!£ COST EVALUATION OF POINT-OF-USE
AND POINT-OF-ENTRY TREATMENT UNITS
FQR SMALL SYSTEMS: COST ESTIMATING
TOOL AND USER GUIDE
Office of Water (4607M) EPA 815-B-07-001 April 2007 www.epa.gov/safewater
-------
Acknowledgements
The U.S. Environmental Protection Agency wishes to thank the peer review participants who
provided comments and suggestions on a draft version of the Point-of-Use (POU) and Point-of-
Entry (POE) model and documentation: Dr. Regu P. Regunathan (Regunathan & Associates), Dr.
Jerry Lowry (Lowry Environmental Engineering, Inc.), Joseph Cotruvo (Joseph Cotruvo &
Associates), and Glen Latimer (Kinetico).
Disclaimer
This document provides a description of a costing tool developed to assist stakeholders with
estimating costs for a centrally managed POU or POE strategy to comply with drinking water
regulations. This document does not establish regulatory requirements, nor is it a regulation,
itself. With respect to the costing tool, neither the U.S. Environmental Protection Agency nor any
of its employees makes any warranty, express or implied, regarding the cost estimates generated
using this tool. Its cost outputs do not represent formal offers to provide equipment or services
for products or services included in the price tables. In addition, mention of specific products
does not constitute an endorsement by either the U.S. Environmental Protection Agency or any
of its employees.
-------
Table of Contents
1. Introduction 1
1.1 Purpose 1
1.2 Document Organization 1
2. Cost Model Description 3
2.1 Model Inputs - Input and User Inputs Worksheets 4
2.1.1 Target Contaminant and Treatment Technology 4
2.1.2 Service Population 5
2.2 Capital Assumptions Worksheet 6
2.2.1 Equipment Installation 7
2.2.2 Public Education Program 8
2.2.3 Initial Water Quality Monitoring 10
2.2.4 Indirect Assumptions 11
2.2.5 Other Capital Assumptions 12
2.3 O&M Assumptions Worksheet 13
2.3.1 Equipment Replacement and Maintenance 14
2.3.2 Education Updates 14
2.3.3 Monitoring 15
2.4 Control Worksheet 15
2.5 Price Tables Worksheet 15
2.5.1 POU and POE Device Prices 15
2.5.2 Labor Costs 17
2.5.3 Monitoring Costs 17
2.5.4 Other Annual Costs 18
2.6 Output Worksheet 18
2.7 References 19
3. Cost Model Operation 21
3.1 Getting Started 21
3.2 User-Defined Model Inputs 22
3.2.1 Target Contaminant 22
-------
3.2.2 Treatment Technology 23
3.2.3 System Size 23
3.2.4 Additional User-Defined Inputs 25
3.3 Understanding the Output Worksheet 30
-------
List of Exhibits
Exhibit 2-1. Cost Estimates Generated by the POU/POE Model 3
Exhibit 2-2. Summary of Contaminants and Treatment Technologies in the POU/POE Model ... 5
Exhibit 2-3. Summary of Design Assumptions for Capital Costs with Example Default Values
for Two POU Technologies 7
Exhibit 2-4. Indirect Cost Multipliers 12
Exhibit 2-5. Summary of Design Assumptions for O&M Costs with Example Default Values for
Two POU Technologies 13
Exhibit 2-6. Example of Equipment Replacement for a Purchased RO Device 14
Exhibit 2-7. Example of Unit Prices for POU Adsorptive Media NSF53 Equipment 16
Exhibit 2-8. Average Cost Data for Laboratory Analyses 17
Exhibit 2-9. Costs of Shipping Samples Using U.S. Postal Service 18
Exhibit 2-10. Example of Cost Details Provided on the Output worksheet 19
Exhibit 3-1. Step 1: Selecting the Operating Mode 21
Exhibit 3-2. Step 1: Additional Instruction for User-Defined System Mode 22
Exhibit 3-3. Step 2: Select a Target Contaminant Using the Contaminant Drop-Down List 22
Exhibit 3-4. Step 3: Select a Treatment Technology Using the Technology Drop-Down List.... 23
Exhibit 3-5. Step 4: Selecting System Size Option and Entering Size Data 24
Exhibit 3-6. Step 4: Selecting the Household Connections Option for System Size Information 24
Exhibit 3-7. Cost Summary and Link to Cost Breakdown 25
Exhibit 3-8. Step 5: Enter User-Defined Values to Replace Model Default Values 25
Exhibit 3-9. List of POU/POE Model Parameters for Equipment Costs 26
Exhibit 3-10. List of POU/POE Model Parameters for Initial Public Education Costs 26
Exhibit 3-11. List of POU/POE Model Parameters for Initial Monitoring Costs 27
Exhibit 3-12. List of POU/POE Model Parameters for Indirect Capital Costs 28
Exhibit 3-13. List of POU/POE Model Parameters for Equipment Maintenance Labor Costs.... 28
Exhibit 3-14. List of POU/POE Model Parameters for Annual Educational Costs 29
Exhibit 3-15. List of POU/POE Model Parameters for Annual Monitoring Costs 29
Exhibit 3-16. List of POU/POE Model Parameters for Annual Equipment Replacement Costs . 30
Exhibit 3-17. Example of Cost Details Provided on the Output Worksheet 31
in
-------
Acronyms
ANSI American National Standards Institute
BLS Bureau of Labor Statistics
GAC granular activated carbon
gpd gallons per day
mgd million of gallons per day
NSF NSF International (formerly National Sanitation Foundation)
O&M operating and maintenance
POE point-of-entry
POU point-of-use
RO reverse osmosis
SDWA Safe Drinking Water Act
SSCT Small System Compliance Technologies
SOC synthetic organic compounds
U.S. EPA U.S. Environmental Protection Agency
UV ultraviolet
VOCs volatile organic compounds
IV
-------
1. Introduction
The Safe Drinking Water Act (SDWA) Amendments of 1996, as well as a number of other
statutes and executive orders, require that the U.S. Environmental Protection Agency (EPA, or
the Agency) estimate regulatory compliance costs as part of its rulemaking process. EPA
estimates these costs at a system level for central treatment technologies and for point-of-use
(POU) and point-of-entry (POE) devices, which EPA has listed as Small System Compliance
Technologies (SSCTs) for several contaminants.1 Although POU and POE devices use the same
contaminant removal processes as central treatment technologies, these devices treat smaller
water volumes because they are installed at the user location and treat only the water that must
meet certain health standards. A POU device treats the water used primarily for drinking and
cooking and is often installed at a single water tap. A POE device is installed at the water entry
point for a home or nonresidential facility to treat all the water entering the facility for use.
EPA estimates costs at the system level by identifying cost components such as equipment and
labor requirements and applying unit costs based on vendor data and national labor rates. This
cost build up method can also be used by small drinking water systems to estimate site-specific
costs for a POU or POE compliance strategy. Consequently, EPA is making the cost estimating
tool it developed for national cost analysis available to small systems that want to use it to
develop site-specific estimates. In addition, EPA added features to enhance the tool's flexibility
for site-specific use.
1.1 Purpose
This report serves as documentation for EPA's cost tool that estimates system-level costs for a
POU or POE compliance strategy (the POU/POE model). The model was developed as a
companion tool for the Point-of-Use or Point-of-Entry Treatment Options for Small Drinking
Water Systems (U.S. EPA, 2006; hereafter called "EPA Guidance"). The cost estimates
generated by the POU/POE model include capital and operating and maintenance (O&M) costs
to implement and maintain an effective program that uses POU or POE devices to comply with
drinking water standards. EPA based the cost components in the POU/POE model on
information contained in the EPA Guidance. This guidance provides additional information
regarding regulatory requirements, technology options and considerations, and POU and POE
program elements and considerations. In addition, document appendices to the EPA Guidance
provide sample documents (e.g., model ordinance language and access agreements, sample
monitoring and maintenance logs, and sample public education flyers). Finally, it contains case
study summaries for several active compliance programs.
1.2 Document Organization
The documentation requirements for the POU/POE model are two-fold given its distribution as a
tool for system-level cost analysis. Chapter 2 provides a description of the model structure and
1 EPA defines small systems as those serving 10,000 or fewer people.
-------
contains documentation for the default assumptions and data in the model. These assumptions
and data provide examples of the types of values that EPA may use when it develops cost
estimates for policy analysis; they may not reflect conditions for a specific system. Chapter 3
contains a user guide that explains how to use the model to develop costs for a user-defined POU
or POE program. It is important to note that the POU/POE model does not replace the need for
systems to obtain professional engineering services noted in EPA Guidance. It is first and
foremost a tool for EPA to use in conjunction with other cost tools for its national cost analysis.
EPA has added the features described in Chapter 3 to allow stakeholders to easily replace default
inputs with site-specific information, thereby creating a tool for rapid system-level cost
modeling.
-------
2. Cost Model Description
The cost tool for POU and POE devices provides cost estimates for the elements of a POU/POE
compliance strategy as identified in EPA Guidance. The capital costs identified in the guidance
pertain to device purchase and installation, public education, and various engineering,
permitting, and legal activities, which are often referred to as indirect capital costs. The O&M
costs in the guidance document include device maintenance, water quality monitoring, residual
disposal, public education, and insurance coverage.
