
Oi CO
co r~-
~
00 CD
~
o> »
17
— .
c?
CD
Confidence In
Observations
rivate Systems
a.
IO "°
^ CM
T-~ 1
+
CM CM
S§
O" W
to o
~
Sc^
O to
CM"
CD CM
CM •«-
CO 10
eou-f
+
^ "O
CM ,
~+~
r- T-
CD CO
m" i
CD rj-
~
CM T}-
~
S^
£ "—
W o5
c. •*-•
Mean Populatiol
Confidence In
S
CO
co
0)
CO
CO
CM
CM
CM
CM
CO
IO
o>
CO
CO
Observations
0
O
i--"
§
of
o
IO
CO
CO
IO
h-"
o
CM
IO
CO
£
CM
O
CO
CM
0>
CO
10"
CO
O>
C0_
CM
IO
1
•f
§5
co »-
,
f
CM O
oo »-
~+~
O ^
+~
<0 T"
CM ^
^
J«
—
1
CD
Confidence In
Observations
g
CM
a
s
as
O
-------
.0
a>
0
O
o>
o ^_
S 2
« 1
fef '
JtA
0 0
"rt W £ W
m 3 * -0
.** t * f\ c
h- u w =
w ^-.
1 I
0
~
S^
C
c
**
s
N
w
^
14
s§
o5
u>
• 0
o ^
o §
o g
£,
is§
« 0^ o
O o o
c w *~
o
"TO
^ * C5
o. ^~ ^
Q O O
n ° cT
»° •»
CD ,-
O
S
CD ' o
V) T- 0
c 0 0
C CO ?•»
in
>*
1 0
T- 0
o CO
• o
T- O
0 U>
O 0)
-1
0
'wnership Type
o
CO rj-
CO -^.
T"
g^
+
^
T—
~T~
CO c\]
8CNJ
^J
T—
o co
ss
•T~
CO ^
__
*"~ 1
O) to
^- co
T"
1
ublic Systems
Mean Residential Deliv
Confidence Interval
Q.
CO h- «-
I-- CO •«.
CO CM
*
"*"
O> CO o
^ CO 2
'— _i_
+
00 CM (^
T- T-_ CM
T- i
+
r~- to (o
o en 05
T— O. CO
1 — ^" ^-
O> 10
^
t O rj.
00 CO o
CO co
T"
CD O en
r-- co o
IO 10
1
+
O |v. co
l»- O> CM
S Sco-
CM CO
r
+
liveries
Deliveries
Median Residential Del
Mean Non-Residential
Confidence Interval
80
CO
CM CO
n §
o
*—
CM O
OO CO
CO '-
o en
CO
CM CO
en o
Tj- T—
IO IO
CO ^
IO h-
CM CM
T~ T—
0 0
10
O T-
il Deliveries
Median Non-Residentis
Observations
St
T"
^ CM
T"
T~ CO
T—
"V~
CM CM
O co
O CM
+
0 o>
O CM
~
O> CM
•T-
Oc^
+"
!J £
•^~
0)
rivate Systems
Mean Residential Deliv
Confidence Interval
Q.
O CO -nj.
CO Y-
1
"1"
5 °§
T- 10 ii
1
+
^ IO o
OO '-CM
CO. «
^ i
+
CM CO N.
^ £ CO
i
00 00 Co
oo "'fr o
CO (O
^>
co en r^
00 J- 0)
Is- CO
T"
CM CD i*
00 10 |CI
CM v-
+
CM CM
r-- 10 o
CD CO ^f
T"
iveries
Deliveries
Median Residential Del
Mean Non-Residential
Confidence Interval
CO O
CO CD
CM
S ">
IO
10 CD
5) ^~
CO IO
to r~-
CO '-
to
T- CO
CO ^~
O CO
CM CM
T —
O T-
co
O IO
co
il Deliveries
Median Non-Residentia
Observations
CO co
"*— CM
~
SZ
+~
T~ XJ-
T— (
"^~
__
IO ^~-
^3 v--
^~
h*- CO
CM?
~
CO o
co i;
•T-
O> co
^ CO
T"
CD to
•x~
CO
•8
ID
II Systems
Mean Residential Deliv
Confidence Interval
<
IO CO CM
r- CD Si
10 ,-
1
"*"
S Sf;
T- CM. -^-
'— JL
+
h^ O> (o
'- IO K.
T- '-.CM
T- 1
+
T- 0 co
+
CO C3 CT»
00 ^ o^
O •$
"^"
oo -jr co
CO CM
•^~
CD CO Q
h- CO •».
IO xj.
1
+
^ s g
IO T- T-.
co co co
•f-
m
ID
1 1
(D d)
Median Residential Del
Mean Non-Residential 1
Confidence Interval
O O
CO CM
T- en
r-~ to
o
'—
IO CO
en •»»
CD •<-
IO t
co en
CO
10 m
en CM
T CO
CO CO
cO T—
o o
CM IO
T— T—
O T-
CO
O CD
il Deliveries
Median Non-Residentia
Observations
CD
CD
s
Q) Q.
8 S
.> Q.CM
15 o .g "5
^^ i CD r?
C O > In"
O O '~ If)
:= O 0) t—
o" i_r
*- « .£
CO " '•«-• m
CM Q) c 5
^- O. ® >
^3 t_ C
£ o o =5;
co o en
2> *- -o c
o>^- ID e>
o "2 =
!2 « -g E
CO CO ^
rt •— — *~
t . "D
••-• O) -JB m
•> 0) CO
ilff
•2 ° * 5
CO '^ u> *•-
0 >,_CO £ 0
S " Q-E =
°- 1 g.2^
T- D. O> TJ O>
0 g| g.§
<3 t/> IB -o E
a 1
CO O
Q Z
-------
to
5
o
£
2
•£
o
o
.0
1
la
38f
O L. 2
«° *
-c m
1$
10
1
co
it—
o
§>
fl
c
d>
o
£
IL
O
D)
B
ro
c
o
"5
a
0
a.
a>
o
'E
stem Sei
co
to
0)
N
to
»r 0
15
0
10
^ o
o o
s|
1 0
•«- 0
§0,
. 0
o o
U) T-
1 0
T- O
0 0
" S
*-
1 0
T- O
0 0
n, 0
ro *-
1 0
•t- 0
Seo
w
1 0
5g
o at
? S
_l
O
u
Ownership Ty
••- o
5*
O 10
CM -q:
o> ,
~
T— *-
Si0?
+
CO 0
fe°?
»- ••£
S co c
Public Systen
Percentage of
Confidence 1
in
CD
CO
o
m
O)
*-
00
en
CD
CM
I
I--
co
CO
T~
Observations
p
Private Systei
m i^.
0X1 _L
p o
"+~
f- CM
a> ^~~
^
in CM
^ 5
(
+
00 (o
Is- CM
CM CO
T
•* c->
££
~
O> (£>
m' o
^~ 7
f
CO
|1
CO C
Percentage of
Confidence 1
t^
ID
o>
CD
CO
CM
0
CM
CM
CO
S
t^
J2
Observations
All Systems
t- N-
•t ,
CO o
~+~
<=>. 0>
en ^
^
s ^
Is— tQ
s«
•* co
O> ->3-
m ^
T"
N. ^
co T-:
O> i~
« 0
+~
CO
ll
co c
Percentage of
Confidence 1
CM
CM
CD
00
O
?
I
s
o>
o
CM
t^
T~
in
in
Observations
CM
CM
O
m
Ta
Q
-------
D)
c
!E
o
S
o:
E
Q
* «>
O 0>
*: in
o E
re CD
;*
C 10
0) V.
> »
£ E
Q- o
O 10
13 O
a> n
c ™
O) C
'to IE
ol
S> c
*- 0
ible 43
ol Program tha
ovides Protect!
wnership
e~£o
c3E£
i- «
ave a Cross Connectioi
ed Distribution System
X C
» 15
o 1
*j 3
2.°-
co a
•5
1
c
CO
a.
o
IS
o
c
o
£3
ra
a
o
a.
a>
o
'?
System Sei
»
Q>
.N
^
* 0
> 0
O -.-
§
in
1 o
§8
- 0
is
1 o
T- 0
o o
0 0
00
IO T-
,18
0 0
o" §
1 o
T- O
0 0
**! o
CO T-
1 0
T- O
O CO
10 p»
• 0
T- O
0 U>
§ I
o
Ownership Type
I — CO
CM "^
T"
in co
•^~
CO to
17
to oo
^ 1
^.
T CM
~
CO ^
T"
CO CO
011 _!.
~
0> T.
c\i "°
in
Public Systems
Percentage of System
Confidence Interval
r~
CO
CM
o
co
oo
o
0
o
CO
co
Observations
Private Systems
r~- co
" ^
i
+
o o
0 0
•^~
t- t^
O> co
~
CD o
CO t~
*~-
'
+
CO If)
~
CO CO
06 «-
CM CM
T"
CO CM
1
CO |^
co" •«:
CM CM
~
CO N.
O o
T"
CO
Percentage of System
Confidence Interval
in
CM
in
i^
T~
CO
0)
co
0>
Observations
All Systems
CO t^
CM "^
~
CO r~
s ^
CM ,
—
CM \f
~
h- 00
C^ i
'+"
CM T-
"T~
CO Tf
N _!.
CO o
O> |<
o> oo
*- CM'
T"
CA
Percentage of System
Confidence Interval
CM
'"-
!5
^
^
CJ)
en
CM
CM
CO
s
in
CM
Observations
Q.
0 ?
•S3 -0 •-
O CO 73
s i >
o > o
**• >. °~
^ "O m
15 •§ S
i. ™ 1
M g>,2
So.-
IN
d) co
£ 4: »
£
0 ij "
~ S •-
to c
^ *rf C
•C c m
1- CD g
> CL
ff 2 CO
M Q. J—
TO *-•
0 T3 ^
Q. CO ~
nt and Isolation
rams are design
/ide protection v\
0) D> o
III
*rf -i* c;
.23
lese systems have a Cor
the tap. Containment-or
stribution system, but do
otection up to the meter.
0 P S =5 Q.
Data:
Definitions:
-------
CO
E
2
u>
r£
S
'c
o
o
c
Q
O
C t-
0.2
0 ra
S "3
S§
o °-
.h *
^ ^ CO
.a Z e
CD c E
K o »
E (A
CO >
u] «
£ CO
o -a
ffll
•c >
g m
M
E
1
s>
•5
§>
c
1
0.
^,
o
O)
B
CD
O
C
o
'•ii
3
a
o
a.
a>
o
CO
w
E
S
w
CO
N
55
m °
,1 °
0 o
0
in
,1 0
§8-
O Q
O fft
*~
1 0
£g
o o"
o o
in T-
1 0
T- o
0 CS
°. o
0 U>
*~
• 0
T- 0
0 0
"V o"
CO *™
1 o
o «o_
1 o
T- O
o m
g 8
O
E
2
I
n_
Cross Connection Control 1
Element
CD K
CO CM
O> ,
+
CO r»-
S «_
-I-
CM CO
in M.
oo ,
~+~
OO co
f
CD CO
a> 10
CO ,
+
CO TJ-'
o> ,
—
co 10
8 7
+
CO |<
o> ,
f- 0
m o
co ,
+
Right of entry
Confidence Interval
mo CM o
CM co CO CM
+ +
CO CM CO T-
cd 10 csi w
00 , O> (
-t- .f-
f~: CM *- r>-
§
+ +
in co c3> co
o 10 m' co
C7> , O> (
qf- —
CO r- ^ T-
f- CM OO CM
, ^i
+" +"
•*. CM 0 0)
~ ~
CM CM •* CM
O) T— C7> ^~.
T T~
»& |
2 *& % E
.^'g «"o 8
Surveys/inspections to identil
connections within the systen
Confidence Interval
Policy specifying which servit
connections must be equippe
backflow prevention device/a
Confidence Interval
O) M.
CM CM
o> t
+
•* K
CM 10
CO ,
_)-
m o
s t
"^~
CD CO
O K
oo ,
~
00 o
O 10
o> ,
+
m CM
00 "^
i- ^
0) ^
T
§S
T
CN N.
§ 5
T
"m
Enforcement authority to insti
devices/assemblies
Confidence Interval
q CM
T~ CO
05 1
+
O> o
OO CO
^.
0>
CD to
1
-f-
CN 0,
in ^
•
+
10 00
S *
+
0> CO
in co
oo ,
~
CM K
S *f
~
OO co
i^
3 £
~
in
CD
CD
CO
CD
Enforcement authority to test
Confidence Interval
00 co
^" ^fr
00 ,
+
^- o>
CM CO'
~
CM 0
£ 2
1
-)-
CM CO
CO CO
~
CO 10
o K:
oo ,
+
CO co
•^~
CO 0
g *
+
I *
i-
CO N.
OO |<
~+~
JZ
S
Penalties for non-compliance
ordinance
Confidence Interval
CO CM
oS ccj
T"
CO l^
CN Is."
m ,
^.
•* co
S £!
i
+
00 CO
§ S
_l_
+
to -^
S °*
+
O CO
CM W
,
+
CO CO
!* t
1
+
CM N.
2 o>
_!_
CM o>
CO o3
T"
Public education programs
Confidence Interval
CN 0)
CM s;
CO ,
+
r-. co
00 to
^.
OO CM
CO S
i
+
CD CM
^~
CO T}.
csi oi
CD ,
+
CO oi
co ,
•* ro
m T~
_i_
+
o o>
CN CO
CD T-.
CO 10
~+~
E
CD
CO
Training/certification of tester
inspectors
Confidence Interval
t---
co
_
m
^
^
O)
CD
CM
CM
CO
^
in
a
Observations
CN
d
S
Q
-------
c
o
o
1
0)
3
Q
"5
3
c
c
^
<5
u.
•O
c
ra
•O
s
to
m
H
£
ra Q
** ic
^ £
i » c
.0 » >
£BO
E >
0) CD
»
tn
.0
c
o
£
i
u
a
GO
"S
§,
c
a
a.
c
£>
O
O)
1
O
_
JS
a
o
a.
d>
u
'E
0)
U)
£
S
(A
><
tn
V)
£
!§
> 0
o 0-
o
10
•r- °
|i
1 o
T- O
0 0
°. 0
0 0
10 T-
' 0
T- 0
0 0
0 o
o 10
1 0
T- O
o o
«"> 0
CO T-
1 o
T- O
§ £
1 0
0 S
O (0
s
$.
Ownership T]
csi
0
d
*
CO
CD
CM
CD
,_
•*
CD
^
d
0
CO
CO
£
*-
M> >«
S w
Public Syster
Percentage of
00
O
CM
CM
~
CM
~
00
•f-
\~-
,
~
CM
T"
o
+
c>
^
M-
CO
^
CO
^
s
Confidence ,
-
CO
CSI
h-
CM
O
CD
CO
CO
CSI
O
00
O
CM
CO
Observations
to
Private Syste
CSI o>
*- o
m 10
,
~
CO 10
CO CM
T"
•* oo
CO T_
T"
"* o
CD co
,
"+"
IO cr\
<0 00
_i_
+
CSI xj.
0 0
+
«=> s
^
0 o
^
tf>
m co
-2 ^
CO Ic
Percentage of
Confidence *
0>
CO
?
m
CD
CM
CM
oo
CO
Observations
O (o
csi o
in ^
0> cvj
,
+
Is- t-
CO ^
~
00 to
-f-
esi ^
CD T-
,
~
m ^
•* CM
T"
f~: o>
0 o
+
0 o
T"
T CM
0 0
*
CO
m ^D
•S ^
cof
All Systems
Percentage of
Confidence i
in
rg
CO
?
o>
CD
O)
^.
OO
S!
00
CO
CO
Observations
in
CM
0
S
co
O
m
r-
-------
«
11
£ o ^
at = * °
<£ m O .._
"* > 3 ^
•S i? o «
•8-wI
Hl « 1
£H
ffl
S
N
CO
<
j- 0
|8
0 o
O
u>
i°
g§
° u>
§,- o
ST- O
o o
O °- 2
C U> T~
o
«
lit
(L ^. o
o
'e
0)
CO ' 0
E o o
fl> ff^ Q
£"'"
CO
sl
M> (X1
o *o
T"
o «
o «
tT" w
O
^
5
•5
a.
CD csj CO *O fs^ IO *~ 10
CO r* CM CM CO O> fv,.
CM , T~ "* ^ "°
+
CO CO C*s| O O h- Cs| Ql
•o co CD" + t-" ,^:
Is- V- CO IT) ^CM COCO ^ COcM
^~L "^" °9. ~T~ '"I 'O
CM" _;_ f— co" _i_
+ +
CM t- O CO CD n-
CD v~ CD -I- O) x-
+
_ __
'"_!_'" T" et> _L
+ +
COtoOCMcoC5 »OO
CM.CMCOWCO meg
+" -L _L
IOCMCMCOK.CO ^-O
— - , ,
+~ +~
£ Q S
£ * S.
to £ »
a ^ ^^ a ^
n to o (0 w 03
S 1 £1 S |
1 1 ^1 8 c
2 Q) .S CD co iS o c ^ o
CD CD *0
CO
o>
oo
CO
o
o"
10
m
CD"
CD
in
CO
CM"
CD
IO
CO
o>
CD
CM
CO
o
1
10 xj
1
+
o c
T"
o c
_.
CM r-
+
0 c
-c-
CO O-.
IO cc
If"
CM CO
CM c\l
,
"+"
,
£
"G ~o.\
ll
% not chargin
Qnnf/rtenrp
o
in
CD
CO
CD
O
CO
in
CO
CM
CO
CO
o
IO
5
Observations
T- •» CO
'—CO *O
*-" _!_
+
S3 S
CO CM ^*
m" o in
T— t T—
+~
O CM *°
CO (Q CO
CM" CM" o"
CM , CM
CO LO °>
*~~ co CT>
h- o> CO
T~ CM" ^c>
^~
st P
10_ o> •*
CO _i_ CM
!T co 7^
en *° rC
"+"
CM 10
1
+
CM If) 0
— -
OT
£
S -
_ r-
T— ^
T
O d *O *~~ '3Q
+ o> >:o
O O) I"*"- CO v—
CM IO co
— °* *-
oo" t-"
4~
— CO i-T)
+ ^ _i-
•* 0) -S £
•I 8 1? 8
j£ c o « c
tu (j) ^3 C Q)
|i 1 III
3? 0=5
<
CO
o
OO
en"
CO
co"
CO
CO
r--"
CM
0
CO
CM"
oo
CD
?
o
CM
CM
Median
0> 10
1
+
o o
_!_
+
CM co
•«• 10
CO CO
8£
_l_
CM CO
,
T"
B
i
£
73 ^5
i|
o>-S
% not chargin
Confidence
E
, —
T-"
S
?
o
en
•*
o
CO
o
CM
8
co
^
Observations
CO
CO T3
^ CO
— o
. O
O
•o ^
H s
"to o
^1
§ r?
T5 "5
"to £
£ S
C CO
If
f i
CO >-o
Q. "> O
0 CO °
Q. "m "^
ill
C 3 CO
CO O N
01 o •—
S *• s
— w) «*-
O> CD O
Is? S
£ CD C f~-
w £ "55
c c to"
S "> —
S £ "5
CO _W H-
|1 i
C m ®
^ | ?.§
LJ t^
-------
^^
CO
i .e- o
„. i^o
fe » C -
<" » £ I
Soc | ^
H m* 3
b
^*
o
O)
£
co
t)
c
0
re
Q.
£
o
o
'E
0>
E
In
CO
s
N
CO
<
*- O
> 0
0 o
0
IO
1 o
is
§1
*"
1 o
T- 0
0 0_
o" o
10 T-
1 0
T- 0
0 0
^. o
o 10
^
1 o
5§
n.o
n *-
1 o
T- O
85
^§
0 IO
*~
0 V>
o
wnership Type
O
g§
CO CM
1O 00
u ^
00 10
J CM
~
§g
CD T-
CT>" •<}."
~
10 °°
co c\i
oo" w
1
^ °
CO CO
CM" _i_
IO ^
i^- -<~
•jr
00 co
T"
S^
+
jr oo
"5
&
Q>
ublic Systems
Mean
Confidence Int
Q.
CO CO CM IO
T— -i-
r*- o o T~
o> ( 10
o" +
00
•* — T-
co"
CD 0 0 10
CM , O)
1*-
CM r- CM CM
O CO
CO -£-T-
CM +
CD — T-
CO h~-
CD IO (o CM
CM , h-
+
0 O co CO
•^ CM CO --
+~
£
CO
3
£
|"-5
co ^
;— CD
Median
% not charging c
Confidence ln\
Observations
rivate Systems
a.
CO x[!
^
0 r}.
CM °°
8~£
T~
CO tjj
^J" lr>
CM ,
~
00 CO
a>" CM"
_!_
y> r^
OO CM
CM" w
~+~
Is- CO
~+~
OO to
™1
+
CD CM
"5
&
CD
Mean
Confidence Int
CO CD Q> •*
CO -t- O>
T"
m o o 10
o —
CM" +
CM
O O o !O
£
CO h*- c\i C^J
co ?: T-
° T"
1*- CM T- N-
CM" Tf-
m a> -a- co
CO T~ ^~
CD ,
~
•* CD ^ O
O) T- CM
r^ o> irj oo
T- CO T- t^
•5-
CM CO to O
• -^
^
8§
O> co"
~
CO ^
85
CO" T-.'
1
Is- OO
00 CO
CM" _i_
0 -^
~
OO CM
CM *?
+
CD CM
1
"B
^
CD
II Systems
Mean
Confidence Int
<
$2^Z
• T—
^~
[x. o o ^D
CO i ***
0>" +
0)
r^ o o *-
CO , O>
4.
co +
co T- »_ r-
•r- O
CO -^ T-
r^
CMCMo,C»
CM" +
•* ^ "^ co
CD -i- T-
*-;i
O CD T- O
CM CM v- 10
~
CM CM to CO
•^ T- CO
s
CO
3
£
"t> CO
2 c
•— CD
Median
% not charging c
Confidence Int
Observations
£
to tn
E <>>
1 i
CO
O CO
||
C 0
118
CO Q. O
CO i_ >C
-^£-
C* ffl fB
£ i co
8 S 8
CO CO '«
§ c •—
-S tt) ^—
W £ —
>»** ^
CO TJ O
s« i
-q § S
g |.i%
^ 3 B-
O f— co co
. . CO
3 £
CO O
O "Z.
to
E
£
to
tf>
jD
CO
,>
Q.
cz
o
S
CD
73
dditional
CO
£
5
S
?
c
Refer to the
-------
™
o
C>
o
n
o
c
o
ts
a
o
a.
0
g
o>
E
w
OT
co
co
=
<
i
•
0
o
in
a>
>*
co
0
«•>
0
u>
i
L.
o
s
£
£
JC.
c
o
O) Co <*} O> to **>
TJ- £? CN T ^~
~
O CNJ fi ^ O *^
v- IT) ^ O o ^
^~ _!_ ^~ +~
ON. °o co -^j- oo
CO ^ CM ,
~ +~
O> if) O O if) ^~
oi co to ^ o ^
T" -!-
s
(0
£
CD "t> co
^ ® C
S S •£
g £ o,-S
c CD .£ q> w
1 8 ^ °
W 03 «* (]> ^
t?" g c "gg S
1 »cS ' ic§ i
5 CD O CD O Ji
•g S 5 ^ 0
a.
C3> t^ O CD
CD 0 C t^
i °°
CD CO C O
cb
^~
CO v- C1 ^
O -r^ C CM
V
r^ o> o co
CD cS O CO
i °°
•^~
to
£
75 tg1
C ®
CO CD -~
% £ 2
** a> .E
S, g s>
J 1 c. 1
I s§ II l
1 1° 1 5
<
O CM CM 10 0
,- °° W o CD
•^ CO v-
T" T"
h^ CO ^ CO CO
o> £° "o S
Y~ l^ 10 cj)
-j- ^ ^ ^
co o o r^ r-
oi " -
Vf **
£ £ S.
CD to 1* 0)
t> c >> o
a o O) c
•2 Hs CB -2
U CO JS u
<£ > «» -c c
c % % c % .2
o 5£ So "g
0=2 2
<
i^
CO
CO
cb
^
g>
C3)
c\i
s
CO
£
1
s
% not charging i
O
^~
T"
CM
-±-
00
'
00
cb
i
+
"5
^
•£
Confidence In
CM
CM
en
CM
CO
h-
CD
CM
Observations
o
CD
CN
d
Jo
"5
Q
-------
10
a>
(0
c
o>
Q.
X
UJ
•o
c
re
§
c
0)
o:
CD
^
O
0> 0.
•2 o
A
S
"•
«
I
0)
£
o
3
1
a>
oz
!
1-
2!
re
tl
» o
Jin
I
re
>, in
m =i
o
£
O
IS
o
0
i
a
S.
CD
.0
O
co
S
u\
'
co
N
(O
fl> ^_>
> 0
Oe
o
10
• 0
*~ ^5
o ^-
• o
o ^
T-
1 0
T- O
0 0
°. 0
0 0
IO T-
1 0
T- O
0 0
0 0
0 U>
1 0
5§
«•> o
<•> T-
1 o
T- O
S 3
• 0
SS
II
*~ _i
0
0>
"•
Q.
1
a>
O
co ^ CD r~-
o> -^ •* o
CO CM *~ CO
^~ _!_
•t o f— Is-
^— O O> "^"
CM CO 00
o" ^- o>"
if) CM 00
T
co i^ Tf en
h~ CO ^" *&
^- CO ^ "*~
T-~ T-' CD"
CM , T-
~^~
O> 00 ^~
co" T-" h-
3 oo o CM
t^ CT) fr> T-
CM _i_ CM
SR|S S
1^- ^^ CD ^~
-^~
T~ CM ^ ^^
o> oo ^ ^
T"
«> 0> CD •*
CO , CM CD
oo i^ o -^
~
I
"c
§a> co
_ g §
" ^c 1
-ill i
11 I o
QL
h^dCDO CDooCOh-
10 r^ a> h- T- ^j- o>
CM , T- O) -,- O>
~ _!_
•^co^5
*~ T" ^ ^~ -L.
IT) 0 CO 10 r- CM
r*— \~. r*^- ^~ •* — iS oo oo
to to o> o^ t^, •*" T —
CO LrT ^ ^~ t~-~ to
C\J , T- (\J , T-
__ _
f>ir}CO O> CNJoo^Q
oo oo f oo csi "*" o
a> c\j- o oo T- h-
h1- o^ O) ^~ IT) CO O CO
^ CO ^ ^^ CO ^ T~~
co T-" CM" CNJ" _i_ oT
T"
^3 O CO (O {O ifi OO OO
00 CO *^ Is- ^- (O •«-
-^- -^-
5fec| S 5JM? 8
+ +
CO , T- TJ- CO "f CM O
T- ^f CD •"" °? l^~
T" ~
I 1
q> Q)
» S £ o>
E CD CO Cb ^ CO
a> o c o c c
•55 c on c CD o
ocl-s i "s's^ *
ot <
Zoo
o -s "t?
« co £
5 co E
® = a.
CD C o
n- ™ W
O £ CO
CD m S
S -^ §
Q> — o
If 5
'" o S
** « f» *••
5 ° = * a
T3 C 3 « £.
0) CO Q. » o
111 |i
5 g-S °i
> i > D> *=
g £
s c S « S
§• IE i S
a>
i £
!£
O H O S= < S:
-------
§
c
0)
£
0
a>
Q.
C K~
l*i
^^ ^ rr in
U) " C
!» |i
*l°l
Is!
(£ 5t*
*t
o
s,
I
o
s.
"*•*
8
f^
55
fc§
a I
o
IO
T^- °
0 °-
o" 2
? ">
C? ' °
Si I
,™ 0 0
O 10 T-
c
o
^5 • o
O ^» C3
£ 0 10
.0
* ' §
w£ §
E ". o
1" "
1 o
§ "
1 0
o u>
o w
? 8!
o
wnership Type
O
q o
O Q
5 _i_
+
q o
8 ^
Tf~
0 0
8?
T— _.
0 o
8?
1— __
0 o
11
"~ +
0 0
11
O o
8°
^ +
q o
° _s_
+
O o
li
+
ublic Systems
Water Sales
Confidence Interval
Q.
O> o>
O rx.
CD ,
+
CD CM
f2 ^
~
^ N.
CNJ if)
Is" v~-
_i_
+
CJ> oj
0> CJi
+"
•*~ CM
ST
+
N. 0
00 ^
•^ rj-
06 •»•'
10 »1
i
C3> -q.
CD ^
CO v
~
CO io
58 «•>
CM CO
•^~
M
£Z
_o
Water-Related Operat
Confidence Interval
•*t CO !*-_ IO
r^- co f^ to
1 *° i
+ +
^ O CM CO
CD to g to
+" +"
co o co o
£ "f " S
T" -!-
+
•* ix. rx. to
co - o
+~ _!_
+
• O
CD -^J.' CO CM
T
CM -^ 0 to
r^ K CM o
T T
O> CM * O
m s; o 06
_!_
+ _!_
o CM Q
*" J_
+
q o
8 ^
+~
0 o
8?
^~ ~-£~
0 O
§'?
T — --
0 o
8?
^ +
0 o
8 =
0 c,
8 =
"" +
q 3
O> ^j
O) ,
+
0 o
8 =
+
rivate Systems
Water Sales
Confidence interval
a.
