United States
Environmental Protection
Agency
Water Engineering Research
Laboratory
Cincinnati OH 45268
EPA/600/9-85/007
February 1985
Research and Development
Plumbing Materials and
Drinking Water Quality:
Proceedings of a
Seminar
Cincinnati, Ohio
May 16-17, 1984
-------
EPA/600/9-85/007
February 1985
PLUMBING MATERIALS AND DRINKING WATER QUALITY:
PROCEEDINGS OF A SEMINAR
Cincinnati, Ohio May 16-17, 1984
Co-sponsored by
Office of Drinking Water
U.S. Environmental Protection Agency
Washington, D.C. 20460
and
Drinking Water Research Division
Water Engineering Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
Coordinated by:
Eastern Research Group, Inc.
Cambridge, MA 02138
Contract No. 68-03-3156
U.S. Environmental Protection Agency
Region 5, Library (,3;JL-15)
230 S. Dearborn Street, Room 1670
.Chicago, IL 60604
WATER ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
All papers and reports contained in these proceedings have not been
reviewed in accordance with U.S. Environmental Protection Agency's peer
and administrative review policies, and therefore, their contents do not
necessarily reflect the view of the Agency and no official endorsement
should be inferred. Also, mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
ii
-------
PREFACE
The National Interim Primary Drinking Water Regulations define a
Maximum Contaminant Level (MCL) as "the maximum permissible level of a
contaminant in water which is delivered to the free-flowing outlet of the
ultimate user of a public water system, except in the case of turbidity
where the maximum permissible level is measured at the point of entry to
the distribution system." With this definition, the Federal regulations
clearly recognize that the quality of drinking water can be affected by
its distribution lines and that it is the responsibility of the water
purveyor to consider these problems in providing water to its customers.
Traditionally, water utilities have taken control measures to prevent
or correct problems associated with water distribution mains. Studies and
reports during the past several years have clearly illustrated that distri-
bution problems are not limited only to distribution mains, but are also
associated with laterals and service lines as well as the interior plumbing
systems in dwellings and buildings. Some of these problems, such as lead
leaching from lead service lines and plumbing, have been common knowledge
for years. Other problems, such as the migration of gasoline and other
petroleum distillates through plastic pipe, have been more recently
identified.
On May 16-17, 1984, the Office of Drinking Water and the Drinking
Water Research Division of the U.S. Environmental Protection Agency
cosponsored a seminar to review drinking-water problems related to plumbing
materials (including those used in service lines) and to identify solutions
for dealing with these problems. In attendance were approximately 150
people representing government, manufacturers, trade associations,
consultants, and public interest organizations. During the first day,
speakers addressed three general topics: (1) the use, application, and
availability of plumbing materials (metallic and plastic); (2) the impact
of these materials on water quality; and (3) solutions to plumbing problems
related to water quality.
On the second day, four concurrent panel sessions were held for all
attendees to share their knowledge and to express their viewpoints on the
information presented. These panel sessions proved to be very worthwhile.
A wide variety of opinions and viewpoints were expressed on the four
panel topics: (1) joining alternatives for copper pipe; (2) metal pipe
and fitting alternatives for plumbing; (3) plastic pipe and fittings; and
(4) regulatory and compliance aspects.
iii
-------
The first section of these Proceedings includes the papers given by
the speakers on plumbing material applications, problems, and solutions.
The latter part of this section includes as a highlight "A Review of
European Developments in the Use of Plumbing Materials," presented by
Ivo Wagner of the Engler-Bunte Institute, Karlsruhe, West Germany. The
second section contains the reports of the panel sessions prepared by
the panel chairman. Although these reports vary somewhat in format,
each contains a summary of the opinions, conclusions, and recommendations
of the panel. The Proceedings conclude with a summary of the conclusions
and recommendations of the panel meetings and a synopsis of the seminar
presentations. This seminar was timely and successfully presented many
views on the complex issue of plumbing materials and water quality.
Frank A. Bell, Jr., Co-chairman
Criteria and Standards Division
Office of Drinking Water
Thomas J. Sorg, Co-chairman
Drinking Water Research Division
Water Engineering Research
Laboratory
Office of Research and Development
iv
-------
ABSTRACT
The seminar on Plumbing Materials and Water Quality was held at
the Andrew Breidenbach Environmental Research Center in Cincinnati, Ohio
on May 16 and 17, 1984. The purpose of the seminar was to review drinking
water problems related to plumbing materials and to identify alternative
solutions for dealing with these problems.
These proceedings are a compilation of speaker's papers and panel
session reports. The speaker's topics covered: 1) Plumbing Materials:
Extent of Use, Application and Availability, 2) Impact of Metallic
Plumbing Materials on Water Quality, and 3) Impact of Plastic Pipe and
Fittings on Water Quality.
Panel session reports are also presented on four topics: 1) Joining
Alternatives for Copper Pipe, 2) Metal Pipe and Fitting Alternatives for
plumbing, 3) Plastic Pipe and Fittings, and 4) Regulatory and Compliance
Aspects. These reports summarize the discussions that were held during
the panel sessions including the conclusions and recommendations made by
each group.
The document concludes with a general summary of conclusions and
recommendations of the entire seminar.
-------
CONTENTS
Preface ill
Abstract v
Plumbing Materials; Extent of Use, Application, and
Availability;
1. Metal piping and joining materials and fittings 1
2. Plastic piping and joining materials 8
Impact of Metallic Plumbing Materials on Water Quality;
3. Impact of lead piping and fittings on drinking
water quality 25
4. Impact of copper, galvanized pipe, and fittings on
water quality 35
5. Summary of impact of metallic solders on water quality . 59
Impact of Plastic Pipe and Fittings on Water Quality;
6. Impact of leaching by plastic pipe, fittings, and
joining compounds 65
7. NSF Standard and certification program for plastics pipe:
Standard 14 74
8. Evaluation of the permeation of organic solvents through
gasketed jointed and unjointed poly(vinyl chloride),
asbestos cement, and ductile iron water pipes 79
Solutions to Plumbing-Related Water Quality Problems;
9. Treatment or water quality adjustment to attain
MCLs in metallic potable water plumbing systems. ... 82
10. Drinking water regulations and their impact on
materials used in water systems 104
11. European developments in use of plumbing materials . . . 110
Panel Sessions;
1. Joining alternatives for copper pipe 124
2. Metal pipe and fitting alternatives for plumbing .... 129
3. Plastic pipe and fittings 147
4. Regulatory and compliance aspects 155
5. Summary of conclusions and recommendations 167
vii
-------
Summation - Plumbing Materials and Drinking Water Quality 172
Appendices
A. List of participants 177
viii
-------
METAL PIPING AND JOINING MATERIALS AND FITTINGS
by: Paul A. Anderson
Vice President
Copper Development
Association, Inc.
Greenwich Office Park No. 2
Greenwich, Connecticut 06836
Copper is the oldest engineering metal, and has been in continuous use
for plumbing since the time of the Pharaohs. Copper sheet rolled into
tubular form to conduct water has been found in excavated Egyptian tombs.
Over the centuries water has been conducted to the consumer through stone
aqueducts, hollow wooden logs, iron and steel pipe, and today's highly
engineered copper tubular products.
The Copper Development Association analyzes the 25 largest application
areas for copper and its alloys. Arranged in order of rank, based on 1983
data, plumbing and heating is the most important use for copper, followed
by building wiring. These two major building construction applications
accounted for 27 percent of U.S. copper and copper alloy shipments last
year. This is the first time plumbing and heating has led the list.
Each year in the United States, well over half a billion linear feet
of copper water tube are installed in water service and distribution
systems. This is equivalent to about 125,000 miles of hot and cold water
systems. Since 1946, when such statistics began to be gathered in the
United States, 17 billion feet of copper tube have been installed in water
service and distribution systems for buildings. That is more than 3
million miles.
But what do these copper shipments mean? The market can be measured
by usage intensity. Is copper usage in plumbing growing or declining
compared to building construction in general? To find out, we need to
compare copper plumbing tube usage to building construction activity.
Housing starts seem to be the best measure of total building
activity. Plumbing is predominantly a residential market in the United
States, and using square feet of total construction makes the index too
diffuse. The two curves in Figure 1 compare housing starts and copper
plumbing tube shipments over the last 12 years on an index basis, with 1972
set equal to 100. There is no doubt that copper plumbing tube consumption
-------
140
Index
U.S. Plumbing & Heating Market
COPPER PLUMBING TUBE VS. HOUSING STARTS
120
100
80
60
100
72
75
80
83
FIGURE 1
180
Index
160
Copper Usage Intensity
U.S. PLUMBING & HEATING MARKET
140
120
100
100
72
75
80
83
FIGURE 2
-------
follows housing starts. To convert these data to usage intensity, we can
divide the index for copper shipments by the index for housing starts.
This gives an intensity index — usage per housing start, in effect
(Figure 2). The result highlights something that was difficult to
appreciate when it was happening in 1974 and 75 and in 1980, 1981, and
1982. It shows that plumbing and heating has been a very healthy usage for
copper over the last 10 years, outperforming housing starts. How can this
be explained?
The major factors include the introduction of copper fire sprinkler
and solar heating systems, which brought copper technology into new
applications in the market in the mid-1970's. There has been increased
usage of large diameter copper tube starting in 1981. Large diameter tube
has, in the past few years, penetrated plumbing applications held by steel,
largely in commercial and industrial buildings. Finally, there seems to be
a trend, as in the 1960s, for more plumbing per dwelling unit.
Through the late 1960s and the 1970s, increasing copper usage
intensity in the plumbing and heating market trended downward. This
measure reached a low point in 1977. It reflected the well-known shift to
thinner wall tube in domestic plumbing systems. This was plumbing progress
that improved copper's usage efficiency — more plumbing per pound of
copper. Since 1977, however, the downward trend has been overpowered by
increasing use of larger diameter and thus heavier weight copper tube.
Fire sprinklers and solar heating use larger diameters than domestic water
systems. Large diameter tube usage is increasing in commercial and
industrial water supply systems.
To appreciate the relationship between the end-use market and the
reliability of the plumbing tube product, understanding of the manufacture
and quality assurance operations is needed. Copper metal is derived from
two sources, the mine and recycled scrap, each accounting for about 50
percent of total U.S. production.
