&EPA
United States
Environmental Protection
Agency
Res National
5rogram Office
536 South Clark Street
Chicago, Illinois 60605
905R85110
The Environmental and Economic
Impacts of Detergent Phosphorus
Bans on Great Lakes Municipal
Wastewater Treatment Systems
Volume I
-------
-------
EPA-68-04-5017
March, 1985
PEER REVj£yy
Draft Report
The Environmental and Economic Impacts
of
Detergent Phosphorus Bans on Great Lakes
Municipal Wastewater Treatment Systems
Project Managers: King K. Moy, P.E. ESEI/EcolSciences, Inc.
John E. Racek, P.E. Williams and Works
Technical Advisors: Charles LaFrance, Ph.D., ESEI/EcolSciences, Inc.
Albert Posthuma, P.E., Williams and Works
EPA Project Monitor: Paul Horvatin
Great Lakes National Program Office
^Region V -
U.S. Environmental Protection Agency
536 South Clark StrSet
Chicago, Illinois 60605
-------
Volume 1
Table of Contents
Section Page
1.0 SUMMARY AND CONCLUSIONS 1
2.0 INTRODUCTION 2
2.1 Background 2
2.2 Purpose of Study 2
3.0 SELECTION OF REPRESENTATIVE MUNICIPAL WWTPs 5
3.1 Preliminary Screening 5
3.2 Final Screening 7
4.0 EVALUATION METHOD 8
4.1 Quality Control Protocol 8
4.2 Data Request Checklist 8
4.3 Technical Assumptions 9
5.0 ASSESSMENT OF ENVIRONMENTAL AND ECONOMIC IMPACTS OF 11
DETERGENT PHOSPHORUS BAN
5.1 Validity of the Selected Sample Plants 11
5.2 Flow Evaluation 15
5.3 Influent Pho'sphorus Loading Evaluation 15
5.4 Effluent Phosphorus Evaluation 18
5.5 Cost per Pound of Phosphorus Removal 18
5.6 Cost per Capita per Year Evaluation 22
6.0 TEST CASE 24
6.1 Phosphorus Loading Evaluation 24
6.2 Cost Evaluation 26
6.3 Conclusion 26
-------
List of Tables
Table
1
2
3
4
5
6
7
8
9
Data Summary of Selected WWTPs
Validity of the Selected Sample Plants
Comparison of Flow (mgd) Between P-Ban and
Non P-Ban States
Comparison of Influent Total-P (ppm) Between
P-Ban and Non P-Ban States
Comparison of Influent Total-P (Ib/capita/year) . . .
Between P-Ban and Non P-Ban States
Comparison of Effluent Total-P (concentration in ppm).
Between P-Ban and Non P-Ban States
Cost Comparison ($/lb P Removal) Between P-Ban and . .
Non P-Ban States
Cost Comparison (Cost of P Removal $/capita/yr) . . . .
Between P-Ban and Non P-Ban States
Comparison of Influent Phosphorus Between Pre and . .
Page
12
14
16
17
19
20
21
23
25
Figure
1
Appendix
A
B
C
D
Post Bans at Midland WWTP, Michigan
List of Figures
Great Lakes Basin
List of Appendices
Status of Detergent Phosphorus Legislation in the
Great Lakes Basin
Data Quality Assurance Protocol
Data Request Checklist
Statistical Evaluation
Page
3
-------
-------
1.0 SUMMARY AND CONCLUSIONS
A total of 24 representative WWTPs were selected in the Great Lakes Basin
to determine the environmental and economic impacts resulting from detergent
phosphorus bans. The sampled WWTPs were selected on the basis of effluent
compliance, laboratory quality control, record keeping and limited industrial
waste contribution. The WWTPs were also visited and data were statistically
analyzed to ensure the validity of the results which included considerable
inputs from individual plant personnel.
^. .
The analysis included comparisons of influent phosphorus loading, effluent
phosphorus loading, and costs between WWTPs in both states with and without
detergent phosphorus bans. The findings of this study are as follows:
• Detergent phosphorus bans have produced a 23 percent reduction
(from 5.6 ppm to 4.3 ppm) in total influent phosphorus concen- \
tration and a 32 percent reduction (from 3.10 to 2.10 lb/capita/ \
year) in influent phosphorus loading per capital per year. c f v^. I'
Therefore, it can be extrapolated that phosphorus loading to the ' • L^V-^
Great Lakes due to combined sewer overflows (CSO) will be reduced • oA^'v.j <-^
r as a result of detergent phosphorus bans. , / , ,. -y^
s yufcl V\ K ^
t There is no reduction in effluent phosphorus concentration from 2 ///<
/ WWTPs with detergent phosohorus bans comoared to WWTPs without ^^
'/ detergent phosphorus bans (0.71 ppm vs. 0.70 ppm).
• There is a 47 percent reduction in average cost per capita per
year for phosphorus removal attributable to detergent phosphorus
' ban (from 2.44 to 1.29 $/capita/yr) . ' However, the cost per Ib of
^tiosphorus removed is approximately the same J,$0.88/lb P-removal)
regardless of detergent phsophorus bairs. &&-> j ._ / /. (I
, O t''T • > . . / j.'l -^.M' X"Si_^ c*r~ 7X i*-*~ t~
-------
2.0 INTRODUCTION
2.1 Background
Pursuant to the Great Lakes Water Quality Agreements of 1972 and 1978, the
United States and Canadian Governments agreed to develop and implement programs
to reduce phosphorus loading in order to control accelerated eutrophication of
the Great Lakes. Measures undertaken by the two governments include the
following:
• Construction and operation of municipal wastewater treatment
plants (WWTPs) with phosphorus removal;
• Implementation of industrial pretreatment program;
• Control of nonpoint sources;
• Limitation of phosphorus in detergents sold for use within
the Great Lakes Basin.
The United States did not enact legislation to limit phosphorus content of
detergents on a national basis. Each state was allowed to pass its own
legislation on the basis of costs and benefits in limiting phosphorus content
in detergents. The eight states which border the Great Lakes are Illinois,
Indiana, Michigan, Minnesota, New York, Ohio, Pennsylvania, and Wisconsin (see
Figure 1). Currently, all Great Lakes States except Ohio, Illinois and
Pennsylvania have legislation limiting laundry detergents to 0.5 percent.
Cities of Chicago and Arkon which have local bans are the exception in the
States of Illinois and Ohio. The status of past and current legislation to
limit the phosphorus content of detergents in the Great Lakes States are
summarized in Appendix A.
2.2 Purpose of Study
The purpose of this study is to quantify the environmental and economic
benefits of phosphorus ban (P-Ban) in detergents by analyzing the influent
phosphorus loadings and operation and maintenance (O&M) costs of representa-
tive sample municipal WWTPs in the Great Lakes Basin. The selected sample
\
\\\^ \ 'VV ~^*-s \
I \ J
-2-
-------
&>
ui;
III
UI
UI
.UI
UJ|_-
-
<
O
CO
O
O
(E
z
CO
z
o
o
V)
t I
-3-
-------
WWTPs will be evenly distributed between P-Ban and Non P-Ban states so that
comparisons could be made to determine the following impacts (if any) as a
result of detergent P-Ban:
• Impaction of WWTP influent phosphorus loadings in ppm and Ib/capita/yr;
0 Impact on WWTP effluent phosphorus loadings in ppm;
• Impact on cost per Ib of phosphorus removed;
• Impact on cost per capita per year for phosphorus removal at WWTPs.
Statistical analyses will be employed to ensure the validity of the
results. Results derived from this study in conjunction with previous studies
would enable USEPA as well as other State and local government agencies to make
sound legislative decisions concerning phosphorus ban in detergents.
-4-
-------
f
I
3. Ql SELECTION OF REPRESENTATIVE MUNICIPAL WWTPs
-. •/? - *•
*-'-* A total, of 287 U.S. Municipal WWTPs were identified in the Great Lakes
Basin by the Water Quality Board of the International Joint Commission (IJC).
