UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D C. 20460

EPA-SAB-EHC-89-006

October 6, 1988

Honorable Lee M. Thomas
Administrator

U.S. Environmental Protection Agency
401 M street, s.W,

Washington, D.C, 20460

Subjects Science Advisory Board*s Review of the issues relating
to the regulation of LEAD in drinking water.

Dear Mr. Thomas:

The Drinking Water Subcommittee of the Science Advifeory
Board*s Environmental Health Committee has completed its review
of t^e issues pertaining to lead. These issues relate to the
proposed regulations from the Office of Drinking Water at its
meeting in Cincinnati, Ohio, June 2-3, 1988. The Subcommittee
was requested to review the procedure for determining what an
adequate tap sample is for determining the drinking water
concentration of lead. Although not part of the charge, the
Subcommittee also commented on treatment technology.

The major recommendations of the Subcommittee include:
basing compliance sampling upon random sampling at customer taps
before requiring that the utility demonstrate optimal corrosion
treatment; providing better justification for the definition of
the worst case; providing better documentation of the relevance
of the two-stage sampling plan; and allowing the utility to
implement appropriate controls, when controls of some type are
' found to be needed.

We appreciate the opportunity to conduct this particular
scientific review. We request that the Agency formally respond
to the scientific advice provided herein.

Chairman, Executive Committee

Richard A. Griesemer
Chairman


-------


SUBJECT; SCIENCE ADVISORY BOARD'S REVIEW OF ISSUES RELATING TO
TOE REGULATION OF LEAD IN DRINKING WATER

SCIENCE ADVISORY BOARD COMMITTEE! DRINKING WATER SUBCOMMITTEE OF
THE ENVIRONMENTAL HEALTH COMMITTEE

DATE OF REVIEW; JUNE 2-3, 1988

PLACE OF REVIEW; EPA LABORATORY, CINCINNATI, OHIO

The Subcommittee was requested to review the procedure for
deteriming what an adequate tap sample is for determining
drinking water concentrations of lead. Although not part of the
charge, the Subcomittee also commented on treatment technology.
In addition, the Subcommittee takes issue with the MCLG of zero
as possibly being too conservative imd with the low MCLs; e.g. 5
ug/L in some cases. These levels may or may not be realistic
considering the existing information on health effects,
especially that relating to the central nervous system.

This review involved part of the proposed regulation; Maximum
Contaminant Isxsl SOllff lad ffafcicnal Primary Drinking Water

Regulations £ez im<& aM Swper
general gwuwatg

The MCLG for lead may be overly conservative. For example, the
document does not present sufficient justification for concluding
that there is no threshold for the inhibition of delta-
aainolevulinic acid anhydrase especially since it gave four
references, all with thresholds.

The document itself is persuasive in indicating what can be
done in removing sources of lead contributing to that present in
* the water, such as the removal of lead pipe and lead connectors
from distribution systems, the ban on lead solder, and the
control of corrosivity. While these efforts are not simple, in
many cases they are feasible and should result in a significant
decrease in lead exposure via the drinking water.

Much of the documentation for indicating how systems might
reach compliance and how stabilization of the water and other
treatment technologies could be achieved are not targeted at the
worst cast. They are based on the old level of 50 ug/L and/or
the possibility that the new level might be 20 ug/L, as opposed
to 10 ug/Ii (i.e., the 10 ppb of the proposed standard). Indeed,
the data presented would suggest that the action levels of 10 and
20 ug/dL will be very difficult for the water suppliers to meet.
The Subcommittee recommends that EPA make a stronger and clearer
justification for its selection.

The Subcommittee is concerned that the water quality
requirement does not consider lead levels in the distribution
system (at the tap) before requiring corrosion treatment. We

1


-------
recommend that systems that are in compliance with the lead "no-
action level#" not be required to meet arbitrary water quality
requirements. Obviously, if the lead level requirements were not
met, utilities would be required to demonstrate optimized control
treatment.