IO ^.
o> oS
0X1 _!_
+
rx. ^.
IO to
|x- CM
'
+
O rj.
CO t
•
+
CM o>
CO CM
i
+
q co
rx- CM
:j-
O \j.
CM rj;
|x- CM
r
O -.
^"
C7) |x.
co c\i
T"
CO
.0
Water-Related Operat
Confidence Interval
rx- o
rx. cj>
^ _!_
+
10 LO
d ,-
OO CM
1
+
CD TJ-
0 0
rx- CM
_;_
+
CD 10
CD •«]-
rx- CM
_i_
+
q o
^~ to
CD CM
~
CO -q-
10 CO
10 CM
T"
o> ^
10 CM
O> IX,
T— fs^
CO -r-
~
T <3>
O |x^
T"
co
S o
i
T"
0 o
!
+
0 o
8?
T— -—-
0 0
8?
T— _—
0 o
11
"~ +
0 0
1 J
q o
8 ^
^ T
10 o
O) ^
01 _!_
+
0 o
IS
+
II Systems
Water Sales
Confidence Interval
<
CD o>
CO 10
^ i
+
CO 0
" ^
+
CO CM
O> Tf
CO -S
i
+
'-. 1O
~T~
0) o
s^
+
o co
ES^
0> 00
CO r-'
10 -^
^~
cq io
^- ^;
CO -^-
T"
^ o
O> CM
T"
CO
C
o
Water-Related Operat
Confidence Interval
10 co TO
•* c\i CM tb
i ^ i
+" T"
CM to CO v-
•0 u-j 0 to
T" T"
^ to 10 oo
Oco CO^
+ _^
+
rx- TJ. en o,
r-' ^ cooS
~ ~+~
CD to O o
CMto CMrx.
T" +~
"^ v- G> v-
U> ^ *— Y-"
^
O> 0) >0 CM
10 ifj goi
T" 7"
CJ> lo CD CM
CM CO ^ •»
"V"
T- -^ CO rf
O QJ O co
T" +"
s
General Fund
Confidence Interval
Other Revenue Souro
Confidence Interval
>•
c
0
in
8 §
T- *=
O l—
E |
I/) 'D
1 1
x £
CO i_
en g
S
2 ?
cr ^
E "
1 |
£ S
£ ^
o E
S Q
. . M—
1 §
0) J
o «)
(D ^
i |
c S
5 c
e 1
g §,
E ">
I 1
ID »
i— 3
i!
CD JJ ^
°J !J -2
d w |2
. . in
CO CD
"5 'o
o z
o
CO
-------
0)
o
CO
0)
3
C.
CO
CD
C <*
i*r°
•o -- o>
^f -» Q.
5> s 2 £•
« * >•
"S a> ^ m
(_ 3 O fB
m ^* C
> S
a. ,g
1^
o
g,
to
c
0)
a
o>
a
'
to
a
5
o" ' o
Q. 0 0
0) °. 0
o o in
C.
§«
m
_J
O
Ci
1—
O.
I
e
O
O> K.
oo *7
+
^
87
+
°> 0)
8 ^
T"
r*- LO
£ 7
+
10 ^
S 7
T"
CO 7
'T"
o n-
0 CM
i
+
^ CO
s t
+
O u>
iS w
CO CM
•T-
. 1
(0 i_^3
i 11
0.
TT o> CM Oo N- co
CO 0 -^0 10 T-
+ + +
LO CM O LO ^t" OO
^7 " ? °° 7
~+~ T" ~
CO CO CD IO f*~ Co
•^t t- O o IO v~
f T~ T
CO v- O> oo O> f^
IO CM O O LO T-
T" ~+~ ~
CO Ol CO ^J- M- CM
IO it- O o "O t~
1 1 1
+ + +
CO O "O CO CM \J-
1O CM •>- -r- TT T-
~ ~ ~
0 o, •<»• CO 0 LO
CM o ^ W CO W
1 1 1
+ + +~
CM oo CJ) v- CM N.
CO CM O W to xj-
1 1 (
+ + +
q 10 co N. to LO
OO co O o
^~ -!__!_
~+~ + +
CO
c
o J?
CO c
§! ! ^!
Zl -S -£ o> -ffi w
^ C "O C 3 C £
So o o £ o 2
§ 0 O 0 O O £
n.
COQ oof^ O>Q Oco f^io> 10 ^j- c\Jo
lOcsi CNJ>- -r^W Csit- COo OQ> Tr-t-
o> , , , o> , , , ,
+ + -f + + + +
T— v- OOif) COlf) Oo "^"t— O>TJ- O>J^
tOCNJ fOc\| OQ f^-CM ^ i- Oo f^-v-
o> ( , ( oo , , , ,
~ ~+~ ~+~ "+" ~+~ ~+~ ~+~
OOcsj O>(^ "^"O COto f^CM °OlO CON,
I^-T^ Oo *~W O)c\j ^"W OQ lO^—
o> V , , °° , , ,
~ ~ ~ T" T" T" T"
fOc\j °O>- ^CQ ^^J- fH O OO|^ lOlO
tX>t~- Ov- f~O CO CM Oo lO-r-
o> t , ( co , , t ,
+~ "+" ~ ~ T" T" T"
^\f h-co r-^^o ^o U>N- 'Oco ^c\j
Olfj ^"C\j "^ ^- OOCM Ift *r~ Oo ^O'x—
til 1 1 1 1
+ + + + + + +
IOLO o>
CNloo ^"C\j COc\i O>CNJ lOv-" ^~ v- ^ v-
o> t , , oo , , ( ,
T" T" T" T" T" ~ T
cocsi OCQ ^co f^co "^ CM "*~CM f>Tj-
CNJ ^"io fOco T~c\j cxj-r- ^~ T- tf> W
^ i i 1^1 i i i
+ + + + + + +
^;co °.N- *^c\i ^^^ coco f^*^}- f^.oo
IO'CM N t-' C^CM f^ico c^Jt-; o'o ^"CM
______ ________
tpco ^CM ^P^o o*^i- fS^- cjo OQJ
_>_ ___ _!_ °* _J_ _L -L. *
•*• + + + + + + CO
c c
-2m .2 S
_2^|_ _?_ _1_
"c59?"e6Q"55 "t5?"c5 "CDQ"^
S 5 o> -2 S ^ .£ §
•2-g i-g ^-gs ^-g ?•§ s-g ^-g
S«g B'g S'gS.S'g S'g ?! »>§
.go -go £§<&.»§ .go coo £o ai
5o §o oo= §o 5o oo oo -g
< o
-------
£
o
O!
co
O
0>
. i
I I
0> O..C
3 ^ °
So § ClS
»|S|
J2 o: j o
§•• 2 ^^
•S > 3
?ml
S >
ra DC
5 -5
CD
C!
5
C
I
!
CO
0
0 §
o
u>
• 0
^^ ^
§1
l§
£•
o
0) ' 0
o) T- o
oSo
c «> T-
o
•5
co
IT! 0
|«"S
O *~
d) • 0
= s§-
C « 0
^> CO ^
(A
CO
1 0
T- 0
o n
u> „-
1 o
T- O
0 U>
§ g
- 2
^
o
0>
a
O
CO CO
+
„. ^
n 1
~+~
iq co
uo ^
*
>O N.
•v~
CO T-
510
,
~
CO in
eju.
"*"
O> (o
10 ^
oo ,
~+~
CO o
CO h^.
O T-.
1
+
1
S
C;
<0
ublic System
Residential
Confidence
D.
O CM
CO CO
+
O -^
0X1 1
"+"
O> CO
cQ ^
^"
10 CM
CM ,
~T~
CT> csi
CD «.
CM ,
cn 10
"*"
CO CO
^ <*>
-(-
CO o
? d
+
00 ^
t—
t
+
1
S
!ss •£
Non-ResidenI
Confidence
CO V- T-
«- g
+
CM ^~ in
CO ,
~+~
^ »
•T"
CM CO '-
CO CO ^
T"
r-- oo CM
t'cM 2
"*"
*w S
^
^~ c\i "*
1
0 o CM
o o T"
1
+
1
"c
Wholesale
Confidence
Observations
'": o>
+
iq CM
r- ,
~+~
cp »~
S 2
•jr
CO CO
1
+
cp o
S ^
1
ro rf
CO ^
*
CM CM
Sg
_*_
+
CM !O
P CM
C3> ^
T"
|
~ CO
~
O K.
CO K.
~
0 K.
+
CM N.
CO 05
^.
CO CO
o d
T"
p •*-
~
1
3
~m c
•.^ •
Non-ResidenI
Confidence
22 |
+
•O; o •*
,
"+~
CM N. 1-
CO CO ^~
~
CO co
_1_
+
O 05 CO
CO N. *~
+~
cp 10 co
+
v- og CO
+
-5 *
+"
CM rj- f-.
dS <°
T"
1
"c
Wholesale
Confidence
Observations
•* CM
ro ^
+
T CM
•* ,
~f
CM CO
S ^
^
O CM
T"
O CO
CO Xh'
CO ,
T"
CM -r-
jjtri
+
O> o
ID M.'
CO ,
+
O T-
CO uS
O) (
+"
iq co
-fT
1
s
c
II Systems
Residential
Confidence
<
CM o
d CM
~
CO 0)
1X1 1
"+"
iq N.
CM" ^
*"
r^ oo
CM ,
~
•* o>
8?
+
0 co
0 Tf
CM ,
O> CNJ
0 co
-^
0 c.
T"
•* co
"+~
1
•2
~ta c
Non-Resideni
Confidence
CO c» T-
<^o g
T"
CM K. O)
0X1 .
~+~
p -^ co
CO CO JE
^
i^- co 2
Jco °°
~r~
O> o ^
^ co £!
+"
in LO co
•* CM ^
"*"
22 S
^
co rj. co
•^ c\i "*
T'
CM rj. O>
T"
1
S
c
Wholesale
Confidence
Observations
c
TJ
C
S
CD
3
T3
O
O
0
|
in
'o
^^
CO
E
CO
1
C
(O
^ "5
0 0
. . CO
to tj
Q Z
CM
CO
-------
o
to
CO
o a. o
_ c -c 15
S i go
il|§.
K 5 O ^
O ,*?* Q)
h- DO Q
2
1
O
O)
a
to
o
o
IS
3
a
o
D.
CD
Servic
E
5
in
oo
i
CO
5
S; o
> °
o °-
u>
i|
81
1 O
T- 0
0 0
°. o"
0 0
in T-
1 0
T- 0
^3 O
O O~
0 S
1 o
5§
n. 0
CO *~
1 o
T- O
O CO
^^ CO
0 IO
is
*
o
&
wnership Tyi
o
to
CO
0)
*-
CSI
o>
CM
1^
in
CM
to
•*
csi
CO
00
CSI
to
5j
^
to
i
>.
OS
li
a.
CO
o
~
CM
o
T"
to
o
+~
CO
CM
+
CO
o
1
CM
CO
o
~
d
T"
in
S.
T"
CM
v^
T"
5
Confidence
m co
CM
CD T-
co in
'-
cq t^
co in
h- 0>
CO O
CM CO
CM T~
•* CJ>
»- CO
csi ^~
co r~-
co to
tsi T~
CM CO
to to
O 0
in
Median
Observations
oS
•*' o
^
p in
to o
T"
S3
csi o
csi o
+
oo in
csi o
i
CO o
T"
S8
•*' -r-
"V~
CO S.
CO Q)
T"
g> CM
CO CM
^ W
T"
S
v> S
rivate Systen
Mean
Confidence
a.
T- in
•* en
csi *~
S10
csi
co m
csi
0 T-
co
0 CO
csi
T- r--
co
•* in
CO CM
CO
CM CO
p •^-
csi
o to
cq in
T—
Median
Observations
00 o>
f CM
CO o
,
^
0 ^
csi o
+~
OO -q-
l-~. CO
•^ o
~+~
JO CM
CM CM
csi o
+
S^
csi o
i
CM CM
to to
CM 0
~
So?
to o
T"
to •«-
in in
tO Q>
~+~
to t
^ w
7-
5
II Systems
Mean
Confidence
<
co to
m csi
CM -
^~
t — to
co m
'-
CM O>
CO CO
•* to
OO O
' T—
CM tO
to •*
CM T~
csi ^
00 CM
p o>
CO ^~
r- T—
csi T~
o> r^
o> to
T—
Median
Observations
i
*•—
"c
to
c
T3
C
CO
c3
I
0)
(D
CD
-&
O
IO "£
o *
t\ *
\J ^y
CO 3
CM o
O =
. . in
to "o
O Z
to
0
o
in
CM
c
^sterns servi
lively.
CD CL
1 s
Q.CO
c in
0 C
O CO
CO -
«*> 0
o ^~
10 o>
TJ IO
C «X
CO ".
o to"
o •<*•
T to
m CD
csi S:
O) a
.— ^
CD ®
in CD
to <°
"S jg
to "co
o E
^3 "«
l|
•g "g
o.x>
0.^
T3 .E
s-
* o
(5 co
.2 «?
= 0
oS
CO
CO
-------
§
A "
0 o
0
in
1 o
° §
|§
1 0
T- O
o o_
o" o
10 »-
1 0
T- 0
0 0
o o
o u>
^
• 0
T- 0
0 0
*t o
CO ^~
1 o
T- O
0 CO
•» co"
• 0
T- 0
3 U>
gtn
O
8
Ownership Tyi
,_ *
00 CM
csi o
17
CO §
••- o
1
+
CO o
•* co
*- d
^" N.
CO U
*- ci
^r
CO M-
CM 4
CM o
_L
+
O if)
'- M.
O s.
co ci
i
~
CO -q-
CO T-'
^~
1
•> -S
Public System
Mean
Confidence
oo o
CO CO
csi «>
CO T-
m m
•^
0> Tt
m t^
^^
,_ ^.
CD CO
in co
CO CSI
in r~-
oq co
CM CO
CD CO
csi *~
CM 00
CO CO
csi
o o>
in
Median
Observations
2
Private Systen
O o
O CO
"»•' o
T"
S|S!
csi ci
•
+
CO CXJ
•* co
CM ci
T"
00 to
CM 0
~
CO o
m 10
csi ci
1
+
CO rj-
CO OO
CO ci
~
§R
CO o
1
CO Tt-
CO g
CO ci
1
CO??
^ w
-£-
£
s
Mean
Confidence
h- h-
CO O>
csi ^~
co m
CO
csi
i^- m
csi
^ ^
CO T-
csi
oo co
CO T-
csi
CM 00
CO
•
o> m
T-
Median
Observations
m o>
CM csi
CO 0
T"
CO T-
•^ o
i
+
CO o
m co
*- o
CSI OO
°1 T^
•^ o
+
CSI co
csi o
'
+
CD |^
CM CM
CSI o
T"
CM CO
CO ci
~
0 TJ.
in <&
co o
I
^
«S
^ -r-'
^-
(b
^
All Systems
Mean
Confidence
t~ t~~
CO CM
CSJ°.
CM CO
co in
•^
•* CO
CO CO
^
m m
CD O
in •*
o> ^r
•<]- in
o in
csi *~
00 O>
csi T~
TJ- co
CM T-
csi '"
CO CO
co co
•«—
Median
Observations
c
CO
0
o
in
CM
O)
c
0)
CO
c/)
E -S-
0) CX
(D (jj
Q. CO
-D °*
m
ro T3
O C
O CO
CO .
o •"
m ,0
T3 CSI
i "
88
T co
m m
csi E
0) ">
c ">
1|
£03"
a> en
o E
II
E £
»t
~° "2
(D *D
CX 3
O Q
•o .E
g> S=
in d>
ci * 5
- 2 c*}
cS .£ co"
N- 13 CD
o o S
CO O
o z
-------
c
>
o
•O
1 1
m in
I a.1
jiff
S 5 §
i* = O T-'
* m •
1 1
co 9
B
5
1
co
5
*- O
l§
0 o
o
IO
• o
5 g
o -
go
° ">
1 0
fro§
O 0 o
Woo
JJ U> T-
n
O
'•=00
5 o 0"
3 d u>
a T-
o
a.
0
|S 1
s""
CO • o
T- O
o co
in „-
• o
58
1 £
*~ _l
^
o
Ownership Type
T- rn 10 cxj cxj •«.
co co o> co in N.
co o cxi ^ co o
~ ~+~
T- CNJ 1O IO *- |X»
p i-. oq co oq T-
cxi o W T- o
+ +
CO o CM CM (O n
p v- p CM P O
(MO CM ^~ co •*—
_i_ _i_
+ +
OO fry O> O CO QQ
CN i o C\i CM O
_i_ ,'
+ +
CO Co CM "^ (D Tt-
oo to r^ a> co co
CM o ^ CM O
1 1
+ 4.
CO 10 CD IO IO c\|
CM O CM CO W
^ ^
CO oj CO O CO oo
CO to O *~ CM ^j-
CO O CO ^~ ^ W
-!- -!-
O> (o CO OO CO (o
CO o> ^" ^" O if)
CO O CO CO W
T" T"
IO o O 00 CM CO
CM S; T-; CO 10
co c\j »ri T-^ Qj
1 1
+~ T"
£ £
UL a> co co q>
2 o c *= o
= c .9 S c
•K <0 -Q to T3 "O
i I c | 1 g Se'g
-§g(3 •§» ^So
= S5 50 §5
•§ £t Z
a.
in co
in i-~
cxi •*
CD r~-
cp co
CO CXJ
T— T~
o en
in o
o o>
cxi
CXJ 00
*~
o> o
N- N-
cxi
cxi
OO CXI
o
cxi
Median
Observations
h-cocom cococnio cxiQit~cxi h-oi^-cxj
r^-coioo cocococxi IOQCDCD coiococo
•r^ d •<- n cxiocxi^ co cxj ^ ^"O^
T" ~ ~
lO^^CN CXIr^Ot COloCDCO •* co 05 CO
•^CXjCXI^ pKCO -r^cxjO Cf>o^ :Q'
T— o T~~ COcSCXI COQCO T— T— T— CD
11^
+ + + + -C:
c
o
o
^f^j-lOI^ CXIonCOCXI COcomCXJ ~*t ir\ OD O)
in cxi ^ ° ^- in CM T- coioco-r- oLO lOrj--^O l^-cof^-CO
COloCOCO Oco^"^" OcoCDT— h-coCD
*— O T~ COWCO CM Y~^— T~OT~
"+" +" "nT ~
COcor»CMO OOcoCMCO CO'CT-CMCM
O>coO>T- CM IS. OO CM CXJcop i-loCX|
^— o "*~ •^j:o'^: coxj-'^o •*— -»^'cxi
-!- _!__!__!_
** *o *t m *t co cOoj^-co ** *o
05 C) •<- 05 O» co °0
cxi ^ cxi cxi o cxj
~ ~
* • * O 05CXJO-* Oo°-* 0>loO>CM
OOY~OOO OoO Oco1^
cxiwcxi a i o o ^^T-'
i i i
5 555
Stft to CD COCOQ) CO Cb CO
cp oc=Oc Sc
cog co£co CO
0> ^J5«CD ^O>Q> ^ 03 '£Z
®"D (8W.5T3 CO"O"O COfl>"D CD
Scg .sgcogcg i g 8cg J g Scg .II
^eo,o -a m a, -Sja2 -jr®0 jz
|5 50« ig5 50 §5 50 |5 50
i ! * z f
10
00
-------
o
•D
J£
n
CD
m
i_
S
55 (Cont.)
e-Custorr
1 =
ll
^
yL
a
(0
1
—
2
o
v>
c
wnership
000 Galloi
O •«-
Z»
co
I
£•
o
C3)
£
CO
O
o
ys
re
3
a
0
a.
CD
o
t
CD
co
E
1
£
1
O
T-
o
0
o"
o
1
5
o
in
o
o
o"
1
^
o
(0
eo"
i
o
16
i
O
g
^™
(0
.»
o
0
o
o
o
0
«o_
n
o
g
8
*
o
£
£
a
I
o
All Systems
"~ 05
•^ co
CO o
+
CM M.
cxi o
1
+
CO v-
*~ "r~
cxi o
+~
CO -q-
CO c\i
CM o
T
ss
cxi o
•
+
m (Q
co K
cxi o
"+"
fc c»
oo o
CM W
T"
"3
fe
4B
c;
8
c
Residential
Mean
Confide
CM 1-
r- o
cxi «>
o> o>
CO CO
T-
in •*
O CO
cxi T~
0 0
CO OO
cxi
CM CO
r-- o
cxi *~
is. o
•* T-
cxi *""
rs. O
CXI CO
CO "~
O> CO
r- r-
cxi ^~
SS
cxi
in
c
o
Median
Observati
CMo *~ OO DOco OO|s-
mK mco r^co mco
co o cxi "> <-' o T-" °°
+ +
*~(O ^~O OO^j- OO)
O)^~ h-Tl- ^CXl CXJCO
^- o *- «- o ^
1 1
^- ^
O>CO COCO 1OCO COT
CJ>Cft COCO iOCM TCO
cxicjTjT~ >-O'-:T~
~ ~
coo cor- mrs. r--o
«•$ cnco o>co coco
cxi o ^ >- o r~
T" +"
cocn'no <«• 10 CM co
OONTOO ^lf)COO
CxioCM'" •-O'-'"
i •
+ +
COo> T-00 'X3(O h-O
CJ>Q) O)CO iD\|- CO*—
cxi o ^ ^ o <-' "~
T" +"
COin C»OO CM •a- COO
io\j. r^r~- a>co oco
•* W cxi •- o ^ ^
+~ +"
T- co ^ CM « * « co
p cxi co T- T-
co w cxi *~
+"
S§ 8° §^ SS
O o CD r- -^ T-
T" T"
"5 "5
& ci
S .£
C C
CO 0) 0) Q) tf>
*= o c o c
<= c o co
CD CO ^ CO ^
"D T3 CO CD "O CO
8cl «g Se| .l|
DicflO ^jg CDCO,O -o«
gl° 18 11° IS
z §
2
CD
•5
o
c
to
E
fl)
Jfj
S2
>^
CO
*o
CD
to
3
O
>v
tr
3
o
o
^,
c
o
T>
CD
C
S
c
o
o
CD
CL
o
CD
Q.
g
in
.'
o
ro
c
"e
0)
CO
. to
w P
fl» •*•*
S «0
w >»
>» w
(/> (J
8 1
CO =
r- a.
i *
S i
a> c
O CD
= i
8 CD
S g
CO m
C^^f
a ? 2
S 1
o i
_>»
C
O
fl5
^f
&
03
"^*
52
^Z
^~
,O
3
Tn
0
CO
co"
i
o
in
03
Q
T3
2
>
o
to
S
to
3
o
to
!g
"co
£
•i
8
CD
T3
u=
cz
8
^
in
a>
•D
ca
1
o>
to
•o
(D
^:
*2
"5
§
S
CD -n
3 ? -
= m CD
s5«
CD o E
'-OS
2 o S
C i CD
CD *- i:
T3 O *-
« M- c
(D CO
e: l * -D
. O) Q)
O "> 3
"- S "5
co S .E
o E ~"
10 CD 0
1^§
CO U) Q
o o o
o = m
T- J3 i
in D.O •^>
CM ^O CD
cj) o o"
C *" 0 0
> CD T- CD
C 3 __ O_
0 C T3 «
to CD £ a)
« S
p CD o Is-
|l§-5
S •= o co
« CD 2 -o
•y 1 - ci
i£g,o
Q. ' _r OO
C O
J- o IO 'N
ox
*- 0)(N
§ _- .E o
1 ° £ (0
CD M- CD .
> CO » 0
£ 6 •» '*•
— o E •*
.2 m a> -
IE SL«
^3 co to m
"to O CD |C
CD O CO k
E 5 5 «
P CM o m
j- __ £ |S^
^^ D) > ' .
If* m-
O. £ C IO
Q. CD to |s.
| » 0 «
^ | 0 |
1— U/ •• TO
||,g»
» « v 8
C CD ^- °
S |SS
3 .2 _r CD
0 O.S:S
-------
_
.o
0)
c
c
o
Ql
8 Jf a
» i §5
i2 > O Q
Dfft
>S
c
T3
8
£•
£
o
c
2
S
a
o
Q.
0)
o
'E
a
CO
S
to
co
V)
0)
N
CO
—
^
*- o
>§
o 0-
o
in
' o
II
<= 0
C3 »ft
1 o
T- 0
0 0
O Q"
0 0
IO T-
1 0
T- O
0 0
° o"
o m
1 o
T- O
0 0
1 0
CO T-
• 0
T- O
SCO
CO
_
*"
§10
m
T-
S
wnership Type
O
Ss
+
£^
+
<0 CO
CM0?
+
CD \(.
CM *7
+
m CM
CM °7
-T~
CO CO
CO CO
CM ,
CM 10
r*- f*S
CM ,
+
CO o
m tX
CM ,
T"
CO o,
CD CO
CO CM
^
o
1
o
O
m
Q.
CO
_.
IO
i- CO
CM
O O
CD CO
CM
r-- co
I-- 1-
gg
o •«•
CM
S5
c
o
B
c?
c
o
o
i_
0>
Q.
to
c
§
<&
EL
"5
Median Residentil
Observations
CO
CM
S
CO
CO
CO
CO
o>
CO
o
s
CO
CO
00
CM
0
CO
CM
o
1
o
o
Q.
CO
CD
c
CD
>
CD
CL
rivate Systems
Mean Residential
Q.
gftS
, CM CM
~+~
to m •*
-i-
+
£??
+
S5°
-^
10 fc CM
CO CO-
•^r-
^ CO CD
^ CO ^
CM CO -
CO
CM
CM
CO
CM
CO
CM
O
'ts
CD
C
0
O
d>
Q.
to
C
CD
>
0)
OL
II Systems
Mean Residential
<
CO CO 0
, CM O>
~
CM ? CO
-^-
^§1
+
CO CO CM
+
o P"- m
"V"
•q- co r-
'f CM —
•»- m CD
10 CO CM
+"
OD m co
5J ^- csi
+
c
o
1
o
o
t_
S.
m
CD
3
C
?
fc 15
Confidence inte
Median Resident!.
Observations
CD
Q.
to
~J
c
CD
c
CD
•a
'to
a>
73
73
_3
~
dj
to
~
CO
CD
0)
0
*»—
CD
CL
Q.
73
to
CO
—
0
8.
o
o
in
CM
O)
£
M
_
< 1
CM to
6 £
- .a
CM 3
CM 2
0 0
. . in
'S 'o
Q Z
•
CO
f
c
.0
CD
c
c
8
CO
0)
t—
CD
>
"ra
'*-
0)
73
'to
CD
73"
CD
73
"O
C
CD
to
to
CD
CD
E
£
73
CD
Q.
Q.
O
73
CD
1
CD
Q.
O
2.
O CO
O CM
CM CD
O CO
c
£ o
CO c
E c
c
•c c
08
3 2.
-------
c
o
t>
c g.
§£
o 7^
^ .Q
® "*
111!
^ c. *•* —
I-- > >* O
SS-
I|
i £
"5
2
= m
i E
O M n
O) >-
5 w
(Q
? 0 CO
2 «> M
•43 n
jo
a
d> o u>
O T-
'E
CO
|| |
>,
CO 0
wnership Type
O
O CO
CO , CM T-
+
CO oo (O CO
CO cvj E:
c 0
o 0
l_ ®
Q. CA
(0 ®
~ 0)
u >
a) "co Q^
^ C «
rivate Systems
Mean Residential
Confidence inte
Median Resident!;
Observations
Q.
f- -^ 0 CM
T- c\l t—
+
O o O T-
~
CM i^ o m
T- -r. CM
+
CM8°$
•v~
c
c .2
.2 -S
i i
. 0)
co Q-
Q. (O
(0 -ij
— » C
i <»
1^1
Q) CO ^
QL p -=
ncillary Systems
Mean Residential
Confidence inte
Median Resident!;
Observations
<
O> (o (NJ O)
CO lo (N| O
CM , (N CM
+
CO CM <"O ^
in co "O CM
CO -^ CO
j|-
OO ^ O CM
T- t^ 'O !•-
CM (M
O -rj- <3> CO
•V"
C
c .2
.2 5
1 §
i §
o 0
l_ ^
•=
II Systems
Mean Residential
Confidence inte
Median Resident!;
Observations
<
JUT*
Q. T3 c
O 3 O
ill
o ~ c
0 S= c
T • °
!0 aj o
CM n 55
O) E Q.
— «•»>:
S S S c5
25 S g>
C *^ m In
fc- E c o
a> *—
£ Q.-D 0
Is gg
Q^ "O f^.
S 0 £ ra **
53 a) g>
0 £• S E So
. . in
TO 'S
O Z
CO
CO
-------
Q) "flT
E |
o 3
D) +Z
.£ co
•0 m
% 2 £
s el
3 c **
3 ? |
V) >,>
o> OQco
^0 **"
tL
_•> ®
S s
C c
« §
I 1
(0
NJ
'in
-0
> 0
0 o
0
IO
1 o
s §
p -
£•'"
o
O)
S • 0
nj T- o
Or** r*\
0 o"
ra
3
0 ' °
P r- O
O. o o
fl) ^^ O
_o o" u>
CO
to i o
ET- O
o o
JJ «_ o"
W CO *~
to
• 0
»- o
u> n
• 0
5 &
O (0
o
nership Type
1
ilic Systems
.0
d!
