United States
Environmental Protection      Office of Water       EPA 812-B-96-002
Agency            4604          January 19%


REGULATORY IMPACT ANALYSIS


ADDENDUM




Proposed Changes to National Primary

Drinking Water Regulations for Lead

and Copper

-------

-------
     Regulatory Impact Analysis Addendum:
              Proposed Changes to
National Primary Drinking Water Regulations For
                 Lead and Copper
                     Prepared By:

                 The Cadmus Group
            4900 Seminary Road, Suite 600
              Alexandria, Virginia 22311
               EPA Contract No. 68-C4-0007
                  Work Assignment 2-18

                 Ms. Patricia Minami
                    Project Officer

                  Ms. Judy Lebowich
                 Technical Project Monitor

          U.S. Environmental Protection Agency
       Office of Ground Water and Drinking Water
                  Washington D.C.

                    January 26, 1996

-------
                              TABLE OF CONTENTS
                                                                              PAGE
1.0    Executive Summary  	  1

2.0    Problem Statement	  3

3.0    Market Imperfections, the Need for Federal Regulation, and Consideration of
       Alternatives	  3

4.0    Assessment of Total National Costs	  3
       4.1    Costs of Monitoring	  5
       4.2    Costs of Coping With Source Water Contamination	  19
       4.3    Costs of Coping With Lead Leached From Solder  	  19
       4.4    Costs of Coping With Lead Leached From Lead Pipes	  19
             4.4.1  Cost of Corrosion Control and Public Education to Cope With
                   Leaching from Pipes	  19
             4.4.2  Cost of Lead Service Line Replacement  	  23
       4.5    State Implementation Costs  	  24

5.0    The Benefits of Reducing Lead in Drinking Water	  27

6.0    Regulatory Flexibility Analysis and Paperwork Reduction Analysis  	  28

-------
1.0    Executive Summary

EPA is proposing several minor revisions to the final Lead and Copper Rule (LCR), which was
promulgated on June 7, 1991 (56 Federal Register 26460).  These minor revisions eliminate
unnecessary requirements and clarify other requirements.  The Agency's intent in proposing them
is to promote consistent implementation of the LCR nationwide, streamline c ic reduce reporting
burden. EPA is confident that the proposed minor revisions will not change the level of public
health protection afforded by the final rule. Accordingly, EPA has updated the 1991 Regulatory
Impact Analysis (RIA), Regulatory Flexibility Analysis, and Information Collection Request
(ICR)1 prepared in support of the final LCR. This RIA addendum updates the impacts estimated
in the 1991 RIA and Regulatory Flexibility Analysis. A separate ICR document presents the
recordkeeping and reporting impacts of the revisions that EPA is proposing.

At the time of promulgation, based on data from a select group of U.S. cities, the Agency
estimated that the cost for water systems to comply with the various treatment requirements
would total between $500 and $790 million per year. Household costs were estimated to range
from less than $1 per year for large systems (serving over 50,000 people) to $2 to $20 per year
per household in smaller systems.  Now that water systems have collected lead samples from
hundreds of thousands of household taps around the country to comply with the monitoring
requirements of the rule, much more reliable predictions of costs (and benefits) can be made. It is
clear that significantly fewer systems will be required to install corrosion control and, therefore,
both costs and benefits associated with the rule are less than originally predicted. EPA would now
estimate that costs associated with the rule are roughly $200 million per year, resulting in reduced
lead exposure for approximately 40 million Americans. Health benefits associated with these
exposure reductions would still be substantial, totaling over $1 billion per year and resulting in an
estimated 200,000 young children whose blood lead levels are reduced to below the Centers for
Disease Control (CDC)/EPA action level of 10 micrograms per deciliter.

To calculate the relative magnitude of the regulatory revisions that EPA is proposing, the original
cost model and  the same basic assumptions regarding impacts of the individual rule components
were used.  It is estimated that the proposed revisions to the LCR will result in a decrease in
national costs of $1.88 million annually. Regardless of the baseline used, it is clear that the
projected impacts of the proposed regulatory revisions will be minimal compared to the total
national costs associated with the lead and copper regulations. Most of the reduction results from
proposed changes in monitoring and public education delivery requirements. Proposed revisions
to the information States are required to report to EPA also contribute to the reduction in State
implementation costs.  Overall, EPA estimates the proposed changes will result in a very minor
reduction (less than one percent) to the annual drinking water program implementation costs and
the Agency does not believe this percentage reduction  would change substantially if costs for the
entire rule were recalculated.
       Final Information Collection Request for National Primary Drinking Water Regulations for Lead and
       Copper. U.S. Environmental Protection Agency. April, 1991.

-------
     a;
     u
X
       .

     11
        g
        o
     «  s
<- ._ «
s|l
^ .2

=*'i «
flj S ts
gm v) Si
<-> < i
o
u £
<-> J
^^ ftS
1 u
*^
o
y ^*>
-<• «
2 •-
_ to
Sv (*l
<7\

tf
y
u-
o
M

Cost Componen
,
_o
'c?
























V)
•**
e
U
i
**
£
L.
*rf
as
3
04


S
1





g



g










00
Monitorin



o






o
fN
fN
*rr


fN



1
Q
o
DO
p^
Treatment (Includil
«•
i
c
2
U



o






S



I
w>







1
rt
H
1




1
•





0



o










1
(J
uu
o
1
CL



o
o
+



^
2
3
a
a

5
-„
2
S
1

-------
 Overall, the LCR revisions would result in only slight changes in national costs. Exhibit 1 is a
 detailed comparison of cost estimates for various rule components. Unless otherwise noted, the
 costs presented in Exhibit 1 represent costs to public water systems. State implementation costs
 associated with the minor LCR revisions are described in Chapter 4.5 of this document.

 2.0    Problem Statement

 The Final Regulatory Impact Analysis of the National Primary Drinking Water Regulations for
 Lead and Copper, (Lead RIA of April 1991), which accompanied the final rule, discussed the
 problem of lead in drinking water and how to control it  Health effects, occurrence, and
 treatment technologies were among the topics covered.

 Although EPA, the states, and the public water systems themselves today have a more thorough
 understanding of the issues associated with implementing this rule, the need to regulate lead and
 copper and the underlying approach to doing so discussed in the April 1991 RIA have not
 changed. Consequently, Chapter 2 of the April 1991 RIA, the Problem Statement, is not revised
 in this RIA addendum.

 3.0    Market Imperfections, the Need for Federal Regulation, and Consideration of
       Alternatives

 As states implemented the provisions of the final LCR, they identified unnecessary requirements
 that should be eliminated and other changes that should be made to improve the consistency of
 the rule's implementation nationwide. The proposed revisions discussed in this document
 incorporate the findings of a work group comprised of EPA Headquarters and regional staff and
 state drinking water officials. EPA developed the proposed revisions after extensive consultations
 with state agencies charged with implementing the LCR. Further, several of the proposed
 revisions, some of which eliminate certain system reporting requirements, are the result of
 Stakeholder meetings that the Office of Ground Water and Drinking Water (OGWDW) held  in
 response to Agency-wide burden reduction initiatives. The market imperfections, need for federal
regulation, and EPA's approach to regulating lead and copper in drinking water have not changed
since the April 1991 RIA. Consequently, Chapter 3 of that document is not revised in this RIA
addendum.

4.0    Assessment of Total National Costs

Chapter 4 contains a detailed description of how the proposed modifications are expected to
impact the total national costs associated with LCR compliance. These modifications will impact
both public water systems and states. State cost and burden impacts that could be quantified are
described separately in Chapter 4.5,  although state activities are described minimally in the
Chapters 4.1 through Chapters 4.4 to the extent this description is needed to complete the
discussion associated with system impacts.

-------
To identify relevant assumptions for analyzing the impacts of EPA's proposed revisions to the
LCR, the April 1991 RIA and ICR were reviewed. The May 1988 RIA, which the April 1991
RIA revised, was reviewed as well.

The minor revisions now proposed by EPA are estimated to result in a cost decrease of $1.88
million annually, which is a change of approximately 0.4 percent.  This estimated impact can be
considered negligible given the underlying uncertainties in model input parameters. Annual cost
estimates for the proposed revisions are averaged over the 13 years remaining in the 18-year LCR
planning period. The cost estimates for the revisions include increased labor rates for both public
water systems and states.

The current revisions were assessed  in light of the assumptions used in preparing the 1991 RIA
and ICR documents and, to the extent possible, were evaluated using models developed for these
analyses. Where the RIA and ICR models did not support analysis of the current revisions, some
additional basic analyses were conducted to determine the cost of the change. Given the
simplifying assumptions used to estimate national cost impacts in the 1991 RIA, the impacts
associated with some of the minor revisions could not be estimated quantitatively and are instead
described qualitatively. For example, some of the LCR modifications provide states with
authority to require water systems to perform additional activities such as conducting public
education for systems that typically would not be subject to this requirement or increased
monitoring.  For the most part, the cost impacts associated with such state-specified requirements
were not included in this RIA.  Proposed rule changes that do not have any impact on national
costs have been identified as well.