Exhibit 2.1 shows the costs included in the POU/POE model and provides general descriptions
for cost calculations. In most instances, the model cost categories correspond with the EPA
Guidance cost categories. Among the notable differences are that POU/POE model's capital
costs include initial water quality monitoring costs and the model's O&M costs do not include
line items for residual disposal or insurance coverage. Including an initial round of sampling in
the capital cost section allows for differences between sampling requirements during the first
year following installation and long-term monitoring requirements. Incremental insurance
coverage costs are not likely, especially if POU or POE devices are installed by vendor
representatives or licensed plumbers. Similarly, EPA assumes that most residual disposal will
involve very small increases in traditional household waste streams (e.g., replaced filters are
disposed in the household trash and sporadic flows for liquid wastes enter either a sanitary sewer
system or a septic system).
Exhibit 2-1. Cost Estimates Generated by the POU/POE Model
Item
Description
Capital Costs1
Treatment device
purchase
Treatment device
installation
Educational materials
Initial water quality
monitoring
Indirect capital costs
Cost to purchase treatment devices (cost per device x number of devices)
Labor costs for time to schedule and install treatment units (hours per device x number of
devices x wage rate)
Labor costs and material costs to prepare and distribute materials that explain the treatment
program to customers (hours or items per task x task frequency x wage rate or unit cost)
Labor, shipping, and analysis costs for the initial sample taken during the first year following
equipment installation (hours or costs per sample x number of samples x wage rate)
Costs for permitting, pilot study, engineering, legal, and contingency (either cost multipliers
or fixed dollar values)
O&M Costs2
Device maintenance
Educational materials
Annual water quality
monitoring
Labor time and material costs to replace treatment device parts such as filters (labor or
material costs per replacement x replacements per year)
Labor time and material costs to update and distribute materials that explain the treatment
program to customers (hours or items per task x annual task frequency)
Labor time, shipping, and analysis costs for the annual samples taken to monitor water
quality (hours or costs per sample x number of samples per year)
1 . Capital costs occur in the initial year, but do not occur again until complete replacement of the treatment units.
2. O&M costs occur annually and include costs for regular equipment replacement such as filter cartridges.
-------
These cost estimates are transparent because the model allows a user to see the calculations and
trace formula inputs back to their source data.2 Source data include user-provided values and
default values contained in the model. Input values identify the type of contaminant of concern
and the type of technology used for treatment as well as factors that affect the number and cost of
the items listed in Exhibit 2-1 such as the number of affected households.
The POU/POE model is a Microsoft Excel workbook that contains seven interactive worksheets,
which are described in the following sections.3 The names on the worksheet tabs are:
• Input
• User Input
• Capital Assumptions
• O&M Assumptions
• Control
• Price Tables
• Output.
2.1 Model Inputs - Input and User Inputs Worksheets
The POU/POE model operates in two modes. The default mode, which is primarily intended for
policy analysis, uses standard system sizes and default input assumptions to estimate costs. A
second mode allows a user to define a system by specifying input assumptions and unit costs.
Each mode has its own input worksheet. The Input worksheet controls model settings for the
default mode and the User Input worksheet controls model settings for the "user-defined' mode.
Mode selection occurs on the Input worksheet. This chapter primarily focuses on the data and
assumptions that apply in the default mode. These values can also be used in the user-defined
mode when site-specific information is missing. Chapter 3 provides supplemental information on
how to use the model and, in particular, how to enter user-defined values.
2.1.1 Target Contaminant and Treatment Technology
Regardless of operating mode, the initial input requirements are the same. The first input is the
target contaminant, which is selected from the drop-down list provided in the model. The second
input is a treatment technology. The list of applicable technologies depends on the contaminant
selected because treatment effectiveness varies by contaminant. Furthermore, the types of
devices with third-party certifications for meeting ANSI/NSF standards also vary by
contaminant. ANSI-accredited certification organizations include NSF International, the Water
Quality Association, Underwriters Laboratories, and CSA International. Relationships between
contaminants and approved technologies will evolve over time as new technologies are tested
Viewing formulas and using the Formula Auditing toolbar to help identify precedent and dependent cells requires a
worksheet be unprotected. To unprotect a worksheet, select Tools, Protection, and Unprotect sheet. If you are
prompted for a password, do not enter one; just leave the password space blank and click on OK.
3 Development versions of the model will contain additional documentation sheets and may contain Visual Basic
modules required by the model.
-------
and approved and as the list of regulated contaminants changes. Exhibit 2-2 provides a summary
of the contaminants and technologies in the current version of the POU/POE model. The model
also provides an option to add UV disinfection to any POU/POE device. The model does not
include pretreatment technologies that are not included in POU/POE devices. Such costs, if
applicable, would need to be accounted for separately.
Exhibit 2-2. Summary of Contaminants and Treatment Technologies in the POU/POE Model
Contaminant
Radium
Arsenic
VOCs
SOCs
Nitrate
Barium
Cadmium
Chromium
Copper
Fluoride
Lead
Mercury
Types of Devices Certified under ANSI/NSF Standards
POU Reverse Osmosis, POE Cation Exchange
POU Adsorptive Media NSF53, POU Reverse Osmosis
POE GAC1
POU GAC, POE GAC1
POU Reverse Osmosis
POU Reverse Osmosis, POE Cation Exchange
POU Reverse Osmosis
POU Reverse Osmosis
POU GAC, POU Reverse Osmosis
POU Reverse Osmosis
POU GAC, POU Reverse Osmosis
POU GAC
VOCs = volatile organic compounds
SOC = synthetic organic compounds
GAC = granular activated carbon
1 . There are no certified devices in this category so the price data indicate $0/unit. To prevent this price from
generating incorrect compliance cost estimates, POE GAC is not included in the treatment technology list on the
Control worksheet. Similarly, to avoid missing technology data errors, VOCs will not be explicitly included in
contaminant list until certified devices are available.
The default assumptions such as installation time and default prices such as the price to purchase
a treatment device correspond primarily to the treatment technologies. Nevertheless, some
default assumptions or prices such as lab analysis costs are contaminant specific.
2.1.2 Service Population
The model estimates costs on a per-household basis. For example, the cost estimate for a POU
compliance strategy for 50 households will include equipment purchase and installation costs for
all 50 households, as well as annual maintenance and monitoring costs for 50 households.
In the Standard System operating mode, the model derives service population estimates based on
the design flow information provided on the Input worksheet. The model uses statistical
relationships between the population and flow values that EPA estimated from survey data
collected through the Community Water System Survey. These relationships vary by source
-------
water.4 Consequently, the Input worksheet contains a drop-down list with water source options
(i.e., ground water or surface water). Chapter 3 describes two options for user-defined service
population data in the User-Defined System mode.
2.2 Capital Assumptions Worksheet
The Capital Assumptions worksheet contains default values for the nonprice parameters or
assumptions required to estimate capital costs; all price information is on a separate worksheet
addressed in a later section. The default assumptions can vary by treatment technology, so in
most cases there is a column of values for each treatment technology in the POU/POE model.
Exhibit 2-3 provides a list of the capital assumptions in the POU/POE model. Chapter 3
describes how a user can override default assumptions such as these to include site-specific
information.
Exhibit 2-3 also illustrates default values for two of the POU treatment technology options. For
example, the model calculates equipment installation costs for the POU/POE treatment device
selected and for UV if that option is selected. One assumption needed to estimate installation
costs is the amount of time required to install the devices. The default values in Exhibit 2-3
indicate that the model includes a 1-hour installation time for the POU/POE device and a 30-
minute period for administrative time and scheduling time (i.e, contacting the household to
schedule installation) if the treatment technology is POU Adsorptive Media NSF53, which is
approved for arsenic removal to meet the arsenic standard. If, however, the treatment technology
is the rental option for POU reverse osmosis (RO), then there are zero hours for scheduling and
installation because EPA assumed that these costs are included in a monthly or annual service
contract. Consequently, they enter the cost analysis as O&M costs rather than capital costs.
The technology selection on the Input worksheet (or the User Input worksheet when operating
the model in User-Defined System mode) dictates which column of default values appear in the
Output worksheet and, therefore, generate capital costs. The following sections provide
background information for the input parameters and the default assumptions in the POU/POE
model. Note that the Capital Assumptions worksheet contains only quantity assumptions (e.g.,
number of hours or number of flyers). The prices assigned to various activities are on the Price
Tables worksheet, which is addressed in a later section.
4 For ground water systems, the relationship is population =[ (design_flow*1000) / 0.54992]A(1 / 0.95538), where
design_flow is measured in million gallons per day (mgd). For surface water systems, the relationship is
population=[(design_flow* 1000) / 0.59028]A(1 / 0.94573). In both instances, the population estimate is then
converted to an estimate of households: household connections = population / hh_size, where hh_size is a parameter
representing average household size. Systems that also serve nonresidential customers may need to incorporate costs
for these customers one of two ways. The first method involves converting these customers to residential equivalents
to use in the POU/POE model (e.g., if the POU device purchase and installation costs for each nonresidential
customer are approximately three times higher than a residential customer, then convert each nonresidential
customer to three household equivalents for modeling purposes). The second method is to estimate nonresidential
costs separately without the model and aggregate residential and nonresidential costs.