O CSI •* CO
csi in to CD
in csi T-
^ "t ^ f~:
O> CO ^ O5
CO »- CO
T- o T- en
T- T-: TT cb
m csi csi
Tf h- CO CO
m W cri co
• •*' o
in csi T-
CSI r- CSI 0
Csi •*' 0 CD
m T- T-
to p to p
*- CD CSI CD
Metered Charges
Uniform Rate
Declining Block Rate
Increasing Block Rate
Seasonal Rate
CO O
in csi
CO O
•^ CD
h- co
in csi
CSI
in en
•^ i^
CSI
co r-
CO CO
CSI
h- CO
W CO
CSI
CO T-
cb CD
O> 0
^ CD
CSI
p m
5 co
<5
6
73
C
a
a&
If
Unmetered Charges
Separate Flat Fee for
Combined Flat Fee foi
Services
CO
^_*
m
CO
to
ix.
csi
o
CSI
in
^.
,_;
,—
CD
O
CO
Other billing methods
m
o>
CSI
m
CO
CO
,_
0
CO
CO
CO
en
r-
CD
CO
Observations
ate Systems
_2
CL
•t CO CSI T-
cri iri iri o
CSI
m o o m
CD cri CD cri
CO CO T-
«D co en o
CO CO CO O
•^- co »-
•* co o o
o cri o o
r- csi
co r^ r^- c?>
•* •* cri csi
csi to in o
cri csi •*" CD
CO ^ T-
co cp in p
h^ CT» h^ O
Tf T- T-
CSI h- Cp T-_
CO ^" CO O
CSI
CO O Tj- O
in CD to ci
CSI
Metered Charges
Uniform Rate
Declining Block Rate
Increasing Block Rate
Seasonal Rate
to •*
to csi
•* o
v-' CD
^
o o
cri CD
CO O
Csi CD
CSI
CO 1—
co oi
CO
in in
csi ^
•* T-
cb CD
o o>
1^ CD
m T-
tb h-I
CD
Q
T3
CO
H) -Jij
I?
Unmetered Charges
Separate Flat Fee for
Combined Flat Fee foi
Services
0
^
o
CD
O
CD
O
O
O
O
O
CD
0
CD
10
0
1---
Other billing methods
T-
m
to
0)
*—
(D
•«—
CO
CSI
CSI
CSI
CSI
CO
in
o
co
to
Observations
T3"
3
C
c;
0
0
-------
1
Q.
O)
m
-?~o
*• c
c a
52
— ' 3
81
5 i:
||
"a
c
o
•O
1
1
0
CO
JI
o
(0
™ £
25
IE
O -2
m«
•5
0
S
C
o
1
o
0)
S
10
O
.O
3
a
S.
a
a
CD
(0
S
M
CO
»
w
55
« 0
Q> o
O
0
IO
• o
§8
g§
°. 10
1 o
•v O
o o_
0" 0
IO T-
1 0
T- 0
0 0
° 0
0 U>
1 o
T- O
0 0
«1o
1 o
T- O
o <*>_
1 0
T- O
0 10
0 W
SJ
b.
o
nership Type
1
in co
o in
CO CM
CM CD
•* CM
CD T-
O CM
in CM
-^ CD
CO CM
•* CM
h- 00
CO CO
m co
r-- o
CM CO
in co
O CO
CM l~-'
in CM
CO O
CD CO
co
in o
•* o
CM
Systems
Metered Charges
Uniform Rate
Declining Block Rate
5
CM
O>
CM
^ —
CO
CM
CO
CM
CO
in
CM
S
fx_
CM
CD
CD
in
co
Increasing Block Rate
TT
O
m
0
CM
CO
CM
CO
CD
C>
0>
CO
O
O
C)
O
O
Seasonal Rate
CM
CO"
o>
CO
CM
CM
CO
CM
CM
in
CM
O)
CO
,_
oo'
CO
o>
L.
-S
CO
§
Unmetered Charges
Separate Flat Fee for
CO
h~'
o
o
c\i
o
h-
m
o
CO
0
N-'
o>
o>
£
O
c.
CO
S
m
5
Combined Flat Fee foi
Services
Tf
T-'
O
in
o
CO
.^
CM
^
O)
CO
^
o
CO
Other billing methods
CD
CM
*~
00
in
CM
0
CM
r^
^
CO
0
r-
,_
CM
t—
CD
in
Observations
«
0)
o
(/)
3
O
!s
"c
0>
CO
C.
i §
~Q_
Q. £
m c
I ^
» °
2 o
5 c
8E
^
'N » -§
o i= o
. . in
CO CD
CO "5
Q Z
o
o>
-------
1-
i e
. *
(A »t— —
8 0 o
o> c S Q
K o " -s
a, S. 3 °
•S x o w
.0 UJ CO -0
"!«!
"- i o
•— f
£ H
>> "•
m
>,
c
o
O)
s
ro
o
c
0
IS
3
Q.
0
a.
0)
.0
a>
CO
B
v>
£
8
CO
<
!r, <=>
Q> C3
> 0
0 o
o
1O
ii
Is
° 10
1 o
T- O
o o
° 0
0 0
in T-
1 o
T- 0
O 0
» 0
0 «>
T—
1 O
T- 0
0 0
w o
co »-
^§
O CO
"> «
• 0
T- O
0 U>
^~
O 10
5 8
T- «
o
1
*c
tt
g
c
co
£
K
•£
0.
CO o
m •«.
CM ~
~+~
5§
CM R;
CM LO-
CO 1~
~
jn a,
>"- o>
"f" CO
co" CM"
+~
P »-
h- LO
i — ^.
CD" _i_
.f.
!5S
^~ M-
CM i
+
CO o
co ic
in r-
(
+"
S o
CO co
-£-
a«
^.
T^- co
~
n
ts
S!
1 !
? 1
rimarily Groui
Mean
Confidence ;
£
CM 0
CM CO
•>T
r>- co
CM
r-
o"
CO
CM CD
T- CO
•*
co"
O CD
CD CM
CD
CD"
m co
°. *
T- 0
O CD
in
CD co
o oo
00 CM
T- CO
CO T-
0>
Median
Observations
CD CO
m £:
00 LO
CM i
CO LO
°§
885
+
SK.
. 10
CO
CM CD
r^
a> co
r- CD
CM
S §
•I- 1^
CM CO
Median
Observations
00 I^
m o
CM ^
+
r~ •«-
jej
fc o>"
CO CO
+
£! <">
853
oo CM"
T"
S 0)
CO CD
m N.
o CN"
+"
•* o
0 io
^ CM
CO" T--
~+~
co o
00 CO
i
~+~
S?
CM ,
~
CO o
CM -S
~+~
t~- t^
+"
1
>,
co
£
S i
Si i
? .s
rimarily Purcl
Mean
Confidence
Q.
co en
t 0)
CO CM
m T-
r-"
oo
CM CO
O> Tt
o>
i^."
T- m
CXI CM
r~
co"
co •>!•
CM CM
• CM"
CM CM
+
ss
in CM
oo w
~
Is- 0
^ CM
CO 0
r--" t--
,
~f~
CO LO
f~ CO
CD CO
CM" i
+
S5S
CD ^
i
+
CD LO
CD CM
+"
CM "7
+
r- co
+~
!
3
.c
II Systems
Mean
Confidence
<
•* CD
•* 0
^
CO O
oo m
CM
CM"
CO
O> CO
$ 2
•*"
C5> CO
CD O
O> r-
CD"
Cf> O
0 •*
CM *-
CM"
OO T-
oo m
IT) T-
CO T-
CM O
T- CM
0> 1^
»- •*
« s
Median
Observations
te
gy
1
E
'w
CD
CD
ii
E
o
T3
CD
D-
0.
e
•o
(O
CO
00
o
CO
CO
o
•
-------
ible 60
Expenses
wnership
ids of Dollars)
•!? — O c
1- to *•* ra
•s >• •>
O OQ 3
i- " o
£
H
£•
o
0)
1
c
o
!S
5
3
O.
&
O
O
'E
0)
co
a
n
d
if
a
h
c/
<
!r o
$ 0
> 0
0 o
o
U<
is
o 2
o 2
° 10
1 o
T- 0
g 9.
°. 0
o o
IO T-
1 o
T- 0
0 0_
°- o
O IO
' O
T- O
0 O
«•> o
n *-
' o
T- 0
SCO
CO
• 0
55
^
^ 10
3 tf)
1
5
a
K
,a
1
o
CM ir> -. h- r--
o ,i CM •*
•* r^ t^
T-" co" O
CO CM 00
T~ ,
~
§§ CO?
^r co p t-
OO T-" •*
~
O Y- CF> if
co o CD en
CD ;; •*
h-" T-" o co
CO CM ^ N;
~+~
co o\ o o>
CO , CO CO
+
10 o co co
t
+
5
6
&
to -~
III
*sl jl
— 0) U CD ^3
•§5 SO
a.
i
rivate Syste
Q.
IO •».
CO CO
~
O) CO
55
S»"
—
CD ??
TT CO
o in"
CM
+"
0 co
— °>
•* oo
h-" W
+
*f (O
CO c\i
CM T-"
!
~
rx
IO T-
oo
T-~
10 00
O> T—
•<*
CO OO
1^ CM
T CO
V- t^
CO T-
CM
T~
CO
Median
Observation:
•\f irj
g?*5
+
C) K
£i "^
CO 00
O) c\T
oi c\!
~
£ R
in CM
00 T."
^r
1^ 0
^- CM
CC 0
r^" w
_J_
CO 10
Ji CO
CO CO
CM i
+
°> 10
o> ^^
co ,-
+"
S ">
CD CM
~
CM1?
+
r^ co
1
~
"5
fe
s
.5
i 1
£§§
to So
<
^ o
•* 0
•^
CO O
OO IO
CM
CM"
00
O> CO
g?
•*r"
CO CO
CO O
O> T-
co"
co o
en • r^
*- •*
"S
"t"
tfl
Median
Observation:
tn
in
CD
s
CO
0
(D
p
8
CD
g>
5
o
CD
3
T3
o
o
o
o"
0
s
CD 0
> a
o o
*j— tD
o to
in cy>
S a>
c co
CD -
Q. C^
less that has ex
> for this categoi
— If*
L- d)
o n
o S
0 {D
X
O> CD
•> TJ
C_ d)
S-n
3
E "
CD —
w s=
"Ti
O (D
- ^ =
CO Q. 0
d < CD
. . CO
s s
CD O
Q Z
-------
to
0)
10
c
S.
X
m
c
ra
9>
c
2"
8.S-1
eporting PC
ly Ownersh
isands of D
U. "J r
o
£ H,
>v
(O
£
Fotal Expenses
£.
o
en
O
c
0
•2
a
o
a.
tu
o
a>
I
m
8
55
Jr 0
2 °
> Q
O o
O
in
1 o
if
° 10
1 o
T- 0
0 0
°. 0
0 0
IO T-
• 0
T- O
0 0
0 10
1 0
58
«•> 0
n *-
0 CO
CO
• o
T- O
0 U>
o m
0 M
T- ^>
^.
O
$
i
wnership 1
O
^_ (0
as 5
*~ :£-
O -." ^
_!_
in Y- o>
CO CM CO
to •<- •*
K" w to
-J-
m en °>
en o o)
to TJ- CM
CM" _i_ CM"
§£8
h- -T-: to
oo o> o
(0 C\j ^
"T"
to o o
(O v- CO
+~
U> -r- CO
.*._
+
—
1
M S
E g
S c
die
w c c ra
- Sol
1I°I
QL
r-
o
oo
^
C3>
to
T_
en
CM
CM
CO
1
S
to
c
Observatio
T- Q 0
CM (o «-
CM ,
~
?2S
"* o °9.
^ to" o
g^£
^" fy\ if!
^3 iri ^
_!_
O •«- CD
^Ic^
r-~ t~" r-
-£-
S§!8
O ^-- 00
co" »-" v-"
T"
to co 10
co K en
Is- CM •*
T"
S ^ J?
oo 10 i^
~^~
o irj co
~+~
76
1
m H
E Q>
O O
OT 1 c
•2 ra § %
a
o
CO
CO
en
*-
'$•
co
CM
co
Ti-
tO
c
Observatio
h- CM ^~
to o "*
oo ;;
-£•
o en co
CM ri to
CO co CM
o> CM" CM
CM CM 00
4-
•^ 10 °>
tO CM tJ)
oo" w •*(•"
-I-
oo OQ 00
o .£ oo
(-- v- m
O (0 Op
h- CM CM
~
0 10 CO
~
"S
!c
8
i 1
I § c
•JJ c c ,ra
<
i^.
en
o
in
CM
CO
0
o
00
CO
oo
10
en
o
m
c
Observatio
^
o
to
u>
to
c
CD
Q.
X
0)
to
o
Q.
T3
C
ra
to
c
CD
£
CO
cl
£
'o
XI
1
o
0.
£
o
.c
ports systems w
£
CD
CO -g
O 1-
. . in
'm ~o
Q Z
—
in
o
Sf
5 S
O io
si
3 system serving
3d, expenses for
I =
Q. <=
< ^
CO
en
-------
10
.a>
o 'Z
O ti>
S *>
3 §
J- X
O LU
S B 2
>, 2 ,2
J"0 *-5
. g
S m -^
£ §
m %
r-
n
= • o
Q. ^ o
0 0 0
Q. o_ 0-
0) O It)
O T-
d> ,
E o o
» « °
(0
• 0
^~ O
o <*^
• o
58
0 &
^7 CW
t.
o
wnership Type
0
T- 0,
C>J CM
10 _i_
+
O"> o
fsi ^
•^~
^ O)
CO 2
T
N: co
h- csj
+
•<- CM
O CO
f 1
~
q to
^
•* CO
S T
+
CD N.
CO ^
^.
CM CM
SG
t
+
jblic Systems
Employee Expenses
Confidence interval
o_
^~ o
0 CO
10 ,
+
^ ^
§ 5
m co
to" "°
^
CO CO
10 co
^- ,
+
o o
Tj IO
^" ,
-__
IT
~
(N ^
CM ^
+
IO to
CM r-^
^-
en s.
5!S*
*
+
CO
0>
m
1
Routine Operating Exf
Confidence interval
10 o> CO OS
CD co O ^j-
^ *
t^ ~^ f^ to
o co m csi
CM , ,
T" ~
IO CM CO rj.
o 10 m' csi
i i
~ ~+~
CO •q. CO cj)
IO ^j. IO csi
T" ~
CM T- •* CO
CM CO CO CSJ
T" T"
•*- •» i^o>
o co m M-'
i i
+ +
CM to t-- •»
CD' N: •* N;
_>_ ~
+
io
Debt Service Expense
Confidence interval
Other Expenses:
Confidence interval
CO XT
tti io
CM ,
+
S! w
•£-
r- co
csi "
*
CSI N.
f^ CO
-t-
•^ N.
^ -r~
CO v-
i
+
O> -q.
+~
CO 10
ss
-£-
CD CO
CM °*
.f.
N- o
Esif
_^.
rivate Systems
Employee Expenses
Confidence interval
0.
q ^ co ^
CO 10 CM \~
to , ,
T" ~
CM 10 h- ,-
S? S CM °°
-L. T"
CM to "^ O
CO O CO CO
CO ,i ,
O 10 CM CO
CM csi CO to
•
CM oi oS to
~+~ T"
CO o> -
CO M-
~
q to
+
o> to
o w
T"
*~. "**
"" 'T
-4.
Other Expenses:
Confidence interval
ip o
^3 rr\
C° _,_
+
CM o
CM ^
^~
10
CO S
•jr
CD csi
N- csj
+
tp csi
O co
CO ,
"^~
CO \f
^
CO v-
csi •«;
CO ,
+
CO to
CM" "*>
^
q ^
CM ^
-|-
II Systems
Employee Expenses
Confidence interval
<
f~ co
CO rv^
10 i
+
>n o
Oco
CO CO
co^
~
T Is-
f^- co
+
co r^
•^ -^j-
^" ,
"^~
tp 10
+~
CO CM
s t
+
T co
+"
in o
OCO
-f.
to
c
S
Routine Operating EXF
Confidence interval
•* CO CM CM
CO W CO ».
+" T"
^ o> 10 10
J^ »-' CO csi
^- -£-
CO to ^ CD csi
*~ 1 ,
If" "^~
CO O O to
•* M-' l° Csi
T" +~
CO Co ^ ^»
•^ csj co csi
T~ 1 t
+ +
CO o co r».
in csi csi csi
T" +~
.
£ i"
tt)
"D HJ
° 8
S CO
i "§
» "O
c m
>v S
2 c
— x"
•§ *
<= ><
E S
O O LU
js 1
CD O
O Z
-------
•o
a>
o
3
il
5 g
8*5 o
w _
*0 a) TS
S g 20
CD O. 3 "D
I— 75 >, §
^\ CO O
H EjE
*fc
>> a
O M
a
~5
Q
£•
o
ra
re
o
0
H
a
o
a
0)
o
'E
0>
co
E
m
CO
(A
8
N
CO
<
*- O
> §
o§
o
u>
• 0
5 g
0 -
|i
1 o
T- O
0 0
° o"
0 0
u> T-
1 o
T- 0
0 0
0 o
o~ v>
*~
1 o
5§
«"» d
n T-
1 o
T- 0
o eo
10 eo"
1 o
T- O
0 K>
o m
SJJ
o
^
a
•5
3
O
co
a
CL
5g
to o
^
CO CM
-- o
1
+
CM o
p co
CM 0
O CO
*~ o
*
TT O
O CO
CM o
_^
+
m CM
O> TJ-
T~ o
~
CM o
+~
co o
m Co
CM 0
~
<0 co
*~t o>
•v~
eo
E
3
co
k.
CB to
1 &
•o •§
c -5
1 i
|| cS
CL
o oo
0 0
CM "*
a> co
*-
m co
a> co
T—
h- tO
CO CM
O CO
O> ••»
cn o
in CD
5cg
CM
•t- O>
CT> l~-
•* CM
CM 1^
CM
ian
srvations
TJ 3>
ID .£)
50
I"*- r^
CO o
T"
CO S
*- o
'
+
ss
^- o
-£-
p~ CO
CD CM
^- o
T"
!T 10
O CM
CM 0
_!_
"*"
CM CO
CM 00
CO o
^~
O) CM
•^ co
T~
oo J
CO W
i
~+~
CM to
N: o
CM »-:
T"
co
E
to
co
S -
(0 CD
! -
8 .£
1 1
LI
n a ,°
g 0) O
£
CM t-
Tt tO
CM "*
0) CM
in co
T~
^- CO
•* o
T—
in m
O CM
en h-
Tt 05
CO CO
Csi
in co
0> CD
CM
CM •*
CO
in o
T- CO
oo
ian
arvations
is
IT CO
•* 10
•* o
~*
m oo
O T~
CM o
'
+
CM CO
0 CM
CM o
:j-
co ,-.
to 10
CM o
*
O CM
cn ^j-
to -^
_J_
"*"
CM 10
m co
CM 0
•f-
g^2
m" w
+"
" t
!
T"
CM 0
~
n
E
S
n
CO
S
5 __
1 |
1 |
i i
~ c C
CB IB P
Ed} N-)
P
CL
oo oo
Tj- O>
CO T~
tD CM
0 T-
CM
00 CO
o M-
CM
CM •*
m CM
CM
CM CM
CO
CO CM
CO CM
CM
CM •*
h- 0
CO CM
to
1^ CD
CM
CO
C
o
g'l
TJ (0
CD -Q
5 0
CO -q-
CM -q-
CO o
^
O T—
r*-- T—
•^ o
-1—
+
CM co
OO t-
*~ O
•^~
^ o
O) CM
^ O
~
CO (O
CM 0
-L-
"•
CM rj-
CO CO
CM o
^~
co -^
CD 10
CO o
~
o to
O> ^)-
CM o
+"
OO \J-
O 00
"V"
^5
1
.c
8
I.I
>» « f\
CO « O
<
T- CO
•* I--
CM °-
T- O
N- m
T~
CO CM
cn oo
^~
to o
T- O>
T- CO
c\i T~
0 T-
o m
CM '"
o o>
o> o>
CM ^~
m to
CM •*
CM ''"
in co
CM O
ian
srvations
"O CO
5 O
to
c:
"ca
cn
c
1
o
*- o>
CD if>
a. •
co ta
£ to
r a>
S c
CO CD
CD CL
0) g
M
SE
9- CD
X Ji;
CD £
's ^
CD 3
Q. 0
0 C
® Zl
Q. —
O
O jj
V CO
« E
CM ^
£ *
Ss
c
» 0
fc -t
t/) CD
>. Q.
to D.
CD °
IE
8 S
ۥ2
§ «
1
O 1- D.
. . io
B S
CB O
Q Z
m
o
-------
Table 64
CD
O
3
1
Q.
V>
CD °-!
rotal Expens
By Ownershi
housand Gal
1
C
™
o
Q
O
D)
£
CO
O
o
to
—
3
Q.
O
O.
0)
0
'E
0)
co
E
j>
(A
c/>
®
CO
5
ll
o
10
^- °
0 °.
go
1 0
T- 0
0 0
0 o
0 0
IO *~
1 o
T- O
0 0_
^^. o
0 U>
1 o
T- O
0 0
<•» o"
m T-
1 o
T- O
S £
1 o
T- O
O IO
O M
°8j
o
$
Ownership Typ
CO U)
O CO
CO o
T"
S£
^~ o
T"
0 N.
CO T-
*- o
'
+
CO ^
CO CM
*- o
_!_
^
CM o
i
+
CM N.
CM CO
CM 0
1
+
m v-
CO (o
co o
CO CO
r-- co
CD CD
m ^
O CM
csi ^~
h- CO
O> CO
CD *-
I-- t--
csi T~
co
CD CO
CO T—
csi
Median
Observations
CD |^
•* CO
CO o
T
CD T-
CO CM
»- o
+
co in
O CO
CM 0
*
+
CM in
csi o
_L-
^ — t^
CO ID
csi o
1
+
S ^
Csi C3
'
+
S «">
O> CM
SK
CM o
_!_
+
O M-
•*' W
~+~
i
S
*> .s
Private System
Mean
Confidence
m h-
co m
CM' N
in co
CD
csi
O) CO
O T-
CM
^- CO
csi
o in
CO T-
csi
CM CO
csi
in oo
CM CM
CO
o t^
o t^
csi
CM CO
csi
Median
Observations
CD
CM
CO
0
CM
CO
o>
CO
CM
CO
csi
CO
CD
CO
o
0>
csi
co
p
S
V> c:
>> CO
eo m
<
^}-
0
T"
o
~
CO
o
1
+
o
CM
o
1
cr»
J
1
Xf
f>
d
1
+
•^l-
10
T"
*
_J_
+
?
1
'S
CD
^
Confidence
O) CO
CO I--
CM P.
T- 0
h-; IO
CO CSI
CD 00
r- T-
co o
T- CD
T- CO
csi ^
0 T-
o m
csi '"
csi ^~
m co
CS| •*
csi T~
•* co
CSI O
csi *~
Median
Observations
c
jO
"to
0)
T3
C <0
= =
CO O
•o tT
CO Q-
o °°
£, ^
0> CD
||
I 1
•- 0)
3 c
CO CO
£CD
r-
Ol C
C/> <1>
CD .C
tn •*-•
5 T3
x"S
11
CD
Q. 0>
o IS
ing 25-100 pe
om the estimi
CD "O
in CD
» &
£ "D
>•» ®
t/) «
£ 9
t- CO -T3
°°- •§ 8
O D. §
C3 ^ Q.
(A
"5 "5
Q Z
CD
a>
-------
S
CO
c
CD
I
Q > ^
J>
2£
S
o
o
0
Dl
4)
o
.
Q
"5
a
0
a.
O>
o
'E
CD
,
(A
CD
N
55
S§
o °-
0
u>
• o
1°
_r 0
|s
1 o
T- O
0 0
<=> o"
0 0
U> T-
1 o
T- O
0 0
=t 0
0 U>
*~
1 0
58
't o
1 o
T~ c^
Sn
n
' 0
0 S
0 0>
?8j
o
SL
_Q
1
O
^ CO O> T— ^" O> N- ^J° CM 1 CT> T— CD to
Q CO ^
i i
+ +
CMC, 0> 0 - CVJ CO N- CM,- _ ,- CM ,f -*
T-0 OT-^^^ ^-Q ^^^^^
"+~ T~
CM
•^0 OT^^T^T- ^£ WQ W^^T^T^
• I
+ +
COcM O> O T— CM CD ^— COcsj O^— CM^T—
t-Q O T— T— T— T- ^~O T-T— T— T— CM
1 I
+ +
•r^o COO-^Cv]^ ^ CNJCVJ ppCNJCOCO
T— Qj Of— T— T— T— ^_ ^"O T— T— T— T— T—
1 1
+ +
c\Jt- o>OT-cot^ eg coo> COT-CMCOCM
T-Q OT-T-X-T- ^2 ''"O O * r- r- f- CSJ
1 1
COCM 00 0> ^ Tf CO 0> r- ^ O> O r- CM CO
^ Ci 0 0 ^ ^ CXJ £ ^o 0^^^-^
+~ T"
^~CM CD cj> T— ^ CD ^ ify co co co T— co u>
"- 0 ° ° ^~ ^ "~ *° ^ ^ 0 0 ^ ^ CM
^ *"
T-Q OOCMN-^^ T- U> c^> U>OCMOOOO
+" 4~
(0 CD
1 cB
« o -^ MM&MM |°^ MMM2M
£•08 ccccc c *ti8 'cc^'c'c
SKS 88888 S |.Kg 88888
^> (U a m m m m m ® 2* ® "O *T m m !n *~
.ago Hill | |S§ Hill
^> o»ootoo :S >> omoioo
(f ol ^~
O CO^~ COO)^~^"CM
5 ---o dd^-'-^cM
i
+
co rg0 CDO^CMCD
T- 0 0 -^ ^ •r- r-
~+~
CO
••" T-: 0 0 r-' v-' r- v^
'
+
O) CO-r— OOr-CMCD
T— CO O i— ^- ^— T-
i
+
-* T-. 0 COOT^CM-*
"*~ O O T— T— T— T—
_l_
+
(O CNiT- OOt-COOO
^ - 0 0 W W W W
>_
CD COv- COO-^COO
" ^^ o^^^^i
+
co cov- r^o>T-mo
•* -0 OO^^CM
T"
Stf> CO >O O CM CO O
.-•••-
+"
-_
1
O '~ CO d>
o
00
0
T-
co
^_
00
•*
in
2
^
^~
N.
Observations
10
co
CO
c
C .CO T3
CD Q. CD
Us
O O "D
•B° S
5 c 5
S g CD
CD CO i?
• B. "> co
2 o o> a.
E CD E
rt\ r^ *—
Oj
«i "• 0)
rn r~» *^
» o
O)
-S
OT *- tf)
^> C O
OT *" •—
5 -i =8
CD £ ™
.Q
3
*
-S? £
•2 E -E
O .c „,
I "§ B
•*-* n CO
o o: o ^t o
Q Z
-------
in
0)
in
1
III
5
O
p™1
8 2 g
o> S >
— 3 i_
•° C >
£ I01
_
1
*T
O
o
i
^
o
D)
*2
n
O
c
J5
Q.
o
D.
u> n
_
^3 in
? 0
o
wnership Type
O
•* CM
*- o
_!_
+
T- T~
*- o
•£-
10 co
*~ o
+
10 XJ.
*- Ci
*"
•g
^
«
rivate Systems
Average Ratio
Confidence in
Q.
co en T- co 10
CD C> T- T- CM
en o T- CM co
C3 T-' T^ T^ T-'
co oo »- co in
o ci ^ W CM
10 O> CM O O
O O T- CM CO
1 0th Percentile
25th Percentile
50th Percentile
75th Percentile
90th Percentile
CO
CM
in
CM
T—
^
CM
co
Observations
CM
*-
(O
ci
CD
d
CO
••--
in
ncillary System
Average Ratio
<
co r>- o
o o <-
_!_
+
O CD CD
0 CD 0
^-
i^ CD CD
0 CD O
T~
•q. p p
o »- •«-
T"
"CD
1
Confidence in
10th Percentile
25th Percentile
i~» CO O h^ ^" CM CD O5 T~ f^- IO 1O
^ ^ IO ^ — Q O C5 *~~ ^ CM
^~ '
+
CDCDCD *- I--T- OOT-CMCO CD
0 Ci Ci -^ o 0 -r-1 ^ T-' T-' ^
^-
CDf^-h~ CM lOco COCOT— COlO CO
CJCici --QCiOT-T-CM''*'
T"
•fcoo •* ioco mocMcoo co
-r^^lO -r-'odT-^^^CO^
+"
"<6
» |
"c 'c: "c c 5j»"ajS "c'c'c'c'c c
S. S. S. | § rj'c £ a £ £ £ |
o K o 5 tt->^ oinoioo S
ioi^cj> O =< T-CMIOI^O) O
<
0)
3
CD
>
£
CD
£
">
o
o
CM
O)
»
I
T— >»
CO *>
d £•
- .2
^ c
O CO
d 6
08 'S
Q Z
CD
CD
E
CD
£
E
2
CD
§•
-J5
1
(O
CO
«
**-
o
"5
CO
c
CD
g-
-fr
T3
T3
T3
.F=
»*-
.2
^
CD "S
.£ .«2
||
"D c
C CB CO
~ 15 CD
5 £
I i
~m ""
-------
In
O
y _
— m «.
g§ o
§ "S "0
< » S
*^ ?! ™
CD T3 " «
(0 d) ^ ^~
X 18 —
o £ yt
Q. 'E in
£0-0
™* 2^ .
