conomic
conomic
Analysis
Analysis
for
the
for
the
ong
ong
Term
2
Enhanced
Surface
Water
Treatment
Rule
Term
2
Enhanced
Surface
Water
Treatment
Rule
E
L
PREPARED
FOR:

U.
S.
ENVIRONMENTAL
PROTECTION
AGENCY
Office
of
Ground
Water
and
Drinking
Water
PREPARED
BY:

THE
CADMUS
GROUP,
INC.
1901
North
Fort
Myer
Drive
Suite
900
Arlington,
VA
22209
US
EPA
CONTRACT:
68­
C­
02­
026
Work
Assignment:
1­
21
June
2003
Proposal
Draft
LT2ESWTR
EA
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
i
Contents
Appendices
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
v
Exhibits
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
vii
Acronyms
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
xii
Health
Risk
Reduction
and
Cost
Analysis
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
xv
Executive
Summary
ES.
1
Need
for
the
Rule
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
1
ES.
2
Consideration
of
Regulatory
Alternatives
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
2
ES.
3
Summary
of
the
LT2ESWTR
Requirements
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
3
ES.
4
National
Benefits
and
Costs
of
the
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
5
ES.
4.1
Benefit
Estimates
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
5
ES.
4.2
National
and
Household
Cost
Estimates
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
9
ES.
5
National
Net
Benefits
and
Summary
of
Comparison
of
Alternatives
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
11
Chapter
1:
Introduction
1.1
Summary
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1­
1
1.1.1
Monitoring
and
Treatment
Requirements
for
Filtered
Systems
.
.
.
.
.
.
.
.
.
.
.
.
.
1­
1
1.1.2
Monitoring
and
Treatment
Requirements
for
Unfiltered
Systems
.
.
.
.
.
.
.
.
.
.
.
1­
6
1.1.3
Requirements
for
Existing
Uncovered
Finished
Water
Reservoirs
.
.
.
.
.
.
.
.
.
.
1­
6
1.1.4
Disinfection
Profiling
and
Benchmarking
Requirements
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1­
7
1.1.5
Implementation
Timeline
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1­
7
1.2
Document
Organization
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1­
8
Chapter
2:
Statement
of
Need
for
the
Proposal
2.1
Introduction
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
1
2.2
Description
of
the
Issue
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
1
2.3
Risk
Balancing
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
2
2.4
Public
Health
Concerns
to
Be
Addressed
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
3
2.4.1
Cryptosporidium
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
3
2.4.2
Uncovered
Finished
Water
Reservoirs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
6
2.5
Regulatory
History
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
7
2.5.1
Statutory
Authority
for
Promulgating
the
Rule
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
7
2.5.2
1979
Total
Trihalomethane
Rule
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
8
2.5.3
1989
Total
Coliform
Rule
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
8
2.5.4
1989
Surface
Water
Treatment
Rule
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
8
2.5.5
1996
Information
Collection
Rule
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
9
2.5.6
1998
Interim
Enhanced
Surface
Water
Treatment
Rule
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
9
2.5.7
1998
Stage
1
Disinfectants
and
Disinfection
Byproducts
Rule
.
.
.
.
.
.
.
.
.
.
.
.
2­
10
2.5.8
2000
Proposed
Ground
Water
Rule
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
10
2.5.9
2001
Filter
Backwash
Recycling
Rule
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
11
2.5.10
2002
Long
Term
1
Enhanced
Surface
Water
Treatment
Rule
.
.
.
.
.
.
.
.
.
.
.
.
2­
11
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
ii
2.5.11
2003
Proposed
Stage
2
Disinfectants
and
Disinfection
Byproducts
Rule
.
.
.
.
.
2­
11
2.6
Economic
Rationale
for
Regulation
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
12
Chapter
3:
Consideration
of
Regulatory
Alternatives
3.1
Introduction
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3­
1
3.2
Development
Process
for
Regulatory
Alternatives
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3­
1
3.3
Specific
Regulatory
Alternatives
Considered
in
this
EA
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3­
2
3.3.1
Summary
of
Bin
Classification
and
Treatment
Requirements
for
Regulatory
Alternatives
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3­
2
3.3.2
Additional
Treatment
for
Direct
Filtration
Systems
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3­
4
3.4
Alternative
Monitoring
Approaches
Considered
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3­
5
3.4.1
Indicators
of
Microbial
Contamination
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3­
5
3.4.2
Cryptosporidium
Monitoring
Strategies
for
Bin
Classification
.
.
.
.
.
.
.
.
.
.
.
.
.
3­
6
Chapter
4:
Baseline
Conditions
4.1
Introduction
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
1
4.2
Data,
Tools,
and
Processes
Used
in
Baseline
Development
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
2
4.2.1
ICR
and
ICRSS
Observed
Data
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
3
4.2.2
ICR
and
ICRSS
Modeled
Data
and
Method
for
Estimating
Source
Water
Occurrence
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
6
4.2.3
Surface
Water
Analytical
Tool
(
SWAT)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
8
4.3
Industry
Profile
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
8
4.3.1
Public
Water
System
Characterization
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
8
4.3.2
Systems,
Plants,
and
Population
Subject
to
the
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
10
4.3.3
Water
Treatment
Plant
Design
and
Average
Daily
Flows
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
17
4.4
Baseline
for
Unfiltered
Plants
(
Pre­
LT2ESWTR)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
20
4.4.1
Treatment
Characterization
for
Unfiltered
Plants
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
20
4.4.2
Number
of
Unfiltered
Systems,
Plants,
and
Population
Served
.
.
.
.
.
.
.
.
.
.
.
4­
21
4.4.3
Source
Water
Cryptosporidium
Occurrence
for
Unfiltered
Plants
.
.
.
.
.
.
.
.
.
4­
22
4.4.4
Finished
Water
Cryptosporidium
Occurrence
for
Unfiltered
Plants
.
.
.
.
.
.
.
.
.
4­
23
4.5
Baselines
for
Filtered
Plants
(
Pre­
LT2ESWTR)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
25
4.5.1
Treatment
Characterization
for
Filtered
Plants
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
25
4.5.2
Number
of
Filtered
Plants
and
Population
Served
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
28
4.5.3
Source
Water
Cryptosporidium
Occurrence
for
Filtered
Plants
.
.
.
.
.
.
.
.
.
.
.
4­
31
4.5.4
Finished
Water
Cryptosporidium
Occurrence
for
Filtered
Plants
.
.
.
.
.
.
.
.
.
.
4­
34
4.5.5
Comparison
of
EPA
Finished
Water
Cryptosporidium
Estimates
with
Aboytes
et
al.
(
2000)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
40
4.5.6
Predicted
Bin
Classification
for
Filtered
Plants
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
41
4.6
Baseline
for
Uncovered
Finished
Water
Reservoirs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
42
4.7
Households
Incurring
Costs
Due
to
the
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
45
4.8
Summary
of
Uncertainties
in
Development
of
LT2ESWTR
Baselines
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
48
Chapter
5:
Benefits
Analysis
5.1
Introduction
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
1
5.2
Quantified
Health
Benefits
from
Reduction
in
Exposure
to
Cryptosporidium
.
.
.
.
.
.
.
.
.
5­
2
5.2.1
Overview
of
Risk
Assessment
Methodology
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
3
5.2.2
Hazard
Identification
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
5
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
iii
5.2.3
Dose­
Response
Assessment
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
7
5.2.4
Exposure
Assessment
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
14
5.2.4.1
Distribution
of
Infectious
Cryptosporidium
in
Finished
Water
.
.
.
.
.
.
5­
14
5.2.4.2
Distribution
of
Individual
Daily
Drinking
Water
Consumption
.
.
.
.
.
.
5­
21
5.2.4.3
Population
Affected
by
the
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
22
5.2.5
Risk
Model
Structure
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
25
5.2.6
Individual
Annual
Risk
Distributions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
31
5.2.7
General
Population
Risk
 
Number
of
Cases
Avoided
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
33
5.2.7.1
Unfiltered
Systems
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
35
5.2.7.2
Filtered
Systems
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
39
5.2.8
Reduction
in
Sensitive
Subpopulation
Risk
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
39
5.3
Monetized
Benefits
from
Reduction
in
Exposure
to
Cryptosporidium
Resulting
from
the
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
39
5.3.1
Value
of
Reduction
in
Cryptosporidiosis
Cases
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
40
5.3.1.1
Value
of
Illnesses
Avoided
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
40
5.3.1.2
Value
of
Avoiding
Fatal
Cases
of
Cryptosporidiosis
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
50
5.3.1.3
Measuring
Benefits
Over
the
LT2ESWTR
Implementation
Schedule
.
5­
50
5.3.1.4
Adjustment
for
Income
Elasticity
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
51
5.3.1.5
Present
Value
of
Future
Benefits
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
53
5.3.1.6
Summary
of
Quantified
Benefits
of
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
53
5.3.2
Monetization
of
Benefits
to
Sensitive
Subpopulations
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
56
5.4
Other
Benefits
of
LT2ESWTR
Provisions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
56
5.4.1
Reduction
in
Outbreak
Risk
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
56
5.4.2
Costs
to
Households
to
Avert
Infection
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
57
5.4.3
Enhanced
Aesthetic
Water
Quality
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
58
5.4.4
Risk
Reduction
from
Co­
occurring
and
Emerging
Pathogens
.
.
.
.
.
.
.
.
.
.
.
.
5­
58
5.4.5
Benefits
from
Other
Rule
Provisions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
58
5.4.6
Summary
of
Nonquantified
Benefits
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
59
5.5
Summary
of
Uncertainties
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
60
5.6
Comparison
of
Regulatory
Alternatives
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
60
Chapter
6:
Cost
Analysis
6.1
Introduction
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
1
6.1.1
Cost
Description
and
Assumptions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
2
6.2
Summary
of
the
National
Costs
of
the
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
5
6.3
Rule
Implementation
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
11
6.3.1
PWSs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
11
6.3.2
States
and
Other
Primacy
Agencies
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
12
6.4
Source
Water
Monitoring
for
Initial
Bin
Classification
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
13
6.4.1
PWSs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
13
6.4.2
State
and
Other
Primacy
Agency
Cost
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
13
6.5
Treatment
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
14
6.5.1
Toolbox
Options
and
Unit
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
16
6.5.1.1
Toolbox
Options
Not
Considered
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
16
6.5.1.2
Technologies
Considered
for
the
LT2ESWTR
Cost
Analysis
.
.
.
.
.
.
.
6­
17
6.5.2
Compliance
Forecast
and
Technology
Selection
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
21
6.5.3
Capital
and
Annual
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
22
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
iv
6.6
Costs
of
Treatment
for
Unfiltered
Plants
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
23
6.7
Costs
for
Benchmarking
and
Technology
Reporting
Requirements
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
24
6.8
Costs
of
Treatment
for
Uncovered
Finished
Water
Reservoirs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
24
6.8.1
Unit
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
25
6.8.2
Compliance
Forecast
and
Technology
Selection
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
25
6.8.3
Total
Annual
Treatment
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
26
6.9
Future
Source
Water
Monitoring
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
27
6.10
Household
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
28
6.11
Unquantified
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
29
6.12
Summary
of
Uncertainties
and
Sensitivity
Analyses
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
30
6.12.1
Cryptosporidium
Occurrence
Data
Sets
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
31
6.12.2
Sensitivity
Analysis
of
Influent
Bromide
Levels
on
Technology
Selection
for
Filtered
Plants
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
32
6.13
Comparison
of
Regulatory
Alternatives
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
33
Chapter
7:
Economic
Impact
Analysis
7.1
Introduction
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
1
7.2
Regulatory
Flexibility
Act
and
Small
Business
Regulatory
Enforcement
Fairness
Act
.
.
.
7­
1
7.3
Small
System
Affordability
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
3
7.4
Feasible
Treatment
Technologies
for
All
Systems
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
3
7.5
Effect
of
Compliance
with
the
Proposed
LT2ESWTR
on
the
Technical,
Managerial,
and
Financial
Capacity
of
Public
Water
Systems
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
4
7.5.1
Requirements
of
the
Preferred
Proposed
Regulatory
Alternative
.
.
.
.
.
.
.
.
.
.
.
7­
5
7.5.2
Systems
Subject
to
the
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
6
7.5.3
Impact
of
the
LT2ESWTR
on
System
Capacity
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
6
7.5.4
Derivation
of
Proposed
LT2ESWTR
Scores
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
6
7.6
Paperwork
Reduction
Act
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
12
7.7
Unfunded
Mandates
Reform
Act
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
13
7.7.1
Social
Benefits
and
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
15
7.7.2
Disproportionate
Budgetary
Effects
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
17
7.7.3
Macroeconomic
Effects
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
21
7.7.4
Consultation
with
Small
Governments
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
21
7.7.5
Consultation
with
State,
Local,
and
Tribal
Governments
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
22
7.7.6
Regulatory
Alternatives
Considered
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
22
7.7.7
Impacts
on
Small
Governments
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
22
7.8
Indian
Tribal
Governments
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
23
7.9
Impacts
on
Sensitive
Subpopulations
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
28
7.9.1
Impacts
on
the
Immunocompromised
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
28
7.9.2
Protection
of
Children
from
Environmental
Health
Risks
and
Safety
Risks
.
.
.
7­
28
7.10
Environmental
Justice
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
30
7.11
Federalism
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
31
7.12
Actions
Concerning
Regulations
That
Significantly
Affect
Energy
Supply,
Distribution,
or
Use
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
31
Chapter
8:
Comparison
of
Benefits
and
Costs
of
the
LT2ESWTR
8.1
Introduction
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
1
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
v
8.2
Summary
of
National
Benefits,
Costs,
and
Net
Benefits
of
the
Preferred
Regulatory
Alternative
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
1
8.2.1
National
Benefits
Summary
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
4
8.2.2
National
Cost
Summary
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
7
8.2.3
National
Net
Benefits
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
9
8.3
Comparison
of
Regulatory
Alternatives
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
11
8.3.1
Comparison
of
Benefits
and
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
12
8.3.2
Cost­
Effectiveness
Measures
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
31
8.4
Effect
of
Uncertainties
on
the
Benefit­
Cost
Comparisons
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
35
8.5
Summary
of
Benefit
and
Cost
Comparisons
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
37
Chapter
9:
References
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
vi
Appendices
Appendix
A:
Pre­
LT2ESWTR
Removal
Credit
Appendix
B:
Characterizing
Cryptosporidium
Concentration
and
Methods
for
Predicting
Plant
Binning
Appendix
C:
Benefits
Appendix
D:
National
Costs
for
Rule
Implementation
and
Monitoring
Appendix
E:
Unit
Costs
for
Technologies
Considered
in
the
LT2ESWTR
Appendix
F:
Technology
Selection
Forecast
Methodology
Appendix
G:
Technology
Selection
Results
Appendix
H:
Treatment
Costs
for
Filtered
and
Unfiltered
Plants
Appendix
I:
Unit
Costs
for
Uncovered
Finished
Water
Reservoirs
Appendix
J:
Estimation
of
Household
Costs
Appendix
K:
Additional
Information
on
the
Approach
For
Valuing
Time
Losses
Appendix
L:
Calculations
Supporting
The
Cost
of
Illness
(
COI)
Analysis
Appendix
M:
Small
Community
Surface
Water
and
GWUDI
Systems
by
State
Appendix
N:
Dose­
Response
Infectivity
Analysis
Appendix
O:
Assigning
LT2ESWTR
Costs
and
Benefits
Appendix
P:
Sensitivity
Analyses
for
Cost
of
Illness
Values
Appendix
Q:
Treatment
Costs
for
Filtered
and
Unfiltered
Plants
Appendix
R:
Sensitivity
Analysis
for
AIDS­
Related
Mortality
Rate
Appendix
S:
Analysis
of
Individual
Risk
by
Initial
Bin
Appendix
T:
Risk
Assessment
Model
 
Program
and
Data
Files
Appendix
U:
Cost
Models
Appendix
V:
Total
Annualized
Costs
for
Rule
Alternatives
Appendix
W:
Screening
Analysis
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
vii
Exhibits
Exhibit
ES.
1
Overview
of
Key
LT2ESWTR
Requirements
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
4
Exhibit
ES.
2
Implementation
Timeline
for
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
5
Exhibit
ES.
3
Summary
of
Nonquantified
Benefit
and
Groups
Affected
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
6
Exhibit
ES.
4
Summary
of
Annual
Avoided
Illnesses
and
Deaths
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
7
Exhibit
ES.
5a
Summary
of
Monetized
Benefits
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
8
Exhibit
ES.
5b
Summary
of
Monetized
Benefits
 
Traditional
Cost
of
Illness
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
8
Exhibit
ES.
6
Summary
of
System
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
10
Exhibit
ES.
7
Summary
of
Annual
Household
Cost
Increases
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
11
Exhibit
ES.
8a
Comparison
of
Mean
Net
Benefits
for
All
Regulatory
Alternatives
 
Enhanced
Cost
of
Illness
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
12
Exhibit
ES.
8b
Comparison
of
Mean
Net
Benefits
for
All
Regulatory
Alternatives
 
Traditional
Cost
of
Illness
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
12
Exhibit
ES.
9a
Comparison
of
Mean
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
14
Exhibit
ES.
9b
Comparison
of
Mean
 
Traditional
Cost
of
Illness
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ES­
15
Exhibit
1.1
Filtered
Systems
Bin
Classification
and
Treatment
Requirements
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1­
3
Exhibit
1.2
Microbial
Toolbox
Components
for
the
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1­
5
Exhibit
1.3
Implementation
Time
Line
for
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1­
8
Exhibit
2.1
Reported
Cryptosporidiosis
Outbreaks
in
U.
S.
Drinking
Water
Systems
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2­
5
Exhibit
3.1
Summary
of
Bin
Requirements
for
Filtered
Systems
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3­
4
Exhibit
3.2
Probability
of
Misclassification
for
Monitoring
and
Binning
Strategies
Considered
for
the
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3­
7
Exhibit
4.1
ICR
and
ICRSS
Comparison
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
4
Exhibit
4.2
Methodology
for
"
Linking"
Consecutive
Surface
Water
CWSs
and
NTNCWSs
to
Their
Selling
Systems
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
13
Exhibit
4.3a
Number
of
Unlinked
and
Linked
Surface
Water
and
GWUDI
CWSs,
Number
of
Plants,
and
Population
Served,
by
Ownership
and
System
Size
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
14
Exhibit
4.3b
Number
of
Unlinked
and
Linked
Surface
Water
and
GWUDI
NTNCWSs,
Number
of
Plants,
and
Population
Served,
by
Ownership
and
System
Size
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
14
Exhibit
4.3c
Number
of
Unlinked
and
Linked
Surface
Water
and
GWUDI
TNCWSs,
Number
of
Plants,
and
Population
Served,
by
Ownership
and
System
Size
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
15
Exhibit
4.4a
Filtered
Plant
Average
Daily
and
Design
Flow
by
System
Size
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
19
Exhibit
4.4b
Unfiltered
Plant
Average
Daily
and
Design
Flow
by
System
Size
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
20
Exhibit
4.5
Treatment
Baseline
for
Unfiltered
Plants
By
System
Size
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
22
Exhibit
4.6
Observed
ICR
Source
Water
Total
Oocyst
Occurrence
for
Unfiltered
Plants
.
.
.
.
.
.
.
.
.
.
.
4­
23
Exhibit
4.7
Modeled
Source
Water
Cryptosporidium
Occurrence
ICR
Data
for
Unfiltered
Systems
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
24
Exhibit
4.8
Predicted
System
Binning
for
Unfiltered
Systems,
Based
on
Central
Tendency
of
Cryptosporidium
Occurrence
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
25
Exhibit
4.9
Percentage
of
Plants
with
Pre­
LT2ESWTR
Cryptosporidium
Log
Reduction
Credits
for
Existing
Technologies
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
26
Exhibit
4.10
Predicted
Percentage
of
Plants
Using
Advanced
Technologies
Following
Implementation
of
the
Stage
2
DBPR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
27
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
viii
Exhibit
4.11
Treatment
Baseline
for
Filtered
Plants
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
30
Exhibit
4.12
Summary
of
Observed
Source
Water
Cryptosporidium
Total
Oocyst
Occurrence
 
Filtered
Plant
Data
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
31
Exhibit
4.13
Modeled
Source
Water
Cryptosporidium
Occurrence
ICR
Data
for
Filtered
Systems
.
.
.
.
4­
32
Exhibit
4.14
Modeled
Source
Water
Cryptosporidium
Occurrence
ICRSSM
Data
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
33
Exhibit
4.15
Modeled
Source
Water
Cryptosporidium
Occurrence
ICRSSL
Data
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
33
Exhibit
4.16
Comparison
of
Modeled
Source
Water
Cryptosporidium
Occurrence
by
Data
Set,
Median
Curves
Only
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
34
Exhibit
4.17
Predicted
Ranges
of
Cryptosporidium
Reduction
Pre­
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
36
Exhibit
4.18a
Distribution
of
Cryptosporidium
Reduction
in
Small
Systems
Pre­
LT2ESWTR,
Standard
Estimate
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
38
Exhibit
4.18b
Distribution
of
Cryptosporidium
Reduction
in
Large
Systems
Pre­
LT2ESWTR,
Standard
Estimate
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
38
Exhibit
4.18c
Distribution
of
Cryptosporidium
Reduction
in
Small
Systems
Pre­
LT2ESWTR,
Estimate
With
0.5
Log
Reduction
Credit
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
38
Exhibit
4.18d
Distribution
of
Cryptosporidium
Reduction
in
Large
Systems
Pre­
LT2ESWTR,
Estimate
With
0.5
Log
Reduction
Credit
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
38
Exhibit
4.19a
Predicted
Finished
Water
Cryptosporidium
Occurrence
Pre­
LT2ESWTR,
Small
Systems
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
39
Exhibit
4.19b
Predicted
Finished
Water
Cryptosporidium
Occurrence
Pre­
LT2ESWTR,
Large
Systems
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
39
Exhibit
4.20
Predicted
System
Binning
for
Preferred
Alternative,
Based
on
Central
Tendency
of
Cryptosporidium
Occurrence
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
41
Exhibit
4.21
Baseline
Numbers
of
Uncovered
Finished
Water
Reservoirs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
42
Exhibit
4.22
Baseline
Number
of
Uncovered
Finished
Water
Reservoirs
in
Systems
of
Each
Size
Category
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
43
Exhibit
4.23
Surface
Area
and
Flows
for
Uncovered
Finished
Water
Reservoirs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
45
Exhibit
4.24
Universe
of
Households
Affected
by
Rule
Provisions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
46
Exhibit
4.25
Baseline
Numbers
of
Households
Incurring
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
47
Exhibit
4.26
Households
Paying
Treatment
Costs
for
Uncovered
Reservoirs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
47
Exhibit
4.27
Mean
Household
Water
Usage
Rates
by
System
Size
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
48
Exhibit
4.28
Summary
of
Uncertainties
Affecting
LT2ESWTR
Baseline
Estimates
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4­
49
Exhibit
5.1
Risk
Assessment
Model
Categories
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
3
Exhibit
5.2
Health
Risk
Assessment
Framework
for
Cryptosporidium
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
5
Exhibit
5.3
Symptoms
of
205
Patients
with
Confirmed
Cases
of
Cryptosporidiosis
During
the
Milwaukee
Outbreak
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
7
Exhibit
5.4
Percent
of
Plants
With
Pre­
LT2ESWTR
Treatment
Credit
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
19
Exhibit
5.5a
Actual
Log
Removal
Achieved
for
Systems
without
Credits
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
20
Exhibit
5.5b
Actual
Log
Removal
Achieved
for
Systems
with
Credits
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
21
Exhibit
5.6
Distribution
of
Individual
Daily
Drinking
Water
Consumption
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
22
Exhibit
5.7
Number
of
Systems,
Population
Served,
and
Annual
National
Exposure
by
System
Type
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
23
Exhibit
5.8
Individual
Exposure
by
System
Type
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
24
Exhibit
5.9
Flowchart
of
Risk
Model
 
Step
1:
Computing
Annual
Individual
Risk
of
Illness
.
.
.
.
.
.
.
.
.
5­
26
Exhibit
5.10
Overview
of
Risk
Assessment
Model
Parameters
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
29
Exhibit
5.11
Secondary
Spread
and
Secondary
Attack
Rates
Associated
with
Cryptosporidiosis
Outbreaks
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
32
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
ix
Exhibit
5.12
Annual
Individual
Risk
Distributions
Based
Upon
ICR
Occurrence
Data
Set,
Filtered
CWSs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
34
Exhibit
5.13
Annual
Individual
Risk
Distributions
Based
Upon
ICR
Occurrence
Data
Set,
Unfiltered
CWSs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
35
Exhibit
5.14
Annual
Cases
of
Illness
and
Deaths
Avoided
for
the
LT2ESWTR,
Preferred
Alternative,
Unfiltered
Systems,
by
Data
Set
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
37
Exhibit
5.15
Annual
Cases
of
Illness
and
Deaths
Avoided
Due
to
the
LT2ESWTR,
Preferred
Alternative,
All
Filtered
Systems,
by
Data
Set
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
38
Exhibit
5.16
Direct
Medical
Costs
of
a
Case
of
Cryptosporidiosis
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
44
Exhibit
5.17
Days
Lost
and
Days
with
Lost
Productivity,
by
Severity
of
Illness
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
46
Exhibit
5.18
Weighted
Average
Days
Lost
for
Work,
Caregivers,
and
Productivity
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
46
Exhibit
5.19
Value
of
Time,
2000
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
47
Exhibit
5.20
Total
Loss
Per
Case,
Enhanced
and
Traditional
COI,
2000
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
48
Exhibit
5.21
Yearly
Total
Loss
Per
Case,
Enhanced
and
Traditional
COI
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
49
Exhibit
5.22
Mean
of
Yearly
Values
for
a
Statistical
Life
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
52
Exhibit
5.23
Annualized
Benefits
of
Illnesses
and
Deaths
Avoided,
Preferred
Alternative,
Enhanced
Cost
of
Illness
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
54
Exhibit
5.24
Annualized
Benefits
of
Illnesses
and
Deaths
Avoided,
Preferred
Alternative,
Traditional
Cost
of
Illness
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
55
Exhibit
5.25
Summary
of
Nonquantified
Benefits
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
59
Exhibit
5.26
Summary
of
Uncertainties
Affecting
LT2ESWTR
Benefits
Estimates
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
62
Exhibit
5.27
Annual
Cases
of
Illness
and
Deaths
Avoided
from
the
LT2ESWTR
for
Regulatory
Alternatives
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
63
Exhibit
5.28a
Summary
of
Estimated
Present
Values
of
Annual
Illnesses
and
Deaths
Avoided
from
LT2ESWTR
for
Regulatory
Alternatives,
Enhanced
Cost
of
Illness
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
64
Exhibit
5.28b
Summary
of
Estimated
Present
Values
of
Annual
Illnesses
and
Deaths
Avoided
from
LT2ESWTR
for
Regulatory
Alternatives,
Traditional
Cost
of
Illness
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
65
Exhibit
6.1
E.
coli
and
Cryptosporidium
Laboratory
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
4
Exhibit
6.2
Number
of
Systems
and
Plants
Expected
to
Incur
Costs,
Preferred
Alternative
.
.
.
.
.
.
.
.
.
.
6­
6
Exhibit
6.3
Initial
Capital
and
One­
Time
Nominal
Costs,
Preferred
Alternative
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
8
Exhibit
6.4a
Annualized
Total
Costs,
Preferred
Alternative,
at
3
Percent
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
10
Exhibit
6.4b
Annualized
Total
Costs,
Preferred
Alternative,
at
7
Percent
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
10
Exhibit
6.5
Present
Value
of
System
Implementation
Total
Costs
by
System
Size,
Preferred
Alternative
6­
12
Exhibit
6.6
Initial
Source
Water
Monitoring
Present
Value
Costs
at
3
and
7
Percent,
Preferred
Alternative,
by
System
Size
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
14
Exhibit
6.7
Methodology
for
Estimating
Treatment
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
15
Exhibit
6.8
Advanced
Technologies
Considered
for
the
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
19
Exhibit
6.9
Technology
Selection
Forecast
for
Filtered
Plants
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
22
Exhibit
6.10
Treatment
Costs
for
Filtered
and
Unfiltered
Plants,
Preferred
Alternative
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
23
Exhibit
6.11
Technology
Selection
for
Uncovered
Finished
Water
Reservoirs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
26
Exhibit
6.12
Total
Annualized
Cost
for
Uncovered
Finished
Water
Reservoirs
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
27
Exhibit
6.13
Present
Value
Costs
for
Future
Monitoring,
Preferred
Alternative
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
27
Exhibit
6.14
Summary
of
Annual
Per­
Household
Cost
Increases,
Preferred
Alternative
.
.
.
.
.
.
.
.
.
.
.
.
6­
29
Exhibit
6.15
Summary
of
Uncertainties
Affecting
LT2ESWTR
Cost
Estimates
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
30
Exhibit
6.16
Sensitivity
of
Treatment
Cost
to
Selection
of
Cryptosporidium
Occurrence
Distribution
.
.
6­
32
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
x
Exhibit
6.17
Sensitivity
of
Technology
Selection
to
Influent
Bromide
Concentration
for
Filtered
Plants
.
6­
33
Exhibit
6.18a
Comparison
by
Regulatory
Alternative
of
Total
Costs,
Annualized
at
3
Percent
for
Filtered
Plants
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
34
Exhibit
6.18b
Comparison
by
Regulatory
Alternative
of
Total
Costs,
Annualized
at
7
Percent
for
Filtered
Plants
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6­
34
Exhibit
7.1a
Estimated
Impacts
of
the
Proposed
LT2ESWTR
on
Small
Systems'
Technical,
Managerial,
and
Financial
Capacity
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
8
Exhibit
7.1b
Estimated
Impacts
of
the
Proposed
LT2ESWTR
on
Large
Systems'
Technical,
Managerial,
and
Financial
Capacity
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
9
Exhibit
7.2
Average
Annual
Burden
Hours
and
Costs
for
the
LT2ESWTR
Information
Collection
Request
3­
Year
Approval
Period
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
13
Exhibit
7.3
Annualized
Value
of
Public
and
Private
Costs
for
the
LT2ESWTR
(
Annualized
at
3
and
7
Percent)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
14
Exhibit
7.4
Total
Annualized
Benefits
and
Costs
of
Regulatory
Alternatives
($
Millions,
2000$)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
16
Exhibit
7.5
Percent
of
Population
of
CWSs
Served
by
Small
Surface
and
GWUDI
Systems
by
State
.
.
7­
20
Exhibit
7.6
Number
of
Small
Surface
and
GWUDI
Systems
by
State
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
21
Exhibit
7.7
Numbers
of
Indian
Tribal
Public
Water
Systems
Using
Surface
Water
Sources
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
25
Exhibit
7.8
Number
of
Tribal
Systems
and
Percent
of
Systems
Nationally
That
Will
Incur
Costs
Due
to
the
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
26
Exhibit
7.9a
Estimates
of
the
Total
Annualized
Costs
Incurred
by
Indian
Tribal
Public
Water
Systems
Due
to
the
LT2ESWTR
(
Annualized
at
3
Percent)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
27
Exhibit
7.9b
Estimates
of
the
Total
Annualized
Costs
Incurred
by
Indian
Tribal
Public
Water
Systems
Due
to
the
LT2ESWTR
(
Annualized
at
7
Percent)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
27
Exhibit
7.10
Total
Increased
Annual
National
Energy
Usage
Attributable
to
the
LT2ESWTR
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
33
Exhibit
7.11
Sample
Calculation
for
Determining
Increase
in
Energy
Usage:
Chlorine
Dioxide
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7­
35
Exhibit
8.1a
Summary
of
Nominal
Benefit
and
Cost
Estimates
by
Year
Incurred,
Preferred
Alternative,
ICR
Data
Set,
Enhanced
COI
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
2
Exhibit
8.1b
Summary
of
Nominal
Benefit
and
Cost
Estimates
by
Year
Incurred,
Preferred
Alternative,
ICR
Data
Set,
Traditional
COI
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
3
Exhibit
8.2
Summary
of
Nonquantified
Benefits
and
Groups
Affected
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
4
Exhibit
8.3
Summary
of
Annual
Avoided
Illnesses
and
Deaths,
Preferred
Alternative
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
5
Exhibit
8.4a
Summary
of
Quantified
Benefits,
Preferred
Alternative
 
Enhanced
Cost
of
Illness
.
.
.
.
.
.
.
8­
6
Exhibit
8.4b
Summary
of
Quantified
Benefits,
Preferred
Alternative
 
Traditional
Cost
of
Illness
.
.
.
.
.
.
8­
6
Exhibit
8.5
Summary
of
the
Costs
for
the
LT2ESWTR
Preferred
Regulatory
Alternative
.
.
.
.
.
.
.
.
.
.
.
.
8­
8
Exhibit
8.6a
Mean
Net
Benefits,
Preferred
Alternative
 
Enhanced
Cost
of
Illness
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
9
Exhibit
8.6b
Mean
Net
Benefits,
Preferred
Alternative
 
Traditional
Cost
of
Illness
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
10
Exhibit
8.7a
Breakeven
Points,
Enhanced
COI
(
Number
of
Avoided
Illnesses
Needed
to
Break
Even
with
Cost
Estimates)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
11
Exhibit
8.7b
Breakeven
Points,
Traditional
COI
(
Number
of
Avoided
Illnesses
Needed
to
Break
Even
with
Cost
Estimates)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
11
Exhibit
8.8
Summary
of
Binning
and
Treatment
Scenarios
for
Filtered
Systems
for
All
Regulatory
Alternatives
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
12
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
xi
Exhibit
8.9
Comparison
of
Number
of
Illnesses
and
Deaths
Avoided
for
All
Regulatory
Alternatives
.
.
.
8­
14
Exhibit
8.10a
Comparison
of
Annualized
Value
of
Illnesses
and
Deaths
Avoided
for
All
Regulatory
Alternatives
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
15
Exhibit
8.10b
Comparison
of
Annualized
Value
of
Illnesses
and
Deaths
Avoided
for
All
Regulatory
Alternatives,
Traditional
COI
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
16
Exhibit
8.11
Comparison
of
Costs
for
All
Regulatory
Alternatives
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
17
Exhibit
8.12a
Comparison
of
Mean
Net
Benefits
for
All
Regulatory
Alternatives
 
Enhanced
COI
.
.
.
.
.
8­
18
Exhibit
8.12b
Comparison
of
Mean
Net
Benefits
for
All
Regulatory
Alternatives
 
Traditional
COI
.
.
.
.
8­
19
Exhibit
8.13a
Incremental
Net
Benefits
for
All
Alternatives,
By
Data
Set,
Enhanced
COI
.
.
.
.
.
.
.
.
.
.
.
8­
21
Exhibit
8.13b
Incremental
Net
Benefits
for
All
Alternatives,
By
Data
Set,
Traditional
COI
.
.
.
.
.
.
.
.
.
.
8­
22
Exhibit
8.14a
Upper
End
of
90
Percent
Confidence
Bound
as
a
Percent
of
Mean
Estimate
of
Benefits,
By
Data
Set,
Annualized
at
3
Percent
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
24
Exhibit
8.14b
Upper
End
of
90
Percent
Confidence
Bound
as
a
Percent
of
Mean
Estimate
of
Benefits,
By
Data
Set,
Annualized
at
3
Percent,
Traditional
COI
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
25
Exhibit
8.15a
Comparison
of
Regulatory
Alternatives
Ranked
by
Net
Benefits,
3
Percent
Cost
­
Enhanced
COI
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
27
Exhibit
8.15b
Comparison
of
Regulatory
Alternatives
Ranked
by
Net
Benefits,
7
Percent
Cost
­
Enhanced
COI
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
28
Exhibit
8.16a
Comparison
of
Regulatory
Alternatives
Ranked
by
Net
Benefits,
3
Percent
Cost
­
Traditional
COI
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
29
Exhibit
8.16b
Comparison
of
Regulatory
Alternatives
Ranked
by
Net
Benefits,
7
Percent
Cost
­
Traditional
COI
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
30
Exhibit
8.17a
Range
of
Annualized
Costs
at
Mean
Benefit
Level,
All
Regulatory
Alternatives
 
­
Enhanced
Cost
of
Illness
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
32
Exhibit
8.17b
Range
of
Annualized
Costs
at
Mean
Benefit
Level,
All
Regulatory
Alternatives
 
Traditional
Cost
of
Illness
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
33
Exhibit
8.18
Cost
per
Illness
or
Death
Avoided,
By
Data
Set,
COI
Type
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
35
Exhibit
8.19
Effects
of
Uncertainties
on
the
National
Estimates
of
Benefits
and
Costs
.
.
.
.
.
.
.
.
.
.
.
.
.
8­
36
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
xii
Acronyms
and
Notations
AIDS
Acquired
Immunodeficiency
Syndrome
AIPC
All
Indian
Pueblo
Council
AMWA
Association
of
Metropolitan
Water
Agencies
ASDWA
Association
of
State
Drinking
Water
Administrators
AWWA
American
Water
Works
Association
CC­
PCR
Cell
culture
and
polymerase
chain
reaction
CCR
Consumer
Confidence
Report
Rule
(
1998)
CDBG
Community
Development
Block
Grant
CDC
Centers
for
Disease
Control
and
Prevention
CFE
Combined
Filter
Effluent
CFU
Colony
forming
unit
CL2
Chlorine
CLM
Chloramines
CLO2
Chlorine
Dioxide
COI
Cost
of
Illness
CPI
Consumer
Price
Index
CSFII
Continuing
Survey
of
Food
Intakes
by
Individuals
CWS
Community
Water
System
CWSS
Community
Water
Systems
Survey
DBPs
Disinfection
Byproducts
DWSRF
Drinking
Water
State
Revolving
Fund
EA
Economic
Analysis
EO
Executive
Order
FACA
Federal
Advisory
Committees
Act
FBRR
Filter
Backwash
Recycling
Rule
(
May,
2001)
FR
Federal
Register
FS
Flowing
stream
FTE
Full­
time
Equivalent
Employee
GDP
Gross
Domestic
Product
GWR
Ground
Water
Rule
(
proposed
2000)
GWUDI
Ground
Water
Under
the
Direct
Influence
of
Surface
Water
HAA5
Haloacetic
Acids
[
total
of
five]
HPC
Heterotrophic
Plate
Count
HRRCA
Health
Risk
Reduction
and
Cost
Analysis
ICR
Information
Collection
Request
ICR
Information
Collection
Rule
(
1996)
ICRSS
Information
Collection
Rule
Supplemental
Survey
ICRSSM
Information
Collection
Rule
Supplemental
Survey
Medium
Systems
ICRSSL
Information
Collection
Rule
Supplemental
Survey
Large
Systems
ICMA
International
City/
County
Management
Association
IDSE
Initial
Distribution
System
Evaluation
IESWTR
Interim
Enhanced
Surface
Water
Treatment
Rule
(
1998)
IMS
Immunomagnetic
separation
IRFA
Initial
Regulatory
Flexibility
Analysis
IRIS
Integrated
Risk
Information
System
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
xiii
LCR
Lead
and
Copper
Rule
(
1992)
LRAA
Locational
Running
Annual
Average
LSP
Lab
Spiking
Program
LT1ESWTR
Long
Term
1
Enhanced
Surface
Water
Treatment
Rule
(
January,
2002)
LT2ESWTR
Long
Term
2
Enhanced
Surface
Water
Treatment
Rule
(
under
development)
MCL
Maximum
Contaminant
Level
MCLG
Maximum
Contaminant
Level
Goal
MCMC
Markov
Chain
Monte
Carlo
[
algorithm]
M­
DBP
Microbial­
Disinfectants/
Disinfection
Byproducts
[
Advisory
Committee]
MF
Microfiltration
MG
Million
gallons
MGD
Million
Gallons
per
Day
mg/
L
Milligrams
per
Liter
MRDL
Maximum
Residual
Disinfectant
Level
MRDLG
Maximum
Residual
Disinfectant
Level
Goal
MRL
Minimum
Reporting
Level
MWRA
Massachusetts
Water
Resources
Authority
NCSL
National
Conference
of
State
Legislatures
NCWS
Noncommunity
water
system
NF
Nanofiltration
NGA
National
Governors'
Association
NIH
National
Institutes
of
Health
NLC
National
League
of
Cities
NODA
Notice
of
Data
Availability
NPDWR
National
Primary
Drinking
Water
Regulations
NRDC
Natural
Resources
Defense
Council
NRWA
National
Rural
Water
Association
NTNCWS
Nontransient
Noncommunity
Water
System
NTU
Nephelometric
Turbidity
Unit
O3
Ozone
OGWDW
Office
of
Ground
Water
and
Drinking
Water
OMB
Office
of
Management
and
Budget
O&
M
Operations
and
Maintenance
O&
P
Overhead
and
Profit
POE
Point
of
entry
POU
Point
of
Use
ppb
Parts
per
Billion
ppm
Parts
per
Million
PWS
Public
Water
System
PWSS
Public
Water
Systems
Supervision
[
Grants
Program]
QA/
QC
Quality
Assurance/
Quality
Control
RAA
Running
Annual
Average
RF
Roughing
Filter
RFA
Regulatory
Flexibility
Act
RIA
Regulatory
Impact
Analysis
RL
Reservoir/
Lake
RO
Reverse
Osmosis
RUS
Rural
Utility
Service
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
xiv
SAB
Science
Advisory
Board
SBAR
Small
Business
Advocacy
Review
SBREFA
Small
Business
Regulatory
Enforcement
Fairness
Act
SDWA
Safe
Drinking
Water
Act
(
1974)
SDWIS
Safe
Drinking
Water
Information
System
SER
Small
Entity
Representative
SF
Secondary
Filter
SW
Surface
Water
SWAT
Surface
Water
Analytical
Tool
Stage
1
DBPR
Stage
1
Disinfectants
and
Disinfection
Byproducts
Rule
(
1998)
Stage
2
DBPR
Stage
2
Disinfectants
and
Disinfection
Byproducts
Rule
(
under
development)
SWTR
Surface
Water
Treatment
Rule
(
1989)
T&
C
Technologies
and
Cost
TCR
Total
Coliform
Rule
(
1989)
TMF
Technical,
Managerial,
and
Financial
TNCWS
Transient
Noncommunity
Water
System
TOC
Total
Organic
Carbon
TT
Treatment
Technique
TTHMs
Total
Trihalomethanes
TTHMR
Total
Trihalomethane
Rule
(
1979)
TWG
Technical
Workgroup
UF
Ultrafiltration
UMRA
Unfunded
Mandates
Reform
Act
USDA
United
States
Department
of
Agriculture
UV
Ultraviolet
[
Light
Disinfection]
µ
g/
L
Micrograms
per
Liter
VSL
Value
of
a
Statistical
Life
WC
Watershed
Control
WTP
Willingness
to
Pay
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
xv
Health
Risk
Reduction
and
Cost
Analysis
Under
the
Safe
Drinking
Water
Act
(
SDWA)
Amendments
of
1996,
when
proposing
a
national
primary
drinking
water
regulation
that
includes
an
maximum
contaminant
level
(
MCL),
the
U.
S.
Environmental
Protection
Agency
(
EPA
or
the
Agency)
must
conduct
a
health
risk
reduction
and
cost
analysis
(
HRRCA).
A
HRRCA
addresses
seven
requirements,
all
of
which
are
addressed
in
this
Economic
Analysis
(
EA)
for
the
Long
Term
2
Enhanced
Surface
Water
Treatment
Rule
(
LT2ESWTR).