-------
Exhibit 2-3. Summary of Design Assumptions for Capital Costs with Example Default Values for
Two POU Technologies
Parameter Units POU RO Rental
Equipment Installation
POU/POE Installation hrs/household 0.00
Scheduling Time hrs/household 0.00
UV Installation (when UV is hrs/household 0.00
excluded)
Public Education Program
Technical Staff Time:
Develop materials hrs 10.0
Nitrate health effects hrs 5.0
Meetings hrs 2.0
Post-meeting hrs 2.0
Clerical Staff Time:
Develop materials hrs 6.0
Nitrate health effects hrs 5.0
Meetings hrs 2.0
Post-meeting hrs 2.0
Education Materials:
Meeting flyers flyers 10
Meeting ads ads 1
Nitrate awareness flyers flyers 10
Meeting handouts pages/household 3
Billing mailers pages/household 2
Monitoring
Sampling time hrs/sample 0.252
Sampling scheduling time hrs/sample O.OO2
Analysis samples/household 1
Frequency % households 100.0%
sampled
POU Adsorptive
Media NSF53
1.001
0.50
0.00
10.0
5.0
2.0
2.0
6.0
5.0
2.0
2.0
10
1
10
3
2
0.252
O.OO2
1
100.0%
1 . Peer review comments indicated that default installation time assumptions should be 1 hour for most POU devices and 2
hours for POU RO and POE devices.
2. Peer reviewers recommended that the initial monitoring sample be taken during the
incremental sampling time and avoid the need to schedule a sampling visit. The mode
installation and sampling to occur independently.
installation visit, which would minimize the
retains the option, however, for
2.2.1 Equipment Installation
Installation of the POU and POE devices will be the responsibility of the local water system or
utility. The utility can, however, hire a licensed plumber or representative of the product
manufacturer to install the devices. Based on the variety of plumbing issues encountered among
older housing units in a rural community, NSF (2005) recommends using an experienced
plumber to perform the installations. Units requiring power connections (e.g., aeration towers)
may also require a professional electrician assist with the installation.
-------
Equipment installation costs POU/POE model include labor costs for installation as well as the
costs to purchase the devices. The model contains a default estimate of two hours per household
to install most POU and POE devices. A variety of factors such as existing plumbing conditions
and travel distance will affect installation times across sites. This estimate represents an average
time across sites, taking into account differences in installation conditions and location. Section
2.5 describes the wage rate used to value installation costs as well as the purchase prices for the
POU or POE devices.
The estimate is consistent with case study data. In the Grimes, CA arsenic demonstration
program (NSF, 2005), POU adsorptive filter installation times ranged from 15 minutes to 3 hours
depending on the accessibility of piping and the need for additional lines (e.g., to provide treated
water to ice-makers). The mean installation time was one hour, but total plumber billing records
indicated twice as much time spent on all installation-related activities. Study authors attributed
the additional cost to time spent obtaining special plumbing fittings and return visits to homes
when residents missed their appointments. Based on peer review suggestions, EPA selected a
default average installation time value of one hour for less complex POU devices and two hours
for more complex POU devices and POE devices. The model includes a contingency factor to
account for unanticipated costs such as the need for special plumbing fittings.
The POU/POE model reflects the assumption that when UV devices are installed, they are
supplemental treatment equipment, requiring incremental installation time. The default
assumption for the incremental amount of on-site time to install a UV device is one hour per
household for a POU device and two hours for a POE device. These installation times are only
applicable when UV treatment is selected. If UV treatment is not selected, then the installation
time will be zero on the Output worksheet despite any nonzero value in the assumptions
worksheet.
Installation costs also include administrative time for utility staff for planning and to contact
homeowners to schedule an installation appointment. The default assumption is an average of 30
minutes (0.5 hours) per household to contact homeowners. Scheduling effort is likely to vary
across customers, with some being relatively easy to schedule while others require multiple calls
to identify and contact the correct homeowners or to handle situations such as homeowner
reluctance to participate or language barriers (U.S. EPA, 2006).
2.2.2 Public Education Program
EPA Guidance (2006) recommends that systems "plan on investing resources in a public
education program to obtain and maintain customer participation and long-term customer
satisfaction" with the POU or POE program. The two main program elements recommended in
U.S. EPA (2006) are:
• conduct one or more public meetings with all customers prior to installing any POU or POE
devices to educate customers regarding the regulatory compliance requirements and the role
of the POU/POE devices
• provide information updates in billing mailers and on information flyers posted in public
locations such as a post office, a public library, or a website.
In addition, the EPA Guidance recommends having someone available to answer questions either
by phone or through web-based communication.
-------
The POU/POE model generates capital (or start-up) cost estimates for the two main educational
program elements. These costs include:
• time for the system's technical and clerical staff to prepare information for at least one
public meeting
• time for the system's technical and clerical staff to attend at least one meeting
• time for the system's technical and clerical staff to prepare an additional billing mailer
following the meeting
• materials for information sheets provided to each household at the meeting
• materials for meeting announcements such as newspaper advertisements, posted fliers, and
billing mailers.
EPA Guidance recommends that system staff be prepared to cover the following topics during
the public meetings:
• inform customers of the current situation (e.g., the contaminant of concern, the applicable
standard, current exposure levels, and potential health impacts)
• describe compliance options and the system's decision process (e.g., centralized treatment,
source water blending, source water changes, and POU or POE alternatives)
• explain what POU/POE devices are and what they will do to improve water quality (e.g.,
the treatment devices will be installed in homes, but owned and maintained by the system,
which will require access for installation, monitoring, and maintenance)
• address ownership or replacement issues regarding any existing POU/POE devices that
homeowners previously installed in their homes
• explain the purpose and scope of a pilot study if one is planned or needed.
The default assumption in the POU/POE model is a total of 16 hours (10 hours for technical staff
and 6 hours for administrative or clerical staff) to develop handout materials for a public meeting
and advertise the meeting. This assumption reflects a scenario in which system staff:
• develop and insert a 1-page meeting announcement in a regularly scheduled billing one
month prior to the meeting and a subsequent 1-page billing insert with program updates
• post meeting announcement flyers in prominent locations
• place an advertisement in the local newspaper and through public service announcements
• develop a 3-page information handout with information on POU or POE devices and
general information on installation, maintenance, and monitoring of the devices for meeting
attendees.
Associated material costs in the POU/POE model include copying costs for the billing insert (1
page per household) and the meeting handout (3 pages per household), printing costs for the
posted flyers (10 full-color copies), and a public meeting announcement in the local newspaper.
The model does not include material costs for public service announcements on the local radio
and television stations, which are free of charge. It also does not include rental fees for a meeting
site because EPA assumes that the meeting can be held in a public building.
-------
The POU/POE model includes 4 hours (2 for technical staff and 2 hours for clerical staff) for
public meeting attendance and another 4 hours (2 for technical staff and 2 hours for clerical staff)
for post-meeting activities such as developing an additional 1-page billing flyer announcing
POU/POE program updates.
EPA Guidance indicates that additional public education efforts may be needed when the
regulated contaminant is nitrate because of the potential for nitrate to cause serious illness among
infants (e.g., blue baby syndrome). The POU/POE model includes additional labor time and
material costs to develop and communicate information on health impacts of acute exposure to
elevated nitrate levels. The default assumptions in the POU/POE model include 10 hours (5 for
technical staff and 5 for clerical staff) to prepare additional materials regarding nitrate health
impacts for a public meeting and to develop and post 10 additional health information flyers. To
assist systems with nitrate compliance needs, EPA Guidance provides a sample public education
flyer with relevant health information and safety instructions for homeowners.
2.2.3 Initial Water Quality Monitoring
Systems that implement a POU or POE compliance strategy will need to monitor the quality of
water produced by the treatment devices to demonstrate compliance with the drinking water
standard for the contaminant of concern. EPA Guidance notes that the system will need to work
with the appropriate regulatory agency to establish an approved compliance-monitoring
schedule. The resulting monitoring schedule may have sampling rates in initial year that differ
from sampling rates in subsequent years. Consequently, the POU/POE model includes cost
calculations for initial year monitoring as well as annual monitoring.
EPA Guidance provides an example of a monitoring schedule in which samples are taken from
every unit during the first year to confirm that the units are working properly, and then
monitoring frequency declines to one-third of units each subsequent year. EPA designed the
POU/POE model to accommodate this kind of variation in monitoring frequency over time.
Furthermore, the default assumptions in the model reflect this example schedule.
Initial water quality monitoring costs can include four types of costs: the time required to obtain
samples from each location, time required to contact households to schedule sampling, shipping
costs to send the samples to a certified laboratory, and fees paid to the laboratory for analysis and
results. The POU/POE model generates estimates for each of these costs.
First, the model calculates the total number of initial sampling events. Two assumptions in the
model affect the total number of sampling events in the first year - the share of households
monitored and the number of samples per household. The default values are based on the
scenario in the EPA Guidance: sampling occurs at 100% of the households receiving a POU or
POE device, and there is one sampling event at each household. These frequency assumptions
combine with the household estimate on the Output worksheet to establish the total number of
sampling events.5
5 The results on the Output worksheet include an adjustment to address the overlap that occurs when capital costs for
initial year monitoring are added to annual monitoring costs. The number of initial sampling events used in the
capital cost estimate is the total initial year events minus the number of annual sampling events to prevent double
counting that portion of the initial year sampling events.
10
-------
System labor costs depend on the total number of sampling events and the average amount of
time per event required to schedule and take samples. The default values reflect an assumption
recommended by peer reviewers to have initial sampling occur during the installation visit.