5"~
5 w
1 ""
3
Z
S
CO
"*
§
o§
o
u>
T^ °
o ^
0 §
o S
o
O)
» • 0
° o °-
.11 U> ^~
I
Q. . o
0 T- 0
Q. O O
a> P. o
o o in
'E"~
o>
w • o
jr o o
2 o
CO
1 0
T- O
S£
" Ff
1 O
5§
o v>
?8
o
e of Water
o
co
ro
CL
CD CO
•»' o
'
+
SS
CO T~
1
CO ^
CO CO
~
^- N.
5! 52
CM
CD ?
CM »C
^~
IO ^
IO co
r- CM
+
CO CO
CD U
,
~
O> ^j-
^~ _i.
8»
CM M-
~
10 CO
-j-
Labor Costs
interval
Mean Annual
Confidence
10 o
1
+
OO CNJ
O ^
'
CD CM
T— *~
T- ^
^
CD M-
10 cb
00 co
1
+
CM 10
00 ^
•f-
CD CO
co o
^—
co o
co oS
l
T"
CM CO
•^ ci
•^"
1
ice Water Systc
r of Employees
interval
rimarily Surf:
Mean Numbe
Confidence
Q.
CO S
,
T"
CO <0
£> co
CO CM
1
CM K.
CO 10
10" i
+
10 10
CD >-
00 TJ.
+
CO rj-
O> T—
f- T-
,
~
0 o>
M T
t^ CO
+•
8^
+
CO CO
•±-
Labor Costs
interval
Mean Annual
Confidence
CM 0)
00 CM
1
+
CD CO
^" CO
CO T-
1
+
p CM
[x_ ^~~
,
~
CO (O
% *
s a
~
co w
~
CM o
•^~
*- O
1
"+"
* *
u>
E
1
based Water Sj
r of Employees
interval
rimarily Pure
Mean Numbe
Confidence
Q.
co °°
~T~
CO •«.
CO co
•* CM
_j
? S
r— 10
CO i
+
CO to
s?
+
00 10
CO CM
,
+~
oo o
CO lf>
4"
sg
+
•* co
~
* HI
Labor Costs
interval
Mean Annual
Confidence
CO 0)
00 o
1
+
CD CO
+
co n-
00 2
+"
0 o>
CM 10
•* CO
1
+
•* co
CD O
:j-
O> rj.
CM o
~j~~
10 10
CM ^
(
~
CO CO
»- o
~+~
r of Employees
interval
II Systems
Mean Numbe
Confidence
<.
s°
~T~
S o
o" -q.-
co ,
CO 10
CJ> CO
CO" _i_
+
O CO
SRi
CO in
00 o,
CD ,
•^-
CD CO
^~
CDCO
CM CO
~+
10 CM
-^-
Labor Costs
interval
Mean Annual
Confidence
0}
Q.
O
0>
Q-
0
o
ro
SL
tn
£
1 8
Q. *^
0 T3
Q- C
CO ^
£ J2
w S
'K S3,
>* O)
0) **~
II
TJ iS
SCO
CO
ro
.C tn
" S>
o_ "
•2 g
"O "O
13 o
^ .E
8 3
. •§ 8
rf^ •
? v 0
co o xa
C» ? 3
S 1
CO O
Q Z
£
ro
i
V)
s
CD
CO
c
g
"co
"5
o
S
CD
£
O
TD
Q)
Q.
0
T3
^
CD
m
"S m
9- c
o o
a> S
o. ro
83
_.
CM CD
D)£
Is
«) -K
CO ^
ȣ
E a>
f.*
l_ ~~
B'S
CO £
§ ^
& en
0> Jjj
> ^
^ ^
P- O
0>
o>
-------
to
o
o
o
^
JS
3
c _
c 5-
»<^
(0 TJ C
• c c
•° n S
2 5 O
• -^
>» "
O CD
'a
UJ
•5
5
E
3
Z
e
_fO
o
a
•5
T(
I
3
O
C.
c
tn
O
3
""*
^
0
B)
*8
ro
O
c
.2
5
a.
o
a.
a>
o
o>
0)
1
M
g
N
55
^
Jr 0
|§
0
in
1 o
It
o g
1 o
T- O
0 0_
o o
in T-
1 o
T- O
0 0
°_ 0
o u>
1 o
T- 0
0 0
£o
1 o
T- O
SCO
n
1 o
T- O
0 U>
gn
m
T- CD
_i
O
wnership Type
0
T- .q.
d T.
"" •£-
00 0,
00 oo
g°?
m. co
+~
T- [«
CO CO
~
N- CM
i
+
•* Ol
CD o
CO rj.
CM 0
T"
CD oo
»- o
"+~
CD N.
•*-' ci
1
CO
o"
CL
ublic Systems
Mean Number of Em
Confidence interva,
Q.
CD 00
CO 'f
~^~
CD 0)
CO
+
CM (o
o> oo
CO _i_
+
CO f\l
CO CM
-__
+
CO -st-
O Q)
!
__
s|
COCO
CM to
T"
CO to
+
Annual Labor Costs
Confidence interva*
*-. CM
•£-
•*
•* oo
CM 16
r~- CM
~
00 |^
CO ^
Tp~
CD 10
CM CO
1
CD -^
CO •*)•
CO W
+
10 CM
CO CO
CO CO
*- ci
1
0>
m
£?
Q. —
rivate Systems
Mean Number of Em
Confidence interva
CL
S? <°
00 CM
~
CO (O
s§
^ s
_L
+
§ 9
oo s.
_L
^ Ol
CC' o
:j-
cr T-
cc 06
' •
CC ^J.
oc 2
~
0 o>
CM ,0
~+~
•*t <*>
1
+
•* co
co o
4-
O) -q.
CM ci
~
tO 10
CM W
CO CO
•^ Ci
1
CO
i
o'
D. —
II Systems
Mean Number of Em
Confidence interva
<
CD o
CN °°
•T"
Pi
CO
+~
CO •q.
CO lo
en oo
CO i
+
O> CO
00 CM
T~ -^~
"*"
CO ^
CO CO
COCO
T
CM CO
10 CM
+
Annual Labor Costs
Confidence interva
to
o
CO
c
8
73
c.
CO
c
0>
JD
0 g
-o °
to .!5
o "^
"m O
Is
CO O
t^-
cS
ISL
s °
co a.
^
<2 E
E ">
to 2
H
^3 3
3 C
°- C
O CO
p E
o
o
-------
£
(V
o>
IO
4^
to
ra
Q.
0)
c
to
*i
0>
E __
Q- §
C i^.
— O
sp
o
'E
a>
co
E
S
in
>»
CO
s
l§
> 0
0 o
0
u>
,1 0
o^
g g
1 o
5§
O Q
o" o
IO T-
1 o
T- 0
0 0
°. 0
0 10
^~
1 o
T- 0
0 0
CO *-
1 o
T- 0
O CO
<° n
' O
T- O
0 IO
o m
t.
O
eofWatei
0
co
«
•c
Q.
*~ „.
5?
CO [^
U> 10
CO T-
'
+
<0 - 10
CM o
m v-
~^-
tO CM
T"
F Systems
interval
II Systems
Percentage o
Confidence
<
co
CM_
CO
in
CM
o
CM
t
s
o
T-
CM
E
S
Observations
CO
ci
a
n
O
-------
£
Hi
V
^
m
»
IB
0.
0>
£
c
c
CB
0
>
o
a
I Q.
_ "m S
°f e
« ml
A O %
£00
"^^ ^s
i"
?
c
!£
m
S
»
s
n
>»
co
"5
§>
1
o
o
a.
o
O)
»
+rf
ra
O
c
0
i
^
a
0
a.
0>
.0
o
CO
E
2
w
>,
eo
1
>
O
1
I
d
o
o
o
o"
in
0
o
0
^
0
P)
<">
1
o
in
i
^
Q
^
O
O
T"
m
a
CO
<
o
o
o
o"
o
m
0
o
o
o
o
in
o
o
o
o
0
o
o
o
c>
in
0
o
0
o"
o
o
to
p>
0
s
w
at
2
o
wnership Type
O
ubllc Systems
D.
CM o
S'j:
~
CO 0)
CM co
O> ,
~
Is- C\J
S o
CO v-
~
CO co
s°?
+
to t^
•
J 10
h- t-.
'
+
in co
T~ to
co ?:
+~
h* CO
S^
1
+
•* -f-
m' CM
tO CM
+~
o> o
in w
m CM
+"
"* tv-
2 °i
^ T-
+~
1
c
»
Mean
Confidence in
t^.
s
to
o
to
CO
CM
CM
CM
D!
CO
in
o>
Observations
CO
to
in
•r-
h-
•*
CM
to
to
CO
CO
O>
CM
00
• «
N~ ff)
+~
MS
«5 CO
T"
£8
T"
^s
O> CM
~
CO f^~
Is- ^
+~
T_
-------
f2
IS
0)
>-
IO
4-* ^
* 'n —
^2
£ ^ .2
£OQ
C | 3
"5 °- t
•~ ^*»
ra
O
B
£
r-
£•
o
D)
B
o
c
o
I
Q.
O
'E
0)
CO
E
ff
tf)
3)
i
CO
5
•j- 0
$ 0
> 0
O o
o
10
•r- °
O ^
g"§
° u>
' 0
T- O
0 0
° 0
o o
IO »-
1 o
T- O
0 0
°. 0
0 U>
1 0
T- O
0 0
n. o
«o »-
• 0
T- 0
o CM
,
+"
en co
00 2.
^
en S
i
+
CO t
T"
m
^
co
rimarily Ground Wai
Mean
Confidence interva
Q.
in r-
m CT>
CM
o co
en
en
CM"
CO
co m
en co
CO
•*
CM CO
00 CM
r-
co
O 00
O CO
r—
CM
0 CD
m TJ-
m
85
T— T-
CM in
en o>
Median
Observations
CM to
•*t in
CD" v-"
~^~
fe^
T— Qj
g o>"
CO ,
+
OO Xf
00 Co
£
h* ^j-
CM CO
^ co"
•^~
•* co
Sfe
CO i
co n-
CO hi
y-T ,
T"
ss
00 ^
"+"
•"- O
OO CO
CO CM
-L-
H?
+
m
>.
co
1_
rimarily Surface Wa
Mean
Confidence interva
Q.
en co
O i-
co •<»•
0 CO
O CO
in
m"
CO
T—
co en
co en
en
CM
m en
co_
CO
en h~
CM"
en m
T- m
en
0 0
I-- in
CO
o co
•* co
*-
r-- co
T- CM
Median
Observations
<0 CO
§8
co" T-"
~
JTCO
m !o
CM" CM
CM ,
+
CD
CO <0
o" w"
CM ,
O Y-
to" 10"
^r
!-•_ CO
CO Co"
'
+
8 £:
en (o
,
~
COg
CO ^
"^~
o> o
CN CTV
CM CM
:^-
CD CM
+
S
tft
m
L.
ff
i -
rimarily Purchased '
Mean
Confidence interva
D.
O CD
in in
CM •<-
CO CM
CO *-
en
co"
CO
t-- en
r-- co
o
in
Sm
CM
co"
r- co
CM CM
CM
in CD
CM T-
co
O 1"-
0 CM
CM CM
CO CM
CM
Median
Observations
en co
*~ JL
CO TJ.
CO IO
en co"
CM ^f
~
5S
CO o
m" en
CO_ CO
T— f^
~
CO ^
CD CO
^- ^
_!_
+
O> in
•^ CM
+"
CM T-
•° -L.
^" CO
^ to
+"
00 t
+
II Systems
Mean
Confidence interva
<
CO T-
cn h-
co
in m
T—
t--"
in
^~
en co
O T—
r-"
oo r-
co en
r^
CM T-
CM"
CM r-
m T—
en CD
OO CM
T— T—
CO T-
CM 0
^~
CD •*
1^
Median
Observations
^r
co
d
is
^
Q
CO
o
-------
e
re
o>
>-
u>
^
g _
Q. g
£ °-3
K~i£
-ill
•8 E s|
,™ ** o c
1- m *•' n
» >. m
> m 3
£ O
B t
Q.
n>
o
s
o
i-
ategory
o
c
0
I
a
o
Q.
O
'E
a>
co
|
(A
>i
co
»
fl>
N
co
5
•- 0
°t
il
is
• 0
T- O
0 0
°. 0
0 O
1 o
5§
°. 0
0 U>
1 o
5§
« 0
CO »-
1 0
T- O
O CO
U> ft)
• o
S3
wnership Type
O
^J.
T—
CO
00
r^
8"
CM
CM
CO
r-
CM
5>
*~
o
o>
^
CO
CM
^-T
in
co
CO
o>
CM
s
CM
ublic Systems
Mean
Q.
„
to
1
T"
to
CM
T"
o
to
CO
to"
CO
10
CM"
~+~
1O
to
CM"
S
oo
~
s
+
s
T"
00
CM
I
+
"§
«
Confidence in
0 CO
in h-
co to
T- CO
r-"
m
SO)
in
O> T-
m"
oo m
r~ oo
CD"
CM m
CO
CM
rivate Systems
Mean
o.
CM
^_
1
+
co"
in
T"
CO
o"
t—
CO
to"
T
CO
CO
CO
CO
to
,
"
CO
CM
-f-
CM
_j_
+
j:
,
f
"5
1
Confidence in
OO CO
T—
CO CO
O)
I
CM Tf
h-"
CM
0 CM
CM T-
to_
0>"
0 N.
O T-
CM"
o m
O T-
m
CO N-
CO
2fe
oo co
CO
Median
Observations
CO r\i
O> co
T~ |
+~
CO ^}.
CO ^
0>" co"
in •«.
CM ,
^
CD §
S J
in co
0> CB
00 CO
^~ CM"
T"
CO t-
^" to
to co
^ T
O> 10
T. CM
"+"
CM T-
+
58
1
+
cot
,
~+~
^
-ffi
II Systems
Mean
Confidence in
<
CD T-
o r-
CO
in in
m
o> co
co iv-
O T-
"
co h-
(D C3>
IC
CM CM
CO CO
CM i-
CNl"
CM r~-
m T-
0) CO
CO CM
co T-
CM O
0 Tt
Median
Observations
^
CO
d
1
o
-------
S
(0
0)
10
*rf ^_
Q. (0
1?
_ °
CO ~ g
h- W £l
Vf I_
«u
** m *^
UJ 2
« C
"5.^
n ^
o m
•s
>.
H
01
£
0>
a
X
UJ
«^
a
J^
iS
n
c
1
i
n
CO
•5
g
2
c
a>
Q.
0
O)
re
O
0
1
3
a
o
a.
0)
O
CO
co
E
S
10
co
to
co
^ o
? 0
0 o
o
IO
•f- °
0 °
|s
1 0
T- O
So
. 0
0 0
10 T-
1 o
0 0
o o~
o" 10
T-
1 o
0 0
1. o"
CO T-
' 0
^ o
o n
10 „-
1 0
0 10
o to
° 8
k.
0
1
0
d)
o
co
a
tx
o o m
r- o ••-
•t co
o o o
CO 00 O
o co o
00 t^ CD
CM •<»• o
CO CO •<*
CO t' CD
T- CD O>
t^ oo a>
r~- h- •*
CM •q- oo
in CD co
CD 0) 0
CM •* r~-
O O CM
in *- co
CO h-
m (^ i^
CM OO T-
TJ- CO
T- CO 1^
T- 0 0
co t
E
B
Jco
c
to ^
«J? I
* I
1 1
1 T!
rimarily Grou
Land
Water Source
Distribution ar
o.
O O CO l~-
^ 1 — CJ> f*,
CO CO T- ^
o T- o co
CO O> CO
CO ••» CO
O CM CD IO
csi o in "
^ oo ^r
O> T- CO CO
-' CD N- 00
cri ^ T-' «
lO Tj- Tj-
10 T- r>- co
oS co co ^
CO •* CM
o o in o>
•* o r^ •*
CO •* T-
CO T- OO "*
eo co T-
CO ••» O> O>
1^ 0 r- •*
CM CO T-
Treatment
Storage
Other
Observations
O CO
CO ^t
T- CM
r-- co
5f£
f- CO
00 CM
co in
o in
as
OO CO
2 CM
•* CM
CM CO
oo r-
CO CM
CM
CD CO
CO CM
0 CM
CM
0k
E
4>
CO
CO*
fe
5
8
rimarily Surfa
Land
Water Source
Q.
IO CO CO CO OO
CD T- co in ^
r- CD co CM ^
O O O CM CO
8 8 £ a! "
O ^~ CO OO O>
O) O) •* CO ^
OO CO CO CD
T- co o> o 05
s § s ? ^
f*-.. ^" (O tO ^"~
f\J If) QQ ^^ ^
co r- co co in
OO CO CO CM
CM m m co o
CM CO 00 O> ^
CD CO CM
CM IO r- CM OD
ST- 06 co •**
r^ co T^
1^ 0 T- t^ CO
CO IO CM CM
E
f
>v
co
C
.2
»
CO
s
T3
CO <2
If. |
Q K 53 O O
O> OO C«J (£> CO O <£>
CO CO
CD CNJ O CM OO T- CSJ
f2S|S^g"
co CD oo in co o o
co cri r-^ r- in f- n
t^ tf) T- n in *t ir>
S S S ^ r3 ? ^
CO CO O CM •* CM CO
CO CO CD CM CM CM ^
^~
•* co co co cn rr CD
CM co in CM
0 00 CO CM 0> 0> I--
CO ^ CM
O O CM CO O> T- CM
1^ CO
O O O O O O CM
o o o o o o
o
^~
•> 1
2 "CO
s >>
W CO
M? C
CO o
,2! co
5 E
s »
•) H
5 E
I | | c I
^ co "3 p fl> >
•§J§QHCo6O
Q.
CO CO OO
CO CM IO
CO CD
o •»»• o
CO •* O
CD CO O
^ CO ^
CO ^ O>
CO O T-
T-: co •*
CM t O>
CO ^ CD
co o r--
CM CO OO
m o> o
in t^ co
CM CO h-
CO IO CD
CO CO TJ-
CM r-
CO t~- T-
CM 00 CM
CO CO
T- •*• in
T- 0 O
CO ^t
E
s
CO
w
c
o
'co
.<£
CO
CO
\ —
T3
II Systems
Land
Water Source
Distribution ar
<
CD CO O T-
co co o> t;
CO CO T- °°
co r^- co ^~
m CM co *"
CO 1^ OO
1O CM CM CO
CM o CD t:
co r~- m ^
O> CO CD 1^-
W co csi ^
CD M" m
•* o> oo CM
co co ^r S2
m ^t co
•* oo CM r^-
O> CM CO 31
CO •* CM
CM h- CO CD
CD CO CM 5!
CM CO T- ^
CM T- r^ •<-
CM CD CO °
CO CO T- ^
^f co co ^"
CO CD CM ^
CM CO r-
Treatment
Storage
Other
Observations
4*
CO
ci
is
"5
0
-------
e
re
0)
in
t»
re
a
£
.£
(D
i
EL
UJ
1
o
0
£"
0>
••-» p
i !2
^ =
O o
0> Q
0 *-
So
o to
CO "O
fc* (0
re §
M
m
o
O)
i
CO
o
c
0
I
3
Q.
£
0)
o
0)
co
V)
co
10
8
to
« °
>g
°g
in
it
|s
1 O
£g
58
in T-
1 o
5g
°.o
o m
*~
1 o
5g
« 0
CO *-
si
"* ro
1 O
"- 9
o m
0 «
O (A
0
1
"5
0
|
O
co
i
OL
r»- T— ^ co CD co h-
O O CM 00 T- r- o>
T- CM •<}• CO CM CO CM
T- CD CM 00 CO O to
CO CO CO O CM CO
CM r~- I*- •* co co
CO T- CO O T- o
co in M- T- in
*- r-- oo to r- o m
Sco oo r-- o> to co
to •* CM co co
•* CO CM T- ^1-
CM m m T- m CM co
CO h- 00 O O> O CM
CM TJ- CM co r- co
CM" co" T-" T-"
r^- co CM i- o> o co
•* CM r- CD t^ to co
T- co r- CM en m
CM N- O i- i- 00 to
O O O> O CO 1 — ^~
^~ CO ^" ^" ^ CM
O r- to T- CM ^ O
in i^ r«- o> to m •*
CO T- T- t-
i^ T— o> in T— in
CM
T— co *~ o> co co o>
E
S
£/)
E w
& C
Tj O
CO* .«
* i
1 1
1 J
15? 1
5 g .2 = S
>, W -5 | ID ^
fe-oSSfflSSS
eEtutooosS
|JS§5K556o
Q.
0
^
h-
O>
t--
CO
?
^
^.
^
cn
co
o
vt
>.
co
a
$
S
3
co
•c -1
Q.
Sen
CO
CM CO
O CM
h- CO
co co
CM m
r~- co
h- CO
05
CM CO
T- in"
Is
T-" CM"
CM CM
5 S
B2
CM tO
T- m
T —
CM •*
E
H
>»
W
^
o
1
m
§
TJ
Water Source
Distribution a
o cb oo TJ-
CO C\l T-
co n co co
O CO CO CO
M- ci> r^
r~- t-- CM
O> CO CO O>
m CD o
CMC,.
CO CI3 CO
r--" T- v-"
1^5^
^
o T- to in
to ^- in in
^_"
CO CO CM O
O) CM CM
CM T- CO CO
CO T-
co cj T- co
*- .2
§03 5
^ C3> !_ £
00 O CO
CO" CM
CM co in
CM O
CO CO CM
T-" to O
r^ o> co
oo CD cn
•* o" m"
CM CO T-
r--"
in co in
CO tD
co"
0 0) CO
•* o
o o t~-
•r—
o o o
» |
Sz >*
« to
.5^ =
CO 0
S 8
I "E
> OT
"g i
<0 ~D
£ „ c
rimarily Pure
Land
Water Source
Distribution a
Q.
r~- •* co co
to" »-~ CM"
cn o o CM
co to to
to cn to
o>" co" CM"
CD cn o> CM
CM m cn
r-~ CM" •*"
CO CO CM CO
,_"
F^ CD CO CO
CO CO T-
co"
O O CM N-
CO CM 00 CM
CO 00 tD CM
O tO O CM
Treatment
Storage
Other
Observations
no cn oo co r— oo T-
^~ T~
m m T- -.j- Tt o T-
m o cj> oo - Tt tO O O> T-
•— CD CM CO 1^- CO
— CM CM tO in CO
.— co in o r~- co T-
,-KOCMCM,
r— •>* oo to •* co
'— to •* in T- CM
CM co r*- CM in co CM
to co o f- in co co
CO" !-" T-~
to to co cn to r^- t^-
° CO S K 5 CM ^
CI> CD OO O> CO OO ID
CO CM CO CM r-
co ^ o T- co o
CM •>- T- r-
^•COCM ^
E
j>
to
c
o
'(0
»
to
r—
m
T3
c
i Us I
o to "5 ® n> 5
•S w .g E D> ,_ C.
= 555Htooo
<
CO
CJ
1
O
-------
"5
0)
a
B fi
CO JJf
* ^
>- °
to £
m » 1-
S. ra £
> o
2 ^ w
£ 'S "J
c v o
U i* ^
UJ |<
flS*
Q. C
« -5* 3
Q CD LX.
u_ •*•
o o
a- 1
^ c
0
1
^
o
CI
1o
o
c
o
1
a
o
a.
a>
o
*>
C
CO
co
E
S
(0
co
4>
>
O
,
o
o
^
i
T—
O
o
0
u>
1
0
o
1
^r
n
n
•
i
i
^
^
o
?
tn
i^
to
o
o
Q
o
o
10
o
o
Q
o
o
in
0
o
o
§
^r~
O
O
0_
o"
o
o
Q
*~
0
o
co"
o
S
r^
cxi cxi T-' CD CM o> ^
i-~ ^r CD co m CM co
O CD CD CO CO CD
T- CO »- CM
-* co r- m N. m m
T— T— h- O) CO T— ^
T- CO CM t-
in cq CM in co p co
CM •* ^ "^ T-
CM CO CO CXI CO O> CO
^~ ^ Tf CM " *°
O> O CO CO O"> -^- CO
Cxi CD CO (D oS CO ^
V- CO T- ^~
r-^ CM cq T- T- T- o>
5 CM CM
T— in CM i^ T- ^ T—
h>^ in v- CO CD ^~ "^
CD •<-
O CO O) t CM CXI O)
•M- CM CM ^~
E
B
JOT
c
io .2
>. CO
co .«
S E
m c
^ CO
1 ^
l||i||l|
CL
T— t — CM CD T— co co mco
cxi^cxicocxir^^ cxiin
CMO>-*I^O>COCO coco
coin^T— cbf^^ T—CO
•* CM T-
COCOh-COincOCD CMO
T-^TrcbcMinoS01 cxicxi
T— CO CO -^
CMh^cM^T— coo mco
co ••*• CM
h-COOCOv-COT— COCO
in CM T-
omcor--r--coin coco
CM h-' cb T— h~ CM" "^ ^~cb
CO •<»• T-
cxiT-coocn^o oco
cDr^^orv-'co'10 oo
cocomi^t^cxtco oo
o o cxi co T— T— ^ oo"
CM CO T-
O-*C»CD^TI-CO oo
o'o'cM-r-cxicxi™ oo
CO CO
£
$ i
1 CO f
* .1 *
>> » i:
to .« 2
*- f£ rO
me *
5 co -o
i i l
III- ill
CO o -= c 73 CL o
— ^"^IS) L*-^" W
«"oB:l"cSSc5® «"pB
sS'—sssS pS"
CL CL
o m
cxi v^
CO «-
o •*
•* co
•* CM
m en
K CO
CO T-
co oS
CO
co m
CO ^~
CO CD
CM •*
CM T-
co o
0> t-
CD" o
0 0
CD O
E
t
.1
1
CO
Distribution and Ti
Treatment
•^J" CD CD
•^ CO J»
m co CM
co cxi *~
^~ T~
T- CM
cxi N- ^
CM co m
in in ^
O> CD CO
t^ CD" N
O) CO CD
CD CO ^
CM
CM t- N-
in o" ^
CM
O> T- CM
CO O ^~
CM
0 O CM
§°'
T-
Storage
Other
Observations
CM •<»•
CM co'
t-- CM
Cxi CD
in co
T-1 CO
T- CO
CO CO
CO CO
T- in
•* T-
CXI CO
CO N-
CM in
••t •*
•«t co
O CO
o co
II Systems
Land
Water Source
<
CO CM
i< cxi
•* CM
o> r--
CO -^
•* cxi
t~- CO
co oS
CO CM
p cxi
CO O
CO CM
O) Tj-
S£
O CO
co in
CO CM
o> o>
CO T-
t CM
o> o>
o o>
m CM
t CM
co co
•<»• CM
E
JD
10
:>,
co
§
8
E
s
CO
Distribution and Ti
Treatment
T- CO T-
cxi ^ £
T CO
T- m T-
iri CD ">
*~ T~
^- m co
f- OS J±
o t r-
cb o °>
CD 0 CM
o M- £
r~- m h-
iri in ^
CM in CD
co csi J!
CO CM T-
CD r— ^
T— *""
m co •*
d T- •"-
CM
Storage
Other
Observations
o>
c
TJ
C
o
0
cu
TJ
8
E
v>
'o
c
>^
CO
tn
0
c
a i
m. o
a o
. . to
'm 'o
Q Z
t--
o
-------
£
10
v>
1o
10
Q.
0)
c
If
o C
o *
3|
® >.
c
X
—
*•»
Q.
8
Q
5
£
o
O)
£
10
O
LU
_C
(A
•0
C
"fa
"5.
a
O
c
s
(0
>
_
M
i
w
•s
g,
•g
CD
£•
o
O)
£
o
o
«
3
D.
O
O.
0)
O
'£
O
(O
I
w
(O
i
55
*- o
«=f
^ o
O Q"
0
u>
• 0
o o
o -
§1
1 o
s§
° 0
0 0
U) T-
1 o
T- O
0 0
°. 0
0 U>
T-
1 0
T- O
0 0
•?. o
m *-
• 0
T- O
o «*>
*O ft)
' O
0 S
^5 M
O 0)
^ •?
o
wnershlp Type
O
0
CO
CO
o
CO
r-
oS
f*.
h-
m
CO
CO
in
CO
CO
CO
o>
cb
CM
o>
o>
Si
0>
0)
1
ublic Systems
Water Quality Imp
Q.
CO CM
CO CO
N- T-
CM OS
O> h-
CO O>
CO CO'
r-- co
0") h-
CO h-
h- CO
in TT
i-^ in
CO CO
CD 0
id oS
CD •*
O> CD
^•' co"
1^ CM
•* co
CO r^
CO
'co
Q.
ID
u.
o.
Replacement or M
System Expansior
o
CM
CO
^"
CJ>
CO
T~
CM
CO
O>
0
O
C3>
CO
O
CM
^
Observations
T- T- CM CO
m CM CM 2
•* 1— CO
0 O 0 CO
c> d o
o o o
T 1 T
T- T- *- CO
Is*- ^ l1^ "^
CO O> CO
T- O T- T-
•^ d -^ ^~
CO O CO
(O CO CO h-
•* W co ^~
m i— co
CO 't O> CO
[•*. m o *~
r- co co
CM O> O> N-
co eo co T~
co in i^
T- r-- CM co
S' id CM "*
I-- CM
r--; if) CD T-
O> ^ h*
CM 1^ ^
•2-5
^ Q.