Information. for each facility was compiled in an inventory of the November,
1983 IJC Study (Appendix A of a Review of the Municipal Pollution Abatement
Programs in the Great Lakes Basin). This document formed the basis for the
selection of the representative municipal WWTPs for this study.
The selection process consisted of two phases: preliminary and final
screenings. The goal of the preliminary screening was to reduce the number of
facilities to a manageable number (i.e. under 75 plants) where more detailed
analysis could be conducted as part of the final screening. The goal of the
final screening process was to further reduce the number of plants to
approximately '25 so that an analysis (including a field trip to each facility)
could be performed on the environmental and cost impacts of detergent
phosphorus ban. \ ^ , ^ ,V';U^ v^u^ b'T
^
"'
3.1 Preliminary Screening "^
The first step of the preliminary screening process was to review the
followingxfocuments to derive certain information for screening:
A Review of Municipal Pollution Abatement Program in the Great
Lakes Basin (November, 1983), Municipal Abatement Task Force of
Water Quality Program Committee.
• Results of Round Robin on-Total Phosphorus in Municipal WWTP
effluents. -. _ "7 -•- '
• Detailed Review of Thirty Municipal Wastewater Treatment
Facilities in the Great Lakes Basin; Canviro Consultants LTD.
• CSO loadings inventory for Great Lakes Basin; GCA Corporation.
• Evaluation of High-Performance Phosphorus Control POTWs in the
Great Lakes Basin; Illinois Institute of Technology. ^
t Municipal October, 1981 -.September, 1982 Total Phosphorus--
Control. 5 " '
• 1983 U.S. Great Lakes Point Source Discharge Inventory f ,-
(October, 1983), Great Lakes National Program Office.^ - '
-5-
-------
Information derived from the above documents included: percentage of
industrial flows, general performance of facilities, laboratory quality
^.control, and type of sewerage facilities (combined or sanitary system). The
above information was then incorporated into Appendix A of the November, 1983
fIJC Report. As a supplement, field trips and telephone contacts were made to
EPA Region II (NYC and Edison, NJ) and Region V, Michigan DNR, Wisconsin DNR
and New York DEC to-collect the following information:
t
t
DMR-QA\fo^ 1981, 1982 and 1983. MJith the exception of a few
planish most of the 1981 DMR-QA were not available. )
Discharge Monitoring Report Results (Files of EPA region II and V,
Michigan DNR and Wisconsin DNR).
User Charge System Summary Sheets (EPA, Region V).
1984 Needs Survey for New York State WWTPs (EPA, Region II) and
copies of New York State permit applications.
t *•. *
t EPA Field Inspection Forms and Handwritten Comments
(£PA Region II and V).
V.^
-------
Application of the above rules resulted in the elimination of 220 WWTPs:
Total
WWTPs
18
97
6
63
65
3
35
287
WWTPS
Eliminated
12
83
3
48
48
2
24
220
WWTPs Survived First
High Rating
1
6
0
2
8
0
7
24
Screening
Marginal
5
8
3
13
9
1
4
43
Indiana
Michigan
Minnesota
New York
Ohio
Pennsylvania
Wisconsin
Total
It should be noted that the DMR-QA and Round Robin Laboratory results were
not used to eliminate any facilities. They were used in conjunction with other
criteria to support the selection or elimination process. Those plants that
survived the preliminary screening process were then yanked as "High Rating" or
"Marginal" for final screening. , fti^^'
, 'M • v '•'•
"AA/ f
c-r /" - • ' * i '
3.2 Final Screening ' /- _.,'••
Once the preliminary screening process was completed, DSE^A requested
inputs from individual State^ concerning the suitability of the "High Rating"
plants and requested recommendations for alternative (backup) plants. As a
result of their inputs, one "High Rating" plant was deleted and six "Marginal"
plants were added resulting in a total of 30 final plants.
Following the selection of the final plants, notifications were forwarded
to the municipal WWTPs detailing the intent of the study and requesting
additional data (i.e. monthly operation reports, cost records, chemical useage,
etc.) for further evaluation. As a supplement, telephone contacts were made to
provide the chief operators or superintendents with more details on the study.
Subsequent to the telephone conversations, three more plants were eliminated
because of their'nutrient deficient wastes, resulting in a total of 27 plants
for further field investigation.
-7-
-------
4.0 EVALUATION METHOD
Upon completion of the screening process, field investigations were made
to individual plants to collect operation/cost data for the analyses. Prior to
initiation of the field investigations, three major tasks were developed to
ensure the effectiveness of the trips and the validity of the analyses, to be
performed later. These three tasks were:
• Determination of the adequacy of the plant quality control program
(formal or informal)
• Preparation of a Data Request Checklist
• Outlining the technical assumptions which might affect the data
collection procedures and results of the analyses.
4 . 1 Quality Control Protocol /v « -*•" ? ' ''
A protocol was developed to evaluate the quality control/assurance of each
selected WWTP. A copy of the protocol is provided in Appendix B of this
report. During each plant field investigation, the protocol was used as a
guide for inspecting lab equipment, standard lab analytical procedures,
sampling procedures, chemical feed rate meteri-ng, sludge metering, record
keeping procedures and other features which might affect the study.
The infection staff consisted of either one licensed engineer with one
licensed operator or chemist. One person addressed data collection and the
other for quality control inspection. All inspection teams realized that no
plant would achieve a perfect score. • In
• -.-: - .^0.
fojrmaljQiiaJ-i±y~-A&sur>ance ^Programs. The adequacy or suitability of a plant was
a subjective assessment of the team members after consulting with the Project
Management Team. The quality control assurance inspections in conjunction with
the availability of operation and cost data resulted in the elimination of
three more plants, leaving 24 plants remaining for cost analyses.
4.2 Data Request Checklist
A Data Request Checklist was developed to help reduce the time needed to
collect the data during the field investigation. A copy of the Data Request
Checklist is provided in Appendix C. Copies of the checklist were forwarded to
-8-
-------
individual plants so that available data could be gathered by plant personnel
prior to the arrival of the inspection team. Once again, no plant would have
all the data outlined in the checklist. Many items were optional which would
help the analyses but not essential to the study. Major items which were
requested during the field trips included:
0 Process flow diagram including actual operating modes versus
designed modes
t Operation records including flow rates, TP/TSS/BOD5
concentrations, and sludge quantities
• Cost records including chemical, labor, fuel oil, natural gas,
electricity, maintenance and contract costs (i.e. sludge
hauling and/or landfill ing)
t Population of service area
• Quantity of industrial flow.
When available, data requested were for both fiscal year 1982 and 1983.
However, only single plant year data were considered suitable in many plants as
a result of operational problems, change of operation modes, and expiration of
phosphorus ban (Wisconsin). Fiscal year 1984 data were not available during
the field investigation period (June to October, 1984).
4.3 Technical Assumptions
In order to perform the analyses, many assumptions were made. These
assumptions were made prior to the field trips and then modified to reflect
inputs from the plant personnel at the end of the field investigation task.
These major technical assumptions are:
a. Sludge/ generated by phosphorus removal were calculated using
stoichiometric equations between phosphorus and iron or alum.
Plant operation records were used to determine the overall
percentage of chemical or P sludge to total sludge.
•Z* L,"^'"
b. Labor''associated with phosphorus laboratory testing and O&M of
chemical feed systems were derived from estimates provided by
plant personnel.
-9-
-------
i?;
-<-<•- .'. -r '•//•" s ' \ ^'•?*£
d. All energy costs for chemical pumping were considered for
phosphorus removal. Energy costs associated with the sludge
handling operation were calculated using chemical sludge
percentages derived under Item a.
,e. No capital costs for phosphorus removal equipment were considered.
f. Sludge hauling costs were considered in cases where hauling was
contracted to private contractors (payment on basis of sludge
quantity). In cases where sludge hauling was performed by plant
personnel, labor and or energy costs were considered based on
inputs from plant staff.
g. Costs for phosphorus removal were determined on a per capita per
year and on a per pound (Ib) of P removal basis. Fiscal year 1982
: costs were indexed to 1983 dollars.