National state implementation costs are estimated to decrease by approximately $0.31 million
annually. These costs have been estimated taking into consideration the changes proposed and
described in the preamble to the Notice of Proposed Rulemaking. The most significant cost
impacts are associated with proposed changes in  monitoring and public education.  These changes
affect water systems as well as states with primary authority to enforce the drinking water
regulations (also known as primacy).7 Changes to lead service line requirements are expected to
result in a minor cost decrease for systems and states. The proposed revisions are not expected to
impact corrosion control or source water treatment. In presenting these cost estimates, the RIA
addendum identifies, to the extent possible, the cost impacts associated with each proposed rule
change.

To aid comparisons with the April 1991 RIA, the cost changes in this addendum follow the
outline and numbering scheme used in the 1991 RIA.
       For simplicity, the remainder of this document will use the term state to mean a state with primacy for the LCR
       or the appropriate EPA Region (where the state does not have primacy for the rule) or any federally-recognized
       Indian Tribe that has been granted primacy for the program.

-------
4.1    Costs of Monitoring

Many of the proposed changes to the final rule address monitoring requirements. Each proposed
change is discussed along with the analytical approach adopted to evaluate its impact on national
costs.

S 141.81(bV2)

This change would require systems optimizing corrosion control under § 141.81(bX2) (or "b2"
systems)* to comply with the requirements for continuing tap water and water quality parameter
monitoring specified in § 141.86(d)(3) and § 141.87(d), respectively.  The current regulation
initially requires tap water monitoring by all systems. It also requires water quality parameter
monitoring by all large systems and by small and medium systems during the monitoring periods in
which they exceed either the lead or copper action level. The current rule inadvertently omitted
language that specifically required "b2" systems to continue lead and copper tap water monitoring
and water quality parameter monitoring after the state set water quality parameters.  Therefore,
this revision proposes language to correct this omission.

The April  1991 RIA also explicitly assumed that "b2" systems would perform lead and copper tap
water monitoring, but not water quality parameter monitoring.  Therefore, this modification
results in a monitoring cost  increase only for small and medium Hb2" systems that exceed the lead
or copper action level.  There is no water quality parameter monitoring cost increase for large
systems because the 1991 RIA assumed all large systems, including "b2H systems, were
performing water quality parameter monitoring. EPA believes, however, that the water quality
parameter monitoring cost increase will be small because: 1.) few small and medium systems are
expected to have installed corrosion control treatment prior to promulgation of the LCR and 2.)
few "b2" systems are expected to exceed an action level since they already have corrosion control
treatment in place.

§ 141.81(bX3)

Attainment of Copper Action Level

This change would require systems seeking to be deemed to optimize corrosion control under
§ 141.81(b)(3) to meet the copper action level of 1.3 mg/1 for two consecutive monitoring
periods, in addition to the current requirement for  lead.  The current lead requirement is to
demonstrate that the difference between the ninetieth percentile lead level as  measured at the tap
and the lead level measured  in source  water is below the lead  Practical Quantitation Limit of
0.005 mg/1 for two consecutive 6-month monitoring periods, before being considered to have
optimized corrosion control under § I41.81(b)(3).
       A system optimizing treatment under § 141.81(b)(2) (i.e., a "b2" system) has demonstrated that it has
       conducted activities equivalent to corrosion control treatment steps in the rule.

-------
 The April 1991 RIA incorporated EPA's assumption at the time that systems meeting the lead
 action level would also meet the copper action level.  Furthermore, the April 1991 RJA explicitly
 assumed that each time lead was analyzed, copper was also analyzed.  Therefore, no RIA cost
 increase is imposed by this change.

 Continued Tap Water Monitoring

 This change would require systems meeting the § 141.81(b)(3) criteria (or "b3" systems) to
 continue to conduct lead and copper tap water monitoring to ensure that lead and copper
 concentrations remain at levels that are protective of public health. Although "b3H systems are
 not required to install corrosion control treatment, they may be currently treating their water (or
 later install treatment) for other reasons that may indirectly affect the water's corrosivity.  Under
 the proposed modification, all Hb3"  systems would be required to collect lead and copper tap
 water samples triennially (once every three calendar years) at the reduced number of sites
 specified in § 141.86(c).

 The April 1991 RIA did not distinguish between "bl,"9 Hb2," or "b3" systems; instead, the RIA
 evaluated the percentages of systems that would adopt different compliance pathways (for
 examples, see Exhibits 4-3 and 4-4 in the April 1991 RIA) and hence different monitoring
 schedules. The number of systems assuming a particular pathway was based on estimates  of the
 percentage of systems that met or failed the lead and/or copper action levels.

 No cost impact is estimated for this change because of the monitoring assumptions contained in
 the 1991 RIA model.  In particular, the model assumed that a set of large systems (i.e., "pathway
 1" systems) would monitor in a similar fashion to medium and small systems that do not exceed
 the lead action level. Although the model does not specifically identify "bl," "b2," and "b3"
 systems, EPA assumed "pathway 1" large systems to be "b3" systems. According to the model,
 these systems would have continued monitoring and would have already qualified for triennial
 monitoring in 1996 (before the changes become effective).

 § 141.86

Holding Time for Acidified Lead or Copper Samples

 EPA is proposing to modify the current analytical methods used for measuring lead and copper.
 This change would decrease the holding time specified in § 141.86(b)(2) for acidified lead or
 copper samples from 28 to 16 hours after acidification in the  laboratory. This change is assumed
 not to have a measurable effect on the monitoring burden or cost of the rule, although reducing
 the holding time will increase the number of samples that can be processed in a day.  Further,
 because EPA has already modified the holding times for other inorganic contaminants to 16
 hours, some water systems may realize a cost savings because this provision eliminates the need
       "b I" systems are those systems that are deemed to optimize corrosion control under § 141.81 (b)( 1) because
       they meet the lead and copper action levels for two consecutive monitoring periods

-------
to coordinate separate acidification holding times for lead and copper with other inorganic
contaminants

Non-Transient, Non-Community Water System Sampling Sites

Some non-transient, non-community water systems (NTNCWSs) do not have enough taps that
can provide first-draw samples (i.e., one liter of tap water collected in accordance with the
provisions in
§ 141,86(b)(2) that has been standing in plumbing pipes at least six hours and is collected without
flushing the tap). This change would allow such systems to collect"non-first-draw" samples, if
the sampling locations and sampling times are first approved by the state.  Systems would be
required to submit written requests to the state before including such taps in their sampling pool
This change is necessary because a number of non-transient, non-community water systems
operate 24 hours a day, and therefore, cannot collect a sufficient number of first-draw samples.

Systems using non-first-draw samples would be required by § 141 90(aX2) to provide the state
with a letter requesting approval of proposed sampling locations and times for non-first draw
samples that are expected to result in the longest standing times.  EPA estimates that
approximately 500 systems will seek state permission to use non-first-draw sampling sites.
However, no cost impact is estimated for this change because the cost for the system to prepare a
letter for the state to review is assumed to be more than offset by the cost now avoided because
the system would no longer need to disrupt its normal operations to obtain a sample meeting the
first-draw criteria.

This provision may result in an additional cost decrease for systems. In particular, some systems
should no longer have to sample at non-representative sites and times because of an insufficient
number of first-draw sites; consequently, these systems could  potentially avoid the costs
associated with corrosion control treatment installation.

Reduced Tap Water Monitoring Sites

Under § 141.86(d)(4), systems serving more than 100 people that qualify for reduced lead and
copper tap water monitoring are allowed to reduce the number of sites from which they must
draw samples. Section 141.86 does not specify which sampling sites should be included in the
reduced sampling pool. EPA is concerned that some systems may select the sample sites with the
lowest concentrations of lead and copper, thus yielding artificially low lead and copper monitoring
results.  To prevent this, EPA proposes a minor change to § 141.86(c) that:  1.) specifies that
systems must choose representative sample sites and 2.) gives states the authority to designate,
when they feel such designation is appropriate, which sampling sites a system must use when
conducting reduced monitoring.

The April 1991 RIA assumed states would not designate the sites that systems must use for
resampling. Since under the proposed revision, states would be given the option of designating
resampling sites, but are not required to designate resampling sites, additional costs are not

-------
 directly imposed by this change. It is reasonable to assume, however, that some states will choose
 to designate resampling sites for certain systems.  The affected systems would incur additional
 costs associated with reviewing the state's letter, which would specify the sites that can be
 included in the reduced sampling pool.