HRRCA
Crosswalk
Summary
HRRCA
Requirement
Addressed
in
Economic
Analysis
Quantifiable
and
nonquantifiable
health
risk
reduction
benefits
Chapter
5
(
All
sections
and
exhibits)
Chapter
8
(
Sections
8.2,
8.3;
Exhibits
8.1
 
8.4,
8.6,
8.7,
8.9,
8.10,
8.12­
17)

Quantifiable
and
nonquantifiable
health
risk
reduction
benefits
from
co­
occurring
contaminants
Chapter
5
(
Section
5.4.4)

Quantifiable
and
nonquantifiable
costs
Chapter
6
(
All
sections
and
exhibits)
Chapter
7
(
Sections
7.2,
7.6,
7.7,
7.8,
7.12;
Exhibits
7.2,
7.3,
7.4,
7.9)
Chapter
8
(
Sections
8.2,
8.3;
Exhibit
8.1,
8.5,
8.8,
8.11,
8.18)

Incremental
costs
and
benefits
associated
with
regulatory
alternatives
Chapter
5
(
Section
5.6;
Exhibit
5.28)
Chapter
6
(
Sections
6.12;
Exhibit
6.18)
Chapter
8
(
Section
8.3;
Exhibits
8.13,
8.17)

Effects
of
the
contaminants
on
the
general
population
and
sensitive
subpopulations
Chapter
5
(
Sections
5.2.2
and
5.2.8)
Chapter
7
(
Section
7.9)

Increased
health
risk
that
may
occur
as
a
result
of
compliance
Chapter
2
(
Section
2.3)

Other
relevant
factors
(
quality
and
uncertainty
of
information)
Chapter
4
(
Section
4.8;
Exhibit
4.28)
Chapter
5
(
Section
5.5;
Exhibit
5.26)
Chapter
6
(
Section
6.11;
Exhibit
6.15)
Chapter
8
(
Section
8.4;
Exhibit
8.19)
1The
key
outcomes
of
the
1992­
1993
regulatory
negotiation
effort
were
to
proceed
with
rules
addressing
DBPs
and
microbial
pathogens
in
two
stages
and
to
collect
relevant
information
from
PWSs
for
use
in
the
development
of
these
rules
and
the
analysis
of
their
impacts.
This
two­
stage
approach
was
subsequently
incorporated
into
the
1996
Safe
Drinking
Water
Act
(
SDWA)
Amendments.
The
first
stage
of
the
M­
DBP
rulemaking
process
culminated
with
the
joint
promulgation
of
the
Stage
1
DBPR
and
the
Interim
Enhanced
Surface
Water
Treatment
Rule
(
IESWTR)
by
EPA
in
December
1998.

Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
1
Executive
Summary
This
document
presents
the
Economic
Analysis
(
EA),
prepared
by
the
U.
S.
Environmental
Protection
Agency
(
EPA),
of
the
benefits
and
costs
of
the
proposed
Long
Term
2
Enhanced
Surface
Water
Treatment
Rule
(
LT2ESWTR).
Executive
Order
12866
requires
federal
agencies
to
conduct
an
analysis
of
the
benefits
and
costs
of
proposed
and
final
rules
that
cost
over
$
100
million
annually.

ES.
1
Need
for
the
Rule
Over
14,000
public
water
systems
(
PWSs),
serving
approximately
180
million
people
in
the
United
States
and
its
Territories,
use
surface
water,
including
ground
water
under
the
direct
influence
(
GWUDI)
of
surface
water,
as
their
source.
These
sources
often
carry
microbial
pathogens,
such
as
Giardia,
E.
coli,
and
Cryptosporidium.
Among
pathogens
in
drinking
water,
Cryptosporidium
is
of
particular
concern
because
it
is
resistant
to
standard
drinking
water
disinfectants,
such
as
chlorine.
Ingestion
of
Cryptosporidium
causes
cryptosporidiosis,
a
gastrointestinal
illness,
and
health
effects
in
sensitive
subpopulations
may
be
severe,
including
risk
of
death.
There
is
no
effective
drug
or
treatment
available
to
cure
cryptosporidiosis
(
Framm
and
Soave
1997).
The
LT2ESWTR
provides
additional
protection
of
public
health
by
requiring
PWSs
with
the
highest
measured
source
water
levels
of
Cryptosporidium
to
provide
additional
treatment
for
this
pathogen.
These
vulnerable
systems
include
both
filtered
systems
with
elevated
levels
of
Cryptosporidium
in
their
source
water
and
all
unfiltered
systems
(
i.
e.,
systems
meeting
the
filtration
avoidance
criteria
of
the
Surface
Water
Treatment
Rule
(
40
CFR
141.71)),
which
currently
provide
little
treatment
for
Cryptosporidium.
The
LT2ESWTR
will
also
reduce
the
public
health
risk
associated
with
uncovered
finished
water
reservoirs,
which
are
susceptible
to
microbial
contamination,
including
Cryptosporidium.

EPA
will
also
be
proposing
the
Stage
2
Disinfectants
and
Disinfection
Byproducts
Rule
(
Stage
2
DBPR),
which
addresses
disinfection
byproduct
(
DBP)
formation.
These
rules
are
to
be
promulgated
concurrently
to
ensure
that
protection
against
microbial
pathogens
is
not
compromised
by
efforts
to
reduce
DBP
formation.
The
Stage
2
DBPR
and
LT2ESWTR
represent
the
final
stage
of
a
two­
stage
strategy
to
reduce
risk
from
microbial
pathogens
and
DBPs
that
was
developed
in
a
regulatory
negotiation
effort
in
1992
and
1993.1
These
rules
reflect
recommendations
presented
by
the
Stage
2
Microbial
and
Disinfection
Byproducts
(
M­
DBP)
Federal
Advisory
Committee
Agreement
in
Principle,
signed
in
September
2000
(
USEPA
2000i).
2The
term
"
log
treatment"
is
used
to
express
the
expected
percent
reduction
of
a
contaminant.
For
example,
1
log
treatment
is
expected
to
provide
90
percent
reduction
of
a
contaminant
and
2
log
treatment
provides
99
percent
reduction.
Compliance
with
the
log
treatment
requirements
is
not
based
on
quantifying
the
actual
reduction;
instead,
other
finished
water
quality
or
operational
conditions
are
used
to
determine
compliance.
This
is
consistent
with
the
previous
SWTRs
where
EPA
assumes
systems
achieve
a
given
log
removal
of
Giardia
and
viruses
if
their
effluent
turbidity
meets
certain
standards.

Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
2
ES.
2
Consideration
of
Regulatory
Alternatives
The
Stage
2
M­
DBP
Advisory
Committee
met
from
March
1999
to
September
2000
to
evaluate
whether
and
to
what
degree
EPA
should
revise
or
establish
additional
microbial
standards
to
protect
public
health.
The
committee
reached
consensus
on
an
approach
for
addressing
Cryptosporidium
risk
in
unfiltered
systems
and
microbial
risk
in
uncovered
finished
water
reservoirs
without
formally
identifying
alternative
regulatory
approaches.
However,
for
filtered
systems,
several
alternatives
were
considered.
All
alternatives
involved
a
treatment
compliance
scheme
where
systems
have
a
"
toolbox"
of
treatment
and
managerial
options
for
meeting
additional
Cryptosporidium
treatment
requirements.
Three
of
the
four
alternatives
(
A2­
A4)
base
the
treatment
requirements
on
a
plant's
Cryptosporidium
levels
in
source
water,
thus
requiring
a
period
of
source
water
monitoring.
The
fourth
would
require
a
uniform
treatment
target
regardless
of
source
water
Crytosporidium
levels.
The
four
alternatives
considered
for
filtered
systems
are
the
following:

°
Alternative
1:
For
all
systems,
2
log
Cryptosporidium
treatment2
in
addition
to
that
required
for
the
IESWTR
and
Long
Term
1
Enhanced
Surface
Water
Treatment
Rule
(
LT1ESWTR).

°
Alternative
2:
Systems
having
mean
source
water
Cryptosporidium
levels
of
0.03
to
less
than
0.1
oocysts/
L
must
achieve
an
additional
0.5
log
Cryptosporidium
treatment.
Systems
with
0.1­
1.0
oocysts/
L
must
add
1.5
log
treatment,
and
systems
with
greater
than
1.0
oocysts/
L
must
add
2.5
log
treatment.
Additional
treatment
requirements
as
stated
are
for
systems
using
conventional
treatment
or
equivalent.
Requirements
for
systems
with
other
treatment
types
may
differ.

°
Alternative
3
(
Preferred
Alternative):
Systems
having
mean
source
water
Cryptosporidium
levels
of
0.075­
1.0
oocysts/
L
must
achieve
an
additional
1.0
log
treatment
for
Cryptosporidium.
Systems
with
1.0­
3.0
oocysts/
L
must
achieve
2.0
log
additional
treatment,
and
systems
with
greater
than
3.0
oocysts/
L
must
achieve
2.5
log
additional
treatment.
Additional
treatment
requirements
as
stated
are
for
systems
using
conventional
treatment
or
equivalent.
Requirements
for
systems
with
other
treatment
types
may
differ.

°
Alternative
4:
Systems
having
mean
source
water
Cryptosporidium
levels
of
0.1­
1.0
oocysts/
L
must
achieve
0.5
log
additional
treatment
for
Cryptosporidium
and
systems
that
have
greater
than
1.0
oocysts/
L
must
achieve
1.0
log
additional
treatment.
Additional
treatment
requirements
as
stated
are
for
systems
using
conventional
treatment
or
equivalent.
Requirements
for
systems
with
other
treatment
types
may
differ.
Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
3
In
this
EA,
EPA
estimates
and
compares
the
costs
and
benefits
of
all
regulatory
alternatives.
The
cost
and
benefit
data
presented
in
section
ES.
4
are
for
the
Preferred
Regulatory
Alternative,
while
section
ES.
5
presents
comparisons
of
the
cost
and
benefit
estimates
for
the
regulatory
alternatives.

ES.
3
Summary
of
the
LT2ESWTR
Requirements
The
LT2ESWTR
applies
to
all
PWSs
that
use
surface
water
or
GWUDI
(
excluding
those
that
purchase
all
their
water).
It
builds
on
the
SWTR,
IESWTR,
and
the
LT1ESWTR
by
improving
control
of
microbial
pathogens,
specifically
Cryptosporidium.
Unlike
the
previous
rules,
the
LT2ESWTR
bases
treatment
requirements
on
a
system's
source
water
Cryptosporidium
concentration
and
type
of
treatment
provided.
This
rule
requires
systems
to
monitor
their
source
water
for
Cryptosporidium,
and
based
on
the
results,
to
meet
one
of
four
levels
of
treatment
for
Cryptosporidium
(
with
the
first
level
requiring
no
additional
treatment).
The
levels
of
treatment
needed
will
be
reassessed
in
the
future
based
on
a
second
round
of
source
water
monitoring.

For
systems
that
do
not
provide
filtration,
the
LT2ESWTR
requires
2
or
3
log
inactivation
of
Cryptosporidium,
depending
on
source
water
monitoring
results.
The
rule
also
requires
systems
with
uncovered
finished
water
reservoirs
to
either
cover
the
reservoirs
or
provide
additional
treatment
to
the
reservoir
effluent
unless
the
state
determines
that
existing
risk
mitigation
is
adequate.
There
are
also
requirements
for
calculating
benchmarks
of
disinfection
performance
to
help
evaluate
future
changes
to
the
disinfection
process.

Exhibit
ES.
1
illustrates
each
rule
activity
for
the
Preferred
Regulatory
Alternative.
The
implementation
schedule
for
small
and
large
systems
differs,
and
Exhibit
ES.
2
presents
the
implementation
timeline.
Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
4
Systems
subject
to
the
LT2ESWTR
Public
water
systems
using
surface
water
or
ground
water
under
direct
influence
of
surface
water
that
are
subject
to
the
SWTR.

Disinfection
Profiling
and
Benchmarking
Community
and
nontransient
noncommunity
water
systems
must
calculate
disinfection
profiles
for
both
Giardia
and
viruses,
except
for
small
systems
with
E.
coli
and
total
trihalomethanes
(
TTHMs)
and
five
haloacetic
acids
(
HAA5)
below
certain
levels
and
systems
that
provide
at
least
5.5
log
treatment
for
Cryptosporidium.
Monitoring
may
be
required
to
establish
a
benchmark
if
historical
data
cannot
be
used.

Requirements
for
Uncovered
Finished
Water
Reservoirs
All
systems
with
uncovered
finished
water
reservoirs
must
either
cover
the
reservoir
or
treat
reservoir
discharge
to
the
distribution
system
to
achieve
4
log
virus
inactivation,
unless
the
state/
primacy
agency
determines
that
existing
risk
mitigation
is
adequate.

Monitoring
and
Treatment
Requirements
for
Unfiltered
Systems
Systems
that
do
not
provide
filtration
and
meet
all
the
filtration
avoidance
criteria
of
the
SWTR
must
provide
2
log
of
Cryptosporidium
inactivation
if
their
Cryptosporidium
occurrence
level
is
less
than
or
equal
to
1
oocyst
in
100
liters,
or
provide
3
log
of
Cryptosporidium
inactivation
if
their
Cryptosporidium
occurrence
level
is
greater
than
1
oocyst
in
100
liters.
Large
systems
monitor
their
source
water
monthly
for
Cryptosporidium
for
2
years
and
small
systems
monitor
twice
per
month
for
1
year.
Initial
Monitoring
for
Filtered
Systems
Systems
must
conduct
initial
Cryptosporidium
monitoring
to
establish
bin
classification
under
the
LT2ESWTR.
Systems
purchasing
all
of
their
water
or
providing
at
least
5.5
log
of
treatment
do
not
have
to
monitor
or
provide
additional
treatment.
Also,
systems
with
at
least
2
years
of
historical
Cryptosporidium
data
that
are
equivalent
in
sample
number,
frequency,
and
data
quality
to
data
collected
under
the
LT2ESWTR
may
submit
their
data
to
EPA
for
consideration
in
determining
bin
classification
in
lieu
of
additional
Cryptosporidium
compliance
monitoring.

Medium
and
Large
Systems
(
serving
at
least
Small
Systems
(
serving
<
10,000
people)
10,000
people)
Source
water
monitoring
will
begin
2
years
after
the
large
and
Systems
will
be
required
to
medium
systems
initiate
source­
water
Cryptosporidium
monitor
their
raw
water
sources
at
least
once
per
monitoring.
Monitoring
is
on
a
delayed
schedule
while
EPA
month
for
a
minimum
of
2
years
for
continues
to
evaluate
indicators
and
system
characterization
Cryptosporidium,
E.
coli,
and
turbidity.
scenarios
for
predicting
Cryptosporidiumoccurrence.
In
the
absence
of
a
new
surrogate
indicator,
small
systems
will
conduct
1
year
of
biweekly
E.
coli
source
water
monitoring
and
will
conduct
Cryptosporidium
monitoring
if
E.
coli
levels
exceed
specified
trigger
values.

Bin
Classification
and
Treatment
Requirement
for
Filtered
Systems
For
those
required
to
monitor
for
Cryptosporidium,
the
results
of
initial
monitoring
will
be
used
to
place
them
into
an
"
action
bin,"
which
defines
the
additional
levels
of
Cryptosporidium
removal
that
must
be
achieved
Systems
whose
Cryptosporidium
occurrence
is
less
than
0.075
oocysts/
L
have
no
additional
treatment
requirements.
Systems
whose
oocyst
occurrence
is
0.075
or
more,
but
<
1.0/
L
must
provide
an
additional
1
log
of
treatment.
Systems
whose
oocyst
occurrence
is
equal
to
or
greater
than
1.0/
L
but
less
than
3.0/
L
must
provide
an
additional
2
log
of
treatment.
Systems
whose
oocyst
occurrence
is
equal
to
or
greater
than
3.0/
L
must
provide
an
additional
2.5
logs
of
treatment.

Microbial
Toolbox
for
Meeting
Additional
Treatment
Requirements
Systems
can
select
from
a
"
toolbox"
of
treatment
or
management
options.

Reassessment
and
Future
Monitoring
Six
years
after
initial
bin
classification,
filtered
systems
will
be
required
to
conduct
a
second
round
of
monitoring
to
reasses
s
the
source
water
quality.
Two
years
prior
to
this
reassessment
(
4
years
after
initial
binning),
EPA
plans
to
initiate
a
stakeholder
process
to
review
available
methods
to
determine
the
appropriate
analytical
method,
monitoring
frequency,
and
monitoring
location
for
the
second
round
of
national
assessment
monitoring.
In
addition,
the
state/
primacy
agency
will
assess
any
significant
changes
in
the
watershed
and
source
water
as
part
of
the
sanitary
survey
process
and
determine
appropriate
follow­
up
action
in
response
to
source
water
changes,
which
could
include
actions
from
the
microbial
toolbox.
Unfiltered
systems
must
meet
Cryptosporidium
inactivation
requirements
using
O
3,
ClO
2,
or
UV.
Exhibit
ES.
1
Overview
of
Key
LT2ESWTR
Requirements
Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
5
States
Crypto
Monitoring
Apply
for
Primacy
Comply
with
Stage
2A
Year
1
Year
2
Year
3
Year
4
Year
5
Year
6
Year
7
Year
8
Year
9
Year
10
Year
11
Year
1
Year
2
Year
3
Year
4
Year
5
Year
6
Year
7
Year
8
Year
9
Year
10
Year
11
Large
and
Medium
Filtered
and
Unfiltered
Systems
(
Serving
>
10,000
People)

Bin
Determination
Treatment
Installation
Small
Filtered
Systems
(
Serving
<
10,000
People)

E.
Coli
monitoring
CryptoMonitoring
Bin
Determination
Treatment
Installation
Bin
Determination
Crypto
Compliance
Crypto
Compliance
Crypto
Compliance
Small
Unfiltered
Systems
(
Serving
<
10,000
People)
CryptoMonitoring
Final
Stage
2
DBPR
&
LT2ESWTR
Rules
Promulgated
Treatment
Installation
(
possible
2
­
year
extension)

(
possible
2
­
year
extension)

(
possible
2
­
year
extension)
Exhibit
ES.
2
Implementation
Timeline
for
LT2ESWTR
ES.
4
National
Benefits
and
Costs
of
the
LT2ESWTR
The
benefits
resulting
from
implementation
of
the
LT2ESWTR
are
due
to
reductions
in
infectious
Cryptosporidium
oocsysts
reaching
the
consumer.
EPA
quantified
the
benefits
of
this
rule
in
terms
of
avoided
endemic
cryptosporidiosis
illnesses
and
associated
deaths
avoided.
Cryptosporidium
can
reach
the
consumer
when
there
is
a
significant
breakdown
in
the
treatment
process
or
during
normal
operating
conditions.
This
EA
focuses
on
the
benefits
resulting
from
reducing
the
continuous,
relatively
low
levels
of
Cryptosporidium
exposure
that
can
occur
even
under
normal
operating
conditions.
Although
requiring
additional
treatment
is
expected
to
also
reduce
the
likelihood
of
an
outbreak,
these
benefits
are
not
quantified.

The
costs
incurred
for
LT2ESWTR
activities
are
associated
with
rule
implementation,
source
water
monitoring,
disinfection
benchmarking,
and
adding
treatment.
Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
6
ES.
4.1
Benefit
Estimates
The
quantified
benefits
of
the
LT2ESWTR
are
the
estimated
reduction
in
the
number
of
endemic
illnesses
and
deaths
associated
with
Cryptosporidium
under
LT2ESWTR
conditions.
There
are
also
benefits
that
cannot
currently
be
quantified
but
are
likely
to
be
substantial.
These
are
summarized
in
Exhibit
ES.
3
and
discussed
in
Chapter
5.

Exhibit
ES.
3
Summary
of
Nonquantified
Benefit
and
Groups
Affected
Type
of
Benefit
Nonquantified
Benefits
Group(
s)
Affected
Health
benefits
Reduction
in
risk
to
sensitive
subpopulations
(
mortality
for
those
with
AIDS
and
other
sensitive
subpopulations
has
been
included)
Immunocompromised
individuals
served
by
systems
that
make
changes
to
or
add
treatment.

Reduction
in
health
risk
during
outbreaks
(
and
response
costs)
All
individuals
served
by
systems
that
make
changes
to
or
add
treatment,
including
those
now
served
by
uncovered
finished
water
reservoirs,
(
between
34
and
55
million
people).
Reduction
in
co­
occurring/
emerging
pathogen
risk
Reduction
in
endemic
morbidity
and
mortality
risk
associated
with
uncovered
finished
water
reservoirs
All
individuals
receiving
water
from
uncovered
finished
water
reservoirs.

Nonhealth
Benefits
Improved
aesthetic
water
quality
All
individuals
served
by
systems
that
make
changes
to
or
add
treatment
that
is
likely
to
reduce
taste
and
odor
problems
(
e.
g.,
ozone).

Costs
of
averting
behaviors
Potentially
all
households
served
by
systems
covered
by
the
rule,
either
because
monitoring
confirms
low
levels
of
risk
or
because
the
addition
of
treatment
assures
low
levels
of
risk.

EPA
developed
a
risk
assessment
model
to
incorporate
several
factors
and
their
uncertainties
for
predicting
the
illnesses
and
deaths
avoided.
Specifically,
the
variables
in
the
model
are
source
water
occurrence
of
Cryptosporidium,
infectivity,
treatment,
consumption,
morbidity,
and
mortality.
To
allow
a
comparison
of
benefits
with
the
cost
of
implementing
the
rule,
the
benefit
estimates
are
monetized
by
calculating
a
cost
of
illness
(
COI)
and
using
a
value
of
a
statistical
life
and
applying
those
factors
to
the
number
of
avoided
illnesses
and
deaths,
respectively.

For
source
water
occurrence,
there
are
three
data
sets
 
labeled
as
ICR,
ICRSSM,
and
ICRSSL.
The
ICR
data
set
is
from
the
Information
Collection
Rule
data
set
of
all
large
systems.
The
ICRSSM
and
ICRSSL
are
the
supplemental
survey
data
sets
of
40
medium­
size
systems
and
40
large
systems.
EPA
judges
each
of
these
data
sets
to
be
equally
likely
to
represent
the
true
distribution
of
Cryptosporidium
in
Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
7
Lower
(
5th
%
ile)
Upper
(
95th
%
ile)
Lower
(
5th
%
ile)
Upper
(
95th
%
ile)

ICR
1,018,915
169,358
2,331,467
141
25
308
ICRSSL
256,173
45,292
560,648
37
7
78
ICRSSM
498,363
84,724
1,177,415
70
13
157
ICR
720,668
119,694
1,647,796
100
18
218
ICRSSL
181,387
32,179
396,845
26
5
55
ICRSSM
352,611
59,942
833,290
50
9
111
Annual
Average
over
25
years
Total
after
Full
implementation
Data
Set
Annual
Illnesses
Avoided
Annual
Deaths
Avoided
Mean
90%
Confidence
Bound
Mean
90%
Confidence
Bound
source
waters
for
all
systems.
All
benefit
and
cost
analyses
are
carried
out
using
each
data
set
to
provide
a
range
of
possible
benefits
and
costs.
In
addition
to
using
the
three
occurrence
data
sets,
this
EA
monetizes
benefits
with
two
values
of
cost­
of­
illness
 
referred
to
as
Enhanced
and
Traditional.

Exhibit
ES.
4
summarizes
the
estimates
of
avoided
illnesses
and
deaths
as
a
result
of
the
LT2ESWTR.
Exhibits
ES.
5a
and
ES.
5b
summarize
the
monetized
value
of
those
estimates
for
Enhanced
and
Traditional
cost
of
illness
values
(
annualized
over
a
25­
year
period
and
discounted
at
3
percent
and
7
percent).

Exhibit
ES.
4
Summary
of
Annual
Avoided
Illnesses
and
Deaths
Source:
Chapter
8,
Exhibit
8.3.
Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
8
Lower
(
5th
%
ile)
Upper
(
95th
%
ile)

ICR
1,445
$
198
$
3,666
$
ICRSSL
374
$
52
$
959
$
ICRSSM
715
$
96
$
1,849
$

ICR
1,230
$
168
$
3,120
$
ICRSSL
318
$
44
$
816
$
ICRSSM
609
$
81
$
1,577
$
Annualized
Value
(
at
7%,
25
Years)
Annualized
Value
(
at
3%,
25
Years)
Data
Set
Value
of
Benefits
($
Millions,
2000$)

Mean
90%
Confidence
Bound
Lower
(
5th
%
ile)
Upper
(
95th
%
ile)

ICR
967
$
105
$
2,713
$
ICRSSL
253
$
27
$
713
$
ICRSSM
481
$
50
$
1,372
$

ICR
826
$
89
$
2,315
$
ICRSSL
216
$
23
$
610
$
ICRSSM
411
$
43
$
1,172
$
Annualized
Value
(
at
3%,
25
Years)

Annualized
Value
(
at
7%,
25
Years)
Data
Set
Value
of
Benefits
($
Millions,
2000$)

Mean
90%
Confidence
Bound
Exhibit
ES.
5a
Summary
of
Monetized
Benefits
 
Enhanced
Cost
of
Illness
[
1]

Exhibit
ES.
5b
Summary
of
Monetized
Benefits
 
Traditional
Cost
of
Illness
[
1]

Notes:
[
1]
The
traditional
COI
only
includes
valuation
for
medical
costs
and
lost
work
time
(
including
some
portion
of
unpaid
household
production).
The
enhanced
COI
also
factors
in
valuations
for
lost
personal
time
(
non­
work
time)
such
as
child
care
and
homemaking
(
to
the
extent
not
covered
by
the
traditional
COI),
time
with
family,
and
recreation,
and
lost
productivity
at
work
on
days
when
workers
are
ill
but
go
to
work
anyway.
Source:
Chapter
8,
Exhibits
8.4a
and
8.4b.
3Derived
from
Exhibit
6.3
using
ICR
data.

Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
9
ES.
4.2
National
and
Household
Cost
Estimates
The
total
national
costs
of
the
LT2ESWTR
include
costs
to
systems
and
States/
Primacy
Agencies
for
implementation
and
compliance.
Specifically,
this
EA
estimates
costs
for
all
rule
activities
including:
rule
implementation,
source
water
monitoring,
disinfection
benchmarking,
adding
treatment,
and
compliance
reporting.
EPA
assumes
nearly
all
surface
water
and
GWUDI
systems
will
incur
rule
implementation
and
initial
source
water
monitoring
costs.
Disinfection
benchmarking
and
compliance
reporting
are
estimated
for
only
those
systems
predicted
to
add
treatment.

Approximately
90
percent3
of
the
estimated
total
national
costs
are
for
systems
to
meet
additional
treatment
requirements.
EPA
developed
a
"
least­
cost"
approach
to
modeling
treatment
costs.
The
following
series
of
steps
were
used
to
develop
treatment
cost
estimates
for
compliance
with
the
rule.

1.
Predict
the
percent
of
systems
falling
into
each
bin
from
modeled
source
water
occurrence
2.
Model
unit
costs
for
each
treatment
technology
3.
Develop
a
technology
forecast
for
each
bin
using
the
least­
cost
approach
and
estimates
of
maximum
use
of
any
one
technology
4.
Calculate
the
number
of
plants
selecting
each
technology
5.
Multiply
the
number
of
plants
per
technology
by
the
technology
unit
cost
Exhibit
ES.
6
summarizes
the
system
costs
associated
with
the
LT2ESWTR.
Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
10
Lower
(
5th
%
ile)
Upper
(
95th
%
ile)
Lower
(
5th
%
ile)
Upper
(
95th
%
ile)

ICR
1,795
$
1,549
$
2,042
$
52
$
49
$
54
$
ICRSSL
1,208
$
1,041
$
1,379
$
32
$
31
$
34
$
ICRSSM
1,412
$
1,218
$
1,612
$
39
$
36
$
41
$

Lower
(
5th
%
ile)
Upper
(
95th
%
ile)
Lower
(
5th
%
ile)
Upper
(
95th
%
ile)
ICR
111
$
98
$
123
$
121
$
107
$
135
$
ICRSSL
73
$
65
$
82
$
81
$
71
$
91
$
ICRSSM
86
$
76
$
96
$
94
$
83
$
106
$
Nominal
Costs
Total
Annualized
Costs
at
7%
Mean
90%
Confidence
Bound
Mean
90%
Confidence
Bound
Data
Set
Total
Annualized
Costs
at
3%
Data
Set
Capital
and
One­
Time
(
Nominal
at
Full
Implementation)

Mean
90%
Confidence
Bound
Mean
90%
Confidence
Bound
Operations
and
Maintenance
(
Nominal
at
Full
Implementation)
Exhibit
ES.
6
Summary
of
System
Costs
($
Millions,
2000$)

Source:
Exhibit
8.11.

EPA
assumes
that
systems
will
generally
pass
the
costs
of
a
new
regulation
on
to
their
customers
in
the
form
of
rate
increases.
Household
costs,
which
are
in
units
of
$
per
household
per
year,
are
estimated
to
provide
a
measure
of
the
increase
in
water
bills
that
could
be
expected
to
result
from
the
LT2ESWTR.
Exhibit
ES.
7
summarizes
household
costs
for
those
systems
predicted
to
require
additional
treatment.
Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
11
System
Type/
Size
Households
Mean
Median
90th
Percentile
95th
Percentile
Percent
of
Systems
with
Household
Cost
Increase
<
$
12
Percent
of
Systems
with
Household
Cost
Increase
<
$
120
All
CWS
65,816,979
$
1.68
$
0.13
$
4.06
$
7.57
98.37%
99.99%
CWS
 
£
10,000
3,318,012
$
4.61
$
1.34
$
13.04
$
14.92
87.88%
99.88%

All
CWS
65,816,979
$
1.07
$
0.03
$
3.24
$
5.43
99.31%
100.00%
CWS
 
£
10,000
3,318,012
$
2.68
$
0.80
$
6.10
$
9.39
95.71%
99.95%

All
CWS
65,816,979
$
1.28
$
0.03
$
3.48
$
6.47
99.07%
100.00%
CWS
 
£
10,000
3,318,012
$
3.27
$
0.80
$
6.62
$
13.04
93.90%
99.93%
All
Systems
­
ICR
All
Systems
­
ICRSSL
All
Systems
­
ICRSSM
Exhibit
ES.
7
Summary
of
Annual
Household
Cost
Increases[
1]

($
per
Year,
2000$)

[
1]
Annualized
at
discount
rates
varied
by
system
size
and
ownership
(
see
Appendix
J,
Exhibit
J.
2).
Source:
Exhibit
6.14.

ES.
5
National
Net
Benefits
and
Summary
of
Comparison
of
Alternatives
The
national
net
benefits
for
each
of
the
regulatory
alternatives
are
shown
in
Exhibit
ES.
8.
The
annualized
net
national
benefits
for
the
Preferred
Alternative
range
from
$
181
million
to
$
1.34
billion,
depending
on
occurrence,
cost
of
illness
assumptions,
and
discounts
rates
used.

From
Exhibit
ES.
8,
several
important
economic
questions
can
be
answered.
First,
the
Preferred
Alternative
(
A3)
has
positive
net
benefits,
a
key
threshold
test
of
the
reasonableness
of
a
regulation.
This
is
also
true
for
Alternatives
A2
and
A4
under
all
combinations
of
occurrence,
cost
of
illness,
and
discount
rates,
and
true
for
Alternative
A1
for
most
of
the
combinations.
Second,
this
exhibit
shows
that
the
Preferred
Alternative
(
A3)
is
the
superior
alternative
(
shows
the
highest
net
benefits)
under
more
of
the
combinations
than
any
other
alternative.
Alternative
A4
has
the
highest
net
benefits
under
most
conditions
with
the
Traditional
cost
of
illness.
Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
12
3%,
25
Years
7%,
25
Years
A1
1,121
$
873
$
A2
1,327
$
1,098
$
A3
­
Preferred
1,335
$
1,109
$
A4
1,290
$
1,083
$

A1
96
$
1
$
A2
298
$
230
$
A3
­
Preferred
300
$
237
$
A4
291
$
238
$

A1
435
$
289
$
A2
630
$
509
$
A3
­
Preferred
629
$
514
$
A4
592
$
493
$
ICRSSL
Data
Set
Rule
Alternative
Annualized
Value
ICR
ICRSSM
3%,
25
Years
7%,
25
Years
A1
628
$
457
$
A2
843
$
688
$
A3
­
Preferred
856
$
705
$
A4
848
$
710
$

A1
­
56
$
­
128
$
A2
168
$
120
$
A3
­
Preferred
180
$
135
$
A4
188
$
151
$

A1
170
$
66
$
A2
386
$
303
$
A3
­
Preferred
395
$
317
$
A4
388
$
321
$
Data
Set
Rule
Alternative
Annualized
Value
ICRSSM
ICR
ICRSSL
Exhibit
ES.
8a
Comparison
of
Mean
Net
Benefits
for
All
Regulatory
Alternatives
 
Enhanced
Cost
of
Illness
($
Millions,
2000$)

Exhibit
ES.
8b
Comparison
of
Mean
Net
Benefits
for
All
Regulatory
Alternatives
 
Traditional
Cost
of
Illness
($
Millions,
2000$)

Notes:
[
1]
The
traditional
COI
only
includes
valuation
for
medical
costs
and
lost
work
time
(
including
some
portion
of
unpaid
household
production).
The
enhanced
COI
also
factors
in
valuations
for
lost
personal
time
(
non­
work
time)
such
as
child
care
and
homemaking
(
to
the
extent
not
covered
by
the
traditional
COI),
time
with
family,
and
recreation,
and
lost
productivity
at
work
on
days
when
workers
are
ill
but
go
to
work
anyway.
Source:
Exhibits
8.12a
and
8.12b.
Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
13
Another
key
economic
test
is
whether
the
proposed
rule
is
cost
effective.
Exhibits
ES.
9a
and
ES.
9b
each
show
four
graphs
that
provide
information
on
this
test.
The
mean
benefit
for
each
alternative
is
plotted
against
the
range
of
costs
that
correspond
to
the
mean
benefit
level.
The
graphs
show
all
alternatives
are
cost
effective
and
no
alternative
provides
additional
benefits
at
the
same
or
lower
cost.
Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
14
$
1,125
$
1,150
$
1,175
$
1,200
$
1,225
$
1,250
$
1,275
Alt
4
Alt.
3
Alt.
2
Alt.
1
$
0
$
50
$
100
$
150
$
200
$
250
$
300
$
350
$
400
$
450
$
500
$
300
$
325
$
350
$
375
$
400
$
425
$
450
$
475
Alt.
4
Alt.
3
Alt.
2
Alt.
1
$
250
$
270
$
290
$
310
$
330
$
350
$
370
$
390
$
410
Alt.
4
Alt.
3
Alt.
2
Alt.
1
3
Percent
Discount
Rate
7
Percent
Discount
Rate
ICR
Data
Set
Costs
($
Millions)

ICRSSL
Data
Set
Costs
($
Millions)

Mean
of
Benefits
($
Millions)
Mean
of
Benefits
($
Millions)
$
0
$
50
$
100
$
150
$
200
$
250
$
300
$
350
$
400
$
450
$
500
$
1,325
$
1,350
$
1,375
$
1,400
$
1,425
$
1,450
$
1,475
$
1,500
Alt.
4
Alt.
3
Alt.
2
Alt.
1
$
0
$
50
$
100
$
150
$
200
$
250
$
300
$
350
$
400
$
450
$
600
$
650
$
700
$
750
$
800
$
850
Alt.
4
Alt.
3
Alt.
2
Alt.
1
$
520
$
560
$
600
$
640
$
680
$
720
Alt.
4
Alt.
3
Alt.
2
Alt.
1
ICRSSM
Data
Set
Costs
($
Millions)
Exhibit
ES.
9a
Comparison
of
Mean
 
Enhanced
Cost
of
Illness
[
1]
Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
15
$
750
$
775
$
800
$
825
$
850
Alt
4
Alt.
3
Alt.
2
Alt.
1
$
0
$
50
$
100
$
150
$
200
$
250
$
300
$
350
$
400
$
450
$
500
$
200
$
225
$
250
$
275
$
300
$
325
Alt.
4
Alt.
3
Alt.
2
Alt.
1
$
180
$
200
$
220
$
240
$
260
$
280
Alt.
4
Alt.
3
Alt.
2
Alt.
1
3
Percent
Discount
Rate
7
Percent
Discount
Rate
ICR
Data
Set
Costs
($
Millions)

ICRSSL
Data
Set
Costs
($
Millions)

Mean
of
Benefits
($
Millions)
Mean
of
Benefits
($
Millions)
$
0
$
50
$
100
$
150
$
200
$
250
$
300
$
350
$
400
$
450
$
500
$
900
$
925
$
950
$
975
$
1,000
Alt.
4
Alt.
3
Alt.
2
Alt.
1
ICRSSM
Data
Set
Costs
($
Millions)

$
0
$
50
$
100
$
150
$
200
$
250
$
300
$
350
$
400
$
450
$
400
$
425
$
450
$
475
$
500
$
525
$
550
Alt.
4
Alt.
3
Alt.
2
Alt.
1
$
520
$
540
$
560
$
580
$
600
$
620
$
640
$
660
$
680
$
700
Alt.
4
Alt.
3
Alt.
2
Alt.
1
Exhibit
ES.
9b
Comparison
of
Mean
 
Traditional
Cost
of
Illness
[
1]

Notes:
[
1]
The
traditional
COI
only
includes
valuation
for
medical
costs
and
lost
work
time
(
including
some
portion
of
unpaid
household
production).
The
enhanced
COI
also
factors
in
valuations
for
lost
personal
time
(
non­
work
time)
such
as
child
care
and
homemaking
(
to
the
extent
not
covered
by
the
traditional
COI),
time
with
family,
and
recreation,
and
lost
productivity
at
work
on
days
when
workers
are
ill
but
go
to
work
anyway.
Source:
Exhibit
8.7.
Economic
Analysis
for
LT2ESWTR
Proposal
June
2003
ES­
16
The
Preferred
Alternative
(
A3)
was
also
evaluated
against
the
other
alternatives
with
respect
to
key
uncertainty
variables.
EPA
conducted
a
series
of
sensitivity
analyses
to
examine
the
effects
of
exaggerating
one
uncertain
variable
while
holding
all
others
constant.
These
analyses
tested
the
following
uncertainty
variables:
the
assumptions
regarding
loss
of
productivity
and
the
value
of
nonwork
time
in
computing
the
cost
of
illness,
the
value
of
AIDS­
related
mortality
rate,
and
the
overall
value
of
benefits
(
presented
in
Appendix
P,
Appendix
R,
and
Chapter
8,
respectively).
In
addition
to
the
sensitivity
analyses,
the
main
body
of
the
EA
carries
two
possible
values
for
the
cost
of
illness.
Further,
uncertainty
in
the
occurrence
of
Cryptosporidium
in
source
water
is
addressed
by
carrying
out
separate
analyses
throughout
the
EA
for
the
three
possible
occurrence
distributions
as
well
as
sensitivity
analyses
on
the
occurrence
distributions.
The
results
of
all
tests
show
that
benefits
still
exceed
costs
for
the
Preferred
Alternative,
it
remains
the
favored
alternative
in
the
majority
of
conditions
analyzed,
and
no
other
alternative
performs
as
well
across
the
range
of
possible
occurrence
and
values
for
benefits.