In fact, copper is the most recycleable of all the engineering metals,
with the use of recycled scrap averaging between 44 and 54 percent of total
production over the past decade; the figure was 53 percent last year. The
reason: recycled copper is identical chemically and metallurgically with
new refined copper.
Copper cathode from the refinery is blended with recycled No. 1 scrap
from either wire and cable mills or brass mills, then remelted to produce
the copper billet. These billets are held in the cast shop until the
production quality control laboratory has completed its quality analysis
and releases the billet for conversion into tubular products in the mill.
Massive equipment is involved in the conversion of the billet into
tubular products. Following inspection, each billet is heated to the
proper temperature prior to insertion into the extrusion press for
conversion into a tube shell. The ram is inserted into the billet cavity
-------
and pressure applied to the heated billet, moving it to the die surface.
The billet is then pierced prior to operation of the main ram. The main
ram then pushes the hot billet through the die with the center punch of the
ram acting as the internal core around which the tube forms as it is
extruded through the die. Surface defects are prevented from being
extruded into the tube shell by proper ram and extrusion chamber clearance.
The tube shell is further reduced to final size through a series of
diameter/wall thickness reductions on bull blocks or draw benches. Drawing
lubricants used by seamless tube manufacturers are lead-free,
water-soluble, lightweight mineral oils. The nature of tube size reduction
is such that the internal surface is wiped clean of any residue by a
floating plug and tool steel die.
These tubular products are produced either as copper or copper alloys
by the seamless or welded manufacturing methods. A total of 12 ASTM
specifications have been developed to cover these commercial products.
Ninety-eight percent of water tube is manufactured according to ASTM
Standard Specification B88.
ASTM B88 describes three types of copper tube: Types K, L and M.
Each of these tube types is a series of standard diameter sizes with
specified wall thicknesses. Type K is the thickest wall, Type M the
thinnest, L intermediate. A lightweight distribution tube is also
available. About three quarters of all copper water tube installed is
either half-inch or three-quarter-inch in nominal size.
Prior to packaging and shipment, all tube must undergo product
inspection to ensure that the finished tube meets quality assurance
requirements. This is accomplished by the eddy-current, nondestructive
inspection operation, which is not subject to operator fatigue, so exacting
acceptance criteria are continually maintained on the production line.
Immediately after the eddy-current inspection station, the name of the
manufacturer and tube type are marked on the tube surface by inking and
incise impressions. This identifies the tube as to source, type, and
specification and assists the plumbing inspector in verifying that it meets
code requirements.
Solder joint plumbing fittings are produced according to the same
exacting manufacturing standards as tube (ANSI B16.18 or B16.22). They are
made in a wide variety of shapes and sizes, including couplings, T's,
elbows, and reducing and special fittings. All quality assurance
operations applied to tube apply here as well.
Up to now, my remarks have been directed to the role of copper tube in
use, its production and quality assurance. Let me now comment on why
copper has achieved such a remarkable level of acceptance by the building
community.
Copper is the dominant material used in potable water systems. Its
combination of corrosion resistance in hot and cold systems and ease of
-------
installation is unmatched by any other plumbing material in use today.
Service experience in general indicates that copper plumbing tube will
outlast the building in which it is installed. Further, the combination of
nontoxicity and nonflammability properties makes it the leading choice.
As with all materials, copper can be attacked by some highly
aggressive environments, and problems arise occasionally with certain well
waters, with very soft waters high in carbon dioxide, and in systems where
the water flow velocity in excess of about 5 fps causes erosion.
In a modern waterworks, the chemist in charge not only prescribes
treatments to make the water conform to health regulations, such as to free
it of pathogenic organisms, but for well waters he must also make
arrangements, if necessary, for chemical treatments designed to reduce the
corrosiveness of the water. The water purveyor may not fully recognize
this responsibility or the economic advantages of providing a water
suitably adjusted chemically to reduce its corrosiveness. All of this is
directly relevant to one of the topics of this conference: lead pickup
from soldered plumbing systems.
Potable waters are developed from a variety of sources, including
lakes, rivers, streams, and wells. Reservoirs are used to store water
where there is seasonal or annual variation. Surface waters tend to be
saturated with air. Waters taken directly from wells may be low in oxygen
but may contain much higher levels of carbon dioxide and minerals than
surface supplies. Unlike surface waters, well waters are usually free of
algae and organic matter.
Waters are routinely treated for biological purity, but in addition,
each water source has to be examined to determine whether some form of
correction or treatment is necessary to control its corrosiveness. Often,
all that is required is pH adjustment. The form of corrective treatment
may be to control scaling or to reduce corrosion. Treatments with other
objectives must be evaluated for the possible effect on the corrosiveness
of the water. For example, water must meet the new EPA limit of one
turbidity unit. Flocculation, the treatment to clarify turbid water, often
uses a coagulant such as alum to remove suspended matter. Water treatment
can also be used to protect consumers from lead pickup by aggressive acid
waters.
Suitable treatments are routine in modern waterworks. Where the water
purveyor continuously controls the character of the water fed to the mains,
he not only ensures optimum corrosion performance of his own supply system
but also greatly reduces costly customer-related water problems.
Unfortunately, smaller water purveyors in many cases do not provide
corrective treatments, either from lack of understanding of the need or
from lack of the proper facilities.
Corrosion and metal pickup, including lead pickup from poorly made
solder joints, can be prevented in domestic plumbing systems by refraining
from the use of aggressive waters or with effective water treatment of
-------
either surface or well waters. This has been the case for over 50 years of
highly serviceable use of copper tube and fittings for water distribution
systems. Examples of successful treatments that have been introduced
include simple aeration, the addition of caustic soda, soda ash or lime.
Neutralizing filters have also been used in private dwellings. Up to now,
these treatments have been aimed primarily at protecting consumers against
pitting and general corrosion, but the same principles apply for protection
against lead pickup by acid soft waters from poorly made joints. Consider
a few case histories that demonstrate the effectiveness of this approach to
consumer protection.
At one location in California, simple aeration of the water reduced
the free carbon dioxide content from over 20 mg/1 to about 10 mg/1 and was
adequate to alleviate a corrosion problem in the system served. This was
done simply by spraying the well water into the water supplier's large
storage tank via a large-diameter perforated pipe at the top of the tank
reservoir.
In another area treatment with caustic soda (NaOH) was effective in
neutralizing the free carbon dioxide in a pitting water serving a
particular housing development, causing the pitting attack to cease. When
treatment was stopped, the pitting resumed. When a nonpitting water was
finally introduced into the system, pitting stopped altogether. In this
case, the treatment consisted of the addition of caustic soda to raise the
pH and eliminate the free carbon dioxide.
Another example of the effectiveness of water treatment is in the
community of Fort Shawnee, Ohio. Here, an isolated housing development of
about 100 residences and condominiums is served by a single well-water
source. About five years ago, an epidemic of cold water pitting in these
homes was brought to our attention. Of the 100 or so dwellings, 16
suddenly experienced 28 pitting perforations in their cold water
distribution systems.
Equipment was installed in the waterworks to inject sodium carbonate
(N32C03 H20) into the water to raise the pH and neutralize the free
carbon dioxide. Since the treatment began there has not been a single
additional corrosion failure within the buildings served by the treated
water.
In Suffolk County, New York, a water supply system with about 150 well
fields (two to three wells per field) has for many years treated the
combined output of more than 350 wells by injecting a slurry of lime (CaO)
into the supply. The remaining wells are treated with caustic soda. When
properly treated, this raises the pH to about 7.5, reducing the C02 to
less than 5 mg/1, and eliminates the copper water tube corrosion problems
on the system almost completely.
A housing development in San Bernardino County, California, started to
experience corrosion approximately 1.5 years after the residences were
initially occupied. Because of the reluctance of the water utility to
-------
introduce the needed treatment, neutralizing filters were introduced by the
developer at each residence. These units contained a bed of dolomite
(CaC03 MgO) which functions by reacting with and removing the free
carbon dioxide from the water with concurrent increase of pH of the water.
The excess hardness (Ca and Mg) is then reduced by passing the
first-stage treated water through a second tank containing an ion exchange
resin.
To summarize, copper tube and fittings have been the dominant accepted
plumbing products in the United States for several decades, providing
trouble-free service to residential and commercial users.
It would be foolhardy to claim that corrosion has not occurred in
copper plumbing systems, but our studies show this situation to be highly
localized and preventable. Other than for faulty design and poor
workmanship, corrosion and metal pickup are invariably associated with
aggressive water compositions. Successful treatments utilized by local
water utilities have corrected the water chemistry problems. This has
resulted in trouble-free plumbing systems for the consumer and high quality
drinking water. The same approach can protect consumers from lead pickup
from poorly made solder joints in those few — really extremely few and
isolated — potable water systems that are aggressive as a result of
acidity and lack of hardness.
-------
PLASTIC PIPING AND JOINING MATERIALS
by: Stanley A. Mruk
Technical Director
Plastics Pipe Institute
A Division of The Society of the
Plastics Industry, Inc.
355 Lexington Avenue, 6th Floor
New York, New York 10017
INTRODUCTION
First use of thermoplastics piping for conveying potable water dates
back to the early 1950s. Since then, thermoplastics piping has grown to
become a most widely used material for a broad range of applications
including water distribution, sewerage, drainage, plumbing, gas
distribution, industrial uses, and power and communications ducting.
Experience has established that thermoplastics piping very often provides
the most economical and satisfactory long-term solution.
The broad acceptance of thermoplastics piping not only evidences the
inherent capabilities and versatility of these materials, but it also
reflects the attention that has been given, from the start of the plastics
industry, to the development of a standardization system that suitably
addresses all the major performance requirements, including preservation of
water quality. The key element of the standardization system covering
thermoplastics piping for potable water applications is that it scrutinizes
each and every commercial formulation for two most important
characteristics: no adverse effect to water quality, and long-term
strength under end-use conditions.
AVAILABLE MATERIALS FOR PLUMBING APPLICATIONS
The major thermoplastic materials used for potable water transport and
plumbing applications include polyvinyl chloride (PVC), chlorinated
polyvinyl chloride (CPVC), polyethylene (PE), polybutylene (PB), and
acrylonitrile-butadiene-styrene (ABS). As indicated by Table 1, all of
these materials are used to manufacture pressure piping. However, some
materials and products are used preferentially over others depending on the
application, and Table 1 also shows current market preferences.