 It is assumed the cost for a system to review the state's letter would be offset by the cost now
 avoided because the system would no longer need to determine which locations to include in its
 reduced sampling pool. Therefore, no cost impact is estimated for this change.

 Sufficient information is not available to determine precisely the number of states that will
 designate resampling sites or the number of systems that will be affected. In the 1991 RIA,
 79,000 water systems were assumed  to be regulated under the LCR. Of these, 46,000 systems
 are assumed to be capable of reducing the number of sampling sites used.  EPA believes,
 however, that no more than twenty percent of the systems capable of reducing the number of
 sampling sites used would be affected by this proposed revision. Note that there are about 33,000
 systems serving fewer than 100 people that cannot reduce the number of sampling sites used.
 These systems are required to sample only five sites—the minimum required by the LCR.

 Because the assumptions described above are considered to be conservative, the number of
 systems that incur costs due to this revision may actually be smaller.  The true monitoring burden
 may be significantly lower because some states may issue broad guidance regarding reduced
 sampling sites.

 Systems Unable to Locate Sufficient  Tier 1, 2 and 3 Sites

 EPA is proposing revisions to § 141.86(a)(5) and § 141.86(a)(7) to clarify that community water
 systems and non-transient, non-community water systems, respectively, having an insufficient
 number of tier 1, 2, or 3 sites can complete their sampling pool with representative sites
 throughout their distribution system.  This provision is needed because the current LCR does not
 address situations where a system is unable to identify a sufficient number of sites meeting the
tiering criteria specified in § 141.86(a).  Consequently, this has resulted in confusion about
whether such systems are required to monitor, and if so, how they should select their sampling
 sites.

No  cost impact, however, is expected to result for this revision because the original RIA cost
 models assumed all systems subject to the LCR would have a sufficient number of sampling sites
 meeting the tiering criteria.  Further, no cost impact is anticipated because the system is not
 required to first obtain state approval before including non-tiered sites in their sampling pool.

Sample Site Justification

 Current LCR provisions at §§ 141.90(a)(2) and (3) and § 141.90(a)(4), respectively, require
 systems to notify the state if they are  unable to complete their sampling pool with Tier 1 sites or
 are  unable to collect 50 percent of their samples from sites served by lead service lines.  EPA is

-------
proposing to eliminate these requirements because states can use other mechanisms such as on-
site inspections or file reviews to determine if systems are sampling at appropriate locations. No
cost impact is estimated for these modifications because, in general, systems are assumed to have
submitted these justifications prior to the start of initial monitoring, the date of which has already
passed. Additional justifications may have been required to be submitted by some systems that
could not gain entry to the same sampling locations for subsequent monitoring periods.  However,
the costs models were not designed to estimate the number of systems fitting this scenario.

Timing of Reduced Monitoring

As required by § 141.86(d)(4)(iv), systems that monitor tap water for lead and copper annually or
less frequently must sample during the months of June, July, August, or September. However,
some non-transient, non-community water systems are closed during those months. Therefore,
samples taken in months when the system is not operating would result in data that are not
representative of water actually used for drinking. To remedy this problem, EPA is allowing
systems that do not operate from June through September to monitor during their warmest
month(s) of operation.  The system, not the state, would be responsible for determining the
month(s).

The April 1991 RIA did not account for off-season monitoring costs, and the cost model used in
the 1991 RIA is not precise enough to quantify this level of detail.  Consequently, no cost impact
is estimated for this proposed revision, although some cost savings could be expected. Off-season
monitoring may have imposed an extra burden on systems that do not operate during the warmest
months of the year, but were required by the LCR to take samples in these months.  Specifically,
there may have been additional costs associated with taking special steps to collect samples; this
change would eliminate such costs.

Accelerated Reduced Lead and Copper Monitoring

This  change would allow water systems with very low lead and copper levels (i.e., those systems
having ninetieth percentile lead levels of less than or equal to 0.005 mg/1 and ninetieth percentile
copper levels of less than or equal to 0.65 mg/1) for two consecutive six-month monitoring
periods to assume a triennial monitoring schedule for lead and copper tap sampling more rapidly
than  permitted by the current rule.

The April 1991 RIA assumed that three years of annual monitoring would be required before a
system could assume a triennial monitoring schedule.  If this proposed change becomes effective
in 1997 (based on EPA's current schedule for promulgation),  alt systems that completed their first
two rounds of monitoring on schedule and met the action levels will not benefit from this change.
However, systems that will have to install corrosion control treatment because they exceed the
lead and/or copper action levels will benefit by being able to reduce monitoring sooner than
projected in the April 1991 RIA. The model indicates that  these systems would be able to

-------
 eliminate an average of approximately 7,077'° lead and copper samples per year with a monitoring
 cost of $50 per lead and copper sample. This would result in a reduction in system monitoring
 costs of $353,800.ll

 In addition, a few large "b2" and "b3" systems that did not continue lead and copper tap
 monitoring beyond initial monitoring may also benefit. However, the 1991 RIA monitoring cost
 models are not sufficiently precise to calculate the cost savings for these categories of systems.

 Sample Invalidation

 EPA recognizes that there may be occasions when a monitoring sample is not collected or
 analyzed properly or would otherwise not provide reliable results.  In such cases, EPA believes,
 the state should have the authority to exclude the test results from such samples  Consequently,
 the Agency proposes to allow States to invalidate lead and copper monitoring samples under four
 circumstances:

       •       When the laboratory establishes that improper sampling analysis caused erroneous
              results;

       •       When the state determines that the sample was taken from a site that does not
              meet the site selection criteria of §  141.86;

       •       When the sample container is damaged in transit; and

       •       When the state has substantial reason to believe that the sample was subject to
              tampering.

 Systems are required to report all sample results to the state. If a system seeks to have any
 samples invalidated, the system must provide supporting evidence to the state in writing as
required by § 141.90(a)(l)(ii). The  state also must document in writing all decisions to invalidate
samples.  The proposed § 141.86(f)(3) prohibits states from invalidating a sample solely because  a
subsequent sample taken at the same site is higher or lower than the original sample.

Some systems take only the required number of samples in each monitoring period, so any
invalidated samples would decrease  the number of samples below the minimum sampling
requirement.  Under such circumstances, a system would incur a monitoring violation, but EPA
       The annualized number of samples avoided each year (7,077) was derived by multiplying the average number
       of samples collected (8.2 samples) by the number of systems affected (5,600) and by the number of rounds of
       monitoring avoided (2). The total of 92,000 samples was divided by 13, the remaining years in the 18-year
       LCR planning period. It should be noted, however, that the 92,000 eliminated samples would not be evenly
       distributed over this 13-year period.

       This cost includes a $15 per hour labor rate for junior staff and is consistent with the Information Collection
       Request for the 1995 Drinking Water Needs Survey, OMB No. 2040-0176.

                                            10

-------
believes that it is reasonable to allow these systems to replace any invalidated samples.  The
proposed § 141.86(f)(4) requires that any replacement samples be taken at the same locations as
the invalidated sample(s) or, if not possible, at locations other than those already used for
sampling during the monitoring period. Any replacement samples also must be collected by the
end of the monitoring period or, if that is not possible because the invalidated sample was drawn
too late in the monitoring period, up to 20 days after the state invalidates the sample.

A water system would need to replace invalidated samples only if it did not have sufficient valid
samples to meet minimum sample requirements.  EPA encourages systems to take more samples
than required by the LCR and not to wait until the end of the monitoring period to complete
sampling.  In this way, if any samples are invalidated, the system has a "cushion" and should not
need to take replacement  samples.

The April  1991 RIA did not assume that samples could be invalidated. Because systems may be
required to replace invalidated samples in order to have  a sufficient number of samples for
calculating ninetieth percentile values, systems may incur some monitoring costs in addition to
those presented in the April 1991 RIA.

EPA assumes that less than 0.5 percent of the approximately 79,000 systems will seek to have
samples invalidated each monitoring period.  The burden incurred by a system in writing a letter
to the state requesting that a samples) be invalidated is  assumed to be approximately one hour.
Assuming  that the system labor cost is approximately $15 per hour, the cost of preparing each
letter is $15.

Many water systems and states have indicated their desire to modify  the LCR to allow sample
invalidation, despite the modest increase in transactional costs associated with this provision. The
burden imposed by this provision could be offset significantly if a water system is able to avoid
having to install corrosion control treatment because of an erroneous sample result that would
otherwise  have to be included in their ninetieth percentile calculation. The potential treatment
cost avoided will vary based on system size.  In particular, the 1991 RIA estimated the following
corrosion control treatment costs ranges for small, medium, and large systems, respectively:
$2,000 to  $6,000; $20,000 to $75,000, and $120,000 to $3 million.  Other potential cost savings
associated with this change include the costs avoided for systems that would have had to resume a
standard monitoring schedule because an erroneous sample resulted in an action level exceedance.