(D 0>
> g^
rivate Systems
Water Quality Imp
Replacement or M
System Expansior
Observations
Q.
O)
CO
CO
^
oS
r-
co
oS
N.
^.
m
r-
co
CO
Tf
C3>
r-'
CM
CD
o
co
id
CO
-B
c
01
E
o>
1
II Systems
Water Quality Imp
<
r- T-
£?
^m
eo d
O) CO
CO CM
o> •*'
t-- CO
CM •*
cri id
co r-
a> T-
cd co
r- co
r— •
CD
CM
CO
0
CO
CM
m
o>
CM
Observations
^
c
o
0)
CO
(D
D.
X
tt)
~m
.•ti
Q.
s
1
CO
o
o_
r-
1
in
CD 3
CM in
S!
si
co 75
O £
to
Q Z
CO
o
-------
2
flj
(ft
c
Q.
x
I
5
^
^5
^
^*-
e
jj
"5
o
•5
in
c
IS
in
o
£
1—
£
o
Cl
a>
^^
fo
A ^
c i3
o —
>. E
m o
(0
(0
a>
a.
*
£
in
>
^
0)
c
~
11_
S
"5
(0
o
c1
o
O)
1
o
_
A
_
3
Q
O
O
'E
d>
(0
E
S
M
CO
in
S
55
0
o
0
cT
o
IO
0>
CD
• 0
2 ^
x* ^
g g
»-
1 o
T- 0
0 0
o o"
0 0
U> T-
1 0
T- O
0 0
°. d
0 10
T"
1 o
T- 0
0 0
CO o
co »-
1 o
T- 0
o
o »
? S
J
o
wnership Type
O
^
in
CO
r-
m
co"
co
CO
co"
§
,_r
r-
h-
0
S
CM
S
O
o>
1
g
m
>
ubllc Systems
Water Quality Impro
Q.
T— O>
CM *-
0 •*
~*~ ^~
5 CO
0> 0
CM" co"
en CM
0> T-
m in
r- co
co" T-~
CO •<}•
m co
N- CO
TT" in"
co M-
CM 0
t-~ CO"
CD CO
O 0
m CM
CO T-
c\l co
i^ i^
OO CM
'5
Q.
CD
O
Replacement or Maj
System Expansion
o
CM
co
*
en
CO
CM
00
0>
o
g
§
CM
^_
»~
Observations
CO •*— h- OO
GO -*t h- CO
cn to in co
in en o
co co T-
m" o T-"
•^ •* CO
r- r- o co
CO «- CO T-
oo in co
oo" r-" o"
CO CM Is- T-
oo m en •»—
O CO 00
m" CD" CM"
8m cr> F —
co co ^-
co cn ^~
,_r
m o h- co
co co •«}• T~
•<* CD r- i^
^coco ,-
CO O CM CO
(M CM m
CM 1^ CM t-
•<- co
ft
0) 0)
E M-
m »—
> 0
rivate Systems
Water Quality Impro
Replacement or Maj
System Expansion
Observations
Q.
CM
CM
CO
CM
CO
^-T
^-
CD
CM
CM
^j-"
1
CM"
co
r-
0
S
S
cn
CO
o
CD
E
>
II Systems
Water Quality Impro
<
o r~-
CO CO
CD CO
m co
f- CO
t o>
Cn Tf"
00 CM
*-
CO *-
•* r-
cn" i-~
*-
Cn *-
co ••-
O> T-
1-" in
T- ID
Cn OO
T- I--
T-" CM"
CD in
CO i-
•* co
CO CM
in h-
CM CO
co •*
in co
*~ T~
I
OL
0
Replacement or Maj
System Expansion
CO
o
00
^
"*
CM
m
T —
CO
o>
CD
CM
CO
O
CO
CM
0>
CM
Observations
^
—
o
{/)
$
c
CD
Q.
X
CD
75
'5.
8
>
'«>
0
Q.
CO
E
m S
CM »
CO >>
0 £
-------
5 Years
+J
n
CD
Q.
0>
£
c
8
5*
ii
pS
»
w
c
3-
I
a
m
O
b.
Q
'CB*
Category)
£
o
_c
•O
"7™ d)
fH
o> —
Jm
73
§
^r"-
ms
a
ID
o
mtage of '
S
£
I
lo
O
c
0
I
a
o
a.
09
O
'>
C
09
CO
E
tt
co
v>
S
co
^
•r 0
> §
0
in
1 o
5 §
|1
• 0
T- O
0 0
o 0-
in T-
1 o
T- 0
0 0
°. 0
o m
1 o
T- 0
0 0
«"» o"
O CO
c^ m
o m
0 «
r- »
o
Ownership Type
CO
t
m
CO
CD
m
"
en
CO
in
CO
oS
CM
CO
CM
O)
CM
p
c
CD
>
Public Systems
Water Quality Impro
CO T-
Tt CO
CO TJ-
•^ •*
in co
co •<*
o> in
in oo
CO •*
0 T-
d in
•* CO
"»' CD
CM CO
in h-
't in
•^ CM
CO 0
in cb
•* CO
o> o
CM CO
•* CM
CO CM
CM CO
t
(D
C£
o
Replacement or Maj
System Expansion
0
CM
co
^
^f
05
CO
CM
CO
05
O
O
O>
CO
O
CM
-
Observations
in CM CM co
o in •* °°
CM CO Tt ^
t in T- co
•^ o> cri
T- CM m
o> Tt h- co
co r^ co ^~
CM -t CM
co CM in ^~
CO -^ T-
1 — CO T— h-
OS CM OO ^~
»- CO •*
N- CO O CO
0) 0> W ^~
CM CM •*
m CD o> r-
o o cb '"
T- CM CO
CM 00 O CO
cb o> •* "°
•>- CO •*
00 • o
Private Systems
Water Quality Impro
Replacement or Maj
System Expansion
Observations
(D 1 — N>
T- CO •«!•
CO O> 00
CD T CO
T- CO ^>
CO CM CM
cb i^ cb
T- CO TC
CO CO O
f*- CO ^~
OO CM O
in in o>
T- CM in
I-- co r-
o> CM r^
CM •* CM
co o m
CD O> CO
T- CO •*
co co T-
cb CM T-
CM •>»• co
o r- co
cb o> •*
CM CO CO
~C Q.
CD (D
EO:
CD *-
> 0
All Systems
Water Quality Impro
Replacement or Maj
System Expansion
CO
o
CO
•o-
CM
in
CO
CD
CM
CO
o
CO
CM
in
CO
CM
Observations
O)
c
"D
0
0
to
o
0
o
m
c.
CO
E
m v>
CM m
« %
o ~
c\i ^
"? 0
0 0
£
o
c
CO
—
£
as
O>
0)
O)
O)
CO
(D
^
c
10
c
g
ID
c
CD
r-
C
0
D)
£
8 E
O CO
CD >*
^ C
O O
CD 73
CD *
.2 CO
i >
CD CO
0.73
reports th<
itage of fu
— 8
£ il
. . io
IS 'o
Q Z
-------
m
a>
in
lo
CO
a>
**
~
1
I
a.
(8
o
•s
4)
S
3
co
£
3
C
S
£
ID
Q
.S-o
e .c
||
*!
co
•5
1
(D
2
0)
a.
*"
^
o
O)
B
(0
o
c
o
JO
a
o
a.
a>
U
'E
a>
co
E
w
co
CO
N
co
^
^
® 0
oS
O
in
1 o
0 §
o S
*~
1 o
^ o
0 0_
°, o"
0 0
m T-
1 o
° o"
0 U)
' §
T- O
•"I o
1 0
•^ o
i^ CO
«> eo
1 g
T~ O
0 U>
O (A
® 0
_l
0
wnership Type
O
in eo T- ro
ro ro T — CM
CD *-
ro r~- T- T-
m m CM CM
ro CM
•^ co ro •^-
o ro CM r-^
h- oq in co
T— CO CO 00
CO
ro r— co ^~
ro t^ ^r CD
r- T-
ro -^ m CM
CO CO ^ — ^°
t~- T-
»- co ro co
O CM CO CM
CO •* 0 ^~
co ro •* co
CO T- T-
CO ^" CO O
CD •* CD O
CM T-
CO
IO
CD
C
CD
£
C
a>
E CA
||
CO ~t^
ublic Systems
Current Revenue
DWSRF Loans
DWSRF Principal
Other Governmel
Q.
oo in eo
^ ro co
CM T-
CM •«- O
co m CM
^
O CO CO
CD O CO
oo in i-~
CO C3 CM
•<- in
CD O T-
T^ CD ^
CM CO
o co in
co r^- co
CM CM
ro eo o
t~- CM CO
^ ^
co o o
CO 0 O
CM T-
m CM ro
eo co eo
0
•5
m co
O m
Other Governmei
Borrowing from P
Other
ro
in
CD
m
^
^
in
in
oo
CO
-
^
ro
,_
•r-
co
"*
*
Observations
CN T— in co co •** o
•*— o o co TJ- co co
CO T-
o o o o o o o
•t CD id d o co CD
CO CO CD
ro T- o co o CD p
CM ^t O in O O O
ro t- T- r--
CD o o eo co co o
co o o h- co in o
r- T- in
ro CM o co co CD -t
in co o CM co CD o
CO CM T- CO T-
m o eo •* CD T- co
r- o T- CM T- •<- ro
eo T- T- m
CD o o co oo T— ro
T- o o oo •*»• •* CD
^ *- CM
m o o co ro T- ro
eo o o o •* eo co
00
CM o T- eo o •* in
ro o o co o CM h-
CD
CD
I
'c o
CD "5
E M* CO CD
>• c t; *"
"^ CO CO Jg
IS —1 **s 5
CO — tz t? "C
CD CO yt o> O-
2 c <2 75 E E E
1 S S .i | g £
33.3 o: I > »
*l&fe??i.
2£t»tOO)CD2
c
0)
o
u.
(D "cj
E « <2 9
>• c "c OT
|g 23
fV^ . . . . >
(A e e c
0) (D a fe Q-
§ Jifiil
'•§ i^~lQ"c5oi>
o =oQQo6t§6
<
CM
in
CO
00
^"
ro
co
m
ro
00
CM
00
o
1^.
CM
CM
°
m
^
Observations
0
,_
0
E
^
CO
t5
c
>,
CO
CO
s
p
c
3
o
o
CD
£
CD
CD
£
?
^
O
CD
O
C
CO
a>
o
>N
TJ
*t-
.1
8
§
CO
0 1
CO ^
O e/>
. . CO
S £
CO O
Q -Z.
CO
CD
m
CO
Q.
0)
£
—
CO
CD
^
T3
c
(D
Q.
X
0)
"S
±i
Q.
CO
O
ro
c
"co
E
E
3
to
>,
CO
o
£L
l_
£
"E
"
CO
"D
C
^
'Q.
8
H-
co
3
g
•c
o
a.
S
m
^
-------
TJ
0>
E
o>
co
co
0>
0)
g
J3
•«-»
e!
ra c
> |
B> 0
U
l||
? u. i
£-50
5 o CD
a£
a 3
O M
§£
3 "5
O •-
80 I
'•^
3
•c
(A
Q
I
C
H
I?
U.
£•
O
O)
•S*
(S
o
o
J5
3
?•
Q)
_0
0)
CO
|
»
CO
(O
^
» °
oS
w o
o
in
,1 g
0 0
0 *J
X" o
O f*
o S
' 0
T- O
0 0
in T-
• 0
^ o
O 0
o r-T
0 S
*~
1 o
0 0
•"I o
rt *-
1 g
T- O
o n
W ,0
1 0
*— o
0 U>
T-
§8
to
"5
0
ership Type
,-
m
to
CM
m
to
r-
to
CO
d
m
00
to
m
,-
o
in
00
CM
in
CO
o
m
T-
CD
CO
0)
ic Systems
rrent Revenu
So
Q.
O CO 00 Of
I--" ^- h-" esi
•* o •<)• r-
•^ d T^ d
o> to en CM
CO O T- T-
t~- to CD o>
CO W ^- -^
CM CO O T-
co T- o in
0 T- Tf CO
co o T- in
CO CM CM 00
o - •* CM
CM
^" co o ^
Tf tD O f
CM T-
0)
c
0>
?
o
U.
*s
>« c: c
O. Q S
rf
— "c c
CO m m
VSRF Loans
VSRF Princip
ler Governmi
ier Governmi
5565
CO O> O>
JN 8
o T- in
in d •*
•o-
to co r-
co ,- «
CM *^
o> to m
m W w
CO
CO CM CO
to' d r;
CM ^
^ ° S
t^ CM "^
o> m •<-
d CM j;
o> o co
o> d *
T- CO T-
CMCO -
O
K
?
la
£
rrowing from
ler
iservations
mo O
N-T-CMCMh-CD^- CO
i^ddcMco'oiin °?
t^ ^
Ocoooor-o co
to •* o o o o> o
m co
OOCOO^tOCMO CM
h^ddiridcbo ^
CD CM
CnOOCOtDCMO O
So d in W cb d T~
CO
CMOOOO>O>T- m
cdcodiridTrco *~
m T- CM
T-or^cOT-coco TT
CMO'J-'-TtOO'oS ^
in CM
T-oor^ocoo> CD
iriddM-'dco'co ^~
CD t- r-
oooocooocoo> o>
Trddd^coco *>
OOr-COOCMCO •*
co'ddcodcMcd ^
IA
IA
0)
m co co CM o
co to CM in co CM f
CD T-
oo in o co to co T-
CM T- O T- o M- O
CO t^ CD T— •*— 00 CM
h- CO O CM -^ CO i W
tD CM
CM CO CM C~ O) T- •^J•
W CO T-^ -^ v-' CO »-'
in co
N- T- T- to o m r-
tb to T- o> to co d
in CM
• CD o
O CM O O CO CO CO
m T- T- T-
co co ^ in co ^ •^-
in co co r^ o t- co
in ^- *-
•* t~- m o o> oo •*
T- T- co CM ^ m CM
m CM to o •* co o>
T; CM ^ CO 0 T- 0)
at
0>
'1
o
LL
"c o
0 t)
t g ^ CO
» * 0 TO
n^ t . t . >
S "5 m S CL
i « -9- E E E
I S i E 11
1 1 3 £ 1 1 1
— OQQOOcoO
<
CM
00
CO
"*
0>
CO
^
in
0
CO
CM
00
o
^
CM
CM
O
in
iservations
JJ
0
to
m
~m
to
Q.
d>
r—
c,
IA
3
73
0)
X
— -
"o.
S
D)
C
CO
IA
-2
>,
IA
d>
CA
O
•*"*
-2
"E
o
IA
73
C
Is
I
*O
ft)
8
t_
0
"C
S.
d>
o S
« J
0 K
. . io
"S "5
O Z
j£ 3
IA ^
E -5
d) JC
IA _
>> ro
IA Q.
"O O
O> 'C
TO Q.
CO _CA
-2 3
3-f
"D ftn
>> 73
S £
E §
ro ,„
s- s
2 g
to .52
o. J2
CO t-
O D)
CA
£ w
o g d
III
o E> ™
01 tl CO
i_ C —
111
s^l
II!
m~ o
E ^-S
E ° 5
ra t~~
c .«
C CO
"5
ro £
0) CO
E S1
o *-
**" O>
E o>
3 £
(A C
c
.3 .—
^ "S
c c
S" £
(D i_
•— o
S 3
— o
jS m
5 D)
(/) —
"E§
IA "~
» s
oT ®
O) C
||
-2- $
|'|
™ CO
•s,
ol
3 S
E -5
% (D
S S
o o
tO A
-g 0)
5 S
^ g
H -S3
* 3 *•
-------
.0
(0
0>
r-
O
**—
"c
£l
CO *-
(!) (0
* 5
u> c
+J
re o
f^ *^
(A
£ g .9
tf> *S At
Is*
5 O CD
"5. w
re <0
ii£
^" «*•
a> c
>f O
§1
w ^
CO £
5
c
8
m
S.
^
o
O)
re
o
c
Is
a
0)
u
'E
CO
co
(0
10
s
0)
=
s °
^ ^?
O
o
IO
1 o
0 g
° S-
§1
1 0
^ o
o o
°. 0
0 0
U> v
1 o
T- O
0 0
|8
1 0
T- O
Ofl
^^
*\ o
1 0
T— O
o CO O CM
CO •* T^ N- ^ CNI I-1
CO •*
o CM o co in i*- *-
O> T^ CD O ! C3 CO O
CM CO
m co co CM o T- o
••t CO C> CM CO TT CM
in co
O> CO CO CM CO T- T—
co m o co ^- to 't
CO •*!•
ro o co r- o to o
ai co CD ci r-^ cb o
TT CM T-
o •* to co r>- co co
•* •* o r^ co co co
CO T- T- CM
CM t T- CO CD CO CM
t^ co i- Tr o o CM
CM CM T- 1- T- T-
m ^ co ^t to C3> o
M- CM CO CO CM CO O
CO IO CO O Tf 1^ O»
CM 00 00 O O> CO TJ-
CO T- CM
10
s
c
a>
'§)
o
u.
lw J
2. 1 1 s
Sd2S
co c c i-
a> co o) CD °-
illtftj
£ co Q ^ S S
WC^K^^I^
mo^^JOa>cD2d>
•§t3QQOOt§6
Q.
O> COTjoCT>^fOO>
J2 CM CMCJ^-cb-^CM
to cotooopv-^p
^ r-r^ooci^cD
CO CO
r-- incoocoooo
12 CM i O O IO O -^ O
CO CO
in inoococMh~-o
00 T-OOCOCM'I~-'O
CM CD
co r-t^-ocO'-coo
II o^cbcbocbcM
CO T- ^
•** coo^-cMr--ino>
01 iti i ci o csi T- CD co
co »- •*
T- COOOCMOT-O)
II CMOcioOTtCMT-
•^ CO T- T-
CO COOOt^O(^OO
^ T-OCDOCM^h^
to T- CM
*- T-OCOCJ>OCMO>
^~ TtocicMcicir-
CO CM T-
to
c
CD
'5>
£
"c 0
CD 15
>> " -^ w
S.| as
fl) _j CO (0
a: „ :: .>
to — 15 c •-
CD CD CD a> a.
O lit CD ^5 -^ 5 Si
1 "cafc^^l^
^ SK^^CDCDScD
6 £ooaoo<§6
a.
CO
0>
CO
CM
T~
o
m
^-
1
to
o>
to
to
c
o
1
$
.a
o
T- CO T-
O> CO •>—
CO
•*— m o
oi W o
CM
co in co
to co o
m
CM T- h-
r^ m o
CO
T- CM CO
CO CO O
CM co m
-* CO 0
CO
in co T-
co o> o
CO T-
m o oo
r- O O
CM T- T-
m T- 10
CO O> ^~
CO T-
a>
c
CD
'E>
o
u_
1r
» t ^ *2
= 000
<
r^- o o> co
r^ in ^ ^
co in ^ T—
o o co o
CD
m h- co oo
CM CM CO T-
co
^" ^" in r*-
co v-' m co
CO CO CM CM
a> r- T- o
*- CM
•^j- to a> o>
m co oo co
»- T- CN
m •* T- T-
co ^ W CM
o oo to CM
T- co t in
^- CO
in co to co
T— ^ 00 OO
1
10 H *
* a s
° o 5
In A fi
III
5 S
o o •§
CO d> 2 CD
O O m O
CM
to
CO
00
^°
O)
CD
'
m
CO
CM
S
o
,
CO
a.
ff
C;
CO
£
^
^
c
CD
o
X
0)
—
5.
s
~
CO
E
to
>,
en
a>
en
o
£
fi
_>»
c
o
in
"^
c
*:.
'a.
8
"S
8
3
O
to
*C
0 §.
CM a>
CO *-
_; m
Q.32A, C
This tablf
to
to 'o
Q Z
^ CO
CO ^
EI
1o ^
CO Q.
T3 '0
CD C
D> C
CO Q.
'E .C
CO 10
> '5
"O ^.
CO °>
to E
TJ 1»
>* "O
S S
^
CD £
E o
>.T>
£ 1
"5 .2
o 1—
i o>
"c
c CO
CO ™
O O)
"o> "*
li .
0 | §
u_> .5 —
o E> ™
Q-£ £
i_ C "~"
0 CD •£
= E £
> * g-
flE
^.c »
1 'i. J
§s *
O *m ^
^ o'l
U. H: "•
o: '5 E
0 B S
.§a
H §•£
—
1
o
.E a>
"D f>
3 •*-•
**~ c
.C '~
0 -D
CO fl>
CD t±
E i.
.t £
CD !2
E •£
85
^—
CO
£ co
,- O)
o SS
3 CD
E S
§§
" i
O (O
Q. 5^
CD »
CD O
S g
S ,_
_to .p
<—
^ Ji
-^ 3
CO 0
C »
fij
c S
c E
|i
-e "
IB
8- E
X
o> $
If
Of all caf
not repor
-------
w
fl>
0
O)
_c
3
1
£ •-
c< -5 •£
4) ft r-
S3 i
|5 ;_ O
•ScS
1
1
1
o
D)
2
o
>
£•
0
D)
s
CO
O
c
O
15
3
a
o
a.
CD
O
'E
a>
CO
CD
CO
"*
co
0
*
ublic Systems
DWSRF Loans
a.
in
CM
CO
CO
CO
CM
CO
in
CO
o
CM
c\!
*
c
CO
c
m
Other Governmi
O) CO
t' T-'
CM «
"*. "^
in in
T— *
in
CO «
co in
CD C3>
^ 0
00 *
CO
o *
ct
Borrowing from
Other
in CD •* CM *
JQ CM in CM
O CO * O •
<"> •*' in
T- O> < CD *
f- ,-• in' r-
^- * co co *
^ co in
CM m CM to *
* tsi CD to
o * * m *
n 0)
CM * co in *
n in" o
co * in ^ *
T- * * CO *
0
,. a>
g W
gs
•»-» -^
§ £
g » E E
1 S§H
~ n _i > o>
1 *fe5|
S! l^lil
xi 2 S S o S
O ~ Q O CO O
a.
T- ^ CO T- T-
n CO CM TT T^
CM T}; CO CM *
coiriiri
m o en co m
to co in in
in p o> CM «
co to m
CO CO O) CM *
CO CO ^~
t- co m o m
T^ co in T-
CO CO CD CM O
•* CM CO CD
^ CO CO CO *
c> •«-: co
CM * « oq *
o
t3
,n «
g CO
co a>
°1
ICL
w E E
10 E E o
0 (ft § I*
O = Q 6 co O
<
CO
CO
CM
CNI
CO
CO
f-
CO
CO
CO
in
CO
in
CO
Observations
to
o
c
CB
.C
O)
c
'•e
0
Q.
cu
TD
CO
CO
CD
in
CO
Q.
xi
2
*i
C
o>
Q.
X
0)
.1
Q.
9 S
1 c?
w i
i? |
8C
»
» E
^ ct>
o SJ
a> ""
« 'g
m ^-
1 £
10 CD
O 5 S
CM M^
co o .52
ci ? i-
. . in
IS "o
Q Z
-------
Part 2
Methodology Report
115
-------
1. INTRODUCTION
1.1 Study Background
In compliance with Executive Order 12866, the Regulatory Flexibility Act, and the Safe
Drinking Water Act (SDWA), the U.S. Environmental Protection Agency's Office of Ground Water
and Drinking Water (OGWDW), Standards and Risk Management Division (SRMD) conducts
periodic surveys of the financial and operating characteristics of community water systems. These
Community Water System Suveys (CWSSs) supply information that is essential to support economic
analyses of the costs and benefits of new regulations and changes to existing regulations on consumers,
the water supply industry, and the nation. The information also will be used to measure the financial
burden of EPA's regulations on consumers and the industry. Furthermore, data from the survey will
help EPA identify, evaluate and develop guidance on Best Management Practices used in water
treatment and distribution systems. Previous surveys of Community Water Systems were conducted in
1976,1982, 1986, and 1995.
1.2 Survey Overview
This section is intended to provide the reader with an overview of the design and conduct of the
CWSS. The topics presented in this section will then be discussed at greater length in the following
chapters.
The CWSS was designed to collect operating and financial information from a representative
sample of community water systems. In order to reduce the burden of the survey on small systems, the
data were collected from systems serving 3,300 or fewer people through site visits by water system
professionals. Systems serving over 3,300 people received the questionnaire in the mail. Water system
professionals were assigned to each system that received the mailed questionnaire to help the systems
respond to the survey's questions. A toll-free telephone number and an e-mail address also were
provided to the systems to provide technical support.
The Community Water System survey was based on a nationally representative sample of
community water systems. The sample was drawn from a list of approximately 53,000 systems in the
Safe Drinking Water Information System (SDWIS). The survey used a stratified random sample
design to ensure the sample is representative. The sample was stratified by several characteristics of
water systems to increase the efficiency of estimates based on the sample. To limit the travel costs
involved in visiting each small system in the sample, they were selected in geographic clusters in a two-
stage design. A sample of 1,806 systems was selected, including a census of all systems serving
populations of 100,000 or more.
117
-------
A separate version of the questionnaire was developed for systems serving more than 500,000
people. Additional questions were asked of these very large systems regarding concentrations of
several contaminants in raw and finished water and average well depth. Questions that would not apply
to very large systems were excluded from their version of the questionnaire.
Water system professionals contacted the small systems in the sample to schedule appointments
for the site visits. Upon mailout, each system receiving the questionnaire was notified by telephone that
they would receive the questionnaire in the mail. Phone calls were made throughout the data collection
period to encourage non-respondents to participate and to provide technical support when needed.
Requests to re-mail the questionnaire were received through the toll-free support line and during the
phone calls to the system; the questionnaires were re-mailed as the requests were received.
As completed questionnaires were returned, they were logged into a receipt control system
using an on-line data tracking system. The completed questionnaires went through an extensive data
quality review. Water system analysts reviewed each mailout questionnaire and contacted the systems
to clarify answers, correct anomalous items, or collect missing responses. The questionnaires were then
reviewed by senior engineering staff. The senior staff also reviewed the site visits reports for each small
system. The questionnaires were then key-entered using independent double-key entry. Finally, the
electronic form of the data were run through automated cleaning and editing programs.
A series of sample weights, non-response adjustments, and other statistical techniques were
created and applied to the final set of sampled respondents. These weights allow for extrapolation from
the sampled systems to the universe of Community Water Systems in the nation. The sample design
and weights also allow for the calculation of confidence intervals for each estimate.
Planning and design of the survey began in August of 1999. A pre-test of the questionnaire was
conducted in July 2000. The pilot test was conducted from April to May, 2001, and the final design
was developed in May 2001. Data collection took place from June through October, 2001. Data
processing and analysis continued through February 2002.
EPA secured the services of several contractors who performed a variety of tasks in support of
the survey design, survey adrninistration, data processing, and analysis. The Cadmus Group, Inc., was
the prime contractor. Abt Associates, Inc. served as a subcontractor, as did Norfolk Data, Inc. The
site visits were conducted through subcontracts with several experienced water system professionals.
The Cadmus Group has been supporting EPA and other clients in the assessment and analysis of the
water industry for over 20 years. Cadmus' primary responsibilities were for overall project
management; design of the questionnaire; sample design; selection of the medium, large, and very large
system sample; design, administration, and management of the data collection; technical support to
water systems in the sample; expert quality assurance review of the survey data; data tabulations; and
report preparation. Abt Associates assisted with the development of the survey instrument, selected
the small system sample, developed and maintained the on-line data tracking system, edited and
118
-------
prepared the data into final form for data entry, managed the data entry, calculated the sample weights,
assisted in the data tabulations, and developed the final data with documentation. Norfolk Data keyed
the data into the electronic data base.
EPA also requested comments on the survey from several independent reviewers. EPA
consulted with Jeanne Bailey in the office of Regulatory Affairs at the American Water Works
Association in Washington, DC. Drafts of the questionnaire were reviewed by Diane Moles, from the
fowa Department of Natural Resources, in Des Moines, I A, and James K. Cleland of the Drinking
Water and Radiological Protection Division of the Michigan DEQ in Lansing, Mf. EPA also consulted
with Robert W. Mann of the Water Supply Engineering Bureau of the State of New Hampshire's
Department of Environmental Services.
119
-------
2. SAMPLE DESIGN AND WEIGHTING
2.1 Sample Design and Selection
This section describes the sample design for the 2000 Community Water System Survey
(CWSS). It includes a description of the sampling frame, target sample size, stratification variables, and
sampling methods.
The survey relied on a probability sample of Community Water Systems. For small systems
(those serving populations of 3,300 or less), a two-stage cluster sample was used. A stratified random
sample was used for systems serving populations of between 3,301 and 100,000. Systems serving
populations of over 100,000 were selected with certainty. The strata were defined by the combinations
of the size of the residential population served by the water systems and the source of water (ground or
surface).