Therefore, the clerical labor to schedule each sampling event is zero and the technical labor to
conduct the sampling is 15 minutes (0.25 hours). The scheduling labor assumption is consistent
with EPA Guidance recommendations that systems attempt to coordinate monitoring activities
with other site visits to minimize the burden on households as well as system resources. The
Grimes case study cost analysis (NSF, 2005) used an estimate of 15 minutes per sampling event,
which is consistent with the default assumption. The labor rates that the POU/POE model uses to
value labor costs are addressed in the Price Tables section.
Shipping and analysis costs depend on the total number of samples and unit costs for shipping or
analysis. Analysis costs include analytical fees for the contaminant of concern and total coliform
as a pathogen indicator. Section 2.5 (Price Tables) addresses the default unit costs in the
POU/POE model.
2.2.4 Indirect Assumptions
EPA Guidance identifies four types of indirect capital costs to include in a system-level cost
estimate:
• permitting costs in cases where an operating permit or compliance plan review is required
by local and/or State agencies
• pilot testing costs in cases where a system tests one or more devices evaluate their
effectiveness under local conditions6
• legal costs, which may include developing an ordinance that defines the water system's and
homeowners' responsibilities for POU or POE devices, providing advice on options for
households that prefer not to participate, and developing agreements that will grant system
personnel or contractors legal access to all compliance devices for maintenance and
monitoring
• engineering costs to evaluate the system's compliance options (e.g., centralized treatment,
new water source, and POU program) and select the appropriate POU or POE treatment
technology.
The POU/POE model contains these costs and a fifth one - contingency - in the indirect portion
of capital costs. Contingency cost accounts for unforeseen costs in installing POU or POE
devices such as additional installation costs due to extra carpentry work or unusual plumbing
conditions. Exhibit 2-3 shows that the default values for indirect costs are multipliers or
percentages of installed equipment costs. For example, for every $1,000 spent to purchase and
install POU or POE devices, the model adds $30 (3% x $1,000) for permit costs.
6 U.S. EPA (2006) notes that "if the system uses a POE device, some form of field testing is required under 40 CFR
Section 141.100. If POU or POE devices are used under a variance or exemption, 40 CFR Section 142.62(h) also
requires field testing."
11
-------
Exhibit 2-4. Indirect Cost Multipliers
Parameter
Permitting
Pilot Testing
Legal
Engineering
Contingency
Units
multiplier - percent of installed equipment cost
multiplier - percent of installed equipment cost
multiplier - percent of installed equipment cost
multiplier - percent of installed equipment cost
multiplier - percent of installed equipment cost
Default Assumption
3.0%
3.0%
3.0%
15.0%
10.0%
Three of these estimates are based on information in Guide for Implementing Phase I Water
Treatment Cost Upgrade (EPA, 1998): 3% for permitting, 3% for legal fees, and 15% for
engineering. The engineering multiplier is also consistent with the results of the Grimes case
study (NSF, 2005), which had engineering costs equal to 15% of installed costs ($41 for
engineering on installed costs of $275). EPA assumed that the default multiplier for pilot testing
was similar to permitting and legal fees because the devices being installed are NSF certified for
drinking water treatment. Finally, the contingency cost multiplier adds 10% to the installed
equipment cost estimate to account for unknown factors that increase costs. These might include
additional time and materials to handle unusual plumbing conditions. For example, in the Grimes
case study, some homes had water shut-off valves located underground in alleyways instead of
inside the home.
2.2.5 Other Capital Assumptions
The capital assumption worksheet has four additional assumptions used in the capital cost
analysis. The first one is the average number of people per household, which the model uses to
convert population estimates to household connection estimates. The default assumption is 2.6,
which is the average household size reported for the most recent national census (U.S. Bureau of
the Census, 2001).
The second and third assumptions are a discount rate and a discount period. The model uses
these assumptions to calculate annualized capital costs, which affects total annual costs
(annualized capital costs plus annual O&M costs). The default discount rate value is 7%, which
EPA often uses for policy analysis. The default annualization period is 10 years, which is based
on case studies in EPA Guidance and vendor information. This period reflects the expected time
the devices will remain installed and operating given regular maintenance (e.g., filter
replacement).
The fourth assumption is an average total daily per capita water consumption estimate of 100
gallons. This value is used only to calculate a cost per thousand gallons ($/kgal) estimate, which
can be compared with $/kgal costs for centralized treatment options in a cost-effectiveness
analysis. The estimate uses total consumption per person per day instead of the actual POU or
POE treated volume because the cost-effectiveness comparison is between alternatives that
achieve the same health risk reduction benefits. Because the benefits of the POU or POE strategy
are the same as the benefits for central treatment, the cost-effectiveness measure must use
comparable treated water quantities to generate comparable $/kgal values.
12
-------
2.3 O&M Assumptions Worksheet
This worksheet contains several matrices of default parameters for O&M costs. There is one
large matrix with default values for parameters that are common to all technologies and several
smaller matrices with technology-specific assumptions for equipment replacement. The default
values for equipment replacement are based on vendor recommendations. Chapter 3 describes
how a user can override these default values to include site-specific information.
Exhibit 2-5 provides a summary of the types of assumptions that affect system O&M costs for
labor and materials, other than replacement parts. These include labor costs to maintain the POU
or POE equipment, prepare and distribute educational program updates, and monitor water
quality.
Exhibit 2-5. Summary of Design Assumptions for O&M Costs with
Example Default Values for Two POU Technologies
Parameter Units POU RO Rental
Equipment Maintenance
POU/POE maintenance hrs/visit 0.00
POU/POE replacement frequency visits/household/yr 0.00
UV maintenance (when UV is excluded) hrs/visit 0.00
UV maintenance frequency visits/household/yr 0.00
Scheduling time hrs/visit 0.00
Education Program
Information updates- technical labor hrs 12.0
Nitrate information updates- technical labor hrs 20.0
Information updates- clerical labor hrs 12.0
Nitrate information updates- clerical labor hrs 20.0
Nitrate awareness flyers flyers 10
Nitrate billing mailers pages/household 3
POU Adsorptive
Media NSF53
0.50
1.00
0.00
0.00
0.50
12.0
20.0
12.0
20.0
10
3
Monitoring
Sampling time hrs/sample 0.251 0.251
Sampling scheduling time hrs/sample O.OO1 O.OO1
Analysis sample/household 1 1
Frequency % households/yr 33.3% 33.3%
1 . Peer reviewers recommended that the monitoring samples be taken during maintenance visits, which would minimize the
incremental sampling time and avoid the need to schedule a sampling visit. The model retains the option, however, for
maintenance and sampling to occur independently.
The exhibit also provides examples of the default values for two of the POU treatment
technology options. For example, the default values POU for the Adsorptive Media NSF53
technology option indicate that the model includes a 30-minute site visit per maintenance trip for
the POU/POE device and a 30-minute period to contact the household to schedule maintenance.
The assumptions for the POU RO Rental treatment show zero hours for scheduling and
maintenance because these costs are included in a single cost for annual service. This rental cost
enters the O&M cost estimate through an equipment price, which is described below.
13
-------
2.3.1 Equipment Replacement and Maintenance
Because a treatment technology can have multiple components that require replacement, the
O&M assumptions worksheet contains several treatment-specific matrices with replacement
assumptions. Exhibit 2-6 shows an example matrix for the POU RO treatment technology
(purchase option instead of rental option). It shows that there are four items replaced at different
frequencies based on vendor recommendations. The RO membrane is replaced once every three
years based on average replacement schedules across vendors, and the other filter cartridges are
changed once per year.
The cost model reflects the assumption that a cost-effective maintenance schedule will overlap
equipment replacement trips whenever possible. Therefore, membranes will be replaced during a
maintenance visit to change the filter cartridges.
Exhibit 2-6. Example of Equipment Replacement for a Purchased RO Device
Equipment Replacement Frequency Units
Sediment Pre-Filter 1 units/year
Pre-GAC Filter Cartridge 1 units/year
Post-GAC Filter Cartridge 1 units/year
RO Membrane 0.33 units/year
The default labor hours include 30 minutes of technical labor time per visit for equipment
replacement and 30 minutes of clerical labor time for scheduling appointments and tracking the
replacement schedule. The Grimes case study cost analysis (NSF, 2005) used an estimate of 15
minutes per filter change out, which suggests EPA's default assumptions (totaling one hour per
visit) are potentially conservative (i.e., tending to overstate costs).
2.3.2 Education Updates
EPA Guidance recommends providing ongoing communication with the community to maintain
a successful POU or POE program and to educate new customers. The POU/POE model
contains costs for annual outreach activities and, if applicable, nitrate awareness. The general
outreach activities are likely to involve responding to phone or email inquiries regarding the
POU/POE program and devices and mailing an annual update on the program. EPA's default
estimate is that these activities require an average of two hours per month or 24 hours per year.
This incremental time requirement is minimal because EPA assumes than most consumer
questions will arise during scheduling calls and maintenance visits. EPA evenly split these hours
between technical and clerical staff for the default cost assumptions. In addition, the POU/POE
model has a default copying cost for sending each household up to 3 pages of POU/POE
program outreach material as a billing inserts each year.
When the contaminant of concern is nitrate, The POU/POE model includes additional labor time
and material costs to update and communicate information on health impacts of acute exposure
to elevated nitrate levels. The default assumptions in the POU/POE model include an additional
week or 40 hours of labor time throughout the year to contact health care professionals and
families with infants to remind them of the risks among infants of acute exposure to nitrate. EPA
assumed these hours were evenly split between technical and clerical staff (20 hours each) for
14
-------
costing purposes. The default costs include updating the 10 nitrate flyers posted in public areas
each year and sending up to 3 additional pages of nitrate information as billing mailers.