Monitoring Waivers for "All Plastic" Systems

In § 141.86(g) EPA is proposing to allow small systems (i.e., those serving 3,300 or fewer) to
obtain a monitoring waiver if they meet certain criteria.  In particular, the system must:
                                           11

-------
       provide the state with a certification that the system itself and all buildings connected to
       the system are free of all lead-containing and copper-containing materials, as specified in §
       141.86(g)(l);and

       demonstrate that the ninetieth percentile lead and copper levels for any and all rounds of
       monitoring performed since the system became "all plastic" do not exceed 0.005 mg/1 of
       lead and 0.65  mg/1 of copper.

The conditions of the waiver would require tap water monitoring once every nine years at the
reduced number of sampling sites specified in § 141.86(c).  If the results of this monitoring
indicate that the concentrations of lead exceed 0 005 mg/1 or concentrations of copper exceed
0.65 mg/1, the state may require the system to resume regular tap water monitoring pursuant to
§§ 141.86(d)(3) or (4) or take other appropriate remedial action(s) deemed necessary.  A state
may grant a waiver for some or all of the current monitoring requirements, and as a condition of
the waiver may require the system to perform specific activities (e.g., limited monitoring or public
education regarding faucet materials).

It is conservatively estimated that approximately 7,000 small water systems (or about 10 percent
of the approximately 71,100 small systems) will be able to obtain a monitoring waiver.12  Applying
these assumptions to the 1991 RIA monitoring cost models yields an annual national  monitoring
cost decrease of $504,800.

In addition to monitoring costs, this provision will also affect transaction costs. The April 1991
RIA did not account for state review of monitoring waiver requests.  A conservative estimate of
the average time for a system to prepare a transmittal letter to the state is approximately one hour.
Assuming that the average water system labor cost is $15 per hour, the cost to generate a letter
for the state to review is $15.  If approximately 7,000 systems apply for a monitoring waiver,
systems will incur a total transaction cost of $105,000. This corresponds to an annual cost  of
approximately $8,100; the annual cost is derived by dividing the total cost of $105,000 by the 13
years remaining in the 18-year LCR planning period. Consequently, the net effect of this
provision is a decrease in annual system cost of $496,700; this corresponds to the difference
between the monitoring cost savings of $504,800 and the transaction cost of $8,100.

Systems issued a waiver that can no longer certify that the system and all building connections are
free of materials containing lead or copper must resume the monitoring schedule specified in §
       To estimate the number of small systems expected to qualify for this provision, EPA reviewed sampling results
       submitted by large- and medium-size systems (those serving > 3,300 persons), for the initial two rounds of
       monitoring. These data indicated that over twenty percent of systems reported ninetieth percentile lead levels
       of 0 005 mg/1 or less for two consecutive six-month monitoring periods. EPA anticipates similar results for
       small-size systems. Of these small systems, EPA assumes approximately one-half will be able to certify that
       their system and all buildings connected to their system are free of lead-containing and copper-containing
       materials.

                                             12

-------
 141.86(d)(4).  Examples of such systems include those with new construction or repair work.
EPA does not believe many systems will be required to resume monitoring under
§ 141.86(d)(4) because the 1986 amendments enacted to the Safe Drinking Water Act prohibit
use of pipe, solder, or flux in a public water system that is not "lead free" and because of the wide
availability of lead-free materials.  Therefore, this potential cost increase is not estimated.

System Source or Treatment  Changes

There is a concern that the current LCR does not adequately address systems deemed to optimize
corrosion control that later make changes to their source or treatment (e.g., because of additional
drinking water regulations).  It is important that corrosion control treatment be maintained when
source or treatment changes occur to ensure that lead and/or copper is not released from the
interior surfaces of pipes and other plumbing or distribution system materials.

To address this concern, EPA is proposing to require such systems to notify the state within 60
days of such changes. The state, in turn, may require the system to resume monitoring in
accordance with §§ 141.86(d) and 141.86(d)(3).  This modification applies to all systems that are
monitoring at a reduced frequency, regardless of whether the system is on a monitoring waiver.

In order to calculate the cost  impact for this modification, EPA assumed the following percentage
of systems would modify either their source or treatment in each of the three-year intervals
remaining the LCR planning period.
                                        Exhibit 2
System Size Category
Small
Medium
Large
Percentage of Total
Systems for First 3- Year
Interval
5
10
25
Percentage of Total
Systems for Subsequent 3-
Year Intervals
5
5
5
The transaction burden is assumed to be one hour for the system to draft a letter notifying the
state of a change in either source or treatment.  Applying these assumptions to the 1991 RIA
costs models, systems are expected to incur an annual increase of $19,900.   Costs associated
with increased monitoring based on state discretion are not estimated because EPA believes the
large majority of systems will have already collected sufficient monitoring data.
                                           13

-------
 State Approval for Reduced Monitoring

 Systems maintaining optimal water quality parameters specified by the state under § 141.82(f) for
 two consecutive 6-month monitoring periods are no longer required to request state approval
 before assuming an annual monitoring schedule at the reduced number of sites specified in §
 141.86(c). Similarly, water systems are currently required by § 141.86(d)(4XHi) to request state
 permission to adjust their monitoring to once every three years after they have completed three
 years of annual monitoring that demonstrates maintenance of optimal  corrosion control.  EPA is
 also proposing to eliminate this requirement.

 EPA assumes approximately 12,000 systems in total (or approximately 920 systems in each of the
 13 years remaining in the LCR planning period) would no longer need to obtain state permission
 to assume a reduced monitoring schedule.  This number includes large systems other than "b3"
 systems, and small and medium systems that continue to exceed either action level after the
 installation of treatment. EPA also assumes all systems that install treatment will meet the water
 quality parameters specified by the state. The transaction burden reduction is assumed to be one
 hour for systems. Consequently, since system costs are assumed to be $15 per hour, annual
 system costs are  expected to decrease by $13,800.

 § 141.87

A ccelerated Reduced Water Quality Parameter Monitoring

EPA is also proposing to allow accelerated reduced water quality parameter monitoring at the tap
for systems meeting the criteria for accelerated  reduced lead and copper tap water monitoring.
This change applies only to large systems, because small and medium  systems with very low lead
and copper levels are not required to perform water quality parameter monitoring. The  April
 1991 RIA assumed that three years of annual monitoring would be required before a system could
assume a triennial monitoring schedule.

Large systems  that have installed corrosion control treatment and those "b2" systems that did not
continue monitoring beyond initial monitoring may be eligible for accelerated reduced water
quality parameter tap monitoring once they have completed the monitoring required by
§ 141.87(d). When the 1991 RIA monitoring cost models were adjusted to account for this
modification only a negligible cost reduction resulted because only a few systems are expected to
benefit from this change.

Biweekly Entry Point Monitoring for Water Quality Parameters

EPA is proposing to allow ground water systems that are required to conduct water quality
parameter monitoring to limit the entry point locations at which they must monitor. In particular,
biweekly entry point monitoring  would be required only  at entry point locations that are
representative of water quality conditions. As stated in § 141.87(c)(3) systems wishing to qualify
                                           14

-------
for this provision must provide states with an identification of selected entry points and
information sufficient to demonstrate that sites are representative of water quality and treatment
conditions throughout the distribution system.

EPA assumes that the number of entry points that must be sampled will be reduced by two-thirds
for large ground water systems and by one-half for medium ground water systems. No reduction
is assumed for small systems because of their limited number of entry points.  A monitoring cost
savings of $83,000 is expected for these systems. EPA also assumes that the system will require
one hour to request this monitoring reduction. The transaction burden for systems is calculated to
be $5,800 per year; this burden is offset by the monitoring savings of $83,000 and results with a
total annual cost reduction of $77,200.

S 141.88
Reduced Source Water Monitoring

The final regulation published in June 1991 permits a reduction in the source water monitoring
frequency by water systems that exceed the lead or copper action level and for which the state has
specified maximum permissible levels of lead and copper in source water.  The final rule does not,
however, permit water systems exceeding the lead or copper action level to reduce the frequency
of source water monitoring, if the state has not specified maximum permissible source water
levels.  In such cases, the state has effectively determined that source water concentrations of lead
and copper do not contribute significantly to concentrations at the tap.

EPA proposes to revise the final regulations to allow reduced monitoring frequency for systems
with low source water lead and copper levels (i.e., below EPA's source water treatment threshold
of 0.005 mg/1 for lead and 0.8 mg/1 for copper). To qualify for this reduced monitoring, such
systems would need to maintain low source water lead and copper levels for three consecutive
periods, if using an exclusively groundwater source, or three consecutive years, if using a surface
water or combined surface/groundwater source.