-------
TABLE 1
THERMOPLASTIC PIPING PRODUCTS
FOR POTABLE WATER APPLICATIONS
KEY: X - Most Frequently Used
0 - Less Frequently Used
PRODUCT
MATERIAL STANDARD
PVC
AWWA
ASTM
ASTM
ASTM
ASTM
ASTM
C
D
D
D
D
D
900
2241
2466
2672
1785
2740
TITLE (ABBREVIATED)
OF STANDARD
PVC Pressure Pipe for Water
PVC Plastic Pipe SDR/PR
PVC Fittings, Sch. 40
PVC Pipe, Belled End
PVC Plastic Pipe, Sch. 40-80
PVC Plastic Tubing
SIZE
MAINS
RANGE
(INCHES) RURAL MUNICIPAL
4 to
1/2
1/2
1/2
1/2
1/2
12
to
to
to
to
to
24
8
8
12
0 X
X 0
X 0
X 0
0 0
SERVICES
0
O
0
O
0
YARD
PIPING
X
X
X
0
HOT/COLD
DISTRIBUTING
1-1/4
PE
PB
CPVC
ABS
AWWA
ASTM
ASTM
ASTM
ASTM
ASTM
ASTM
ASTM
AWWA
ASTM
ASTM
ASTM
ASTM
ASTM
ASTM
ASTM
ASTM
ASTM
ASTM
D
D
D
D
D
D
D
F
C
D
D
D
D
D
F
F
F
D
D
901
2239
2609
2737
2683
3281
3035
714
902
2662
2666
3000
3309
2846
441
442
438
2282
468
PE Pipe 4 Tubing for
Services
PE Plastic Pipe, SDR/PR
Plastic Insert Fittings
PE Plastic Tubing
PE Fittings, Socket Type
PE Fittings, Butt Type
PE Pipe, SDR/PR, OD
Controlled
PE Pipe, SDR/PR, Large Diam.
PB Pipe & Tubing for Services
PB Pipe, SDR/PR
PB Tubing
PB Pipe SDR/PR, OD
Controlled
PB Hot/Cold Water Systems
CPVC Hot/Cold Water Systems
CPVC Pipe, Sch. 40-80
CPVC Pipe, SDR/PR
CPVC Fittings, Sch. 40
ABS Pipe, SDR/PR
ABS Fittings, Sch. 40
1/2
1/2
1/2
1/2
1/2
1/2
1/2
3 to
1/2
1/2
1/2
1/2
1/2
1/2
1/2
1/2
1/2
1/2
1/2
to
to
to
to
to
to
to
63
to
to
to
to
to
to
to
to
to
to
to
3
6
4
2
4
10
6
3
6
2
6
2
2
12
12
6
12
8
0 0
0
0 0
0 0
o o
0
0
X
X
X
X
0
0
0
X
X
X
0
0
0
0
X
X
0
0
0
0
o
0
0
0
X
X
0
0
0
-------
The major products used for the various non-pressure plumbing
applications are presented in Table 2. Pressure and non-pressure piping
made from ABS, PVC, and CPVC can be jointed by solvent cementing. Table 3
identifies the standard cements. The polyolefins, PE and PB, are most
often joined by mechanical systems, primarily insert, compression, and
flare. They can also be joined by heat fusion, but this technique is
usually used for larger pipe and by major users such as gas distribution
companies.
A very popular technique for joining buried PVC pipe is the
elastomeric gasket. Almost all PVC American Water Works Association (AWWA)
C-900 water distribution pipe as well as American Society for Testing and
Materials (ASTM) D 3034 sewer pipe is manufactured with integral bells with
gaskets, into which the spigot end may be pushed for easy joining in the
field. The various available joining methods are summarized in Table 4.
Estimates of the approximate market share held by the various
thermoplastic materials in water distribution and plumbing applications are
given in Table 5. These are very rough estimates because accurate
statistics on actual shipments to different market sectors presently are
not collected by government or industry. The estimates presented in Table
5 have been developed by consolidating the best judgment of a number of
knowledgeable people in the industry.
ATTRIBUTES OF PLASTIC PIPING
When evaluating the capacities and limitations of any piping material,
it is important to examine its performance in an operational system rather
than to compare certain individual properties with those offered by
alternate materials. Individual properties of materials, such as shown in
Table 6, are measured under conditions that do not sufficiently approximate
those encountered in actual use. Furthermore, each material can have its
own particular problems, all of which need to be recognized and resolved if
the material is to be used successfully in a working system. With
traditional materials, the resolution of many of these problems has come
about very gradually through years of experience. With the newer
thermoplastics materials, the evolution to mature engineering materials has
been accelerated through an emphasis on the technological assessment of
potential performance limitations, particularly long-term durability. That
this evolution is proceeding most successfully is indicated by the quite
successful application history and the excellent acceptance of plastics for
a broad range of applications.
ADVANTAGES
The principal advantages of thermoplastics piping are as follows:
Immunity to "corrosion". Plastics do not rust, pit, or corrode in the
sense that metals do. Being nonconductors, they are not susceptible to
galvanic or electrochemical effects and, therefore, are unaffected by
10
-------
TABLE 2
THERMOPLASTIC PIPING PRODUCTS
FOR NON-PRESSURE PLUMBING APPLICATIONS
KEY: X - Most Frequently Used
0 - Less Frequently Used
BUILDING
DRAINAGE & VENT
PRODUCT
MATERIAL STANDARD
PVC
ABS
ASTM
ASTM
ASTM
ASTM
ASTM
ASTM
ASTM
ASTM
ASTM
D
D
D
D
D
D
D
F
D
2665
2949
3311
2729
3033
3034
2661
628
2751
TITLE (ABBREVIATED)
OF STANDARD1
PVC
3-In
DWV
PVC
PVC
PVC
ABS
ABS
ABS
DWV Pipe &
. PVC Thin
Fittings
Wall DWV Pipe
Fitting Patterns
Drain Pipe
Sewer Pipe
PSP
Sewer Pipe
PSM
DWV Pipe
-------
TABLE 3
SOLVENT CEMENTS FOR THERMOPLASTIC PIPING
MATERIAL
PRODUCT
STANDARD
TITLE (ABBREVIATED) OF STANDARD
ABS
PVC
ABS/PVC
CPVC
ASTM D 2235
ASTM D 2564
ASTM D 3138
ASTM F 493
Solvent Cements for ABS Piping
Solvents for PVC
Solvent Cements for ABS/PVC
Transitions
Solvent Cements for CPVC Piping
TABLE 4
JOINING METHODS
METHOD
THERMOPLASTIC PIPE
ABS
PVC CPVC
PE
PB
Solvent Cementing
Heat Fusion
Threading (Sch. 80 only)
Insert
Mechanical Compression
Flaring
Elastomeric Seal
0
-
0
-
0
-
0
0
-
0
-
0
-
0
0
-
0
-
0
-
0
-
0
0
0
0
0
0
-
0
-
0
0
o
0
12
-------
TABLE 5
ESTIMATED MARKET SHARES
OF THERMOPLASTIC PIPING IN WATER
DISTRIBUTION AND PLUMBING APPLICATIONS
MATERIAL
PVC
PE
PB
CPVC
ABS
Others
PRESSURE
MAINS
RURAL MUNICIPAL
80 25
1
19 75
100 100
PIPE
SERVICES
10
30
10
50
100
YARD HOT/COLD
PIPING DISTRIBUTING
20
60
6
9
20 85
100 100
NON-PRESSURE PIPE
(FOR HOME CONSTRUCTION ONLY
MATERIAL
PVC
ABS
PE (Corrugated)
Others
BUILDING
DRAINAGE & VENT
ABOVE BELOW
GROUND GROUND
60 60
20 20
20 20
100 100
BLDG.
SEWER
70
10
20
100
)*
BLDG.
STORM
SEWER DRAIN
70 40
10
60
20
100 100
*Market shares for plastics are substantially less for other
types of construction.
13
-------
TABLE 6
TYPICAL PHYSICAL PROPERTIES OF
MAJOR THERMOPLASTIC PIPING MATERIALS
ASTM
PROPERTY AT 75° F TEST NO. ABS PVC CPVC PE PB
Specific Gravity D-792 1.08 1.40 1.54 0.95 0.92
Tensile Strength
psi (103) D-638 7.0 8.0 8.0 3.2 4.2
Tensile Modulus
psi (105) D-638 3.4 4.1 4.2 1.3 0.55
Impact Strength,
Izod ft-lb/inch
notch D-256 4 1 1.5 |10 |10
Coeff. of Linear
Expansion in/in-
°F (105) D-696 6.0 3.0 3.5 9.0 7.2
Thermal Conductivity
Btu-in/hr-ft-
°F D-177 1.35 1.1 1.0 3.2 1.5
Specific Heat
Btu/lb-°F - 0.34 0.25 0.20 0.55 0.45
Approx. Operating
Limit1
°F, nonpressure - 180 150 210 160 210
°F, pressure - 160 130 180 140 180
ifixact operating limit will vary for each particular
commercial plastic material. Effect of environment should be
considered.
14
-------
acids, bases, salt, "corrosive" waters, or "hot" soils. They can be,
however, susceptible to other forms of degradation, such as direct chemical
attack and solvation.
Nonconductivity of electricity. This capacity reduces the problem of
electrolytic corrosion of connected metallic piping.
Light weight; ease of use. These features simplify handling and
installation with attendant cost savings and worker accident reduction.
Their light weight also reduces dead-load forces on the structure.
Hydraulically smooth surfaces. Results in most efficient flow and
minimal accumulation of deposits.
Elimination of brazing and soldering. During construction, fire
hazards and work accident hazards from torches used for brazing and
soldering of metal piping are eliminated.
Low stiffness. Allows for the accommodation of deformations with
minimal stress development. For certain materials, i.e., PE and PB,
stiffness is low enough to allow for coiling of smaller diameter pipe.
Decreases sound transmission.
Ductility. The capacity of thermoplastics to undergo significant
permanent deformation without fracture allows them to shed off, by means of
localized deformation, concentrated earth and other loading.
Durability. Properly selected and installed thermoplastics piping can
provide very high reliability and long service life.