2.1.1 SDWIS Sampling Frame and Coverage
The sampling frame is developed from the federal Safe Drinking Water Information System
(SDWIS/Fed). SDWIS is a centralized database of information on public water systems, including
their compliance with monitoring requirements, maximum contaminant levels (MCLs), and other
requirements of the Safe Drinking Water Act (SDWA) Amendments of 1996. The following
information was extracted from SDWIS for the statistical survey:
Name of system
Address of system
Population served
Primary source (surface water or ground water)
Public water system identification number (PWSID)
Ownership type
Consecutive system (i.e., does system purchase or sell water)
From these data, EPA developed a sample list from which it (1) calculated summary statistics
for use in calculating sample size, and (2) randomly chose systems within the design strata which will
take part in the survey.
120
-------
SDWIS is the appropriate sampling frame because:
It fully covers the target population.
It contains no duplication.
It contains no foreign elements (i.e., elements that are not members of the population).
It contains information for identifying and contacting the units selected in the sample.
It contains other information that will improve the efficiency of the sample design.
SDWIS is the best choice for a sample frame because of its inclusive coverage of all units of
observation for this survey, hi addition, SDWIS has two other advantages: it contains information that
will facilitate contacting the respondents, and it contains other information that is useful in stratifying the
sample, thereby improving the efficiency of the sample design. However, SDWIS is not designed to be
such a sample frame; many properties of SDWIS, and some lingering problems of system classification
in SDWIS, can result in many inaccuracies for such sample frame applications and sample selection.
For EPA's 1999 Drinking Water Infrastructure Needs Survey (Needs Survey), EPA and
Cadmus made a considerable effort to improve the SDWIS information and create an inventory list
more suitable as a sample frame. The Needs Survey took several steps to prepare SDWIS for use as
a sample frame. Problematic data were first identified based on the experience of the 1995 Needs
Survey. EPA then provided the confirmed inventory data to the States (including the Virgin Islands and
Puerto Rico) for review and asked the States to provide any necessary changes.
EPA also worked with the States to identify the total "consecutive" population served (including
the population of retail buyers) by many prominent large systems, to group systems into size and type
categories that more accurately reflect actual populations served by a particular water system. For
instance, the reported population served by the Metropolitan Water District of Southern California
(MWD) is categorized as a small system serving less than 3,300 people in SDWIS; in fact, MWD
serves nearly 16 million consumers. Therefore, the system was reclassified as a large system, which
accurately reflects the way this system is regulated under the Safe Drinking Water Act.
Criteria used to determine the accuracy of SDWIS data include: (1) 1995 inventory
verification showing a discrepancy rate greater than 1 percent; or (2) the number of community water
systems in a State in SDWIS as of March 1998 being at least 3 percent greater than in the sampling
frame used for the 1995 Needs Survey. On site inventory verifications were also conducted for States
that contributed to at least 0.8 percent of total national need in the 1995 Needs Survey, and if EPA
determined that SDWIS inventory may not accurately reflect a State's inventory, based on experience
with the 1995 Needs Survey. Inventory verifications were conducted in Arizona, Arkansas, California,
121
-------
Colorado, New York, North Carolina, Ohio, Oklahoma, and Tennessee. SDWIS-Fed inventory
information for Virginia was replaced with SDWIS/State inventory information, since SDWIS/Fed was
known to be an inaccurate source of current inventory.
The process of State corrections included a variety of inventory review procedures and data
verification:
A stratified random sample of systems was used to select systems within each State to
verify the inventory. (A two-staged cluster sampling approach was used to select
systems in New York, since data in this State are managed by numerous district
offices.)
Sanitary survey information, bacteriological results, or other chemical records in State
files and/or databases were reviewed on site to ensure that inventory data were
accurate. If inventory information was different between SDWIS and the State files
and/or database, a discrepancy was issued. Each State so identified was then given an
opportunity to provide monitoring results or other documentation of a system's
characteristics, and in some cases, a system's actual existence. Systems that were
inactive were removed from the Needs Survey sampling frame, while other systems
were re-categorized if necessary.
Based on results of the inventor/ verification, the total inventory for each State was
further refined. The inventory verification results were extrapolated to all systems in
each State to estimate the number of active systems in each size and type category.
In addition to these changes, the inventoiy of systems serving Alaskan native populations in
SDWIS was replaced with data developed by EPA Region 10 and the State of Alaska.
The Needs Survey sample frame was further refined during the course of the data collection
period. System status as of January 1,1999 was; used to determine inclusion and placement within the
sample frame.
The sample frame used for the Needs Survey is based on the SDWIS from 1998; to account
for changes in SDWIS, we compared the SDWIS available from the first quarter of 2000 to the Needs
Survey list. New systems were verified and added. This revised list of systems will be used as the
sample frame for the CWSS.
122
-------
2.1.2 Sample Design and Selection
Sample Eligibility
To be eligible for the CWSS, a water system must meet several criteria. First, it must meet the
CFR definition of a community water system; principally, a water system providing drinking water to 25
or more permanent residents or to 15 permanent connections. (See 40 CFR 141.2 for the complete
definition.) In addition, the CWSS excluded federal- and state-owned or operated systems; because
these are not affected by regulatory and economic forces in the same way as other systems. To the
extent possible, all ineligible systems were identified in SDWIS and removed from the frame; however,
many ineligible systems could not be identified and were therefore left in the frame. If systems were
clearly identified as ineligible during data collection (e.g., they are no longer an active water system, they
no longer meet the CFR definition of a Community Water System, or they are owned by the federal or
a state government), the data were excluded from analyses based on the sample.
After it drew the initial sample, EPA decided to exclude systems in the trust territories. The
original sample included 64 systems in Puerto Rico, the U.S. Virgin Islands, and the Pacific territories.
EPA decided not to select a new sample because this would affect the site visit schedule. Instead, EPA
dropped the systems in the trust territories, and classified these systems as non-respondents when
calculating the sample weights.
Sample Design
The CWSS analytical plan specified precision level targets for subpopulations of systems,
which required minimum sample sizes be achieved for each subpopulation. The precision targets for
each subpopulation were 95 percent confidence intervals of ± 10 percentage points for estimated
proportions. The domains of the population of interest for EPA are based on two characteristics of the
systems:
1. The source of water. Systems that rely on ground water are distinguished from
surface water systems.
2. The size of the population served by the system. Eight size categories will be used:
systems that serve less than 100 people; systems that serve 101 to 500 people; systems
that serve from 501 to 3,300 people; systems that serve from 3,301 to 10,000 people,
systems that serve from 10,001 to 50,000 people; systems that serve 50,001 to
100,000 people; systems that serve from 100,001 to 500,000 people; and systems
serving more than 500,000 people.
The two water sources and the eight system sizes produce sixteen strata.
123
-------
A system is classified as a surface water system in SDWIS if any of its water is surface water.
Ground water under the direct influence of surface water is classified as surface water. Systems that
rely on purchased water systems are included in the ground water strata because we assume the
characteristics of the water and the treatment requirements will be more similar to ground water than to
surface water. (While some untreated surface v/ater is purchased, the majority is treated and therefore
more similar to ground water than surface water.)
The sample is stratified to achieve two goals. First, stratifying the data allows us to draw
inferences about specific population domains. For example, EPA may wish to draw conclusions about
systems serving populations of less than 10,000 or 3,300. We can ensure that estimates of the sub-
populations will meet the required levels of precision by drawing the necessary number of observations
for each stratum.
The second goal achieved by stratifying the data is that we can increase the efficiency of our
estimates by grouping systems into relatively homogeneous strata. The strata were chosen to minimize
the differences among systems within strata, and to maximize the differences among strata. The results
of previous surveys indicate there are important differences in the way systems are operated and in their
finances across the strata selected. The operating characteristics and treatment requirements of ground
water systems tend to be different from surface water systems. The operating and financial
characteristics of large systems tend to be more complex than small systems. System management, and
the resources available to it, also may vary by system size. The regulatory impact models require
reasonably precise parameter estimates from each of these domains. The sample size in each domain
should be large enough to provide a sufficient mimber of completed questionnaires to obtain estimates
with reasonable precision.
Table 2-1 shows the number of systems in the sample frame and the minimum sample size
required to obtain an estimate for a proportion of 50 percent with an error not exceeding ±10
percentage points (except for a 1 in 20 chance) in each domain. (A 50 percent statistic was used
because the standard error is largest when the population percentage is 50 percent. The error will be
smaller for other population percentages.) Systems with populations served of over 100,000 were
selected with certainty.
Sample Selection
For Community Water Systems serving, 3,300 or fewer people (small CWSs), a two-stage
sampling design was used to reduce field data collection costs. Field data collectors were sent to the
clusters of six systems at a time to collect data. The primary sampling unit (1*811) was a county or a
group of counties. (Each county with fewer than six small systems was combined with geographically
adjacent counties to form the primary sampling units.) At the first stage of sampling, a sample of 100
PSUs were selected with probabilities proportional to size. The measure of size was the number of
small systems in the PSU. A large PSU could be selected ("hit") more than one time. PSUs were re-
124
-------
sampled to account for potential non-response. The over-sampling rate was determined based on
EPA's experience with the 1999 EPA DWINS and the 1995 CWS Survey. States were provided
with a list of small CWSs in the counties selected, and EPA asked States to verify that the systems on
the list are active and serve populations of 3,300 or fewer.
To select the second stage sample of small systems, the overall selection rate for each small
system stratum was calculated as the target initial sample size in the stratum divided by number of
systems in the stratum. The expected frequency of selection was calculated for each PSU in the first
stage sample. For each PSU selected, the second stage selection rate for a stratum equaled the overall
selection rate for the stratum divided by the first-stage expected frequency of selection. That second
stage selection rate for a stratum was applied to the count of systems in that county to determine the
fractional sample size. The fractional sample sizes was converted to integer sample sizes using
stochastic rounding and with the constraint that the total integer sample size for a county hit equals six
systems. To measure composite sample size in selecting counties or PSUs, an "overall stratum
selection rate" was multiplied by the number of systems in the stratum in that PSU, and summed over all
strata in each primary sampling unit (county or group of counties).
125
-------
Table 2-1. Frame and Sample Sizes by Strata
Source of Water
Ground
Surface
Population Served
100 or less
101-500
501 - 3,300
3,301 - 10,000
10,001-50,000
50,001 - 100,000
100,001 -500,000
More than 5 00,000
100 or less
101-500
501 - 3,300
3,301 - 10,000
10,001-50,000
50,001 - 100,000
100,001 - 500,000
More than 500,000
All
Frame Size
14,972
16,025
12,341
3,300
1,921
260
126
17
538
789
1,551
988
970
215
189
66
54,268
Required
Sample
95
95
95
92
89
60
126
17
88
91
93
91
91
74
189
66
1,452
For systems serving populations of 3,301 - 100,000, the sample was obtained by drawing a
random sample of systems from the cleaned SDWIS frame, within each sampling stratum serving
populations of this size. Systems in these strata were oversampled to account for non-response. As
with the small systems, the over-sampling rate was based on EPA's experience with the 1999 DWINS
and the 1995 CWSS. Systems serving populations of more than 100,000 were selected with certainty.
The resulting increase in sample size is warranted for the following reasons:
126
-------
Each of the larger systems has a more significant impact on the total costs and benefits
of regulations.
Because of the small numbers of systems in many of the larger strata, precision can be
increased at comparatively lower cost than it can be for smaller systems. Other things
being equal, doubling precision will quadruple sample size in strata with 5,000 systems
or more. Many of the larger strata, however, have only hundreds of systems, hi a
stratum of 750 systems, one could double precision by only tripling sample size. In a
stratum of 200 systems, one could double precision by doubling sample size.
A total of 1,870 systems were selected into the sample. The sample size by strata and the and
sampling rate are shown in table 2-2.
2.1.3 Stratum Migration
Errors in the SDWIS frame classification of the water systems by population served and water
source introduces inefficiency in the sample design through a loss of sample size and/or by introducing
unequal sampling rates. Among the respondents, 87 percent reported the same population served
category as indicated by the frame. Just over 91 percent reported the same source as the frame.
Population Served by the System
Table 2-3 compares the classification of systems by their population served using the population
data from the frame and from the systems' responses to the survey. In all size categories, more than 80
percent of systems confirmed their original size category. Within each size category, over 96 percent of
systems were either in their original size category or in the adjacent class. While migration across size
categories is small, several systems reported to serve 3,300 or fewer people in the frame moved into
larger categories; 20 moved into the 3,301-10,000 category, and 4 moved into larger population
categories. Also, 19 systems moved into small system categories from the larger categories.
127
-------
Table 2-2. Sample Size and Sampling Rate by Strata
Source of
Water
Ground
Surface
Population Served
1 00 or less
101-500
501 -3,300
3,301 - 10,000
10,001-50,000
50,001 - 100,000
100,001 -500,000
More than 500,000
1 00 or less
101-500
501-3,300
3,301 - 10,000
10,001-50,000
50,001 - 100,000
100,001 - 500,000
More than 500,000
AH
Sample
Size
128
124
124
155
152
116
126
17
94
85
90
145
144
110
194
66
1,870
Sampling
Rate (%)
0.9
0.8
1.0
4.7
7.9
44.6
100.0
100.0
17.5
10.8
5.8
14.7
14.8
51.2
102.6
100.0
3.4
128
-------
o
U
03
i
c
5
S-i
h o
> <=>
o
1 O
"^ O
® °-
O o
0 ^
1 O
5 g
o
o" 3
in ^^
1—1
1 o
s §
o
0 £
1 O
_ o
f"} ^r-
- cT
1 O
o
in "^1
O
O
in
0
0
>n
(N
a
_o
&1 <8
Cw * r™^
-o o
£
Wrt
, 0
^ '"O
p, a>
S b
03 (U
00 00
o o
— 1 ^
o o
o o
o o
o o
m oo
^ ON
-^ "S
0 fe
O OH
o
0
1
(N
o o
o o
0 0
o o
_ , 1
o ^o
^H
•^- oo
^H
-<-< C
§ p
§ fe
U OH
o
o
in
1
o
o o
— ' O
o o
^ o
>n r--
— *
r- -*
t~- oo
^^
' '
^ G
c g
U OH
0
o
»\
ff)
1
0
0
O 0
— H •—!
o o
o ^o
• — '
ON (N
m oo
^^
ON ^
o o
^
11
O OH
o
o
§
,__<
o
^
o o
— ' •— '
ON ^O
^O ^d"
m oo
CO OO
^
, , , !
o o
+-» s
0 §
O OH
0
o
o
0s
1
o
o"
o o
^ CO
OO (N
O ON
^- ^-
o o
o o
o o
-*— »
1 1
O OH
o
0
o"
0
»— I
*— I
o
n m
in ON
oa m
o o
o o
o o
o o
^_,
G
-------
Source of Water
Table 2-4 shows the cross-tabulation of the frame-based and response-based water source
classifications. Approximately 86 percent of the systems classified as ground water systems in the frame
confirmed that status in the sample. Ninety-four percent of surface water systems in the frame were
also classified as surface water systems in the sample.
Table 2-4. Survey Respondents by the Frame-Based and the Sample-Based Source
Categories
Sample-Based
Water Source
Ground
Surface
Count
Percent
Count
Percent
Frame-Based Water Source
Small
Ground
323
88
9
5
Surface
42
12
164
95
Medium
Ground
146
92
10
6
Surface
12
8
163
94
Large
Ground
123
87
15
6
Surface
19
13
220
94
All
Ground
592
89
34
6
Surface
73
11
547
94
Impact of Strata Migration on the Accuracy of Domain Estimates
The sample was designed to estimate a 50 percent statistic with a 95 percent confidence
interval of ± percentage points. One measure of the impact of the strata migration on the efficiency
estimates is to calculate the size of the confidence interval given number of observations in each stratum
for the sample collected Table 2-5 shows the minimum sample required to estimate a 50 percent
statistic with a 95 percent confidence interval of ± 10 percentage points. The planned sample size is the
sample that would be needed if the frame had correctly identified all subdomain members. The
required sample size is the sample needed, given the inaccuracies in the frame. The table also shows
the half-width of the 95 percent confidence interval that results from the acted sample selected, given
the sample's estimate of the number of systems in each subdomain. The increase in the half-width for
mid-sized ground water systems is modest. Because the sample was designed to collect data on all
systems with populations of more than 100,000, the width of the confidence interval for these systems
would have been zero. The width increased substantially for ground water systems in these strata and
slightly for surface water systems serving greater than 500,000 people categories due to the strata
migration, as the number of systems in these strata is larger than expected. It also increased for surface
water systems serving populations of more than 500,000.
130
-------
Table 2.5. Sample Sizes and the Impact on Precision of Estimates of Strata Migration
Source
Ground
Surface
Population Served
1 00 or less
101 -500
501 -3,300
3,301 - 10,000
10,001 -50,000
50,001-100,000
100,001-500,000
Over 500,000
1 00 or less
101 -500
501 -3,300
3,301 - 10,000
10,001 -50,000
50,001-100,000
100,001-500,000
Over 500,000
Required
95
95
95
93
91
70
125
17
79
85
90
88
87
66
191
66
Planned
95
95
95
92
89
60
73
8
88
91
93
91
91
74
257
50
Half width of
95% Confidence
Interval
0.100
0.100
0.100
0.101
0.101
0.111
0.074
0.260
0.094
0.096
0.099
0.098
0.098
0.092
0.000
0.069
2.2 Weighting and Estimation
A sampling weight is attached to each responding water system record to (1) account for
differential selection probabilities, and (2) reduce the potential bias resulting from nonresponse. The
sampling weights are necessary for estimation of the population characteristics of interest. The sample
variance is then used to calculate 95 percent confidence intervals for the estimates.
2.2.1 Derivation of Base Weight and Nonresponse Adjustment
The calculation of the sample weight reflects the complex nature of the sampling design. The
community water system sample consists of a stratified element sample of medium and large water
systems. Systems were stratified by water source and their population served. For small water
systems a two-stage cluster sample design was used.
1. At the first stage geographic clusters (counties or county groupings) were sampled using
probability proportional to size sampling. The measure of size was a function of the
number of small systems in the cluster.
131
-------
2. Within clusters a stratified element sample of small systems was drawn.
After the initial sample was drawn, it was decided to go back to some of the clusters and draw
additional small systems from specific strata to make up for a short fall in sample size due to a larger
than expected number of ineligible small systems. Also, sample clusters located in U.S. territories were
not included in the actual data collection for cost reasons and are treated as nonrespondents.
The EPA SDWIS data file was used as the sampling frame for sample selection. Sixteen
sampling strata were defined based on systems' population served and source of water; all weight
calculations use this sample stratum variable.
Base weights
The first step was the calculation of a base sampling weight for each sample system. For the
medium and large systems the base sampling weight equals the number of systems in the stratum
divided by the number sampled from that stratum, hi other words the base weight for the hth stratum,
Bh, is:
n N*
where Nh represents the number of systems in the stratum in SDWIS, and ^ represents the number of
systems sampled from the stratum.
For the small systems the base sampling weight equals the product of the reciprocal of the
probability of selection of the cluster times the reciprocal of the within cluster probability of selection of
the small system. For large clusters selected with certainty the cluster base sampling weight component
equals one. The reciprocal of the within cluster selection probability equals the number of small systems
in a stratum divided by the number selected from that stratum. The base weight for a sample system in
the hth stratum in the mth cluster is given by B^:
where Pm is the probability of selection of the mh cluster.
Nonresponse adjustment
The second step in the weighting methodology was to make a unit nonresponse adjustment to
132
-------
the base sampling weights. For each medium and large system stratum, the nonresponse adjustment
factor is equal to the ratio of the number of systems that completed the survey plus the number of
nonrespondents to the number of systems that completed the survey (i.e., the reciprocal of the stratum
response rate). Ineligible systems are not incorporated into the unit nonresponse adjustment. The
adjustment factor for the hth stratum is given by h:
Where rh is the number of refusals and other nonrespondents in the h"1 stratum.
For the small system sample the unit nonresponse adjustment was not implemented within each
cluster because the sample sizes were too small. Rather the adjustment was carried out within each
small system stratum at the total sample (i.e., national) level.
Final weights
The nonresponse adjustment factor h was multiplied by the base sampling weight, Bh, to
obtain the nonresponse adjusted base sampling weight. The nonresponse adjusted base sampling
weight for the medium and large systems that completed the survey is the final weight for use in analysis.
The nonresponse adjusted weights can be written as:
for medium and large systems, and
= Bmh8h
for small systems.
The final step in the weight calculations for small systems was a ratio adjustment to the SDWIS
data file count of small systems in each small stratum at the national level. This step was carried out
because the two-stage sample of small systems, drawn from 104 sample clusters, may not have the
same stratum distribution as the entire EPA data file of small systems. For each small system stratum,
the sum of the nonresponse adjusted base sampling weights for systems with a completed survey was
added to the sum of the base sampling weights for the ineligible systems. The count of small systems in
SDWIS was then divided by this sum. This yielded a ratio adjustment factor for each small system
133
-------
stratum, h:
Where: Rh is the set of systems that responded to the survey, and
Ih is the set of systems sampled that were ineligible.
j designates the jlh sample system.
For the small systems with a completed survey their nonresponse adjusted base sampling
weight was multiplied by the ratio adjustment factor to yield a final weight for use in analysis:
2.2.2 Variance Estimation
The estimate of the variance must account for the sampling design. Weights are used to
produce estimates for the population as a whole - for example, the proportion of treatment facilities that
use a particular treatment practice, or the mean water-sales revenue of a system. Weights also affect
the standard error of the estimates, and therefore the confidence intervals.
The 2000 CWSS sampling design was relatively complex; medium and large systems were
selected by strata; small systems were selected in clusters of counties (or, in some cases, groups of
counties) using a probability proportional to size sampling. This sampling design also affects the
estimate of the standard error. The stratification of the systems by water source and population served
will tend to reduce the overall sample variance, !is systems within a stratum tend to be similar to each
other and different from systems in other strata. The clustering will likely increase the sampling
variance, as systems within a cluster may be similar to each other. This effect of clustering may not be
large; while systems within a county share some; characteristics, the often are a diverse group in terms of
population served and water source, as well as revenue, expenses, and operating characteristics. But
ignoring the clustering may lead to an underestimate of the sampling variance, so it must be taken into
account.
The treatment facilities in the sample were not selected independently; rather, they were
selected in clusters in a two-stage process. For medium and large systems, the stratified random
sample of systems were selected in the first stage; every treatment facility in each system was selected
in the second stage. Facilities in small systems were selected in a three-stage process: counties (or
groups of counties) were selected in the first stage; a sample of systems withim each county was
selected in the second stage; every facility within each system was selected in the third stage. The
calculation of the sample variance of estimates regarding treatment facilities also must take into account
134
-------
this sampling design.
Variance Estimator
The variance is estimated using a first-order Taylor expansion. The variance is calculated in
Stata. The variance estimator is given by:
•^ / ^ \ 1 f -^ / •"* \ ^ /•"* ^ \ -^ "i •"* / ^ \1
V(R)= -^{v(Y)-2RCov(Y,x) + R2V(X}\
where R - / ~ , the ratio of estimates of two population totals. Y is equal to £ £ £ *%,%
./i A=] ,= ] ;=]
L mi, nh,
X is equal to £ £ £ w/;i JCA/ • L is the number of strata, rr^ is the number of primary sampling units in
strata h, and r^, is the number of elements in the fh primary sampling unit in the hth strata.
Most of the estimates produced in this volume are either means or proportions. A mean is
simply a ratio in which x,^ is equal 1. A proportion is simply a mean in which yhy is equal to a 0/1
variable.1
Finite Population Correction
A finite population correction factor was derived for medium and large systems in the sample.
The factor is the ratio of systems in the sample to the number of systems in each stratum. Because the
primary sampling units for small systems were selected with replacement, the finite population
correction factor is set equal to zero for small systems.
To estimate the variance, we first define the following ratio residual:
We then define the weighted total of the ratio residual as
'See Cochran, W.G. 1977, Sampling Techniques, New York: John Wiley & Sons for amore
information about variance estimates..
135
-------
"*./
Zdln =:
and the weighted average of the residual as:
1
>=\
We can then define the variance estimate as:
where if, is the finite population correction.
The estimate of the variance is used to estimate 95 percent confidence intervals in the detailed
tables of this report. An implicit assumption is that the average values presented in each table are
normally distributed. When the estimate is based on a large number of systems, this will generally be
true; in cases where the estimate is based on a small number of systems, the assumption may not hold.
The confidence interval in these cases may be larger than the mean itself. The confidence interval is not
adjusted in these cases; to compute the correct confidence interval requires examination of the empirical
distributions for each variable in the calculation and is beyond the scope of thus study.
136
-------
3. SURVEY DESIGN AND RESPONSE
The survey was administered through site visits to small systems (those serving populations of
3,300 or less), and through a mail survey to medium and large systems (those serving more than 3,300
people). This chapter discusses the survey instrument, the processes for conducting the site visits and
distributing the questionnaires, as well as the process to assure sufficient response rates and the handling
of returned questionnaires.
3.1 Questionnaire Design
The Cadmus Group, working closely with EPA staff responsible for regulatory development,
developed the questionnaire. The process began with a meeting of EPA staff to discuss their data
needs, distinguishing core needs required for regulatory development from other data needs. Based on
these discussions, some of the questions that were in the 1995 CWSS were eliminated from the 2000
questionnaire. Other questions - especially those focusing on treatment - were further developed. A
slightly modified version of the questionnaire was developed for systems that serve populations of over
500,000; this version of the questionnaire included additional questions on source and finished water
contaminant concentrations, and excluded questions that only would apply to small systems. The
survey instrument is in Appendix A. The questionnaire in Appendix A is a composite of the two
questionnaires used; the questions that are asked only of systems serving over 500,000 people or only
systems serving up to 500,000 people are noted.
Cadmus worked with EPA on the wording and organization of the questionnaire. It was
responsible for the design and layout of the questionnaire form, and for documenting and incorporating
all revisions to the several design and test versions of the questionnaire. Throughout the design process,
the EPA project officer consulted with the full range of EPA regulatory and analytical staff, representing
expert advisors and future users fo the data, to identify and correctly present the broad survey topics
and specific survey questions to be included in the survey instrument. These covered such areas as
water production, storage, distribution, treatment, and cross connection control, as well as financial
information regarding water sales revenue, customer data, operating expenses, and capital investment.
EPA went to great lengths to attempt to reduce the burden to respondents while collecting
complete, accurate, detailed data. EPA decided to conduct site visits to small systems because of the
difficulties they faced in responding to past Community Water System Surveys. EPA also established a
process to provide extensive technical assistance and guidance to recipients of the mailed questionnaire.
As discussed in chapter 4, EPA conducted a pre-test of the questionnaire to identify questions that
posed potential problems for respondents. EPA also conducted a pilot test of the data collection
methods. As a response to both tests, EPA made several changes to the questionnaire, reducing the
scope of several questions. For example, as a result of the pre-test, the number of age and diameter
categories was reduced in the question regarding the length of the distribution system, was simplified.
137
-------
3.2 Data Verification
EPA forwarded the list of water systems selected in the sample to the states. For small
systems, the states were asked to verify that the systems were active systems serving up to 3,300
people, as well as the address, telephone, and the contact information. The states identified 27
systems that were not active Community Water Systems. To replace the systems that were inactive, 22
additional systems were selected from each of the clusters that contained inactive systems. Not every
inactive system could be replaced because each cluster did not have enough systems in the sample
frame to replace all the inactive systems.
3.3 The Pilot Test
Approximately 50 systems were selected from the sample for a pilot test. Two clusters of small
systems were selected for site visits by senior Cadmus water system professionals. Ten systems
participated in the pilot test. (One system was inactive, and one refused to participate.) Approximately
40 systems serving more than 3,300 people received the questionnaire by mail. The pilot tested the site
visit and mail-out process, and the technical support system. The pilot systems were included in the full
sample.
3.4 Site Visit Operations.
Three contractors were selected to conduct the site visits. The contractors were:
International Studies and Training Institute, Inc.,
Southwest Environmental Engineering, and
McNenny Environmental Engineering and Consulting.
Cadmus also conducted several site visits.
Cadmus trained the site visit staff. The training covered the survey, the information required
from the systems, and the data collection protocol. The training included on-site inspections with
Cadmus staff of a cluster of systems in the sample, as well as detailed instructions on the conduct of the
visit.
The states were contacted ahead of time to confirm the systems in the sample and to review
information on the system contacts. Site visitors were told to let state contacts know they were in their
area and what they were doing, as a courtesy. The surveyor extended the opportunity to the states to
attend the survey. Otherwise, the surveyors were told to not burden the states with requests for
assistance.
138
-------
As part of the training, site visitors were instructed as follows:
The survey is voluntary and not to be misrepresented as mandatory. It is an
opportunity to provide information to be used by EPA to make sound, informed
decisions and regulations.
Obtain the operating and financial information for the same rime period of time, if at all
possible
If information is not available for the separate classes of system (for example, water
deliveries by customer class), then collect the totals (e.g., total deliveries).
Indicate the system has a treatment objectives only if the facility was "designed" for that
purpose. For example, if the facility was designed for particulate removal and removed
arsenic in the process, the surveyors were to only check particulate removal.
Complete the sequence of treatment after a walk-through of the treatment plant. If
available, collect a schematic.
Operators on-call are not the same as on-site; therefore, if a system only has an
operator on-call, it should not be classified as having an operator on-site 24 hours a
day.
SCADA for process monitoring was defined as information on values (i.e. elevations,
pressures, pH, c!2, etc.). It was classified as process control if it had the capability to
automatically control equipment (i.e. pump controls, feed equipment controls, pacing
control, etc)
Related questions should be checked for consistency. For example, questions on water
produced should be consistent with deliveries and unaccounted for water. Water
delivered should be consistent with the number of customers and connections.