Uniform unit costs J5or chemical, energy and labor were not
' considered because-- they would not reflect the actual plant
'T—,experiences in the Great Lakes Basin.
""ir During the study period, two plants in the Non P-ban states did
hot have to pay for all of their chemicals used for phosphorus
/ removal. This may have produced a slightly lower cost difference
/ for phosphorus removal between P-Ban and Non P-Ban states. Sf^"-"V •?
.
> n
-10-
-------
5.0 ASSESSMENT OF ENVIRONMENTAL AND ECONOMIC IMPACTS OF DETERGENT
PHOSPHOROUS BAN
Once the cost and environmental data were developed for the individual
plants (see Volume 2), analyses were performed to evaluate the impacts of
detergent phosphorous ban and the validity of the data using the analysis of
variance (ANOVA). Items evaluated in this study include: , .
f • / -- •',...- U "* -
/ ^
* Validity of the selected sample plants
•"BlaT'due to plant flow (mgd) for selected P-Ban and Non P-Ban WWTPs
• Impacts on municipal WWTP influent loadings (ppm and Ib per capita/yr)
t Impacts on municipal WWTP effluent loadings (ppm)
• Impacts on cost of phosphorus removal ($/lb P Removal and $/capita/
year)
Prior to the evaluation and statistical analyses, operation data of the 24
selected municipal WWTPS were tabulated (Table 1) for comparison. Data
provided in Table 1 represents the average of Fiscal Years 1982 and 1983*
although onTy^single7 ph&qt ,year da^a was' used for some of the 'selected
facilities. °"""' * *" ''\ I
ft -'<
The following sections address the results of the above-subjected items
individually.
5.1 Validity of the Selected Sample Plants
The validity of the selected sample plants was tested by comparing the
flow (mgd) data for the 24 sample plants and the flow data for all WWTPs with
flows of 1 mgd or higher in the entire Great Lakes Basin. The statistical test
demonstrates that the selected sample of 24 WWTPs is a valid, representative
sample of Great Lakes Basin WWTPs.
-- - ,., ,,
The data are summarized in Table 2. There are 275 - WWTPs in the Great
Lakes Basin with flows of 1 mgd or higher. Of these, 24 are the "sample
population" for this study. To insure independence of the data the sample
population was removed from the "statistical universe" for the test of the
validity of the sample, leaving 251 WWTPs in the statistical universe. The
-11-
-------
Table 2
Validity of the Selected Sample Plants
Sample Population
n= 24 Iy=400.67 2^=14511.54
y= 16.69 S2= 340.11
Statistical Universe (excluding sample population)
n=251 Ey=2907.69 Iy2=612799.74
y= 11.58 S2= 2316.46
Total
in=275 ZEy=3308.36 Z£y2=627311.28
y= 12.03 S2= 2144.20
o Test for Homogeneity of Variance
Fs = 6.8109 *(therefore, ANOVA cannot be used)
o Cochran and Cox t = 1.056 n.s.
t05 = 2.034
-14-
-------
statistical test was set up to determine (shether/or not the sample population
of 24 WWTPs are taken from the same popuratton as the statistical universe.
Since the test for homogeneity of variances detected heterogeneity, the ANOVA
could not be used. Instead, Cochran and Cox's procedure for a t-test with
unpaired observations and unequal variances was used. This test result was not
~X f.
significant; t=1.0564 is well below the critical value, tg5=2.0339. '<• ,• <..,/,
5.2 Flow Evaluation
There is no reason to believe that detergent phosphorus bans will have any
causal relationship to flow. Statistical comparison confirms that there is no
difference in flow (mgd) between the two groups of WWTPs. Flow was tested as a
means discovering any inadvertent biases introduced by the sampling design.
Although the mean flow at WWTPs with phosphorus bans of 21.2 mgd seems
intuitively to be different from the mean flow at WWTPs without phosphorus bans
of 12.0 mgd, the underlying WWTP flows are not different. The F ratio of 2.96
for the analysis of variance (Table 3) is less than the critical value of 4.08
associated with the 95 percent confidence level.-, Therefore, there is no reason
to reject the hypothesis that the flow of WWTPs with phosphorus bans and the
flow of WWTP's without phosphorous bans are the same. ,
5 . 3 Influent Phosphorus Loading Evaluation W l
Detergent phosphorus bans have produced a 23 percent reduction in the
total phosphorus^entering the WWTPs of the Great Lakes Basin. The average
loadings on WWTPs/ in states with detergent phosphorus bans are 4.3 ppm compared
to 5.6 ppm in states without detergent phosophate bans. This difference is
significant (probability greater than 95 percent), with an F ratio of 10.68
(Table 4). The, assumptions of the analysis of variance (ANOVA) are met because
the test of homogeneity of variances, yielded an Fs^ratio of 1.13 which is well
below the critical value of 2.12 at the 95 percent! confident level. The other
ANOVA assumptions are discussed in the section entitled "Statistical Evalua-
tion" (Appendix
-15-
-------
Comparison of Flow (mgd)
P-Ban
Table 3
Between P-Ban and Non P-Ban States
Non P-Ban
WWTP
Indiana
Ft. Wayne
South Bend
Michigan
Ann Arbor
Benton Harbor-
St. Joseph
E. Lansing
Gr. Rapids
Midland
New York
Monroe NWQ
N. Tonawanda
Tonawanda
Monroe GCO
1982
y
34.1
43.6
18.3
7.5
11.1
56.9
6.8
11.1
-
17.2
12.2
1983
y
35.1
38.0
-
8.2
11.3
56.1
6.5
12.3
6.9
17.8
12.8
n=20 Iy=423.8 Ey2=14169.84 n=19
y=21.2 S2=273.13
IN=39 ZZy=650.8 Z£y2=22098.90
y=16.7 S2=295.76
o Test for Homogeneity
WWTP
Ohio
French Cr.
Lima
Lorain
Maumee River
Oregon
Painesville
Willoughby-
E. Lake
Wisconsin
De Pere
Fond Du Lac
Manitowoc
Milwaukee-
S. Shore
Oshkosh
Sheboygan
ly=227.00
y=12.0
1982 1983
y y
1.8 1.9
16.0 12.7
18.4 16.4
6.7
2.6 4.0
3.9 3.3
8.5 7.5
3.9
7.5
9.9
79.0
11.2
11.8
Ey2=7929.06
S2=289.83
Fs = 1.0611 n.s. variances are equal
o ANOVA Table
Source df ss
B/T 1 832.36
W/I 37 10406.52
Total 38 11238.88
ms
832.36
281.26
2.9594 n.s.
means are
different
-16-
-------
Table 4
Comparison of Influent Total-P (ppm) Between P-Ban and Non P-Ban States
P-Ban
n=20
Non P-Ban
WWTP
Indiana
Ft . Wayne
South Bend
Michigan
Ann Arbor
1982 1983
y y
5.7 7.7
2.1 1.9
5.0
1982
WWTP y
Ohio
French Cr. 7.1
Lima 4.2
Lorain 6.7
Maumee River
1983
y
6.5
4.1
6.2
4.4
Benton Harbor- 5.1 4.3
St. Joseph
E. Lansing
Gr. Rapids
Midland
New York
Monroe NWQ
N. Tonawanda
Tonawanda
Monroe GCO
Zy=85.40
y=4.3
5.2 5.3
3.8 3.5
4.1 3.9
4.7 4.3
2.4
3.9 3.0
5.0 4.5
Zy2=398.94
S2=1.80
Oregon 4.8
Painesville 5.2
Willoughby- 6.1
E. Lake
Wisconsin
De Pere
Fond Du Lac
Manitowoc
Milwaukee-
S. Shore
Oshkosh
Sheboygan
n=19 !y=107.10
y=5.6
4.0
5.3
6.2
6.7
8.3
7.2
5.4
4.0
4.7
Iy2=632.49
S2=1.60
ZN=39 EZy=192.50 E£y2=1031.43
y=4.9 S2=2.14
o Test for Homogeneity
Fs = 1.1250 n.s.
o ANOVA Table
Source df ss ms
B/T
W/I
Total
1 18.20 18.20
37 63.07 1.70
38 81.27
F
10.68 *
-17-
-------
The test for homogeneity of variances yielded an Fs ratio of 1.43 which is
well within the critical value of 2.12, so the analysis of variance is a valid
test.