Energy efficiency. Significantly less energy is required to
manufacture a given length of plastic pipe than that needed to produce most
metal pipes. For example, it has been estimated that the energy
requirements for the manufacture of 2-inch, 160-pound-per-square-inch (psi)
PE pipe are 60 percent of those for the same size copper pipe.
LIMITATIONS
The principal limitations of thermoplastics piping include:
Time, temperature, and environment sensitivity. All the engineering
properties, including strength, stiffness, and strain at failure, are
dependent on load, duration of loading, temperature, and environment.
Furthermore, the sensitivity can vary not only from one type of plastic to
another (i.e., PVC and PE), but also' can significantly differ within the
same material type depending on the exact choice of polymer and additives
as well as on processing conditions. These limitations must be recognized
in product design and standardization, particularly when limiting
conditions, such as higher temperature, are likely to be encountered in
actual service.
15
-------
High coefficient of expansion/contraction. The thermal
expansion/contraction of unrestrained plastics can be from 6 to 10 times
greater than with metals. This requires greater attention to support
spacing in horizontal runs. However, because of their significantly lower
stiffness, expansion/contraction can often be restrained without the
development of excess forces.
Light weight. Because of its low mass, thermoplastic piping offers
little acoustical dampening to airborne sound. The light weight, in
combination with low stiffness, can also result in greater movement of
piping from hydraulic loads. This necessitates that more attention be
given to proper pipe restraint in above-ground installations.
Electrical insulator. Grounding of electrical appliances and
electrical circuits must be accomplished via a separate conductor and
ground spike.
Combustibility. Control of the spread of fire, and toxic gases from
plumbing chases, or walls containing thermoplastic materials, may require
special attention to installation details, particularly when penetrating
fire-rated building components. However, modern fire technology indicates
that such attention to effects of wall penetration is appropriate to all
piping materials.
DESCRIPTION OF THERMOPLASTIC PIPE MATERIALS
"Plastic" is as indefinite a term as "metal." First, plastics consist
of two basic groups, thermosetting and thermoplastics, which are both used
in the manufacture of piping. Thermoplastics, as the name implies, soften
upon the application of heat and reharden upon cooling. This
characteristic enables them to be readily formed into a wide variety of
different shapes. Thermosetting plastics, on the other hand, form
permanent shapes when cured by the application of heat or a "curing"
chemical.
Secondly, among the thermoplastics the characteristics and properties
depend both on the specific chemical composition of the base resin and on
the kind and amounts of additives included in the commercial composition.
For example, it is possible to formulate mixtures of polyvinyl chloride
(PVC) resins plus appropriate additives to yield "vinyl" compositions that
range from a clear, soft, and pliable product (such as is used to produce
laboratory tubing and upholstery) to a rigid and strong material (such as
for pressure pipe). For these reasons, standardization of plastic pipe
materials not only classifies the base resin, but also defines the
composition by a combination of specific end-property requirements that
take into account possible property variability contributed by formulation
additives of a plastics composition. Additives for plastics can consist of
antioxidants, stabilizers, colorants, ultraviolet protective screens,
lubricants, property modifiers (i.e., impact and stiffness), and fillers.
16
-------
In the case of thermoplastic pipe materials intended for pressure
piping and potable water service, the current material classification
system requires that each commercial composition used for such product be
evaluated to: (1) demonstrate that no deleterious substance will leach
from the plastic into the potable water; and (2) establish the material's
long-term strength under conditions of service. These requirements
recognize that, because of variabilities in polymer and additives, plastics
compositions cannot be adequately defined solely on the basis of the
"short-term" physical property requirements that are part of ASTM material
standards. These requirements alone are not a source of sufficient
assurance that potable water quality will not be adversely affected and
that the material has adequate long-term strength capacity for the intended
service. Consequently, for a plastic material to be acceptable for potable
water pressure piping, it must satisfy, under requirements in effect since
the early 1960s, two additional criteria: (1) it must be approved for
contact with potable water based on requirements not less strict than those
in National Sanitation Foundation (NSF) Standard No. 14; and (2) it mist
have a recommended hydrostatic design stress (HDS) based on long-term
pressure testing. The NSF requirements and programs are described in
detail in a companion presentation by Dr. Nina McClelland of NSF.
The recommended HDS for individual commercial formulations is
established from long-term pressure testing data on the basis of ASTM D
2837, "Standard Method for Obtaining Hydrostatic Design Basis for
Thermoplastic Pipe Materials." The Hydrostatic Stress Board of the
Plastics Pipe Institute (PPI) has been issuing since the mid-1960s
recommendations of HDS for commercial grades of thermoplastic piping
materials. These recommendations, which are based on the methods of D 2837
and the additional requirements given in PPI TR-3, "Policies and Procedures
for Developing Recommended Hydrostatic Design Stresses for Thermoplastic
Pipe Materials," appear in the regularly updated report PPI TR-4,
"Recommended Hydrostatic Strengths and Design Stresses for Thermoplastic
Pipe and Fittings Compounds." Nearly all U.S. thermoplastic pipe pressure
standards reference the PPI recommendations. NSF policy requires such
recommendation for each approved pressure piping material.
The distinctive characteristics of the principal thermoplastic piping
materials are as follows:
Polyvinyl Chloride (PVC). PVC piping is made from compositions
containing no plasticizers and only minimal quantities of other
ingredients. To differentiate these materials from flexible, or
plasticized, PVCs (from which are made such items as upholstery, luggage,
and laboratory tubing), they have been labeled rigid PVCs. Rigid PVCs used
for piping range from a Type I to Type III as identified by a previous
classification system that is still in use. In this system the Type
designations are .supplemented by Grade designations (i.e., Grade I or II)
which further define the material's properties. Type I materials, from
which most pressure and non-pressure pipe is made, have been formulated to
provide optimum strength as well as chemical and temperature resistance.
Type II materials are those formulated with modifiers that improve impact
17
-------
strength but that also somewhat reduce, depending on modifer type and
quantity, the aforementioned properties of Type I materials. There is
little call for Type II pipe, as the impact strength of the stronger Type I
pipe is more than adequate for most uses. Type III materials contain some
inert fillers which tend to increase stiffness concomitant with some
lowering of both tensile and impact strength and chemical resistance. Some
non-pressure PVC piping, such as that used for conduit, sewerage, and
drainage, is made from Type III PVCs.
The currently used classification system for rigid PVC materials for
piping and other applications is described in ASTM D 1784, "Standard
Specification for Rigid Polyvinyl Chloride and Chlorinated Polyvinyl
Chloride Compounds." This specification categorizes PVC materials by
numbered cells that designate value ranges for the following properties:
impact resistance (toughness), tensile strength, modulus of elasticity
(rigidity), deflection temperature (temperature resistance), and chemical
resistance. The following table cross-references the designation of the
principal PVC materials from the older to the newer classifications system.
OLDER SYSTEM
CELL CLASSIFICATION SYSTEM OF
ASTM D 1784 MINIMUM CELL CLASS TYPE AND GRADE DESIGNATION
12454-B Type I, Grade I PVC 11
12454-C Type I, Grade II PVC 12
14333-D Type II, Grade I PVC 21
13233 Type III, Grade I PVC 31
The designation of PVC materials that have been formulated for
pressure piping also includes a code indicating their maximum recommended
HDS for water at 73° F as determined from long-term pressure testing. For
example: PVC 1120 is a Type I, Grade I PVC with a maximum HDS of 2,000 psi
for water at 73° F; and PVC 2110 is a Type II, Grade I PVC with a 1,000 psi
HDS. The last two digits in the number code identify, in hundreds of psi,
the material's maximum recommended design stress. Table 7 presents all the
PVC designations used in pressure piping.
The combination of good long-term strength with higher stiffness
explains why PVC has become the principal plastic pipe material for both
pressure and non-pressure applications. Major uses include: water mains;
water services; irrigation; drain, waste, and vent (DWV) pipe; sewerage and
drainage; well casing; electric conduit; and power and communications
ducts. A much broader range of fittings, valves, and appurtenances of all
types is available in PVC than in any other plastic.
Chlorinated Polyvinyl Chloride (CPVC). As implied by its name,
chlorinated polyvinyl chloride is a chemical modification of PVC. CPVC has
properties very similar to PVC, but the extra chlorine in its structure
extends its temperature limitation by about 50° F (28° C), to nearly 200° F
(93° C) for pressure uses and about 210° F (99° C) for non-pressure
applications. ASTM D 1784, the rigid PVC materials specification, also
18
-------
covers CPVC which it classifies as Class 23477-B. By the older designation
system, it is known as Type IV, Grade I PVC. CPVC's for pressure pipe are
designated CPVC 4120 (i.e., Type IV, Grade I CPVC with a maximum
recommended hydrostatic design stress of 200 pounds per square inch
(lb/in2) for water at 73.4° F. At 180° F (82° C) the maximum recommended
hydrostatic design stress for CPVC is 500 lb/in2 (3.4 MPa).
Principal applications for CPVC are for hot or cold water piping and
for many industrial uses that take advantage of its higher temperature
capabilities and superior chemical resistance.
Polyethylene (PE). Polyethylene is the best-known member of the
polyolefin group (plastics that are formed by the polymerization of
straight-chain hydrocarbons, known as olefins), which includes ethylene,
propylene, and butylene. Polyethylene plastics are tough and flexible even
at subfreezing temperatures. They are generally formulated with only an
antioxidant (for protection during processing) and some pigment (usually
carbon black) or other agent designed to screen out ultraviolet radiation
in sunlight which, over long-term exposure, could be damaging to the
natural-color polymer.
ASTM D 1248, the PE molding and extrusion materials specification,
classifies these materials into three types depending on the density of the
natural resin. Type I consists of lower-density materials which are
relatively soft and flexible and have low heat resistance. Type II PEs are
of medium density, slightly harder, more rigid, and more resistant to
elevated temperatures; they also have better tensile strength. Type III
materials show maximum hardness, rigidity, tensile strength, and resistance
to the effects of increasing temperature. Pipe is made almost exclusively
from Type II and Type III PEs. ASTM D 1248 also provides for grade
designations to further classify PEs according to other physical
characteristics. PE piping materials for pressure piping are classified by
a designation system that combines the type and grade coding with that for
the maximum HDS for water at 73.4° F (23° C). PEs used for pressure piping
are listed in Table 6.