Alternative
A3
was
recommended
by
the
Stage
2
Microbial
Disinfectants
and
Disinfection
Byproducts
(
Stage
2
M­
DBP)
Advisory
Committee.
Based
on
the
recommendation,
and
supported
by
EPA's
evaluations
of
benefits
and
costs,
EPA
selected
Alternative
A3
as
the
proposed
rule.
1
IESWTR
(
63
FR
69477
December
1998),
LT1ESWTR
(
67
FR
1811
January
2002)

Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
1­
1
1.
Introduction
This
document
presents
an
analysis
of
the
costs
and
benefits
of
the
proposed
Long
Term
2
Enhanced
Surface
Water
Treatment
Rule
(
LT2ESWTR).
The
analysis
is
performed
and
presented
in
compliance
with
Executive
Order
12866,
Regulatory
Planning
and
Review
(
58
Federal
Register
(
FR)
51735),
which
requires
that
the
U.
S.
Environmental
Protection
Agency
(
EPA)
estimate
the
economic
impact
of
rules
costing
over
$
100
million
annually
and
submit
the
analysis
in
conjunction
with
publishing
the
rule.

This
chapter
provides
a
summary
of
the
LT2ESWTR
in
section
1.1
and
describes
the
organization
of
this
Economic
Analysis
(
EA)
in
section
1.2.

1.1
Summary
The
LT2ESWTR
builds
on
the
Interim
Enhanced
Surface
Water
Treatment
Rule
(
IESWTR)
and
the
Long
Term
1
Enhanced
Surface
Water
Treatment
Rule
(
LT1ESWTR)
by
improving
control
of
microbial
pathogens,
specifically
the
contaminant
Cryptosporidium1.
The
LT2ESWTR
also
addresses
the
risk­
risk
tradeoffs
posed
by
the
simultaneous
control
of
microbial
pathogens
and
disinfection
byproducts
(
DBPs).
The
disinfectants
commonly
used
to
kill
microorganisms
react
with
naturally
occurring
organic
and
inorganic
matter
in
source
water,
forming
DBPs
that
are
known
to
have
adverse
health
effects
(
including
cancer,
developmental,
and
reproductive
effects).
In
order
to
balance
the
risks
posed
by
DBPs
and
microbial
pathogens
and
to
facilitate
compliance
decisions
by
water
systems,
the
LT2ESWTR
will
be
promulgated
concurrently
with
the
Stage
2
Disinfection
Byproducts
Rule
(
DBPR).
The
LT2ESWTR
applies
to
all
community
water
systems
(
CWSs)
and
noncommunity
water
systems
(
NCWSs)
that
use
surface
water
or
ground
water
under
the
direct
influence
of
surface
water
(
GWUDI)
as
a
source.

The
intent
of
the
LT2ESWTR
is
to
supplement
existing
microbial
treatment
requirements
for
systems
where
additional
public
health
protection
is
needed.
The
rule
will
require
filtered
systems
to
monitor
their
source
water
for
Cryptosporidium.
Based
on
monitoring
results,
filtered
systems
must
meet
one
of
four
levels
of
treatment
for
Cryptosporidium
(
with
the
first
level
requiring
no
additional
treatment).
All
unfiltered
systems,
which
are
not
currently
required
to
provide
any
treatment
for
Cryptosporidium,
must
achieve
2
or
3
log
Cryptosporidium
inactivation,
depending
on
their
source
water
Cryptosporidium
levels.
The
rule
also
requires
systems
with
uncovered
finished
water
reservoirs
to
either
cover
the
reservoirs
or
provide
additional
treatment
to
reservoir
effluent.
Rule
provisions
are
described
in
detail
below.

1.1.1
Monitoring
and
Treatment
Requirements
for
Filtered
Systems
Systems
first
monitor
source
water
Cryptosporidium
concentrations
and
based
on
those
results,
are
assigned
to
different
treatment
"
bins."
Within
each
bin,
systems
will
choose
technologies
from
a
"
toolbox"
of
options
for
ensuring
Cryptosporidium
removal
or
inactivation
from
treated
water.
The
bins
2
Systems
must
meet
all
requirements
of
the
analytical
methods
for
Cryptosporidium,
which
include
analysis
of
two
matrix
spiked
samples.

Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
1­
2
for
source
waters
with
higher
concentrations
of
Cryptosporidium
involve
treatment
options
providing
higher
levels
of
inactivation
and/
or
removal.

Initial
Monitoring
for
Bin
Classification
 
Systems
Serving
at
Least
10,000
People
Medium
and
large
filtered
systems
(
those
serving
at
least
10,000
people)
will
be
required
to
monitor
their
raw
water
sources
for
Cryptosporidium
at
each
plant
at
least
once
per
month
for
a
minimum
of
2
years.
Bin
classification
will
be
based
on
one
of
the
following:

°
The
highest
12­
month
running
annual
average
Cryptosporidium
concentration
(
in
oocysts/
liter)
if
samples
are
taken
monthly
(
24
samples
total),
or
°
The
2­
year
mean
Cryptosporidium
concentration.
The
facility
may
conduct
monitoring
twice
per
month
for
24
months
(
48
samples
total)
or
perform
additional
sampling
and
include
these
results
in
the
calculation
of
the
mean,
but
the
additional
samples
must
be
evenly
distributed
over
the
2­
year
monitoring
period.

Cryptosporidium
analysis
must
be
conducted
in
accordance
with
EPA
Method
1622/
23
using
a
sample
volume
of
at
least
10
liters.
2
Samples
must
also
be
analyzed
for
E.
coli
and
turbidity.
The
E.
coli
and
turbidity
data
will
be
used
by
EPA
and
stakeholders
to
evaluate
methods
for
predicting
Cryptosporidium
occurrence.

Systems
with
at
least
2
years
of
historical
Cryptosporidium
data
that
are
equivalent
in
sample
number,
frequency,
and
quality
to
data
required
under
the
LT2ESWTR
may
use
these
data
to
determine
bin
classification,
in
lieu
of
additional
Cryptosporidium
compliance
monitoring,
if
EPA
approves
the
use
of
these
data.

Monitoring
for
medium
and
large
systems
starts
no
later
than
6
months
after
promulgation
of
the
LT2ESWTR.
Systems
will
submit
monitoring
data
to
EPA
on
an
ongoing
basis,
(
i.
e.,
as
data
are
generated,
they
will
be
entered
into
an
EPA
database).
At
the
end
of
the
2­
year
monitoring
period,
EPA
will
give
the
results
to
the
States/
Primacy
Agencies,
who
will
then
work
with
their
systems
to
determine
appropriate
compliance
steps.

Initial
Monitoring
for
Bin
Classification
 
Systems
Serving
Fewer
than
10,000
People
Source
water
monitoring
for
small
systems
(
those
serving
fewer
than
10,000
people)
will
begin
2
years
after
the
medium
and
large
systems
initiate
source
water
Cryptosporidium
monitoring.
The
required
monitoring
is
on
a
delayed
schedule
so
EPA
can
incorporate
information
on
E.
coli
and
turbidity
collected
by
the
medium
and
large
systems
into
monitoring
requirements
(
EPA
will
examine
these
data
and
their
use
as
indicators
of
Cryptosporidium).
In
the
absence
of
a
new
indicator,
small
systems
will
conduct
1
year
of
biweekly
E.
coli
source
water
monitoring
and
will
be
required
to
conduct
Cryptosporidium
monitoring
only
if
E.
coli
concentrations
exceed
the
following
levels:

°
An
annual
mean
concentration
greater
than
10
E.
coli
per
100
mL
for
lake
and
reservoir
source
waters
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
1­
3
°
An
annual
mean
concentration
greater
than
50
E.
coli
per
100
mL
for
flowing
stream
source
waters
Systems
that
do
not
exceed
these
levels
are
assumed
to
have
a
Cryptosporidium
concentration
of
less
than
0.075
oocysts/
L
and
are
placed
in
Bin
1
(
see
Exhibit
1.1).
Small
systems
that
exceed
the
E.
coli
levels
mentioned
above
would
be
required
to
conduct
semimonthly
Cryptosporidium
monitoring
for
a
1­
year
period,
beginning
6
months
after
the
conclusion
of
E.
coli
monitoring.
Bin
classification
for
small
systems
conducting
Cryptosporidium
monitoring
is
based
on
the
arithmetic
mean
concentration
of
the
24
Cryptosporidium
samples.

All
filtered
systems
that
provide
5.5
log
of
treatment
for
Cryptosporidium
by
the
date
they
are
required
to
start
source
water
monitoring
are
exempt
from
monitoring
and
subsequent
bin
classification.
To
meet
the
requirement
for
5.5
log
of
treatment,
systems
using
conventional
treatment
would
be
required
to
provide
2.5
log
of
additional
treatment,
and
systems
using
direct
filtration
would
be
required
to
provide
3
log
of
additional
treatment.

Bins
and
Treatment
Requirements
 
All
System
Sizes
Exhibit
1.1
presents
the
bins
for
filtered
systems
according
to
treatment
plant
type.
Systems
must
meet
Cryptosporidium
treatment
requirements
by
using
one
of
the
treatment
options
in
the
"
microbial
toolbox"
or
by
demonstrating
performance
equivalent
to
or
exceeding
the
required
treatment.
Systems
have
3
years
after
initial
bin
classification
to
meet
the
treatment
requirements
associated
with
the
bin.
States/
Primacy
Agencies
may
grant
systems
a
2­
year
extension
to
comply
if
capital
investments
are
necessary.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
1­
4
Exhibit
1.1
Filtered
Systems
Bin
Classification
and
Treatment
Requirements
If
your
source
water
Cryptosporidium
concentration
(
oocysts/
L)
is...
Your
bin
classification
is...
And
if
you
use
the
following
filtration
treatment
in
full
compliance
with
existing
regulations,
then
your
additional
treatment
requirements
are...

Conventional
Filtration
Direct
Filtration
Slow
Sand
or
Diatomaceous
Earth
Filtration
Alternative
Filtration
Technologies
<
0.075
1
No
additional
treatment
No
additional
treatment
No
additional
treatment
No
additional
treatment
>
0.075
and
<
1.0
21
1
log
treatment
1.5
log
treatment
1
log
treatment
As
determined
by
the
State2
>
1.0
and
<
3.0
33
2
log
treatment
2.5
log
treatment
2
log
treatment
As
determined
by
the
State4
>
3.0
43
2.5
log
treatment
3
log
treatment
2.5
log
treatment
As
determined
by
the
State5
[
1]
Systems
may
use
any
technology
or
combination
of
technologies
from
the
microbial
toolbox.
[
2]
Total
Cryptosporidium
treatment
must
be
at
least
4.0
log.
3Systems
must
achieve
at
least
1
log
of
the
required
treatment
using
ozone,
chlorine
dioxide,
ultraviolet
light
(
UV),
membranes,
bag/
cartridge
filters,
or
bank
filtration.
[
4]
Total
Cryptosporidium
treatment
must
be
at
least
5.0
log.
[
5]
Total
Cryptosporidium
treatment
must
be
at
least
5.5
log.

The
total
Cryptosporidium
treatment
required
for
Bins
2,
3,
and
4
is
4.0
log,
5.0
log,
and
5.5
log,
respectively.
The
additional
treatment
requirements
in
Exhibit
1.1
are
based
on
a
determination
that
conventional,
slow
sand,
and
diatomaceous
earth
filtration
plants
in
compliance
with
the
IESWTR
or
LT1ESWTR
achieve
an
average
of
3
log
removal
of
Cryptosporidium
(
the
2
log
credit
for
Cryptosporidium
under
the
IESWTR
and
LT1ESWTR
is
based
on
the
minimum
removal
expected
with
these
types
of
filtration).
Therefore,
conventional,
slow
sand,
and
diatomaceous
earth
filtration
plants
will
require
an
additional
1.0
to
2.5
log
additional
treatment
to
meet
the
total
removal
requirement,
depending
on
the
bin
they
are
placed
in.

EPA
has
determined
that
direct
filtration
plants
achieve
an
average
2.5
log
removal
of
Cryptosporidium
(
their
removal
is
less
than
in
conventional
filtration
because
they
lack
a
sedimentation
process).
Consequently,
under
the
LT2ESWTR,
direct
filtration
plants
in
Bins
2
 
4
must
provide
0.5
log
more
in
additional
treatment
than
conventional
plants
to
meet
the
total
Cryptosporidium
removal
requirement.

Microbial
Toolbox
for
Meeting
Additional
Treatment
Requirements
To
meet
the
Cryptosporidium
treatment
requirements
for
the
bin
in
which
they
are
classified,
filtered
systems
can
select
from
a
"
toolbox"
of
treatment
or
management
options.
The
technologies
and
management
strategies
in
the
microbial
toolbox,
along
with
their
log
treatment
credits,
are
presented
in
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
1­
5
Exhibit
1.2.
Systems
do
not
get
the
log
credit
automatically
when
they
install
these
technologies;
they
must
show
they
are
meeting
certain
operational
or
other
criteria
specific
to
the
technology.
Log
credit
under
existing
rules
(
e.
g.,
the
IESWTR
and
LT1ESWTR)
works
much
the
same
way.
Systems
currently
using
ozone,
chlorine
dioxide,
ultraviolet
light
(
UV),
or
membranes
in
addition
to
conventional
treatment
may
receive
credit
for
those
technologies
toward
meeting
bin
requirements
if
they
meet
the
LT2ESWTR
criteria
for
the
chosen
technology.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
1­
6
Exhibit
1.2
Microbial
Toolbox
Components
for
the
LT2ESWTR
(
To
be
used
in
addition
to
existing
treatment)

Toolbox
Option
Log
Credit
Source
Toolbox
Components
Watershed
control
program
0.5
Alternative
source/
intake
management
None,
but
conduct
source
water
monitoring
concurrently
at
both
sources
or
under
both
intake
management
plans
and
determine
bin
based
on
the
lower
mean
concentration
Pre­
Filtration
Toolbox
Components
Presedimentation
basin
with
coagulation
0.5
Two­
stage
lime
softening
0.5
Bank
filtration
0.5
or
1.0,
depending
on
setback
Treatment
Performance
Toolbox
Components
Combined
filter
performance
0.5
Individual
filter
performance
1.0
Demonstration
of
performance
State
approved1
Additional
Filtration
Toolbox
Components
Bag
filters
1.0
Cartridge
filters
2.0
Membrane
filtration
As
demonstrated2
Second
stage
filtration
0.5
Slow
sand
filters
2.5
Inactivation
Toolbox
Components
Chlorine
dioxide
As
demonstrated3
Ozone
As
demonstrated3
UV
As
demonstrated2
[
1]
State
must
approve
the
method
used
to
demonstrate
performance
and
must
approve
the
log
credit
claimed
by
the
system.
[
2]
Credit
for
membrane
filtration
and
UV
is
based
on
the
results
of
challenge
testing
and
reactor
validation
testing,
respectively.
[
3]
Credit
for
chlorine
dioxide
and
ozone
is
based
on
CT
values
achieved
(
CT
is
the
product
of
disinfectant
concentration
and
contact
time).
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
1­
7
Reassessment
and
Future
Monitoring
Six
years
after
initial
bin
classification,
systems
will
be
required
to
conduct
a
second
round
of
monitoring
to
reassess
source
water
conditions
for
bin
assignments.
Two
years
before
this
reassessment
(
4
years
after
initial
binning),
EPA
plans
to
initiate
a
stakeholder
process
to
review
available
analytical
methods
for
detecting
Cryptosporidium.
If
there
are
new,
improved
methods,
EPA,
with
stakeholder
input,
will
determine
the
appropriate
analytical
method,
monitoring
frequency,
and
monitoring
locations
for
the
second
round
of
national
assessment
monitoring.
In
the
absence
of
an
improved
Cryptosporidium
detection
method,
monitoring
will
follow
EPA
Method
1622/
23.
Systems
that
provide
a
total
of
5.5
log
treatment
for
Cryptosporidium
are
not
subject
to
future
monitoring.

In
addition
to
the
reassessment
and
re­
binning
described
above,
the
State/
Primacy
Agency
will
assess
any
significant
changes
in
the
watershed
and
source
water
as
part
of
the
sanitary
survey
process.
They
will
then
determine
what
follow­
up
action
is
appropriate
in
response
to
source
water
changes,
which
could
include
actions
from
the
microbial
toolbox.

1.1.2
Monitoring
and
Treatment
Requirements
for
Unfiltered
Systems
Unfiltered
systems
that
already
have
3
log
Cryptosporidium
treatment
in
place
prior
to
the
date
they
would
have
to
start
monitoring
are
exempt
from
monitoring
and
additional
Cryptosporidium
inactivation
requirements.
Otherwise,
large
unfiltered
systems
must
monitor
Cryptosporidium
in
their
source
water
monthly
for
at
least
2
years,
and
small
unfiltered
systems
must
monitor
semimonthly
for
12
months.
All
unfiltered
systems
must
determine
their
treatment
requirements
based
on
the
arithmetic
mean
Cryptosporidium
concentration.
If
their
average
Cryptosporidium
concentration
is
less
than
or
equal
to
0.01
oocyst/
L,
systems
must
provide
2
log
Cryptosporidium
inactivation.
If
their
average
Cryptosporidium
concentration
is
greater
than
0.01
oocysts/
L,
they
must
provide
3
log
Cryptosporidium
inactivation.

Monitoring
for
unfiltered
systems
will
be
based
on
the
same
schedule
as
monitoring
for
filtered
systems,
although
unfiltered
systems
are
not
required
to
monitor
E.
coli
or
turbidity.
As
with
the
filtered
systems,
unfiltered
systems
must
conduct
a
second
round
of
Cryptosporidium
monitoring
six
years
after
the
initial
bin
assignment.

In
addition
to
the
new
Cryptosporidium
inactivation
requirements,
the
LT2ESWTR
will
require
unfiltered
systems
to
continue
to
meet
the
filtration
avoidance
criteria
under
the
1989
SWTR
and
to
continue
provide
inactivation
for
Giardia
and
viruses.
The
overall
inactivation
requirements
(
i.
e.,
4
log
virus,
3
log
Giardia,
and
2
or
3
log
Cryptosporidium)
must
be
met
using
a
minimum
of
two
disinfectants.
Additionally,
each
of
two
disinfectants
must
meet
the
total
inactivation
for
one
of
the
three
pathogens.
For
example,
a
system
could
use
UV
to
inactivate
2
log
of
Cryptosporidium
and
Giardia
and
use
chlorine
to
inactivate
4
log
of
viruses
and
1
log
Giardia.

1.1.3
Requirements
for
Existing
Uncovered
Finished
Water
Reservoirs
The
LT2ESWTR
builds
on
the
IESWTR
and
LT1ESWTR,
which
require
covers
only
for
new
finished
water
reservoirs.
The
LT2ESWTR
will
establish
requirements
for
all
systems
with
existing
uncovered
finished
water
reservoirs.
Systems
must
either
cover
the
reservoir
or
treat
reservoir
discharge
to
the
distribution
system
to
achieve
4
log
virus
inactivation,
unless
the
State/
Primacy
Agency
determines
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
1­
8
that
existing
risk
mitigation
is
adequate.
Risk
mitigation
plans
must
address
physical
access,
surface
water
run­
off,
animal
and
bird
waste,
and
continuing
water
quality
assessment
at
a
minimum.

1.1.4
Disinfection
Profiling
and
Benchmarking
Requirements
The
LT2ESWTR
includes
a
disinfection
profile
and
benchmark
requirement
to
ensure
that
any
significant
change
in
disinfection,
whether
for
byproduct
control
under
the
Stage
2
DBPR,
improved
Cryptosporidium
control
under
the
LT2ESWTR,
or
both,
does
not
significantly
compromise
existing
Giardia
and
virus
protection.
A
disinfection
profile
is
a
graphical
representation
of
a
system's
level
of
Giardia
and
viral
inactivation
measured
during
the
course
of
1
or
more
year(
s).
A
benchmark
is
the
lowest
monthly
average
of
microbial
inactivation
during
the
disinfection
profile
period.

The
profiling
requirements
under
the
LT2ESWTR
are
similar
to
those
promulgated
under
the
IESWTR
and
LT1ESWTR
and
are
applicable
to:
1)
systems
required
to
conduct
Cryptosporidium
source
water
monitoring
and
2)
small
surface
water
systems
that
do
not
have
to
conduct
Cryptosporidium
source
water
monitoring
but
that
have
Stage
1
DBPR
TTHM
annual
average
results
of
at
least
56
µ
g/
L
or
HAA5
annual
average
results
of
at
least
42
µ
g/
L.
The
LT2ESWTR
requires
these
systems
to
prepare
a
disinfection
profile
that
characterizes
current
levels
of
Giardia
lamblia
and
virus
inactivation
throughout
the
plant
over
the
course
of
one
year.
The
profile
may
be
developed
using
equivalent
historical
data.
If
a
system
subsequently
proposes
to
make
a
significant
change
to
its
disinfection
practice,
then
the
LT2ESWTR
requires
the
system
to
calculate
a
disinfection
benchmark
and
consult
with
the
state
regarding
how
the
proposed
change
will
affect
that
benchmark.

1.1.5
Implementation
Timeline
Exhibit
1.3
shows
the
timeline
of
LT2ESWTR
activities.
The
schedule
for
monitoring
and
compliance
with
treatment
requirements
differs
between
large
and
small
systems.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
1­
9
States
Crypto
Monitoring
Apply
for
Primacy
Comply
with
Stage
2A
Year
1
Year
2
Year
3
Year
4
Year
5
Year
6
Year
7
Year
8
Year
9
Year
10
Year
11
Year
1
Year
2
Year
3
Year
4
Year
5
Year
6
Year
7
Year
8
Year
9
Year
10
Year
11
Large
and
Medium
Filtered
and
Unfiltered
Systems
(
Serving
>
10,000
People)

Bin
Determination
Treatment
Installation
Small
Filtered
Systems
(
Serving
<
10,000
People)

E.
Coli
monitoring
Crypto
Monitoring
Bin
Determination
Treatment
Installation
Bin
Determination
Crypto
Compliance
Crypto
Compliance
Crypto
Compliance
Small
Unfiltered
Systems
(
Serving
<
10,000
People)
Crypto
Monitoring
Final
Stage
2
DBPR
&
LT2ESWTR
Rules
Promulgated
Treatment
Installation
(
possible
2­
year
extension)

(
possible
2­
year
extension)

(
possible
2­
year
extension)
Exhibit
1.3
Implementation
Time
Line
for
LT2ESWTR
1.2
Document
Organization
This
EA
is
organized
into
the
following
chapters:

°
Chapter
2
identifies
public
health
concerns
addressed
by
the
rule
and
provides
a
20­
year
regulatory
history
that
includes
a
description
of
relevant
National
Primary
Drinking
Water
Regulations
(
NPDWRs).
It
also
explains
the
statutory
authority
for
promulgating
the
LT2ESWTR
and
economic
rationale
for
choosing
a
regulatory
approach.

°
Chapter
3
describes
the
regulatory
alternatives
considered
for
the
LT2ESWTR
and
the
process
for
developing
them.

°
Chapter
4
characterizes
the
baseline
conditions
that
are
expected
to
exist
(
including
system
inventory,
treatment,
and
water
quality
data)
before
systems
complete
the
monitoring
and
begin
making
treatment
changes
to
meet
the
LT2ESWTR
requirements.
Because
of
the
3
The
compliance
deadline
for
the
IESWTR
and
Stage
1
DBPR
occurred
recently
(
January
2002)
for
large
and
medium
surface
water
systems.

Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
1­
10
timing
of
the
IESWTR,
LT1ESWTR,
and
Stage
1
DBPR3,
EPA
had
to
predict
the
changes
in
treatment
and
water
quality
made
by
systems
as
a
result
of
these
rules
to
characterize
Pre­
LT2ESWTR
baseline
conditions.

°
Chapter
5
reviews
available
toxicological
and
epidemiological
data
related
to
Cryptosporidium
and
presents
the
public
health
and
economic
benefits
(
both
quantifiable
and
unquantifiable)
of
this
rule.
It
compares
the
benefits
of
the
four
regulatory
alternatives,
and
presents
several
sensitivity
analyses.

°
Chapter
6
presents
an
estimate
of
the
costs
of
implementing
the
rule
to
the
drinking
water
industry,
households,
and
States/
Primacy
Agencies.
It
also
compares
the
costs
of
the
four
regulatory
alternatives.

°
Chapter
7
discusses
distributional
analyses
performed
to
evaluate
the
effects
of
the
rule
on
different
segments
of
the
population.
It
also
considers
various
executive
orders
and
requirements,
including
the
Regulatory
Flexibility
Act
(
RFA)
and
Unfunded
Mandates
Reform
Act
(
UMRA).

°
Chapter
8
summarizes
and
compares
the
rule's
benefit
and
cost
estimates.
(
Note:
It
does
several
things,
not
limited
to
whether
$
B
>
$
C
­
and
not
focused
on
that
comparison
specifically.)
The
results
for
the
Preferred
Regulatory
Alternative
are
discussed
and
compared
to
the
other
regulatory
alternatives
considered.
Economic
Analysis
for
the
LT2ESWTR
Proposal
2­
1
June
2003
2.
Statement
of
Need
for
the
Proposal
2.1
Introduction
This
chapter
presents
the
need
for
the
LT2ESWTR
by
identifying
the
public
health
concerns
that
this
rule
will
address.
Included
is
a
discussion
of
related
regulations
that
shows
that
these
public
health
concerns
are
not
adequately
addressed
by
other
rules.
Lastly,
this
chapter
discusses
the
economic
rationale
for
the
rule.
The
remaining
sections
are
organized
as
follows:

2.2
Description
of
the
Issue
2.3
Risk
Balancing
2.4
Public
Health
Concerns
to
be
Addressed
2.4.1
Cryptosporidium
2.4.2
Uncovered
Finished
Water
Reservoirs
2.5
Regulatory
History
2.5.1
Statutory
Authority
for
Promulgating
the
Rule
2.5.2
1979
Total
Trihalomethane
Rule
2.5.3
1989
Total
Coliform
Rule
2.5.4
1989
Surface
Water
Treatment
Rule
2.5.5
1996
Information
Collection
Rule
2.5.6
1998
Interim
Enhanced
Surface
Water
Treatment
Rule
2.5.7
1998
Stage
1
Disinfectants
and
Disinfection
Byproducts
Rule
2.5.8
2000
Proposed
Ground
Water
Rule
2.5.9
2001
Filter
Backwash
Recycling
Rule
2.5.10
2002
Long
Term
1
Enhanced
Surface
Water
Treatment
Rule
2.5.11
2003
Proposed
Stage
2
Disinfectants
and
Disinfection
Byproducts
Rule
2.6
Economic
Rationale
for
Regulation
2.2
Description
of
the
Issue
Over
14,000
public
water
systems
(
PWSs),
serving
approximately
180
million
people
in
the
United
States
and
its
territories,
use
surface
water
or
ground
water
under
the
direct
influence
of
surface
water
(
GWUDI)
as
their
source.
These
sources
carry
microbial
contaminants,
some
of
which
pose
significant
risks
to
public
health.
The
U.
S.
Environmental
Protection
Agency
(
EPA
or
the
Agency)
is
particularly
concerned
about
Cryptosporidium
because
it
is
resistant
to
many
commonly
used
drinking
water
disinfectants,
such
as
chlorine,
and
it
poses
significant
health
risks,
including
death.
Moreover,
there
is
no
effective
drug
available
to
cure
cryptosporidiosis,
the
health
condition
caused
by
Cryptosporidium
infection
(
Framm
and
Soave
1997).
The
primary
issue
of
concern
addressed
by
this
rule
is
the
risk
to
public
health
in
water
supplies
with
inadequate
Cryptosporidium
treatment.

The
1989
SWTR
requires
most
surface
water
and
GWUDI
systems
to
physically
remove
microbial
contaminants
through
filtration.
(
Exemptions
to
this
filtration
requirement
are
granted
to
systems
that
meet
specified
avoidance
criteria.)
Types
of
filtration
systems
include
the
following:
Economic
Analysis
for
the
LT2ESWTR
Proposal
2­
2
June
2003
°
Conventional
treatment
 
coagulation,
flocculation,
and
sedimentation
of
particles,
followed
by
granular
media
filtration.

°
Direct
filtration
 
coagulation
and
flocculation
followed
by
rapid
sand
filtration,
but
no
sedimentation
basin.
This
type
of
filtration
system
is
designed
for
low­
turbidity
waters.

°
Slow
sand
and
diatomaceous
earth
filtration
 
filters
that
work
at
very
low
flow
rates
without
the
use
of
a
coagulant
in
pretreatment.

°
Alternative
filtration
 
other
technologies,
including
membranes
and
bag
and
cartridge
filters.

Current
regulations
specify
the
performance
of
filtration
systems
in
terms
of
filtered
water
turbidity
limits.
Turbidity
is
a
measure
of
the
clarity
of
water
and
is
quantified
in
nephelometric
turbidity
units
(
NTU).
The
1989
SWTR
required
all
surface
water
and
GWUDI
systems
using
rapid
sand
filtration
technologies
to
meet
combined
filter
effluent
turbidity
limits
of
0.5
NTU,
95
percent
of
the
time.
The
1998
Interim
Enhanced
Surface
Water
Treatment
Rule
(
IESWTR)
requires
improved
filtration
performance
by
lowering
the
turbidity
standard
to
0.3
NTU,
95
percent
of
the
time
(
with
a
maximum
of
1
NTU
at
any
time)
for
large
systems
using
rapid
sand
filtration.
The
2002
LT1ESWTR
extended
this
requirement
to
small
systems.
At
this
lower
limit,
EPA
believes
that
systems
are
generally
achieving
a
minimum
of
2
log
(
99
percent)
removal
of
Cryptosporidium.
Slow
sand
and
diatomaceous
earth
filtration
systems
can
achieve
at
least
2
log
removal
at
a
higher
effluent
turbidity
of
1
NTU
95
percent
of
the
time
because
of
differences
in
their
removal
mechanisms.
While
the
degree
of
Cryptosporidium
reduction
achieved
under
these
standards
may
provide
adequate
public
health
protection
for
some
source
waters,
EPA
recognizes
that
some
systems
may
have
higher
levels
of
contamination
where
additional
protection
is
warranted.
Methods
such
as
additional
filtration,
the
use
of
alternative
disinfectants
such
as
ozone
or
ultraviolet
light
(
UV),
improved
source
water
protection,
or
other
treatment
and
management
initiatives
can
help
systems
achieve
additional
protection
against
Cryptosporidium.

The
LT2ESWTR
also
addresses
the
risk
of
microbial
pathogen
contamination
in
unfiltered
systems,
which
lack
the
protective
barriers
from
Cryptosporidium
that
filtered
systems
provide.
The
rule
requires
unfiltered
systems
to
provide
at
least
2
log
inactivation
of
Cryptosporidium;
the
amount
of
disinfection
will
depend
on
the
results
of
source
water
Cryptosporidium
monitoring.
The
rule
also
requires
the
use
of
two
disinfectants
to
meet
Cryptosporidium
and
existing
(
those
for
Giardia
and
viruses)
inactivation
requirements
.

Lastly,
the
LT2ESWTR
addresses
health
risks
posed
by
uncovered
finished
water
reservoirs.
There
are
approximately
140
uncovered
reservoirs
that
hold
finished
water,
not
including
those
that
are
scheduled
to
be
covered
or
taken
off­
line,
ranging
in
size
from
a
few
thousand
to
more
than
three
billion
gallons.
While
the
IEWSTR
and
LT1ESWTR
require
systems
to
cover
all
new
finished
water
reservoirs,
the
LT2ESWTR
builds
on
these
rules
by
addressing
existing
uncovered
finished
water
reservoirs.

2.3
Risk
Balancing
EPA
expects
some
systems
to
change
treatment
practices
in
response
to
the
Stage
2
Disinfectants
and
Disinfection
Byproducts
Rule
(
DBPR)
requirements.
These
changes
have
the
potential
to
increase
the
occurrence
of
microbial
pathogens
in
drinking
water
as
systems
alter
the
use
of
disinfectants
to
comply
with
the
new
disinfection
byproduct
(
DBP)
requirements.
DBPs
result
from
Economic
Analysis
for
the
LT2ESWTR
Proposal
2­
3
June
2003
chemical
reactions
between
disinfectants
and
organic
and
inorganic
compounds
in
the
water.
Some
DBPs
are
associated
with
adverse
health
risks,
including
adverse
developmental
and
reproductive
health
effects
and
cancer.
The
LT2ESTWR,
therefore,
has
additional
disinfection
profiling
and
benchmarking
provisions
to
help
ensure
that
systems
maintain
control
of
microbial
risks
as
they
take
steps
to
reduce
the
formation
of
DBPs.

EPA
is
making
a
concerted
effort
to
understand
and
balance
risks
from
DBPs,
risks
from
microbes,
and
the
costs
and
benefits
of
addressing
those
risks
in
its
rulemaking
efforts.
To
allow
for
simultaneous
compliance
and
balancing
of
risks
between
microbial
pathogens
and
DBPs,
EPA
is
promulgating
the
LT2ESWTR
concurrently
with
the
Stage
2
DBPR.
For
detailed
information
regarding
the
Stage
2
DBPR,
see
the
draft
Economic
Analysis
for
the
Stage
2
Disinfectants
and
Disinfection
Byproducts
Rule
(
USEPA
2003d).

2.4
Public
Health
Concerns
to
Be
Addressed
In
1990,
EPA's
Science
Advisory
Board
(
SAB),
an
independent
panel
of
experts
established
by
Congressional
mandate,
cited
drinking
water
contamination
as
one
of
the
most
important
environmental
risks
and
indicated
that
disease­
causing
microbial
contaminants
(
i.
e.,
bacteria,
protozoa,
and
viruses)
pose
a
particularly
high
health
risk
due
to
the
large
populations
that
are
directly
exposed
to
them
(
SAB
and
USEPA
1990).
Information
on
waterborne
disease
outbreaks
from
the
U.
S.
Centers
for
Disease
Control
and
Prevention
(
CDC)
underscores
this
concern.
Data
collected
by
CDC
indicates
that
between
1971
and
1998,
689
waterborne
disease
outbreaks,
caused
by
various
types
of
contamination,
were
reported
(
Craun
and
Calderon
1996;
Levy
et
al.
1998;
Barwick
et
al.
2000).
However,
since
the
effective
date
of
the
SWTR
(
1993),
the
rate
of
outbreaks
has
decreased.
From
1993
to
1995,
10
to
15
outbreaks
were
reported
annually.
From
1996
to
1998
(
the
most
recent
year
for
which
data
are
available),
there
were
five
to
seven
outbreaks
per
year
(
Kramer
et
al.
1996a,
Levy
et
al.
1998,
Barwick
et
al.
2000).

The
effects
of
waterborne
disease
are
usually
acute,
resulting
from
a
single
or
small
number
of
exposures.
Most
waterborne
pathogens
cause
gastrointestinal
illness
with
diarrhea,
abdominal
discomfort,
nausea,
vomiting,
or
other
symptoms.
Most
such
cases
involve
a
sudden
onset
and
generally
are
of
short
duration
in
healthy
people.
Some
pathogens
(
e.
g.,
Giardia
and
Cryptosporidium),
however,
may
cause
extended
illness,
lasting
weeks
or
longer
in
otherwise
healthy
individuals.
The
infection
can
prove
fatal
for
members
of
sensitive
populations,
such
as
the
immunocompromised
or
the
elderly.
Other
waterborne
pathogens
cause,
or
at
least
are
associated
with,
more
serious
disorders
such
as
hepatitis
(
hepatitis
A)
(
Moore
et
al.
1993),
peptic
ulcers
and
gastric
cancer
(
Helicobacter
pylori)
(
Park
et
al.
2001,
Sepulveda
and
Graham
2002),
myocarditis
(
group
B
coxsackievirus)
(
Kim
et
al.
2001),
meningitis
(
group
B
coxsackievirus
and
echoviruses)
(
Lee
and
Kim
2002,
Amvrosieva
et
al.
2001),
and
many
other
diseases.

2.4.1
Cryptosporidium
Cryptosporidium
is
of
particular
concern
to
EPA
because,
unlike
pathogens
such
as
viruses
and
bacteria,
Cryptosporidium
oocysts
are
resistant
to
inactivation
using
many
common
disinfection
practices.
Since
the
oocyst
is
especially
resistant
to
chlorine
disinfection,
simply
increasing
existing
chlorination
dosage
levels
or
contact
time
above
those
most
commonly
practiced
in
the
United
States
is
not
effective.
Other
emerging
disinfectant­
resistant
pathogens,
such
as
Microsporidia,
Cyclospora,
and
Toxoplasma,
are
also
a
concern
for
similar
reasons.
Economic
Analysis
for
the
LT2ESWTR
Proposal
2­
4
June
2003
Cryptosporidiosis
is
a
protozoal
infection
that
usually
causes
7­
14
days
of
diarrhea,
possibly
accompanied
by
a
low­
grade
fever,
nausea,
and
abdominal
cramps
in
individuals
with
healthy
immune
systems
(
Juranek
1998).
It
is
caused
by
the
ingestion
of
infectious
oocysts,
which
are
readily
carried
in
water.
The
most
common
source
of
oocysts
in
water
is
the
feces
of
infected
hosts
(
Perz
et
al.
1998;
Rose
1997).
Although
cryptosporidiosis
often
occurs
through
ingestion
of
the
infective
oocysts
from
contaminated
food
or
water,
it
may
also
result
from
direct
or
indirect
contact
with
infected
people
or
animals
(
Casemore
1990;
Juranek
1998;
Rose
1997).
Infected
humans
and
other
animals
excrete
oocysts,
which
can
then
be
transmitted
to
others.
Okhuysen
et
al.
(
1998)
and
Dupont
et
al.
(
1995)
found
through
human
volunteer
feeding
studies
that
a
low
dose
of
Cryptosporidium
parvum
(
or
C.
parvum)
is
sufficient
to
cause
infection
in
healthy
adults.

Some
populations
are
at
greater
risk
of
serious
illness
or
death
from
waterborne
disease
than
the
general
population
(
Frost
et
al.
1997).
The
sensitive
populations
include
children
(
especially
the
very
young),
the
elderly,
pregnant
women,
and
the
immunocompromised.
These
sensitive
groups
account
for
almost
20
percent
of
the
population
in
the
United
States
(
Gerba
et
al.
1996;
USEPA
1998a).
The
severity
and
duration
of
illness
are
often
greater
in
immunocompromised
people
than
in
healthy
individuals,
and
death
may
result.
For
instance,
of
the
people
who
died
in
the
1993
Milwaukee
cryptosporidiosis
outbreak,
85
percent
had
AIDS
as
the
underlying
cause
of
death
(
Hoxie
et
al.
1997).

Cryptosporidium
has
caused
a
number
of
documented
waterborne
disease
outbreaks.
However,
it
is
important
to
note
that
C.
parvum
was
not
identified
as
a
human
pathogen
until
1976,
and
outbreaks
attributed
to
cryptosporidiosis
were
not
reported
in
the
United
States
prior
to
1984.
The
first
report
of
an
outbreak
caused
by
Cryptosporidium
was
published
during
the
development
of
the
SWTR
(
D'Antonio
et
al.
1985).
EPA,
CDC,
and
the
Council
of
State
and
Territorial
Epidemiologists
have
maintained
a
collaborative
surveillance
program
for
collection
and
periodic
reporting
of
data
on
waterborne
disease
outbreaks
since
1971.
The
CDC
database
and
biennial
CDC
 
EPA
surveillance
summaries
include
data
reported
voluntarily
by
the
States
on
the
incidence
and
prevalence
of
waterborne
illnesses.

Between
1991,
the
first
year
the
SWTR
and
Total
Coliform
Rule
were
in
effective,
and
2000,
the
most
recent
year
for
which
data
are
available,
95
drinking
water­
related
outbreaks
associated
with
confirmed
or
suspected
microbiological
causes
occurred
in
PWSs.
Twenty
outbreaks
occurred
in
PWSs
with
surface
water
sources;
the
remainder
had
wells
or
springs
as
sources.
Etiology
of
outbreaks
included
Cryptosporidium;
Giardia;
bacteria
such
as
Campylobacter
jejuni,
Shigella
sonnei,
and
E.
coli
O157:
H7;
Norwalk­
like
viruses
and
small
round­
structured
viruses;
and
acute
gastrointestinal
illness
of
unknown
etiology
(
AGI).
These
outbreaks
are
listed
individually
in
Appendix
A
of
the
Occurrence
and
Exposure
Assessment
for
the
Long
Term
2
Enhanced
Surface
Water
Treatment
Rule
and
are
based
on
CDC
surveillance
summaries
(
Moore
et
al.
1993,
Kramer
et
al.
1996a,
Levy
et
al.
1998,
Barwick
et
al.
2000,
Lee
et
al.
2002).