-18-
When influent phosphorus is expressed in pounds per capita (3.10
Ib/capita/yr in Non P-Ban States and 2.10 Ib/capita/yr in P-Ban States), the ^ w
reduction attributable to detergent phosphorus bans is 32 percent. This ^^
difference is also significant (probability greater than 95 percent), with at f
ratio of 3.336 (Table 5). The test for homogeneity of variances yielded an Fs
^-«—-•*'
ratio of 5.59 which is not below the critical value of 2.12 (95 percent
confident level), so Cochran and Cox's t-test was used to test significance of
the means.
5.4 Effluent Phosphorous Evaluation
There is no reduction in effluent phosphorous concentration from WWTPs
with detergent phosphorus bans compared to WWTPs without detergent phosphorus
bans. The average discharge of 0.71 ppm with detergent phosphorus bans is
indistinguishable from the 0.70 ppm discharged without the detergent phosphorus
bans. The analysis of variance (Table 6) yields an F ratio of only 0.03 which
is an order of magnitude less than the critical value of 4.08 associated with
the 95 percent confidence level. The Fs ratio test of homogeneity of variances
produces a value of 1.68 which is well below the critical value of 2.12 at the
95 percent confidence level. Therefore, the ANOVA test is valid. There is no
evidence to reject the hypothesis that all WWTPs in the Great Lakes Basin form
a single population with respect to effluent phosphorus concentration, with a
sample mean of 0.70 ppm, and a sample variance of 0.04.
5.5 Costs per Pound of Phosphorus Removal
Although the WWTPs with detergent phosphorus bans have sligteiy- lower
costs per pound of phosphorus removal than do the WWTPs without detergent
5Ml-<^J-^jc- **j
phosphorus bans, the difference is not .significant. The F ratio of 1.86 (Table
7) is less than the critical value of 4.08 associated with the 95 percent
confidence level. Thus there is 00" evidens^-to-H^e^eet— tte- hypothesis 4&at the \ F'V
/
average cost per pound of phosphorus removed ($0.88/lb) is imtep-endent of- they I1*-"
phosphorus ban. Kcf/ '^^^ ^^/ ** L
-------
Table 5
Comparison of Influent Total-P (Ib/capita/year) between P-Ban and Non P-Ban States
P-Ban
WWTP
Indiana
Ft. Wayne
South Bend
Michigan
Ann Arbor
Benton Harbor-
St. Joseph
E. Lansing
Gr. Rapids
Midland
New York
Monroe NWQ
N. Tonawanda
Tonawanda
Monroe GCO
1982
y
2.33
2.49
2.66
1.81
2.06
2.53
2.17
1.66
-
1.57
2.72
1983
y
3.29
1.90
_
1.67
2.02
2.29
1.96
1.69
1.40
1.25
2.97
Non P- Ban
1982
WWTP y
Ohio
French Cr. 2.48
Lima 3.44
Lorain 4.76
Maumee River
Oregon 1.72
Painesville 3.30
Willoughby- 3.52
E. Lake
Wisconsin
De Pere
Fond Du Lac
Manitowoc
Milwaukee-
S. Shore
Oshkosh
Sheboygan
1983
y
2.49
2.71
3.90
1.40
1.96
2.99
3.10
2.78
4.06
6.70
1.99
2.60
2.94
n=20 Zy=42.02 Zy2=93.23 n=19 Zy=58.84 Zy2=208.45
y=2.10 S2=0.26 (variance) y=3.10 S2=l .46 (variance)
ZN=39 ZZ y=100.86 ZZ y2=310.05
y=2.59 S2=1.30
o Test for Homogeneity
Fs = 5.59*
o Cochran and Cox t = 3.336 *
critical value, tgs = 2.100
-19-
-------
Table 6
Comparison of Effluent Total-P (concentration in ppm) between P-Ban and Non P-Ban States
P-Ban
Non P-Ban
n=20
WWTP
Indiana
Ft . Wayne
South Bend
Michigan
Ann Arbor
1982
y
0.71
0.36
' 0.80
Benton Harbor- 0.80
St. Joseph
E. Lansing
Gr. Rapids
Midland
New York
Monroe NWQ
N. Tonawanda
Tonawanda
Monroe 6CO
Zy=14.19
y=o.7i
0.93
0.91
0.30
0.97
0.62
0.60
0.91
Zy2=11.14
S2=0.06
1983
y
0.91
0.30
-
0.60
0.88
0.93
0.26
0.88
-
0.66
0.86
1982
WWTP y
Ohio
French Cr. 0.80
Lima 0.56
Lorain 0.82
Maumee River
Oregon 0.90
Painesville 0.47
Willoughby- 0.80
E. Lake
Wisconsin
De Pere
Fond Du Lac
Manitowoc
Milwaukee-
S. Shore
Oshkosh •
Sheboygan
n=19 Zy=13.27 zy2=9.87
y=0.70 S2=0.03
1983
y
0.90
0.52
0.84
0.50
0.90
0.47
0.90
0.50
0.64
0.74
0.65
0.41
0.95
ZN=39 ZZy=27.46 zzy2=21.02
y=0.70 S2=0.04
o Test for Homogeneity
Fs = 1.6834 n.s.
o ANOVA Table
Source df ss
B/T
W/I
Total
1
37
38
0.0012
1.6829
1.6841
-20-
ms F
0.0012 0.0263 n.s.
0.0455
-------
Table 7
Cost Comparison ($/lb P Removal) between P-Ban and Non P-Ban States
P-Ban
Non P-Ban
WWTP
Indiana
Ft. Wayne
South Bend
Michigan
Ann Arbor
Benton Harbor-
St. Joseph
E. Lansing
Gr. Rapids
Midland
New York
Monroe NWQ
N. Tonawanda
Tonawanda
Monroe GCO
1982
y
0.26
0.74
1.16
0.45
0.50
0.50
1.02
1.27
-
0.45
1.02
1983
y
0.15
0.46
-
0.54
0.49
0.38
1.31
0.99
2.29
0.57
0.81
1982
WWTP y
Ohio
French Cr. 1.43
Lima 0.64
Lorain 0.52
Maumee River
Oregon 1.70
Painesville 0.88
Willoughby- 0.91
E. Lake
Wisconsin
De Pere
Fond Du Lac
Manitowoc
Milwaukee-
S. Shore
Oshkosh
Sheboygan
1983
y
1.25
0.23
0.57
2.41
2.12
0.92
0.88
1.19
1.25
0.74
0.36
0.71
0.37
n=20 zy=15.36 zy2=16.3910 n=19
y=0.79 S2=0.2418
ZN=39 ZZy=34.44 zzy2=41.7688
y=0.91 S2= 0.2988
o Test for Homogeneity of Variance
Fs = 1.4284 n.s.
o ANOVA Table
Source df ss
Zy=19.08 Zy2=25.3778
y=1.03 S2= 0.3454
B/T
W/I
Total
1 0.5436
37 10.8120
38~ 11.3556
ms
0.5436
0.2922
1.8604 n.s.
-21-
-------
5.6 Cost per Capita per Year Evaluation
There is a 47 percent reduction in average costs per capita per year for
phosphorus removal attributable to detergent phosphorus ban. The average costs
in P-Ban states was $1.29/capita/year compared to $2.44/capita/year. This
difference is statistically significant, with a Cochran and Cox's t test value
of 3.84 (Table 8) compared to a critical value of 2.10 at the 95 percent
confidence level.