Outstanding toughness and relatively low flexural modulus, which
permits coiling of smaller diameter pipe, are large factors in PE's
prominence in gas distribution and water service piping. Other features,
such as heat fusibility, good abrasion resistance, and availability in
large diameters (up to 96 inches) account for the use of PE piping for
chemical transfer lines, slurry transport, sewerage force mains, intake and
outfall lines, and renewal (by insertion into the old pipe) of deteriorated
sewers, gas, water, and other pipes.
Polybutylene (PB). Polybutylene is a unique polyolefin because its
stiffness resembles that of low-density PE but its strength is higher than
that of high-density PE. However, its most significant feature is its
better retention of strength with increasing temperature. Its higher
temperature limit is higher than that of any PE: nearly 200° F for
pressure uses and somewhat higher for non-pressure applications.
19
-------
PE piping materials are covered by ASTM D 2581. Pressure piping PBs
are designated as PB 2110 (see Table 7), which indicates a maximum design
stress of 1,000 psi for water at 73° F. At 180° F this design stress is
500 psi, the same value as for CPVC.
Major applications for PB piping take advantage of its improved
temperature resistance. They include hot/cold piping and industrial uses
such as hot/cold effluent lines. Because of its excellent abrasion
resistance, PB is also used for slurry lines.
Acrylonitrile-Butadiene-Styrene (ABS). ABS comprises a family of
materials that are formed from three different monomers (chemical building
blocks): acrylonitrile, butadiene, and styrene. The properties of the
components and the way in which they are combined can be varied to produce
a wide variety of properties. Acrylonitrile contributes rigidity,
strength, hardness, chemical resistance, and heat resistance; butadiene
contributes toughness; and styrene contributes gloss, rigidity, and ease of
processing.
ASTM D 1788 classifies ABS plastics into numbered cells that designate
value ranges for each of three properties: impact strength (toughness),
tensile stress at yield (strength), and deflection temperature under load.
ABS pipe materials are categorized into Types and Grades in accordance with
established minimum cell requirements for each Type and Grade. Like the
other major thermoplastics, ABS materials for pressure pipe are designated
by a coding that identifies both short-term properties and long-term
strength. ABS pressure-pipe materials are listed in Table 7.
An advantageous combination of toughness with good strength and
stiffness largely accounts for the most common uses of ABS pipe, i.e., for
DWV applications as well as for sewers, well casings, and communications
ducts.
A foam-core construction ABS pipe is commercially available. The wall
of the product consists of thin inner and outer solid skins sandwiching a
high-density foam. The primary benefit of this construction is improved
ring and longitudinal (beam) stiffness in relation to the amount of
material used. Applications of foam-core pipe are for non-pressure uses.
A standard for a PVC pipe with a foam-core wall construction is under
development at ASTM.
Polyacetal, also known as polyoxymethylene (POM), is a strong, hard,
highly crystalline thermoplastic offering good rigidity, strength, and
toughness, and relatively good temperature resistance. Because of these
properties, POMs are increasingly being used to mold various hot/cold water
plumbing components including valves, faucets, and compression fittings for
use with PB and CPVC tubing.
Crosslinked PE. A new standard has recently been issued by ASTM which
..overs hot/cold water piping made from crosslinked PE (XPE). The
crosslinking changes the PE from a thermoplastic to a thermoset and endows
20
-------
TABLE 7
MAXIMUM RECOMMENDED HYDROSTATIC DESIGN STRESSES (RHDS)
FOR THERMOPLASTIC PIPE MATERIALS FOR WATER AT 73° F
MATERIAL DESIGNATION1 MAXIMUM RHDS, IN PSI, FOR WATER AT 73° F2
PE 1404 400
PE 2305 500
PE 2306 630
PE 3036 630
PE 3406 630
PE 3408 800
XPE 330 630
PB 2110 1000
CPVC 4120 2000
PVC 1120 2000
PVC 1220 2000
PVC 2110 1000
PVC 2112 1250
first two digits code the material according to
short-term properties. The last two digits code the maximum
RHDS expressed in hundreds of pounds per square inch.
2Since thermoplastics, even though of the same material
designation, may be affected differently by increasing
temperature, RHDS for higher temperatures must be established
for each specific commercial product.
21
-------
the material with improved higher-temperature strength. Because they are
crosslinked, these materials cannot be heat fused. XPE has not yet been
commercially offered in the United States and it is not presently approved
by any major plumbing code.
Styrene Rubber (SR). Styrene rubber compositions are combinations of
polystyrene, in the greater part, with rubber. SR plastics are relatively
strong and stiff. However, their impact strength is somewhat lower than
other plastics, making SR more susceptible to damage by impact,
particularly during cold weather.
SR plastics are used exclusively for non-pressure applications. Once
relatively popular, they are now used little, having been largely displaced
by the other materials.
Solvent Cement Materials. Solvent cements, which are often used when
joining PVC, CPVC, and ABS piping, are compositions consisting essentially
of a solvent, or combination of solvents, in which has been dissolved a
quantity of the base plastic material sufficient to give the cement the
body and consistency required for proper applications. Small amounts of
inert fillers are sometimes also added to control these properties as well
as shrinkage during drying. The principal solvents used include
tetrahydrofuran (THF) for PVC and CPVC cements and methyl ethyl ketone
(MEK) for ABS cements. To control rate of drying and other properties,
PVC, CPVC, and ABS cements will sometimes include other solvents, most
often cyclohexanone (CH) and dimethylformamide (DMF). As in the case of
pipe, solvent cement compositions for use with potable water piping must be
approved for this purpose by NSF or an equivalent authority.
FUTURE TRENDS
The generally successful use of plastics piping over the past 30 years
has demonstrated that it is a cost-effective alternative for many water
distribution and plumbing applications. As with any new material, some
performance problems have occurred but in each case the underlying causes
have been identified and appropriate correction measures taken. None of
the noted problems suggests any inherent limitations by any of the plastics
piping systems to providing reliable and long-term service for the intended
applications. Enhancement in quality assurance requirements of current
standards and better attention to certain installation practices have
always been satisfactory resolutions to any noted performance inadequacies,
except, of course, for cases of clear misapplication of the product.
Whenever surveys have been conducted comparing the performance of
plastics to traditional piping materials, plastics have always shown up
well. For example, a recent survey conducted by the American Gas
Association indicates that during the year 1981 the reporting gas
distribution companies have experienced leaks in plastic piping at a rate
only 40 percent of that experienced for metallic piping, with no evidence
22
-------
that any significant problems are indicated with currently available
materials.
A survey during 1981 by the American Water Works Association (AWWA)
Standards Committee on Thermoplastic Pressure Pipe revealed no reported
problems with PVC C 900 pipe in water distribution. In an AWWA survey
during 1982 for polyolefin water service pipe, 80 percent of the
respondents replied that their experience with these materials shows them
to be as good as, and more often than not better than, conventional
materials. An investigation of the other 20 percent by a private
consultant disclosed that improper installation and a problem with a
limited production run of one pipe product accounted for a higher incidence
of problems. Both of these causes have been addressed by the addition of
appropriate quality assurance requirements and broader adoption of good
handling and installation practices.
From the first introduction of plastics piping, producers and users
have been alert to identify and correct problems with plastics. This is
evident by the intense activity in such standard writing organizations as
ASTM, AWWA, API, and NSF in enhancing plastic piping product standards by
adopting the lessons of technology and experience to better define the end
product and its proper application. The result is the steadily improving
performance reliability and durability of plastics piping.
The primary obstacle to increased utilization of plastics is not any
inherent lack of capability but rather the exploitation of public and
worker concerns over safety and health risks that are often associated with
synthetic materials. The following allegations have been offered in
support of restricting the acceptance of plastics piping: The possible
leaching of toxic substances that could be present in plastic materials
into drinking water; health risks to the worker and user created by the use
of solvent cements to join plastics piping; health risks by the possible
permeation through the pipe wall of toxic substances that might be present
in the soil in contact with buried piping; fire and accompanying toxic gas
hazards posed by the combustibility of plastics; and decreased labor
utilization because of the greater ease of installing plastics. Obviously,
any concern dealing with public and worker health and safety deserves to be
duly considered and any necessary precautionary measures must be
incorporated into product standardization and utilization. For plastics
such scrutiny and action, which shall be elaborated by others in this
workshop, is a matter of record since the first use of these materials for
piping. For example, the NSF requirements date back to the late 1950s.
All the objective scrutiny that plastics piping has been receiving
since that time points to the conclusion that these materials, when
properly used, present no significant health and safety risks to the public
and to the installer. Consequently, there is no apparent basis for
speculating that the acceptance of plastics in general, or any material
specifically, may be curtailed by water quality/material problems. Rather,
future trends in market share shall be largely influenced by technical
developments in materials and products which will affect their ease of
23
-------
application, performance, and ultimate best effectiveness. To this end
there is considerable activity by a number of plastics piping interests.
There is also considerable activity, as evidenced by this workshop, in
scrutinizing the effects of plastics and other materials on water quality.
Such balanced scrutiny is appropriate and welcomed. In the case of
plastics this might result in revision or additions to current material
requirements which, if needed, can readily be included into an existing
system, an opportunity that does not exist for the other piping materials.
24
-------
IMPACT OF LEAD PIPING AND FITTINGS
ON DRINKING WATER QUALITY
by: Peter C. Karalekas, Jr.
Sanitary Engineer
Region I, Water Supply Branch
U.S. Environmental Protection Agency
Boston, Massachusetts 02203
HISTORICAL PERSPECTIVES
Lead pipe for conveying drinking water has been used for centuries
(1). Among the known advantages of lead are its flexibility, durability,
and long life. For these reasons, lead was used extensively in the United
States beginning in the nineteenth century. Unfortunately, the hazards of
using lead pipe, which were known at the time, did not deter many public
water systems from using the material.
For example, in 1845 a report on water supplies for the city of Boston
(2) concluded that:
Considering the deadly nature of lead poison, and the fact that so
many natural waters dissolve this metal, it is certainly [in] the
cause of safety to avoid, as far as possible, the use of lead pipe for
carrying water which is to be used for drinking. The best substitute
is found in a pipe well coated with pure tin on the interior — such
pipes being quite safe and in every way preferred.