2.3.3 Monitoring
As noted in the initial monitoring section, the POU/POE model incorporates annual monitoring
costs, and the default assumptions in the model reflect the example monitoring scenario in EPA
Guidance. A system's approved monitoring plan may differ from this example, and the model
can accommodate such differences (see Chapter 3). In the example scenario, all units are
monitored during the first year of operation, and one-third of a system's connections are sampled
in each subsequent year. Aside from the change in the number of households monitored per
year, the default assumptions are identical to the initial monitoring assumptions.
The exception to this default frequency assumption is nitrate, which has annual monitoring for
all devices. This assumption reflects the more frequent monitoring requirement for nitrate
compared to other inorganic contaminants (40 CFR Section 141.23). Because of the health risks
associated with acute nitrate exposure, more frequent sampling is needed to ensure adequate
public health protection.
2.4 Control Worksheet
As the name suggests, the Control worksheet controls the operation of the POU/POE model. It
identifies which input model, contaminant, and treatment technology control the assumptions on
the Output worksheet. It contains the list of the contaminants that appear in the drop-down lists
on either of the input worksheets. It also contains the contaminant-specific treatment
technologies. Finally, it contains parameters that pass the correct input information to the Output
worksheet. The technology options include placeholders for technologies for which devices are
not yet ANSI/NSF certified. These placeholders were included during the development phase so
they will be easier to integrate should certified devices become available.
2.5 Price Tables Worksheet
The Price Tables worksheet contains the default price data used in the POU/POE model. These
data include prices for the POU or POE devices, replacement parts, labor rates, and prices for
sample analysis and shipping. All prices have been escalated to current dollars using applicable
price indices. Adjustments to labor rates utilize the Employment Cost Index (BLS, 2006b), and
all other prices have been adjusted using the Consumer Price Index (BLS, 2006a). Chapter 3
describes how a user can override default prices to include site specific information.
The products listed in the Price Tables worksheet are randomly selected from among those that
meet ANSI/NSF certification requirements. Information in this worksheet does not constitute an
endorsement by EPA for these products. The average prices and price ranges across comparable
products in the model are comparable to those reported in Raucher et al. (2004), which is a study
of POU treatment costs sponsored by the American Water Works Association Research
Foundation (AWWARF).
2.5.1 POU and POE Device Prices
The price data for POU and POE devices are grouped by treatment technology. For each
technology, there are two price tables. The first table contains the unit prices for complete
15
-------
treatment systems that are used in the capital cost estimates. Each table is a schedule of unit
prices for different purchase quantities. The unit price declines when vendors provide quantity
discounts. These discounts can vary across vendors. Exhibit 2-7 shows an example of the price
table for the POU Adsorptive Media NSF53 equipment. The first two columns show the range of
households to which the unit prices apply and the third column shows the unit cost. For example,
a system installing units in 20 households would pay $263 per unit, while a larger system with
120 households would pay $210 per unit.
Exhibit 2-7. Example of Unit Prices for
POU Adsorptive Media NSF53 Equipment
Range for Number
Households
1 to 4
5 to 15
16 to 24
25 to 50
51 to 500
501 to 1,000
Avg. Unit Cost
525
289
263
236
210
170
EPA Guidance notes that the POU and POE devices used as part of a compliance strategy must
be ANSI/NSF certified. Therefore, the prices on the Price Table worksheet are for POU and POE
devices that have been certified under one of the following ANSI/NSF standards.
• ANSI/NSF Standard 53 - Health Effects: This standard addresses POU and POE devices
designed to reduce specific contaminants such as lead, VOCs, and arsenic. For example, the
unit prices in Exhibit 2-7 are for an adsorptive media device that is certified for arsenic
removal.
• ANSI/NSF Standard 44 - Residential Cation Exchange Water Softeners: This standard
pertains to residential cation exchange water softeners that reduce hardness, and it also
verifies a system's ability to remove radium and barium.
• ANSI/NSF Standard 58 - Reverse Osmosis Drinking Water Treatment Systems: The
contaminant removal capabilities for POU RO systems are certified under this standard.
• ANSI/NSF Standard 55 -Ultraviolet Microbiological Drinking Water Treatment
Systems: This standard applies to POU and POE ultraviolet systems that are designed to
provide safe levels of microbial disinfection in drinking water.
The Price Tables worksheet will contain average unit costs for POU and POE devices based on
vendor data documented in the worksheet.7 The mention of specific products or prices does not
constitute an endorsement by EPA.
The format for the vendor section is the same database format used in the cost models for central treatment
although some fields are not used. This format allows equipment price data to be adjusted for installation costs,
transportation costs, and inflation factors. The POU unit prices are delivered prices so the transportation cost
multipliers used for central treatment components are not applicable. Because installation costs are modeled
explicitly in the POU/POE model, the installation cost columns in the database are not used. Nevertheless, the
16
-------
The second price table for each technology comprises of replacement parts for the POE/POU
units. To simplify the cost analysis, EPA did not develop price schedules with volume discount
information for replacement parts. Therefore, O&M costs may be overstated. The volume
discounts for replacement parts, however, are not likely to have the same impact as volume
discounts for initial device purchases. First, replacement parts are less expensive and, depending
on replacement frequency, are likely to be purchased in smaller quantities with purchases spread
over time to minimize inventory costs.
2.5.2 Labor Costs
Costs for labor hours attributed to system technical and clerical staff are based on a wage rate
study prepared for EPA (SAIC, 2003). The loaded hourly rates (wages plus benefits) for
technical staff ($26.06) and clerical staff ($18.54) are based on national wage data for treatment
operator and secretarial occupations, respectively, in the water supply industry. The default
assumptions in the POU/POE model assign the operator wage rate to maintenance and sampling
activities, and public education program activities. Cost estimates for the clerical hours depend
on the clerical hourly rate.
Labor costs for device installation in the POU/POE model use an average loaded wage rate of
$33.04, which is based on national average wages for plumbers ($21.40) and electricians
($21.73) (BLS, 2006c) and a benefits multiplier of 1.48 (BLS, 2006b).
2.5.3 Monitoring Costs
Costs for laboratory analysis are based on average fees across a sample of laboratories. Exhibit
2-8 provides a sample of fees in the POU/POE model. The Price Tables worksheet documents
data sources.
Exhibit 2-8. Average Cost Data for Laboratory Analyses
Contaminant
Arsenic
Copper
Nitrate
Cost of Analysis
per Sample
$25.75
$19.50
$24.25
Contaminant
SOCs
Radium (226 and 228)
Total Coliform
Cost of Analysis
per Sample
$183.33
$257.50
$16.50
Applying these analysis unit costs to every sample may overstate monitoring costs for one of two
reasons. First, laboratories may offer quantity discounts for shipments with multiple samples.
Second, systems may be able to submit composite samples across multiple devices rather than a
single sample for each device. The latter strategy helped reduce analysis costs in the Grimes case
study (NSF, 2005).
database format for POU/POE prices must be consistent with the database format used in other models. Vendor
names have been removed from the version of the model distributed to the public to avoid the appearance of product
endorsement.
17
-------
Shipping costs for samples are based on U.S. Postal Service rates for 2-day delivery. Exhibit 2.9
shows shipping costs for packaging containing selected ranges of samples. The average cost per
sample declines as the number of samples increases.
Exhibit 2-9. Costs of Shipping Samples Using U.S. Postal Service
Number of Samples
15
16 to 50
51 to 100
101 to 200
201 to 500
Radium 1
$11.00
$24.00
$48.00
$96.00
$250.00
Nitrate, Arsenic, and
other Inorganics 2
$9.00
$11.00
$21.00
$42.00
$100.00
SOCs3
$9.00
$11.00
$21.00
$42.00
$100.00
1 . Assumes sample is taken in a one-liter bottle and shipped on ice.
2. Assumes sample is taken in 250 ml sample bottle and shipped on ice.
3. Assumes sample is taken in two 40-mL vials for SOCs. All samples are shipped on ice.
2.5.4 Other Annual Costs
There are several material costs associated with the educational program. These include $0.08
per page to cover cost of copying the billing inserts and meeting handouts, $2.00 per flyer for 10
posted flyers, and $40 for an announcement in the local newspaper. The model does not include
material costs for public service announcements on the local radio and television stations, which
are free of charge. It also does not include rental fees for a meeting site because EPA assumes
that the meeting can be held in a public building.
EPA assumed that disposal costs for POU and POE device residuals are negligible and,
therefore, did not include them in the POU/POE model. For example, the media filters used in
the Grimes case study were determined to be safe for disposal in the household trash by the
California Department of Health, which tested the arsenic concentrations in the filters (NSF,
2005). The EPA Guidance provides information on disposal issues and considerations.
As noted above, the POU/POE model does not include a line item for annual incremental
insurance to cover any increased liability associated with entering private residences. The capital
cost estimate, however, includes installation by a plumber, which is intended to cover the cost for
a licensed, bonded, and insured plumber. This assumption captures costs associated with the
most likely source of risk to private residences - a poor installation.
2.6 Output Worksheet
The Output worksheet shows all of the cost components and the cost build-up. A partial example
in Exhibit 2-10 provides shows the build-up for equipment costs. The top section is a summary
of the contaminant, treatment, and system size information provided on the Input worksheet (or
the User Input worksheet in the User-Defined System mode).