From
1984
to
2000,
there
were
10
reported
outbreaks
of
cryptosporidiosis
associated
with
drinking
water
in
PWSs
in
the
United
States
(
Moore
et
al.
1993;
Kramer
et
al.
1996a;
Craun
1996;
Levy
et
al.
1998;
Barwick
et
al.
2000,
Lee
et
al.
2002).
An
additional
outbreak
occurred
in
a
private
well,
and
another
41
outbreaks
occurred
in
recreational
waters.
Exhibit
2.1
summarizes
the
cryptosporidiosis
outbreaks
associated
with
drinking
water.
Economic
Analysis
for
the
LT2ESWTR
Proposal
2­
5
June
2003
Exhibit
2.1
Reported
Cryptosporidiosis
Outbreaks
in
U.
S.
Drinking
Water
Systems
Year
Location
and
System
Type
Cases
of
Illness
Source
Water
Treatment
Suspected
Cause
1984
Braun
Station,
TX,
CWS
117
(
confirmed)
2,000
(
estimated)
Well
Chlorination
Sewagecontaminated
well
1987
Carrollton,
GA,
CWS
13,000
(
estimated)
River
Conventional
filtration/
chlorination;
inadequate
backwashing
of
some
filters
Treatment
deficiencies
1991
Berks
County,
PA,
NCWS
551
(
estimated)
Well
Chlorination
Ground
water
under
the
influence
of
contaminated
surface
water
1992
Medford
(
Jackson
County
and
Talent),
OR,
CWS
3,000
(
estimated);
combined
total
for
Jackson
County
and
Talent
Spring/
River
Chlorination/
package
filtration
plant
Source
not
identified
for
Jackson
County;
treatment
deficiencies
at
water
treatment
plant
in
Talent
1993
Milwaukee,
WI,
CWS
403,000
(
estimated)
Lake
Conventional
filtration
High
source
water
contamination
and
treatment
deficiencies
1993
Cook
County,
MN,
NCWS
27
(
confirmed)
Lake
Filtered,
chlorinated
Possible
sewage
backflow
from
toilet/
septic
tank
1994
Clark
County,
NV,
CWS
103
(
confirmed);
many
were
HIV
positive
River/
Lake
Prechlorination,
filtration
and
postfiltration
chlorination
Source
not
identified
1994
Walla
Walla,
WA,
CWS
134
(
confirmed)
Well
None
reported
Sewage
contamination
1998
Williamson
County,
TX,
CWS
1,400
(
confirmed)
Well
Chlorinated
Sewage
contamination
2000
Florida,
CWS
5
Well
Chlorinated
Broken
well,
treatment
deficiencies
Source:
Craun
et
al.
(
1998),
Barwick
et
al.
(
2000),
and
Lee
et
al.
(
2002).

Five
of
the
10
outbreaks
in
Exhibit
2.1
originated
from
surface
water
or
possibly
GWUDI
supplied
by
PWSs
serving
fewer
than
10,000
people.
In
total,
the
10
outbreaks
caused
an
estimated
421,337
cases
of
illness,
the
majority
occurring
in
Milwaukee
in
1993.
These
outbreaks
demonstrate
that
when
treatment
is
not
operating
optimally
or
when
source
water
is
highly
contaminated,
Cryptosporidium
can
be
present
in
the
finished
drinking
water
and
infect
consumers,
ultimately
resulting
in
disease
outbreaks.

The
National
Research
Council
concluded
that
the
number
of
identified
and
reported
outbreaks
in
the
CDC
database
(
both
for
surface
and
ground
waters)
represents
a
small
percentage
of
actual
waterborne
disease
outbreaks
(
National
Research
Council
1997).
Most
outbreaks
in
CWSs
are
not
recognized
until
a
sizable
proportion
of
the
population
is
ill
(
Perz
et
al.
1998,
Craun
1996).
In
addition
to
Economic
Analysis
for
the
LT2ESWTR
Proposal
2­
6
June
2003
the
complications
involved
in
identifying
waterborne
disease
outbreaks,
some
States
do
not
have
active
outbreak
surveillance
programs.
Those
that
do
exist
are
based
on
voluntary
and
confidential
responses
by
State
and
local
public
health
officials.
Even
when
outbreaks
are
recognized,
few
are
successfully
traced
to
the
drinking
water
source.
Physicians,
for
instance,
may
not
have
sufficient
community­
wide
information
to
attribute
gastrointestinal
illness
to
any
specific
origin,
such
as
a
drinking
water
source.
Many
people
who
experience
gastrointestinal
illness
(
predominantly
manifested
as
diarrhea)
do
not
seek
medical
attention,
and
some
healthy
adults
with
cryptosporidiosis
may
not
suffer
severe
symptoms
from
the
disease.
Even
if
infected
individuals
consult
a
physician,
Cryptosporidium
is
not
identified
by
routine
diagnostic
tests
for
gastroenteritis
and,
therefore,
tends
to
be
under­
reported
in
the
general
population
(
Craun
1996).

The
limited
number
of
reported
cases
of
waterborne
disease
such
as
cryptosporidiosis
may
be
due
to
the
fact
that
a
significant
portion
of
these
illnesses
may
be
endemic
(
i.
e.,
not
associated
with
an
outbreak),
and
thus
are
even
more
difficult
to
recognize.
One
study,
for
example,
found
that
14
to
40
percent
of
the
normal
gastrointestinal
illness
in
a
community
in
Quebec
was
associated
with
treated
drinking
water
from
a
surface
water
source
(
Payment
et
al.
1997).

2.4.2
Uncovered
Finished
Water
Reservoirs
Many
PWSs
store
treated
drinking
water
in
some
type
of
reservoir
before
delivering
it
to
their
customers.
Although
good
engineering
practice
dictates
that
such
reservoirs
be
covered
to
prevent
recontamination,
there
are
currently
no
regulations
that
require
existing
reservoirs
to
be
covered
(
the
IESWTR
and
LT1ESWTR
require
new
reservoirs
to
be
covered).
The
use
of
uncovered
finished
water
reservoirs
has
been
questioned
since
1930
because
of
their
susceptibility
to
contamination
and
subsequent
threats
to
public
health.
Many
sources
of
contamination
can
lead
to
the
degradation
of
water
quality
in
uncovered
finished
water
reservoirs.
These
include,
but
are
not
limited
to,
surface
water
runoff,
algal
growth,
insects
and
fish,
bird
and
animal
waste,
airborne
deposition,
and
human
activity.
Algal
blooms
are
the
most
common
problem
in
open
reservoirs
and
can
become
a
public
health
risk.
Algae
growth
leads
to
the
formation
of
DBPs
and
causes
taste
and
odor
problems.
Algae
also
provide
a
food
source
for
bacteria
that
decompose
plant
matter.
Some
blue­
green
algae
(
actually
a
type
of
bacteria
called
cyanobacteria)
contain
toxins
that
can
induce
headaches,
fever,
diarrhea,
abdominal
pain,
nausea,
and
vomiting.
Bird
and
animal
wastes
are
other
common
and
significant
sources
of
contamination.
These
wastes
may
carry
microbial
contaminants
such
as
coliform
bacteria,
viruses,
and
human
pathogens,
including
Vibrio
cholera,
Salmonella,
Mycobacteria,
bacteria
that
cause
typhoid
fever,
and
Giardia
in
addition
to
Cryptosporidium
(
USEPA
1999c).
Microbial
pathogens
can
also
be
found
in
surface
water
runoff
along
with
agricultural
chemicals,
automotive
wastes,
turbidity,
metals,
and
organic
matter
(
USEPA
1999c;
LeChevallier
et
al.
1997b).
In
an
effort
to
minimize
contamination,
systems
have
implemented
controls,
such
as
reservoir
covers
and
liners,
regular
draining
and
washing,
proper
security
and
monitoring,
bird
and
insect
control
programs,
and
drainage
designed
to
prevent
surface
runoff
from
entering
the
reservoir
(
USEPA
1999c).

Few
studies
quantitatively
evaluate
the
impacts
of
uncovered
finished
water
reservoirs
on
public
health.
LeChevallier
et
al.
(
1997b)
compared
the
influent
and
the
effluent
water
quality
from
six
New
Jersey
reservoirs
for
a
1­
year
period
to
determine
the
impact
of
uncovered
finished
water
storage
reservoirs
on
water
quality.
There
were
significant
increases
in
turbidity,
particle
counts,
total
coliform,
fecal
coliform,
and
heterotrophic
plate
count
bacteria
in
the
effluent
compared
to
the
influent.
There
was
Economic
Analysis
for
the
LT2ESWTR
Proposal
2­
7
June
2003
also
a
significant
decrease
in
the
chlorine
residual
in
the
effluent
samples,
meaning
little
chlorine
would
be
left
to
provide
continued
disinfection
in
the
distribution
system.

2.5
Regulatory
History
The
primary
responsibility
for
regulating
the
quality
of
drinking
water
lies
with
EPA.
The
Safe
Drinking
Water
Act
(
SDWA)
establishes
this
responsibility
and
defines
the
mechanisms
at
the
Agency's
disposal
to
protect
public
health.
EPA
sets
standards
by
identifying
which
contaminants
should
be
regulated
and
by
establishing
the
maximum
levels
of
the
contaminants
allowed
in
drinking
water
specifying
treatment
techniques
to
reduce
contaminant
levels.

2.5.1
Statutory
Authority
for
Promulgating
the
Rule
Section
1412(
b)(
1)
of
the
1996
SDWA
reauthorization
mandated
new
drinking
water
requirements.
EPA's
general
authority
to
set
Maximum
Contaminant
Level
Goals
(
MCLGs)
and
develop
National
Primary
Drinking
Water
Regulations
(
NPDWRs)
was
modified
to
apply
to
contaminants
that
"
may
have
an
adverse
effect
on
the
health
of
persons,"
are
"
known
to
occur
or
there
is
a
substantial
likelihood
that
the
contaminant
will
occur
in
PWSs
with
a
frequency
and
at
levels
of
public
health
concern,"
and
for
which,
"
in
the
sole
judgment
of
the
Administrator,
regulation
of
such
contaminant
presents
a
meaningful
opportunity
for
health
risk
reductions
for
persons
served
by
public
water
systems"
(
SDWA
1412(
b)(
1)(
A)).

To
regulate
a
contaminant,
EPA
first
sets
an
MCLG
at
a
level
at
which
no
known
or
anticipated
adverse
health
effects
occur.
MCLGs
are
established
solely
on
the
basis
of
protecting
public
health
and
are
not
enforceable.
EPA
simultaneously
sets
an
enforceable
Maximum
Contaminant
Level
(
MCL)
as
close
as
technologically
feasible
to
the
MCLG,
while
taking
costs
into
consideration.
If
it
is
not
feasible
to
measure
the
contaminant
at
levels
presumed
to
have
impacts
on
health,
a
treatment
technique
can
be
specified
in
place
of
an
MCL.
Water
systems
comply
with
a
drinking
water
regulation
by
not
exceeding
the
MCL
or
by
meeting
treatment
technique
requirements.

In
addition
to
the
general
authorities
cited
above,
SDWA
Section
1412(
b)(
2)(
C)
requires
specifically
that
EPA
promulgate
the
Final
Enhanced
Surface
Water
Treatment
Rule
(
ESWTR).

The
Administrator
shall
promulgate
an
Interim
Enhanced
Surface
Water
Treatment
Rule,
a
Final
Enhanced
Surface
Water
Treatment
Rule,
a
Stage
1
Disinfectants
and
Disinfection
Byproducts
Rule,
and
a
Stage
2
Disinfectants
and
Disinfection
Byproducts
Rule
in
accordance
with
the
schedule
published
in
volume
29,
Federal
Register,
Page
6361
(
February
10,
1994),
in
Table
III.
13
of
the
proposed
Information
Collection
Rule.

The
promulgation
of
the
IESWTR
and
LT1ESWTR
satisfied
the
statutory
requirement
for
an
interim
rule,
and
the
LT2ESWTR
satisfies
the
requirement
for
a
final
rule
and
the
Congressional
intent
to
review
and
revise
the
IESWTR
and
LT1ESWTR
based
on
data
available
from
the
Information
Collection
Rule
(
ICR)
and
Information
Collection
Rule
Supplemental
Survey
(
ICRSSs)
(
see
section
2.5.5).
Also,
to
achieve
the
goals
of
the
Stage
2
DBPR,
the
LT2ESWTR
must
be
promulgated
to
achieve
a
balance
between
the
risks
of
microbial
pathogens
and
DBPs.
Economic
Analysis
for
the
LT2ESWTR
Proposal
2­
8
June
2003
The
following
sections
summarize
the
development
of
NPDWRs
over
the
past
20
years.

2.5.2
1979
Total
Trihalomethane
Rule
Under
the
Total
Trihalomethane
Rule
(
44
FR
68624
November
1979),
EPA
set
an
MCL
for
total
trihalomethanes
(
TTHM),
the
sum
of
chloroform,
bromoform,
bromodichloromethane,
dibromochloromethane,
of
0.10
mg/
L
as
a
running
annual
average
(
RAA)
of
quarterly
samples.
This
standard
applied
to
CWSs
using
surface
or
ground
water
that
served
at
least
10,000
people
and
that
added
a
disinfectant
to
the
drinking
water
during
any
part
of
the
treatment
process.
This
1979
rule
was
superseded
by
the
1998
Stage
1
DBPR
(
section
2.5.7).

2.5.3
1989
Total
Coliform
Rule
The
Total
Coliform
Rule
(
TCR)
(
54
FR
27544
June
1989)
applies
to
all
PWSs.
Because
monitoring
PWSs
for
every
possible
pathogenic
organism
is
not
feasible,
coliform
organisms
are
used
as
indicators
of
possible
distribution
system
contamination.
Coliforms
are
easily
detected
in
water
and
are
used
to
indicate
a
system's
vulnerability
to
pathogens.
In
the
TCR,
EPA
set
an
MCLG
of
zero
for
total
coliforms.
EPA
also
set
a
monthly
MCL
for
total
coliforms
and
required
testing
of
total
coliform­
positive
cultures
for
the
presence
of
E.
coli
or
fecal
coliforms.
E.
coli
and
fecal
coliforms
indicate
more
immediate
health
risks
from
sewage
or
fecal
contamination
and
result
in
an
acute
MCL
violation
(
acute
MCL
violations
require
immediate
public
notification).
Coliform
monitoring
frequency
is
determined
by
population
served,
the
type
of
system
(
community
or
noncommunity)
and
they
type
of
source
water
(
surface
water
or
GWUDI,
or
ground
water).
In
addition,
the
TCR
required
sanitary
surveys
every
5
years
(
or
10
years
for
NCWSs
using
disinfected
ground
water)
for
systems
that
collect
fewer
than
5
routine
total
coliform
samples
per
month
(
typically
serving
fewer
than
4,100
people).

2.5.4
1989
Surface
Water
Treatment
Rule
Under
the
SWTR
(
54
FR
27486
June
1989),
EPA
set
MCLGs
of
zero
for
Giardia
lamblia,
viruses,
and
Legionella,
and
established
treatment
requirements
for
all
PWSs
using
surface
water
or
GWUDI
as
a
source.
The
SWTR
includes
treatment
technique
requirements
for
filtered
and
unfiltered
systems
that
are
intended
to
protect
against
the
adverse
health
effects
associated
with
Giardia
lamblia,
viruses,
and
Legionella,
as
well
as
many
other
pathogenic
organisms.
These
requirements
include
the
following:

°
Maintenance
of
a
disinfectant
residual
in
water
entering
and
within
the
distribution
system.

°
Removal/
inactivation
of
at
least
99.9
percent
(
3
log)
of
Giardia
and
99.99
percent
(
4
log)
of
viruses.

°
Filtration,
unless
systems
meet
specified
avoidance
criteria.

°
For
filtered
systems,
meet
a
turbidity
performance
standard
for
the
combined
filter
effluent
consisting
of
a
5
NTU
maximum
and
95
percent
of
measurements
in
one
month
not
to
exceed
0.5
NTU,
based
on
4­
hour
monitoring
for
treatment
plants
using
conventional
treatment
or
Economic
Analysis
for
the
LT2ESWTR
Proposal
2­
9
June
2003
direct
filtration
(
with
separate
standards
for
other
filtration
technologies).
These
particular
requirements
were
superseded
by
the
1998
IESWTR
and
the
2002
LT1ESWTR.

°
Watershed
control
programs
and
water
quality
requirements
for
unfiltered
systems.

2.5.5
1996
Information
Collection
Rule
The
ICR
(
61
FR
24354
May
1996)
applied
to
PWSs
serving
more
than
100,000
people.
A
more
limited
set
of
ICR
requirements
pertained
to
ground
water
systems
serving
50,000
to
100,000
people.

The
ICR
authorized
EPA
to
collect
occurrence
and
treatment
information
to
help
evaluate
the
need
for
possible
changes
to
the
microbial
requirements
and
microbial
treatment
practices
and
to
help
evaluate
the
need
for
future
regulation
of
disinfectants
and
DBPs.
The
ICR
provided
EPA
with
additional
information
on
the
national
occurrence
in
drinking
water
of
(
1)
chemical
byproducts
that
form
when
disinfectants
used
for
microbial
control
react
with
naturally
occurring
compounds
present
in
source
water;
and
(
2)
disease­
causing
microorganisms,
including
Cryptosporidium,
Giardia,
viruses,
and
coliform
bacteria.
The
ICR
also
required
water
systems
to
collect
plant
configuration
data
showing
the
type
of
treatment
provided.
The
ICR
monthly
sampling
data
provided
a
total
of
18
months
of
influent
and
treated
water
quality
data
including
pH,
alkalinity,
turbidity,
temperature,
calcium
and
total
hardness,
total
organic
carbon,
UV
254,
bromide,
ammonia,
and
disinfectant
residual.
These
data
provided
an
indication
of
the
"
treatability"
of
the
water,
the
occurrence
of
contaminants,
and
the
potential
for
DBP
formation.
The
data
collected
under
the
ICR
were
used
in
analyses
supporting
development
of
the
LT2ESWTR
and
Stage
2
DBPR.

2.5.6
1998
Interim
Enhanced
Surface
Water
Treatment
Rule
The
IESWTR
(
63
FR
69478
December
1998)
updated
the
1989
SWTR
for
large
systems.
It
applies
to
PWSs
serving
at
least
10,000
people
and
using
surface
water
or
GWUDI
as
a
source.
These
systems
were
to
comply
with
the
IESWTR
by
January
2002.
The
primary
purpose
of
the
IESWTR
is
to
improve
control
of
Cryptosporidium
and
to
address
tradeoffs
between
the
risks
from
microbial
pathogens
and
those
from
DBPs.
The
requirements
and
guidelines
include:

°
An
MCLG
of
zero
for
Cryptosporidium
°
Removal
of
99
percent
(
2
log)
of
Cryptosporidium
for
systems
that
provide
filtration
°
For
treatment
plants
using
conventional
treatment
or
direct
filtration,
a
turbidity
performance
standard
for
the
combined
filter
effluent
consising
of
a
1
NTU
maximum
and
95
percent
of
measurements
in
one
month
not
to
exceed
0.3
NTU,
based
on
4­
hour
monitoring
°
Continuous
monitoring
of
individual
filter
effluent
turbidity
in
conventional
and
direct
filtration
plants
and
recording
turbidity
readings
every
15
minutes
when
these
filters
are
on­
line
°
A
disinfection
benchmark
to
assess
the
level
of
microbial
protection
provided
before
facilities
change
their
disinfection
practices
to
meet
the
requirements
of
the
Stage
1
DBPR
°
Inclusion
of
Cryptosporidium
in
the
definition
of
GWUDI
and
in
the
watershed
control
requirements
for
unfiltered
PWSs
Economic
Analysis
for
the
LT2ESWTR
Proposal
2­
10
June
2003
°
Covers
for
all
new
finished
water
storage
facilities
°
A
primacy
provision
that
requires
States
to
conduct
sanitary
surveys
with
a
minimum
frequency
for
all
surface
water
systems,
including
those
serving
fewer
than
10,000
people
EPA
promulgated
the
IESWTR
concurrently
with
the
Stage
1
DBPR
so
that
systems
could
coordinate
their
response
to
the
risks
posed
by
DBPs
and
microbial
pathogens.

2.5.7
1998
Stage
1
Disinfectants
and
Disinfection
Byproducts
Rule
The
Stage
1
DBPR
(
63
FR
69390
December
1998)
applies
to
all
CWSs
and
NTNCWSs
that
add
a
chemical
disinfectant
to
their
water.
Certain
requirements
designed
to
provide
protection
against
acute
health
effects
from
chlorine
dioxide
also
apply
to
transient
noncommunity
water
systems
(
TNCWSs).
Surface
water
and
GWUDI
systems
serving
at
least
10,000
people
were
required
to
comply
with
the
rule
by
January
2002.
Surface
water
and
GWUDI
systems
serving
fewer
than
10,000
people
and
all
ground
water
systems
must
comply
by
January
2004.

The
Stage
1
DBPR
sets
maximum
residual
disinfectant
level
goals
(
MRDLGs)
for
chlorine
(
4
mg/
L
as
Cl
2),
chloramines
(
4
mg/
L
as
Cl
2),
and
chlorine
dioxide
(
0.8
mg/
L
as
ClO
2);
and
MCLGs
for
bromodichloromethane
(
0
mg/
L),
bromoform
(
0
mg/
L),
dibromochloromethane
(
0.06
mg/
L),
dichloroacetic
acid
(
0
mg/
L),
trichloroacetic
acid
(
0.3
mg/
L),
bromate
(
0
mg/
L),
and
chlorite
(
0.8
mg/
L).
The
rule
sets
MRDLs
for
chlorine
(
4
mg/
L
as
Cl
2),
chloramines
(
4
mg/
L
as
Cl
2),
and
chlorine
dioxide
(
0.8
mg/
L
as
ClO
2);
and
MCLs
for
TTHM
(
0.080
mg/
L),
HAA5
(
0.060
mg/
L),
bromate
(
0.010
mg/
L),
and
chlorite
(
1.0
mg/
L).
The
MRDLs
and
MCLs,
except
those
for
chlorite
and
chlorine
dioxide,
are
calculated
as
RAAs.
For
conventional
surface
water
and
GWUDI
systems,
a
treatment
technique
 
enhanced
coagulation/
softening
 
is
specified
for
the
removal
of
DBP
precursors.

As
noted
in
section
2.5.6,
the
Stage
1
DBPR
was
promulgated
concurrently
with
the
IESWTR
to
coordinate
the
control
of
DBPs
and
microbial
contaminants.

2.5.8
2000
Proposed
Ground
Water
Rule
The
proposed
Ground
Water
Rule
(
65
FR
30194
May
2000)
addresses
fecal
contamination
in
ground
water
systems.
The
rule
also
builds
on
the
TCR
through
provisions
based
on
further
evaluation
of
E.
coli
monitoring
results
measured
under
the
TCR.
Key
components
of
the
multibarrier
approach
for
protection
of
ground
water
included
in
the
proposed
rule
are:

°
Sanitary
surveys
for
all
ground
water
systems
conducted
at
the
same
frequency
as
in
surface
water
systems
°
Hydrogeologic
sensitivity
assessments
to
identify
ground
water
sources
that
are
susceptible
to
fecal
contamination
°
Source
water
monitoring
for
an
indicator
of
fecal
contamination
for
systems
drawing
from
susceptible
ground
water
sources
Economic
Analysis
for
the
LT2ESWTR
Proposal
2­
11
June
2003
°
Correction
of
significant
deficiencies
and
fecal
contamination
by
eliminating
the
source
of
contamination,
correcting
the
deficiency,
providing
an
alternative
source
of
water,
or
providing
inactivation
and/
or
removal
of
99.99
percent
(
4
log)
of
viruses
°
Compliance
monitoring
to
ensure
that
disinfection
treatment
is
reliably
operated
when
it
is
used
2.5.9
2001
Filter
Backwash
Recycling
Rule
The
Filter
Backwash
Recycling
Rule
(
FBRR)
(
66
FR
31086
June
2001)
regulates
systems
where
filter
backwash
is
returned
to
the
treatment
process.
The
rule
applies
to
surface
water
and
GWUDI
systems
that
use
direct
or
conventional
filtration
and
recycle
spent
filter
backwash
water,
sludge
thickener
supernatant,
or
liquids
from
dewatering
processes.
The
rule
requires
that
these
recycled
liquids
be
returned
to
a
location
such
that
all
steps
of
a
system's
conventional
or
direct
filtration
are
employed.
The
rule
also
requires
systems
to
notify
the
State
that
they
practice
recycling.
Finally,
systems
must
collect
and
maintain
information
for
review
by
the
State.

2.5.10
2002
Long
Term
1
Enhanced
Surface
Water
Treatment
Rule
The
LT1ESWTR
(
67
FR
1812
January
2002)
is
an
extension
of
the
1998
IESWTR
to
small
systems.
LT1ESWTR
extends
control
of
Cryptosporidium
and
other
disease­
causing
microbes
to
surface
water
and
GWUDI
systems
that
serve
fewer
than
10,000
people.
Key
provisions
in
the
LT1ESWTR
are
very
similar
to
those
for
the
IESWTR,
but
provide
additional
flexibility
for
small
systems.

2.5.11
2003
Proposed
Stage
2
Disinfectants
and
Disinfection
Byproducts
Rule
The
requirements
of
the
Stage
2
DBPR
apply
to
all
CWSs
and
NTNCWSs
that
add
a
disinfectant
other
than
UV
or
that
deliver
water
that
has
been
treated
with
a
disinfectant
other
than
UV.
The
Stage
2
DBPR
builds
on
the
1979
Total
Trihalomethane
Rule
and
the
1998
Stage
1
DBPR
by
requiring
reduced
levels
of
DBPs
in
distribution
systems.
Each
rule
activity
for
the
Preferred
Regulatory
Alternative
and
the
associated
rule
schedule
are
described
below.

The
Stage
2
DBPR
is
designed
to
reduce
DBP
occurrence
peaks.
in
the
distribution
system
by
changing
compliance
monitoring
requirements.
Compliance
monitoring
will
be
preceded
by
an
initial
distribution
system
evaluation
(
IDSE)
to
identify
distribution
system
locations
that
represent
high
total
trihalomethane
(
TTHM)
and
haloacetic
acids
(
HAA5)
levels.
Systems
may
perform
an
IDSE
either
by
completing
a
system­
specific
study
(
SSS)
or
a
standard
monitoring
program
(
SMP).
NTNCWSs
serving
fewer
than
10,000
people
are
not
required
to
conduct
an
IDSE,
and
other
systems
may
receive
waivers
from
the
IDSE
requirement.

The
Stage
2
DBPR
changes
the
way
sampling
results
are
averaged
to
determine
compliance.
The
compliance
determination
for
the
Stage
2
DBPR
is
based
on
a
locational
running
annual
average
(
LRAA)
instead
of
the
system­
wide
RAA
used
under
the
Stage
1
DBPR.
LRAAs
are
RAAs
calculated
separately
for
each
sample
location
in
the
distribution
system.
With
the
Stage
2
LRAA
requirement,
the
Economic
Analysis
for
the
LT2ESWTR
Proposal
2­
12
June
2003
TTHM
and
HAA5
MCLs
must
be
met
at
each
monitoring
location,
while
the
Stage
1
RAA
requires
a
system
to
average
results
over
all
monitoring
locations.

The
Stage
2
DBPR,
which
is
being
promulgated
simultaneously
with
the
LT2ESWTR
to
address
complex
risk
trade­
offs
between
DBPs
and
microbial
pathogens,
will
be
implemented
in
two
stages.

Stage
2A:
Starting
[
3
years
after
rule
promulgation],
all
systems
must
comply
with
TTHM/
HAA5
MCLs
of
120/
100
µ
g/
L
measured
as
LRAAs
at
each
Stage
1
DBPR
monitoring
site
and
must
continue
to
comply
with
the
Stage
1
DBPR
TTHM/
HAA5
MCLs
of
80/
60
µ
g/
L,
measured
as
RAAs.

Stage
2B:
Starting
[
6
years
after
rule
promulgation],
systems
serving
at
least
10,000
people
must
comply
with
TTHM/
HAA5
MCLs
of
80/
60
µ
g/
L
measured
as
LRAAs
at
the
monitoring
sites
identified
during
an
IDSE.
For
small
systems
required
to
perform
Cryptosporidium
monitoring
under
the
LT2ESWTR,
compliance
with
the
80/
60
µ
g/
L
MCLs,
measured
as
LRAAs,
will
begin
[
8.5
years
after
rule
promulgation].
For
all
other
small
systems,
compliance
with
the
80/
60
µ
g/
L
MCLs,
measured
as
LRAAs,
will
begin
[
7.5
years
after
rule
promulgation].
If
a
system
requires
capital
improvements,
the
State/
Primacy
Agency
may
grant
up
to
an
additional
24
months
for
compliance.

2.6
Economic
Rationale
for
Regulation
This
section
addresses
the
economic
rationale
for
choosing
a
regulatory
approach.
Such
a
rationale
for
the
rule
is
required
by
Executive
Order
Number
12866,
Regulatory
Planning
and
Review
(
The
White
House
1993),
which
states
the
following:

...[
E]
ach
agency
shall
identify
the
problem
that
it
intends
to
address
(
including,
where
applicable,
the
failures
of
the
private
markets
or
public
institutions
that
warrant
new
agency
action)
as
well
as
assess
the
significance
of
that
problem.
(
Section
1,
b(
1))

In
addition,
Office
of
Management
and
Budget
(
OMB)
guidance
dated
January
11,
1996,
states
that
"
in
order
to
establish
the
need
for
the
proposed
action,
the
analysis
should
discuss
whether
the
problem
constitutes
a
significant
market
failure"
(
USEPA
1996b).

In
a
perfectly
competitive
market,
prices
and
quantities
are
determined
solely
by
the
aggregated
decisions
of
buyers
and
sellers.
Such
a
market
occurs
when
many
producers
of
a
product
are
selling
to
many
buyers,
and
both
producers
and
consumers
have
perfect
information
on
the
characteristics
and
prices
of
each
firm's
products.
Barriers
to
entry
in
the
industry
cannot
exist,
and
individual
buyers
and
sellers
must
be
"
price
takers":
i.
e.,
their
decisions
cannot
affect
the
price.
Several
properties
of
the
public
water
supply
do
not
satisfy
the
conditions
for
a
perfectly
competitive
market
and,
thus,
lead
to
market
failures
that
require
regulation.

First,
many
water
systems
are
natural
monopolies.
A
natural
monopoly
exists
when
it
is
impossible
for
more
than
one
firm
in
each
area
to
recover
the
costs
of
production
and
survive.
There
are
high
fixed
costs
associated
with
reservoirs
and
wells,
transmission
and
distribution
systems,
treatment
plants,
and
other
facilities.
For
other
potential
suppliers
to
enter
the
market,
they
would
have
to
provide
Economic
Analysis
for
the
LT2ESWTR
Proposal
2­
13
June
2003
the
same
extensive
infrastructure
to
realize
similar
economies
of
scale
and
be
competitive.
A
splitting
of
the
market
with
increased
fixed
costs
(
for
example,
two
supplier
networks
in
a
single
market)
usually
makes
this
situation
unprofitable
for
one
or
both
suppliers.
The
result
is
a
market
suitable
for
a
single
supplier
and
one
that
is
hostile
to
alternative
suppliers.
In
such
natural
monopolies,
suppliers
have
fewer
incentives
for
providing
high­
quality
service
or
maintaining
competitive
prices.
In
these
situations,
governments
often
intervene
to
help
protect
the
public
interest.

Because
PWSs
are
legal
as
well
as
natural
monopolies,
they
are
often
subject
to
price
controls,
if
not
outright
public
ownership.
While
customers
may
demand
improvements
in
water
quality,
the
regulatory
regime
may
not
transmit
that
demand
to
the
water
supplier
or
allow
the
supplier
to
raise
its
price
to
recover
the
cost
of
the
improvements.
If
consumers
do
not
believe
that
their
drinking
water
is
safe
enough,
they
cannot
simply
switch
to
another
water
utility.
Other
options
for
obtaining
safe
drinking
water
(
e.
g.,
buying
bottled
water
or
installing
point­
of­
use
filtration)
cost
consumers
more
than
the
purchase
from
public
water
supplies.
Therefore,
the
water
supplier
may
have
little
incentive
to
improve
water
quality.

Second,
the
public
may
not
understand
the
health
and
safety
issues
associated
with
drinking
water
quality.
Understanding
the
health
risks
posed
by
trace
quantities
of
drinking
water
contaminants
involves
analysis
and
synthesis
of
complex
toxicological
and
health
sciences
data.
Therefore,
the
public
may
not
be
aware
of
the
risks
it
faces.
For
example,
cases
of
waterborne
disease
are
likely
to
be
under­
reported
since
a
significant
portion
may
be
endemic,
making
them
more
difficult
to
recognize.
There
is,
therefore,
a
lack
of
occurrence
data
and
related
cost
information
on
endemic
waterborne
disease
available
to
the
public.
EPA
has
implemented
a
Consumer
Confidence
Report
(
CCR)
Rule
(
USEPA
1998)
that
makes
water
quality
information
more
easily
available
to
consumers.
This
rule
requires
CWSs
to
publish
an
annual
report
on
Local
drinking
water
quality.
Consumers,
however,
still
have
to
analyze
this
information
for
its
health
risk
implications.
Even
if
informed
consumers
are
able
to
engage
water
systems
in
a
dialogue
about
health
issues,
the
costs
of
such
interaction
(
measured
in
personal
time
and
monetary
outlays)
present
a
significant
impediment
to
consumer
expression
of
risk
reduction
preferences.
Moreover,
these
reports
typically
contain
no
information
about
the
risk
associated
with
Cryptosporidium
and
most
other
microbial
pathogens,
because
PWSs
are
not
required
to
analyze
for
them.

SDWA
regulations
are
intended
to
provide
a
level
of
protection
from
exposure
to
drinking
water
contaminants
by
setting
minimum
performance
requirements.
These
regulations
are
not
intended
to
restructure
market
mechanisms
or
to
establish
competition
in
supply;
rather,
they
establish
the
level
of
service
to
be
provided
that
best
reflects
public
preference
for
safety.
The
Federal
regulations
reduce
the
high
information
and
transaction
costs
by
acting
on
behalf
of
consumers
in
balancing
risk
reduction
and
the
social
costs
of
achieving
this
risk
reduction.
1
The
Stage
2
M­
DBP
Advisory
Committee
was
composed
of
representatives
from
a
variety
of
stakeholder
organizations.
A
complete
list
of
participating
members
(
as
well
as
a
summary
of
committee
findings)
is
included
in
the
docket
for
the
LT2ESWTR.

Economic
Analysis
for
the
LT2ESWTR
Proposal
3­
1
June
2003
3.
Consideration
of
Regulatory
Alternatives
3.1
Introduction
The
U.
S.
Environmental
Protection
Agency
(
EPA
or
the
Agency)
evaluated
a
number
of
regulatory
alternatives
that
could
mitigate
the
health
concerns
addressed
by
the
Long
Term
2
Enhanced
Surface
Water
Treatment
Rule
(
LT2ESWTR).
These
evaluations
took
place
during
a
regulatory
negotiation
process
that
began
in
the
Spring
of
1999,
and
included
consultation
with
the
Stage
2
Microbial
and
Disinfection
Byproducts
(
Stage
2
M­
DBP)
Advisory
Committee
that
was
convened
under
the
Federal
Advisory
Committees
Act
(
FACA).
This
chapter
summarizes
the
alternatives
considered
and
develops
a
context
for
the
proposed
regulatory
approach.
The
remainder
of
the
chapter
is
organized
as
follows:

3.2
Development
Process
for
Regulatory
Alternatives
3.3
Specific
Regulatory
Alternatives
Considered
in
this
EA
3.3.1
Summary
of
Bin
Classification
and
Treatment
Requirements
for
Regulatory
Alternatives
3.3.2
Additional
Treatment
for
Direct
Filtration
Systems
3.4
Alternative
Monitoring
Approaches
Considered
3.4.1
Indicators
of
Microbial
Contamination
3.4.2
Cryptosporidium
Monitoring
Strategies
for
Bin
Classification
3.2
Development
Process
for
Regulatory
Alternatives
Two
efforts
in
the
regulatory
development
process
for
the
LT2ESWTR
are
particularly
relevant
to
evaluation
of
alternatives
discussed
in
the
Economic
Analysis
(
EA):
(
1)
the
data
synthesis
and
analysis
resulting
from
the
Information
Collection
Rule
(
ICR)
and
ICR
Supplementary
Surveys
(
ICRSSs),
and
(
2)
the
deliberations
and
recommendations
of
the
Stage
2
M­
DBP
Advisory
Committee.
1
EPA
held
14
formal
negotiation
meetings
of
the
Stage
2
M­
DBP
Advisory
Committee,
between
March
1999
and
September
2000.
Before
convening
the
committee,
EPA
also
held
three
preparatory
stakeholder
meetings
on
pathogen
and
disinfection
byproduct
(
DBP)
health
effects,
occurrence,
and
treatment.
The
objective
of
the
committee
meetings
was
to
reach
a
consensus
regarding
recommended
provisions
for
the
two
rules
(
Stage
2
Disinfectants
and
Disinfection
Byproducts
Rule
(
DBPR)
and
LT2ESWTR).

Technical
support
for
the
Stage
2
M­
DBP
negotiation
meetings
was
provided
by
the
Technical
Work
Group
(
TWG),
which
was
established
by
the
committee
at
its
first
meeting
and
which
was
composed
of
EPA
and
drinking
water
industry
experts.
The
TWG's
activities
resulted
in
the
collection,
development,
evaluation,
and
presentation
of
key
data
related
to
the
LT2ESWTR,
including
new
data
on
pathogenicity,
occurrence,
and
treatment
of
microbial
contaminants,
specifically
Cryptosporidium.
Economic
Analysis
for
the
LT2ESWTR
Proposal
3­
2
June
2003
The
ICR
database
provided
much
of
the
information
evaluated
for
the
LT2ESWTR.
EPA
promulgated
the
ICR
in
1996
pursuant
to
the
Safe
Drinking
Water
Act
(
SDWA)
requirements.
The
ICR
required
approximately
300
large
public
water
systems
(
PWSs)
with
approximately
500
separate
water
treatment
plants
to
conduct
18
months
of
sampling
for
water
quality
and
treatment
parameters
related
to
DBP
formation
and
the
occurrence
of
microbial
pathogens.
Subsequent
to
the
ICR
data
collection,
EPA
obtained
additional
data
on
pathogen
occurrence
through
the
ICRSSs.
These
involved
127
water
treatment
systems,
including
40
small
systems.
Large
and
medium
systems
collected
semi­
monthly
samples
for
Cryptosporidium,
Giardia,
and
other
water
quality
parameters
for
1
year.
Small
systems
(
those
serving
fewer
than
10,000
people)
collected
monthly
water
quality
data,
but
did
not
sample
for
protozoa.

EPA,
in
consultation
with
nationally
recognized
experts
in
statistics,
evaluated
ICR
and
ICRSS
data
to
generate
estimates
of
national
occurrence
of
Cryptosporidium
in
surface
water
and
finished
water.
These
data
were
evaluated
under
various
regulatory
scenarios
to
estimate
the
reduction
of
Cryptosporidium
reaching
consumers.

3.3
Specific
Regulatory
Alternatives
Considered
in
this
EA
The
recommendations
of
the
Advisory
Committee
are
described
in
a
document
called
the
Agreement
in
Principle
(
USEPA
2000g).
The
Advisory
Committee
reached
consensus
on
the
issues
of
uncovered
finished
water
reservoirs
and
treatment
of
unfiltered
water
without
formally
identifying
regulatory
alternatives
other
than
the
proposed
approaches.
Consequently,
no
formal
alternatives
were
presented
for
these
requirements.
The
committee's
recommendations
to
address
these
issues
are
reflected
in
the
proposed
LT2ESWTR.
For
control
of
Cryptosporidium
in
filtered
systems,
however,
several
alternatives
were
considered.
The
committee
discussed,
but
quickly
found
impractical,
alternatives
based
on
monitoring
for
Cryptosporidium
in
finished
water.
The
occurrence
of
Cryptosporidium
in
finished
water
is
so
low
that
the
volume
of
water
required
for
analysis
would
make
monitoring
costs
prohibitive.
Thus,
all
the
alternatives
based
on
monitoring
directed
that
monitoring
be
performed
on
the
source
water.

The
following
subsections
detail
the
differences
between
the
alternatives
and
explain
the
rationale
behind
EPA's
selection
of
the
Preferred
Alternative.
Section
3.3.2
also
describes
the
additional
treatment
requirements
proposed
for
direct
filtration
systems.

3.3.1
Summary
of
Bin
Classification
and
Treatment
Requirements
for
Regulatory
Alternatives
In
considering
different
approaches
for
filtered
systems
under
the
LT2ESWTR,
the
M­
DBP
Advisory
Committee
focused
on
four
regulatory
alternatives
(
hereafter
referred
to
as
Alternatives
A1
through
A4).
Alternative
A1
requires
the
same
amount
of
reduction
of
Cryptosporidium
for
all
systems,
while
the
other
three
base
their
treatment
requirements
on
the
amount
of
Cryptosporidium
found
in
a
system's
source
water
through
monitoring.
These
measurements
place
a
system
in
one
of
several
"
bins,"
ranging
from
"
no
action"
to
an
additional
2.5
log
crypto
treatment.
Further,
all
alternatives
allow
systems
to
select
treatment
technologies
based
on
the
amount
of
crypto
treatment
needed
to
meet
requirements
and
the
effectiveness
of
each
technology.

In
evaluating
each
binning
scenario,
the
committee
asked
the
following
questions:
Economic
Analysis
for
the
LT2ESWTR
Proposal
3­
3
June
2003
°
Do
the
treatment
requirements
adequately
reduce
Cryptosporidium
concentrations
in
finished
water?

°
How
many
systems
would
be
required
to
add
treatment?

°
What
is
the
likelihood
of
bin
misclassification?

°
What
are
the
chances
that
systems
with
high
source
water
concentrations
would
be
placed
in
the
bin
requiring
no
action?

The
predicted
finished
water
Cryptosporidium
concentrations
and
percentages
of
plants
adding
treatment
are
shown
for
the
Preferred
Regulatory
Alternative
in
Chapter
4.
The
likelihood
of
classification
in
a
certain
bin
for
a
given
concentration
and
the
predicted
percentages
of
plants
in
each
bin
for
all
the
regulatory
alternatives
are
shown
in
Appendix
B.