However, this warning was not heeded, and in 1899 Clark (3) reported
that in addition to Boston, 70 other communities in Massachusetts reported
the use of lead or lead-lined service pipes. Commenting at length on
Clark's work, Forbes (4), in an address to the members of the New England
Water Works Association in 1900, concluded that Clark's work showed:
The many causes which make it possible for water to dissolve lead, and
should teach us to look with extreme caution to the use of lead
service pipes. It is also quite a serious matter, from a financial
standpoint, to invest thousands of dollars in lead pipe, and then find
that we have an element of serious danger in our midst which, sooner
or later, must be remedied. The health of a community must first be
considered; this goes without saying.
25
-------
Now, in light of so much conclusive evidence which is within our
reach, what position should we take in this matter in constructing and
maintaining a system of water works? Is it not wiser and better to
eliminate all possible elements of danger from a thing so vital as the
water which we must daily take into our bodies? We must bear this
fact in mind; that ordinary citizens know little about these things,
and trust to those who have charge of the water supplies to furnish
them with a good and safe water and we should not be unmindful of the
confidence placed in us.
Again the warning was not heeded. In a survey of water utilities in
1924, Donaldson (5) reported that 51 percent of the 539 cities surveyed
across the country indicated they used lead or lead-lined services to some
extent.
More recently, two nationwide surveys have demonstrated that lead is
still a significant problem for some water supplies. In a study of 969
water supplies, McCabe et al. (6) concluded that 2 percent of the
population served was receiving lead in excess of the mandatory limit of
0.05 milligram per liter (mg/1). Patterson (7), in a survey of 580 cities,
found that 2.5 percent of the water samples exceeded 0.05 mg/1.
In a more recent survey (8) conducted by the U.S. EPA of 153 of the
largest water systems in 41 states, 112 systems indicated they had used
lead services during some period of time. However, at the time of this
survey only the City of Chicago continues to install lead services, as
mandated by their planning code. Lead goosenecks have also been widely
used and 94 systems indicated their use in the past.
MONITORING FOR LEAD IN DRINKING WATER
Current EPA regulations (9) for monitoring trace metals in drinking
water require that one sample per year be taken for lead at a consumer's
tap. In the author's opinion, a single sample per year is inadequate to
assess the concentrations of lead in drinking water from the corrosion of
lead pipe. Among the variables that have to be considered in a sampling
scheme to detect lead are contact time, water temperature, length of lead
pipe, and water quality. In order to assess the variation in lead
concentration in drinking water that had been standing for varying lengths
of time in piping, it was considered necessary to collect three samples at
each home. Samples were collected by the homeowners who followed the
instructions in Table 1.
Samples were collected at the kitchen sink the first thing in the
morning before any water was used in the home. Sample 1, the interior
plumbing sample, was collected immediately upon opening the faucet. This
represented water that had been standing overnight in the fixture and
interior plumbing serving the faucet. Sample 2, the service line sample,
was collected after the sample collector felt a temperature change in the
26
-------
TABLE 1
SAMPLING INSTRUCTIONS
After 11:00 p.m., do not use the kitchen cold-water faucet
until collecting the water samples the next morning. Using the
following directions, in the morning, collect the water samples
at the faucet before using any faucet or flushing any toilets
in the house. Fill the provided containers to one inch below
the top and put the caps on tightly.
Sample 1: Open the cold water faucet, immediately fill
bottle No. 1 and turn off the water. Recap
this bottle.
Sample 2: Turn the faucet on and place your hand under
the running water, and immediately upon
noticing that the water turns colder, fill
bottle No. 2. Recap this bottle.
Sample 3: Allow the water to run for three additional
minutes and then fill bottle No. 3. Recap
this bottle.
The man from the Environmental Protection Agency will stop at
your house on the morning of , to pick up the
samples. if you do not expect to be home, please leave the
samples outside the front door.
27
-------
water from warm to cold. Since water would be expected to warm slightly
after standing in interior plumbing, this cold water would represent water
that had been standing overnight just outside the foundation of the house
in contact with the interior of the lead service line underground. Sample
3, the water main sample, was collected after the water was allowed to run
for several minutes. This represented water that would have a minimum of
contact time with the service line and interior plumbing.
The author feels that this method of sample collection is a realistic
representation of the range of lead concentrations which the consumer is
likely to find in the water he drinks. It is certainly conceivable that
the first sample of water might be drawn directly by the homeowner in the
morning and used to make juice or coffee, or drunk directly. The second
sample, representing the usually colder water standing in the service line,
could be drawn by a person trying to obtain the coldest water possible to
drink. The third sample, taken after running the water, represents water
that would be found more commonly during the day when water is used
frequently for flushing toilets and cooking.
WATER QUALITY RESULTS
In order to assess the contribution to lead in drinking water from
lead pipe, water samples were collected at consumers' taps supplied through
a lead gooseneck or service line in Boston, Chatham, and New Bedford,
Massachusetts, and Bennington, Vermont.
Although it is recognized there may be other contributors of lead,
such as lead/tin solder, brass or galvanized pipe, or bronze fittings in
the systems, all homes selected for sampling in Boston, New Bedford, and
Bennington had lead service lines and those in Chatham had lead goosenecks,
which are expected to be the major contributors of lead in these systems.
Table 2 lists raw water quality data for various parameters for each
of these systems. It should be noted that all are very low in alkalinity,
hardness, and pH, with lead concentrations in the sources of supply below
the detection limit of 0.005 mg/1.
Table 3 shows the quality of water at consumers' taps after that water
passed through the distribution system, the lead service line or gooseneck,
and the internal plumbing. In each system, significant numbers of samples
contained lead in excess of the Maximum Contaminant Level (MCL) of 0.05
mg/1. The lowest number of samples exceeding the standard occurred in
Chatham, where there were lead goosenecks of approximately 2 feet in
length. The other systems, where there were longer lead service lines
running from the water main in the street to the house, had a greater
proportion of samples in excess of the MCL.
In Bennington, where the water was extremely corrosive, 28 of the 30
samples exceeded the MCL. The percentage of samples containing lead
28
-------
TABLE 2
RAW WATER QUALITY DATA*
pH
Hardness as
CaC03
Alkalinity
as CaCO3
Sodium
Zinc
Copper
Lead
BOSTON
6.5
14
8
5
<0.02
<0.02
<0.005
CHATHAM
6.3
20
3
11.9
<0.02
<0.02
<0.005
NEW BEDFORD
6.0
10
8
6.5
<0.02
<0.02
<0.005
BENNINGTON
5.5
6
2
1
<0. 02
<0.02
<0. 005
*A11 values except pH in mg/1
29
-------
TABLE 3
DRINKING WATER QUALITY AT CONSUMERS'
No. Homes
10
Sampled
No. Samples
No. Samples
Pb >0.05
No. Samples
Pb > 0.005
Avg . Pb
Concentration
Highest Lead
Concentration
pH
TAPS BEFORE CORROSION
BOSTON CHATHAM
14
42 30
28 4
42 20
0.128 0.017
0.870 0.098
6.5 6.3
CONTROL*
NEW BEDFORD
10
30
14
26
0.06
0.26
7.3
BENNINGTON
10
30
28
30
0.158
0.475
5.5
*A11 values except pH in mg/1
30
-------
greater than the detection limit ranged from 66 percent in Chatham to 100
percent in Boston and Bennington, indicating that all water reaching the
tap was picking up some lead. The average lead concentration for all
samples in Boston, New Bedford, and Bennington was above the MCL.
The foregoing illustrates the effect of the corrosive nature of low pH
soft water on various lengths of lead pipe ranging from short, 2-foot lead
goosenecks to service lines as long as 200 feet.
REMEDIAL ACTION
As a result of the identification of the problem of corrosion of lead
pipe and the resulting high lead concentrations in drinking water, Boston,
New Bedford, and Bennington instituted corrosion control in an attempt to
reduce lead concentrations. Table 4 shows the results of corrosion control
in each of these three systems.
In Boston (10), the pH was raised to 8.2 using sodium hydroxide,
resulting in a significant decrease in average lead concentrations and in
the number of samples exceeding the lead MCL. There are still a large
number of samples exhibiting some pickup of lead. A similar situation
occurred in New Bedford, where the pH and carbonate alkalinity were raised
using sodium carbonate (soda ash). Again, significant reductions occurred
in average lead concentrations and in the number of samples exceeding the
MCL. In Bennington, corrosion control involved the addition of a
combination of sodium hydroxide and sodium bicarbonate to raise both the pH
and carbonate alkalinity. This appears to be the most effective treatment
in that the average lead concentration was reduced to the lowest level and
the number of samples with detectable levels was also the lowest. However,
like Boston and New Bedford, there are still a small number of samples
(three in each system) with lead greater than the MCL.
It would appear that the final solution to the problem involves both
corrosion control and systematic replacement of existing lead services over
time, giving priority to the worst-case situations, since corrosion control
alone, while highly effective in reducing lead, does not completely
eliminate the problem.
SUMMARY, CONCLUSION, AND RECOMMENDATIONS
1. Widespread use of lead service lines and goosenecks occurred in
many water systems.
2. Many water systems have less than the optimum treatment for
controlling lead corrosion.
3. High lead concentrations in all communities studied can be found
if water samples are taken in a careful, systematic manner.
31
-------
TABLE 4
DRINKING WATER QUALITY AT CONSUMERS'
TAPS AFTER
ITEM
No. Homes Sampled
No. Samples
No. Samples Pb >0.05
No. Samples Pb >0.005
Average Pb Concentration
Highest Lead Concentration
PH
CORROSION
BOSTON
13
39
3
29
0.020
0.134
8.2
CONTROL*
NEW BEDFORD
8
24
3
21
0.026
0.118
8.5
BENNINGTON
10
30
3
11
0.
0.
8.
014
096
1
*A11 values except pH in mg/1
32
-------
4. The Interim Primary Drinking Water Regulations requiring only one
sample per year are inadequate to identify the presence and
amount of lead resulting from corrosion.
5. The solution to the problem of lead involves systematic removal
of lead materials, and water treatment to reduce corrosion.
6. Although lead services are a major contributor of lead, other
sources such as lead/tin solder must also be considered in any
overall control strategy.