The next section on the Output worksheet shows the cost build-up for direct capital costs by
component. The reporting format shown in Exhibit 2-10 illustrates the cost build-up for capital
18
-------
costs. The cost build-up details include quantities (e.g., number of devices or hours per
household), frequencies (usually the number of households affected), and unit costs (e.g., cost
per device or per staff hour). Output includes total cost per component and, where applicable, a
useful life estimate. This estimate generally reflects the expected life of the installed device. It is
used in the POU/POE model to convert capital costs to annual costs, which can be added to
O&M costs to obtain an annual cost estimate and an average unit cost ($/kgal) designed for
comparison with centralized treatment costs.
Exhibit 2-10. Example of Cost Details Provided on the Output worksheet
System Description
Households Served
Contaminant
Treatment Technology
Treatment Location
UV Treatment
Capital Costs
WBS # Item
25
Arsenic
POU Reverse Osmosis
POU
no
Quantity
Design
Frequency
Unit Cost Total Cost
23.1.
23.1.
23.1.
23.1.
23.1.
23.2.
23.2.
23.2.
23.2.
,1
2
3
4
,5
111
.1.
1
1.
1.
HI
.1
.2
3
4
W/'
POU/POE Unit Purchase
POU/POE Installation
Scheduling Time
UV Purchase
UV Installation
Develop materials
Nitrate health effects
Meetings
Post-meeting
V/1"''1' '• ' -'''''''"l'"w//$////r/
1
2.00
0.50
0
0.00
10.00
0.00
2.00
2.00
'////////////////////////////////////
unit/household
hours/household
hours/household
unit/household
hours/household
hours
hours
hours
hours
'//////////////////////////////
25
25
25
0
0
'////////////////////////
households
households
households
households
households
'//////////////////////
$
$
$
$
$
'////////////////////////
561
33.12
17.89
25.07
25.07
25.07
25.07
'//////////////////
$ 14,023
$ 1 ,656
$ 224
$
$
^i™lfl
"$" " 251
$
$ 50
$ 50
'//// ////// i/ u'li ^y ''"•''•
17.89 $
17.89 $
17.89 $
23.2.2.1 Develop materials 6.00 hours $ 17.89 $ 107
23.2.2.2 Nitrate health effects 0.00 hours $
23.2.2.3 Meetings 2.00 hours $
23.2.2.4 Post-meeting 2.00 hours
23.2.3.1 Meeting flyers 10 flyers
23.2.3.2 Meeting ads 1 ads $
23.2.3.3 Nitrate awareness flyers 0 flyers $
23.2.3.4 Meeting handouts 3 pages/household 25 households $
23.2.3.5 Billing mailers 2 pages/household 25 households $
2.12 $
42.46 $
2.12 $
0.08 $
0.08 $
36
36
II!
21
42
6
4
2.7 References
Bureau of Labor Statistics. 2006a. Consumer Price Index. Available at
ftp://ftp.bls.gov/pub/special.requests/cpi/cpiai.txt.
BLS. 2006b. Employer Costs for Employee Compensation Historical Listing, 2004-2005.
Available at ftp://ftp.bls.gov/pub/special.requests/ocwc/ect/ececqrtn.pdf.
19
-------
BLS. 2006c. November 2004 National Occupational Employment and Wages Estimate.
Available at http://www.bls.gov/oes/current/oes_nat.htm.
NSF International (NSF). 2005. Feasibility of Economically Sustainable Point-of-Use/Point-of-
Entry Decentralized Public Water System. Ann Arbor, MI: NSF International.
Raucher, B., J. Cotruvo, R. Narashimhan, R. Regunathan, and J. Drago. 2004. Conventional and
Unconventional Approaches to Water Service Provision. Report 91031 prepared for AwwaRF
Project 2761. Denver, CO: American Water Works Association Research Foundation.
Science Applications International Corporation (SAIC). 2003. Labor Costs for National Drinking
Water Rules. Report prepared for U.S. EPA, Office of Ground Water and Drinking Water.
Reston, VA: SAIC.
U.S. Bureau of the Census. 2001. Census 2000 Demographic Profile: Summary File 1. Available
online at http://censtats.census.gov/data/US/01000.pdf.
U.S. Environmental Protection Agency (EPA). 2006. Point-of-Use or Point-of-Entry Treatment
Options for Small Drinking Water Systems. EPA 815-R-06-01 O.Washington, D.C.: U.S. EPA
Office of Water.
U.S. EPA. 1998. Guide for Implementing Phase I Water Treatment Cost Upgrades. Washington,
D.C.:U.S. EPA.
U.S. Postal Service. 2006. Notice 123 Ratefold Effective January 8, 2006. Available at
http://pe.usps.gov/cpim/ftp/notices/Notl23/Notl23.pdf.
20
-------
3. Cost Model Operation
This chapter contains a user guide that explains how to use the POU/POE model to develop costs
for a user-defined POU or POE program. As stated in Chapters 1 and 2, the POU/POE model is a
Microsoft Excel workbook that combines user-defined inputs with pre-programmed calculations
to estimate total costs for a POU or POE strategy. These costs include upfront capital costs (to
purchase and install POU or POE devices, develop and implement an educational program, and
conduct initial water quality sampling to test the operating effectiveness of the devices) as well
as O&M costs (for equipment replacement, customer education, and water quality monitoring).
EPA designed the cost model to be easy to operate while also giving the user enough flexibility
to alter key inputs such as unit prices and assumptions.
3.1 Getting Started
To use the POU/POE model as a stand-alone tool, copy the file called POU_POE model.xls to
any location on your computer's hard drive. You will need to have Microsoft Excel (Office 2000
or a later version) to open and operate the model. You can open the file by double clicking on the
filename in a file browser, or open Excel first and browse for the file. If prompted, select "Enable
Macros" to ensure the design macros used to run the model are functional.8
The file opens to show a Cover worksheet, which includes disclaimer information. Clicking on
the Start link at the bottom of this worksheet will take you to the Input worksheet. The first step
is to select the operation mode, shown in Exhibit 3-1. Click on the gold cell next to Step 1, and
then click on the small down arrow at the lower right corner of the cell, to see a drop-down list
with two options: Standard System or User-Defined System. Choose the type of system you want
to model. If you choose Standard System, you can view costs for a system that matches a
midpoint flow size in one of EPA's standard system size categories. Choosing User-Defined
System, however, will allow you to enter data and estimate costs for a particular system. When
you choose User-Defined System, a new instruction appears directing you to go to the User
Input worksheet (see Exhibit 3-2). Clicking on the new orange cell ("To User-Defined System
Page") will take you to the correct worksheet to use for user-defined inputs.
Exhibit 3-1. Step 1: Selecting the Operating Mode
Step 1:
Select a User-Defined System to estimate costs for a
single system using site-specific information in addition
to the default assumptions and prices, as needed. The
EPA Standard System option will generate costs for
the six standard designs shown in Step 4.
EPA Standard
System
If you are not prompted to enable macros, your Excel security settings may be set to automatically disable macros.
If this is the case, you will need to reset your security settings, close the workbook, and open it again. To check your
security settings, go to the Tools menu and select Macro, then click on Security. In the Security dialog box, on the
Security Level tab, make sure the radio button for "Medium" is selected. If the radio button "High" was selected
when you first opened the dialog box, Excel automatically disabled macros when opening the workbook, and you
must now close the POU-POE model and open it again (and choose to enable macros).
21
-------
Exhibit 3-2. Step 1: Additional Instruction for User-Defined System Mode
click here -->
To User-Defined System Page
3.2 User-Defined Model Inputs
Once you have chosen the operating mode, you can enter model inputs in the User Input
worksheet. This section describes the contents of the User Input worksheet. On this worksheet,
you will identify the following for the system you want to model:
• target contaminant
• treatment technology
• system size using either an average daily water flow rate or number of households
• assumptions to use for cost parameters and prices.
3.2.1 Target Contaminant
After selecting the operating mode, the next step is to select a target contaminant. Click on the
orange cell shown in Exhibit 3-3, then click on the down arrow to the right of the cell, to activate
the contaminant drop-down list. You can use your mouse pointer to move the scroll bar on the
right of the drop-down box to view the entire contaminant list. Select the contaminant you want
by clicking on it.
Exhibit 3-3. Step 2: Select a Target Contaminant Using the Contaminant Drop-Down List
Step 2:
Status: Contaminant selection OK
If you select a valid target contaminant from the list, the message below the contaminant name
will say "Contaminant selection OK." One of two warning messages will appear if your selection
is not valid:
• Please select a target contaminant indicates the contaminant cell is blank; click on the cell
again to activate the drop-down list and select one of the contaminants
• Warning: Target contaminant is not valid indicates the contaminant cell value is not
valid; click on the cell again to activate the drop-down list and select one of the
contaminants.
22
-------
Warning messages may appear elsewhere after you select a target contaminant. These will vanish
when you successfully complete Step 3.