The source water and tap water monitoring cost models included in the April  1991 RIA were
developed independently to simplify model design. As a result, the source water model implicitly
assumed all systems conducting source water monitoring  would eventually be eligible for reduced
monitoring.  This implies that systems were assumed not to wait for states to designate maximum
permissible levels before reducing their monitoring. Therefore, no decrease in national costs is
estimated.

Resampling Triggers for Composited Samples

The proposed revision under  §  I41.88(a)(l)(iii) affects the resampling trigger level for lead and
copper in source water when  composite samples are taken.  Under the final regulation published
in June 1991, any system that exceeds an action level for lead or copper must collect samples
from source water entry points. These samples are used to determine the contribution of source
                                           15

-------
 water to lead and copper concentrations at the tap. Furthermore, under the final regulation, a
 water system may composite as many as five sampling sites.

 In the final regulation, the resampling trigger for composited lead and copper samples was based
 on the method detection level for lead and copper (i.e., 0.001 mg/1 for both lead and copper).  The
 Phase V Rule (57 FR 3 1776, July 17, 1992), modified these provisions to require the resample
 trigger for composited inorganic chemical samples to be no greater than or equal to one-fifth of
 their maximum contaminant level.

 EPA has not specified maximum contaminant levels for lead or copper,  however.  As described in
 more detail in the preamble accompanying this RIA addendum, EPA is proposing to base the
 composite resample triggers for lead and copper on source water treatment thresholds according
 to Lead and Copper Rule Guidance Manual Volume II:  Corrosion Control Treatment
 (Document No. EPA 81 l-B-92-002). EPA proposes:

 •     To set the lead resample trigger at 0.001  mg/1 (i.e., one-fifth of the 0.005 mg/1 treatment
       threshold), and

 •     To set the copper resample trigger at 0. 160 mg/1 (i.e., one-fifth of 0.8 mg/1 treatment
       threshold).

Therefore, relative to the final regulation published in June 1991, EPA proposes not to change the
lead resampling trigger (i.e., it remains 0.001 mg/1), and to revise the copper resample trigger
upward, from the method detection level of 0.001 mg/1 (or 0.020 mg/1 when atomic absorption
direct aspiration is used) to 0.160 mg/1 (i.e., one-fifth of the treatment threshold).

EPA is modifying § 141.89 to remove the requirements that laboratories achieve the copper
method detection level of 0.001 mg/1 or 0.020 mg/1 because the laboratories will be sufficiently
tested on their capabilities, as required by § 141.90(l)(ii).  Section 141 .90(1 )(ii) requires
laboratories to achieve a quantitative acceptable limit of + 10 percent of the actual amount of a
performance sample when the actual amount is greater than or equal to 0.050 mg/1, which is
below the proposed copper resampling trigger of 0. 160 mg/1

The April 1991 RIA assumed that all systems with multiple entry points which were required to
perform source water monitoring would composite samples. Furthermore, the 1991 RIA did not
evaluate how many systems would be required to resample because they exceed the lead or
copper trigger level. Note, however, that because the trigger levels will not decrease due to the
proposed revision, EPA does not anticipate an increase in costs.  In fact, costs may actually
decrease because the resampling trigger for copper will increase from 0.001 mg/1 to 0.160 mg/1.

S 141.90
EPA is proposing a number of changes to system reporting requirements codified at § 141.90.
With the exception of modifications to the reporting requirements associated with lead service line
                                           16

-------
replacement, these modifications address lead and copper tap water and water quality parameter
monitoring.

First-Draw Certifications

EPA is proposing to eliminate the reporting requirement, currently codified at § 141.90(a)(l)(ii)
that systems submit a certification that each lead or copper tap sample collected by the system
meets the first-draw criteria at § 141.86(b). EPA also is proposing to eliminate the reporting
requirement at § 141.90(a)(l)(iii) that systems certify that tap  samples collected by residents
occurred after the water system informed the residents of the proper sampling procedures.

EPA assumes that this change will eliminate approximately 535,200 certifications in total (an
average of 41,170 per year) for the remaining 13 years in the planning period and that the burden
incurred by a system to prepare the certification is 10 minutes. Assuming the system labor cost is
approximately $15 per hour, the cost associated with each certification is $2.50. EPA estimates
that eliminating these requirements will save water systems approximately $102,900 annually.

Sample Invalidation Requests

EPA is proposing to add a new reporting requirement at § 141.90(a)(l)(ii) that systems
requesting sample invalidation in accordance with § 141.86(f) submit supporting documentation
demonstrating the need to have the sample invalidated.  The estimated costs associated with this
reporting requirement are discussed above under § 141.86 in this chapter.

Sample Site Justification

As described above under § 141.86, EPA is proposing that water systems no longer be required
to submit a letter to the state demonstrating why sufficient Tier 1 sites cannot be located or why
50 percent of the sites sampled are not  served by lead service lines. EPA is therefore proposing to
eliminate the corresponding reporting requirements currently codified at  §§ 141 90(a)(2) - (4).
The estimated cost savings associated with these reporting requirements  are discussed as a part §
141.86 in this chapter.

NTNCWS Sampling Sites

EPA is proposing to codify at § 141.90(a)(2) the reporting requirements for requesting state
approval of alternative sampling sites, pursuant to § 141.86(bX5), by NTNCWSs with too few
taps from which first-draw samples can be drawn.  The estimated system costs associated with
this reporting requirement are discussed as a part of § 141.86 in this chapter.
                                           17

-------
System Source or Treatment Changes

EPA is proposing to codify at § 141.90(a)(3) the reporting requirements for reporting changes in
system source or treatment, pursuant to § 141.86(d)(4)(vii), by systems subject to reduced
monitoring. The estimated system costs associated with this reporting requirement are discussed
as a part of § 141.86 in this chapter.

Monitoring Waivers for "All Plastic" Systems

EPA is proposing to codify at § 141.90(a)(4) the reporting requirements for small "all plastic"
systems to request monitoring waivers from the state pursuant to § 141.86(g), The estimated
system costs associated with this reporting requirement are discussed as a part of § 141.86 in this
chapter.

State Approval for Reduced Monitoring

As discussed above, EPA is proposing to eliminate the requirements in § 141.86(d)(4) that water
systems request state approval before reducing the number and frequency of sampling. EPA also
is proposing to eliminate the corresponding reporting requirement currently codified at §
141.90(a)(5).  The estimated cost savings associated with eliminating this reporting requirement
are discussed as a part of § 141.86 of this chapter.

Biweekly Entry Point Monitoring for Water Quality Parameters

EPA is proposing to codify at § 141.81.90(a)(5) the new requirement for ground water systems to
provide the state with information identifying the selected entry points and demonstrating that
these entry points are representative of water quality and treatment conditions throughout the
systems in conjunction with requesting state approval to limit biweekly entry point water quality
parameter monitoring to representative locations pursuant to §  141.87(c)(3). The estimated costs
associated with this reporting requirement are discussed as a part of § 141.87 in this chapter.

Rebutting Presumption of Lead Service Line Control

As discussed in chapter 4.4.2, EPA is proposing to eliminate the presumption that the water
system controls the entire length of the lead  service line  This presumption is no longer required
since EPA is proposing a revised definition of "control" of lead service lines.  EPA also is
proposing to modify § 141.90(e)(4) to remove the reporting requirement associated with
rebutting the presumption of lead service line control since the revised definition is self-
implementing and it will no longer be necessary to demonstrate to the state why the water system
cannot replace the entire length of the line. The cost savings associated with removing this
reporting requirement are discussed as a part of chapter 4.4.2.
                                            18

-------
4.2    Costs of Coping With Source Water Contamination

None of the proposed changes are expected to affect the number of systems that the April 1991
RIA estimated will be required to install treatment equipment to remove lead from source water.
Consequently, source water treatment costs are not expected to change because of the proposed
revisions.

4.3    Costs of Coping With Lead Leached From Solder

The proposed changes will not have a direct impact on the costs of coping with lead leached from
solder. Public education tasks may be required due to a lead exceedance resulting from lead
solder; however, such costs are described in Chapter 4.4.1.

4.4    Costs of Coping With Lead Leached From Lead Pipes

These costs are discussed in sub-parts 4.4.1 and 4.4.2 as follows.

4.4.1   Cost of Corrosion Control and Public Education to Cope With Leaching from Pipes

Cost of Corrosion Control

The proposed revisions are not expected to affect the cost of installing corrosion control
treatment for an individual water system, but some systems that no longer meet the §
141.81(b)(3) criteria may now be required to begin or complete corrosion control treatment
Conversely, some systems may be able to  avoid the cost associated with corrosion control
treatment installation because of the new sample invalidation provision that would allow
erroneous sample data to be removed from a system's ninetieth percentile calculation.