7. The author recommends the use of lead as an acceptable material
for conveying drinking water be removed from plumbing codes.
REFERENCES
1. Frontinus, S.J. 1973. The Water Supply of the City of Rome. New
England Water Works Association, Boston.
2. Report of Commissioners, Appointed By Authority of the City Council to
Examine the Sources From Which a Supply of Pure Water May be Obtained
for the City of Boston. J.H. Eastburn, City Printer, Boston (1845).
3. Clark, H.W. 1899. An investigation of the action of water upon lead,
tin and zinc, with especial references to the use of lead pipes with
Massachusetts water supplies. Annual Report of the Massachusetts
State Board of Health, Boston, Massachusetts.
4. Forbes, F.F. 1900. A very brief discussion of lead poisoning caused
by water which has been drawn through lead service pipe. Journal
NEWWA 15:58-60.
5. Donaldson, N. 1924. The action of water on service pipes. Journal
AWWA 11:649.
6. McCabe, L.J., J.A. Symons, and G.G. Robeck. 1970. Survey of
community water supply systems. Journal AWWA 62:670.
7. Patterson, J.W. 1981. Corrosion in Water Distribution Systems.
Office of Drinking Water, U.S. Environmental Protection Agency,
Washington, D.C.
8. Chin, D., and P.C. Karalekas, Jr. 1984. Lead production utilization
survey of public water supply distribution systems throughout the
United States. U.S. Environmental Protection Agency, Boston,
Massachusetts.
33
-------
9. National Interim Primary Drinking Water Regulations, EPA, Federal
Register (December 24, 1975).
10. Karalekas, P.C., Jr., et al. 1983. Control of lead, copper, and iron
pipe corrosion in Boston. Journal AWAA 75:93.
34
-------
IMPACT OF COPPER, GALVANIZED PIPE,
AND FITTINGS ON WATER QUALITY
by: Chester H. Neff
Principal Scientist
Illinois State Water Survey
605 E. Springfield
Box 5050, Station A
Champaign, Illinois 61820
INTRODUCTION
The stated purpose of this seminar is to review drinking water
problems associated with various plumbing materials. This presentation
will explore the impact of galvanized steel and copper piping materials on
water quality. This is a concern of environmental scientists who are
evaluating the health hazard associated with the corrosion of these
materials in a drinking water system.
For a corrosion engineer, the first instinct is to investigate the
reverse issue; what effect does the water quality have upon copper or
galvanized pipe? Although this is a subtle difference, both viewpoints
have a common focal point. Each is concerned with the chemical or physical
reaction of a metal with its environment, a sample definition of
corrosion. Corrosion engineering is the application of science and art to
prevent or minimize corrosion damage economically and safely. Therefore,
the corrosion engineer will select the best available piping material that
is compatible with the water quality, while considering other factors such
as cost, safety, and regulations. This includes the National Drinking
Water Regulations, which establish Maximum Contaminant Levels (MCLs) for
several metals found in copper or galvanized steel systems.
Copper and galvanized steel have found widespread use in potable water
systems for many years. Experience has shown that both materials have good
corrosion resistance to drinking water when properly selected and properly
installed. Either can be subject to corrosion failures or contribute to
the health hazards problem when abuses occur in designing the system or by
installation procedures. The National Association of Corrosion Engineers
(NACE) has published guidelines (4) for architects, design engineers, and
plumbing contractors to call attention to the corrosion problems in these
systems which they should be aware of.
35
-------
Table 1 presents some of the constituents found in drinking water that
define water quality. Many investigators have reported on the impact of
the various constituents on the corrosion of copper piping materials
(5-11), galvanized steel (12-21), and copper alloys (22-25). With the
exception of lead, only in recent years has the impact of piping materials
upon water quality been investigated to any extent. This presentation will
review recent literature and report recent findings of the author and
coworkers on the subject.
CORROSIVE QUALITIES OF DRINKING WATER
Some investigators have attempted to classify drinking waters
according to their potential to be corrosive to piping materials. Campbell
(5) classified waters by the water source and treatment rather than by
chemical analyses. Types of water recognized by Campbell were: deep well
waters, which are not usually corrosive to galvanized steel but may produce
pitting in copper; upland surface waters, which are soft, have a low pH,
and contain organic matter, and which are aggressive to galvanized steel
and nonaggressive to copper; slow sand filtered surface waters, which
contain a natural organic corrosion inhibitor for copper and are not
corrosive to galvanized steel; and lime-softened waters with a high pH,
which are nonaggressive to copper, but can be aggressive to galvanized
steel in the absence of carbonate hardness. In similar matter, Obrecht and
Myers (17) outlined fifteen categories for potable water, based on the
corrosiveness and the scaling potential of the water supply. Calcium,
silica, sulfate, dissolved oxygen concentrations, and temperature were
employed to categorize the water. Lane and Neff (18) classified Midwest
water supplies into five types, based on hardness, chloride and sulfate;
pH; and alkalinity in a similar manner as Campbell.
Other investigators have attempted to employ predictive corrosion
indices to identify waters that are aggressive. The Langelier Index (26)
is the U.S. Environmental Protection Agency (EPA) recommended parameter for
estimating corrosivity of public water supplies, although many corrosion
engineers also employ the indices developed by Ryznar (27), Larson and
Skold (28), Dye (29), McCauley and Abdullah (30), and others. These
indices have been reviewed by Rossum and Merrill (31) and Singley (32).
The chemical constituents that are commonly cited by investigators as
influencing the corrosivity of a water supply are calcium, alkalinity, pH,
carbon dioxide, sulfate, oxygen, silica, and temperature. The unknown
influences (organics, polyphosphates) and the interaction of all the
constituents are not understood. None of the previously referenced papers
provide information on the solubilization of metals from galvanized steel
and copper pipe or their impact on water quality.
Lead concentrations exceeding the MCL value were reported in the
Boston, Massachusetts, water supply by Karalekas et al. (33) and in
Worcester, Massachusetts, by O'Brien (34) in the 1970s. Those studies,
36
-------
Table 1. Common Constituents Found in Drinking Water
Calcium
Magnesium
Sodium
Potassium
Barium
Strontium
Trace Metals
Oxygen
Carbon Dioxide
Corrosion Products
Unidentified Organics
Dissolved minerals
Hypochlorite
Fluoride
Phosphate
Borate
Hydroxide
Iron
Manganese
Dissolved gases
Hydrogen Sulfide
Nitrogen
Suspended Matter
Mineral Deposits
Miscellaneous
Trihalomethanes
Bicarbonate
Carbonate
Sulfate
Chloride
Nitrate
Silica
Methane
Ammonia
Microbial Organisms
37
-------
along with the development of analytical procedures capable of detecting
trace metal concentrations (e.g., atomic absorption) and the passage of the
National Safe Drinking Water Act, prompted several researchers to study the
impact of plumbing materials on water quality.
METAL CONCENTRATIONS FOUND IN SAMPLES FROM HOUSEHOLD PLUMBING
The metals associated with copper or galvanized steel piping materials
and fittings are zinc, copper, lead, tin, cadmium, and iron. Copper, zinc,
and lead concentrations found in samples taken from household taps are
presented in Table 2. All of the studies (33-39) referenced employed
procedures for determining running and standing sample metal
concentrations. Some of the studies were primarily concerned with lead
concentrations contributed by lead service lines, although the samples were
also exposed to the building plumbing materials.
Lead, copper, and zinc were found in concentrations that exceeded the
MCL value. As anticipated, this occurred most frequently in the standing
samples. Cadmium was not found in significant enough concentrations in
household plumbing to be considered a potential health problem. Tin
concentrations were not commonly reported and are not likely to be a
potential problem in drinking water. Iron concentrations frequently exceed
the MCL value in many water supplies as a natural occurring mineral and are
a factor in corroding galvanized steel plumbing.
Newly installed plumbing materials that have not formed protective
oxide films or mineral deposits are found to have very high zinc, copper,
and lead concentrations when in contact with water for long periods of
time. An example of this phenomenon occurred very inconveniently during a
cooperative corrosion study the Illinois State Water Survey (ISWS) is
conducting with EPA assistance. A corrosion test loop (Figure 1) was
installed with new copper and galvanized steel piping in a public water
supply to simulate normal household water usage. Approximately 40 feet of
galvanized steel pipe was installed by licensed plumbers ahead of
approximately 40 feet of copper tube employing 50 Sn:50 Pb solder
fittings. Chrome-plated brass sample valves, brass shut-off valves,
corrosion specimens, bronze water meter, and a PVC pipe nipple (to separate
materials) were also employed in the test loop. Water flow was controlled
by an electrically operated solenoid valve activated by a timer. The timer
was programmed to allow water to flow 5 to 10 minutes every hour for 16
hours and then to shut down the system for 8 hours. Standing and running
samples were taken after 7 to 8 hours of contact with the plumbing
materials. When this test loop was first placed in operation, the timer
failed and did not operate during the first 6 weeks. Two sets of samples
were collected during the 34-day period in which the test loop was flushed
one time after 15 days to collect running samples. Extremely high iron,
zinc, copper, and lead concentrations were found in both copper and
galvanized steel piping (Tables 3 and 4). Even after 5 minutes of
flushing, the running samples continued to have a high zinc concentration.
38
-------
Table 2.
Location
Metal Concentrations Found in Samples from Household Taps
Concentration (yg/L)
Worcester, MA
Bridgeport, CT
Marborough, MA
Chatham, MA
New Bedford, MA
Seattle, WA
Trondheim, Norway
Netherlands
Standing
No. of sites
9
10
9
10
10
34
18 cities
10 cities
Metal
Pb
Pb
Pb
Pb
Pb
Cu
Zn
Pb
Fe
Zn
Cd
Cu
Pb
Zn
Cd
Cu
Pb
Mean
270
10
14
17
76
450
1740
39
1400
348
—
164
9.7
344
<0.5
647
81
Max.
1900
40
250
98
260
2050
5460
170
5400
2125
8.6
1100
110
650
0.5
1870
180
Running
Mean
0
<5
10
15
13
120
150
5
280
126
—
52
2.7
—
—
—
—
Max.