3.2.2 Treatment Technology
After selecting a target contaminant, the next step is to select a treatment technology. Click on
the larger orange cell shown in Exhibit 3-4, then click on the down arrow that appears to the
right of the cell, to activate the treatment drop-down list. This list will vary depending on the
target contaminant because the types of certified treatment devices differ by contaminant (see
Chapter 2 for a list of certified treatment devices for each contaminant). Use your mouse pointer
to click on the technology for which you wish to estimate costs. Then, click on the smaller
orange cell below the treatment technology to show the drop-down arrow for this box, and after
clicking on the arrow, select whether you plan to include ultraviolet (UV) disinfection with your
POU or POE devices. A "yes" response will add UV equipment, installation, and maintenance
costs to your cost estimate. A "no" response excludes UV-related costs.
Exhibit 3-4. Step 3: Select a Treatment Technology Using the Technology Drop-Down List
Step 3:
Choose a treatment technology, and POU Reverse Osmosis
indicate whether treatment includes UV disinfection. no
Status: Treatment technology OK
If you select a valid technology from the list, the message area below the contaminant name will
display the message "Treatment technology OK." If your selection is not valid, one of three
warning messages will appear:
• Please select a certified treatment technology indicates the treatment cell is blank; click
on the cell again to activate the drop-down list and select one of the technologies
• Warning: Treatment technology selected is not certified for target contaminant
indicates the treatment technology is not valid; click on the cell again to activate the drop-
down list and select one of the technologies listed for the contaminant you selected (if a
Microsoft error dialog box also appears, click on the Cancel button, then click on the cell
again to select a technology)
• Warning: Invalid target contaminant; unable to evaluate validity of treatment
technology indicates that you need to return to Step 2 and select a valid contaminant, then
return to Step 3 to select a treatment technology.
3.2.3 System Size
There are two options for entering the information that the POU/POE model needs to estimate
costs for your system. The first option is to enter an average daily flow in gallons per day (gpd);
the second is to enter a number of households requiring treatment devices.9 Click on the gold
9 The POU/POE model is designed to estimate costs for a system that has a homogenous customer base with
uniform costs for treatment devices and services. A system that anticipates substantially nonuniform costs across its
customer base (e.g., with small residential customers and large nonresidential customers) will need to use a modified
23
-------
button next to item (a) of Step 4 (shown in Exhibit 3-5) to choose either "average flow" or
"household connections." After making this selection, you will be prompted to enter the required
information.
Exhibit 3-5. Step 4: Selecting System Size Option and Entering Size Data
Step 4:
Choose a method for entering system size. Then enter the required information in the gold cell(s).
a. Select method for system size input average flow
b. Enter average daily system flow in gallons per day, and 8,800 gallons per day
select a water source. surface water
Exhibit 3-5 shows that if you select the "average flow" option, the model will prompt you to
enter average daily flow in gallons per day and the type of source water. Source water options
are "ground water" or "surface water." The POU/POE model provides an approximate estimate
of the households served by the flow estimate you enter. If this estimate differs from the number
of households you expect to provide with treatment equipment and services, then returning to
Item 4 (a) and selecting the "household connections" option may provide a better cost estimate.
Exhibit 3-6 shows that doing this will remove the entry prompts for system flow and source
information, and activate a prompt to enter the number of household connections. This should be
the number of households requiring a treatment device and related services.
Exhibit 3-6. Step 4: Selecting the Household Connections Option for System Size Information
Step 4:
Choose a method for entering system size. Then enter the required information in the gold cell(s).
a. Select method for system size input household connections
b. Enter number of household connections
At this point, you can end your data entry and use the cost estimates based on default
assumptions and unit prices in the POU/POE model. A summary of the cost estimates appears to
the right of Step 5 (see Exhibit 3-7), along with a link to the detailed cost breakdown at the top of
the summary box. If, however, you have site-specific information, the next section provides
instructions for entering this information.
approach. One option is to convert all customers to a uniform measure (e.g., convert nonresidential customers to a
comparable number of residential customer equivalents based on flow or treatment device needs). Another option is
to run the POU/POE model more than once and combine capital and O&M cost results across customer groups.
24
-------
Exhibit 3-7. Cost Summary and Link to Cost Breakdown
Results summary (see
OUTPUT sheet for details)
Direct Capital Cost: $17,343
Total Capital Cost: $22,750
Annual O&M Cost: $4,471
Total $/kgal: $3.25
3.2.4 Additional User-Defined Inputs
After you have selected a contaminant, a treatment technology, and system size, the POU/POE
model generates costs using its default assumptions and prices (see Chapter 2 for descriptions).
The User Input worksheet reports these default values and also provides opportunities to override
these defaults with user-defined values. Exhibit 3-8 provides an example for the first two
components of capital costs - equipment purchase and installation. The default unit cost for the
POU/POE treatment device ($518.04 in Exhibit 3-8) is the average price from a database of
vendor prices in the model. To override the default cost, enter the unit price you prefer to use in
the blue cell to the left of the default value. Follow this procedure for all of the parameters shown
in Step 5 - enter values in the blue cells when you do not want to use the default values in your
cost estimate and leave the cells blank when you want to accept the default value. The following
tables contain the parameters by cost category. Some parameters are only active when the
treatment devices include a UV device; others are only active when the target contaminant is
nitrate. When these parameters are not active, they appear in a light gray italics text.
Exhibit 3-8. Step 5: Enter User-Defined Values to Replace Model Default Values
Capital Cost Assumptions
Equipment Purchase
Unit cost of POE/POU device without installation
Unit cost of UV system
$/unit
$/unit
Default Values
560.92
Equipment Installation
Wage rate for installation specialist (plumber/electrician)
Wage rate for system technical and maintenance labor
Wage rate for scheduling and administrative labor
POU/POE installation time
POU/POE installation scheduling time
UV installation time
$/hour
$/hour
$/hour
hours/household
hours/household
Default Values
33.12
25.07
17.89
2.00
0.50
hours/household
Exhibit 3-9 contains a list of the POU/POE equipment purchase and installation parameters that
can be revised. They include unit prices for the treatment devices, labor time to install each
POU/POE device, labor time to schedule each installation, and wage rates for three labor
categories: contractors who will install the devices, system staff who will maintain the devices
and system staff who will perform scheduling and administrative activities.
The POU/POE model includes start-up costs for an educational program. This program includes
preparing and distributing educational materials to inform the affected customers about the
POU/POE devices, treatment requirements, and utility activities. Exhibit 3-10 provides a list of
25
-------
the parameters used to estimate costs for this program. They include staff time to implement the
program as well as costs for various educational materials.
Exhibit 3-9. List of POU/POE Model Parameters for Equi
Parameter
Equipment Purchase
Unit cost of POE/POU device without installation
Unit cost of UV system*
Equipment Installation
Wage rate for installation
Wage rate for maintenance and technical labor
Wage rate for scheduling and administrative labor
POU/POE installation time
UV installation time*
Scheduling time
Units
$/unit
$/unit
$/hour
$/hour
$/hour
hours/household
hours/household
hours/household
pment Costs
User-Defined Entry
Enter dollar value
Enter dollar value
Enter dollar value
Enter dollar value
Enter dollar value
Enter average time in hours
Enter average time in hours
Enter average time in hours
* Active only when treatment equipment includes a UV device.
Exhibit 3-10. List of POU/POE Model Parameters for Initial Public Education Costs
Parameter
Technical Labor to Support Educational
Program
Develop technical content for educational
materials
Develop nitrate health impact information for
stakeholders*
Prepare for and attend public meetings
Post-meeting stakeholder communication
Clerical Labor to Support Educational
Program
Prepare educational materials for distribution
Prepare nitrate health impact information for
distribution*
Prepare for and attend public meetings
Prepare post-meeting materials for distribution
Communication Materials for Educational
Program
Print flyers announcing public meetings
Cost per flyer for pringint
Buy ads to announce public meetings
Cost per meeting ad
Print nitrate health impact flyers*
Print handouts for meetings
Print inserts for billing mailers
Cost to print handouts and mailers
Units
total hours
total hours
total hours
total hours
total hours
total hours
total hours
total hours
flyers
$/flyer
ads
$/ad
flyers
pages/household
pages/household
$/page
User-Defined Entry
Enter average time in hours
Enter average time in hours
Enter average time in hours
Enter average time in hours
Enter average time in hours
Enter average time in hours
Enter average time in hours
Enter average time in hours
Enter number of flyers
Enter cost in dollars
Enter number of ads
Enter cost in dollars
Enter number of flyers
Enter average pages per household
for meeting handouts
Enter average pages per household
for billing mailers
Enter cost in dollars
* Active only when target contaminant is nitrate.
26
-------
The POU/POE model includes first-year monitoring costs that are separate from annual
monitoring costs to allow monitoring assumptions to differ between the first year after
equipment installation and subsequent years. Exhibit 3-11 shows the parameters that affect costs.
They include the staff time required to take the sample (including travel time), the staff time to
schedule the sampling event, the sampling frequency (i.e., the number of samples per household
taken during the first year), and the fraction of households to include in the initial sampling
event.10
Exhibit 3-11. List of POU/POE Model Parameters for Initial Monitoring Costs
Parameter
Units
User-Defined Entry
First Year Water Quality Sampling
Time to take sample during first year
Time to schedule sample event at
household
Number of samples per household during
the first year
Fraction of households sampled during the
first year
Laboratory analysis fee
Sample shipping cost (bulk)
hours/sample
hours/sample
samples/household
% households
$/sample
$/bulk shipment
Enter average time in hours
Enter average time in hours
Enter number of target contaminant
samples per household
Enter percent of total households
included in sampling event
Enter cost per sample
Enter package shipping cost
The final set of capital costs includes a series of indirect costs that account for permitting, pilot
testing, legal, engineering design, and contingency costs. Chapter 2 provides descriptions of the
types of costs to include in each category. As Exhibit 3-12 illustrates, there are two options for
entering each of these costs. The first option is to enter a percentage, which will act as a
multiplier on installed equipment costs to represent the indirect cost. For example, an indirect
pilot study cost parameter of 3% will add $3 of pilot study costs to every $100 of installed
equipment costs (including equipment and installation labor costs). The second option is to enter
a dollar value. If pilot study costs are known and fixed, this option will provide a better result. In
this section of the user input worksheet, enter the values in the blue cells and indicate the units
using drop-down lists in the cells to the right of each blue cell. u The units can differ across the
indirect costs, i.e., some can be percentages while others are dollar values.