The test of homogeneity of variances yielded an Fs ratio of 2.34 which is
not within the critical value of 2.12 at the 95 percent confidence level.
Therefore, the Cochran and Cox's t test has to be used to test the significant
difference between the mean costs per capita per year in P-Ban and Non P-Ban
states.
-22-
-------
Table 8
Cost Comparison (Cost of P Removal $/capita/yr) between P-Ban and Non P-Ban States
P-Ban
Non P-Ban
WWTP
Indiana
Ft. Wayne
South Bend
Michigan
Ann Arbor
Benton Harbor-
St. Joseph
E. Lansing
Gr. Rapids
Midland
New York
Monroe NWQ
N. Tonawanda
Tonawanda
Monroe GCO
1982
y
0.52
1.53
2.63
0.68
0.84
0.94
2.04
1.64
-
0.56
2.22
1983
y
0.45
1.53
-
0.77
0.81
0.62
2.37
1.29
2,27
0,54
1.63
n=20 zy=25.88
y= 1.29
1982
WWTP y
Ohio
French Cr. 3.14
Lima 1.88
Lorain 2.13
Maumee River
Oregon 2.41
Painesville 2.63
Willoughby- 2.78
E. Lake
Wisconsin
De Pere
Fond Ou Lac
Manitowoc
Milwaukee-
S. Shore
Oshkosh
Sheboygan
1983
y
2.70
0.54
1.90
2.97
3.04
2.53
2.33
3.06
4.69
4.45
0.62
1.67
0.85
Zy2=43.2322 n=19 Zy=46.32
S2= 0.5128 y= 2.44
ZN=39 IZy=72.20 Zly2=177.7584
y= 1.85 S2= 1.1604
o Test for Homogeneity of Variance
Fs = 2.3403 *
o Cochran and Cox t = 3.8385 *
tos - 2.101
£y2=134.5262
S2= 1.2002
'1,30 '
-23-
-------
6.0 TEST CASE
The previous sections have demonstrated a significant reduction in
influent phosphorus loadings (ppm; Ib/capita/yr) and a significant cost savings
($/capit a/year) as a result of detergent phosphorus bans. Based on the design
of the present study, these conclusions are valid for regional studies within
the Great Lakes Basin, or for the entire Great Lakes Basin.
It is tempting to generalize the conclusions as a prediction of the
effects of detergent phosphorus policy on individual WWTPs. To do so would
require extrapolation from the study data from the pooled regional data to the
individual WWTP operating experience. On the other hand, if it could be
demonstrated that such extrapolations are valid, it would; (1) strengthen the
conclusions, and (2) provide potentially useful guidance in setting detergent
phosphorus policy; i.e., instituting a ban when one is not in place, or
suspending the ban when it is in place.
1 ,;•
'''
A r
." /l< To achieve the verification that the study conclusions can be applied to
.7 individual WWTPs, a representative WWTP was selected. The selected WWTP was
. / '
u the Midland, Michigan facility, for which data were readily available. Only
influent phosphorus loading (ppm) and costs/capita were analyzed to confirm the
results. The data used for this test case study were drawn from different
years than those included in the basin-wide study. • ^^ ~1
:•}• ' i -r '-''•• ' " ; ; : . . , ' f, ., J ,-'•"'
' /, — — /, -, ,*•''•• */ ,
6.1 Phosphorus Loading Evaluation / J'f ~t'l^t'1 ' ('• "' j,£_, ' -^ ••" ' >"'' ' «<
There was a 31 percent reduction in the influent phosphorus loading (ppm) ,^,
at Midland as a result of the detergent phosphorus ban (Table 9). Monthly S '"•-'•
average concentrations of phosphorus in the influent wastewater dropped from
6.4 ppm in July, 1976 - June, 1977 (pre ban) to 4.4 ppm in July, 1978 - June,
1979 (post ban).
The observed reduction in phosphorus loading is significant, with a
Cochran and Cox, t=6.068 compared to a critical value of 2.^201 at the 95
percent confidence level . The Cochran and Cox t-test was used as the test of
significance because the test for homogeneity of variance was significant,
Fs=4.21 compared to a critical value of 2.12.
-24-
-------
Table 9
Comparison of Influent Phosphorus between Pre and Post Bans
at Midland WWTP, Michigan
Pre-Ban (1976-1977)
Post-Ban (1978-1979)
Month
July
August
September
October
November
December
January
February
March
April
May
June
n=12 zy=76.2
y= 6.4
y
5.4
5.8
6.3
6.5
6.5
6.4
6.3
6.7
6.2
6.0
7.4
6.7
Zy2=486.62
S2= 0.25
Month
July
August
September
October
November
December
January
February
March
April
May
June
n=12 Zy=52.2
y= 4.4
y
5.0
4.6
4.7
5.2
4.1
4.8
5.3
5.4
2.1
2.7
3.9
4.4
Zy2=238.66
S2= 1.05
Zn=24 ZZy=128.4 ZZy2=725.28
y= 5.4 S2= 1.67
o Test for Homogeneity of Varience
Fs = 4.2144 *
o Cochran and Cox t = 6.068 *
t05 = 2.201
-25-
-------
-------
6.2 Cost Evaluation
Fiscal year 1976-1977 (pre ban) and 1978-1979 (post ban) data from the
Midland WWTP were compared to determine the impact caused by detergent
phosphorus ban imposed in October, 1977. The 1976-1977 costs were updated by
indexing the unit costs to 1978-1979 level. The indexed pre ban total cost was
$77,156 for phosphorus removal of $2.06/capita/year. The post ban cost was
$57,514 for phosphorus removal or $1.53/capita/year. This comparison indicates
a j:ost reduction of approximately 25.5 percent. Further, the costs per pound
of phosphorus removed between pre ban and post ban were approximately the same
($0.73/lb vs. $0.75/lb, respectively) which concurs with the statistical
conclusion discussed in a previous section.
6.3 Conclusions
The results and conclusions presented in Section 5 were based on a statis-
tically valid sample of basin-wide conditions. As a result, the conclusions
can be used to describe the effects of detergent phosphorus policies generally
in the Great Lakes Basin. • It is clear that regional detergent phosphorus bans
will result in significant reductions in the following parameters:
• Influent phosphorus loading (ppm and Ibs/capita/year)
• Costs ($/capita/year)
Section 6 presented a test case to confirm that these same conclusions can
be generalized to predict what would happen on an individual WWTP case if a
phosphorus ban were adopted. It was confirmed that there was a significant
reduction in influent phosphorus loading (ppm) and that cost savings were
consistent with basin-wide results when a post ban fiscal year was compared
with a pre ban fiscal year for the Midland, Michigan, WWTP.
Therefore, the conclusions of the study are valid equally for individual
WWTPs in the Great Lakes Basin and for basin-wide studies.
-26-
-------
APPENDIX A
-------
DETERGENT PHOSPHORUS LEGISLATION
JURISDICTION
DATE
EFFECTIVE
ALLOWABLE
PU)
DETERGENTS INCLUDED
REFER-
ENCES
Michigan - cont'd.
01/81 to present
Detroit
(07/72)
Minnesota
New York
01/72 to 05/73 '
06/73 to present
Erie County 05/71 to 12/71
01/72
Syrr.;.use 07/71
Ohio
Akron
02/71 to 06/72
None*
07/72 to 12/72
01/73 to present
28.0
(0.5)
01/77 to present 0.5
01/77 to present 11.0
8.7
0.5
8.7
0.5
8.7
8.7
8.7
0.5
- metal brighteners, cleansers & 19
treatment compounds, corrosion
or paint removers, conversion
coating agent,rust inhibitors,
etchant, phosphatizer,
degreasing compound, industrial
or commercial cleansers used
primarily in industrial and
manufacturing projects.
- (City ordinance enacted 10
but pre-empted by Act
226 - State of Michigan-
above).
- total ban
- detergents used for house-
hold and commercial
machine dishwashing.