Failure to Maintain "bB" Criteria

The proposed revisions would require states to review copper tap sample data for initial
monitoring to determine  if a system previously designated as a "b3" system still qualifies for this
designation. States may  direct systems whose initial monitoring results meeting the lead criteria,
but which exceed the copper action level,  to resample for both copper and lead to "re-check" for
low levels of corrosion before requiring a  system to initiate corrosion control treatment.  Small
and medium systems meeting the criteria in § 141.81(b)(l) would not be required to install
corrosion control treatment.

The April 1991 RIA did not account for systems initially meeting, but subsequently failing to
meet, the "b3" criteria. However, EPA expects that few systems will be affected by this
provision.  Furthermore,  EPA believes that the conservative assumption in the 1991 RIA that
                                           19

-------
approximately 50 percent of systems would install corrosion control treatment to address high
lead levels sufficiently encompasses any additional systems that may need to perform corrosion
control treatment due to the proposed revision.

Public Education

The proposed changes to the public education requirements alter the existing rule in four ways.
In particular, the proposed changes provide alternative public education language that can be used
by non-transient, non-community water systems, allow some community water systems, allow
notices to be mailed separately from water bills, affect the tinning of initial public education for
some water systems, modify the public education requirements for community water systems
serving 500 or fewer persons, and change the schedule for reporting completion of public
education tasks.  All of these changes would become effective 18 months after promulgation of
the revised regulation.

Public Education Language

Section 141.85(a) contains specific language that systems must use in all the written materials
they distribute as part of their lead education programs. The public education efforts of systems
that have exceeded the lead action level have done much to inform the public about the issue of
lead in drinking water. Some EPA regions and states, however, are concerned  that the language
specified in this section of the LCR may not be appropriate for non-transient, non-community
water systems and some small community water systems such  as hospitals and prisons.

EPA wants to make the language in public education materials distributed by non-transient, non-
community water systems more relevant to the people they serve.  For example, EPA proposes
replacing phrases such as "some homes in the community" with "some drinking water samples
[taken from this facility]" since an non-transient, non-community water system typically does not
serve water to homes. The Agency also proposes to delete references to lead testing; non-
transient, non-community water system customers are unlikely to have water tested for lead
because they consume the water for only short periods of time and have little or no control over
the distribution system.  For similar reasons, the Agency wants to delete references to having a
plumber check pipes for solder and an electrician test pipes for improper grounding.  Other similar
changes are also being proposed, and their net effect is to make the public education material
more relevant to the customers of non-transient, non-community water systems.

EPA also believes that the current public information language is inappropriate for some
community water systems, such as hospitals and prisons. Patients and inmates, for example, are
not able to install home treatment devices or have their water tested, as the current community
water system public education text encourages.  Similarly, references to "your home" and "your
family's health" are inappropriate. Consequently, EPA proposes to allow community water
systems to request state approval to use the public education language prepared for non-transient,
non-community water systems when the community water systems' service population cannot
make plumbing improvements or install point-of-use treatment devices and the  system provides
                                           20

-------
water as part of the covered services and does not directly bill for water consumption.  The
Agency believes that this modification is appropriate for certain institutions and should be allowed
where the state believes it better protects the public health.

This change is expected to have a minimal cost impact because systems are not required to re-
print existing public education materials. In addition, systems that choose to modify their public
education language will be able to exhaust their existing supply of materials first.

Separate Mailing of Notices

Section 141.85(c) requires water systems that exceed the lead action level to undertake public
education tasks within 60 days of the exceedance. Under § 141.85(cX2XO, community water
systems must include the information specified in § 141.85(a) with the water bill.  In establishing
these requirements, EPA wanted to provide customers with timely notification that their drinking
water system had exceeded the lead action level and with information about health risks, sources
of exposure, measures the system is required to undertake, and ways the customer could reduce
exposure to lead in drinking water.  The Agency believes that including the notice with the water
bill would make customers more inclined to  read the notice, and less likely to throw it away as
"junk mail." These requirements have, however, caused some unintended problems for many
water systems.

Many systems do not bill often enough to meet the 60-day requirement; 90 days is more the norm.
Also, many systems use postcards or computer-generated self-mailers to bill  customers. These
bills do not allow the enclosure of additional materials, nor do they have sufficient space to
include the information at § 141.85(a) on the bill itself.

The Agency believes it would be inappropriate to force these systems  to change their billing
formats to comply with the LCR's public education requirements.  Therefore, EPA is proposing in
§ 141.85(c)(2)(i)(A) to allow systems to mail public education materials on the same schedule as
the system's billing cycle, provided the mailings occur within six months of the lead exceedance.
As noted in § 141.85(c)(2)(i)(B), EPA will give systems the option of distributing public
education materials to billing units through separate mailings if they do not bill within six months
of the exceedance and/or, they cannot insert information in the bill, provided the mailing is
completed within the required time frame and achieves equal coverage.  Systems that use an
alternative delivery method must include an alert with the public education material to minimize
the risk that the material will be discarded as "junk mail."

It is not currently possible to estimate how many systems will be able to benefit from this
proposed change, and consequently, it is not possible to estimate precisely the impact on national
costs.
                                           21

-------
 Systems Serving 500 or Fewer People

 Under § 141.85(c)(2), community water systems must submit notices to the major daily and
 weekly newspapers circulated in their community, in addition to mailing notices to billing units.
 These systems must provide information on lead to facilities and organizations frequented by
 children and pregnant women, and they must submit public service announcements to radio and
 television stations with the largest audience shares in the community served by the water system

 For some systems serving 500 or fewer people, particularly those that provide water only to small
 numbers of people in an urban or suburban area, these requirements have engendered a number of
 unintended consequences.  For example, small systems that serve only a small portion of a
 metropolitan area have been deluged with calls from individuals served by other water systems.
 These small systems often are ill-equipped to handle a large-scale public response to a newspaper
 notice or public service announcement.  Similarly, the requirement to distribute materials to
 locations visited frequently by children and pregnant women imposes a significant burden on small
 systems, since it may involve a large number of locations if the system is located in or near an
 urban or suburban area.

 Because EPA believes it is inappropriate to impose such burdens on systems serving so few
 people, the Agency is proposing in a new § 141.85(c)(8) to limit or omit some of the required
 public education tasks for systems serving  500 or fewer people:

       •     Instead of submitting information to newspapers and broadcast outlets, these
             systems can mail or hand-deliver to all consumers of the water they provide (e.g.,
             tenants of multi-family dwellings whose water is included with their rent) the same
             lead education materials that the system must mail to its billing units.

       •     These systems must furnish informational pamphlets to locations in their service
             areas and only to locations outside their service areas that are regularly visited by
             their consumers.

 Systems performing public education in accordance with the provisions of § 141.85(c)(8) must
repeat the tasks every 12 months for as long as the system continues to exceed the lead action
level.

Although individual system public education costs were included in the April 1991 RIA, the
components of these costs were not itemized.  After evaluating this change, EPA believes there is
probably an overall decrease in cost because approximately 19,000 systems serving fewer than
 500 people will benefit from this provision. The 19,000 system estimate represents the number of
community water systems serving fewer than 500 people that the 1991 RIA models  assumed
would be required to perform public education because of a lead action level exceedance. The
burden associated with § 141.85(c)(2)(ii) and  § 141.85(c)(2)(iv) is assumed to save  each system
 approximately two burden hours. Assuming an hourly labor cost of $15, the per-system savings is
 $30 and the national cost savings is estimated  to be $570,000 annually.
                                           22

-------
Demonstrating Compliance with Public Education Requirements

This change would require each system subject to the public education requirements to submit to
the state a letter demonstrating compliance with those requirements. The letter would be due
within 10 days after the end of the period in which the system was required to perform public
education.  A* is currently required, this letter must include a list of all newspapers, radio and
television stations, and facilities/organizations to which the system delivered public education
materials during the period covered by the letter.  Currently, systems do not have to submit a
letter until December 31 of each year.  EPA wants to change this requirement because the Agency
and the states are not provided information in a timely manner to ensure compliance.
EPA assumes that public water systems will require 10 minutes to prepare the letter/report
demonstrating compliance with public education requirements. EPA estimated the number of
additional reports that would be required by determining the number of systems serving more than
500 persons that would remain above the lead action level for more than one reporting period
and, therefore, be required to continue public education delivery.