Sampling Plan

The subcommittee found that the basic purpose of the
sampling for lead levels is not explicitly stated. There are no
goals, criteria or focus given for this strategy. It is not
clear what it represents. There is apparently more than one
purpose to the sampling, in which case different methods
appropriate for one purpose may be inappropriate for another.

This Subcommittee recommends that the basic purpose of this
sampling strategy be made quite clearly. It is difficult to
evaluate the effectiveness of the sampling strategy if it does
not represent the exposure of the general population or i£ may
not truly represent the high risk population.

* EPA has not demonstrated that the worst case houses give an
appropriate.indication that corrosion treatment is being
optimized. A portion of high risk houses should be sampled to
determine the need for public education, but corrosion control
compliance should be based on random sampling. Random sampling
gives a more reliable indication of the system-wide impact of
corrosion control programs. Thus the Subcommittee recommends
that random sampling be used as the basis of determining the
action needed.

The subcommittee recommends tap sampling for compliance with
lead regulations <10 ug/L average and 20 ug/L 95% of time) be
based upon random sampling at customer taps before requiring
utility demonstration of corrosion treatment optimisation. The
* sampling should be allowed where high lead levels are found.

Thus the Subcommittee concludes that the form of the regulation
is good in using an average and a 95% exceedance level.

As noted above, the design of a sampling plan depends on the
purpose for which sampling is being done. If the purpose of the
plan is to provide information about interventions such as
corrosion control, then random sampling of high-risk targeted
households is needed. If an intervention is required, then
subsequent sampling at the same households is needed to assess
the effectiveness of the intervention. If a system is regarded
as non-compliant over a long time, then extension of the sampling
to other households may or may not be done. No assessment of
these possibilities was presented. The subcommittee recommends
that the statistical analysis and explanation of purpose be
included in the regulations.

The two-stage plan presents a commendable effort to reduce
the burden of sampling for situations in which there is clearly a
problem or clearly not a problem. However, there was no

2


-------
assessment of the performance of any of the sampling plans with
respect to decision errors (requiring intervention when none is
neededr or not requiring intervention when it might be helpful).
Thus, no rational basis was presented for a choice of plans.
Clearly, more detail 01* the applicability of the methods than was
presented either in writing or orally would be needed to make a
logical choice of methods. Another consideration is that no
matter which plan is selected, there is a definite need for
validation. The subcommittee recommends that methods and
validation be clearly developed.

The Subcommittee is concerned about the relevance of the
justification of the two-stage sampling plan in the context of
water lead sampling, which differs from a quality control model
in several ways?

(i) There is a finite, potentially identifiable pool of
high-risk households in each city, which can be sampled
repeatedly over time-

. (ii) The assessment of sampling plan performance by criteria
such as average outgoing quality limit (AQQI.) may not be
appropriate. The AOQL calculations assume that all households in
the targeted population will be brought to an acceptable levels
of water lead. This is not so, as even SAT may not accomplish
that. Some other assessment criteria for performance of the
plans should be calculated under alternative scenarios about
intervention effectiveness.

(iii) The proposed plans reduce the numeric information on
lead levels to a dichotomous pass/fail variable with a somewhat
arbitrary cutpoint (e.g. 15 ug/L). One-step and two-stage
sampling plans for continuously distributed lead measurements
should be evaluated.

The subcommittee recommends that the relevance of the two-
stage sampling plan be clearly explained and justified.

i
-------
carbonate alkalinity). If the inhibitor® are effective in
controlling the lead then there should be no further concern.

Lead is the health concern, not pH and alkalinity.

C.	The EPA proposal presents no evidence that alkalinity itself
is related to lead control.

The studies cited to support the alkalinity requirement do
nothing to justify an alkalinity of 30 ug CaC03/L. The alkalinity in
both studies (after treatment), was IS ug/L or less.

Additionally in these studies the pH was not controlled and in
both cases was raised more than 2 pH units i.e. 5 to 7.2 and 6.2
to 8.2. A field study is needed which compares pH adjustment
alone to both pH and alkalinity adjustments.

D.	The references cited by SPA in other SPA documents also
present no data to support statements made concerning the need to
adjust alkalinity.