- household use, laundry use,
other personal use, indus-
trial uses except those for
machine dishwashers, dairy
equipment, beverage equip-
ment, food processing and
industrial cleaning equip-
ment.
- excluded detergents used
for machine dishwashers;
dairy, beverage, food
processing and industrial
cleaning equipment.
- all cleansers
- excludes machine dish-
washers; dairy, beverage,
food processing and indus-
trial cleaning equipment.
1-6,
11
12
2,3,16
1-5,16
6
13
6
1,3,4
2,14
14
2,4,5,
14
-------
STATUS OF LEGISLATION TO LIMIT THE PHOSPHORUS CONTENT OF DETERGENTS
IN THE GREAT LAKES BASIN
DETERGENT PHOSPHORUS LEGISLATION
JURISDICTION
DATE
EFFECTIVE
ALLOWABLE
Pii)
DETERGENTS INCLUDED
REFER-
ENCES
Illinois
Chicago
Indiana
None
/71 to 06/72
07/72 to present
01/72 to 12/72
01/73 to present
Michigan
07/72 to 09/77
10/77 to present
8.7
0.5
8.7
0.5
8.7
0.5
01/81 to present 14.0
- all cleansers
- detergents.
- all cleansers
- excludes detergents used
for cleaning in-place
food processing and dairy
equipment; phosphorus acid
products including san-
itizers, brightners,
acid cleansers and metal
conditioners; detergents
used in household and
commercial machine
dishwashers; detergents
used in hospitals and
health care facilities;
industrial laundry
detergents; detergents
used in dairy, beverage,
food processing and other
industrial cleaning
equipment.
- all cleansers
- household laundry
detergents
- commercial machine dish-
washers, dairy and farm opera-
tion cleansers; cleansers used
in the manufacturing, prepara-
tion and processing of foods
and food products including
dairy, beverage, egg, fish,
brewery, poultry, meat, fruit
and vegetable processing.
1
2
1-5
2,20
1-7,
20
2,8,17
1-4,
9,18
19
-------
DETERGENT PHOSPHORUS LEGISLATION
JURISDICTION
DATE
EFFECTIVE
ALLOWABLE
Ptt)
DETERGENTS INCLUDED
REFER-
ENCES
Ohio
Akron
None*
02/71 to 06/72 8.7
07/72 to 12/72 8.7
01/73 to present 0.5
Pennsylvania None
Wisconsin 07/79 to 06/82** 0.5
8.7
Canada
03/70 to 12/72
01/73 to present
20.0
8.7
2.2
- excluded detergents used
for machine dishwashers;
dairy, beverage, food
processing and industrial
cleaning equipment.
- all cleansers
- excludes machine dish-
washers; dairy, beverage,
food processing and indus-
trial cleaning equipment.
- laundry detergents
- machine dishwashing
detergents and medical and
surgical equipment cleansers
- chemical water conditioners,
- laundry detergents.
1,3,4
2,14
14
2,4,5
14
1,3
1,2,
21
21
21
15,23
1,3,4
6,23
m/ !' ' '
ft' proposed 2.2% ban is under consideration. Sadewicz, John J. July 11, 1983:
Personal Communication^. Ohio Environmental Protection Agency.
**An reinstatement; of the bans ts undor--conGidGrat'ion> poscib-ly -comfnoncing
January 1, 198^. Schuettpelz, Duane H. July 12, 1983: Personal Communication
Wisconsin Department of Natural Resources.
-------
APPENDIX B
-------
DATA QUALITY ASSURANCE PROTOCOL
FOR
MUNICIPAL POTWS
PHOSPHORUS BAN STUDY
EPA DELIVERY ORDER #16
-------
This report provides protocols for evaluating the data quality control/
assurance of each POTW selected for this study. In order to determine the
validity of the data from the POTW for the purpose of this study, the following
areas will be evaluated.
I. General Laboratory Quality Assurance/Control
II. Quality Control for Analytical Performance -
a. Phosphorus
b. Biochemical Oxygen Demand
c. Total and Suspended Solids
III. Sampling Procedures
IV. Flow Measurement and Recording
a. Raw Wastewater
b. Sludge Handling
'7 c. Chemical Feed System
v. Recordkeeping and Reporting
-------
Name of POTW: _fe4.OcVr%
Date:
Reviewer:
I. GENERAL LABORATORY QUALITY ASSURANCE/CONTROL
Quality control is an on-going laboratory responsibility to assure the genera-
tion of data that is accurate. Quality assurance is the result of a series of
checks and balances monitoring all phases of the measurement process. The
components of the quality control program are described below.
T
-i-
-------
Name of POTW:
Date:
Reviewer:
Yes No
Laboratory Staff Responsibilities and Operational Procedures
a. An individual is responsible to implement and monitor
laboratory quality assurance and control /
Whom? _ /
b. The analytical procedures used are acceptable to USEPA .
Type: _ ^f
c. Laboratory bench sheets and written procedures are used
d. Quality assurance check samples are performed
Frequency:
e. All pertinent data is recorded in laboratory records \/ _
f. Instrument or procedure malfunctions or variances
from acceptable limits are reported \/ _
g. Comments: _
2. Distilled Water
a. Distilled water is produced in laboratory still
Type : — ^" _ _
b. Distilled water is of acceptable quality for phosphorus,
BOD and suspended solids analyses: • /.
- The distilled water is phosphorus free J
- BOD dilution water oxygen depletion is equal to or
less than 0.20 mg/1 after 5 days incubation of 20°C
- Conductivity is greater than 0.2 megohm as resis- /
tivity or less than 2.0 micromhos/cmat 25°C j/
- Maximum total matter of 1.0 mg/1
c. Still is in good working condition / /
d. Preventive maintenance is performed on the still regularly tx
e. Storage of distilled water is acceptable \s
f. Comments-: — "t(*0.. \ naw^-in
C i
-2-
-------
Name of POTW:
Date:
Reviewer:
Yes No
3. Laboratory Services
The following laboratory services are provided in good
working condition?
a. Compressed air , /
b. Vacuum system [/
c. Hood system \/'
d. Electrical services
e. Comments:
Analytical Balance Type:
a. The balance is operated according to manufacturer's *
instructions \/ _
b. Balance is mounted on heavy shock proof table
c. Balance location is appropriate
d. Balance is stored properly when not in use.
e. Balance is checked and calibrated by manufacturer
Frequency:
Standard weights are used to check accuracy
Frequency: _ ' _ \s
Comments:
5. pH/Specific Ion Meters Type:
a. Instrument is in good working condition \/
b.. Instrument is calibrated with two buffers at a
minimum of two points that bracket the expected /v
pH of the sample and are three pH units apart _ V _ .
c. pH/ion probes are properly stored and in good condition
d. Comments: "TV\£. n\-\ m? *s rr>.^rA,r ^r^i
» . i
-j o^(? -Vo
-3-
-------
Name of POTW:
Date:
Reviewer:
Yes No
6.. Spectrophotometer Type: pyaaScV
V
a. Instrument is in good working condition
b. Instrument is used according to manufacturer's
recommendations
c. Instrument is checked and calibrated by manufacturer
Frequency:
d. Instrument checked for wave length alignment
Frequency: ^ •* npo ypaT
e. Proper cells (cuvetts) or sample holders are used
f» Instrument has ultraviolet range
g. Instrument has infrared range
h. Absorption cells are kept clean
i. Matched absorption cells are used for check
j. Comments:
7. Glassware
a. The following volumetric glassware is used:
- volumetric flasks
- pipettes i/
- burets • \/
b. Volumetric glassware is Class A designation
c. Cleaning and storage of glassware is proper, including
warm detergent wash and three distilled water rinses
d. Cleaning and handling of glassware for biochemical oxygen
demand is acceptable
e. Cleaning and handling of glassware for phosphorus
analysis is acceptable
Comments:
W^T or,\
-------
Name 01 rui«:
Date:
Reviewer:
Yes No
8. Reagent Quality
a. Reagents used are of analytical reagent grade
b. Reagents are prepared and standardized with care and ,
proper technique \s
c. Reagents are restandardized or prepared fresh as required /
by their stability i/
d. Stocks and standards checked regularity for deterioration
e. Reagents are prepared in volumetric glassware
f. Reagents are stored in proper glassware
g. Borosilicate glass bottles with ground glass stoppers used
h. Reagents properly labeled with proper identification
i. Comments:
9. Ovens and Furnaces
a. Ovens, incubators, and temperature control instrumentation
are checked against NBS certified thermometers
Frequency:
xw
b. Servicing/recalibration is initiated when problems arise
c. Comments:
-5-
-------
roame or
Date:
Reviewer:
II. QUALITY CONTROL FOR ANALYTICAL PERFORMANCE
Quality assurance control for analytical performance is used to insure that
valid precision and accuracy is generated for each determination and by each
analyst.