Exhibit
3.1
summarizes
binning
and
treatment
scenarios
for
each
specific
regulatory
alternative.
These
alternatives
were
defined
by
two
criteria:
(
1)
bin
boundaries,
as
defined
by
the
results
of
Cryptosporidium
monitoring,
and
(
2)
the
treatment
scenarios
(
log
reduction
requirements
for
Cryptosporidium)
required
for
each
bin.
The
alternatives
were
compared
within
the
context
of
the
economic
analysis
to
assist
EPA
in
selecting
a
Preferred
Regulatory
Alternative.

On
the
basis
of
preliminary
cost­
benefit
analyses,
the
Advisory
Committee
chose
Alternative
A3
as
the
Preferred
Alternative.
This
EA
continues
to
support
Alternative
A3
as
the
best
choice
and
EPA
is
proposing
Alternative
A3
for
this
reason.
Alternative
A3
was
shown
to
be
the
most
cost­
effective
and
the
best
value.
Comparisons
of
the
net
benefits
of
each
alternative
are
summarized
in
the
executive
summary
and
described
in
more
detail
in
Chapter
8.
Economic
Analysis
for
the
LT2ESWTR
Proposal
3­
4
June
2003
Exhibit
3.1
Summary
of
Bin
Requirements
for
Filtered
Systems
Source
Water
Cryptosporidium
Monitoring
Results
(
oocysts/
L)
Additional
Treatment
Requirements
Alternative
A1
2.0
log
inactivation
required
for
all
systems
Alternative
A2
<
0.03
No
action
$
0.03
and
<
0.1
0.5
log
$
0.1
and
<
1.0
1.5
log
$
1.0
2.5
log
Alternative
A3
­
Preferred
Alternative
<
0.075
No
action
$
0.075
and
<
1.0
1
log
$
1.0
and
<
3.0
2
log
$
3.0
2.5
log
Alternative
A4
<
0.1
No
action
$
0.1
and
<
1.0
0.5­
log
$
1.0
1.0
log
Note:
"
Additional
treatment
requirements"
are
for
systems
with
conventional
treatment
in
full
compliance
with
existing
rules
(
the
IESWTR
and
LT1ESWTR).

3.3.2
Additional
Treatment
for
Direct
Filtration
Systems
The
Agreement
in
Principle
(
USEPA
2000g)
recommended
that
EPA
address
direct
filtration
systems
in
connection
with
Bins
2­
4
of
Alternative
A3
in
the
proposed
LT2ESWTR.
Direct
filtration
plants
lack
sedimentation
basins;
their
treatment
processes
move
directly
from
addition
of
coagulant
and
mixing
to
filtration.
Conventional
filtration
plants
use
coagulation,
sedimentation,
and
filtration.
Sedimentation
reduces
the
Cryptosporidium
load
on
the
filters
and
helps
plants
respond
to
sudden
changes
in
influent
water
quality.

EPA
considered
the
effectiveness
of
direct
filtration
in
removing
Cryptosporidium
when
determining
how
to
apply
the
Advisory
Committee's
treatment
technique
recommendations
for
conventional
plants
to
direct
filtration
plants.
EPA
has
consistently
recognized
the
value
of
employing
multiple
barriers
for
pathogen
removal
to
provide
redundancy
and
reliability.
Studies
have
shown
that
a
well­
operated
sedimentation
basin
can
reduce
Cryptosporidium
levels
by
0.5
log
or
greater
(
Dugan
et
al.
1999;
Edzwald
and
Kelley
1998;
and
Patania
et
al.
1995).
The
SWTR
Guidance
Manual
(
USEPA
1991)
also
supports
giving
less
credit
to
direct
filtration
systems;
these
systems
are
eligible
for
0.5
log
less
credit
for
Giardia
than
conventional
filtration
systems.
Economic
Analysis
for
the
LT2ESWTR
Proposal
3­
5
June
2003
Based
on
these
studies,
EPA's
prior
consideration
of
the
effectiveness
of
direct
filtration
systems
for
Giardia,
and
the
Agency's
belief
in
the
multiple
barrier
approach,
EPA
concluded
that
direct
filtration
plants
should
provide
an
additional
0.5
log
of
treatment
beyond
that
required
for
conventional
treatment
plants.
A
more
detailed
discussion
of
Cryptosporidium
removal
by
conventional
and
direct
filtration
can
be
found
in
the
EPA
document,
Occurrence
and
Exposure
Assessment
for
the
Long
Term
2
Enhanced
Surface
Water
Treatment
Rule
(
USEPA
2003c).

3.4
Alternative
Monitoring
Approaches
Considered
EPA
considered
a
variety
of
monitoring
approaches
while
developing
LT2ESWTR
regulatory
alternatives.
These
include
evaluating
other
water
quality
parameters
as
surrogates
for
Cryptosporidium
and
alternative
monitoring
strategies
to
minimize
monitoring
costs,
especially
for
small
drinking
water
systems.
These
issues
are
described
in
the
subsections
below.

3.4.1
Indicators
of
Microbial
Contamination
Due
to
the
cost
associated
with
Cryptosporidium
monitoring,
the
Stage
2
M­
DBP
Advisory
Committee
evaluated
alternative
source
water
quality
parameters
to
determine
if
they
could
be
used
to
identify
water
sources
with
high
Cryptosporidium
levels.
The
committee
assessed
the
12­
month
means
of
monitoring
data
for
turbidity,
total
organic
carbon
(
TOC),
E.
coli,
fecal
coliform,
and
total
coliform
bacteria
as
surrogates
for
Cryptosporidium.
Specifically,
the
committee
evaluated
whether
these
potential
surrogates
could
accurately
assign
plants
to
LT2ESWTR
microbial
framework
bins.
None
of
these
parameters
correlated
well
with
Cryptosporidium
levels
at
all
concentrations.
Evidence
indicated
that
E.
coli
would
be
somewhat
effective
in
identifying
plants
with
Cryptosporidium
levels
above
or
below
0.075
oocysts/
L
in
reservoirs,
lakes,
and
flowing
streams
(
USEPA
2003c).
Under
the
Preferred
Alternative,
this
level
is
the
cutoff
between
the
no­
action
and
the
action
bins
(
Bins
2,
3,
and
4).
Thus,
the
committee
recommended
that
E.
coli
be
used
as
a
screening
test
for
small
systems
under
the
Preferred
Alternative,
A3.

The
selection
of
the
E.
coli
level
for
determining
when
additional
Cryptosporidium
monitoring
in
small
systems
should
be
conducted
was
based
on
limited
data.
Therefore,
the
Advisory
Committee
agreed
that
additional
data
should
be
collected
to
evaluate
the
E.
coli
indicator
criteria
and
to
develop
alternative
criteria,
if
appropriate.
Accordingly,
the
Advisory
Committee
agreed
that
large
systems
would
measure
E.
coli
and
turbidity
in
their
source
water
when
they
sample
for
Cryptosporidium.
The
composite
data
will
be
submitted
to
EPA.
This
will
give
EPA
time
to
develop
possible
alternative
indicator
levels
or
indicator
parameters
(
e.
g.,
turbidity
in
combination
with
E.
coli)
prior
to
the
date
when
small
systems
are
required
to
begin
source
water
monitoring
for
E.
coli.
Following
the
completion
of
1
year
of
monitoring
under
the
LT2ESWTR,
EPA
will
determine
if
alternative
indicators
(
to
the
E.
coli
levels
prescribed
in
the
rule)
are
appropriate
for
determining
classification
into
bins.
Depending
upon
its
findings,
EPA
will
issue
guidance
for
States
to
consider
prescribing
alternative
indicator
requirements
for
small
systems.
Therefore,
the
LT2ESWTR
allows
for
alternative
indicators
to
be
considered
by
the
Primacy
Agency.

The
use
of
E.
coli
as
a
screen
for
Cryptosporidium
monitoring
is
applicable
only
to
Alternatives
A3
and
A4.
Using
E.
coli
levels
to
predict
a
mean
Cryptosporidium
concentration
of
less
than
0.03
oocysts
per
liter
 
the
action
cutoff
level
for
Alternative
A2
 
was
less
reliable
(
USEPA
2003c).
Thus,
Economic
Analysis
for
the
LT2ESWTR
Proposal
3­
6
June
2003
under
Alternative
A2,
no
screening
test
is
available,
and
all
small
systems
must
monitor
for
Cryptosporidium.
Since
Alternative
A1
requires
the
same
level
of
treatment
for
all
systems,
no
monitoring
provisions
for
large
or
small
systems
are
included
for
Alternative
A1.

3.4.2
Cryptosporidium
Monitoring
Strategies
for
Bin
Classification
EPA
and
the
Advisory
Committee
also
evaluated
alternative
monitoring
strategies
to
ensure
that
levels
of
source
water
contamination
would
be
adequately
characterized,
while
minimizing
the
monitoring
burden.
Approaches
considered
included
taking
24,
48,
or
72
source
water
samples
to
determine
bin
classification
using
the
bin
boundaries
in
Alternative
A3.

EPA
chose
to
allow
systems
serving
at
least
10,000
people
to
collect
and
analyze
24
monthly
samples
over
a
2­
year
period
and
base
the
bin
assignment
on
the
maximum
running
annual
average
(
RAA).
(
The
first
RAA
will
be
the
average
of
the
results
of
the
first
12
months
of
monitoring;
the
second
RAA
will
be
the
average
of
results
from
months
2
 
13,
the
third
will
be
the
average
of
concentrations
from
months
3
 
14,
etc.)
Alternatively,
systems
may
collect
two
or
more
samples
per
month
over
the
2­
year
period,
at
regular
intervals,
and
use
the
simple
average
(
the
average
of
all
48
samples)
to
determine
bin
placement.

The
following
paragraphs
discuss
the
methodology
for
choosing
these
monitoring
frequencies.

EPA
knew
that
the
measured
amount
of
Cryptosporidium
in
each
sample
might
be
different
from
the
actual
or
"
true"
concentration
because
of
error
inherent
in
sampling
and
analytical
methods.
For
example,
method
error
can
be
introduced
by
oocyst
recovery,
false
detections,
and
analyst
error.
Sampling
error
is
affected
by
the
sample
size
and
the
fact
that
the
concentration
in
a
given
sample
may
misrepresent
the
concentration
in
the
larger
water
body.
EPA
was
primarily
concerned
that
highoccurrence
sources
could
possibly
be
placed
in
either
the
no­
action
bin
(
Bin
1,
for
mean
occurrence
below
0.075
oocysts
per
liter)
or
bins
that
provided
insufficient
remedies.
This
could
result
in
insufficient
protection
of
public
health.
A
secondary
concern
was
that
systems
could
be
assigned
to
a
higher
bin
than
was
warranted
by
their
true
concentration,
resulting
in
unnecessary
costs
for
systems.

EPA
performed
a
Monte
Carlo
analysis
to
determine
the
probabilities
of
misclassification
based
on
different
monitoring
scenarios
(
see
Appendix
B
for
more
details).
The
analysis
accounted
for
the
volume
assayed,
variation
in
source
water
Cryptosporidium
occurrence,
and
variable
method
recovery.

The
analysis
specifically
considered
the
likelihood
that
a
system
with
a
true
mean
Cryptosporidium
concentration
a
factor
of
3.16
(
0.5
log)
above
or
below
a
bin
boundary
would
be
assigned
to
the
wrong
bin.
Probabilities
were
assessed
for
two
cases:

°
Misclassification
low:
a
system
with
a
mean
concentration
of
0.24
oocysts/
L
(
i.
e.,
factor
of
3.16
above
the
Bin
1
boundary
of
0.075
oocysts/
L)
is
misclassified
low
in
Bin
1.

°
Misclassification
high:
a
system
with
a
mean
concentration
of
0.024
oocysts/
L
(
i.
e.,
factor
of
3.16
below
the
Bin
1
boundary
of
0.075
oocysts/
L)
is
misclassified
high
in
Bin
2.
Economic
Analysis
for
the
LT2ESWTR
Proposal
3­
7
June
2003
Exhibit
3.2
shows
the
error
rates
the
model
predicts
at
concentrations
of
0.24
oocysts/
L
and
0.024
oocysts/
L
(
a
factor
of
3.16
above
and
below
the
Bin
1
boundary
of
0.075
oocysts/
L)
under
different
monitoring
scenarios.

Exhibit
3.2
Probability
of
Misclassification
for
Monitoring
and
Binning
Strategies
Considered
for
the
LT2ESWTR
Monitoring
Strategy
Probability
of
Misclassification
High
Probability
of
Misclassification
Low
48­
sample
simple
mean
1.7%
1.4%

24­
sample
maximum
RAA
5.3%
1.7%

24­
sample
simple
mean
2.8%
6.2%

12­
sample
second
highest
value
47%
1.1%

8­
sample
maximum
value
66%
1.0%

Note:
Probability
of
misclassification
high
into
Bin
2
was
calculated
for
systems
with
true
Cryptosporidium
concentrations
of
0.024
oocysts/
L,
or
0.5
log
below
the
Bin
1
boundary
of
0.075
oocysts/
L.
Probability
of
misclassification
low
into
Bin
1
was
calculated
for
systems
with
Cryptosporidium
concentrations
of
0.24
oocysts/
L,
or
0.5
log
above
the
Bin
1
boundary.

Source:
Appendix
B.

The
first
two
of
the
approaches
shown
in
Exhibit
3.2,
the
48­
sample
simple
mean
and
24­
sample
maximum
RAA,
were
recommended
by
the
Advisory
Committee
and
are
proposed
for
bin
classification
under
the
LT2ESWTR
because
they
have
low
misclassification
rates.
As
shown
in
Exhibit
3.2,
these
strategies
have
misclassification
low
rates
of
1
to
2
percent,
meaning
there
is
a
98
to
99
percent
likelihood
that
a
plant
with
an
oocyst
concentration
0.5
log
above
the
Bin
1
boundary
would
be
correctly
assigned
to
Bin
2.
The
misclassification
high
rate
is
near
2
percent
for
the
48­
sample
simple
mean
and
5
percent
for
the
24­
sample
maximum
RAA.
These
rates
indicate
that
a
plant
with
an
oocyst
concentration
0.5
log
below
the
Bin
1
boundary
of
would
have
a
95
to
98
percent
probability
of
being
correctly
assigned
to
Bin
1.
Bin
misclassification
rates
across
a
wide
range
of
concentrations
are
shown
in
Appendix
B.

The
24­
sample
simple
mean
had
a
slightly
lower
misclassification
high
rate
than
the
24­
sample
maximum
RAA
(
2.8
vs.
5.3
percent)
but
the
misclassification
low
rate
of
the
simple
mean
was
almost
4
times
greater.
Consequently,
a
plant
with
a
mean
Cryptosporidium
level
above
the
Bin
1
boundary
would
be
much
more
likely
to
be
misclassified
in
Bin
1
using
a
24­
sample
simple
mean
than
with
a
24­
sample
maximum
RAA.
To
increase
the
probability
that
systems
with
mean
Cryptosporidium
concentrations
above
0.075
oocysts/
L
will
provide
additional
treatment,
EPA
is
proposing
that
if
only
24
samples
are
taken,
the
maximum
RAA
be
used
to
determine
bin
assignment.

EPA
evaluated
monitoring
strategies
involving
only
12
and
8
samples
to
determine
if
lower
frequency
monitoring
could
provide
satisfactory
bin
classification;
these
lower
numbers
of
samples
are
not
adequate.
For
example,
Exhibit
3.2
shows
that
if
plants
were
classified
in
bins
based
on
the
second
highest
concentration
of
12
samples
or
the
highest
concentration
of
eight
samples,
then
low
misclassification
low
rates
could
be
achieved.
A
system
with
a
mean
Cryptosporidium
level
0.5
log
above
the
Bin
1
boundary
would
have
a
99
percent
chance
of
being
appropriately
classified
in
a
bin
requiring
additional
treatment
under
either
strategy.
However,
a
system
with
a
mean
oocyst
Economic
Analysis
for
the
LT2ESWTR
Proposal
3­
8
June
2003
concentration
0.5
log
below
the
Bin
1
boundary
would
have
a
47
percent
chance
of
being
incorrectly
classified
in
Bin
2
using
the
second
highest
result
among
12
samples,
or
a
66
percent
likelihood
of
being
misclassified
in
Bin
2
using
the
maximum
result
among
8
samples.
Therefore,
these
strategies
were
not
proposed.
Increasing
the
number
of
samples
used
to
compute
the
maximum
RAA
above
24
also
increased
the
number
of
annual
averages
computed,
so
it
did
not
reduce
the
likelihood
of
misclassification
high.
Computing
a
simple
mean
based
on
more
than
48
samples
did
reduce
bin
misclassification
rates,
but
the
rates
were
already
very
small
(
1
to
2
percent
for
plants
with
levels
0.5
log
above
or
below
bin
boundaries).
For
sources
with
Cryptosporidium
concentrations
very
near
or
at
bin
boundaries,
increasing
the
number
of
samples
did
not
markedly
improve
the
error
rates,
which
remained
near
50
percent
at
the
bin
boundaries.

In
summary,
EPA
believes
that
the
prescribed
sampling
designs
in
the
Preferred
Alternative
perform
well
for
the
purpose
of
assigning
source
waters
to
bins.
More
costly
designs,
involving
more
frequent
sampling
and
analysis,
provide
only
marginally
improved
performance,
while
placing
a
greater
burden
on
limited
laboratory
capacity.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
1
4.
Baseline
Conditions
4.1
Introduction
To
estimate
the
impact
of
the
LT2ESWTR
regulatory
alternatives
on
the
water
supply
industry,
it
is
necessary
to
first
establish
the
conditions
that
exist
within
the
industry
just
before
the
regulatory
requirements
take
effect.
This
baseline
allows
a
consistent
comparison
of
public
health
impacts
(
developed
in
Chapter
5)
and
the
economic
and
financial
impacts
(
developed
in
Chapters
6
and
7)
across
regulatory
alternatives.

Because
the
compliance
deadlines
of
recently
promulgated
and
proposed
rules
will
occur
after
the
date
EPA
is
required
to
promulgate
the
LT2ESWTR,
many
of
the
baseline
conditions
must
be
estimated
rather
than
directly
measured.
Thus,
data
on
existing
conditions
are
combined
with
projections
of
changes
to
those
conditions
to
estimate
the
baseline
for
the
LT2ESWTR.
The
steps
required
to
determine
the
baseline
conditions
are
as
follows:

°
Compile
an
industry
profile
 
identify
and
collect
information
on
the
segments
of
the
water
supply
industry
subject
to
the
rule.

°
Characterize
influent
water
quality
 
summarize
the
relevant
characteristics
of
the
raw
water
treated
by
the
industry.

°
Characterize
treatment
for
other
rules
 
predict
what
the
industry
will
do
to
comply
with
the
provisions
of
rules
that
will
precede
the
LT2ESWTR
and
that
may
generate
changes
relevant
to
this
rule,
specifically
the
IESWTR,
LT1ESWTR,
Stage
1
DBPR,
and
Stage
2
DBPR.

°
Predict
occurrence
following
implementation
of
other
rules
 
estimate
what
the
treated
water
quality
will
be
after
the
rules
preceding
the
LT2ESWTR
are
implemented.

This
chapter
presents
an
analysis
that
is
at
a
level
of
detail
and
precision
appropriate
to
support
subsequent
analyses
and
regulatory
decisions
for
the
LT2ESWTR.
An
exhaustive
review
of
the
water
supply
industry,
source
waters,
or
industry
practices
was
not
needed
to
conduct
the
analysis.
The
remainder
of
this
chapter
is
organized
as
follows:

4.2
Data,
Tools,
and
Processes
Used
in
Baseline
Development
4.2.1
ICR
and
ICRSS
Observed
Data
4.2.2
ICR
and
ICRSS
Modeled
Data
and
Method
for
Predicting
Source
Water
Occurrence
4.2.3
Surface
Water
Analytical
Tool
(
SWAT)
4.3
Industry
Profile
4.3.1
Public
Water
System
Characterization
4.3.2
Systems,
Plants,
and
Population
Subject
to
the
LT2ESWTR
4.3.3
Water
Treatment
Plant
Design
and
Average
Daily
Flows
4.4
Baseline
for
Unfiltered
Plants
(
Pre­
LT2ESWTR)
4.4.1
Treatment
Characterization
for
Unfiltered
Plants
4.4.2
Number
of
Unfiltered
Systems,
Plants,
and
Population
Served
4.4.3
Source
Water
Cryptosporidium
Occurrence
for
Unfiltered
Plants
4.4.4
Finished
Water
Cryptosporidium
Occurrence
for
Unfiltered
Plants
1
Throughout
this
document,
the
acronym
"
SDWIS"
represents
"
SDWIS
 
Federal
Version."

2
All
industry
baseline
data
reflect
revisions
to
SDWIS
4th
Quarter
Year
2000
Freeze
to
account
for
reporting
errors
in
Massachusetts
and
Montana.

Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
2
4.5
Baselines
for
Filtered
Plants
(
Pre­
LT2ESWTR)
4.5.1
Treatment
Characterization
for
Filtered
Plants
4.5.2
Number
of
Filtered
Plants
and
Population
Served
4.5.3
Source
Water
Cryptosporidium
Occurrence
for
Filtered
Plants
4.5.4
Finished
Water
Cryptosporidium
Occurrence
for
Filtered
Plants
4.5.5
Comparison
of
EPA
Estimates
with
Aboytes
et
al.
(
2000)
4.5.6
Predicted
Bin
Classification
for
Filtered
Plants
4.6
Baseline
for
Uncovered
Finished
Water
Reservoirs
4.7
Households
Incurring
Costs
Due
to
the
LT2ESWTR
4.8
Summary
of
Uncertainties
in
Development
of
LT2ESWTR
Baselines
4.2
Data,
Tools,
and
Processes
Used
in
Baseline
Development
Several
data
sources
were
used
to
characterize
the
baseline
and
to
predict
changes
in
treatment
technologies
and
water
quality
for
different
regulatory
alternatives.
The
Safe
Drinking
Water
Information
System
 
Federal
Version
(
SDWIS1)
(
4th
Quarter
Freeze
Year
2000
data2)
is
used
to
create
system
and
population
baselines
(
USEPA
2000f).
SDWIS
is
EPA's
national
regulatory
compliance
database
for
the
drinking
water
program.
It
includes
information
on
the
nation's
170,000
public
water
systems
(
PWSs)
and
on
violations
of
drinking
water
regulations.
EPA's
web
site
provides
more
information
on
SDWIS
(
http://
www.
epa.
gov/
safewater/
sdwisfed/
sdwis.
htm).
A
second
key
source
of
data
used
to
develop
the
industry
profile
is
the
Third
Edition
of
the
Water
Industry
Baseline
Handbook
(
Baseline
Handbook)
(
USEPA
2001c)
published
in
May
2001,
which
compiles
data
derived
from
the
1995
Community
Water
System
Survey
(
CWSS)
and
SDWIS.
The
1995
CWSS
was
a
mail
survey
that
covered
ground
and
surface
water
systems
of
all
sizes
(
based
on
population
served).
The
survey
was
based
on
a
two­
phase,
stratified
sample
design.
Phase
1
was
a
telephone
screening
survey
that
provided
a
sampling
frame
for
the
main
data
collection
in
Phase
2.
The
survey
sample
in
Phase
2
was
stratified
according
to
water
system
size
(
residential
population
served),
ownership
(
public,
private,
or
ancillary),
and
primary
water
source
(
ground
or
surface).
A
total
of
3,681
systems
covering
a
range
of
source
water
types
and
system
sizes
were
randomly
selected
to
receive
the
main
survey
questionnaire.
Of
these,
1,980
systems
responded.
See
Community
Water
System
Survey,
Volume
2
(
USEPA
1997c),
for
more
information
on
the
1995
CWSS
sample
design
and
data
evaluation.

EPA
also
used
Geometries
and
Characteristics
of
Water
Systems
(
also
called
the
Model
Systems
Report)
(
USEPA
2000b).
In
this
document,
EPA
used
CWSS
data
to
develop
equations
to
predict
flows
based
on
system
populations.

To
characterize
the
influent
water
quality,
treatment
processes,
and
finished
water
quality,
EPA
primarily
used
data
from
the
1996
Information
Collection
Rule
(
ICR),
for
which
Cryptosporidium
monitoring
requirements
applied
to
all
PWSs
serving
at
least
100,000
people
and
using
surface
water
or
ground
water
under
the
direct
influence
of
surface
water
(
GWUDI)
as
a
source.
The
purpose
of
the
ICR
was
to
collect
DBP
and
microbial
occurrence
and
treatment
information
to
help
evaluate
the
need
for
further
microbial
and
DBP
rules.
The
ICR
gathered
plant­
level
data
from
about
300
water
systems
over
3
Method
1622
was
used
for
the
first
4
months
of
data
collection,
at
which
time
Method
1623
replaced
Method
1622.
The
primary
difference
between
Method
1622
and
Method
1623
is
that
the
Method
1622
immunomagnetic
separation
(
IMS)
kit
includes
only
reagents
for
Cryptosporidium
purification,
whereas
Method
1623
uses
the
Giardia/
Cryptosporidium­
combination
kit,
which
includes
reagents
for
both
Cryptosporidium
and
Giardia
purification.

4
Forty
small
plants
also
were
included
in
the
survey,
but
they
did
not
monitor
protozoa
concentrations.

Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
3
18
months
(
July
1997­
December
1998).
These
data
characterize
the
source
water
occurrence
of
Cryptosporidium,
Giardia,
viruses,
and
indicators
of
microbial
contamination,
along
with
types
of
treatment
in
place.
The
data
used
in
this
economic
analysis
(
EA)
are
from
the
Auxiliary
1
(
AUX1)
database
(
USEPA
2000j).

The
Information
Collection
Rule
Supplemental
Surveys
were
voluntary
surveys,
for
which
40
medium
(
serving
10,000
 
99,999)
and
40
large
plants
(
serving
100,000
or
more)
collected
Cryptosporidium
and
Giardia
source
water
occurrence
data.
These
data
are
presented
in
the
Occurrence
and
Exposure
Assessment
for
the
Long
Term
2
Enhanced
Surface
Water
Treatment
Rule
(
USEPA
2003c).
Treatment
characterization
and
other
information
were
obtained
from
industry
and
State
sources.

Several
analytical
tools
(
models)
also
were
used
to
estimate
the
following:

°
Source
water
Cryptosporidium
occurrence
°
Treatment
changes
due
to
the
Stage
1
and
Stage
2
DBPRs
°
Number
of
plants
falling
into
treatment
bins,
based
on
predicted
source
water
occurrence
°
Finished
water
occurrence
under
Pre­
LT2ESWTR
and
LT2ESWTR
conditions
4.2.1
ICR
and
ICRSS
Observed
Data
Three
sets
of
monitoring
data
are
used
to
characterize
Cryptosporidium
source
water
occurrence:
the
ICR
data
set
and
the
ICRSS
data
sets
for
medium
and
large
systems.
Microbial
analyses
for
the
ICR
were
conducted
according
to
the
ICR
Method
(
USEPA
1996a).
The
ICRSSs
evaluated
source
water
for
Cryptosporidium,
Giardia,
and
coliforms
at
a
sample
of
medium
surface
water
systems
(
those
serving
10,000
to
99,999
people)
as
well
as
a
sample
of
large
surface
water
systems.
EPA
Methods
1622
(
USEPA
1999a)
and
16233
(
USEPA
1999b),
summarized
in
the
Occurrence
and
Exposure
Assessment,
were
used
for
Cryptosporidium
and
Giardia
analyses.
With
Methods
1622
and
1623,
the
volume
of
water
analyzed
was
on
the
average
larger
than
the
volume
analyzed
with
the
ICR
Method,
yielding
better
estimates
of
Cryptosporidium
occurrence
on
a
per­
sample
basis
(
the
volume
analyzed
with
the
ICR
Method
depended
on
the
sample
pellet
volume
after
centrifugation
and
was
based
on
the
volume
needed
to
meet
detection
limits).
The
ICRSS
data
consist
of
semi­
monthly
observations
taken
over
a
12­
month
period
at
40
randomly
selected
large
plants
and
40
randomly
selected
medium
plants4.
Exhibit
4.1
summarizes
the
differences
between
the
ICR
and
ICRSS
data
collection
methods.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
4
Exhibit
4.1
ICR
and
ICRSS
Comparison
ICR
ICRSSM
ICRSSL
System
size
of
plants
participating
(
population
served)
>
100,000
10,000­
99,999
>
100,000
Number
of
surface
water
plants
participating
350
40
40
Sample
frequency
Monthly
Semi­
monthly
Semi­
monthly
Sampling
period
July
1997­
December
1998
March
1999­
February
2000
March
1999­
February
2000
Required
sample
volume
100
L
10
L
10
L
Median
sample
volume
analyzed
3.2
L
10
L
10
L
Average
recovery
rates
for
lab
method
12%
43%

Percentage
of
samples
where
Cryptosporidium
was
not
detected
93%
83%
87%

Percentage
of
plants
with
at
least
one
positive
Cryptosporidium
sample
44%
85%
85%

Source:
USEPA
2003c,
USEPA
2000j,
and
USEPA
2000k.

The
ICR
monitoring
program
resulted
in
nearly
6,000
Cryptosporidium
measurements
from
350
water
sources
(
USEPA
2000j).
The
ICRSS
monitoring
produced
approximately
1,900
measurements
from
80
source
waters.
Cryptosporidium
was
not
detected
in
most
of
the
samples
(
93
percent
of
ICR
samples
and
86
percent
of
ICRSS
samples).
Approximately
44
percent
of
plants
participating
in
the
ICR
program
had
at
least
one
positive
sample,
while
the
increased
sensitivity
of
the
methods
under
the
ICRSS
led
to
a
much
higher
percentage
of
plants
(
approximately
85
percent)
having
at
least
one
positive
sample.

The
detection
of
Cryptosporidium
oocysts
is
complex.
Because
of
the
low
occurrence
of
Cryptosporidium
in
source
waters,
a
sample
may
not
contain
any
oocysts
even
though
the
source
water
does.
Thus,
a
non­
detection
in
a
test
volume
is
not
definitive
evidence
against
occurrence
in
the
source.
In
addition,
the
laboratory
method
is
inefficient
and
may
not
recover
all
the
oocysts
that
were
in
a
sample.
While
underestimation
is
much
more
likely,
when
detections
do
occur,
sample
concentrations
also
may
overestimate
influent
concentrations
because
of
the
small
volume
of
sample
involved
(
i.
e.,
one
oocyst
identified
in
a
10­
liter
sample
may
not
represent
the
true
proportion
of
oocysts
in
the
much
larger
source
water
volume).

During
the
ICR
collection
period,
EPA
implemented
the
ICR
Lab
Spiking
Program
(
LSP)
to
assess
the
recovery
of
Cryptosporidium
oocysts
from
field
samples
analyzed
with
the
ICR
Method.
At
the
time
of
the
ICR
sample
collection,
duplicate
100­
liter
samples
were
collected
and
spiked
with
a
known
quantity
of
Cryptosporidium
oocysts.
Recovery
of
the
oocysts
(
the
detection
of
known
oocysts)
by
laboratories
was
very
low,
with
an
average
of
12
percent
of
the
known
quantity
being
recovered
per
sample.
The
ICRSS
Matrix
Spike
Program
was
used
to
assess
the
recovery
of
oocysts
from
field
samples
using
Methods
1622/
1623
during
the
ICRSS.
Duplicate
samples
were
spiked
with
a
known
quantity
of
Cryptosporidium
oocysts,
filtered,
and
analyzed
using
Methods
1622/
1623.
The
average
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
5
recovery
for
the
ICRSS
was
43
percent.
These
spiking
programs
are
further
described
in
the
LT2ESWTR
Occurrence
and
Exposure
Assessment
(
USEPA
2003c).

Each
set
of
data
has
its
advantages;
for
instance,
the
ICR
data
set
contains
data
from
more
plants
than
the
ICRSS
data
sets
do,
but
the
ICR
data
set
does
not
include
data
for
plants
in
medium­
sized
systems.
The
ICRSS
data
set
does
not
include
enough
data
for
unfiltered
plants
to
be
useful
for
modeling,
while
the
ICR
set
does.
The
ICRSS
sets,
on
average,
had
higher
recovery
rates
and
larger
sample
volumes
than
the
ICR
set.
The
ICR
collected
data
for
a
longer
time
period,
but
the
ICRSS
data
were
collected
more
frequently.
In
considering
which
data
set
best
represents
the
national
distribution
of
Cryptosporidium
occurrence,
none
was
judged
superior
to
the
others;
that
is,
no
one
set
was
considered
to
have
a
greater
likelihood
of
representing
"
true"
occurrence.
In
view
of
this,
each
data
set
was
kept
separate,
and
a
weighting
of
their
relative
values
to
allow
them
to
be
combined
was
not
attempted.

Cryptosporidium
observations
were
characterized
according
to
oocyst
structure
as
observed
under
a
microscope,
as
defined
below:

°
Empty:
Oocyst­
type
walls,
but
not
presently
containing
internal
material.

°
Non­
Empty:
Includes
oocysts
with
internal
structure
and
with
amorphous
structures.
Oocysts
with
amorphous
structures
have
walls
and
internal
material
characteristic
of
Cryptosporidium,
but
the
material
cannot
be
confirmed
as
Cryptosporidium.

°
With
Internal
Structures:
Oocysts
that
have
a
cell
wall
and
recognizable
internal
structures
consistent
with
Cryptosporidium;
subset
of
non­
empty
category.

°
Amorphous
Structures:
Oocysts
that
have
a
cell
wall
and
internal
material
but
no
recognizable
internal
structures;
subset
of
non­
empty
category.

°
Total:
The
combined
count
of
empty
oocysts
and
non­
empty
oocysts
(
those
with
either
internal
or
amorphous
structures).

At
meetings
of
the
Technical
Working
Group
of
the
Microbial­
Disinfection
Byproducts
Advisory
Committee,
participants
agreed
that
the
type
of
oocyst
observed
gives
information
about
the
level
of
confidence
that
the
oocyst
is
actually
Cryptosporidium.
The
presence
of
internal
structures
may
increase
the
confidence
that
the
observed
object
is
indeed
a
Cryptosporidium
oocyst,
and
not
some
other
item
or
organism
with
similar
gross
morphology.
While
oocysts
that
are
empty
are
unlikely
to
be
viable
or
infectious
at
the
time
of
the
laboratory
analysis,
they
still
are
indicators
of
the
presence
of
Cryptosporidium
Oocysts
with
amorphous
structures
give
a
level
of
confidence
between
that
of
empty
oocysts
and
those
with
internal
structures.
A
detailed
presentation
of
observed
Cryptosporidium
occurrence
and
evaluation
of
results
from
the
ICR
and
ICRSS
is
provided
in
the
Occurrence
and
Exposure
Assessment
(
USEPA
2003c).

The
analysis
presented
in
this
document
assumes
that
total
Cryptosporidium
counts
are
the
most
representative
of
the
presence
of
Cryptosporidium
in
source
waters.
While
some
of
these
oocysts
may
not
have
been
infectious
at
the
time
of
analysis,
they
likely
represent
Cryptosporidium
that
were
present
in
the
water.
The
probability
of
oocyst
infectivity
is
addressed
in
the
risk
assessment
model
in
section
5.2.3.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
6
To
account
for
the
limitations
in
observed
data
in
the
three
data
sets,
modeled
estimates
of
the
range
of
underlying
total
"
true"
Cryptosporidium
occurrence,
consistent
with
the
observed
occurrence,
were
developed.

4.2.2
ICR
and
ICRSS
Modeled
Data
and
Method
for
Estimating
Source
Water
Occurrence
An
accurate
representation
of
Cryptosporidium
concentrations
in
source
water
is
important
for
estimating
both
costs
and
benefits.
The
LT2ESWTR
treatment
requirements
are
based
on
the
results
of
source
water
Cryptosporidium
monitoring.
Therefore,
occurrence
estimates
determine
the
number
of
plants
requiring
treatment
and
the
level
of
treatment,
which
in
turn
determines
cost.
The
population
affected
and
the
estimated
reduction
in
illness
also
derive
from
those
same
occurrence
estimates.
This
section
provides
the
rationale
for
using
model­
based
estimates
over
observed
data
and
summarizes
the
modeled
results.

The
ICR
and
ICRSS
provide
source
water
quality
data
for
developing
this
estimated
national
occurrence
distribution.
The
raw
data
include
the
number
of
oocysts
detected
and
the
associated
sample
volume
analyzed.
A
straightforward
approach
to
modeling
is
to
first
divide
the
counts
by
the
volumes
to
obtain
concentrations
and
then
model
the
distribution
of
estimated
concentrations.
However,
there
are
limitations
in
the
data
that
negate
the
usefulness
of
this
approach.

First,
sample
volumes
are
low
relative
to
the
volume
needed
to
consistently
calculate
representative
concentrations
of
Cryptosporidium
in
source
water.
The
volumes
collected
also
are
inconsistent
from
location
to
location
or
even
month
to
month
at
a
given
location.
For
example,
the
majority
of
sample
counts
in
the
ICR
and
ICRSS
were
zeros
(
no
Cryptosporidium
detected),
but
these
zero
counts
are
based
on
widely
varying
sample
volumes.
Common
sense
suggests
that
a
zero
count
from
a
10­
liter
sample
should
be
weighted
more
heavily
than
a
zero
count
from
a
1­
liter
sample.
Based
on
a
straightforward
sample
concentration
calculation,
however,
both
concentrations
would
be
considered
the
same.

Second,
the
majority
of
Cryptosporidium
oocysts
captured
in
samples
likely
were
not
detected
in
testing,
due
to
a
lack
of
precision
in
analytical
methods.
Therefore,
straightforward
concentration
estimates
would
systematically
under­
estimate
the
true
concentration
of
Cryptosporidium
in
national
source
waters.

Finally,
there
are
limitations
in
the
counts
or
concentrations
that
can
be
reported
due
to
the
rare
occurrence
of
oocysts.
Oocyst
counts
can
only
be
whole
numbers
 
it
is
impossible
to
detect
half
an
oocyst
 
and
most
counts
were
zeros
or
ones.
Since
the
volume
analyzed
in
each
sample
was
generally
small,
a
limited
number
of
concentrations
could
be
calculated
from
the
count
and
volume
data.
For
instance,
one
oocyst
in
10
liters
gives
a
concentration
of
0.1
oocysts/
L,
while
one
oocyst
in
5
liters
gives
a
concentration
of
0.2
oocysts/
L.
The
volume
analyzed
would
have
to
be
quite
large
or
the
number
of
oocysts
present
greater
to
enable
more
precise
calculations
of
concentration.

These
limitations,
when
the
observed
data
are
used
to
calculate
concentrations,
result
in
a
large
number
of
individual
sample
values
that
may
each
over­
or
under­
represent
the
true
Cryptosporidium
concentrations.
To
account
for
these
limitations
and
other
sources
of
variability
in
the
data
and
to
be
able
to
estimate
national
occurrence,
model­
based
occurrence
estimates
are
chosen
over
observed
data.
A
more
detailed
discussion
of
this
estimation
procedure
is
provided
in
the
Occurrence
and
Exposure
Assessment
(
USEPA
2003c).
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
7
The
benefits
of
using
model­
based
estimates
are
that
a
model
can
properly
account
for
the
following
conditions:

°
Variability
in
the
data,
based
on
location,
sampling
technique,
turbidity
dependence,
and
other
factors
°
The
low
and
variable
recovery
rates
of
the
measurement
methods
°
The
small
volumes
assayed
and
their
adequacy
in
representing
the
much
larger
volume
of
source
water
at
the
time
of
sampling
°
The
small
number
of
samples
assayed
at
each
location
and
their
ability
to
represent
the
average
concentration
in
that
location's
water
during
the
18
months
of
the
survey
EPA
developed
a
probability
model
that
links
the
survey
data
(
sample
volumes
and
laboratory
oocyst
counts)
to
the
unknown
source
water
concentrations,
the
quantities
of
interest.
The
model
accomplishes
two
tasks.
First,
it
adjusts
concentration
estimates
to
account
for
varying
sample
volumes
and
test
method
recovery
rates.
Second,
the
assumed
probability
structure
makes
it
possible
to
quantify
uncertainty
in
the
concentration
estimates.

To
account
for
varying
sample
volumes
and
recovery
rates,
the
model
defines
an
expected
count
for
each
survey
sample.
This
is
the
average
number
of
oocysts
detected
in
repeated
sampling
from
a
given
survey
location,
assuming
a
particular
source
water
concentration,
sample
volume,
and
test
method
recovery
rate:

Average
Count
=
Concentration
×
Volume
×
Recovery
or,
in
terms
of
units:

oocysts
detected
=
(
oocysts
present/
liter)
×
(
liters)
×
(
oocysts
detected/
oocysts
present)

Concentration
is
the
unknown,
and
estimating
it
from
data
is
the
goal
of
the
modeling.
Count
and
Volume
are
known.
They
are
measured
directly
and
reported
for
each
survey
sample.
Recovery
is
the
ratio
of
oocysts
detected
in
laboratory
testing
to
oocysts
present
in
the
sample.
Since
Recovery
cannot
be
measured
directly
for
an
individual
test,
there
are
no
sample­
by­
sample
data
available
for
it.
For
a
particular
test
method,
however,
the
typical
range
of
recovery
values
can
be
estimated
from
designed
experiments
using
"
spiked"
samples
with
known
concentrations
(
see
Chapter
3).
This
was
done
for
each
of
the
Cryptosporidium
lab
methods
used
in
the
ICR
and
ICRSS,
and
recovery
rate
probability
distributions
were
fit
to
the
results.
Simulated
Recovery
values
were
drawn
from
these
probability
distributions.