The Murrell report (p.19) concludes that "lead leaching is
reduced by increasing pK and alkalinity to result in a less
corrosive water". This conclusion is not supported by the data
presented in the report. The data presented on pp. 76-78 are not
controlled for pH, and none of the data relate to locations where
adjustments were made to increase the alkalinity. The low
alkalinity (8 ug/L) water used in the data base had a low pH
(<6.4). Additionally, the waters with highest alkalinity (30
ug/L, south Huntington Water District) had a high pH (>8.0).

Clearly the conclusions reached regarding alkalinity are
misleading.

E.	A preliminary evaluation of the data from the American water
works service Co. lead survey indicates that pH, but not
alkalinity has a relationship to the lead contamination. A
comparison of average lead levels and the percent of sample

, exceeding 20 ug/L for groupings of pH and alkalinity was made.

In the pH range 1 to <8 the average lead levels were actually
lower (8 ug/L vs. 13 ug/L) when the alkalinity was below 30 ug
CaC03/L. In the samples with pH >8 there was no difference in
the lead levels when alkalinitios above and below 30 ug CaC03/L
were compared* Average lead levels were 5 ug/L in both cases.

Based on the preceding comments the Subcommittee recommends
that EPA simply require "corrosion control" and allow the utility
to implement the appropriate controls for their system.

Additional concern

The Subcommittee expressed the concern that the charge
was too narrow. Specifically, they also discussed certain
health issues, which had been treated in an earlier Agency
document (vizi Air Quality Criteria for Lead, EPA-600/8-83/028)
and examined by the clean Air scientific Advisory Committee of
the science Advisory Board in 1986.

4


-------
U.S. Environmental Protection Agency
Science Advisory Board
Drinking Water Subcommittee

CHAIRMAN AM VICE CHAIRMAN

Or, Gary Carlson [CHAIR], Department of Pharaacology and Toxicology, School
of Pharaacy, Purdue University, West Lafayette, Indiana 47907 (317) 494-1412

Dr. Robert Tardlff, [VICE~CHAIR], Principal, Environ Corporation* 1000 Potomac
St., N.W., Terrace level, Washington, DC 20007 (202) 317-7444

MEMBERS ASP CONSULTANTS

Dr. Julian B. Andelman, Graduate School of public Health, 130 Desoto SCreet*
Parraa Hall - Room A-711, University of Pittsburgh, Pittsburgh, PA 15261

Dr. Rose Dagirmanjian, professor, Department of pharaacology and Toxicology,
University of Louiaville, Louisville, Kentucky 40292 (502) 588-5160

Dr. Millias Glaze, Director, School of Public Health, University of California,
Los Angeles, 650 Circle Drive South, Loa Angeles, CA 90024 (213) 206-1278

Dr. J. Donald Johnson, professor, School of Public Health, University of
North Carolina, Chapel Hill, NC 27514 (919) 966*3856

Dr. David Kaufman, Department of Pathology, University of North Carolina,
Rm. 515 Brinkhous-Sullltt, Chapel Hill, North Carolina 27514 (919) 966-1396

Dr. Nancy Kim, Director, New York Department of Health, Bureau of Toxic
Substance Assessment, Room 359, Tower Building, Empire State Plaxa, Albany,
* New York 12037 (518) 474-2084

Dr. Verne Ray, Medical Research Laboratory, Pflser, Inc. Groton, Connecticut,

06340 (203) 441-3564

Or. Harold Schechter, Professor,. Chemistry Departaent, Ohio State University,
140 Wast 18th Avenue, Columbus, Ohio 43201 (614) 292-6300

Dr. Thomas Tephly, Professor, Departaent of Pharaacology, The Bowen
Science Bldg.» University of Iowa, Iowa City, Iowa 52242 (319) 335-7979

EXECUTIVE SECRETARY

Dr. Ct Richard Cotheru, Executive Secretary, Science Advisory Board
tA-lOlF] U.S. Environmental Protection Agency, Washington, D.C. 20460
(202) 382-2552


-------