Yes _No
1. Standard Acceptance Criteria
a. Acceptance criteria is established for each parameter and ,
procedure \/
b. Quality control charts are used for establishing deviations
c. Mean or central lines are established for upper or lower
limits
d. Comments:
2. Standard Curves
a. Standard curves constructed for all spectrometer
determinations .
b. Curve consists of reagent blank and three to five standards ^/
c. Standards fall on linear portion of curve \/
d. Sample data falls on curves within range of standards \/
e. Curve constructed by plotting instrument response versus
concentration
f. Curve contains proper title and analytical protocol
- dates of analyses
- analysts initials ^/ _
- operating condition of instrument \/ _
g. Curves constructed when new reagents are made or .
standards exceed deviation limits \/
h. Reagent blank and at least two standards used with each
determination
i .
Comments : 5Aa.vvln> v-i4
-6-
-------
name u i i \j i i« .
Date:
Reviewer:
Yes No
3. Quality Control Check Samples
a. USEPA known value samples are used for quality control
check samples
b. Frequency:
c. Analyses: �<- n\e4&l<
9 "
55
t • k
^ L
-------
Date:
Reviewer:
Yes N£
Proper sample dilutions are used
g. Incubation bottles as water sealed ^/
h. Dissolved oxygen analyses performed by acceptable
method: /\* probe winkler i/
i. Samples show residual DO of 1 mg/1 and a depletion of
at least 2 mg/1 after incubation S
j. Depletion of dilution water no more than 0.2 mg/1 after
incubation y/
k. Calculations are correct •
1. Comments:
6. Solids Analyses
a. Procedure as outlined in 15th edition of Standard Methods y _
b. Gooch Crucibles Size #4 are used
c.' Reeve Angel, 934AH, 2.4 cm or equivalent glass filter
filter paper are used [/
d. Analytical balance is used properly (/
e. Muffle furnace operated at 550°C . //
f. Drying oven operated at 103°C ± 1°C _^X
g. Desiccator and desiccant are used properly tX
h. Gooch crucibles are prepared properly tx
i. Two samples are analyzed for each determination _
j. Sample volumes greater or equal to 50 ml are used for all ..
determinations is
k. Calculations are correct _
1. Comments: O^lvj ov\o. sawU \S uiSgrV
-8-
-------
ui ruin;
Date:
Reviewer:
III. SAMPLING PROCEDURES
Yes No
a. Sampling points are representative of waste stream
- Influent
b.
c.
d.
e.
f.
- Final effluent
- Pimary sludge
Composite sampling is used
- based on time *r
- proportion to flow ^/
Samples are properly refrigerated and preserved
Automatic samplers are clean and properly maintained
Sample vessels are properly cleaned
Comments: Py^VA-p y\v"-. ,o>"\\\ ^ \OLO
\\'i<
^
A
* TI
i 0 \ >"V\CA
D«>\
IV. FLOW MEASUREMENT AND RECORDING
1. Raw Wastewater
a. Raw wastewater flows metered
Type: U>ev\\u-v* i Location:
Meters properly maintained
Calibration periodically checked
Frequency: Method:
b.
c.
Regular service/recalibration by service representative
Frequency:
Comments: y\r> •cPovA en.\ \ \\fT*j\~\o ir\ 5cK'gr\ u^l-ff^
IS -Sollon') or Oer^nrmppi1. vY\e^
2. Sludge Handling
a. Sludge handling flow metered
Type: ^
b. Meters properly maintained
-9-
-------
c.
d.
e.
Name of POTW:
Date:
Reviewer:
Calibration periodically checked
Frequency: Method:
Regular service/recalibration by service representative
Frequency:
Comments:
Yes No
Chemical Feed System
a. Chemical feed rates metered
Type: (jjJa 11 <\ cv_ r Ti *v A^ **•
b. Metering pumps are calibrated properly
c. Calibration chambers are used to calibrate pumps
d. Jar testing is used for chemical feed rate determination
e. Comments: r^Sdv v ^ \r-.oS We-^ IA^PC/
~T v, « °
.frV b<
w-^
I/
~7
k
RECORDKEEP.IMG AND REPORTING
a. Data handling in laboratory accomplished with bench sheets
b. Laboratory bench sheets containing information as to
analytical methodology, date of analysis, analyst,
standards and corresponding resposnes, sample identi-
fication, calculation, and quality control measures
c. Raw flow records adequate
d. Sludge quantities record adequate
e. Chemical feed rates properly recorded
f. Chemical feed rates accurate
g. All records stored properly for future reference
h. Comments:
-10-
-------
Date:
Reviewer:
SUMMARY
a. Quality Control and Assurance is Adequate
Laboratory Staff Responsibilities and Operational Procedures:
Impact on Data Quality fl>> r,r\ucr<^
Yes
Distilled Water: (
Impact on Data Quality /A7r> nA\f?,r
pH/Specific Ion Meters:
Impact on Data Quality !{}& add
Glassware:
Impact on Data Quality _ A/o
tmfiac-f-
Reagent Quality:
Impact on Data Quality Wr> g d Perse.
Standard Acceptance Criteria:
Impact on Data Quality
>W
Spectrophotometer: -
Impact on Data Quality j arli/f'sfr ir-^C'r
Ovens and Furnaces: /
Impact on Data Quality /l/p *,efiscr?e. imacT
Standard Curves:
Impact on Data Quality rv/^tx^r «?hr>t*,(d
TTrtr'l r hPc.kecf'
l\.t
/
-11-
-------
Date:
Reviewer:
Yes No
Quality Control Check Samples:
,
Impact on Data Quality A/7 //y.*f-A<'/\ o c'c-fQio
"'..-- i
Phosphorus Analyses:
Impact on Data Quality
run C{y\fi
(/ /
Biochemical Oxygen Demand: /
% Impact on Data Quality
Solids Analyses:
Impact on Data Quality
Sludge Handling:
Impact on Data Quality /\Q
Chemical Feed System:
Impact on Data Quality ftnnrf harrf/e
b. Comments:
'T ,'.<;
_
y ' J. htive. vteihrf
p/o rs*-4()r\! J^'lS A- r fit* fe-
/nino/~
.
Signature A ^-/IAI A ,'
Date:
'
-12-
Raw Wastewater: /
Impact on Data Quality nn nd(/e/-\z i^f&C'
trta,' fioer or\ et^ r/a< l kaSK I H&.