As
the
formula
shows,
the
average
oocyst
count
is
directly
related
to
source
water
concentration,
but
also
to
sample
volume
and
recovery
rate.
For
example,
increasing
the
sample
volume
will
result
in
a
higher
average
oocyst
count
even
if
the
source
water
concentration
is
fixed.
Because
Count,
Volume,
and
Recovery
are
known
or
can
be
simulated,
however,
fitting
the
data
to
this
formula
results
in
estimates
for
the
unknown
Concentration
that
account
for
the
variation
in
these
other
variables.

The
second
part
of
the
model
is
a
probability
structure
for
the
observed
sample
counts.
Each
observed
count
is
assumed
to
come
from
a
Poisson
probability
distribution,
a
fundamental
probability
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
8
model
for
counting
rare
events
(
this
is
the
distribution
that
results
from
the
fact
that
only
zero
or
one
oocyst
is
usually
present
in
a
sample).
The
mean
of
each
distribution
is
the
average
count
as
defined
above.
Incorporating
these
probability
distributions
in
the
model
allows
for
the
calculation
of
uncertainty
bounds
for
the
concentration
estimates.

A
Markov
Chain
Monte
Carlo
(
MCMC)
approach
was
employed
to
fit
this
model
to
the
data.
MCMC
is
an
iterative
technique
for
fitting
statistical
models
to
data.
The
Markov
Chain
is
a
sequence
of
joint
probability
distributions
that
converges
to
a
stable
distribution
for
likely
model
parameter
values.
Monte
Carlo
is
a
computational
technique
for
solving
intractable
integrals
through
extensive,
simulated
sampling.

For
each
primary
data
set
(
ICR,
ICRSSM,
and
ICRSSL),
rather
than
estimating
a
single
mean
and
standard
deviation
that
captures
the
log­
normal
distribution
of
plant­
mean
concentrations,
the
model
produces
distributions
of
possible
means
and
variances
from
which
one
can
sample.
(
EPA
believes
the
distribution
of
plant­
mean
concentrations
is
log­
normal;
this
is
described
in
section
3.3.3.2
of
the
Occurrence
and
Exposure
Assessment.)

The
results
of
the
occurrence
models
employed
in
this
EA
are
documented
in
sections
4.4.3
and
4.5.3
for
unfiltered
and
filtered
plants.
As
described
above,
there
was
no
single
occurrence
distribution
that
served
as
input
but,
instead,
a
collection
of
plausible
occurrence
distributions.
Summary
plots
in
these
sections,
then,
show
both
the
mean
occurrence
distribution
 
which
represents
the
"
middle"
distribution
 
and
confidence
bounds
that
capture
the
range
of
occurrence
distributions
in
a
given
collection.

4.2.3
Surface
Water
Analytical
Tool
(
SWAT)

SWAT
used
source
water
and
treatment
data
collected
under
the
ICR
to
estimate
the
percentage
of
large
surface
water
plants
that
would
require
advanced
technologies
to
meet
Stage
1
and
Stage
2
DBPR
limits
for
disinfection
byproducts
(
DBPs).
Several
advanced
technologies,
namely
chlorine
dioxide,
ozone,
ultraviolet
light
(
UV)
disinfection,
and
membrane
processes,
not
only
achieve
DBP
reduction,
but
also
provide
varying
degrees
of
Cryptosporidium
removal
or
inactivation.
The
characteristics,
costs,
and
effectiveness
of
these
technologies
are
taken
into
account
in
developing
the
baselines
for
this
EA.

A
more
detailed
description
of
SWAT
and
how
it
was
used
to
predict
changes
in
treatment
technologies
under
the
Stage
2
DBPR
is
provided
in
the
Draft
Economic
Analysis
for
the
Stage
2
Disinfectants
and
Disinfection
Byproducts
Rule
(
USEPA
2003d).
In
addition,
a
detailed
description
of
the
SWAT
components
and
its
operation
can
be
found
in
the
document
Surface
Water
Analytical
Tool
(
SWAT)
Version
1.1
 
Program
Descriptions
and
Assumptions
(
USEPA
2000c).

4.3
Industry
Profile
This
section
identifies
the
PWSs
subject
to
the
LT2ESWTR.
Subsequent
sections
identify
the
subset
of
systems
subject
to
specific
provisions
of
the
rule
(
e.
g.,
baselines
for
unfiltered
and
filtered
systems).
The
water
system
baseline
characterizations
presented
here
are
key
inputs
to
the
cost
and
benefit
assessments
described
in
this
EA.
5
Although
the
language
in
EPA
rules
generally
does
not
include
systems
serving
exactly
10,000
people
in
the
small
category
(
or
refers
to
small
systems
as
serving
fewer
than
10,000
people),
analyses
in
this
document
place
them
in
the
small
category
to
be
consistent
with
the
system
and
population
data
categories
from
SDWIS
and
the
Baseline
Handbook.

Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
9
EPA's
categorization
scheme
for
water
systems
is
first
summarized,
followed
by
a
presentation
of
systems,
plants,
and
populations
subject
to
the
LT2ESWTR.
A
summary
of
water
treatment
plant
design
flows
and
average
daily
flows
concludes
the
industry
profile
section.

4.3.1
Public
Water
System
Characterization
Categorizing
water
systems
allows
EPA
to
determine
the
impacts
of
this
rule
on
different
types
of
systems.
For
this
EA,
EPA
sorted
systems
on
the
basis
of
size,
ownership,
and
retail/
wholesale
relationships
provided
in
SDWIS.
This
section
explains
the
classifications
used.

PWS
Type
As
defined
by
the
Safe
Drinking
Water
Act
(
SDWA),
a
PWS
is
a
water
system
that
provides
piped
water
for
human
consumption
and
has
at
least
15
service
connections
or
regularly
serves
an
average
of
at
least
25
individuals
per
day
for
at
least
60
days
per
year.
EPA
classifies
PWSs
into
two
broad
groups:

°
Community
Water
Systems
(
CWSs)
have
at
least
15
service
connections
used
by
yearround
residents
or
regularly
serve
at
least
25
year­
round
residents.

°
Noncommunity
Water
Systems
(
NCWSs)
are
PWSs
that
are
not
classified
as
CWSs.

EPA
further
classifies
NCWSs
into
two
types:

°
Nontransient
Noncommunity
Water
Systems
(
NTNCWSs)
regularly
serve
at
least
25
of
the
same
people
more
than
6
months
per
year.

°
Transient
Noncommunity
Water
Systems
(
TNCWSs)
do
not
regularly
serve
at
least
25
of
the
same
people
more
than
6
months
per
year.

Population
Served
Under
the
LT2ESWTR,
as
with
some
previous
rules,
small
systems
(
those
serving
fewer
than
10,000
people,
as
defined
in
the
rule)
will
face
somewhat
different
requirements
than
larger
systems5.
System
size
is
important
in
the
regulatory
analysis
as
well.
Costs
are
estimated
using
the
size
of
systems
as
a
factor,
household
costs
are
derived
in
part
from
the
number
of
households
in
a
system
size
category,
and
separate
technology
decision
trees
are
used
for
different
sizes
of
systems.
6
In
the
first
part
of
the
cost
model,
cost
equations
are
specified
for
all
nine
subcategories.
Later
these
subcategories
are
combined
into
four
categories.

Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
10
In
some
of
the
analyses
that
follow,
nine
size
categories
are
used
(
these
are
based
on
SDWIS
categories):

C
Small
systems,
which
include
those
serving
10,000
or
fewer
people
and
are
broken
down
into
five
subcategories:
C
<
25
C
26
 
100
C
101
 
500
C
501
 
3,300
C
3,301
 
10,000
C
Medium
systems,
which
are
broken
down
into
two
subcategories:
C
10,001
 
50,000
C
50,001
 
100,000
C
Large
systems,
which
are
broken
down
into
two
subcategories:
C
100,001
 
1
million
C
>
1
million
In
other
analyses,
four
sizes
are
used
(
medium
and
large
categories
include
the
same
systems,
but
smalls
are
subdivided
into
very
small
(<
500)
and
small
(
501
 
10,000))
6.
To
calculate
net
benefits,
only
two
categories
are
used
(
small
and
large
systems
 
those
serving
<
10,000
people
and
those
serving
>
10,000
people).

Source
Water
Systems
are
classified
by
the
source
water
from
which
they
draw.
Surface
water
systems
typically
draw
from
reservoirs,
natural
lakes,
or
flowing
streams.
Ground
water
systems
draw
from
wells.
Some
ground
water
sources
are
under
the
direct
influence
of
surface
water
sources.
These
systems,
called
GWUDI
systems,
are
considered
directly
influenced
if
surface
water
microorganisms
are
present.
This
category
is
important
to
the
extent
that
pathogens,
such
as
Giardia
cysts
and
Cryptosporidium
oocysts,
can
contaminate
the
ground
water
source.
Some
systems
may
have
multiple
source
types
and
are
referred
to
as
"
mixed
systems."
In
SDWIS
and
the
Baseline
Handbook,
a
mixed
system
is
categorized
as
a
surface
water
system
if
it
gets
any
portion
of
its
flow
from
surface
water.
Based
on
the
analysis
in
the
Model
Systems
Report
(
USEPA
2000b),
it
is
estimated
that
21
percent
of
surface
water
systems
obtain
some
of
their
water
from
ground
water
sources.
Of
these
systems,
one­
third
(
8
percent
of
all
surface
water
systems)
get
the
majority
of
their
flow
from
ground
water.
Mixed
systems
may
either
be
systems
that
have
some
plants
that
are
solely
supplied
by
ground
water
and
other
plants
that
are
solely
supplied
by
surface
water,
or
they
may
have
one
or
more
plants
in
which
both
types
of
source
waters
are
mixed.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
11
Ownership
Systems
also
are
categorized
by
ownership.
Private
systems
are
owned
by
private
corporations,
associations,
or
individuals.
Private
systems
are
still
PWSs.
Public
systems
are
owned
by
public
entities,
such
as
a
municipality,
county,
or
special
district.
Ownership
distinctions
are
important
to
the
analysis
since
differences
exist
between
public
and
private
systems
in
their
access
to
capital
and
other
means
of
financing.
This
distinction
becomes
important
in
calculating
household
costs
in
Chapter
6
and
the
Unfunded
Mandates
analysis
presented
in
Chapter
7.

Purchased
and
Nonpurchased
Systems
Systems
are
categorized
according
to
whether
they
provide
or
treat
water
themselves
or
whether
they
purchase
it
from
other
systems.
Treatment
modifications
are
not
expected
to
be
made
by
purchased
systems
under
the
LT2ESWTR;
instead,
purchased
systems
will
absorb
(
through
rate
increases)
the
costs
of
additional
treatment
installed
by
the
sellers.
On
the
other
hand,
nonpurchased
systems
collect
and
treat
the
water
themselves
and
distribute
it
to
their
retail
and
wholesale
customers.
These
systems
are
subject
to
most
of
the
provisions
of
the
LT2ESWTR.

4.3.2
Systems,
Plants,
and
Population
Subject
to
the
LT2ESWTR
This
section
estimates
the
baseline
number
of
systems
subject
to
the
LT2ESWTR.
The
proposed
LT2ESWTR
applies
to
all
PWSs,
regardless
of
type
or
size,
that
use
surface
water
or
GWUDI,
since
a
person
may
need
only
ingest
one
Cryptosporidium
oocyst
to
become
infected.
However,
the
economic
impact
of
the
LT2ESWTR
may
vary
across
the
types
of
surface
water
systems
(
e.
g.,
some
systems
may
purchase
all
their
water
from
other
systems
and
will
not
be
subject
to
monitoring
or
treatment
requirements).

The
baseline
presented
in
this
section
is
used
to
estimate
implementation
costs,
such
as
those
for
training
and
becoming
familiar
with
the
rule
(
see
Appendix
D).
Not
all
of
the
systems
incurring
implementation
costs
will
incur
costs
for
other
provisions
of
the
rule.
Based
on
analysis
of
existing
treatment
technology,
size,
and
other
factors,
only
subsets
of
the
systems
presented
in
this
section
are
subject
to
specific
provisions
of
the
rule.
The
numbers
of
unfiltered
plants,
filtered
plants,
and
uncovered
finished
water
reservoirs
subject
to
specific
rule
provisions
are
presented
in
sections
4.4
through
4.6.

Number
of
Systems
Systems
in
SDWIS
are
listed
according
to
systems'
retail
populations.
The
advantage
of
classifying
them
this
way
is
that
it
appropriately
accounts
for
both
the
total
number
of
individual
PWSs
in
the
United
States
and
the
total
population
served
by
all
of
those
systems.
However,
a
disadvantage
of
this
method
(
especially
for
surface
water
CWSs)
when
estimating
national
costs
of
regulations
is
that
it
does
not
directly
account
for
the
fact
that
the
water
delivered
by
purchased
systems
to
their
retail
customers
is
actually
treated
by
other
systems.
It
is
important
to
recognize
that
the
total
flow
for
systems
supplying
surface
water
is
actually
treated
by
fewer
than
half
of
the
surface
water
systems
accounted
for
in
SDWIS.
Because
of
economies
of
scale,
the
cost
of
treatment
(
in
cents
per
gallon)
is
less
for
systems
treating
larger
flows
than
it
is
for
systems
treating
smaller
flows.
For
example,
it
is
typically
more
expensive
to
build
and
operate
two
treatment
plants
serving
5,000
people
than
one
treatment
plant
serving
10,000
people.
Failing
to
account
for
the
fact
that
surface
water
is
actually
treated
in
larger
quantities
at
a
smaller
number
of
systems
than
SDWIS
suggests
could
result
in
an
upward
bias
in
national
cost
estimates
of
rules
that
affect
a
substantial
portion
of
surface
water
systems.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
12
To
rectify
this
bias,
an
analysis
was
performed
to
"
link"
consecutive,
or
purchased,
surface
water
CWSs
and
NTNCWSs
to
their
respective
wholesale
systems
using
SDWIS
data.
CWSs
were
only
linked
to
other
CWSs
and
NTNCWSs
were
linked
only
to
other
NTNCWSs.
TNCWSs
were
not
linked,
since
they
usually
do
not
purchase
water
from
other
TNCWSs.
If
a
consecutive
system
could
be
linked
to
a
wholesaler,
that
system
was
removed
from
the
system
count
and
its
population
was
added
to
the
population
of
the
wholesale
system.
Consecutive
systems
that
could
not
be
definitively
linked
to
their
wholesalers
were
considered
stand­
alone
purchased
systems.
Although
consecutive
systems
do
not
treat
their
water,
these
systems
are
included
in
the
treatment
baseline
because
their
populations
and
flows
must
be
accounted
for
in
estimating
treatment
costs.
EPA
recognizes
that
including
them
as
separate
plants
overestimates
treatment
costs.
The
decision
process
for
this
analysis
is
shown
in
Exhibit
4.2.

The
number
of
surface
water
(
including
mixed
systems)
and
GWUDI
systems
per
size
category
in
SDWIS
(
pre­
linking)
is
shown
on
the
left
side
of
Exhibit
4.3.
These
numbers
are
modified
to
correct
obvious
population
errors,
such
as
excluding
populations
that
were
counted
by
both
purchasing
and
selling
systems.
Additionally,
systems
whose
ownership
category
was
listed
as
"
other"
in
SDWIS
were
reallocated
to
private
and
public
and
purchased
and
nonpurchased
categories
based
on
the
existing
proportion
of
each
category
to
the
total
number
of
systems.
Note
that
the
total
number
of
nonpurchased
systems
in
columns
H
and
I
is
the
same
as
the
total
number
of
nonpurchased
systems
before
linking
in
columns
C
and
D.
However,
the
numbers
of
public
and
private
nonpurchased
systems
changed
slightly
because
of
how
the
systems
with
"
other"
ownership
types
are
allocated
within
each
size
category.
The
inventory
of
nonpurchased
systems,
unlinked,
or
pre­
linking,
is
used
as
the
baseline
for
determining
implementation
costs.
Purchased
systems
are
not
included
in
this
baseline
because
they
do
not
have
their
own
source
water,
so
they
will
not
be
subject
to
monitoring
or
treatment
requirements
and
do
not
need
to
conduct
implementation
activities.
This
baseline
is
also
used
with
minor
modifications
to
determine
the
number
of
systems
(
and
plants)
subject
to
monitoring
costs.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
13
Exhibit
4.2
Methodology
for
"
Linking"
Consecutive
Surface
Water
CWSs
and
NTNCWSs
to
Their
Selling
Systems
°
If
a
system
has
multiple
sources,
(
e.
g.,
a
system
has
a
primary
source
of
surface
water
in
addition
to
a
purchased
surface
water
source),
it
is
assumed
to
be
adequately
represented
as
a
nonpurchased
surface
water
system,
and
was
not
linked
to
its
seller
(
i.
e.,
only
100­
percent
purchased
surface
water
systems
were
linked).

°
If
a
purchased
surface
water
system
(
System
P)
purchases
all
of
its
water
from
one
nonpurchased
surface
water
system
(
System
S),
its
population
was
added
to
that
of
System
S,
and
it
was
removed
from
the
inventory
of
purchased
systems.

°
If
the
purchased
surface
water
system
buys
water
from
multiple
nonpurchased
systems,
it
was
assigned
to
the
most
directly
related
nonpurchased
seller
with
the
largest
population.
For
example,
a
purchased
system
(
System
C)
purchases
from
a
nonpurchased
system
(
System
B1)
and
a
purchased
system
(
System
B2),
which
in
turn
purchases
from
a
nonpurchased
system
(
System
A).
In
this
case,
System
C
was
linked
to
System
B1,
and
the
population
of
System
C
was
added
to
that
of
System
B1.

°
When
the
purchased
system
and
its
seller
were
not
of
the
same
type
(
e.
g.,
a
CWS
purchasing
from
a
NTNCWS),
they
were
not
linked.
Systems
purchasing
from
sellers
of
a
different
system
type
were
counted
as
separate,
unlinked
purchased
systems.

°
If
the
PWS
identification
number
of
the
seller
did
not
correspond
to
an
active
water
system,
the
purchased
system
was
counted
as
a
separate,
unlinked,
purchased
system.

°
Some
purchased
systems
have
what
is
referred
to
as
"
cascading
provider
relationships."
For
instance,
a
purchased
system
(
System
C)
may
purchase
water
from
another
system
(
System
B).
This
system
(
System
B)
does
not
treat
its
own
water,
but
instead
purchases
water
from
another
system
(
System
A).
For
this
analysis,
the
populations
of
both
Systems
B
and
C
were
added
to
the
population
of
System
A,
and
Systems
B
and
C
were
removed
from
the
inventory
of
purchased
systems.

°
In
a
few
cases,
the
seller
could
not
be
found,
i.
e.,
a
purchased
system
(
e.
g.,
System
C)
could
not
be
linked
to
a
nonpurchased
system
(
e.
g.,
System
A).
These
purchased
systems
were
counted
as
separate,
unlinked,
purchased
systems.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
14
Number
of
Systems
Number
of
Systems
Purchased
Nonpurchased
Purchased
Nonpurchased
Public
Private
Public
Private
Public
Private
Public
Private
A
B
C
D
E
=
A+
B+
C+
D
F
G
H
I
J=
F+
G+
H+
I
K
L=
J*
K
M
N
<
100
445
308
222
308
1,283
18
14
135
321
488
1.0
488
28,999
0.02%
101­
500
850
513
426
331
2,120
47
22
396
338
803
1.1
883
224,777
0.13%
501­
1,000
614
243
372
84
1,313
20
9
344
86
459
1.0
459
353,113
0.20%
1,001­
3,300
1,199
181
977
110
2,467
51
5
904
106
1,066
1.0
1066
2,158,022
1.24%
3,301­
10,000
831
80
961
56
1,928
45
1
951
64
1,061
1.2
1273
6,343,982
3.65%
10,001­
50,000
672
59
880
79
1,690
32
6
984
78
1,100
1.1
1210
25,722,417
14.80%
50,001­
100,000
106
14
165
28
313
6
0
210
30
246
1.8
443
16,976,910
9.77%
100,001­
1
Million
60
11
176
29
276
8
0
215
33
256
1.6
410
70,801,601
40.74%
>
1Million
0
0
13
0
13
0
0
20
2
22
2.8
62
51,175,720
29.45%
Total
4,776
1,410
4,192
1,025
11,403
227
57
4,159
1,058
5,501
6,293
173,785,541
100%
Population
Served
Total
No.
of
Systems
Total
No.
of
Systems
System
Inventory
Before
Linking
Linked
System
Inventory­­
Treatment
Baseline
Plants
per
System
Total
Plants
Population
Served
Percent
of
Total
Population
Number
of
Systems
Number
of
Systems
Purchased
Nonpurchased
Purchased
Nonpurchased
Public
Private
Public
Private
Public
Private
Public
Private
A
B
C
D
E
=
A+
B+
C+
D
F
G
H
I
J=
F+
G+
H+
I
K
L=
J*
K
M
N
<
100
21
80
76
126
303
21
74
76
126
297
1.0
297
15,319
1.71%
101­
500
28
38
60
176
302
28
36
61
174
299
1.0
299
81,388
9.06%
501­
1,000
11
14
19
65
109
12
12
19
66
108
1.0
108
77,627
8.64%
1,001­
3,300
11
8
15
40
74
9
9
14
40
71
1.0
71
124,176
13.83%
3,301­
10,000
5
4
4
9
22
5
4
6
8
23
1.0
23
122,305
13.62%
10,001­
50,000
3
3
2
1
9
2
3
2
1
8
1.0
8
214,174
23.85%
50,001­
100,000
1
0
0
0
1
1
0
0
0
1
1.0
1
93,204
10.38%
100,001­
1
Million
1
0
0
0
1
1
0
0
0
1
1.0
1
169,846
18.91%
>
1Million
0
0
0
0
0
0
0
0
0
0
1.0
0
0
0.00%
Total
81
147
177
416
821
77
138
177
416
808
808
898,039
100%
System
Inventory
Before
Linking
Linked
System
Inventory­­
Treatment
Baseline
Population
Served
Plants
per
System
Total
Plants
Population
Served
Percent
of
Total
Population
Total
No.
of
Systems
Total
No.
of
Systems
Exhibit
4.3a
Number
of
Unlinked
and
Linked
Surface
Water
and
GWUDI
CWSs,
Number
of
Plants,
and
Population
Served,
by
Ownership
and
System
Size
Exhibit
4.3b
Number
of
Unlinked
and
Linked
Surface
Water
and
GWUDI
NTNCWSs,
Number
of
Plants,
and
Population
Served,
by
Ownership
and
System
Size
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
15
Number
of
Systems
Number
of
Systems
Purchased
Nonpurchased
Purchased
Nonpurchased
Public
Private
Public
Private
Public
Private
Public
Private
A
B
C
D
E
=
A+
B+
C+
D
F
G
H
I
J=
F+
G+
H+
I
K
L=
J*
K
M
N
<
100
88
415
172
559
1,234
88
415
172
559
1,234
1.0
1234
53,903
5.69%
101­
500
23
83
137
243
486
23
83
137
243
486
1.0
486
117,344
12.38%
501­
1,000
11
14
35
22
81
11
14
35
22
81
1.0
81
61,426
6.48%
1,001­
3,300
12
2
36
16
67
12
2
36
16
67
1.0
67
155,500
16.41%
3,301­
10,000
5
1
22
2
31
5
1
22
2
31
1.0
31
176,789
18.66%
10,001­
50,000
3
0
8
0
10
3
0
8
0
10
1.0
10
203,456
21.47%
50,001­
100,000
0
0
3
0
3
0
0
3
0
3
1.0
3
179,240
18.91%
100,001­
1
Million
0
0
0
0
0
0
0
0
0
0
1.0
0
0
0.00%
>
1Million
0
0
0
0
0
0
0
0
0
0
1.0
0
0
0.00%
Total
141
515
413
843
1,912
141
515
413
843
1,912
1,912
947,658
100%
System
Inventory
Before
Linking
Linked
System
Inventory­­
Treatment
Baseline
Population
Served
Plants
per
System
Total
Plants
Population
Served
Percent
of
Total
Population
Total
No.
of
Systems
Total
No.
of
Systems
Exhibit
4.3c
Number
of
Unlinked
and
Linked
Surface
Water
and
GWUDI
TNCWSs,
Number
of
Plants,
and
Population
Served,
by
Ownership
and
System
Size
Note:
For
TNCWSs,
"
population
served"
is
actually
the
population
served
at
a
given
time.
These
numbers
are
used
for
calculating
treatment
costs.
The
total
number
of
people
served
over
1
year
(
or
whichever
length
of
time
the
TNCWSs
is
in
operation)
is
generally
much
larger.
To
calculate
benefits,
EPA
adjusted
the
population
to
account
for
the
total
number
of
people
served
per
year
and
for
the
fact
that
each
customer
would
be
served
by
the
system
for
a
shorter
period
of
time
than
1
year.
These
adjustments
are
described
in
Chapter
5.

The
total
number
of
nonpurchased
systems
remains
the
same
before
and
after
linking;
however,
the
number
of
such
systems
in
a
size
category
may
change.
This
is
because
a
nonpurchased
system's
population
changes
if
a
purchased
system
is
linked
to
the
nonpurchased
system,
and
the
population
may
change
enough
to
move
the
system
to
the
next
size
category.

The
number
of
purchased
systems
left
after
the
linking
process
and
the
population
associated
with
these
systems
are
included
in
the
baseline
for
plants
subject
to
treatment
requirements
because
their
population
must
be
accounted
for
in
determining
treatment
costs
(
EPA
realizes
that
including
the
systems
themselves
does
result
in
an
over­
estimate
of
the
number
of
systems
requiring
treatment
under
LT2ESWTR).

Sources:
[
A]­[
D]
SDWIS
September
2000
data
adjusted
for
reporting
errors
in
Massachusetts
and
Montana
(
USEPA
2000f).
Also
includes
systems
not
categorized
as
"
public"
or
"
private"
in
SDWIS,
which
were
redistributed
among
public
and
private
purchased
and
nonpurchased
water
systems
according
to
the
proportions
of
systems
in
these
categories.
[
F]­[
I]
Data
from
Columns
A­
D
modified
using
linking
methodology
described
in
Exhibit
4.2,
except
for
TNCWSs,
which
were
not
linked.
[
K]
Derived
from
CWSS
data
(
USEPA
1997c)
and
Model
Systems
Report
(
USEPA
2000b),
modified
to
exclude
ground
water
plants
and
weighted
for
representativeness
of
each
system
to
all
CWSs.
[
M]
Includes
SDWIS
population
served
by
surface
water
and
GWUDI
systems
for
nonpurchased
and
purchased
systems
in
columns
F
 
I
based
on
SDWIS
(
USEPA
2000f).
Original
SDWIS
population
distribution
was
modified
using
linking
methodology
described
in
Exhibit
4.2,
except
for
TNCWSs,
which
were
not
linked.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
16
The
right
side
of
Exhibit
4.3
shows
the
baseline
number
of
systems,
after
the
linking
process,
that
are
subject
to
the
treatment
requirements
of
the
LT2ESWTR
(
actual
treatment
requirements
will
be
determined
by
results
of
Cryptosporidium
monitoring).
This
baseline
includes
both
purchased
and
nonpurchased
systems.
The
baseline
number
of
systems
was
converted
to
plants
as
described
below.

Number
of
Plants
Water
systems
can
consist
of
one
treatment
plant
supplying
all
of
the
water
received
by
the
population,
or
multiple
plants
treating
water,
possibly
from
different
sources,
or
even
different
source
types
such
as
surface
and
ground
water
sources.
Many
of
the
costs
in
Chapter
6
are
developed
by
estimating
the
costs
of
installing
additional
treatment
at
existing
water
treatment
plants.
Therefore,
for
this
baseline
analysis,
EPA
needed
to
calculate
the
number
of
plants
from
the
number
of
systems
reported
in
existing
databases.
To
determine
the
appropriate
number
of
plants
per
system,
EPA
used
the
1995
CWSS
data.
All
systems
reporting
some
surface
water
use
were
extracted
from
the
database.
Each
system's
plants
were
then
examined
to
determine
the
number
of
plants
that
would
be
subject
to
the
LT2ESWTR.
Any
plant
that
reported
treatment
of
any
amount
of
surface
water
was
included.
Plants
that
treated
solely
ground
water
were
eliminated.
The
CWSS
does
not
differentiate
between
ground
water
and
GWUDI,
so
GWUDI
plants
could
not
be
included
in
the
baseline.
Because
GWUDI
systems
are
generally
less
than
10
percent
of
the
total
number
of
systems
subject
to
the
rule,
it
is
assumed
that
excluding
them
will
introduce
only
a
small
error.

Using
the
number
of
surface
water
plants
and
the
number
of
systems,
a
weighted
analysis
was
performed
to
determine
the
average
number
of
surface
water
treatment
plants
per
system
for
each
size
category.
(
The
sampling
weights
were
developed
in
the
CWSS
to
account
for
the
representativeness
of
each
system
to
the
nationwide
population
of
systems.)

The
ratios
of
plants
to
systems
are
shown
in
Exhibit
4.3.
For
surface
water
and
GWUDI
CWSs,
the
number
of
surface
water
treatment
plants
per
system
varies
from
1.0
to
2.8.

Plant
information
is
not
available
for
noncommunity
water
systems.
Because
they
typically
serve
a
single
building
or
are
located
in
a
small
area,
this
analysis
assumes
that
the
ratio
of
plants
per
system
is
1:
1
for
all
size
categories.
Exhibit
4.3
summarizes
the
total
number
of
plants
for
CWSs,
NTNCWSs,
and
TNCWSs.
The
total
number
of
systems
displayed
in
Exhibit
4.3
is
the
baseline
from
which
regulatory
impacts
are
estimated.

Population
The
total
population
affected
by
the
LT2ESWTR
is
derived
from
SDWIS
data
(
USEPA
2000f).
For
purposes
of
estimating
the
number
of
households
served
(
one
method
for
determining
the
impact
of
the
rule
is
calculated
in
terms
of
costs
per
household),
the
total
population
affected
by
the
LT2ESWTR
is
estimated
to
be
174
million
people,
the
population
served
by
CWSs.
This
number
includes
modifications
to
SDWIS
data
to
correct
double­
counted
populations.
The
breakdown
of
population
by
size
category
is
shown
in
Exhibit
4.3a;
this
breakdown
includes
adjustments
made
to
populations
during
the
linking
process
(
see
Exhibit
4.2).

The
population
served
by
NTNCWSs
and
TNCWSs
was
not
included
in
the
determination
of
the
number
of
households.
People
served
by
NTNCWSs
and
TNCWSs
when
working,
attending
school,
or
traveling
also
are
served
on
a
regular
basis
by
another
source,
such
as
a
private
well
or
a
CWS.
Their
consumption
from
a
NCWS
is
an
incidental
use
in
addition
to
their
regular
service.
Adding
the
population
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
17
served
by
NCWSs
to
that
served
by
CWSs
would
lead
to
double
counting
in
cases
where
the
same
person
was
served
by
both
types
of
systems.
If
some
people
served
by
private
wells
receive
water
from
NTNCWSs
for
part
of
the
year,
the
error
introduced
by
ignoring
this
consumption
from
this
source
is
less
than
if
the
population
served
by
NCWSs
was
included
in
the
baseline.

For
determining
the
benefits
of
the
LT2ESWTR,
the
populations
served
by
CWSs
and
both
types
of
NCWSs
are
included,
since
people
receive
the
benefits
of
the
rule
at
each
system
they
use.

Uncertainty
in
Baseline
Input
Data
EPA
recognizes
that
there
is
uncertainty
in
the
data
sources
used
to
define
the
system
inventory
for
the
LT2ESWTR.
The
uncertainty
is
not
quantified
in
this
EA;
however,
a
qualitative
discussion
of
the
identified
uncertainties
is
provided
below.

As
noted
above,
SDWIS
and
the
1995
CWSS
are
the
primary
sources
of
system
inventory
data.
SDWIS
is
EPA's
primary
drinking
water
database.
SDWIS
stores
State­
reported
information
on
each
water
system,
including
name,
ID
number,
number
of
people
served,
type
of
system
(
year­
round
or
seasonal),
and
source
of
water
(
ground
water
or
surface
water),
along
with
monitoring
and
violation
information.
In
1998,
EPA
began
a
major
effort
to
assess
the
quality
of
its
drinking
water
data
in
SDWIS.
The
results,
published
in
Data
Reliability
Analysis
of
the
EPA
SDWIS/
FED,
found
that
the
quality
of
the
required
inventory
data
was
high
(
USEPA
2000d).
Thus,
EPA
believes
that
uncertainty
in
the
system
inventory
data
from
SDWIS
with
respect
to
numbers
of
systems,
source
information,
and
size
classification
is
low.

The
1995
CWSS
was
developed
to
gather
data
on
water
systems
in
the
United
States.
Of
the
3,681
systems
statistically
selected
to
receive
the
main
survey
questionnaire,
1,980
responded.
These
responses
were
weighted
to
maintain
statistical
representation
of
the
total
universe
of
CWSs
(
USEPA
1997c).

The
1995
CWSS
was
the
primary
data
source
used
to
develop
estimates
of
the
number
of
treatment
plants
per
system
and
of
average
and
design
flows
as
a
function
of
population
served
(
presented
in
section
4.3.3).
Because
the
CWSS
is
a
survey
of
CWSs,
estimates
based
on
the
data
will
contain
uncertainty
because
of
sampling
errors.
To
help
define
these
uncertainties,
the
CWSS
report
provides
the
confidence
intervals
on
certain
parameters.
The
report
does
not,
however,
contain
data
for
the
percentage
of
systems
disinfecting,
the
percentage
of
SDWIS
surface
water
systems
providing
ground
water,
and
the
number
of
treatment
plants
per
system
that
are
used
in
this
EA.
The
confidence
intervals
for
similar
parameters
can
provide
some
information
on
uncertainty.
For
example,
an
analysis
of
the
percent
of
ground
water
systems
with
no
treatment
(
which
uses
similar
data
to
the
analysis
of
percent
disinfecting),
yielded
95
percent
confidence
intervals
of
less
than
+/­
10
percent.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
18
4.3.3
Water
Treatment
Plant
Design
and
Average
Daily
Flows
Treatment
technology
costs
are
based
on
the
volume
of
water
treated
per
day.
The
cost
analysis
described
in
Chapter
6
uses
two
types
of
treatment
plant
flow:

°
Design
flow
 
the
maximum
capacity
at
which
the
plant
was
intended
to
operate,
expressed
in
millions
of
gallons
per
day
(
mgd).

°
Average
daily
flow
 
the
flow
produced
by
a
treatment
plant
in
1
year,
averaged
over
365
days,
expressed
in
mgd.

Design
flows
are
used
to
estimate
the
capital
costs
of
the
technology
that
will
be
installed
to
meet
the
requirements
of
the
LT2ESWTR.
Average
daily
flows
are
used
to
estimate
the
annual
cost
of
ongoing
operations
and
maintenance
(
O&
M).
Average
daily
flows
give
a
better
indication
of
chemical
usage
and
operational
costs
than
do
design
flows.
The
flows
presented
in
this
section
are
used
to
estimate
costs
for
both
unfiltered
and
filtered
plants.

To
derive
flow
information
for
different
sized
plants,
EPA
developed
the
following
regression
equations
relating
design
and
average
daily
flow
to
population
served
for
surface
water
systems
using
data
from
the
1995
CWSS:

Design
Flow
(
MGD)
=
0.36971
X0.97757
/
1,000
Average
Daily
Flow
(
MGD)
=
0.10540
X1.02058/
1,000
Where
X
=
mean
population
served.

The
derivation
of
these
equations
is
presented
in
detail
in
the
Model
Systems
Report
(
USEPA
2000b)
and
summarized
in
the
Baseline
Handbook
(
USEPA
2001c).
EPA
used
these
equations
to
estimate
mean
flows
per
system,
based
on
the
population
per
system
shown
in
column
A
of
Exhibit
4.4,
and
then
divided
by
the
number
of
plants
per
system
to
determine
the
flow
per
plant.

Exhibit
4.4a
summarizes
these
populations
and
design
and
average
daily
flows
for
filtered
plants
at
CWSs.
Exhibit
4.4b
shows
the
flows
and
populations
for
unfiltered
plants
(
all
unfiltered
plants
are
CWSs
except
for
one
TNCWS,
which
was
grouped
with
the
CWSs).
EPA
recognizes
that
there
is
a
range
of
design
and
average
daily
flows
within
each
category,
but
believes
that
using
mean
flow
values
is
adequate
for
the
cost
and
benefit
analysis
in
this
EA.

An
equivalent
regression
analysis
relating
NCWS
flows
to
population
served
was
not
done
in
the
Model
Systems
Report.
Therefore,
average
daily
and
design
flows
for
NCWSs
were
estimated
using
mean
population
served
per
plant
for
NCWSs
substituted
into
the
CWS
regression
equations.
Flows
are
summarized
in
Exhibit
4.4a
for
filtered
NTNCWSs
and
TNCWSs.
Plant
flows
for
filtered
CWSs,
NTNCWSs,
and
TNCWSs
differ
from
each
other
because
of
the
difference
in
mean
population
per
plant
for
each
of
the
three
categories,
and
the
volume
of
water
delivered
to
commercial
and
industrial
customers.
EPA
recognizes
that
using
CWS
equations
to
determine
NCWS
flows
may
overestimate
flows
and,
therefore,
benefits
and
costs.
This
overestimation
is
addressed
as
part
of
the
uncertainties
summarized
in
section
4.8.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
19
For
this
rulemaking,
EPA
considered
estimating
flows
for
NCWSs
according
to
service
category
(
e.
g.,
schools,
restaurants,
hotels,
industry)
instead
of
size
as
has
been
done
in
some
other
rules.
EPA
decided
against
an
approach
based
on
service
category
for
the
following
reasons:

°
Service
category
flows
are
based
on
mean
population
served
for
all
systems
in
that
category,
regardless
of
source
water
type.
EPA
expects
that
surface
water
and
GWUDI
sources
would
be
more
prevalent
in
larger
noncommunity
systems,
but
has
no
basis
for
developing
revised
population
estimates
for
each
service
category
by
source.

°
More
critical
to
the
LT2ESWTR,
the
prediction
of
technology
selection
in
Chapter
6
is
a
function
of
population
served,
and
does
not
directly
apply
to
service
categories
that
may
include
a
wide
range
of
water
system
sizes
(
e.
g.,
schools
can
be
very
small
local
buildings
or
large
universities).
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
20
System
Size
Population
per
System
Plants
per
System
Average
Daily
Flow
(
MGD)
per
Plant
Design
Flow
(
MGD)
per
Plant
A
B
(
0.10540*
A1.02058)/(
B*
1,000)
(
0.36971*
A0.97757)/(
B*
1,000)

<
100
59
1.0
0.01
0.02
101­
500
280
1.1
0.03
0.08
501­
1,000
769
1.0
0.09
0.25
1,001­
3,300
2,030
1.0
0.25
0.63
3,301­
10,000
5,961
1.2
0.63
1.51
10,001­
50,000
23,403
1.1
2.76
6.28
50,001­
100,000
69,010
1.8
5.08
11.04
100,001­
1
Million
277,629
1.6
23.67
48.43
>
1Million
2,158,693
2.8
109.71
205.50
<
100
52
1.0
0.01
0.02
101­
500
272
1.0
0.03
0.09
501­
1,000
719
1.0
0.09
0.23
1,001­
3,300
1,749
1.0
0.21
0.55
3,301­
10,000
5,318
1.0
0.67
1.62
10,001­
50,000
26,772
1.0
3.48
7.87
50,001­
100,000
93,204
1.0
12.43
26.66
100,001­
1
Million
169,846
1.0
22.94
47.93
>
1Million
0
1.0
­
­

<
100
44
1.0
0.00
0.01
101­
500
241
1.0
0.03
0.08
501­
1,000
758
1.0
0.09
0.24
1,001­
3,300
2,321
1.0
0.29
0.72
3,301­
10,000
5,703
1.0
0.72
1.74
10,001­
50,000
20,346
1.0
2.63
6.02
50,001­
100,000
59,747
1.0
7.90
17.26
100,001­
1
Million
0
1.0
­
­
>
1Million
0
1.0
­
­
CWSs
NTNCWSs
TNCWSs
Exhibit
4.4a
Filtered
Plant
Average
Daily
and
Design
Flow
by
System
Size
Source:
[
A]
Population
served
for
each
size
category
(
Exhibit
4.11,
Column
I)
divided
by
number
of
systems
in
each
category
(
Exhibit
4.11
Column
G),
based
on
the
treatment
baseline.
[
B],
[
C]
USEPA
2000b.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
21
System
Size
Population
per
System
Plants
per
System
Average
Daily
Flow
(
MGD)
per
Plant
Design
Flow
(
MGD)
per
Plant
A
B
C
=
(
0.10540*
A1.02058)/(
B*
1,000)
D
=
(
0.36971*
A0.97757)/(
B*
1,000)

<
100
75
1.0
0.01
0.03
101­
500
269
1.1
0.03
0.09
501­
1,000
774
1.0
0.09
0.25
1,001­
3,300
1,592
1.0
0.20
0.50
3,301­
10,000
6,881
1.2
0.87
2.09
10,001­
50,000
22,097
1.1
2.86
6.53
50,001­
100,000
69,138
1.8
9.17
19.91
100,001­
1
Million
209,787
1.6
28.45
58.92
>
1Million
3,386,848
2.8
486.43
893.70
CWS
Exhibit
4.4b
Unfiltered
Plant
Average
Daily
and
Design
Flow
by
System
Size
Source:
[
A]
Population
served
for
each
size
category
(
Exhibit
4.5,
Column
D)
divided
by
number
of
systems
in
each
category
(
Exhibit
4.5,
Column
A),
based
on
the
treatment
baseline.
[
B],
[
C]
USEPA
2000b.