Community water systems serving more than 500  persons are required to perform the public
education tasks associated with public service announcements (PSAs) every six months as long as
they continue to exceed the lead action level.  Under this proposed revision, these systems would
now be required to notify the state twice a year of the completion of these tasks. EPA estimated
the period of time these systems would remain above the lead action level by identifying LCR
activities (e.g., the installation of corrosion control or source water treatment) that should result
in systems no longer exceeding the lead action level.  The total additional annual cost for this
modification is estimated to be approximately $34,200; this burden estimate corresponds to an
additional 13,670 reports.13

Community water systems serving fewer than 500 persons would not be affected by the change in
the reporting frequency for completion of public education tasks because the proposed revision,
discussed above, that these systems no longer be required to submit public service
announcements.  Although the change in reporting frequency would not affect them, these
systems would continue to be required to conduct other public education tasks on an annual basis
for as long as the lead action level is exceeded and to report completion of these tasks to the state
annually They therefore could be affected by the change in timing of such reports.

4.4.2  Cost of Lead Service Line Replacement

Public water systems that exceed the lead action level at the tap after installation of optimal
corrosion control treatment and/or source water treatment are required to replace the lead service
lines they "control," unless they can demonstrate that all samples taken from the line have lead
       The burden for reporting completion of public education activities was not quantified in the 1991ICR.

                                           23

-------
levels of 0.015 mg/1 or less.14  The proposed modifications provide a narrower definition of
"control" than EPA promulgated in June 1991.15 EPA is now proposing that "control" equate to
ownership plus the authority to replace. This narrower "control" definition may result in fewer
total lines (or a smaller portion of these lines) being replaced than would have occurred under the
1991 definition. However, no decrease in public water system burden is estimated for this LCR
modification because the 1991 RIA also used a narrow definition of "control" that is essentially
equivalent to the currently proposed definition.

Under the current rule, systems are presumed to control the entire length of the service line unless
they can demonstrate otherwise to the State.  This requires systems trying to rebut the
presumption to submit documentation to the state to support their claim that a line (or portion of
a line) is beyond their control  and, therefore, the system is  not required to replace it. The Agency
believes that the proposed definition of control is "self-implementing" and is therefore removing
the rebuttable presumption since there should no longer be issue about control. A reduction in
burden is expected for systems due to eliminating the time associated with preparing the
documentation to rebut the presumption that the system controls the entire line .

As a result of this rule modification, EPA assumes that 1,500 systems in total will no longer rebut
lead service line control and that each system would have required eight hours to prepare
documentation supporting its  rebuttal.. This will result in an  annual savings of $13,800 for
systems.

4.5    State Implementation Costs

There are three areas in which state implementation costs are expected to be affected as a result of
the proposed modifications. These areas are discussed separately below.  Note that for purposes
of calculating the change in state costs, EPA used an hourly labor rate (including benefits) of
S31.09.16

Data Entry and Reporting

EPA is proposing to eliminate some of the state milestone reporting requirements contained in 40
CFR § 142.15 because they are redundant or unnecessary  EPA,  however, is also proposing  to
require that additional monitoring and milestone data for some of the current requirements be
   "   A system may avoid replacing a lead service line by demonstrating that the lead concentration in any sample
       collected from that line is 0.015 mg/1 or less. The system may count lines with such levels toward the annual
       seven percent replacement requirements.  None of the proposed rule modifications change the lead service line
       monitoring provisions.

   15   This definition was challenged by the American Water Works Association and in December 1994. In
       December 1991, the court remanded and vacated the definition of control as it applies to portions of line
       beyond a system's ownership. (AWWA v EPA. 40 F 3d 1266 (D.C. Cir. 1994))

   16   EPA revised its estimate of the hourly labor rate contained in the  1991 ICR for state staff from $26.04 to
       $31.09 per hour.  The upward adjustment in the labor rate was made to account for inflation.

                                             24

-------
reported.  The additional data are intended to track the dates of key state decisions and water
system compliance. EPA anticipates the costs for reporting additional data will be offset by
eliminating other requirements for which a substantial volume of data would have been reported.
Specifically, EPA estimates that the annual state data entry and reporting costs will decrease by
approximately $81,300.17 This cost savings assumes states require 0.5 hours to process each
record submitted by systems.

Review System Submittal and Notify Systems Of Requirements

A number of modifications will affect the state costs associated with reviewing systems submittal
and notifying systems of LCR requirements.  In particular:

•     Non-first-draw samples. States will be required to review non-transient, non-community
       water system requests to use non-first draw samples.  It is estimated that this activity will
       require approximately 10 minutes.  Assuming that over the remaining 13 years of the LCR
       planning period, approximately 500 systems will make such a request, annual state costs
       are expected to increase by approximately $200

•     Reduced tap water monitoring sites.  EPA assumes that over the remaining 13 years in the
       LCR planning period, states wilt designate sampling sites for about 9,200 systems and that
       the burden to designate sites for each system is about 0.5 hours.  Thus, the annual costs to
       states is expected to increase by approximately $11,000.  The actual burden states may
       realize may be significantly lower because some states might issue broad guidance
       regarding reduced sampling sites.

•     Monitoring waivers for "all plastic" systems.  The burden incurred by states to review
       public water system requests for monitoring waivers is assumed to be one hour.  It is
       assumed that states will require 0.5 hours to review waiver renewal applications.
       Therefore, if approximately 7,000 systems nationwide seek monitoring waivers over the
       remaining LCR planning period of 13 years, annual state review costs are estimated to
       increase by approximately $8,400.  These costs will be offset by the cost savings
       associated with having to review less monitoring data.  In particular, EPA estimates that
       states will no longer need to review approximately 21,000 rounds of monitoring data; each
       of these rounds of monitoring data are assumed to require one hour of state time.
       Consequently, the cost savings associated with needing to review less monitoring data is
       $50,200 per year. Therefore, the net effect of this modification is an annual state cost
       savings of $41,800.

•     Biweekly entry point monitoring for ground water systems. EPA assumes states will
       require 0.5 hours to review requests to limit biweekly water quality parameter monitoring
       from approximately 387 ground water systems each year.  It is assumed that large and
       For a description of the methodology EPA used to evaluate state data entry and reporting costs, see section
       6a(ii)ofthelCR.

                                           25

-------
 medium ground water systems will be able to reduce the number of monitoring locations
 by two-thirds and one-half, respectively. Small ground water systems are assumed to
 already be monitoring from a small number of monitoring sites due to their size and
 configuration and will be unable to further reduce entry point monitoring. Consequently,
 state costs are estimated to increase by $6,000 per year.

 Changes in treatment or source.  States will be required to review documentation from
 systems on a reduced monitoring schedule that change either their treatment or source.
 For each of these submissions, the state will need to determine if the changes warrant
 having the system return to a standard monitoring schedule.  The state is assumed to need
 10 minutes to make this determination and notify the system if a change in monitoring
 frequency is required.  The number of systems that EPA estimates will change either their
 source or treatment is expressed as a percentage for each major system size category. See
 page 13 of this document for an explanation of how the number of affected systems was
 derived. This translates to an annualized average cost of $6,900; this cost was derived by
 dividing the total cost of approximately $89,700 by the 13  years remaining in the LCR
 planning period.

 Accelerated reduced monitoring.  The 1991 RIA models predict that systems will be able
 to eliminate approximately 7,077 lead and copper samples  per year because of being able
 to proceed to triennial monitoring before completing three years of annual monitoring if
 they meet certain criteria. As a result, states will have less monitoring data to review.
 EPA estimates that the amount of monitoring data  states will need to review will decline
 by approximately 862 rounds.  Assuming the states require one hour to review each round
 of data, this translates into an annual state cost savings of $26,800.

 Elimination of certification for first-draw or customer-collected samples. EPA is
 proposing that systems would no longer be required to submit a certification that each lead
or copper tap sample meets the first-draw criteria at § 141.86(b) or that tap samples
 collected by residents occurred after the water system informed the residents of the proper
 sampling procedures. As a result, states will no longer need to process and review
certifications that formerly accompanied monitoring data submissions. EPA estimates that
annually 41,170 certifications will no longer be required. If state review of each
certification is assumed to require 10 minutes, states should realize an annual $213,300
cost  savings.

 State approval for reduced monitoring schedule when water quality parameters are met.
EPA assumes approximately 12,000 systems in total (or approximately 920 systems in
each of the 13 years remaining in the LCR  planning period) would no longer need to
obtain state permission to assume a reduced monitoring schedule once they meet optimal
water quality parameter ranges. The state is assumed to avoid one hour of review time
per system. Therefore,  state costs are expected to  decrease by $28,700 per year.
                                    26

-------
•     Lead service line control. Under the current rule, systems can submit documentation to
       the state to support their claim that a lead service line (or portion of a line) is beyond their
       control and, therefore, is not required to be replaced under the requirements contained in §
       141.84,  EPA is proposing a narrower definition of "control" where "control" equates to
       ownership plus the authority to replace. As a result of this narrower definition, EPA
       expects  that systems will no longer need to rebut control presumptions and that 1,500
       systems in total (or approximately  115 systems in each of the 13  years remaining in the
       LCR planning period) will avoid having to submit rebuttals.  Therefore, if states are
       assumed to require four hours to review a system's documentation used to rebut control
       presumptions, this modification will result in an annual savings of $14,300 for states.