A - r-aa -ea
/? /. era. // A- rt r{
-------
APPENDIX C
-------
Environmental Protection Agency
Evaluation of Impact of Detergent Phosphorus Bans
on Great Lakes Municipal Wastewater Treatment Systems
REQUEST FOR INFORMATION
INFORMATION
FURNISHED
A. DESCRIPTION OF TREATMENT PLANT
1. Process Schematics
- Wastewater Treatment
- Solids Handling and Disposal
- Recycle Streams
2. Design Data
3. Effluent Limits
4. Site Plan
5. Influent Flow
- Domestic, MGD, BOD, SS
- Industrial, MGD, BOD, SS ;_
- Infiltration & Inflow, Average & Peak MGD
- Major Industrial Flows - Identify and
describe impact on influent flow and quality
- Population Served
- Overflows, Bypassing in sewer system-describe
6. Phosphorus Removal - Describe
- Chemical Used
- Addition Point
- Method of Controlling
Possible Sources:
1. Facility Plan or Design Report
2. NPDES Permit
4/84
-------
Environmental Protection Agency
Evaluation of Impact of Detergent Phosphorous Bans
On Great Lakes Municipal Wastewater Treatment Systems
PROCEDURES FOR DATA COLLECTION AND TECHNICAL ANALYSIS
1. Telephone - State Regulatory Agency
- Review each WWTP
- Obtain name of Superintendent or contact person
2. Telephone Each WWTP
- Review purpose of Study
- Confirm willingness to participate
- Discuss Information Request form to be mailed and availability of data
- Tentative dates for site visit
3. Mail Information Request Form to each WWTP
4. Review data returned with request form
- Make list of incomplete or missing information
- Make list of questions on plant and operation
5. Telephone Each WWTP
- Review list of information still needed and ways to get information
- Confirm date for plant visit
6. Plant Visit
- Evaluate quality of data per protocol
- Collect remaining data from files or from in-plant investigation if not
available in files
- Complete checklist of information needed
7. Review Quality of Plant Data
- Quality of Data per protocol and impact on study conclusions
- Completeness of Information per checklist and impact on study conclus-
ions
- Screen out any plants whose data would adversely affect the validity of
the study conclusions
8. Technical Analysis of Data
- Summarize Data from remaining plants
- Calculate average BOD, Solids and phosphorus removal versus chemical and
energy use
- Estimate solids production and costs associated with phosphorus removal
- Develop list of questions and additional information needed
9. Follow-up Plant Visit
- Telephone WWTP and review questions
- Visit plant and obtain additional information
10. Complete Technical Analysis of Data
- Develop Summary of Data
- Develop matrix to compare plants in different/states.
4/84
-------
B. OPERATING DATA - 1982 & 1983
1. Influent Flow - BOD, SS, Phosphorus
2. Primary Effluent - BOD, SS, Phosphorus
3. Secondary Effluent - BOD, SS, Phosphorus
4. Final Effluent - BOD, SS, Phosphorus
5. Sludge Production - Primary, Secondary, Other
6. Sludge Thickening - Flow, % Solids, Phosphorus
7. Sludge Dewatering - Flow, % Solids, Phosphorus
8. Sludge Disposal - Total, % Solids, Phosphorus
9. Recycle Streams - Flows, % Solids, Phosphorus
10. Operating Parameters - Biological Treatment
11. Chemical Use for Phosphorus Removal
- Chemical, concentration, where added, total Ibs,
or gallons used
12. Chemical Use for Sludge
- Chemical(s), concentration, where added, total
Ibs. or gallons used
Possible Sources:
1. Monthly Operating Reports
2. Annual Summary
3. Operating Logs
4/84
-------
C. ENERGY USE - 1982 & 1983
1. Electricity - Total Plant
- By process if available (Sludge Only)
2. Natural Gas - Total Plant
- By process if available (Sludge Only)
3. Other Energy - Describe
4. Energy Reuse - Describe
Possible Sources:
1. Operating Reports
2. Billing Records
3. , Annual Report
D. COSTS - 1982 & 1983
1. Chemicals - Total Cost of chemical per year
- Cost per unit as delivered
2. Electricity - Total Cost per month
- Cost per KW hr. (average including
all charges)
3. Natural Gas - Total Cost per month
- Cost per 100 cu. ft. (average
including all
charges)
4. Other Energy - Total Cost per month
5. Sludge Disposal - Total Cost per year
6. Overall Operation & Maintenance if available
Possible Sources:
1. Billing Records
2. Annual Report
4/84
-------
E. Comments on plant and information furnished. (Attach additional sheets if
more space is needed).
4/84'
-------
APPENDIX D
-------
APPENDIX D.
Statistical Evaluations
Formal hypothesis testing, or statistical procedures have been employed
here to determine whether or not there is a difference in the means of each
parameter between states with detergent phosphorus bans ("P-Ban" states) and
states without detergent phosphous bans ("non-P-Ban" states). The two primary
options for statistically testing a difference between two means are (1) to
compare an estimate of the difference to an estimate of dispersion (or
variability) of the individual values used to estimate the mean, by a t-test,
the most familiar of which is Student's t; or (2) to compare an estimate of
variance based on the individual values, by an F-ratio (cf.., Steele & Torrie,
1960, p. 72). In either case, the usual null hyptohesis, for testing purposes,
is that there is no different between the means.
Although each test criterion, a t-test on an F-ratio, is employed to test
the same null hypothesis, they differ in the implicit alternative hypothesis
suppported. The underlying model for t-tests is usually that each observation,
Yj, consists of
YI = y + Eis
where y is the true, parametric population mean and E-j is a random variable,
which is normally distributed with a mean of zero and which describes sampling
error. Therefore, the most valid alternative hypothesis is that the two means
being tested are not equal.
On the other hand, the usual underlying model for F-ratios is that each
observation, Y2j> consists of
Y2j = y + Tj + E,
where Tj is a "treatment" component attributable to treatment j. Therefore,
the F-ratio as a test criterion supports the alternative hypothesis that the
treatments (in the present case, whether or not there is a P-Ban) have a
significant effect on the results. Since the present study is concerned with
the alternative hypothesis that P-Ban versus non P-Ban has an effect on the
results, the F-ratio is a more appropriate test criterion as long as the
assumptions of the test are met.
-------
The primary assumptions of the F-ratio test criteria include normally
distributed sampling error, additive treatment effects, independence, and uni-
form or homogeneous variance. Each of these will be discussed below to show
how these assumptions are met for the F-ratio or analysis of variance (ANOVA)
test.
Normal distribution. The raw data for all of the statistical tests in
this study are annual average values. By the Central Limit Theorem, these
averages are normally distributed although the individual values (e.g., 365
values for flow in mgd) are not (Sokal &. Rohlf, 1969, p 130).
Additive treatment effects. Conversely, this assumption can be worded as
"no interaction anong treatments." In the special case of single treatment (as
is the case for this study), this assumption reduces to a tautology so is
necessarily true.
Independence. This assumption is the equivalent of the previous one
applied to residual variance (sampling error); that is, no interaction between
sampling error and treatments. There is no reason to suspect a systematic
error associated with P-Ban on non P-Ban. Therefore it is assumped that the
residual variance (W/I in the ANOVA tables) is independent.
Homogeneous variance. This assumption must be tested individually for
each ANOVA. For a comparison of two means, this assumption is tested by taking
the ratio of the variance about one sample mean to the variance about the other
sample mean. For each statistical test employed in this study, the homogeneity
of variances test is reported as an Fs ratio. Whenever heterogeneous
variances are encountered, a t-test has been used as the test criterion for the
effects of P-Ban. However, the familiar Student's t required homogeneity of
variance. The appropriate t for heterogeneous variances is Cochran and Cox's
(Steele & Torrie, 1960, p 81).
-------
Procedures.
For each statistical test (in the present study), it is necessary to
accumulate the sum of the observations, ZY (P-Ban), ZY (non P-Ban) and ZZ Y
(pooled), and the sum of the observations squared, xzY^ (P-Ban), Z Y^ (non
P-Ban), ZZ y2 (pooled). We then calculate the means Y (and pooled on total
mean, Y) and variances, 82-
/ !
/ 1
The test of omogeneity of jvariance, Fs, is calculated by dividing the
larger of the\J;wo variances X^-Ban or non P-Ban) by the smaller of the two.
The following formulas are used for the ANOVA (if Fs was less than the
critical value):
TT-
ssw/I = ssTOT - ssB/T
and any MS = SS/df. The F-ratio of interest here is MSg/j/MSw/j.
The Cochran & Cox t-test differs from Student's t only in the critical
value of t05, which for Cochran & Cox's t is calculated as the weighted
mean of the appropriate Student's t valves.
------- |