4.4
Baseline
for
Unfiltered
Plants
(
Pre­
LT2ESWTR)

Unfiltered
plants
are
subject
to
different
LT2ESWTR
provisions
than
filtered
plants
(
e.
g.,
all
unfiltered
plants
must
achieve
some
inactivation
of
Cryptosporidium;
binning
only
determines
the
degree
of
inactivation
required).
Therefore,
the
baselines
for
unfiltered
and
filtered
plants
must
be
developed
separately.
The
following
sections
summarize
the
existing
treatment;
system,
plant,
and
population
data;
source
water
Cryptosporidium
occurrence;
and
predicted
finished
water
Cryptosporidium
occurrence
for
unfiltered
plants.

4.4.1
Treatment
Characterization
for
Unfiltered
Plants
EPA
estimates
that
a
number
of
plants
that
currently
are
not
required
to
filter
will
need
to
install
an
advanced
disinfectant
to
meet
the
2
log
or
3
log
removal
requirement
for
LT2ESWTR.
Some
treatment
data
collected
over
time
by
EPA
regional
offices
are
available
on
these
plants.
Most
of
these
plants
are
not
predicted
to
add
treatment
to
meet
the
Stage
1
or
Stage
2
DBPR
requirements,
because
they
generally
have
low
turbidity
source
water
and,
thus,
low
levels
of
precursors
for
DBP
formation.
Unfiltered
plants
are
not
subject
to
IESWTR
or
LT1ESWTR
filtration
requirements.
Thus,
EPA
used
the
existing
regional
data
to
develop
the
treatment
characterization
for
these
unfiltered
plants.

A
review
of
these
data
reveals
that
unfiltered
plants
use
a
variety
of
treatments
to
disinfect
or
control
other
water
quality
problems.
Plants
serving
3,300
or
fewer
people
generally
use
chlorine
as
the
primary
disinfectant,
although
at
least
one
plant
serving
501
to
1,000
people
uses
ozone.
Some
of
these
plants
may
also
employ
corrosion
control,
as
well
as
manganese
and
iron
removal.
Plants
serving
3,301
or
more
people
mainly
use
chlorination
for
disinfection,
and
at
least
one
plant
serving
3,301
to
10,000
people
uses
chlorine
dioxide.
Other
treatment
processes
used
in
medium
and
large
unfiltered
plants
include
corrosion
control,
softening,
fluoridation,
DBP
control,
taste
and
odor
control,
as
well
as
organics
and
iron
removal.
Some
plants
avoiding
filtration
may
have
already
installed
treatment
equivalent
to
filtration,
and
some
systems
may
filter
water
from
some
but
not
all
of
their
sources.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
22
The
unfiltered
plant
database
contains
treatment
data
for
less
than
50
percent
of
unfiltered
plants.
The
ICR
database
also
was
examined.
Only
one
unfiltered
plant
participating
in
the
ICR
was
found
to
use
ozone
(
USEPA
2000j),
and
its
dose
levels
were
not
high
enough
to
meet
the
LT2ESWTR
requirements.
No
ICR
unfiltered
plants
used
chlorine
dioxide.
Because
of
the
high
doses
of
ozone
or
chlorine
dioxide
required
to
inactivate
Cryptosporidium,
it
appears
unlikely
that
any
unfiltered
plants
are
currently
meeting
the
requirements
of
the
LT2ESWTR.
Therefore,
for
this
EA,
EPA
estimates
that
all
unfiltered
systems
will
have
to
add
advanced
disinfection
to
achieve
the
2
log
Cryptosporidium
inactivation
minimum
requirement.

4.4.2
Number
of
Unfiltered
Systems,
Plants,
and
Population
Served
Systems
that
operate
unfiltered
plants
can
be
placed
in
one
of
two
categories:

°
Systems
with
plants
that
are
now
unfiltered
but
are
required
to
filter
under
the
1989
Surface
Water
Treatment
Rule
(
SWTR)

°
Systems
that
meet
the
filtration
avoidance
criteria
of
the
SWTR
Three
of
the
12
unfiltered
plants
in
the
ICR
are
currently
unfiltered,
but
will
be
changing
to
filtration
in
the
future.
These
systems
are
subject
to
requirements
of
the
proposed
LT2ESWTR
for
filtered
systems
because
they
do
not
meet
avoidance
criteria
under
the
SWTR
and
are,
therefore,
included
in
the
baseline
for
filtered
systems
(
section
4.5).
In
addition,
a
fourth
plant,
the
Massachusetts
Water
Resources
Authority,
was
omitted
from
the
unfiltered
baseline
(
and
not
moved
to
filtered
baseline)
because
of
the
uncertainties
regarding
its
filtration
avoidance
status
due
to
ongoing
litigation
at
the
time
of
calculation.
The
purchased
plants
associated
with
this
plant
were
also
removed
from
the
baseline.

The
baseline
for
unfiltered
plants,
therefore,
includes
only
systems
that
meet
the
SWTR
avoidance
criteria.
The
criteria
include
the
following:

°
Disinfect
to
achieve
3
and
4
log
reduction
of
Giardia
and
viruses,
respectively
°
Have
watershed
control
measures
in
place
°
Are
below
source
water
limits
on
fecal
coliform
occurrence
(
20/
100
ml)
and
turbidity
(
5
nephelometric
turbidity
units
(
NTU))

Exhibit
4.5
presents
the
baseline
for
unfiltered
systems
and
plants
that
is
used
for
estimating
costs
and
benefits
in
this
EA.
Data
on
population
and
the
number
of
systems
are
derived
from
SDWIS
and
from
the
ICR
for
large
systems
(
USEPA
2000f,
2000j).
The
number
of
plants
is
calculated
using
the
plants­
per­
system
ratios
given
earlier.
There
is
only
one
TNCWS
unfiltered
system,
and
there
are
no
unfiltered
NTNCWSs.
Maintaining
a
separate
category
in
subsequent
analyses
for
one
system
was
unlikely
to
add
precision,
so
this
TNCWS
was
grouped
with
CWSs
in
all
subsequent
analyses.
These
baseline
values,
in
conjunction
with
flows
presented
in
section
4.3.3,
are
used
to
estimate
costs
for
unfiltered
systems
(
see
Chapter
5).
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
23
System
Size
(
Population
Served)
Number
of
Systems
Plants
per
System
Number
of
Plants
Population
Served
Percent
of
Total
Population
Served
by
Surface
and
GWUDI
CWSs
A
B
C=
A*
B
D
E
<
100
1
1.0
1
75
0.00%
101­
500
5
1.1
6
1,343
0.00%
501­
1,000
7
1.0
7
5,418
0.00%
1,001­
3,300
14
1.0
14
22,281
0.01%
3,301­
10,000
22
1.2
26
150,000
0.09%
10,001­
50,000
16
1.1
18
353,552
0.20%
50,001­
100,000
3
1.8
5
207,415
0.12%
100,001­
1
Million
4
1.6
6
839,148
0.48%
>
1Million
3
2.8
8
10,160,545
5.85%

Total
75
91
11,739,777
6.76%
Exhibit
4.5
Treatment
Baseline
for
Unfiltered
Plants
By
System
Size
Notes:
All
systems
are
CWSs
except
one
TNCWS,
which
was
grouped
with
the
CWSs
for
analysis.

Sources:
[
A]
SDWIS
(
USEPA
2000f)
data
adjusted
by
EPA
to
exclude
systems
that
do
not
meet
filtration
avoidance
criteria.
[
B]
Exhibit
4.3,
Column
K.
[
D]
SDWIS
data
for
the
systems
in
Column
A
(
2000h),
modified
to
include
populations
added
in
the
linking
process
(
see
Exhibit
4.1).
[
E]
Population
served
(
Column
D)
divided
by
total
population
served
by
surface
water
and
GWUDI
CWSs
(
Exhibit
4.3a,
bottom
of
Column
M).

4.4.3
Source
Water
Cryptosporidium
Occurrence
for
Unfiltered
Plants
ICR
data
from
12
plants
that
are
classified
as
unfiltered
surface
water
are
used
to
characterize
Cryptosporidium
occurrence
(
USEPA
2000j).
The
results
of
the
ICR
Cryptosporidium
monitoring
were
evaluated
using
the
model
described
in
section
4.2.2.
Because
a
few
of
these
plants
do
not
meet
the
filtration
avoidance
criteria,
a
sensitivity
analysis
was
performed
to
see
if
results
would
be
significantly
different
if
they
were
excluded.
The
occurrence
distributions
were
nearly
identical;
therefore,
the
original
results
were
used
in
the
analysis.

Observed
Cryptosporidium
Occurrence
Observed
results
for
ICR
unfiltered
plants
are
shown
in
Exhibit
4.6.
A
comparison
of
these
results
to
those
for
filtered
plants
in
the
ICR
(
Exhibit
4.12)
shows
both
a
lower
rate
of
positive
samples
and
a
lower
average
concentration
of
Cryptosporidium
for
unfiltered
plants.
While
too
few
unfiltered
plants
were
sampled
in
the
ICRSS
to
carry
out
a
meaningful
analysis,
the
few
that
were
sampled
also
were
on
the
low
side
of
the
ICR
unfiltered
occurrence
distribution.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
24
Exhibit
4.6
Observed
ICR
Source
Water
Total
Oocyst
Occurrence
for
Unfiltered
Plants
Total
Number
of
Plants
Number
of
Plants
with
at
Least
One
Positive
Sample
(
Percent)
Observed
Plant­
Mean
Data
(
oocysts/
L)

Mean
Median
90th
Percentile
12
7
(
58%)
0.002
0.001
0.005
Notes:
Total
Cryptosporidium
includes
non­
empty
and
empty
oocysts.
Non­
empty
includes
oocysts
with
internal
structures
and
with
amorphous
structures.
For
each
plant,
all
monthly
observations
were
averaged
over
the
sampling
period
(
18
months)
to
produce
plant­
mean
data.
The
mean,
median,
and
90th
percentile
shown
summarize
plant­
mean
Cryptosporidium
for
all
unfiltered
plants.

Source:
USEPA
2000j.

Modeled
Cryptosporidium
Occurrence
The
data
from
these
12
plants
are
used
to
fit
the
unfiltered
plants
occurrence
model,
which
serves
as
input
to
the
EA.
The
modeling
was
carried
out
using
the
approach
outlined
in
section
4.2.2.
As
explained
in
that
section,
the
modeling
produces
a
collection
of
plausible
occurrence
distributions.
Exhibit
4.7
summarizes
this
collection
of
distributions.
The
solid
center
curve
represents
the
mean
distribution
across
the
collection,
and
the
dotted
lines
give
a
90­
percent
confidence
bound
for
the
unfiltered
occurrence
distribution.

4.4.4
Finished
Water
Cryptosporidium
Occurrence
for
Unfiltered
Plants
As
mentioned
in
section
4.4.1,
because
unfiltered
plants
do
not
have
advanced
treatment
technologies
in
place
that
are
capable
of
meeting
2.0
log
Cryptosporidium
removal
or
inactivation,
they
are
not
expected
to
remove
or
inactivate
any
Cryptosporidium
from
their
source
water.
Although
most
unfiltered
systems
chlorinate
their
water
(
a
few
may
use
other
disinfectants
besides
chlorine),
chlorination
is
ineffective
for
inactivation
of
Cryptosporidium.
Therefore,
the
finished
water
occurrence
of
Cryptosporidium
for
unfiltered
plants
is
assumed
to
be
the
same
as
the
source
water
occurrence
shown
in
Exhibit
4.7.

This
occurrence
distribution
is
derived
from
the
ICR
data
set
(
USEPA
2000j).
Although
there
were
unfiltered
plants
in
the
ICRSS
data
sets,
they
were
too
few
to
use
successfully
in
the
model
to
derive
national
distributions.
Thus,
there
are
no
explicit
estimated
occurrence
distributions
for
the
unfiltered
ICRSS
data
sets.
In
order
to
develop
national
benefit
estimates
for
the
ICRSS
data
sets,
EPA
estimated
unfiltered
results
based
on
the
ratios
between
ICR
and
ICRSSM,
and
ICR
and
ICRSSL,
results
for
filtered
plants.
Thus,
although
no
explicit
occurrence
distributions
were
estimated
for
the
ICRSS
data
sets,
the
likely
differences
in
Cryptosporidium
occurrence
between
the
ICR
and
ICRSS
unfiltered
data
sets
are
reflected
in
later
analyses.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
25
1e­
005
0.0001
0.001
0.01
0.1
1
10
Plant
Mean
Cryptosporidium
Concentration
(
Total
oocysts/
L)
0%
20%
40%
60%
80%
100%

Cumulative
Percentage
of
Plants
5th/
95th
%
tile
Median
Example:

Modeled
Percent
of
Plants
with
Average
Source
Water
Cryptosporidium
Concentration
Below
0.01
oocysts/
L:

5th
%
tile
Bound
=
74%

Median
Estimate
=
59%

95
%
tile
Bound
=
39%
This
graph
shows
modeled
variability
and
uncertainty
in
source
water
Cryptosporidium
occurrence
for
unfiltered
systems.
It
summarizes
1,000
plausible
curves
used
in
the
Economic
Analysis.

Ninety
percent
of
the
modeled
curve
values
fall
within
the
dashed­
line
uncertainty
bounds.
The
heavy
center
line
reflects
the
central
tendency
across
all
1,000
curves.

Among
the
1,000
modeled
curves,
any
single
curve
describes
variability
in
occurrence
across
all
filtered
systems.
Steeper
curves
indicate
less
variability
from
system
to
system.

Variability
Uncertainty
Exhibit
4.7
Modeled
Source
Water
Cryptosporidium
Occurrence
ICR
Data
for
Unfiltered
Systems
Source:
USEPA
2003c.

Under
the
LT2ESWTR,
all
unfiltered
systems
must
provide
treatment
for
Cryptosporidium.
Systems
with
Cryptosporidium
concentrations
<
0.01
oocysts/
L
must
provide
2
log
treatment,
while
systems
with
concentrations
>
0.01
oocysts/
L
must
provide
3
log
treatment.
The
predicted
binning
for
unfiltered
systems,
derived
from
the
occurrence
distribution
in
Exhibit
4.7,
is
shown
in
Exhibit
4.8.
Percentages
represent
averages
over
250
simulated
assignments
of
unfiltered
systems
to
2
or
3
log
treatment
bins.
In
the
case
of
systems
serving
fewer
than
100,000
people,
the
simulated
binning
was
based
on
1,000
values
drawn
from
the
modeled
unfiltered
occurrence
distribution.
For
the
small
number
of
systems
serving
100,000
or
more,
the
binning
simulation
drew
Cryptosporidium
concentrations
directly
from
ICR
survey
results.
In
both
cases,
recovery
was
simulated
to
match
recovery
rates
expected
in
Cryptosporidium
monitoring
employing
EPA
Methods
1622/
1623.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
26
Exhibit
4.8
Predicted
System
Binning
for
Unfiltered
Systems,
Based
on
Central
Tendency
of
Cryptosporidium
Occurrence
System
Size
(
Population
Served)
2.0
Log
Treatment
3.0
Log
Treatment
<
100,000
79.2%
20.8%

>
100,000
81.2%
18.8%

Source:
Monte
Carlo
simulation.
For
systems
serving
less
than
100,000,
percentages
based
on
modeled
occurrence
distribution
and
lab
method
recovery
rate
distribution.
For
systems
serving
100,000
or
more,
percentages
are
based
on
actual
ICR
results
and
modeled
lab
method
recovery
distribution.

The
percentages
in
Exhibit
4.8
were
used
to
determine
treatment
costs
incurred
by
each
unfiltered
system.

4.5
Baselines
for
Filtered
Plants
(
Pre­
LT2ESWTR)

This
section
first
presents
the
Pre­
LT2ESWTR
treatment
characterization
for
filtered
plants.
It
includes
estimates
of
number
of
systems,
plants,
and
population
served
for
filtered
systems.
It
also
contains
source
water
and
finished
water
Cryptosporidium
occurrence
and
predicted
bin
classification.

4.5.1
Treatment
Characterization
for
Filtered
Plants
The
treatment
characterization
for
the
LT2ESWTR
must
take
into
account
projected
treatment
modifications
made
to
comply
with
other
existing
and
soon­
to­
be­
promulgated
rules.
This
adjustment
allows
treatment
changes
attributable
only
to
LT2ESWTR
requirements
to
be
isolated
for
benefit
and
cost
analysis.
In
addition,
systems
with
certain
treatments
already
in
place
prior
to
the
implementation
of
the
LT2ESWTR
may
qualify
for
Pre­
LT2ESWTR
credit,
meaning
they
can
get
credit
towards
the
log
treatment
requirements
of
the
LT2ESWTR.
The
rest
of
this
subsection
explains
how
treatment
changes
attributable
to
the
IESWTR,
LT1ESWTR,
Stage
1
DBPR,
and
Stage
2
DBPR
have
been
accounted
for
in
developing
the
LT2ESWTR
baseline.
Although
other
rules
have
been
promulgated
recently
or
are
scheduled
to
be
promulgated
before
this
rule,
they
either
do
not
affect
surface
water
supplies
or
do
not
involve
installation
of
treatment
that
is
expected
to
significantly
remove
or
inactivate
Cryptosporidium.
Therefore,
such
rules
were
not
considered
further.

Post­
IESWTR
and
LT1ESWTR
Treatment
Characterization
For
this
EA,
it
is
assumed
that
all
medium
and
large
systems
(
those
serving
at
least
10,000
people)
using
conventional
or
direct
filtration
meet
the
combined
effluent
turbidity
limit
of
0.3
NTU
95
percent
of
the
time,
as
required
under
IESWTR
and
LT1ESWTR.
EPA
assumed
as
part
of
the
economic
analyses
for
the
IESWTR
and
LT1ESWTR
that
systems
would
achieve
less
than
0.2
NTU
95
percent
of
the
time
in
order
to
operate
within
a
margin
of
safety.
Lower
finished
water
turbidity,
with
a
combined
effluent
turbidity
level
of
0.15
NTU,
is
one
of
the
approaches
in
the
microbial
toolbox
(
described
in
detail
in
Chapter
6)
available
to
systems
for
achieving
an
additional
0.5
log
removal
credit
for
Cryptosporidium.
There
are
several
other
toolbox
technologies
that
plants
may
already
have
installed
that
would
gain
them
Pre­
LT2ESWTR
credit
of
0.5
log
toward
Cryptosporidium
removal
requirements.
These
include
secondary
filters
and
two­
stage
softening.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
27
To
determine
the
number
of
plants
that
might
obtain
credit
for
already
using
some
of
these
toolbox
technologies,
several
data
sources
were
reviewed.
Data
from
the
ICR,
the
American
Water
Works
Association
(
AWWA)
(
AWWA
2000),
a
survey
of
small
systems
conducted
by
the
National
Rural
Water
Association
(
NRWA)
(
USEPA
2001b),
and
the
1995
CWSS
were
reviewed
to
determine
the
percentage
of
plants
that
have
these
technologies
in
such
a
way
as
to
qualify
for
removal
credits.
Exhibit
4.9
provides
an
estimate
of
the
percentage
of
systems
that
would
be
capable
of
obtaining
each
of
these
Pre­
LT2ESWTR
half­
log
removal
credits
with
existing
equipment
and
operations
or
with
equipment
and
operations
predicted
to
be
installed
and
in
use
prior
to
promulgation
of
this
rule.
Details
on
the
derivation
of
these
numbers
are
provided
in
Appendix
A.
Some
of
these
plants
may
be
able
to
obtain
credit
for
having
multiple
technologies.
The
percentage
of
plants
estimated
to
get
1.0
log
removal
credit
is
presented
in
Exhibit
4.9
as
well.
The
estimates
of
1.0
log
credits
are
derived
from
the
individual
credits
for
plants
receiving
the
various
0.5
log
removal
credits,
assuming
complete
independence
of
the
chance
of
having
any
one
of
the
0.5
log
credits.

Exhibit
4.9
Percentage
of
Plants
with
Pre­
LT2ESWTR
Cryptosporidium
Log
Reduction
Credits
for
Existing
Technologies
System
Size
(
Population
Served)
Plants
with
0.15
NTU
Finished
Water
Turbidity
(
0.5
Log)
Plants
with
Multiple
Settling
Basins
(
Conventional
and
Softening)
(
0.5
Log)
Plants
with
Multiple
Filters
(
0.5
Log)
Plants
with
0.5
Log
Total
Credit
Plants
with
1.0
Log
Total
Credit
A
B
C
D=
A
+
B
+
C
­
E
E=(
A*
B)+(
A*
C)+(
B*
C)

Small
(
#
10k)
34%
3%
0%
36%
1%

Medium
(
10k
­
100k)
46%
5%
4%
51%
4%

Large
(>
100k)
46%
5%
7%
52%
6%

Source:
Appendix
A,
Exhibit
A.
7.

Post­
Stage
2
DBPR
Treatment
Characterization
Under
the
LT2ESWTR,
EPA
exempts
plants
from
monitoring,
bin
classification,
and
associated
treatment
requirements
if
the
plants
are
achieving
5.5
log
reduction.
EPA
estimates
that
very
few
plants
met
this
requirement
at
the
time
of
the
ICR.
A
proportion
of
surface
water
plants,
however,
are
expected
to
implement
advanced
technologies
to
meet
the
DBP
requirements
of
the
Stage
1
DBPR
and
Stage
2
DBPR.
Several
of
these
technologies
will
provide
additional
logs
of
Cryptosporidium
removal
or
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
28
inactivation
in
addition
to
reducing
DBPs.
Advanced
technologies
that
will
provide
both
DBP
and
Cryptosporidium
control
include
the
following:

°
Chlorine
dioxide
°
UV
°
Ozone
°
Microfiltration/
ultrafiltration
(
MF/
UF)

SWAT
(
see
section
4.2.3)
used
source
water
and
treatment
data
collected
through
the
ICR
to
predict
the
percentage
of
large
systems
that
would
have
to
add
treatment
to
meet
Stage
2
DBPR
requirements.
SWAT
was
also
able
to
predict
plants'
technology
selection
based
on
source
water
characteristics,
treatment
plant
configurations,
and
other
factors.
SWAT
results
were
extrapolated
to
medium
and
small
systems
using
best
professional
judgment.
EPA's
Economic
Analysis
for
the
Stage
2
Disinfectants
and
Disinfection
Byproducts
Rule
(
USEPA
2003d)
provides
a
detailed
description
of
Stage
2
DBPR
requirements,
how
SWAT
modeled
ICR
systems,
and
how
SWAT
was
used
to
develop
the
Stage
2
DBPR
compliance
forecast
for
plants
of
all
sizes.
Exhibit
4.10
presents
predictions
from
SWAT
of
technology
use
following
Stage
2
DBPR
implementation
for
the
four
technologies
listed
above.

Exhibit
4.10
Predicted
Percentage
of
Plants
Using
Advanced
Technologies
Following
Implementation
of
the
Stage
2
DBPR
System
Size
(
Population
Served)
Chlorine
Dioxide
UV
Ozone
MF/
UF
#
100
0
%
3.1
%
0
%
18.6
%

101­
500
2.4
%
0.4
%
11.9
%
9.9
%

501­
1,000
2.4
%
0.4
%
11.9
%
9.9
%

1,001­
3,300
5.1
%
0.5
%
10.1
%
5.4
%

3,301­
10,000
5.1
%
0.5
%
10.1
%
5.4
%

10,001­
50,000
7.0
%
0.7
%
12.8
%
1.8
%

50,001­
100,000
7.0
%
0.7
%
12.8
%
1.8
%

100,001­
1
Million
7.0
%
0.7
%
12.8
%
1.8
%

>
1
Million
7.0
%
0.7
%
12.8
%
1.8
%

Source:
Economic
Analysis
for
the
Stage
2
Disinfectants/
Disinfection
Byproducts
Rule
(
USEPA
2003d),
Exhibit
6.15a.

EPA
did
not
adjust
the
baseline
for
plants
using
chlorine
dioxide
and
ozone.
Chlorine
dioxide
and
ozone
doses
required
for
Cryptosporidium
inactivation
are
higher
than
SWTR
requirements
for
inactivation
of
Giardia
and
viruses.
To
evaluate
the
use
of
these
technologies
at
doses
that
could
inactivate
Cryptosporidium,
both
incremental
costs
for
the
increased
dose
and
incremental
benefits
would
need
to
be
evaluated.
However,
available
studies
do
not
allow
the
quantitative
evaluation
of
inactivation
based
on
dose
changes.
In
the
absence
of
usable
data,
this
analysis
assumes
that
plants
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
29
predicted
to
have
installed
chlorine
dioxide
and
ozone
would
not
meet
the
monitoring
and
treatment
exemption
requirements
for
the
LT2ESWTR.
The
cost
model
for
this
EA
assumes
that
these
plants
would
install
the
entire
technology
again
for
the
LT2ESWTR
rather
than
simply
increase
the
dose,
resulting
in
an
overestimate
of
costs.
Uncertainties
and
biases
affecting
costs
are
summarized
in
section
4.8.

Currently,
very
few
plants
use
UV.
This
is
partly
because
of
the
need
for
higher
doses
required
to
inactivate
viruses
(
compared
to
that
needed
for
protozoans)
and
the
need
to
maintain
a
disinfectant
residual.
These
factors
result
in
only
about
half
a
percent
of
plants
nationwide
using
UV.
Since
this
is
such
a
small
number
of
plants,
the
LT2ESWTR
baseline
was
not
adjusted
to
reflect
plants
that
already
have
installed
UV.

MF/
UF
is
the
only
treatment
process
assumed
to
achieve
5.5
log
removal.
The
LT2ESWTR
baseline
is
adjusted
in
several
ways
to
account
for
the
percentage
of
plants
that
may
already
be
using
MF/
UF.
CWS
and
NTNCWS
plants
(
both
in
small
and
large
systems)
predicted
to
have
installed
MF/
UF
prior
to
the
Stage
2
DBPR
are
removed
entirely
from
the
monitoring
and
treatment
baselines.
Plants
achieving
greater
than
5.5
log
treatment
of
Cryptosporidium
are
meeting
the
highest
level
of
treatment
that
could
be
required
based
on
source
water
monitoring.
These
plants
would
not
have
to
add
treatment
if
they
did
monitor
and
were
assigned
to
the
most
stringent
treatment
bin.
EPA
anticipates
that
some
plants
will
install
MF/
UF
to
meet
the
requirements
of
the
Stage
2
DBPR.
These
plants
are
included
in
the
rule
implementation
and
initial
monitoring
baselines
(
because
they
are
still
subject
to
the
LT2ESWTR
and
because
the
Stage
2
DBPR
schedule
is
such
that
they
will
not
have
installed
MF/
UF
before
initial
Cryptosporidium
monitoring
begins),
but
not
the
future
monitoring
and
treatment
baselines.
TNCWSs
are
unlikely
to
have
installed
advanced
technologies,
partly
because
TNCWSs
are
not
addressed
by
the
Stage
1
or
Stage
2
DBPRs.

Treatment
plants
with
different
types
of
filtration
systems
have
been
regulated
differently
under
IESWTR
and
LT1ESWTR
and,
to
a
small
extent,
under
LT2ESWTR.
The
types
of
filtration
included
in
this
analysis
are
as
follows:

°
Conventional
filtration
includes
coagulation,
flocculation,
and
sedimentation
of
particles,
followed
by
granular
media
filtration.

°
Direct
filtration
involves
coagulation
and
flocculation
followed
by
rapid
sand
filtration,
but
no
sedimentation
basin.

°
Slow
sand
filtration
works
at
very
low
filtration
rates
without
the
use
of
coagulant
in
pretreatment.

°
Diatomaceous
earth
filtration
works
at
low
filtration
rates
with
the
addition
of
diatomaceous
earth.

°
Alternative
filtration
systems
include
membrane,
bag,
and
cartridge
filters.

Plants
filtering
by
slow
sand
and
diatomaceous
earth
are
not
required
to
meet
new
combined
filter
effluent
provisions
under
the
IESWTR
or
LT1ESWTR
because
they
can
generally
achieve
higher
Cryptosporidium
removals
at
higher
effluent
turbidity
levels.
These
treatment
plants
are
still
subject
to
additional
treatment
requirements
of
the
LT2ESWTR
based
on
the
results
of
source­
water
monitoring
and,
thus,
are
included
in
the
filtered
system
baseline.
Direct
filtration
plants
also
are
included
in
this
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
30
baseline;
however,
they
have
slightly
different
Cryptosporidium
reduction
requirements
under
the
LT2ESWTR.
The
impacts
of
these
additional
requirements
for
direct
filtration
plants
are
addressed
in
Chapter
6.

4.5.2
Number
of
Filtered
Plants
and
Population
Served
Exhibit
4.11
presents
the
baselines
for
filtered
plants.
As
mentioned
previously,
plants
with
MF/
UF
are
assumed
to
achieve
5.5
log
treatment
and,
therefore,
are
not
subject
to
treatment
and,
in
some
cases,
monitoring
requirements.
Note
that
the
baseline
numbers
of
plants
used
to
estimate
benefits
and
treatment
costs
for
filtered
plants
(
columns
G
and
H)
include
nonpurchased
plants
and
purchased
plants
that
could
not
be
linked
with
the
systems
from
which
they
purchased
their
water.
To
capture
the
total
flow
that
must
be
treated
and
the
population
affected
by
the
LT2ESWTR,
EPA
includes
these
purchased
systems
in
the
analysis
as
though
they
are
treating
water
themselves.
Effectively,
this
assumption
places
more
plants
in
smaller
size
categories
rather
than
fewer
plants
in
larger
size
categories,
resulting
in
lower
flows
and
populations
per
plant.
This
bias
was
minimized,
however,
by
the
analysis
described
in
section
4.3.1,
linking
most
purchased
systems
to
their
sellers.
In
the
process
of
linking
purchased
systems
to
sellers,
the
population
served
by
the
purchased
systems
is
added
to
that
of
the
sellers.
This
can
result
in
a
population
change
large
enough
to
bump
the
seller
up
to
a
higher
size
category.
Uncertainties
associated
with
categorization
of
some
purchased
systems
as
retail
are
summarized
in
section
4.8.

Source
water
monitoring
costs
are
not
based
on
flow
or
population,
but
are
calculated
on
a
perplant
basis
based
on
the
number
of
plants
in
column
E
of
Exhibit
4.11.
In
estimating
monitoring
costs,
EPA
assumed
that
purchased
plants
do
not
treat
source
water
directly
and,
thus,
would
not
directly
incur
monitoring
costs.
Further
evaluation
of
the
baseline
for
filtered
plant
monitoring
is
provided
in
Appendix
D.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
31
System
Size
Number
of
Filtered
Systems
Number
of
Filtered
Plants
Number
of
Filtered
Systems
Number
of
Filtered
Plants
Number
of
Filtered
Systems
Number
of
Filtered
Plants
Population
Served
A
B
C
D=
A*(
1­
C)
E=
B*(
1­
C)
F
G
H
I
CWSs
<
100
529
529
3.6%
510
510
18.64%
396
396
23,533
101­
500
752
827
3.6%
725
797
9.93%
719
791
201,247
501­
1,000
449
449
3.6%
433
433
9.93%
407
407
313,169
1,001­
3,300
1,073
1,073
3.6%
1,034
1,034
5.44%
995
995
2,019,557
3,301­
10,000
995
1,194
3.6%
959
1,151
5.44%
982
1,179
5,857,029
10,001­
50,000
943
1,037
0.4%
940
1,034
1.83%
1,064
1,171
24,904,234
50,001­
100,000
190
342
0.4%
189
341
1.83%
239
429
16,462,361
100,001­
1
Million
201
322
0.4%
200
320
1.83%
247
396
68,681,089
>
1
Million
10
28
0.4%
10
28
1.83%
19
52
40,263,981
Total
5,142
5,801
5,001
5,648
5,068
5,816
158,726,200
NTNCWSs
<
100
202
202
3.6%
195
195
18.64%
242
242
12,463
101­
500
236
236
3.6%
228
228
9.93%
269
269
73,307
501­
1,000
84
84
3.6%
81
81
9.93%
97
97
69,919
1,001­
3,300
55
55
3.6%
53
53
5.44%
67
67
117,421
3,301­
10,000
13
13
3.6%
13
13
5.44%
22
22
115,651
10,001­
50,000
3
3
0.4%
3
3
1.83%
8
8
210,252
50,001­
100,000
0
0
0.4%
0
0
1.83%
1
1
91,497
100,001­
1
Million
0
0
0.4%
0
0
1.83%
1
1
166,735
>
1
Million
0
0
0.4%
0
0
1.83%
0
0
0
Total
593
593
572
572
707
707
857,244
TNCWSs
<
100
731
731
0.0%
731
731
0.00%
1,234
1,234
53,903
101­
500
380
380
0.0%
380
380
0.00%
486
486
117,344
501­
1,000
57
57
0.0%
57
57
0.00%
81
81
61,426
1,001­
3,300
53
53
0.0%
53
53
0.00%
67
67
155,500
3,301­
10,000
25
25
0.0%
25
25
0.00%
31
31
176,789
10,001­
50,000
8
8
0.0%
8
8
0.00%
10
10
203,456
50,001­
100,000
3
3
0.0%
3
3
0.00%
3
3
179,240
100,001­
1
Million
0
0
0.0%
0
0
0.00%
0
0
0
>
1
Million
0
0
0.0%
0
0
0.00%
0
0
0
Total
1,256
1,256
1,256
1,256
1,912
1,912
947,658
Treatment
Baseline
Percent
Avoiding
Treatment
Requirements
Implementation
Baseline
Percent
Avoiding
Monitoring
and
Treatment
Requirements
Monitoring
Baseline
Exhibit
4.11
Treatment
Baseline
for
Filtered
Plants
Notes:
[
A]
Number
of
filtered
systems
=
nonpurchased
unlinked
systems
(
Exhibit
4.3,
Columns
C+
D)
­
unfiltered
systems
(
Exhibit
4.5,
Column
A).
[
B]
Number
of
filtered
plants
=
Column
A
*
Exhibit
4.3,
Column
K.
[
C]
Percentage
of
plants
predicted
to
have
installed
MF/
UF
to
comply
with
Stage
1
DBPR,
from
SWAT
technology
results
for
Pre­
Stage
2
DBPR
(
USEPA
2003d).
[
F]
Percentage
of
plants
predicted
to
have
installed
MF/
UF
to
comply
with
Stage
2
DBPR,
from
SWAT
technology
results
for
Post­
Stage
2
DBPR
(
USEPA
2003d).
[
G]
Number
of
filtered
systems
subject
to
treatment
=
linked
nonpurchased
and
purchased
systems
(
Exhibit
4.3,
Column
J)
*
(
1­
Column
F).
This
column
includes
nonpurchased
systems
that
conducted
monitoring,
but
excludes
systems
that
monitored
but
that
are
predicted
to
install
MF/
UF
to
comply
with
the
Stage
2
DBPR.
Purchased
systems
are
included
so
their
populations
can
be
used
to
determine
benefits
of
treatment
installed
under
the
LT2ESWTR
(
customers
of
purchased
systems
will
incur
benefits
of
treatment
installed
at
the
associated
nonpurchased
system).
[
H]
Number
of
filtered
plants
=
Column
G
*
Exhibit
4.3,
Column
K.
[
I]
Population
=
(
Pop.
served
by
linked
nonpurchased
and
purchased
systems
(
Exhibit
4.3,
Column
M)
­
pop.
served
by
unfiltered
systems
(
Exhibit
4.5,
Column
A))
*
(
1­
F).
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
32
4.5.3
Source
Water
Cryptosporidium
Occurrence
for
Filtered
Plants
For
filtered
plants,
the
results
of
plant­
specific
source
water
monitoring
for
Cryptosporidium
will
dictate
the
additional
treatment
required
to
meet
provisions
of
the
LT2ESWTR.
This
subsection
summarizes
observed
Cryptosporidium
data
and
discusses
statistical
modeling
of
the
observed
data
used
to
estimate
underlying
"
true"
Cryptosporidium
concentration
distributions.
The
"
true"
distributions
are
then
used
to
project
bin
classifications
for
filtered
systems.

Observed
Cryptosporidium
Occurrence
Exhibit
4.12
summarizes
the
observed
Cryptosporidium
occurrence
of
the
ICR
and
ICRSS
studies.
See
section
4.2.1
for
a
description
of
the
study,
laboratory
methods,
and
method
used
to
count
oocysts.
Additional
data
on
observed
occurrence
are
available
in
the
Occurrence
and
Exposure
Assessment
(
USEPA
2003c).
The
data
shown
for
"
total
oocysts"
were
used
to
generate
the
modeled
Cryptosporidium
distributions
shown
below.

Exhibit
4.12
Summary
of
Observed
Source
Water
Cryptosporidium
Total
Oocyst
Occurrence
 
Filtered
Plant
Data
Data
Set
Total
Number
of
Plants
Number
of
Plants
with
at
Least
One
Positive
Sample
(
Percent)
Observed
Plant­
Mean
Data
(
oocysts/
L)

Mean
Median
90th
Percentile
ICR
350
154
(
44%)
0.066
0.000
0.190
ICRSSL
40
34
(
85%)
0.040
0.020
0.100
ICRSSM
40
34
(
85%)
0.080
0.020
0.110
Notes:
Total
oocysts
include
non­
empty
and
empty
oocysts.
Non­
empty
oocysts
include
oocysts
with
internal
structures
and
with
amorphous
structures.
For
each
plant,
all
monthly
observations
were
averaged
over
the
sampling
period
(
12
months
for
ICRSSM
and
ICRSSL
and
18
months
for
ICR)
to
produce
plant­
mean
data.
The
mean,
median,
and
90th
percentile
shown
summarize
plant­
mean
Cryptosporidium
for
all
plants.

Source:
USEPA
2003c.

Modeled
Cryptosporidium
Occurrence
Each
of
the
three
data
sets
shown
in
Exhibit
4.12
was
used
to
fit
an
occurrence
model
for
filtered
plants
using
the
approach
outlined
in
section
4.2.2.
As
explained
in
that
section,
the
modeling
produces
a
collection
of
plausible
occurrence
distributions
from
each
of
the
three
data
sets.
Exhibits
4.13
through
4.15
summarize
the
resulting
collection
of
distributions
from
each
of
the
three
models.
The
solid
center
curves
represent
the
mean
distribution
across
a
given
collection,
and
the
dotted
lines
give
a
90­
percent
confidence
bound
for
the
true
filtered
occurrence
distribution
based
on
the
particular
data
set.

As
outlined
in
section
5.2.4.1,
all
three
of
these
models
were
used,
independently,
as
input
to
the
benefits
modeling.
Each
model
is
also
used,
along
with
a
distribution
of
lab
method
recovery
rates,
to
simulate
results
of
initial
LT2ESWTR
Cryptosporidium
monitoring.
The
assumed
recovery
rate
distribution
is
based
on
"
spiked"
sample
evaluations
of
the
lab
method
that
will
be
used
for
initial
monitoring
(
EPA
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
33
1e­
005
0.0001
0.001
0.01
0.1
1
10
Plant
Mean
Cryptosporidium
Concentration
(
Total
oocysts/
L)
0%
20%
40%
60%
80%
100%

Cumulative
Percentage
of
Plants
5th/
95th
%
tile
Median
Example:

Modeled
Percent
of
Plants
with
Average
Source
Water
Cryptosporidium
Concentration
Below
0.01
oocysts/
L:

5th
%
tile
Bound
=
28%

Median
Estimate
=
23%

95th
%
tile
Bound
=
18%
This
graph
shows
modeled
variability
and
uncertainty
in
source
water
Cryptosporidium
occurrence
for
filtered
systems.
It
summarizes
1,000
plausible
curves
used
in
the
Economic
Analysis.

Ninety
percent
of
the
modeled
curve
values
fall
within
the
dashed­
line
uncertainty
bounds.
The
heavy
center
line
reflects
the
central
tendency
across
all
1,000
curves.

Among
the
1,000
modeled
curves,
any
single
curve
describes
variability
in
occurrence
across
all
filtered
systems.
Steeper
curves
indicate
less
variability
from
system
to
system.

Variability
Uncertainty
Methods
1622/
23).
The
overall
result
of
this
Monte
Carlo
simulation
is
a
predicted
distribution
of
systems
assigned
to
each
LT2ESWTR
treatment
bin;
this
distribution
serves
as
input
to
the
cost
model.