Collect a service area map or draw the service area on a map (USGS, MapBlast,
Microsoft, etc) for each system.
The cross connection control program must be more than the plumbing code. The
program should be specifically designed as a backfiow prevention/cross connection
control program.
Financial data. If there is no other information, get the average annual bill
(question26). Again, if the information was not available in a manner that it could be
139
-------
broken down into components, get totals.
Collect financial reports if they are available and if the system will not break down the
costs as requested.
Several issues arose during the site visits; that required consistent responses. They included:
If the system indicated that it merged with another system, the site survey was
conducted.
If a system decreased in size so it was no longer a community water system, the site
visit was conducted to confirm the status.
If the system grew so it was no longer a small system, the site visit was conducted and
data collected.
Each site visitor was given a list of systems to visit. The site visitors contacted the systems to
schedule the on-site inspections; the site visitors were required at times to contact the state to confirm
contact information. Once on-site or in some cases prior to the site visit, the systems were provided
with an letter introducing the site visitor and explaining the survey. The site visitor inspected the system,
interviewed the staff, photographed the system (from source to delivery), and filled out the
questionnaire. The completed questionnaire, inclusive of the pictures, site map, and collected
information and reports, was then submitted to Cadmus. Senior staff at Cadrnus reviewed all surveys
submitted by the site visitors to ensure the site visitors were filling out the questionnaire correctly and to
insure consistent responses from the 6 site surveyors. The questionnaires were then logged into the
tracking system as received and completed.
During the site visits, Cadmus senior staff communicated with the site visitors via telephone and
e-mail to insure consistent and complete results. Group email was used to provide answers and
clarification to the site visitors questions. All site: visitors received the same information.
Site visits were not done in Alaska or Hawaii. For Alaska, two clusters of small systems were
surveyed using a combination of telephone interviews with system, state regulatory, and EPA personnel.
The remoteness and reluctance of some of the systems in Alaska made on-site visits infeasible. The
information that was collected from these system was abbreviated to promote response. The data
reflect the Alaska systems, but may not be easily compared to the remaining small system surveys.
140
-------
3.5 Mail Survey Operations
Cadmus produced camera-ready versions of the two questionnaires, which EPA printed at its
facility. Cadmus produced three mailing information and control labels for each system's questionnaire.
Information for the mailing label was extracted from the sample frame and attached to the envelope for
mailing. A second label that included contact information, the mailing address, and a telephone number
was produced and attached to the questionnaire itself. This label explicitly instructed the respondent to
respond only to the sampled system when answering the questions. Both labels included a public
water system identification number (PWS1D), in both alpha-numeric and bar-code. The third label
contained the toll-free telephone support line and e-mail information. Included in the envelope, along
with the questionnaire, was an introductory letter from the president of the Cadmus Group, a brief list of
instructions, and a pre-addressed pre-paid FedEx envelope for systems to return the completed
questionnaire.
The questionnaires were mailed to approximately 1,200 community water systems over a 2 day
period. Each system then received a telephone call from the analyst at Cadmus responsible for that
system. The call informed the system of the survey, told them they would receive the questionnaire, and
gave the systems a name and telephone number to call with any questions. If a system did not receive
the questionnaire, the analyst responsible for that system sent them another copy of the questionnaire
via FedEx. The analyst continued to follow-up with each system until the system either responded to
the survey or refused to participate. The analysts provided technical assistance as necessary, and in
some cases filled-out the questionnaire through a telephone interview.
As questionnaires were received from the water systems, Cadmus logged them into the on-line
tracking system. The analyst responsible for the questionnaire reviewed it for data quality and to
identify potential problems. When necessary, senior engineering or financial staff were consulted
regarding potential problems. If a problem or question could not be resolved by Cadmus staff, the
analyst contacted the Community Water System. When this initial review was completed, the
questionnaire was forwarded to senior staff for additional review. All changes to the questionnaires
were recorded in a permanent log. After the senior review was completed, the completed
questionnaire was forwarded to data entry.
3.6 Data Entry
Upon review by the senior staff, all questionnaires were logged as completed, and sent to Abt
for data review and editing, and preparation for data processing. The questionnaires were then key-
entered using 100 percent verified double-key entry. After entry, the data were run through automated
cleaning and editing programs that checked each variable for proper values and ranges, and checked
skip patterns. Items failing these checks were examined and either confirmed or corrected.
Questionnaires that reached this stage were considered to be entered and cleaned. (The data were
subject to further intensive checks as part of the quality assurance process, discussed in chapter 4.)
141
-------
Status reports were sent to the EPA project managers every two weeks during the data
collection effort. The report showed the number of questionnaires with each of the following status
codes:
Site visit appointments scheduled or questionnaires mailed
Questionnaires re-mailed
Inactive systems
Questionnaires undeliverable
Refusals
Site visits completed/questionnaire returned
Questionnaires reviewed and ready to enter into database
Completed questionnaires entered into database
Table 3-1 presents an example of the information provided to EPA.
142
-------
*-
o
D.
tu
VI
3
«
)
E
O
*j
Q
o
01
a.
E
X
U
T-*
E
!fl
a>
a
^
T3
e
o
u
A
0
E-
^" o
C3 O
IS
n A
i/l
o
o"
A
1 o
o_ »
o 5
1 0
i §
S o
1 o
o °
cT ^
1 0
s §
*T. cT
*** "
— : o
« 0
o **-
-° «A
i ^
o
1
*— (
0
o
0
V)
Population Served
^j-
ON
r~
NO
^
r^
=
CN
to
>T)
12
NO
m
^
CN
fS
OO
1
"5.
1
CO
m
00
oo
£
CN
0
_
ON
m
oo
w>
ON
ON
ON
Sample needed to meet
precision requirements*
NO
0
ON
NO
tri
r--
NO
Z
_
l/~l
m
!2
5
en
o
2
Appointments scheduled or
questionnaires mailed
o
• — •
0
•*
0
(N
OO
CN|
NO
CN
O
0
o
o
Questionnaires re-mailed
ON
">
O
-
o
_
NO
0
NO
0
£
o
•u
c
1
E >
"K y
r-
w>
o
CN|
0
_
CN
N
O
0
r,
Questionnaires
undeliverable
•^
ON
ON
-
^
(N
(N
u-i
' —
O
CN
O
-
Water systems that refused
to participate
CN
ON
"/">
O
OO
-
oo
NO
3
^
NO
m
r-
CN
m
~
ON
NO
ON
Site visits complete/
questionnaires received
00
f-
^"
CN
">
O
NO
^
^t
ON
NO
O
en
"
ON
ON
Questionnaires reviewed
for quality assurance and
ready to enter into database
0
o
0
o
o
o
o
o
o
o
o
Completed questionnaires
entered into database
m
-------
^
o
o.
V
a
|
a
53
E
o
£
3
CS
a
o
—
a.
E
X
fid
•
^^
(*)
4>
IS
C3
^
E
H
t^
L-
a)
«
^
u
<5
3
tf)
3
o
H
_r o
CB C>
2 *A
t/5
o
Of
o
0
in
A
1 0
•r o
O f--i
0 ^
o" S
o g
1 0
^! o
o ci
^" ^
* o
1 ^
0 °
^ cT
0 tff
> 0
o °
r^ °'
•™T ^5
0} O
-S f)
•Q '
3 rM
C»
T! °
i
i
o
o
0
Population Served
(N
ON
OO
SO
SO
SO
ON
O
^
2
m
2
O\
SO
ON!
0
ON
OO
s;
Sample selected
in
in
i —
o
in
so
sO
OO
ON
SO
SO
^
00
r-
oo
»•
in
CN|
0
3,
r-
Sample needed to meet
precision requirements*
in
00
3
SO
ON
OO
NO
0
oo
m
2
r*"i
.^
(N
m
00
NO
m
so
Appointments scheduled or
questionnaires mailed
3
00
-
a
2
(N
,j.
1 —
O
o
o
0
Questionnaires re-mailed
(N
-
0
-
O
0
o
^«
^~
o
r-
System found to be
inactive
SO
^
0
CN,
o
fN
O
C*l
CN|
O
o
Questionnaires
undeliverable
Tj-
r--
*
0
r-i
2
0
0)
VI
"~~
__ 1
o
o
Water systems that refused
to participate
ON
ON
in
0
f-
o
(N
SO
™
OO
SO
oo
r-
ON
' — '
^
ON
SO
Site visits complete/
questionnaires received
so
S
1^
rNi
CN!
oo
^
m
\o
o
in
r^i
ON
—
r--
K
ON
in
Questionnaires reviewed
for quality assurance and
ready to enter into database
o
o
0
o
0
o
o
o
o
0
o
Completed questionnaires
entered into database
-------
LH
O
a.
01
3
s
Sri
£
|
C5
a
o
_«
"a.
CB
X
W
,H4
P")
5
a
L_
r^
i
+-
>>
to
i>
•«j
SB
£
s
r> j;
1 0
0 §
<=" ^
C1 i/)
' 0
5 §
^i (^
^ ^*
_' o
A ^^
o ••
^ 1
3 01
1 ^
i"H O
1/1 r»T
•0
o
1
^*
0
o
0
Population Served
o
oo
01
n
oo
o
o!
NO
ON
Ol
0
o
f)
^
NO
^-
s
01
Ol
Ol
Sample selected
00
•rj-
o
f>
00
Ol
2
oo
£;
0
oo
s
•n
00
ON
r-
t^
Sample needed to meet
precision requirements*
t —
S
m
oo
2
01
ol
ON
00
ol
NO
ON
Ol
En
NO
0
0)
r-
oo
Appointments scheduled or
questionnaires mailed
^.
—
I
*n
—
5;
NO
ON
"*
0
rf
0
O
O
| Questionnaires re-mailed
*
0
ol
o
_
_
01
o
o
r-
o
13
C
1
s >
^~t f-
C/3 -?2
r^i
ON
0
^
o
m
01
"*
01
o
01
Questionnaires
undeliverable
ON
—
NO
tn
3
NO
u->
C^
in
"*
m
0
Ol
Water systems that refused
to participate
ON
01
00
NO
oo
"n
oo
00
oo
01
ON
ON
ON
0
oo
ON
2
r-
Site visits complete/
questionnaires received
ro
ON
|
0
f-)
01
01
r-
^
o
oo
o
ON
ON
00
ON
in
0
Questionnaires reviewed
for quality assurance and
ready to enter into database
o
0
o
o
o
o
0
o
0
o
0
Completed questionnaires
entered into database
T3
o
<4-l
o
to
§
<" 3
SEX
O
S o.
" 60
fr S
I "5
S V)
— tfl
8 S
.t= »
tfl
0> X
3 "
2; *
-------
3.7 Survey Response
The data collection effort was closed out October 31,2001. Of the 1,807 systems included in
the sample, 1,246 responded to the survey. The overall response rate was 67 percent. Table 3.2
shows the response rate by strata. Excluding the trust territories, the overall response rate was 69
percent.
Table 3-2. CWSS Responses and Response Rate by Strata
Source of
Water
Ground
Surface
Population Served
100 or less
101-500
501 -3,300
3,301 - 10,000
10,001-50,000
50,001 - 100,000
100,001 - 500,000
More than 500,000
100 or less
101-500
501 - 3,300
3,301 - 10,000
10,001-50,000
50,001 - 100,000
100,001 - 500,000
More than 500,000
M
Completed
Question-
naires
107
106
124
77
70
62
69
11
61
62
83
91
82
70
123
48
1,246
Response
Rate (%)
83.6
85.5
100.0
49.7
46.1
53.4
54.8
64.7
64.9
72.9
92.2
62.8
56.9
63.6
63.4
72.7
66.6
146
-------
4. QUALITY ASSURANCE AND PEER REVIEW
The quality assurance plan for the CWSS encompassed specific measures to check and ensure
the validity of the survey data from data collection through data processing and analysis, as well as
measures to assure the quality of other survey components. A Quality Assurance Project Plan (QAPP)
was developed for the survey and was approved prior to the start of data collection. The Office for
Environmental Information reviewed the methodology and the data presented in the body of the report.
The report results and statistical methods also were peer reviewed by subject matter experts. Drafts of
the questionnaire were reviewed by Diane Moles, from the Iowa Department of Natural Resources, in
Des Moines, IA, and James K. Cleland of the Drinking Water and Radiological Protection Division of
the Michigan DEQ in Lansing, Ml. OGWDW also consulted with Robert W. Mann of the Water
Supply Engineering Bureau of the State of New Hampshire's Department of Environmental Services.
Finally, the sampling design was reviewed by senior statisticians at as part of the external QAPP
review, as well as by Cadmus, Abt, and within EPA; it is the same basic design used for the 1999
DWINS.
Section 4.1 discusses the questionnaire pre-test and the survey pilot test. Section 4.2 presents
the measures taken to assure the quality of the statistical sample. Section 4.3 discusses the quality
assurance procedures used during the data collection effort. Section 4.4 describes the data processing
quality assurance procedures. The last section describes the quality assurance steps taken during the
preparation of this report.
4.1 Draft Questionnaire Pre-test and Survey Pilot Test
A significant component of the survey quality assurance plan was to thoroughly test the
questionnaire design, the survey design, and data collection procedures prior to implementing the full
study. Confirming the validity and effectiveness of these designs, or revising them when the tests
revealed problems, errors, or difficulties, led to design and process improvements that would have a
positive effect on the quality of the survey in such areas as data reliability, data completeness, accuracy
of the sample frame, and response rates.
4.1.1 Pre-test
When the initial data collection objectives had been identified and the questionnaire shaped into
a working draft instrument, EPA conducted a pre-test of this draft with 7 water systems in New
England of various sizes, including ground and surface water systems. The pre-test participants were
recruited with the assistance of Ray Raposa of the New England Water Works Association. The main
objective of the pre-test was to gauge the respondents' reactions to the questionnaire itself. The test
did not address any of the actual survey operations and response rate issues that would later be tested
in the full-scale pilot test.
147
-------
The recruited systems received the questionnaire in July, 2000. EPA then convened a focus
group meeting of the 7 water systems, facilitated by survey research staff from Abt Associates and
Cadmus. The focus group explored questions regarding comprehensibility, use of clear and
appropriate terminology, provision of suitable response categories, and questionnaire layout. The focus
group also discussed respondents' ease or difficulty in providing answers, their immediate knowledge of
or access to information requested by the questionnaire, and their overall reaction to the survey.
Overall, the focus group felt the questionnaire was clear and relatively easy to follow. As a
result of the pre-test, some questions were re-worded, and others were shortened. Otherwise, the pre-
test found no systematic problems in the respondents' ability to provide answers to the questions.
4.1.2 Pilot Test
A full scale pilot test was conducted in April and May, 2001. The pilot tested the questionnaire
and the major operational components of the survey design. The results of the pilot, along with the final
version of the questionnaires were delivered to EPA in May, 2001. The full on-line tracking system
was developed during the pilot, and the mail-out and receipt logging procedures were finalized.
Twelve small systems and 42 medium, large, and very large systems were selected from the full
sample for use in the pilot. Of these, 10 of the small systems and 26 of the medium, large, and very
large systems responded. The response rate was consistent with the target rate for the survey as a
whole.
As a result of the pilot, modest changes were made to the mail-out process and the instructions
for systems. The pilot also resulted in changes to several questions in the questionnaires. Questions 19
(length of distribution mains) and 32 (capital improvements) were simplified. Modest changes were
made to several other questions to clarify the question. The pilot also finalized the site visit protocols,
and identified issues that needed to be addressee! when training the site visitors.
4.2 Sampling Quality Assurance
Quality assurance of the sampling process for the CWSS involved three principal areas:
Development of the sample frame
Sampling specifications, and
Use of software designed to draw complex samples.
Development of the Sample Frame. EPA conducted an extensive review of the data used
for the sample frame. By starting with the data used for the 1999 DWINS frame, the 2000 CWSS
was able to take advantage of the extensive data verification effort undertaken the 1999 DWINS. The
1999 DWINS frame was updated with data collected through the survey. This updated frame was
148
-------
then compared to the data in SDWIS in the first quarter of 2000 to identify systems that were either
added or removed from SDWIS since the DWINS was completed. The development of the frame is
discussed in detail in section 2.1.
Sampling Specifications. In order to carry out the sampling processes, the survey statisticians
prepared detailed specifications that served as directions for performing the sampling and as a
permanent documentation of the process. The sampling plan was documented in both the supporting
materials for the Information Collection Request submitted to the Office of Management and Budget,
and in the Quality Assurance Project Plan. The specifications ensured the sample was drawn in
conformity with the sample design and in a statistically valid manner.
Sampling Software. The CWSS sample of systems serving up to 3,300 people was drawn
using SAS-based program designed to draw two-stage cluster samples of this type. The same program
was used to draw the 1999 DWINS. The sample of systems serving population s of 3,301 to 100,000
was a stratified random sample and was drawn using Stata-based program to select random samples.
4.3 Data Collection Quality Assurance
Each component of the CWS survey was implemented pursuant to detailed written
specifications that clearly stipulated how the design was to be implemented.
Questionnaire Design
The various drafts of the questionnaires were the product of close review and
comments by EPA, Cadmus, and outside reviewers. Improvements also were made as
a result of the pre-test and pilot test.
Questionnaire version control was maintained through the various drafts by hand-
writing all changes onto the hard copy master of the current version. After the changes
were made to the master word processing file, the previous hard copy version was
filed. Each version was dated and serially numbered.
The questionnaire form was designed to clarify and simplify for respondents the
provision of the highly detailed and complex data required for the survey. Graphic
devices were used to make the form clearer and simpler to use. The devices included
type fonts and sizes, borders, and text boxes.
Because of the difficulties small systems have with filling out complex questionnaires like
the CWSS, site visitors were sent to small systems to ensure the questionnaires were
filled out correctly.
149
-------
Mail Data Collection
Workers preparing the questionnaire for mailing were provided with detailed written
specifications for the job and were supervised by a mail operations manager.
Mail clerks worked in tandem to produce the packages to be mailed to each
respondent. One clerk applied identification labels and assembled the packet; the
second applied the corresponding address label to the mail out envelope. This
procedure effectively provided a 100% check that the mailing and questionnaire labels
were for the same water system.
A survey manager conducted spot checks of the questionnaires before they were
sealed to ensure the respondent was receiving the appropriate form and that the
address label and CWS information label matched.
The prepared questionnaires were counted prior to mailing to verify that the correct
number of questionnaires were mailed.
Each recipient of the mailed questionnaire was assigned an analyst who maintained
contact with the water system throughout the survey. The analysts provided reminder
calls and technical support to the systems. They also reviewed the data as it was
received, following up with the system if there were any questions.
Senior survey managers reviewed at least the first 25 surveys received to ensure
analysts were using consistent procedures for each survey.
The on-line tracking system ensured proper tracking and control of all questionnaires
from the point of sampling until the data were entered and cleaned. In addition to
supporting overall management of the project, the periodic status reports identified
response rate problem areas which enabled Cadmus to take appropriate follow-up
measures.
Site Visits
Extensive training was provided to the site visitors, including on-site training at a cluster
of small systems.
Detailed instructions were provided to each site visitor regarding the conduct of the on-
site surveys.
Regular contact was maintained with all site visitors. Site visitor questions and Cadmus
150
-------
responses were sent to all site visitors to ensure each received complete and consistent
information.
Each completed survey was reviewed by Cadmus staff as it was received. Follow-up
instructions were provided as needed.
4.4 Expert Review of Responses
Each questionnaire was subjected to a multi-level, detailed review by Cadmus staff as it was
returned by the systems. Cadmus reviewed the questionnaire for completeness and internal
consistency. Systems were called if key questions were not answered or if answers were inconsistent
or unclear.
Upon receipt of the completed questionnaire, the Cadmus analyst responsible for the system
reviewed the survey. They identified missing information and questions or potential problems with
responses. The analysts were provided training on how to evaluate a completed questionnaire, as well
as written guidance for reviewing the responses. The written guidance included rules-of-thumb for
internal consistency checks; these guidelines helped the analyst compare questions and identify
inconsistent answers. For example, guidelines were provided on average annual water consumption
per household, which were used to compare annual water production with the number of connections
reported.
Guidelines were provided regarding follow-up questions for the system. If essential data on
system finance, treatment, and production were missing, or if inconsistencies could not be resolved,
analysts contacted the system. If detailed information was not available (e.g., revenue by customer
class), analysts attempted to collect more aggregate-level data (e.g., total water sales revenue.)
Analysts worked with the systems to resolve inconsistencies. Senior staff contacted systems when
difficult issues arose. Changes to the questionnaire were documented and logged.
The analysts review of the surveys was itself reviewed by senior survey staff. Senior survey
evaluated the analysts' reviews at the beginning of the review process, and and provided feedback to
the analysts. Senior staff and water system experts provided information and answered questions
throughout the data collection period.
Upon the completion of the analyst's review of a questionnaire, the completed questionnaire
was then reviewed by Cadmus water system experts. Each question in the survey was subject to
review. The expert review focused on the validity of the responses to each question (e.g., checking that
the treatment sequence is logical), consistency across quesitons (e.g., the treatment practice is
consistent with the treatment objectives), and that questions were answered and reviewed consistently
across by water systems. Any iurther changes were documented and logged.
151
-------
4.5 Data Processing Quality Assurance
The completed surveys were edited and entered into an electronic database.. The electronic
data were then imported into a hierarchical database for distribution, and a statistical package for
detailed analysis. Procedures were in place at every step to maintain the integrity and quality of the
data.
4.5.1 Manual Editing, Coding, and Data Entry
Following the expert review, the questionnaires were subjected to a 100 percent editing review
in preparation for entering the data. This editing process examined every response field on every form,
to check skip patterns, clarify handwriting that would be difficult for the data entry staff to read,
standardize the recording of quantitative data, and identify any potential problems, such as marginal
notes or potential order-of-magnitude reporting errors in the volumetric questions. General protocols
were developed to guide the data preparation staff in reviewing the forms and in handling generic
problems. Data Preparation supervisors performed a 100% quality control review of edited forms
before they were data entered. During the editing process any questions, including the creation of
open-ended codes, that could not be answered by the general protocols were passed on to Cadmus by
staff at Abt Associates for final decisions. The editing protocol was updated to reflect coding decisions.
After the initial edit, the questionnaires moved to the data entry process. Each form was
entered with 100 percent verification, that is, using independent double key entry. The automated data
entry program was customized to each form.
As the data were entered, the batches of entered records passed through a data cleaning
process, consisting of standard computer edits that examined each variable for conformity to
appropriate values or data ranges and also checked the small number of skip patterns that existed in the
survey instruments. A report identified each variable for each case that failed any of these tests. A data
preparation supervisor then examined the original questionnaire forms to determine whether the
anomaly occurred in the original data and, if so, whether to confirm it as correct or to refer to Cadmus
for resolution as described above. The standard computerized edits were rq^eated for all the data until
no cases failed the edits, except for any that had been specifically confirmed as valid outliers during a
previous review.
After all questions were edited, entered, and cleaned to the degree permitted by these
processes, the resulting keyed data passed to a process of detailed automated logical edits that enabled
expert staff to conduct a highly focused review of data values and relationships.
152
-------
4.5.2 Automated Data Validation Checks
In preparing the final database, EPA, Cadmus, and Abt designed, produced, and analyzed a
series of computer validation checks. These checks were run on the full survey database after the data
were entered and passed the standard computer edits for values and ranges on a variable-by-variable
basis. The checks included the following:
Distribution frequencies for all categorical variables;
Distribution frequencies for all continuous numerical variables formatted into four
categories (non-zero responses, zero responses, legitimately skipped, and missing);
Univariates for each continuous variable;
Item-specific cross-tabulations of categorical variables;
Item-specific cross-univariates of continuous data; and
Item-specific advanced logic edits.
4.5.3 Database Quality Assurance
The final, clean survey database represented the product of the various review, editing, data
entry, and data validation steps described above. Once the database was prepared, there were a
number of subsequent data processing steps required to create a variety of files suitable for analyses
and tabulations for the final delivery of a permanent database to EPA. The principal steps included:
Appending needed variables from external files, including sample and contact
information from SDWIS.
Analyzing the hard copy questionnaires and the frequency distributions of continuous
and categorical variables to devise rules for handling missing data.
Zero-filling blank responses. A detailed series of rules was developed for assessing
blank responses and determining whether to regard these as zeros or missing values. In
general, blank quantity fields were treated as zero, except when there was external
evidence in a logically related item that the response should not be zero. A detailed set
of programming specifications was designed to implement these rules.
Creating new derived variables from the survey data to categorize systems into strata
comparable to the original sampling strata but based on the final survey responses
153
-------
rather than the SDWIS data.
Attaching the sample weights to the analytical file.
For the final delivery of the database to EPA, deriving and attaching the numerous
composite variables created for the production of the analytical tables in this report.
Each step was planned in advance. Detailed specifications were written to guide the
programming and data processing needed to peiform each step. In addition to these specifications, the
processing of files and flow of data throughout these steps were planned, controlled, and documented
through data flow diagrams. The diagrams are schematic representations of how files, data record,
data elements, and individual data point values are handled, combined, extracted, and moved from one
stage to the next. These diagrams are crucial quality assurance tools to help ensure that programmers
and systems analysts have aclear an common understanding of the entire process of data management,
that the processing stages fit together in a logical order and accomplish the intended objectives, and that
there is an unambiguous audit trail of the condition of the data at each stage.
Version control was maintained for all computer programs, an interim stages of all data files
were permanently archived. This meant that when changes were made to a program or process, it was
clear which was the current version and it always was clear of sequential changes that had been made
from one version to the next. It was always possible to restore any earlier version in full or to merge
selected data from the old version to the new version.
The combination of the processing specifications, data flow diagrams, version control, and data
archiving ensured that no process was irreversible, that it was always possible to recover from any
deliberate or inadvertent changes to the data, and that the characteristics of the survey data were fully
known at each processing stage.
4.5.4 Tabulation Quality Assurance
The tabulations of the results presented in the tables in this report are varied and complex.
Rather than being a simple presentation of individual survey variables, each table usually presents the
results of multiple calculations involving several survey variables. Many tables present several such
results in a single table. There often were several different ways of defining or calculating an item of
interest, and sometimes there were different direct or derived sources of data for the calculation
available on the survey database. Hence, the following steps were taken to help assure that each table
accurately summarized and presented the data contained in the final survey database.
Identify important, relevant, and useful information that could be developed from
analyses of the survey data;
154
-------
Design each table to effectively present the results or to juxtapose related results in the
same table;
Clearly describe the contents of each table;
Define in detail the variables, values, formulas, and derivations that went into each
calculation;
Prepare clear and detailed data processing specifications for carrying out the
tabulations according to the calculation definitions;
Develop computer programs to process the data pursuant to the tabulation
specifications;
Review the initial tabular output for:
Consistency with the design of the table of contents;
- Conformity with the definitional and programming specifications; and
- Reasonable agreement with expected valuesbased on external measures and
expert knowledge of water system operations and finance;
Review definitions, specifications, programs, and underlying data for tabulations
exhibiting data anomalies or outliers;
Review any definitions, specifications, or programs if the review process identifies
errors or the need for modifications to previous decisions; and
Repeat previous tabulation quality assurance steps and re-run tabulations until no further
unacceptable data anomalies are found.
The tabulation process was fully automated, from the underlying source data through all
processing stages to the final formatted tables. There were no intermediate stages requiring manual
transfer or entry of data from one stage to the next. This eliminated human transcription error. Of
equal importance, it also expedited the process of successive iterations of the tabulations during the
quality review process, as each time a table was produced the output data automatically were
transferred into the same final table form as on the previous iteration. This ensured that any new
anomalies identified in later iterations did not result from transcription errors, and allowed the review
staff to focus their investigations on the table data, specifications, and programs.
155
-------
4.6 Quality Assurance During Report Preparation
EPA requested comments on the survey from several independent reviewers. EPA consulted
with Jeanne Bailey in the office of Regulatory Affairs at the American Water Works Association in
Washington, DC. Drafts of the questionnaire were reviewed by Diane Moles, from the Iowa
Department of Natural Resources, in Des Moines, IA, and James K. Cleland of the Drinking Water
and Radiological Protection Division of the Micliigan DEQ in Lansing, MI. EPA also consulted with
Robert W. Mann of the Water Supply Engineering Bureau of the State of New Hampshire's
Department of Environmental Services. As noted in the introduction to this part of the report, EPA's
Office of Environmental Information reviewed the data presented in this report to ensure the findings
were appropriately described and presented. Additional peer review was provided by Jan Beecher of
Michigan State University and John Petersen of George Mason University.
156
-------
Appendix A
Community Water System Survey
Questionnaire
157
-------
United States
Environmental Protection Agency
< r**f
5SB
SURVEY OF SMALL, MEDIUM, LARGE,
AND VERY LARGE
COMMUNITY WATER SYSTEMS
(Composite of questionnaire sent to small, medium, and large systems,
and to very large systems)
OMB No. 2040-0227
Expiration date: 2/29/04
-------
Please return this questionnaire in the enclosed Federal Express envelope
or mail to:
EPA Community Water System Survey
c/o The Cadmus Group, Inc.