Reporting Completion of Public Education

One of the proposed changes would affect the deadline for reporting to the state compliance with
public education requirements.  For community water systems serving 500 or fewer persons that
remain above the lead action level for two monitoring periods in a calendar year, states would be
required to review an additional letter/report prepared by the system. Currently, states review
such reports only once at the end of the calendar year.

EPA assumes that the state will require 10 minutes to review the letter/report submitted by the
system demonstrating compliance with public education requirements. Consequently, the total
additional annual cost for this task is estimated to be $70,800; this cost estimate corresponds to
an additional 13,670 reports.

5.0    The Benefits of Reducing Lead in Drinking Water

The minor changes proposed in this rulemaking are intended to improve  implementation of the
LCR. The effect of these changes will be to eliminate unnecessary requirements and to promote
consistent national implementation of the rule.  EPA does not intend these changes to modify the
level of health protection extended by the final rule and no impact on health protection is
anticipated.  There are no known decreases in health benefits associated with these proposed
changes.
                                           27

-------
6.0    Regulatory Flexibility Analysts and Paperwork Reduction Analysis

Regulatory Flexibility Analysis

As indicated in the earlier sections of this RIA addendum, EPA is proposing revisions to the final
LCR designed to eliminate unnecessary requirements, streamline and reduce burden, and promote
consisten. and flexible implementation of the rule nationwide. Such objectives are consistent with
the Regulatory Flexibility Act (Act), which requires that all executive agencies explicitly consider
small entities in their regulatory design and implementation processes.  For regulatory flexibility
analysis purposes, systems serving 50,000 or fewer people are considered to meet the Act's
definition of a small entity.  See Exhibit 3 for a distribution of the number of public water systems
by size category.
Exhibit 3
Public Water System Distribution
Population Served
> 50,000
3,301 to 50,000
< 3,300
Number of Systems"
793
6,768
71,142
The quantifiable cost impacts described in this RIA addendum primarily affect systems serving
50,000 or fewer people. In fact, EPA estimates that most of  the impacts associated with these
revisions will be borne by systems serving 50,000 or fewer people.  Accordingly, costs for these
systems are expected to decrease by approximately $1.57 million per year (see Exhibit 4),  Cost
impacts are summarized in more detail below.
       Represents the distribution of systems included in the 19911CR as subject to the LCR.

                                            28

-------
                                        Exhibit 4
                Cost Impact To Systems Serving Fewer Than 50,000 People
                      (Annual Cost Estimates in Millions of Dollars)
Proposed Change
Monitoring Waivers for "All Plastic" Systems
Request for Reduced Monitoring
Biweekly Entry Point Monitoring
Rebuttable Presumption Regarding Control of Lead Service
Lines
Treatment or Source Water Changes
Certifications of First-draw and Customer-collected Samples
Accelerated Reduced Lead and Copper Monitoring
Public Education
Reporting Completion of Public Education
Total
Cost Impact (SM)19
- 0.497
- 0.013
- 0.074
- 0.014
+ 0.020
- 0.103
- 0.354
- 0.570
+ 0.034
-1.57
Monitorine-Related
       Assuming systems that serve fewer than 3,300 people meet the lead and copper action
       levels at the tap are able to benefit from the modification permitting accelerated reduced
       lead and copper tap water monitoring, national monitoring costs will decrease by
       $353,800 annually.

       The availability of monitoring waivers for "all plastic" systems is a provision restricted to
       systems serving 3,300 or fewer. Consequently, all of the $496,700 annual cost savings
       will accrue to systems serving less than 50,000 persons.

       As a result of the modification that would allow biweekly entry point monitoring for water
       quality parameters at representative sites, systems are expected to save $73,800 in annual
       monitoring costs.
       Total may not add, due to independent rounding.
                                           29

-------
 •     Annual system costs are expected to decrease by $ 13,400 because of a modification that
       would eliminate the requirement that systems first obtain state permission before assuming
       a reduced monitoring schedule.

 •     EPA is proposing to require that systems on a reduced monitoring schedule notify the
       state if there are changes in their treatment or source. Consequently, system costs are
       expected to increase by $19,500 annually.

 •     As a result of a modification EPA is proposing, systems would no longer be required to
       submit a certification that each lead or copper tap sample meets the first-draw criteria at §
       141.86(b) or that tap samples collected by residents occurred after the water system
       informed the residents of the proper sampling procedures. These modifications are
       expected to decrease annual system costs by $102,900.

Lead Service Line Replacement

•     Because of a modification to §§ 141.84(d) and 14l.90(e)(4) that would modify the
       definition of "control" for lead service lines and eliminate the associated reporting
       requirements, respectively, systems are expected to realize an annual cost savings of
       $13,800.

Public Education

Many of the proposed changes are intended to  address the unique circumstances of systems
serving 50,000 or fewer persons. Specifically, the public education modifications provide
alternative language tailored to the unique circumstances of non-transient, non-community water
systems, almost all of which serve 50,000 or fewer persons. In addition, modification to allow
alternate public education delivery methods was intended to address systems that do not bill
frequently or that use billing methods that preclude the insertion of public education materials
Many systems serving fewer than 3,300 persons are expected to benefit from this change because
they tend to bill their customers quarterly or even less frequently.

The public education requirements have been reduced explicitly for systems serving 500 or fewer
people because many such systems are ill-equipped to respond to the level of public response
those requirements generate. In particular, the tasks contained in §§  141,85(cX2)(ii) and
141.85(c)(2)(iv) that  require submission of lead public education information to newspapers and
radio and television stations would be omitted.  Systems also would be allowed to limit the
number of locations to which they must furnish informational materials.  As a result, the 19,000
systems serving fewer than 500 people will incur a $570,000 per year cost-savings.

Reporting Completion of Public Education
                                           30

-------
The proposed change associated with reporting completion of public education will increase water
system costs by approximately $34,200 annually.  This cost increase is conservatively assumed to
be borne by community water systems serving fewer than 50,000 persons.

Paperwork Reduction Act Analysis

One of the purposes of the Paperwork Reduction Act is to minimize the costs of federally
imposed paperwork requirements for individuals,  small businesses, and state and local
governments, as well as minimize the federal government's costs associated with collecting and
maintaining information submitted by such entities.

The Paper Reduction Act requires preparation of an ICR for any activity that involves collecting
information from 10 or more non-federal respondents. As noted in Chapter 1.0, a separate ICR
will address the proposed minor revisions to the LCR. The results of the ICR are discussed in
Chapter 4 of this RIA addendum.  As described in detail in the ICR, many of the proposed
changes to § 142.15 are intended to eliminate duplicative or unnecessary reporting of information
already provided to the Safe Drinking Water Information System.20 As shown in Chapter 4, the
net quantifiable decrease in costs associated with these proposed revisions is minuscule when
compared to the cost of the entire LCR.

Water system cost savings are expected to result from revisions associated with accelerated
reduced lead and copper and water quality parameter tap monitoring, assumption of a reduced
monitoring schedule without first having to obtain state permission, monitoring waivers for "all
plastic" systems, reduction in biweekly entry point water quality parameter monitoring, reduced
source water monitoring for some additional systems, less frequent resampling of composited
source water samples, elimination of certifications pertaining to first-draw and customer-collected
samples, elimination of select public education requirements for small systems, reduction in
system reporting requirements, and elimination of rebuttable presumption regarding  ownership of
lead service  lines. A portion of these cost savings, however, will be offset by transaction costs
incurred in preparing letters associated with monitoring requirements for state review, increased
public education compliance reporting, and reporting changes to treatment or source water.
       Effective August 15, 1995, the Safe Drinking Water Information System became the data base of record for the
       national drinking water program and replaced the Federal Reporting Data System.

                                           31

-------
States are also expected to realize a net reduction in costs associated with the proposed
modifications to the LCR  In particular, state costs are expected to decline for provisions
associated with eliminating the requirement that systems obtain state  permission to assume a
reduced monitoring schedule, accelerated reduced monitoring, monitoring waivers for "all plastic"
systems, reduction in water quality entry point monitoring, elimination of rebuttable presumption
regarding ownership of lead service lines, elimination of certifications pertaining to first-draw and
customer-collected samples, changes in federal reporting requirements, and reduced reporting
resulting from less frequent system monitoring requirements  Although overall state costs will
decrease, states will incur transaction costs associated with reviewing system requests, making
monitoring-related determinations, and more frequent assessments of public education
compliance.
                                            32

-------