Exhibit
4.13
Modeled
Source
Water
Cryptosporidium
Occurrence
ICR
Data
for
Filtered
Systems
Source:
USEPA
2003c.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
34
1e­
005
0.0001
0.001
0.01
0.1
1
10
Plant
Mean
Cryptosporidium
Concentration
(
Total
oocysts/
L)
0%
20%
40%
60%
80%
100%

Cumulative
Percentage
of
Plants
5th/
95th
%
tile
Median
1e­
005
0.0001
0.001
0.01
0.1
1
10
Plant
Mean
Cryptosporidium
Concentration
(
Total
oocysts/
L)
0%
20%
40%
60%
80%
100%

Cumulative
Percentage
of
Plants
5th/
95th
%
tile
Median
Exhibit
4.14
Modeled
Source
Water
Cryptosporidium
Occurrence
ICRSSM
Data
Exhibit
4.15
Modeled
Source
Water
Cryptosporidium
Occurrence
ICRSSL
Data
Source:
USEPA
2003c.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
35
1e­
005
0.0001
0.001
0.01
0.1
1
10
Plant
Mean
Cryptosporidium
Concentration
(
Total
oocysts/
L)
0%
20%
40%
60%
80%
100%

Cumulative
Percentage
of
Plants
ICR
Unfiltered
ICR
Filtered
SSM
SSL
Exhibit
4.16
Comparison
of
Modeled
Source
Water
Cryptosporidium
Occurrence
by
Data
Set,
Median
Curves
Only
Source:
USEPA
2003c
Exhibit
4.16
compares
the
central
tendency
of
the
cumulative
distributions
for
the
ICR,
ICRSSM,
and
ICRSSL
data
sets.
As
indicated
by
the
shape
of
the
curves,
the
filtered
plant
ICR
data
set
describes
a
source
water
occurrence
pattern
with
a
greater
frequency
of
high
oocyst
concentrations
than
either
the
ICRSSM
or
ICRSSL
system
data
sets.
The
relatively
shallower
"
steepness"
of
the
ICR
curve
and
the
resulting
lower
and
higher
limits
on
the
range
of
oocyst
concentrations
represented
in
the
distribution
implies
more
overall
variability
from
plant
to
plant
than
the
other
two
data
sets.
In
other
words,
in
addition
to
suggesting
a
greater
frequency
of
high
concentrations,
the
ICR
data
set
also
suggests
a
greater
frequency
of
lower
values
than
do
the
other
two
data
sets.
Exhibit
4.16
also
shows
for
reference
the
central
tendency
of
the
cumulative
distributions
for
unfiltered
systems
using
the
ICR
data
set.

Even
though
the
ICRSS
data
represent
higher
sample
volumes
and
better
recoveries,
the
ICR
distribution
includes
data
from
more
source
waters.
Each
of
these
three
distributions
represent
a
plausible
picture
of
the
national
occurrence
of
Cryptosporidium.
It
is
difficult
to
weigh
the
relative
importance
of
sample
recovery
versus
representativeness
of
the
sample
population.
Therefore,
all
three
occurrence
distributions
are
used
in
both
the
benefits
analyses
in
Chapter
5
and
the
cost
analyses
in
Chapter
6.
The
three
distributions
taken
together
encompass
a
range
of
values
that
help
characterize
the
actual
occurrence
of
Cryptosporidium
in
surface
waters
and
the
uncertainty
surrounding
it.

4.5.4
Finished
Water
Cryptosporidium
Occurrence
for
Filtered
Plants
Pre­
LT2ESWTR
finished
water
Cryptosporidium
concentrations
(
presented
in
this
section)
will
be
compared
to
predicted
Post­
LT2ESWTR
finished
water
levels
(
presented
in
Chapter
5)
to
assess
the
7
As
a
separate
part
of
predicting
Pre­
LT2ESWTR
baselines,
the
baseline
of
plants
and
systems
subject
to
LT2ESWTR
provisions
was
reduced
to
account
for
the
percentage
of
plants
that
are
predicted
to
have
installed
microfiltration/
ultrafiltration
(
MF/
UF)
before
the
implementation
of
LT2ESWTR.
Plants
with
MF/
UF
will
not
be
required
to
conduct
LT2ESWTR
monitoring
and
will
be
exempt
from
additional
Cryptosporidium
reduction
requirements.

Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
36
benefits
of
the
regulatory
alternatives.
As
with
the
treatment
characterization
presented
above,
the
finished
water
Cryptosporidium
occurrence
estimates
must
account
for
finished
water
improvements
predicted
to
result
from
implementation
of
other
existing
rules
or
rules
under
development
for
implementation
prior
to
LT2ESWTR.
This
section
describes
the
methodology
for
predicting
Cryptosporidium
finished
water
occurrence
based
on
the
treatment
in
place
following
the
IESWTR,
LT1ESWTR,
Stage
1
DBPR,
and
Stage
2
DPBR.

Compliance
with
the
IESWTR,
LT1ESWTR,
and
the
Stage
1
and
2
DBPRs
requires
or
will
require
some
systems
to
modify
their
treatment
processes
to
improve
performance
or
control
formation
of
DBPs.
Other
rules
are
not
expected
to
have
an
appreciable
impact
because
they
affect
mainly
ground
water
systems.
EPA
used
the
EAs
from
the
IESWTR
and
the
LT1ESWTR,
along
with
more
recent
plant
performance
data,
to
predict
the
number
of
systems
in
each
size
category
that
have
made
or
will
need
to
make
treatment
modifications
and
the
effectiveness
of
those
modifications.
(
These
rules
have
not
been
in
place
long
enough
for
all
plants
to
have
actually
implemented
modifications
and
to
have
gathered
information
on
their
effectiveness.)
The
IESWTR
and
the
LT1ESWTR
establish
filtration
requirements
EPA
believes
provide
finished
water
having
a
minimum
2
log
(
99
percent)
reduction
in
Cryptosporidium
concentrations
relative
to
source
water
levels.
Although
systems
are
required
to
meet
only
the
2
log
removal
target,
it
is
recognized
that
most
systems
will
achieve
greater
levels
of
Cryptosporidium
removal.
For
example,
some
systems
have
unit
processes
in
addition
to
conventional
treatment
that
provide
higher
Cryptosporidium
removals.
To
capture
the
different
levels
of
treatment,
EPA
estimated
a
range
of
Cryptosporidium
removals
for
calculating
finished
water
occurrence.
7
Exhibit
4.17
shows
the
range
of
Cryptosporidium
log
reduction
that
treatment
plants
are
expected
to
achieve
just
prior
to
implementation
of
the
LT2ESWTR.
To
account
for
systems
using
conventional
treatment
and
minimally
meeting
the
IESWTR
and
LT1ESWTR
effluent
turbidity
requirements
(
shown
in
the
top
half
of
Exhibit
4.17
as
the
standard
estimate),
the
low
end
of
the
range
is
set
at
2.0
log.
This
reduction
is
based
on
the
requirements
of
the
IESWTR
and
LT1ESWTR
(
which
specify
2.0
log
removal)
and
on
several
studies
that
show
plants
can
achieve
2
log
Cryptosporidium
treatment
credit
even
under
stressed
conditions.
These
studies
are
described
in
Chapter
7
of
the
Occurrence
and
Exposure
Assessment
(
USEPA
2003c).
As
described
in
section
4.5.1,
many
plants
are
expected
to
be
eligible
for
0.5
log
additional
treatment
credits
for
additional
unit
processes
or
for
achieving
effluent
turbidity
below
0.15
NTU.
For
these
systems,
the
range
of
performance
was
shifted
up
by
0.5
log
to
reflect
the
improved
performance,
as
shown
in
the
bottom
half
of
Exhibit
4.17.

Based
on
the
studies
cited
in
the
Occurrence
and
Exposure
Assessment,
large
systems
under
the
standard
estimate
are
thought
to
achieve
a
maximum
Cryptosporidium
log
reduction
of
4.5.
Many
factors
can
negatively
affect
the
log
reduction
systems
are
able
to
achieve,
and
small
systems
are
thought
to
be
more
significantly
impacted
by
these
factors.
For
example,
smaller
systems
typically
have
fewer
filters
than
large
systems.
One
consequence
is
that
a
single
filter
performing
poorly
is
more
likely
to
cause
poorer
overall
system
performance.
In
addition,
backwashing
a
filter
in
a
small
system
with
few
filters
is
more
likely
to
cause
hydraulic
fluctuations
that
could
result
in
poorer
performance
of
the
other
filters.
Small
systems
also
tend
to
have
less
automated
control
and
monitoring
equipment,
which
makes
controlling
temporary
aberrations
more
difficult.
For
this
reason,
the
maximum
log
reduction
small
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
37
systems
are
estimated
to
achieve
is
smaller
than
that
achieved
by
large
systems
(
for
the
standard
estimate,
3.5
log
vs.
4.5
log).
For
systems
eligible
for
an
additional
0.5
log
reduction
credit,
the
top
end
of
the
range
is
increased
to
4.0
log
for
small
systems
and
5.0
log
for
large
systems.

Exhibit
4.17
Predicted
Ranges
of
Cryptosporidium
Reduction
Pre­
LT2ESWTR
System
Size
(
Population
Served)
Exhibit
Range
of
Log
Reduction
Mode
 
Lower
End
Mode
 
Higher
End
Small
(<
10,000)
Standard
Estimate
4.18a
2.0
­
3.5
2.25
log
2.75
log
Large
(
$
10,000)
Standard
Estimate
4.18b
2.0
­
4.5
2.5
log
3.0
log
Small
(<
10,000)
Estimate
w/
0.5
Log
Removal
Credit
4.18c
2.5
­
4.0
2.75
log
3.25
log
Large
(
$
10,000)
Estimate
w/
0.5
Log
Removal
Credit
4.18d
2.5
­
5.0
3.0
log
3.5
log
Source:
Chapter
7,
LT2ESWTR
Occurrence
Assessment
(
USEPA
2003c).

The
log
reduction
values
at
the
low
and
high
ends
of
the
ranges
are
thought
to
be
the
exception
more
than
the
rule.
The
studies
cited
in
the
Occurrence
and
Exposure
Assessment
noted
median
and/
or
average
log
reductions
that
fall
in
the
middle
of
the
ranges
shown
in
Exhibit
4.17.
Although
little
specific
information
is
available
on
how
often
these
values
occur,
there
is
thought
to
be
a
central
tendency.
Therefore,
modes
of
2.5
to
3.0
log
were
chosen
for
the
distribution
of
log
reductions
for
large
systems
under
the
standard
estimate
(
see
second
row
of
Exhibit
4.17).
Values
in
between
2.5
and
3.0
log
are
thought
to
have
the
same
possibility
of
being
modes.

As
described
above,
small
systems
are
more
likely
to
have
filtration
performance
problems.
The
modes
at
small
systems
are,
therefore,
assumed
to
be
lower
than
for
large
systems
 
2.25
and
2.75
logs
under
the
standard
estimate
(
first
row
of
Exhibit
4.17).
The
bottom
half
of
Exhibit
4.17
shows
the
estimated
modes
for
large
and
small
systems
for
systems
getting
a
0.5
log
reduction
credit,
which
was
also
applied
to
the
modes.

An
upper
and
lower
bound,
combined
with
a
selected
mode,
defines
a
triangular
distribution.
In
Exhibit
4.18,
the
minimum
and
maximum
of
the
range
define
the
base
of
the
triangle
and
the
mode
defines
the
top.
Each
graph
in
Exhibit
4.18
corresponds
to
the
range
of
one
of
the
rows
in
Exhibit
4.17.
Two
triangles
are
shown
for
each
range,
illustrating
the
lower
and
upper
modes.
The
triangles
defined
by
all
the
modes
in
between
are
not
shown,
but
are
indicated
by
the
two­
way
arrow
on
each
graph.
Each
possible
triangle,
of
which
there
is
an
infinite
number,
is
considered
equally
likely
to
represent
the
log
reduction
distribution
for
a
given
system
size
category
and
treatment
scenario.
For
example,
Exhibit
4.18a
shows
two
triangular
distributions
for
small
systems
under
the
standard
treatment
estimate,
which,
as
described
in
Exhibit
4.17,
have
a
range
of
2.0
to
3.5
log
reduction
and
modes
of
2.25
and
2.75
log.
For
the
same
group
of
systems,
there
are
an
infinite
number
of
triangles
with
the
same
range
and
with
modes
between
2.25
and
2.75
log.
Taken
together,
the
collection
of
all
the
distributions
for
a
given
group
of
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
38
systems
(
e.
g.,
small
systems
without
a
0.5
log
reduction
credit)
reflects
the
uncertainty
about
the
true
Cryptosporidium
log
reduction.

These
triangular
distributions
are
used
in
benefit
and
cost
modeling.
For
a
given
model
iteration
for
a
given
system
size,
the
model
first
decides
whether
systems
get
the
0.5
log
reduction
credit
(
36
percent
of
small
systems,
55
percent
of
medium
systems,
and
58
percent
of
large
systems
are
assumed
to
qualify).
Then
it
randomly
selects
a
mode
from
the
appropriate
set
of
triangular
distributions.
From
the
individual
distribution
associated
with
that
selected
mode,
the
model
randomly
picks
100
log
removal
values
from
that
distribution
and
uses
these
log
reduction
values
to
predict
finished
water
Cryptosporidium
concentrations.
This
process
is
repeated
for
250
modes.

Exhibit
4.19
shows
an
example
of
the
finished
water
concentration
distributions
generated
with
a
modeling
process
similar
to
that
described
above
(
the
curves
in
4.19
were
generated
choosing
only
250
points
from
each
triangular
distribution).
These
distributions
were
generated
using
the
same
process
as
the
source
water
Cryptosporidium
concentrations
for
different
data
sets
in
Exhibit
4.16
so
that
the
differences
between
them
would
be
directly
comparable.
Because
disinfection
in
unfiltered
systems
is
expected
to
have
a
negligible
effect
on
Cryptosporidium
concentrations,
finished
water
concentrations
in
unfiltered
plants
are
assumed
to
be
identical
to
source
water
concentrations
(
compare
Exhibits
4.16
and
4.19).
For
filtered
plants,
however,
modeled
source
water
concentrations
were
higher
than
those
for
unfiltered
plants
(
see
Exhibit
4.16),
but
modeled
finished
water
concentrations
were
well
below
those
same
unfiltered
plant
concentrations.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
39
2
3
4
5
Log
Removals
Relative
Probability
Mode
=
3.0
Mode
=
2.5
Exhibit
4.18b
Distribution
of
Cryptosporidium
Reduction
in
Large
Systems
Pre­
LT2ESWTR,
Standard
Estimate
2
3
4
5
Log
Removals
Relative
Probability
Mode
=
3.5
Mode
=
3.0
Exhibit
4.18d
Distribution
of
Cryptosporidium
Reduction
in
Large
Systems
Pre­
LT2ESWTR,
Estimate
With
0.5
Log
Reduction
Credit
2
3
4
5
Log
Removals
Relative
Probability
Mode
=
3.25
Mode
=
2.75
Exhibit
4.18c
Distribution
of
Cryptosporidium
Reduction
in
Small
Systems
Pre­
LT2ESWTR,
Estimate
With
0.5
Log
Reduction
Credit
2
3
4
5
Log
Removals
Relative
Probability
Mode
=
2.75
Mode
=
2.25
Exhibit
4.18a
Distribution
of
Cryptosporidium
Reduction
in
Small
Systems
Pre­
LT2ESWTR,
Standard
Estimate
Source:
USEPA
2003c.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
40
0%
20%
40%
60%
80%
100%

1.00E­
10
1.00E­
09
1.00E­
08
1.00E­
07
1.00E­
06
1.00E­
05
1.00E­
04
1.00E­
03
1.00E­
02
1.00E­
01
1.00E+
00
1.00E+
01
Finished
Water
Cryptosporidium
(
Oocysts/
L)
Cumulative
Probability
ICR
FIL
­
SM
ICRSSL
­
SM
ICRSSM
­
SM
ICR
UNF
­
SM
0%
20%
40%
60%
80%
100%

1.00E­
10
1.00E­
09
1.00E­
08
1.00E­
07
1.00E­
06
1.00E­
05
1.00E­
04
1.00E­
03
1.00E­
02
1.00E­
01
1.00E+
00
1.00E+
01
Finished
Water
Cryptosporidium
(
Oocysts/
L)
Cumulative
Probability
ICR
FIL
­
LG
ICRSSL
­
LG
ICRSSM
­
LG
ICR
UNF­
LG
Exhibit
4.19a
Predicted
Finished
Water
Cryptosporidium
Occurrence
Pre­
LT2ESWTR,
Small
Systems
Exhibit
4.19b
Predicted
Finished
Water
Cryptosporidium
Occurrence
Pre­
LT2ESWTR,
Large
Systems
Source:
USEPA
2003c
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
41
4.5.5
Comparison
of
EPA
Finished
Water
Cryptosporidium
Estimates
with
Aboytes
et
al.
(
2000)

A
study
by
Aboytes
et
al.
(
2000)
provides
an
alternative
perspective
on
Cryptosporidium
occurrence
in
finished
drinking
water
and
the
efficacy
of
treatment.
This
study
involved
collecting
100­
L
finished
water
samples
on
a
monthly
basis
from
80
surface
water
utilities.
Samples
were
analyzed
for
infectious
Cryptosporidium
parvum
with
a
cell
culture­
PCR
method.
The
objective
of
the
study
was
to
"
assess
the
adequacy
of
treatment
to
protect
against
infectious
Cryptosporidium
in
drinking
water."
All
utilities
in
the
study
were
enrolled
in
the
Partnership
for
Safe
Water,
a
voluntary
cooperative
program
that
seeks
to
optimize
treatment
plant
performance.
Most
samples
had
turbidity
below
0.1
NTU
and
all
were
below
0.3
NTU,
the
standard
set
by
the
IESWTR.
Among
1,674
samples
of
100
L
each,
24
were
positive
for
infectious
C.
parvum
(
LeChevallier
2001).
The
authors
determined
an
average
CC­
PCR
recovery
efficiency
of
32.3
percent.
Hence,
if
it
is
assumed
that
one
infectious
oocyst
accounted
for
each
positive
sample,
and
the
oocyst
count
is
adjusted
for
average
recovery,
these
results
produce
a
mean
concentration
of
infectious
oocysts
of
4.4
×
10­
4
oocysts/
L,
or
0.044
oocysts/
100
L.

To
compare
results
from
Aboytes
et
al.
with
EPA
finished
water
Cryptosporidium
estimates
based
on
results
from
the
ICR
and
ICRSS,
it
is
necessary
to
consider
the
fraction
of
oocysts
that
are
infectious.
Because
oocysts
lose
viability
in
the
environment,
it
is
expected
that
infectious
oocysts
are
only
a
fraction
of
the
total
number
of
oocysts
in
a
water
sample.
While
the
CC­
PCR
method
registers
only
infectious
oocysts,
the
ICR
Method
and
EPA
Methods
1622/
23
count
total
oocysts
without
regard
to
whether
they
are
viable
and
infectious.
To
estimate
the
fraction
of
oocysts
that
may
be
infectious,
EPA
evaluated
a
study
by
LeChevallier
et
al.
(
2003)
that
analyzed
several
hundred
source
water
samples
from
six
utilities
using
both
the
CC­
PCR
method
and
Method
1623.
Oocysts
were
detected
in
60
of
593
samples
(
10.1
percent)
by
Method
1623
and
infectious
oocysts
were
detected
in
22
of
560
samples
(
3.9
percent
(
by
the
CC­
PCR
procedure.
Recovery
efficiencies
for
the
two
methods
were
similar.
According
to
the
authors,
these
results
suggest
that
approximately
37
percent
of
the
oocysts
detected
by
Method
1623
were
viable
and
infectious.
Based
on
these
results,
as
well
as
consideration
of
the
structure
of
counted
oocysts,
EPA
assumes
that
the
fraction
of
oocysts
that
is
infectious
may
average
from
15
to
25
percent
for
the
ICR
and
30
to
50
percent
for
the
ICRSS
(
described
in
Section
5.2.4).

If
the
estimates
of
mean
large
plant
finished
water
oocyst
concentrations
from
Exhibit
4.19b
are
multiplied
by
40
percent
to
adjust
for
the
fraction
of
oocysts
that
are
infectious,
the
mean
finished
water
concentrations
of
infectious
oocysts
are
as
follows:

ICR
Mean
=
6.1
×
10­
5;
ICRSSM
Mean
=
2.5
×
10­
5;
ICRSSL
=
1.4
×
10­
5
(
oocysts/
L)

Thus,
the
mean
finished
water
infectious
oocyst
level
reported
in
Aboytes
et
al.
of
4.4
×
10­
4
oocysts/
L
is
a
factor
of
7
greater
than
the
EPA
mean
estimate
based
on
the
ICR
and
a
factor
of
30
greater
than
the
ICRSSL
estimate.
While
the
reason
for
this
significant
discrepancy
is
unknown,
it
cannot
be
fully
attributed
to
potential
error
in
factors
such
as
the
fraction
of
oocysts
that
are
infectious.
Rather,
it
may
indicate
that
the
ICR
and
ICRSS
underestimate
source
water
Cryptosporidium
occurrence
and/
or
that
EPA
has
overestimated
treatment
efficacy,
as
discussed
below.

Unfortunately,
source
water
data
are
not
available
for
plants
in
the
Aboytes
et
al.
study,
so
it
is
not
possible
to
directly
determine
their
average
treatment
efficiency.
However,
removal
efficiency
may
be
estimated
based
on
other
survey
data.
The
ICRSS
was
conducted
during
a
time
frame
similar
to
Aboytes
et
al.,
and
the
filtration
and
separation
steps
used
in
the
CC­
PCR
method
of
Aboytes
et
al.
are
similar
to
those
in
Method
1622/
23
used
in
the
ICRSS.
Consequently,
the
ICRSS
source
water
data
may
be
somewhat
comparable
to
the
Aboytes
et
al.
finished
water
data.
The
mean
source
water
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
42
Cryptosporidium
concentration
of
plants
in
the
ICRSS,
adjusted
for
average
recovery
of
43
percent,
was
0.14
oocysts/
L.
If
this
value
is
multiplied
by
40
percent
as
an
estimate
of
the
fraction
of
oocysts
that
are
infectious,
the
result
is
a
mean
source
water
concentration
of
0.056
infectious
oocysts/
L.
If
this
were
the
mean
source
water
concentration
in
the
Aboytes
et
al.
study,
then
the
plants
in
that
survey
achieved
an
average
removal
of
1.7
log
to
produce
the
mean
finished
water
concentration
of
0.0011
oocysts/
L
that
was
measured.
For
the
plants
in
the
Aboytes
et
al.
survey
to
have
achieved
a
mean
oocyst
removal
of
2.5
log,
which
was
the
lowest
mean
removal
assumed
in
EPA
estimates,
the
mean
source
water
oocyst
concentration
would
have
to
have
been
over
twice
that
measured
during
the
ICRSS.
Either
hypothesis
indicates
that
EPA
may
have
underestimated
the
risk
from
Cryptosporidium
in
drinking
water
by
underestimating
finished
water
oocyst
levels.

4.5.6
Predicted
Bin
Classification
for
Filtered
Plants
Under
the
LT2ESWTR,
filtered
plants
will
be
assigned
to
a
treatment
bin
based
on
the
results
of
source
water
monitoring
for
Cryptosporidium.
Each
bin
is
defined
by
a
range
of
oocyst
concentrations
and
determines
the
amount
of
treatment
each
plant
must
provide.
Using
the
modeled
source
water
occurrence
distributions
described
in
section
4.5.3,
EPA
predicted
the
percentage
of
filtered
plants
that
will
fall
into
each
bin.
Appendix
B
presents
the
probability
functions
for
the
occurrence
distributions
used
to
evaluate
bin
classification.
Note
that
the
tables
in
Appendix
B
estimate
the
percentage
based
on
a
prediction
of
the
test
measurement
mean,
which
is
the
estimated
underlying
source
water
concentration
adjusted
for
sampling
and
laboratory
analysis
error.
Bins
are
based
on
the
predicted
lab
results,
not
estimates
of
the
"
true"
concentration.
Exhibit
4.20
presents
the
central
tendency
of
plant
binning
based
on
a
Monte
Carlo
evaluation
of
the
probability
distributions
for
the
three
data
sets
of
Cryptosporidium
occurrence
for
the
Preferred
Regulatory
Alternative.
Note
that
the
estimate
based
on
the
ICR
data
set
is
the
most
conservative
of
the
three
filtered
plant
data
sets;
the
predicted
percentage
of
plants
requiring
treatment
is
largest
for
the
ICR.

The
percentages
shown
in
Exhibit
4.20
were
used
to
determine
costs
of
installing
treatment
for
each
occurrence
distribution.
Plants
were
expected
to
incur
different
costs
depending
on
the
bin
to
which
they
were
assigned.

Exhibit
4.20
Predicted
System
Binning
for
Preferred
Alternative,
Based
on
Central
Tendency
of
Cryptosporidium
Occurrence
Data
Set
No
Action
1
Log
Bin
(
0.075
 
1
oocysts/
L)
2
Log
Bin
(
1
 
3
oocysts/
L)
2.5
Log
Bin
(>
3.0
oocysts/
L)
Total
ICR
65.4%
27.2%
4.4%
3.2%
100%

ICRSSM
72.9%
25.4%
1.4%
0.4%
100%

ICRSSL
77.7%
21.8%
0.5%
0.06%
100%

Note:
Numbers
may
not
add
to
exactly
100
percent
due
to
rounding.
Bin
assignment
is
based
on
the
highest
running
annual
average
(
RAA)
of
concentrations
of
24
influent
samples
taken
over
2
years.

Source:
Appendix
B,
Exhibit
B.
8.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
43
4.6
Baseline
for
Uncovered
Finished
Water
Reservoirs
Number
of
Reservoirs
The
LT2ESWTR
baseline
for
uncovered
finished
water
reservoirs
is
presented
in
Exhibit
4.20.
Data
on
uncovered
finished
water
reservoirs
were
provided
by
EPA
regional
offices
based
on
information
on
States
in
their
region
and
do
not
include
reservoirs
that
are
scheduled
to
be
covered
or
taken
off­
line.
Most
uncovered
finished
water
reservoirs
are
located
in
only
a
few
States;
specifically
California,
New
York,
New
Jersey,
and
Oregon,
and
Puerto
Rico.
The
largest
is
in
Southern
California.
Although
there
are
only
138
uncovered
finished
water
reservoirs,
there
is
a
limited
amount
of
information
on
them.
The
information
provided
differed
by
State
and
region.
Most
systems
only
reported
reservoir
volume
or
surface
area.
Very
few
reported
both
parameters,
forcing
EPA
to
make
assumptions
on
volume
to
calculate
surface
area
(
surface
area
was
later
used
to
determine
costs
for
covering
or
treating
reservoirs).
Ownership
and
population
served
by
each
reservoir
were
also
often
unclear.

For
this
analysis,
the
reservoirs
are
categorized
according
to
usable
volume
(
in
millions
of
gallons,
or
MG).
The
mean
volume,
shown
in
the
last
column
of
Exhibit
4.21,
is
the
average
usable
volume
of
all
reservoirs
in
a
volume
category
and
is
used
to
estimate
costs
in
Chapter
6.

Exhibit
4.21
Baseline
Numbers
of
Uncovered
Finished
Water
Reservoirs
Size
Category
(
Usable
Volume
in
MG)
Number
of
Reservoirs
Mean
Usable
Volume
(
MG)

0
­
0.1
25
0.01
>
0.1
­
1
7
0.43
>
1
­
5
44
2.59
>
5
­
10
12
8.55
>
10
­
20
10
15.32
>
20
­
40
9
28.68
>
40
­
60
4
52.99
>
60
­
80
4
71.44
>
80
­
100
6
93.57
>
100
­
150
6
125.34
>
150
­
200
2
187.00
>
200
­
250
4
210.75
>
250
­
1,000
4
781.50
>
1,000
1
3,313.80
Total
138
­

Source:
EPA
regions.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
44
System
Sizes
for
Uncovered
Reservoirs
In
order
to
apportion
the
costs
to
PWSs
of
treating
water
from
uncovered
finished
water
reservoirs
or
of
covering
them,
EPA
estimated
the
size
of
the
systems
that
the
reservoirs
serve.
To
do
this,
the
average
daily
flow
for
each
system
size
(
USEPA
2001c)
was
used.
The
average
daily
flow
was
then
substituted
into
a
regression
equation
from
the
Model
Systems
Report
(
USEPA
2000b)
that
predicts
flow
based
on
population
(
see
Exhibit
4.4).
The
equation
yielded
the
population
that
the
reservoir
would
be
expected
to
serve
if
the
entire
system
flow
passed
through
the
reservoir.
Many
reservoirs
are
only
one
of
several
in
a
system,
and
in
such
cases,
there
is
no
basis
to
determine
what
percentage
of
the
full
flow
might
pass
through
a
single
reservoir.
Exhibit
4.22
shows
the
distribution
of
uncovered
filtered
water
reservoirs
by
system
size.
Once
size
categories
are
determined,
the
number
of
systems
is
multiplied
by
the
average
population
per
system
for
that
size
category
and
divided
by
2.59,
the
number
of
people
per
household
(
U.
S.
Census
Bureau
2001),
to
obtain
the
number
of
households
served
by
such
reservoirs.

Exhibit
4.22
Baseline
Number
of
Uncovered
Finished
Water
Reservoirs
in
Systems
of
Each
Size
Category
System
Size
(
Population
Served)
Number
of
Reservoirs
<
100
25
101
­
500
0
501
­
1,000
0
1,001
­
3,300
0
3,301
­
10,000
7
10,001
­
50,000
66
50,001
­
100,000
9
100,001
­
1
M
30
>
1
M
1
Total
138
Source:
Exhibit
4.21
and
regression
equation
from
Model
Systems
Report
(
USEPA
2000b).

Uncovered
Reservoir
Surface
Area,
Average
Daily
Flow,
and
Design
Flow
The
costs
for
covering
reservoirs
are
based
largely
on
the
surface
area
of
the
reservoir,
while
costs
for
treating
the
discharge
for
4
log
virus
inactivation
are
based
on
flow
through
the
reservoir.
As
with
technologies
used
for
Cryptosporidium
removal,
the
design
flow
and
average
daily
flow
are
used
to
estimate
capital
and
O&
M
costs
of
reservoir
treatment,
respectively.
The
purpose
of
this
section
is
to
derive
mean
surface
area,
average
daily
flow,
and
design
flow
for
each
size
category
of
uncovered
finished
water
reservoirs.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
45
For
some
cases,
EPA
regions
provided
the
surface
area
for
individual
uncovered
finished
water
reservoirs.
Where
this
information
was
not
available,
engineering
assumptions
were
necessary.
To
calculate
surface
area
if
not
known,
a
representative
reservoir
depth
of
25
feet
was
assumed
based
on
industry
consultation.
The
mean
surface
area
for
each
size
category
presented
in
Exhibit
4.23
is
the
average
surface
area
of
all
reservoirs
in
the
size
category,
whether
based
on
actual
data
or
based
on
volume
with
an
assumed
25­
foot
depth.

Hydraulic
residence
time,
defined
as
the
time
water
spends
inside
a
reservoir,
can
be
used
in
conjunction
with
volume
to
estimate
average
daily
flow
through
a
reservoir.
Hydraulic
residence
times
in
finished
water
storage
reservoirs
vary
greatly
among
systems
and
seasons.
The
shortest
times
are
often
in
the
Summer,
while
the
longest
may
be
during
the
lower­
demand
periods
in
the
Winter.
Because
water
systems
strive
to
maintain
a
certain
volume
of
storage
in
the
distribution
system
for
emergencies
such
as
fire,
residence
times
can
be
as
great
as
several
weeks.
Within
the
last
several
years
water
systems
have
focused
more
on
water
quality
in
the
distribution
system
and
have
decreased
average
residence
time,
striving
to
"
turn
over"
finished
water
in
storage
facilities
on
a
regular
basis.
Considering
these
factors,
typical
hydraulic
residence
times
of
1
to
3
days
were
used
in
conjunction
with
volume
to
estimate
average
daily
flow
for
reservoirs
up
to
100
million
gallons
in
size,
as
presented
in
Exhibit
4.21.
For
the
very
largest
reservoirs,
longer
hydraulic
residence
times
are
more
likely;
residence
times
as
high
as
21
days
are
assumed.

Design
flow,
used
to
estimate
capital
costs,
represents
the
maximum
flow
exiting
the
reservoir.
Finished
water
storage
facilities
respond
to
daily
water
demand
fluctuations
in
the
distribution
system
in
order
to
maintain
a
more
constant
flow
at
the
treatment
plant.
Therefore,
maximum
or
design
flow
from
a
reservoir
may
be
higher
than
the
average
flow
at
the
treatment
plant.
Peak
hourly
flows
in
a
distribution
system
have
been
estimated
as
three
times
the
average
flow
for
hydraulic
modeling
purposes
(
Lindeburg,
1997).
Therefore,
the
design
flow
presented
in
Exhibit
4.23
is
estimated
to
be
three
times
the
average
daily
flow.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
46
Exhibit
4.23
Surface
Area
and
Flows
for
Uncovered
Finished
Water
Reservoirs
Size
Category
(
Usable
Volume,
MG)
Number
of
Reservoirs
Mean
Volume
(
MG)
Surface
Area
(
sq.
ft.)
Estimated
Hydraulic
Residence
Time
(
days)
Average
Daily
Flow
(
mgd)
Design
Flow
(
mgd)

A
B
C
D
E
F
0
­
0.1
25
0.01
176
1
0.01
0.04
>
0.1
­
1
7
0.43
2,952
1
0.43
1.30
>
1
­
5
44
2.59
14,127
1
2.59
7.78
>
5
­
10
12
8.55
45,703
2
4.27
12.82
>
10
­
20
10
15.32
84,165
3
5.11
15.32
>
20
­
40
9
28.68
213,890
3
9.56
28.68
>
40
­
60
4
52.99
283,353
3
17.66
52.99
>
60
­
80
4
71.44
365,472
3
23.81
71.44
>
80
­
100
6
93.57
517,699
3
31.19
93.57
>
100
­
150
6
125.34
670,246
4
31.34
94.01
>
150
­
200
2
187.00
999,934
4
46.75
140.25
>
200
­
250
4
210.75
1,126,931
4
52.69
158.06
>
250
­
1,000
4
781.50
3,482,390
14
55.82
167.46
>
1,000
1
3,313.80
7,666,876
21
157.80
473.40
Note:
Data
shown
are
rounded.

Sources:
[
A]
EPA
regions.
[
B]
EPA
regions.
[
C]
EPA
regions
for
some
reservoirs.
Surface
area,
if
not
provided
for
an
individual
reservoir,
is
estimated
based
on
data
on
volume
(
from
EPA
regions)
and
an
assumed
depth
of
25
feet.
[
D]
EPA
regions.
[
E]
Average
daily
flow
is
equal
to
reservoir
volume
(
Column
B)
divided
by
hydraulic
residence
time
(
Column
C).
[
F]
Design
flow
is
three
times
the
average
daily
flow.

4.7
Households
Incurring
Costs
Due
to
the
LT2ESWTR
Estimating
the
number
of
households
in
systems
affected
by
the
LT2ESWTR
is
necessary
to
derive
a
national
distribution
of
the
rule's
annual
costs.
Exhibit
4.24
shows
the
various
sets
of
households
and
their
relationship
to
one
another.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
47
Implementation
Costs
All
Households
Monitoring
Costs
Uncovered
Finished
Water
Reservoirs
Treatment
Costs
Unfiltered
Treatment
+
Monitoring
+
Implementation
Filtered
Treatment
+
Monitoring
+
Implementation
Unfiltered
Treatment
+
Uncovered
Reservoir
+
Monitoring
+
Implementation
Filtered
Treatment
+
Uncovered
Reservoir
+
Monitoring
+
Implementation
Uncovered
Reservoir
+
Monitoring
+
Implementation
Monitoring
+
Implementation
Implementation
Exhibit
4.24
Universe
of
Households
Affected
by
Rule
Provisions
Exhibit
4.25
presents
estimates
of
the
number
of
households
served
by
surface
and
GWUDI
systems,
further
subdivided
by
those
subject
to
rule
provisions.
The
estimates
were
derived
by
dividing
the
total
population
served
by
CWSs
subject
to
each
rule
provision
per
size
category
by
an
average
number
of
people
per
household
of
2.59
(
U.
S.
Census
Bureau
2001).
Only
CWS
population
is
converted
to
household
estimates
because
it
is
assumed
that
only
CWSs
serve
residential
customers.

Exhibit
4.27
presents
the
estimated
mean
water
usage
rates
(
in
gallons
per
year)
per
household
for
each
system
size
category.
These
estimates
are
based
on
total
residential
consumption
and
number
of
residential
connections
from
the
Baseline
Handbook.
The
consumption
in
the
two
smallest
size
categories
was
adjusted
based
on
analyses
of
CWSS
data.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
48
System
Size
(
Population
Served)
Households
Paying
Implementation
Costs
Households
Paying
Monitoring
Costs
Households
Paying
Treatment
Costs
for
Unfiltered
Systems
Households
Paying
Treatment
Costs
for
Filtered
Systems
A
B
C
D
<
100
9,115
8,787
29
9,086
101­
500
78,220
75,404
519
77,702
501­
1,000
123,007
118,578
2,092
120,915
1,001­
3,300
788,354
759,974
8,603
779,752
3,301­
10,000
2,319,316
2,235,821
57,915
2,261,401
10,001­
50,000
9,752,041
9,716,319
136,507
9,615,534
50,001­
100,000
6,436,207
6,412,631
80,083
6,356,124
100,001­
1
Million
26,841,790
26,743,469
323,995
26,517,795
>
1Million
19,468,929
19,397,614
3,922,990
15,545,939
Total
65,816,980
65,468,597
4,532,732
61,284,247
System
Size
(
Population
Served)
Households
Paying
Uncovered
Finished
Reservoir
Costs
<
100
574
101­
500
0
501­
1,000
0
1,001­
3,300
0
3,301­
10,000
13,469
10,001­
50,000
541,715
50,001­
100,000
133,227
100,001­
1
Million
2,002,187
>
1Million
320,762
Total
3,011,934
Exhibit
4.25
Baseline
Numbers
of
Households
Incurring
Costs
Notes:
All
unfiltered
systems
will
incur
treatment
costs.
Not
all
filtered
systems
shown
above
will
incur
treatment
costs
 
the
households
shown
in
Column
D
represent
all
systems
that
conducted
monitoring
less
systems
that
are
predicted
to
have
installed
MF/
UF
to
comply
the
the
Stage
2
DBPR.
Some
of
these
systems
will
be
assigned
to
the
"
no
action"
bin
and
will
not
incur
treatment
costs.

Sources:
[
A]
SDWIS
population
from
unlinked
inventory
(
USEPA
2000f)
/
2.59
people
per
household
(
U.
S.
Census
Bureau
2001).
[
B]
Implementation
baseline
(
Column
A)
*
percentage
of
systems
subject
to
monitoring
requirements
(
1
­
Exhibit
4.11,
Column
C).
[
C]
Treatment
baseline
population
for
unfiltered
systems
(
Exhibit
4.5,
Column
D)
/
2.59
people
per
household.
[
D]
Treatment
baseline
population
for
filtered
systems
(
Exhibit
4.11,
Column
I)
/
2.59
people
per
household.

Exhibit
4.26
Households
Paying
Treatment
Costs
for
Uncovered
Reservoirs
Source:
Appendix
I.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
49
Exhibit
4.27
Mean
Household
Water
Usage
Rates
by
System
Size
System
Size
(
Population
Served)
Mean
Household
Water
Usage
Rate
(
gal/
yr)

<
100
83,000
101­
500
83,000
501­
1,000
104,000
1,001­
3,300
87,000
3,301­
10,000
97,000
10,001­
50,000
109,000
50,001­
100,000
119,000
100,001­
1
Mil
125,000
>
1
Mil
125,000
Source:
USEPA
2001c,
modified
for
systems
serving
500
or
fewer
people
based
on
CWSS
data.

4.8
Summary
of
Uncertainties
in
Development
of
LT2ESWTR
Baselines
Uncertainties
in
this
baseline
analysis
could
result
in
either
an
overestimate
or
underestimate
of
the
costs
or
benefits
presented
in
Chapters
5
and
6.
Exhibit
4.28
below
presents
a
summary
of
these
issues
and
an
estimate
of
the
effects
that
each
source
of
uncertainty
may
have
on
subsequent
analyses.
Note
that,
in
many
cases,
assumptions
made
in
this
baseline
will
overestimate
both
costs
and
benefits;
however,
costs
are
overestimated
in
more
cases
than
are
benefits.
Economic
Analysis
for
the
LT2ESWTR
Proposal
June
2003
4­
50
Exhibit
4.28
Summary
of
Uncertainties
Affecting
LT2ESWTR
Baseline
Estimates
Assumption
Section
with
Discussion
of
Uncertainty
Effect
on
Benefit
Estimate
Effect
on
Cost
Estimates
Underestimate
Overestimate
Underor
Over­
Estimate
Underestimate
Overestimate
Underor
Over­
Estimate
Uncertainty
in
baseline
data
inputs
(
SDWIS,
ICR,
and
ICRSS
data)
4.3.2
4.4.2
X
X
Point
estimates
instead
of
distributions
for
population
and
flow
4.3.3
X
X
CWS
flow
equations
for
NTNCWSs
4.3.3
X
X
No
unfiltered
plants
have
advanced
disinfection
4.4.1
X
X
Filtered
plants
do
not
get
credit
for
ozone
and
ClO2
from
Stage
2
4.5.1
X
Predicted
Pre­
LT2ESWTR
Cryptosporidium
removal
using
triangular
distributions
(
with
uncertain
modes)
and
log
reduction
achieved
4.5.1
X
X
Uncertainty
in
baseline
surface
area
and
flows
for
uncovered
finished
water
reservoirs
4.6
X
X
Note:
The
uncertainties
associated
with
some
assumptions
are
discussed
in
more
detail
in
Chapters
5
and
6,
and
so
the
summaries
of
those
assumptions
are
reserved
until
the
ends
of
those
chapters.
Those
key
assumptions
include
the
occurrence
in
source
water
and
modeling
of
occurrence
in
finished
water;
risk
parameters,
such
as
infectivity
and
the
percent
of
viable
oocysts;
and
binning
assignments.