10 The model addresses the overlap between initial year sampling and annual sampling by subtracting annual
sampling events from the initial year sampling events to estimate capital costs. If 33% of households are monitored
annually and 100% are monitored during the first year, the overlap during the installation year leads to 133% of
households sampled. By subtracting 33% from 100%, the sum of initial sampling events in capital costs and annual
sampling events in O&M costs is 100% for the installation year.
11 Note that if the unit selected is "% installed equipment cost," the number in the cell will range from 0 to 100
(rather than a decimal between 0 and 1). The POU/POE model detects whether the range should be interpreted as a
percentage or a dollar value based on the unit selected ("% installed equipment cost" or "$ fixed cost").
27
-------
Exhibit 3-12. List of POU/POE Model Parameters for Indirect Capital Costs
Parameter
Units
User-Defined Entry
Cost to obtain operating permit
Cost to conduct pilot test
Cost for legal activities (e.g., ordinance changes)
Cost for engineering activities (e.g., device selection)
Contingency cost (unknown factors)
% installed equipment cost
or dollar value
Enter a dollar value or a
percentage (e.g., enter 5 for
5%) and select the correct units
from the drop-down list that
appears when you click the
units cell
All of the input parameters in the preceding tables affect the POU/POE model's estimate of
capital costs. The remaining parameters, described below, affect annual O&M costs.
The first set of O&M cost parameters define the overall frequency of maintenance visits and the
length of time per visit. Exhibit 3-13 shows that visit frequency and duration can vary for the
primary treatment device and a UV device, if needed. The equations in the cost model reflect an
assumption that these visits overlap. For example, if there are two trips per year for POU/POE
device maintenance and one per year for UV device maintenance, then the total number of trips
will be two. The maintenance labor hours, however, are additive across the POU/POE and UV
devices, if both are used. Parameters for equipment costs are addressed in a later exhibit.
Exhibit 3-13. List of POU/POE Model Parameters for Equipment Maintenance Labor Costs
Parameter
POU/POE maintenance
POU/POE replacement frequency
UV maintenance*
UV maintenance frequency*
Scheduling time
Units
hours/visit
visits/household/yr
hours/visit
visits/household/year
hours/visit
User-Defined Entry
Enter average time in hours
Enter average number of annual
visits per household
Enter average time in hours
Enter average number of annual
visits per household
Enter average time in hours
* Active only when treatment equipment includes UV device.
Exhibit 3-14 shows the parameters used to derive annual educational program costs. The
activities include developing updates to materials that might be distributed during maintenance
visits. If the target contaminant is nitrate, there are a few parameters for annual updates to
information materials distributed to all customers.
28
-------
Exhibit 3-14. List of POU/POE Model Parameters for Annual Educational Costs
Parameter
Technical Labor to Support On-Going Education
Program Activities
Develop updates to technical information
Develop updates for nitrate health effects*
Clerical Labor to Support On-Going Education
Program Activities
Prepare information updates for distribution
Prepare nitrate information updates for distribution*
Communication Materials to Support On-Going
Education Program Activities
Nitrate fliers*
Billing mailers
Units
hours
hours
hours
hours
flyers
pages/household
User-Defined Entry
Enter average annual hours
Enter average annual hours
Enter average annual hours
Enter average annual hours
Enter number of flyers
Enter average pages per
household for billing mailers
* Active only when the target contaminant is nitrate.
The POU/POE model includes annual monitoring costs. The relevant parameters shown in
Exhibit 3-15 affect sampling frequency and duration.
Exhibit 3-15. List of POU/POE Model Parameters for Annual Monitoring Costs
Parameter
Units
User-Defined Entry
Sampling time (including travel)
Sampling scheduling time
Analysis frequency (samples)
Analysis frequency (percent)
hrs/sample
hrs/sample
samples/household/year
% households/year
Enter average time in hours
Enter average time in hours
Enter number of annual
samples per household included
in the sampling event
Enter percent of households
included in the annual sampling
event
The final O&M costs pertain to equipment replacement. The POU/POE model accommodates
replacement schedules and costs for up to four device components, as well as UV lamp and
sleeve replacement, when applicable. User-defined data can override one or more of the default
equipment replacement frequencies or costs. Exhibit 3-16 shows the user input form for
equipment replacement costs.
29
-------
Exhibit 3-16. List of POU/POE Model Parameters for Annual Equipment Replacement Costs
Parameter | Units I User-Defined Entry"
Parts Replacement
Name of Replacement Part 1
Unit Cost of Replacement Part 1
Replacement Frequency
Name of Replacement Part 2
Unit Cost of Replacement Part 2
Replacement Frequency
Name of Replacement Part 3
Unit Cost of Replacement Part 3
Replacement Frequency
Name of Replacement Part 4
Unit Cost of Replacement Part 4
Replacement Frequency
Replacement UV Lamp*
Replacement Frequency*
Replacement UV Quartz Sleeve*
Replacement Frequency*
NA - name of part
$/part
times/household/yr
$/part
times/household/yr
$/part
times/household/yr
$/part
times/household/yr
$/part
times/yr
$/part
times/yr
Enter equipment name
Enter dollar value
Enter average annual
replacement frequency
Enter equipment name
Enter dollar value
Enter average annual
replacement frequency
Enter equipment name
Enter dollar value
Enter average annual
replacement frequency
Enter equipment name
Enter dollar value
Enter average annual
replacement frequency
Enter dollar value
Enter average annual
replacement frequency
Enter dollar value
Enter average annual
replacement frequency
3.3 Understanding the Output Worksheet
The Output worksheet shows all of the cost components and the cost build-up. Exhibit 3-17
provides an example of the elements reported for equipment costs. The top section of the Output
worksheet provides a summary of the target contaminant, treatment, and system size information
provided on the User Input worksheet (or, in the Standard Input mode, the Input worksheet).
The main section of the Output sheet shows the cost build-up for direct capital costs, with a line
item for each component. Exhibit 3-17 illustrates the capital cost reporting elements for the
POU/POE model. Each line item includes the name of the component, the quantity (e.g., number
of devices or hours per household), frequency (usually the number of households affected), and
unit cost (e.g., cost per device or per staff hour).
Each line item also includes the total cost for the component and, where applicable, a useful life
estimate (or a dash indicating the useful life is not applicable). The useful life estimate generally
reflects the expected life of the installed device, taking into account engineering constraints (e.g.,
the expected life of a component until it must be replaced) as well as regulatory change (e.g., the
30
-------
potential for regulations to change, requiring equipment replacement even if the equipment still
functions as designed). The useful life estimate is used to convert upfront capital costs to
annualized capital costs, which can then be added to annual O&M costs to obtain an annual total
cost estimate and an average unit cost ($/kgal) designed for comparison with centralized
treatment costs.
Exhibit 3-17. Example of Cost Details Provided on the Output Worksheet
System Description
Households Served
Contaminant
Treatment Technology
Treatment Location
UV Treatment
25
Arsenic
POU Reverse Osmosis
POU
no
Capital Costs
WBS # Item
Design
Quantity
Frequency
23.1.1
23.1.2
23.1.3
23.1.4
23.1.5
POU/POE Unit Purchase
POU/POE Installation
Scheduling Time
UV Purchase
UV Installation
1 unit/household
2.00 hours/household
0.50 hours/household
0 unit/household
0.00 hours/household
25 households
25 households
25 households
0 households
0 households
Unit Cost Total Cost
561 $ 14,023
33.12 $ 1,656
17.89 $ 224
23.2.1.1 Develop materials
23.2.1.2 Nitrate health effects
23.2.1.3 Meetings
23.2.1.4 Post-meeting
10.00 hours
0.00 hours
2.00 hours
2.00 hours
25.07 $ 251
25.07 $
25.07 $ 50
25.07 $ 50
" '!,:y/', '{/I •••'' ':: " '^".'.i^f'Vllv'ffffffffffffffffW^
23.2.2.1
23.2.2.2
23.2.2.3
23.2.2.4
• i ;?&
23.2.3.1
23.2.3.2
23.2.3.3
23.2.3.4
23.2.3.5
Develop materials
Nitrate health effects
Meetings
Post-meeting
%: " .! ''<
Meeting flyers
Meeting ads
Nitrate awareness flyers
Meeting handouts
Billing mailers
6.00
0.00
2.00
2.00
10
1
0
3
2
hours
hours
hours
hours
flyers
ads
flyers
pages/household 25 households
pages/household 25 households
$
$
$
$
$
$
$
$
17.89
17.89
17.89
17.89
2.12
42.46
2.12
0.08
0.08
$
$
$
WM
$
$
$
$
$
107
-
36
36
21
42
-
6
4
31
------- |