135 Beaver St., Suite 2
Waltham, MA 02452
Participation in the survey is voluntary. However, as a matter of policy, EPA will not disclose the identity of any
respondent to this questionnaire, nor the identity of any participating water system. While no respondent has ever
claimed that the information asked for in this survey contains confidential business information (CBI), EPA will offer
you the opportunity of claiming CBI in the event that we receive a Freedom of Information Act request for any data
that would identify you or your system. It should be noted, however, that EPA has never received a Freedom of
Information Act request for such information in prior surveys.
The public reporting and record keeping burden for this collection of information is estimated to average 2.27 hours
per response or to range from 1 hour to 4 hours per respondent annually. Burden means the total time, effort, or
financial resources expended by persons to generate, maintain, retain, or disclose or provide information to or for a
Federal agency. This includes the time needed to review instructions; develop, acquire, install, and utilize
technology and systems for the purposes of collecting, validating, and verifying information, processing and
maintaining information, and disclosing and providing information; adjust the existing ways to comply with any
previously applicable instructions and requirements; train personnel to be able to respond to a collection of
information; search data sources; complete and review the collection of information; and transmit or otherwise
disclose the information. An agency may not conduct or sponsor, and a person is not required to respond to, a
collection of information unless it displays a currently valid OMB control number. The OMB control numbers for
EPA's regulations are listed in 40 CFR Part 9 and 48 CFR Chapter 15.
If you wish, you may send comments on the Agency's need for this information, the accuracy of the provided
burden estimates, and any suggested methods for minimizing respondent burden, including through the use of
automated collection techniques to the Director, Collection Strategies Division, U.S. Environmental Protection
Agency (2822), 1200 Pennsylvania Ave., NW, Washington, D.C. 20460; and to the Office of Information and
Regulatory Affairs, Office of Management and Budget, 725 17th Street, NW, Washington, DC 20503, Attention: Desk
Officer for EPA. Include the EPA ICR number and OMB control number in any correspondence. Do not send the
completed survey to this address.
-------
Dear Owners and Operators of Community Water Systems:
The United States Environmental Protection Agency (EPA) is conducting a national survey of drinking water
systems using the attached questionnaire. About 1,500 water systems have been randomly selected to participate in
this survey, and yours was one such system. This survey is conducted approximately every five years, the last one
being in 1995. We are sending you this questionnaire because you were identified in your state's database (State
Drinking Water Information System) as the most appropriate person to provide information about your water system.
Participation in the survey is voluntary.
This survey will accomplish a number of important objectives. First, it will give us current data that will allow us to
better consider the costs and benefits to water systems when we develop new national drinking water regulations. It
will also allow us to measure the impact of drinking water regulations that have been put in place since the last
survey. This, in turn, will help us determine more affordable approaches to drinking water treatment. Furthermore,
the answers you provide in this questionnaire will help us in developing more effective programs to safeguard our
nation's drinking water, provide guidance to the states and measure the effectiveness of programs already in
existence, such as the Drinking Water State Revolving Fund.
As we have done in the past, EPA will only make use of the information you provide when it has been aggregated
with the responses of many other water systems in the same size category as yours. We will never disclose your
name or the name of your water system in any public documents. Please see the inside cover of the questionnaire if
you'd like more details on how your privacy will be protected.
Answers to this questionnaire will help EPA to understand your circumstances better than any other single tool we
have. If you have ever wanted to have a larger say in the development of national rules that could directly effect
you and your water system, providing answers to this questionnaire is an important contribution. Because only
1,500 of you are being asked to speak for over 50,000 other systems, your voice is that much more important and will
carry that much more weight. If you have ever felt that Federal regulators don't understand your situation, then
please take this opportunity to tell us, in detail, just what your situation is. It will make a difference.
Sincerely,
Brian C. Rourke
Program Analyst
Standards and Risk Management Division
-------
Please respond about:
Please return you completed questionnaire in the enclosed
pre-paid Federal Express envelope by July 13,2001
-------
2000 Community Water Systems Survey
Small, Medium Large / Very Large Systems Questionnaire
GENERAL INSTRUCTIONS
This questionnaire asks two preliminary questions and then is divided into two parts.
PART I - OPERATING CHARACTERISTICS (Questions 3 25); and
PART II - FINANCIAL CHARACTERISTICS (Questions 26 - 32).
Please complete the questionnaire.
Make a copy of the completed questionnaire for your records before sealing it in the enclosed envelope.
Please include a map of your service area or delineate your service area on the enclosed map.
[VERY LARGE SYSTEMS: ]
Please enclose a schematic of your system, site plans of your treatment facilities, and the latest available financial
report. If the schematics, diagrams, financial or other reports contain the information requested by a question,
you may refer to the documentation rather than fill out the question.
[SMALL/MEDIUM/LARGE SYSTEMS]:
Your are strongly encouraged to enclose schematics, diagrams, financial reports, or other information that will
help provide a complete picture of your water system. If schematics, diagrams, financial or other reports contain
the information requested by a question, you may enclose and refer to the documentation rather than fill out the
question.
If you require more space to answer an question than is provided, please record the information on a photocopy
of the question or use a blank sheet of your own.
Page 1 of 22
-------
1. Please provide the name, title, and telephone number of the most knowledgeable person to contact for
information on:
(A) Part I - Operating Characteristics
Name: Title:
Tel. No. Fax No.
e-mail:
(B) Part II - Financial Characteristics
(Write "SAME" if same as above)
Name: Title:
Tel. No. Fax No.
e-mail:
2. This survey will ask you to provide operating and financial information for your public water system for the
most recent 12-month period for which data are available. Please specify below the end dates for which data are
provided.
A. Operating information / /
(mm / dd / yy)
B. Financial information / /
(mm / dd / yy)
Part I - Operating Characteristics
For Part I of the survey, please use the period indicated in Question 2(a) to report "last year's"
operating data.
3. Please classify your water system using the following criteria (circle one).
Owned or operated by a government or
public agency (including government-owned systems that hire a
private company to operate the system) 1
Owned privately and operated for profit primarily as a
water business 2
Owned privately and not operated for profit (e.g., a homeowners
association or a non-profit cooperative) 3
Owned privately and operated as a necessary
part of another business (e.g., a mobile home park) 4
Page 2 of22
-------
4. If your system is owned or operated by the government, please select one of following that best describes the
form of government (circle one).
A local or municipal government (e.g. towns, townships, cities,
counties, boroughs, parishes, and special districts) 1
State government 2
The Federal government 3
Some other government
(Please specify) 4
A. PRODUCTION and TREATMENT
5. What was the amount of water that was produced and delivered to each of the following customer categories
during the last year [as defined by your answer to Question 2A]? (In millions of gallons. Note: if you cannot
distinguish among the different types of non-residential customers, enter the total for your non-residential
customers in line c.4.)
Customer Type
Total
a. Sold to other public water suppliers:
1. Treated water Million Gallons/Yr.
2. Untreated water Million Gallons/Yr.
b. Residential Million Gallons/Yr.
c. Non-residential
1. Commercial/industrial Million Gallons/Yr.
2. Agricultural Million Gallons/Yr.
3. Other (specify) Million Gallons/Yr
4. Subtotal, non-residential Million Gallons/Yr
d. Unaccounted for water not included above (including
uncompensated usage and system losses) Million Gallons/Yr
e. Total, all customer types (including unaccounted for water) .... Million Gallons/Yr
6. Provide the name of each public water supplier included in the response to Question 5a, above.
a. Treated water
b. Untreated water
Page 3 of 22
-------
7. How much of the water reported in Question 5 came from each of the following sources during the twelve month
period reported in Question 2 A ? (in millions of gallons; answer 'none' if a source does not apply).
a. Surface water (non-purchased) (including Ground Water Under the
Direct Influence of Surface Water) Million Gallons/Yr.
b. Ground water (non-purchased) Million Gallons/Yr.
c. Purchased water
1) Treated water Million Gallons/Yr.
2) Untreated water Million Gallons/Yr.
d. Other (specify) Million Gallons/Yr.
e. Total Million Gallons/Yr
8. What was the maximum daily water production from all sources for this
utility over a single 24 hour period during the twel ve month period
reported in Question 2A? Million Gallons/day
The following definitions of the components of a water system are used in this survey. Figure 1 is
an example of a schematic of a water system, showing water sources, treatment plants,
transmission lines, and the distribution system.
Please refer to these definitions and the schematic for an explanation of the terms used in
questions 10 through 18. Please submit diagrams or schematics, using figure 1 as a guide.
Please note that the identifier numbers used in the questions do not refer to specific items in the
schematic. For example, use 'SI1 to refer to your first surface water source, regardless of whether it
is a flowing stream, as depicted in the schematic, or another surface water source.
Page 4 of 22
-------
Term
Surface water
intake
Ground water
intake
Purchased
water intake
Water
treatment plant
Entry point
Definition
A surface water intake refers to the transmission of untreated water from a
surface water source (flowing stream, lake, reservoir, or ground water under the
direct influence of surface water) to a water treatment plant at the utility (see
accompanying diagram).
A ground water intake refers to the transmission of untreated water from one or
more wells to a water treatment plant or directly into the distribution system.
Where the water from multiple wells flows through a common pipe prior to entry
into the treatment plant or distribution system, the combined flow is considered
one ground water intake (see accompanying diagram).
A purchased water intake refers to the transmission of water from the seller's
utility to a water treatment plant or directly into the distribution system of the
purchaser's utility.
A facility where water is filtered, disinfected, and/or otherwise treated prior to its
transmission into the distribution system (or its conveyance to another
purchasing water utility). For the purposes of this survey, simple disinfection
only or pH adjustment prior to entry into the distribution system are considered
to be a water treatment plant.
An entry point is where treated or untreated potable water enters into the
utility's distribution system.
Code in the
Schematic
S1,S2
G1-G4
P1.P2
WTP 1,
WTP2
E1-E4
Figure 1: Sample Diagram of Intakes, Treatment Plants, and Entry Points
Flowing
Stream
Page 5 of22
-------
9. Please draw your schematic here or submit a schematic on a separate sheet of paper.
Page 6 of 22
-------
10. Provide the following information for the surface water, ground water, and purchased water intakes for this
utility: MOD refers to millions of gallons daily. If the source is used on a seasonal basis, the average daily
amount is for the months in which water is drawn from the source.
A.
Surface Source.
Surface
water
intake
identifiers
What is the source type for each
surface water intake'' (circle the
appropriate number)
Flowing
stream
Reservoir
or lake
Ground-
water
under the
direct
influence
of surface
water
Is this a
seasonal
source'7
(Circle 1
for yes
and 2 for
no)
Yes No
What was the
average daily
amount of water
drawn from each
surface water
intake during the
reporting period
in Question 2A9
(MGD)
What is the estimated
potential maximum
daily amount of water
that can be drawn
from each surface
water intake, given
water availability
constraints only (e g
source capacity
contractual
obligations, permits,
or legal constraints)9
What is the estimated
potential maximum
daily amount of water
that can be drawn from
each surface water
intake, given current
pumping and
equipment constraints
only (e g system
components and
physical constraints)1'
(MGD)
SI 1 2 312 (MOD)*
S2 1 2 312
S3 I 2 312
S4 1 2 312
S5 1 2 312
* 1 f not limited enter "no limit" here
Q10A Totals
B. Ground water source.
Ground
water
wells
How many
individual
wells
supply
each
ground
water
intake?
/,,.:;,•
SVv;.'^
( ?'~, •
H 'i; jx
v.,,:
<< rJ' ' <
U o* '
•-' • •:->•
^ . ,'"'N
, ) , ' >. •;.>
nii.ll.O .i
,-:,irv,!
i,i
-------
C. Purchased water source.
Purchased
water
intake
identifiers
Provide the names of the sellers for this
water
Is this a
sCiisonal
source9
(Circle 1
for yes and
2 for no)
Yes No
What was the
average daily
amount of water
drawn from each
purchased water
intake during the
reporting period in
Question 2Ar>
(MOD)
What is the
estimated
potential
maximum
amount drawn
from each
purchased water
intake given water
availability
constraints only
(e g. source
capacity
contractual
obligations.
permits, or legal
constraints)?
(MOD)*
What is the
estimated
potential
maximum
amount drawn
from each
purchased water
intake, given
current pumping
and equipment
constraints only
(e g. system
components and
physical
constraints)9
(MOD)
PI
P2
P3
P4
P5
1
Ql DC Totals
Totals for all intakes Q10A-Q10C (MGD)
* If not limited enter "no limit" here
11. Does your system provide treatment? (Circle one)
Yes 1
No (go to Question 18) 2
Page 8 of22
-------
12. Provide the following information for each water treatment plant or other facility at this utility. Design Capacity
refers to the maximum amount the plant can produce in a single 24 hour period with all treatment trains operating
at capacity. Peak daily production refers to the maximum amount produced in a single day over the twelve month
reporting period in Question 2A. Please submit site plans and flow charts of each treatment facility in your
water system.
Water treatment plant
identifiers
List all of the surface,
ground, and purchased
water intake identifiers
from Question 10 that
feed into each water
treatment plant or
other facility.
What was the average
daily production for
each water treatment
plant or other facility?
(MGD)
What was the peak
daily production for
each water treatment
plant or other facility?
(MGD)
What was the design
capacity for each water
treatment plant or
other facility?
(MGD)
WTP1
WTP2
WTP3
WTP4
WTP5
Note: WTP refers to a treatment plant or any other facility that provides treatment.
13. Using the water treatment plant identifiers from Question 12, indicate which water treatment objectives apply to
each plant. (Circle 1 for Yes and 2 for No).
Do you have this treatment objective in the following
water treatment plant?
(use the plant number from Question 1 2)
WTP1
Yes No
WTP2 WTP3
Yes No Yes No
WTP4
Yes No
WTP5
Yes No
Algae control
Corrosion control
Disinfection
Dechlorination
Oxidation
Iron removal
Manganese removal
Fluoridation
Taste/odor control
TOC removal
Particulate/Turbidity Removal
Softening (hardness removal)
Recarbonation
Organic contaminant removal (e.g., VOCs, pesticides)
Inorganic contaminant removal (e.g., arsenic)
Radionuclides contaminant removal
Other (specify)
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2 1 2
2 1 2
1
1
1
1
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
Page 9 of 22
-------
14 A. Using the Water Treatment plant identifiers from Question 12, characterize the treatment used and the
sequence of treatment for each plant by entering a number to identify the order in which each treatment
process occurs for each water treatment plant. (See example. If you have the option of more than one
treatment type for a single step — e.g., you can use either chlorine or chlorine dioxide for disinfection —
assign the same sequence number to each alternative).
Water Treatment Plant Number (from
Question 12)
Treatment Category
Example
WTP1 WTP2 WTP3 WTP4 WTP5
fhlorination only
Raw water storagp/Prpsptiimpntation basin
PrpHisinfprtion/o\idation nrior tn spriimpntation
Phlnrinp
Chlorine dioxide
Phlnramines
Potassium nermanoanate
Other PreHisinfertion
PrpHisinfprtinn/O\iriatinn prior tn filtration
Chlorine
Chlorine dirrxifle
rhlnramines
Qynne
Potassium permanganate
Other Predisinfectinn
RapiH
Poaoiilation/ Florrnlation
Polvmprs
Spftlinp/Spriimpntatiftii
Snftpning
T i
ash
Rpnarhonation
Trm
Filtration
Direct filtratinn
Mirrn strainer
/ sand
Rag and
Ranifl tanH
arth
Diial/Miilti media
Pressure filtration
Othe
Pogt-di«infpption aftpr filfprg
Phlr
Hinvide
IIV
Page 10 of22
-------
Water Treatment Plant Number (from
Question 12)
Treatment Category
Vlpmhranps
WTP1 WTP2 WTP3 WTP4 WTP5
Reverse osmosis
Mirrrv filtration
I Iltrafiltratinn
Nanofiltration
("nrrfisinn (""nntrnl
Ion exrhanpe
("irannlar artivaterl rar
Artivaterl Alumina
Aeration
I This is an example of a green sand filter plant for treating ground water for Iron and Manganese removal In this example
'other ' is contact basin
B. Using the Water Treatment plant identifiers from Question 12, indicate which filter backwash techniques
you use for each treatment plant that uses filtration. (Check all that apply).
Filter backwash
Air sronrinp
Slurfarp wash
Reryrlp filtpr harkwash
Filter to waste
Other filter harkwash
ISIrmp
Example
•Y
"X
T
Water Treatment Plant Number (from Question 12)
WTP1
WTP2
WTP3
WTP4
WTP5
Page 11 of22
-------
[QUESTION 15 ASKED OF VERY LARGE SYSTEMS ONLY]
15. Using the identifiers from questions 10 and 12, please list the intake or entry point identifier number and
concentration for each water quality parameter (contaminant) for the reporting period recorded in Question 2A.
If you conducted multiple tests of a source over the reporting period, report the average concentration. If you
did not test for a contaminant mark N/A; if you did not detect a contaminant mark ND. If you require additional
space for entry points/well identifiers please attach an additional sheet.
A. Using the source identifiers from questions 10, please provide raw water concentration in units of parts per
million (ppm) for each compound:
Intake
Identifier
Example 1 :
G-l/W-1
Example 2:
G-l/W-2
Raw Water Concentration (units- ppm, except Radon - pCi/L)
Arsenic
<0.002
0.002
Radon
N/A
100
MTBE
N/A
N/A
Atrazine
N/A
0.002
Metolachlor
N/A
N/A
Boron
N/A
N/A
2,4-D
ND
ND
Stmazine
0.001
ND
Glyphosat
e
0.01
ND
Page 12 of22
-------
B. Using the treatment plant identifiers from question 12, please provide post-treatment concentration in units
of parts per million (ppm) for each compound:
Entry
Point
Identifier
Example 1 :
WTP-1
Example 3:
WTP-2
Treated Water Concentration (units- ppm, except Radon - pCi/L)
Arsenic
<0.002
0.002
Radon
N/A
ND
MTBE
N/A
N/A
Atrazine
ND
ND
Metolachlor
N/A
N/A
Boron
N/A
N/A
2,4-D
ND
ND
Simazine
0.001
ND
Glyphosat
e
0.01
ND
Page 13 of22
-------
16 A. Using the water treatment facility identifiers from Question 12, indicate if the specified residuals
management practices are used and provide the requested information regarding potential discharge.
(Circle I for Yes and 2 for No).
Residual Management
Category
Mechanical dewatering
Non-mechanical dewatering
Chemical precipitation
Land application
Nonhazardous waste landfill
Deep well injection
Evaporation pond
French drain
Direct discharge to surface water
If no, is direct discharge to surface water an option?
Septic system
If no, is discharge to a septic system an option? ...
Sanitary sewer
If no, is discharge to a sanitary sewer an option? . ..
Do you use the following residual management process in the
following water treatment plants?
(use the water treatment plant numbers from Question 12)
WTP1 W1
Yes No Yes
1 2 1
1 2 1
1 2 1
1 2 I
1 2 1
1 2 1
1 2 1
1 2 1
1 2 1
1 2 1
1 2 1
1 2 1
1 2 1
1 2 1
PP2 WTP3
No Yes No
2 1 2
2 1 2
2 1 2
2 1 2
2 1 2
2 1 2
2 1 2
2 1 2
2 1 2
2 1 2
2 1 2
2 1 2
2 1 2
2 1 2
WTP4
Yes No
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
WTP5
Yes No
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
B. Please describe any current limitations on the use of direct discharge to surface water, septic systems, and
discharge to sanitary sewer for residuals management at any of the water treatment plants at this utility.
Page 14 of 22
-------
17. Using the water treatment plant identifiers from Question 12, provide the following information regarding
operators and SCADA usage at each water treatment plant at this utility. (SCADA Supervisory Control and
Data Acquisition system is an automated system for monitoring, controlling, and /or transmitting information
on water treatment plant processes).
Operator Information
Water Treatment Plant
(use the water treatment plant numbers from
Question 12)
WTP1
Yes No
WTP2
Yes No
WTP3
Yes No
WTP4
Yes No
WTP5
Yes No
1. Is there an operator on-site 24 hours per day seven days per
week? (Circle yes or no)
2. If the above answer is no, estimate the total number of hours
per week that an operator is on site
3. Is there a SCADA in use for process monitoring? (Circle ves
or no)
4. Is there a SCADA in use for process control? (Circleyes or
no)
B. STORAGE AND DISTRIBUTION SYSTEM INFORMATION
18. Please indicate whether you have the following types of treated-water storage, and if so, the number of storage
facilities and their capacity, in millions of gallons.
A. Do you have clearwell storage after treatment?
B. Do you have storage after treatment (and after clearwell, if any), but before
the distribution system with dedicated entry and exit points?
C. Do you have storage after treatment (and after clearwell, if any), but before
the distribution system with a common inlet and outlet (i.e., rides the line)
D. Do you have storage within the distribution system with dedicated entry
and exit points?
E. Do you have storage within the distribution system with a common inlet
and outlet (i.e., rides the line)?
Yes
1
No
If yes, what is the
total capacity of the
storage (in millions
of gallons)?
Page 15 of 22
-------
19 A. Estimate the length of the distribution mains pipe in your system, and length of pipe replaced in the last five
years.
Pipe Diameter
Less than or equal to
6"
Greater than 6" but
less than or equal to
10"
Greater than 10"
Length of Pipe
(In Miles)
Length of Pipe Replaced in
the Last 5 Years'
(In Miles)
Total Cost of Pipe Replaced in
the Last 5 Years'
(In Dollars)
1 .Ending on the date shown in your answer to Question 2A.
B. What percentage of the total length of pipe in the distribution system is less than 40 years old, between 40
and 80 years old, and more than 80 years old? (Thepercentages should total 100percent.)
1. Less than 40 years old %
2. Between 40 and 80 years old
3. More than 80 years old
20. A. How many people and connections does your system currently serve year round? Please indicate the
number of connections and number of people served by your system for all customer types that apply. (If
you cannot distinguish among the different types of non-residential customers, enter the total for your
non-residential customers in line c.4.)
b.
c.
Customer Type
Sold to other public water suppliers
1. Treated water
2. Untreated water
Residential
Non-residential
1. Commercial/industrial ..
2. Agricultural
3. Other (specify)
Connections
Number of
People
4. Subtotal, non-residential
Page 16 of22
-------
[QUESTIONS 20 B. AND 20 C. ASKED OF SMALL, MEDIUM, LARGE SYSTEMS ONLY]
B. Does your system serve a residential population that changes on a seasonal basis (for example, is it a
winter or summer resort area)? (Circle one)
Yes 1
No (Go to Question 21) 2
C. If your system serves a population that changes on a seasonal basis, please indicate the highest seasonal
number of people (residential only), the highest number of residential connections and the number of
months each year during which the seasonal population is the highest.
1. Highest seasonal population
2. Highest number of seasonal residential connections
3. Number of months when seasonal population is highest ....
21. Please enclose a map of your service area
C. CROSS CONNECTION CONTROL
22. Does your system have a cross connection control (CCC) program? (Circle one)
Yes 1
No (Go to Question 26) 2
Don't Know (Go to Question 26) 3
23. Please indicated the type of program your system has. (Circle one)
Containment (A program that is designed to prevent backflow from reaching a publicly
owned distribution system, but does not provide protection within the
premises. This often is referred to as providing protection up to the meter.)
Containment (A program that is designed to prevent backflow from reaching a publicly
and isolation owned distribution system and provides protection within a customer's
premises. This often is referred to as providing protection up to the tap.) ...
Page 17 of22
-------
24. Please indicate which elements are included in the CCC program:
(Circle 1 for "yes" and 2 for "no")
Does your program
Cross Connection Control include this element?
Program Elements Yes No
a. Right of entry 1 2
b. Surveys/inspections to identify cross connections within
the system/facility 1 2
c. Policy specifying which service connections must be
equipped with backflow prevention device/assemblies . 1 2
d. Enforcement authority to install devices/assemblies 1 2
e. Enforcement authority to test assemblies 1 2
f. Penalties for non-compliance
with ordinance 1 2
g. Public education programs 1 2
h. Training/certification of testers
and inspectors 1 2
25. What percentage of backflow prevention assemblies that are
tested fail annually during inspection?
(An assembly is a device that can be tested)
Page 18 of22
-------
Part II - Financial Characteristics
Reporting period: For Part II of the survey, please use the period indicated in Question 2(B) to report
"last year's" financial data.
Providing estimates: Please provide exact information from your system's records. Otherwise, provide
your best estimate of financial information that is applicable to your drinking water system.
If your system is a joint drinking water/wastewater facility, please be careful to record only data that are
relevant to the drinking water part of the facility. If data for the drinking water part of the facility are not
kept separately, please provide your best estimate of the share that is attributed to drinking water.
Rounding: Please record your dollar amounts to the nearest dollar. Do NOT record cents.
26 A. During the last year [as defined in your response to Question 2(B)] what were your drinking water system's
revenues from water sales for each of the following customer categories.
(If zero, enter "0". Note: if you cannot distinguish among the different types of non-residential customers,
enter the total for your non-residential customers in line d.)
Water Sales Customer Categories I Water Sales Revenues
1. Sold to other water suppliers
a. Treated water $
b. Untreated water $
2. Residential $
3. Non-residential
a. Commercial/industrial $
b. Agricultural $
c. Other (please specify) $
d. Subtotal, nonresidential (a - c)
4. Total water sales revenues (1-3) ...
Page 19 of22
-------
B. Please indicate your water system's revenues (luring the last year from other water-related revenue sources.
(If zero, enter '0')
I Water-related Revenues
5. Connection fees S
6. Development fees S
7. General fund revenues (e.g., from municipalities) $
Other water revenues not reported in the categories above
(e.g., fines, penalties, other fees)
(Please specifo)
9. Total water-related revenues $_
C. Please indicate your total water system revenues from lines 4 and 9 above.
10. Total water system revenues (lines 4 + 9) $_
[QUESTION 27 ASKED OF SMALL, MEDIUM, AND LARGE SYSTEMS ONLY]
27. If your primary business is not water-related and you reported
no revenues in Question 26 A or B, please indicate the revenues
you receive from your primary, non-water related business,
including rental income and the sale of other goods or services $
Note: Questions 28 - 30 Refer to Residential Customers Only
28. What is the average annual bill for a residential customer? $_
29. Please identify your drinking water system's billing structure for residential customers (circle all that apply).
Billing
structure
(Circle all that
apply)
Metered charges
Uniform rates 1
Declining block rate 2
Increasing block rate 3
Peak period rate (e.g., seasonal) 4
Unmetered charges
Separate flat fee for water 5
Combined flat fee for water and other services (e.g., lental fees, associate feeds, pad fees) 6
Other billing methods (Please specify) 7
Page 20 of 22
-------
30 A. Does your system use rates that may lower the cost of drinking water for low-income or fixed-income
households? (e.g., lifeline rates) (Circle one)?
Yes 1
No (Go to Question 31) 2
B. If your system uses rates that lower the cost of drinking water for low-income or fixed-income households,
please answer the following:
I. How many households receive these rates? (Number of households)
2. What is the highest annual income that qualifies for these
rates? (Dollars per year)
3. How much does it cost your system to provide these rate
reductions (i.e., what is the total dollar amount of the
reductions)? (Dollars per year)
31. Question 31 is intended to account for all of your drinking water expenses related to the revenues referred to in
question 26 A and B.
A. Please enter the number of people employed by your drinking water system and your system's total
compensation expenses (including direct compensation and fringe benefits) in the last year: If your system
is operated by or employs a contractor, enter the number of contract employees and total expenses of the
contractor on line 2.
Last year's employment and compensation
1. Number of Employees
2. Total Expenses,
including fringe benefits
(in Dollars)
1. Employment and employee expenses
2. Contractor expenses
B. Please enter other routine operating expenses in the last year (in dollars) $_
C. Please enter the amount of debt service expenditures in the last year for
borrowing to finance capital expenses (i.e., excluding expenditures for
borrowing to finance operating expenses, in dollars) $_
D. Please enter the amount of other expenses (excluding operating and debt
service expenses reported in Parts A, B, and C, but including any debt
service expenditures for borrowing to finance operating expenses) in the
last year (in dollars) (Please specify) $_
E. Total expenses, Parts A through D. (in dollars) $_
Page 21 of 22
-------
32. A. If you have paid for major capital improvements, repairs, or expansion in the last five years ending on the
date reported in Question 2B, allocate those expenditures to the following categories.
Type of Expense Total
1. Land $
2. Water source $
3. Distribution and transmission system $
4. Treatment $
5. Storage $
6 All other not included above $
7. Total capital expenses
B. What percent of the total capital expenditures from Line 7 of Part A was for water quality improvement,
replacement or major repair, or system expansion.? (Thepercentages should total to 100percent.)
1. Water quality improvements %
2. Replacement or major repair %
3. System expansion %
C. How were the total expenses from Line 7 in Part A funded''
Percentage of
capital expenses
funded from
each source
(Should sum to Average Interest
Source of funds 100%) Rate
1. Current revenues %
2. Grants from the Drinking Water State Revolving Fund %
3. Other government grants (either Federal or State) %
4 Borrowing from the Drinking Water State Revolving Fund % %
5. Borrowing from other public sector sources (e.g., state or regional
authorities) % %
6. Borrowing from private sector sources (e. g., banks or the bond market) % %
7. Other (please specify) % %
Page 22 of 22
-------
U.S. Environmental Protection Agency
Region 5, Library (PL-12J)
7? West Jackson Boulevard, 12th Floor
Cfcicago, II 60604-3590
-------
Office of Water (4607M)
EPA815-R-02-005B
November 2002
www.epa.gov/safewater
-------