0
—
—
—
—
1670
1730
22
1200
—
—
—
—
—
—
—
—
39
-------
Water Meter
jf
SIMULATED CORROSION TEST LOOP
Shut-off Valve
Water Supply
t
1/2 in. Schedule 80, Galvanized Steel Pipe, 30 — 40 ft.
Sampling >
Valve 2—1/2 in. Galvanized Steel Corrosion Testers
IW-
1/2 in. Schedule 80, PVC, 6 in. Length
1/2 in. Type L Copper Tube, 30 — 40 ft.
4-
Sampling
Valve 2—1/2 in. Copper Corrosion Testers
Solenoid
Valve
60 min.
Timer
_L
24 hour
Timer
$ 110 VAC
Threaded
•^ Galvanized
Fittings
Solder
Fittings
W.iste
Figure 1
-------
Table 3.
Parameter 15 days, no flow
Standing
Flowing
Pb
Cu
Zn
Fe
Cd
Pb
Cu
Zn
Fe
Cd
* Note:
11 .0
1.687
70.2
14.3
0.006
1.89
0.490
17.4
1.4H
0.001
Galvanized P
Total Metal Concentrations (mg/L) of Samples
from New Copper Plumbing*
19 days, no flow Normal operation
2.24
0.3^3
78.0
7.43
0.330
0.018
31.0
7.94
0.010
0.005
0.175
0.13
0.004
0.004
0.090
0.13
Galvanized Plumbing Precedes Copper Plumbing
Table 4. Total Metal Concentrations (mg/L) of Samples
from New Galvanized Plumbing
Parameter
Standing Pb
Cu
Zn
Fe
Cd
Flowing Pb
Cu
Zn
Fe
Cd
15 days, no flow
0.770
0.880
67.60
18.30
0.009
0.440
0.013
63.2
2.66
0.001
19 days, no- flow
0.710
2.324
11.2
41.30
0.007
0.005
0.15
0.11
Normal operation
0.001
0.003
0.150
0.14
0.001
0.003
0.09
0.12
41
-------
The metal concentrations increased again during the next 19 days of
exposure to the stagnant water in the loop. Normal operation (Tables 3 and
4) refers to the metal concentrations found in samples from the test loops
after the timer was replaced and programmed water flow was established.
This example dramatically indicates that extremely high lead, copper,
zinc, and iron concentrations can occur within newly installed household
plumbing. Such systems may require long flushing periods to reduce metal
concentrations below MCL values. Stagnant water in household plumbing will
frequently develop a bitter taste or turbidity due to the metal
concentrations. A consumer would probably flush such a system until the
water was palatable; however, the author is familiar with some galvanized
steel systems in which taste problems and high zinc concentrations
persisted for over a year after installation.
Plumbing materials do affect water quality. That impact is influenced
not only by the plumbing materials but by water quality, velocity,
temperature, exposed surface area, and system design. Most homes and
buildings are unique when these influences are considered, and the wide
variation of metal concentrations reported in the literature for drinking
water is not surprising.
IMPACT OF BRASS MATERIALS ON WATER QUALITY
The metal concentrations reporeted in Tables 3 and 4 exceed their
predicted solubility, which indicates suspended corrosion products were the
major component responsible for exceeding MCL values. Lead concentrations
were also seen as an indication that lead-tin solder was affecting water
quality. The chrome-plated brass sampling valves were also suspected as a
source of zinc, lead, and copper.
To determine the effect the brass sampling valves were having on
standing metal concentrations, a series of samples were taken from the
brass valve, which was then replaced with a high-density polyethylene
valve. Another series of standing samples were collected until the
corrosion test loop was permanently shut down. Table 5 shows the sudden
decrease in zinc, copper, and lead concentrations that occurred immediately
with the sampling valve change. Copper and lead concentrations were
reduced to near analytical detection limits. Zinc concentration decreased
to the background level contributed by the galvanized steel piping. The
chrome-plated brass sampling valve was determined by these tests to be a
major source of zinc, copper, and lead, even in the nonaggressive water
employed at this site. The sequential collection of 125- milliliter
samples also indicated that the brass sampling valve was influencing the
standing metal concentrations for more than 20 valve volumes of water.
The chrome-plated brass valves were employed in the ISWS-EPA corrosion
study to simulate the household sampling methods employed by the studies in
Table 2. Other brass valves with copper tubing and solder connections and
42
-------
Table 5. Influence of Valve Material on Total Metal Concentrations of Standing Samples
OJ
Zn (mg/L)
Cu (mg/L)
Pb (yg/L)
Date
3/8
4/6
4/14
1/21
5/3
5/14
5/21
6/13
6/22
6/27
7/12
7/16
7/21
8/5
8/15
8/22
Valve material (a)
Chrome-plated
n
n
n
n
it
n
n
n
ti
n
Polyethylene
»
n
n
n
a
b
c
ND
brass 1
1
2
2
2
1
0
1
1
0
0
0
0
0
0
0
— first
.30
.00
.09
.66
.09
.98
.65
.10
.01
.75
.68
.30
.22
.35
.55
.46
125
(b)
ND
ND
ND
ND
ND
ND
ND
0.41
0.42
0.28
0.69
0.20
0.16
0.26
0.22
0.21
mL of
(c)
ND
ND
ND
ND
ND
ND
ND
0.48
0.30
0.26
0.25
0.24
0.14
0.22
0.01
0.17
water
(a)
0.73
0.31
0.18
0.77
0.06
0.07
0.29
0.04
0.06
0.08
0.03
<.01
<.01
<.01
<.01
<.01
drawn
(b)
ND
ND
ND
ND
ND
ND
ND
0.03
0.03
0.02
0.04
<.01
<.01
<.01
<.01
<.01
from
— second 125 mL of water drawn from
~ third
125
mL of
water
drawn
from
(c)
ND
ND
ND
ND
ND
ND
ND
0.04
0.04
0.04
0.04
<.01
<.01
<.01
<.01
<.01
valve
valve
valve
(a)
99
56
59
60
31
75
21
30
19
15
13
1
1
2
1
1
(b)
ND
ND
ND
ND
ND
ND
ND
11
18
5
23
1
1
1
1
1
(0
ND
ND
ND
ND
ND
ND
ND
8
5
4
4
1
1
2
1
1
— not determined
-------
more surface area may influence metal concentrations to a greater extent.
Samuels and Meranger (40) recently reported on metal concentrations leached
from eight types of new chrome-plated brass kitchen faucets by water of
varying quality. Data for three faucets are shown in Table 6. The authors
concluded that copper, zinc, chromium, cadmium, and lead were leached from
the faucets in varying amounts, depending on the type of faucet and the
water quality. Chromium and cadmium in the leachate were not considered
significant when compared to the MCLs for these metals. Werner (16)
observed that considerable quantities of copper and zinc were being
contributed by brass fittings and may be a significant factor in the
corrosion behavior of galvanized steel. These studies confirms ISWS
observations. Any attempt to correlate standing metal concentrations to
the corrosion of galvanized steel pipe, copper tube, lead service lines,
and solder should be preceded by isolating the system with plastic sampling
valves.
Brass or bronze water meters are also suspected to have similar metal
dissolution patterns as brass valves. Water meters are usually installed
in service line or piping near the entrace of the water into the
household. Standing or running metal concentrations may be influenced by
water meters for several system volume changes. The ISWS-EPA study will
attempt to determine if water meters are, in fact, contributing to the
metal concentrations of drinking water.
IMPACT OF MATERIALS VS. AGE OF SYSTEM
Metal concentrations found in samples exposed to stagnant water have
been discussed. This condition occurs only when homes are first built,
during vacation periods, etc. During normal period of water usage, metal
concentrations in samples are found to be much lower. Figure 2 illustrates
the change in zinc concentration with time of exposure. It also
demonstrates the difference in zinc concentration of established galvanized
steel plumbing (site 313) and new galvanized steel plumbing (site 315).
Zinc concentrations exceeded the MCL value for the first 9 months for the
new plumbing and required the entire 24 months of the study to attain
equivalent zinc concentrations found in the established system. In the
older system, zinc concentrations did not exceed the MCL for any sample
during the 24-month study and remained relatively constant for the entire
period. The decrease in metal concentration with time was also observed
for copper, lead, and zinc concentrations at other sites where new plumbing
materials were employed. The standing and running copper concentrations at
site 314, a copper test loop, are illustrated in Figure 3 showing the same
decline in copper concentration as seen for galvanized steel in this
particular water supply. The "spikes" or wide fluctuation in metal
concentrations was observed for all newly installed piping systems.
With the increased contact time of piping materials with water,
corrosion films may become more adherent and cover more surface area to
reduce metal concentrations found in samples. The "spikes" in
-------
Table 6. Total Metal Concentration Leached from Kitchen Faucets
Faucet No
Sample
Source
Raw Water
Filtered Water
Treated Water
Well Water
Fulvic Acid (2
description
Leaching
1st
2nd
1st
2nd
1st
2nd
1st
2nd
mg/L) 1st
2nd
pH
7.4
7-4
6.3
6.3
8.4-8.6
8.4-8.6
8.1
8.1
6.2
6.2
Cu
mg/L
0.69
0.83
1 .82
1 .40
1.03
0.61
0.85
0.28
0.19
0.35
Zn
mg/L
0.59
0.40
2.75
2.25
0.40
0.18
0.57
0.46
0.55
0.29
. 1
Pb
yg/L
4.3
3.4
1.4
1.5
0.7
0.7
0.8
0.4
4.3
7.4
Faucet No
Cu
mg/L
1.68
0.43
0.78
0.41
0.37
0.16
0.43
0.30
0.65
0.26
Zn
mg/L
0.51
0.65
1.15
0.55
0.39
0.10
0.85
0.51
0.37
0.25
. 5
Pb
yg/L
31-0
20.0
55.0
26.0
49.0
15.0
4.1
1.8
110.0
K?.O
Faucet No
Cu
mg/L
0.65
0.41
2.47
0.74
3.33
1.32
9.65
3-54
1.76
1 .01
Zn
mg/L
2.80
1.70
2.55
1.10
4.70
2.85
4.85
2.00
1.32
0.74
. 6
Pb
yg/L
26.0
24.0
26.0
15.0
45.0
24.0
58.0
45.0
18.0
20.0
Re: Samuels and Meranger
-------
SITE 313 AND SITE 315
20-1
18-
16-
14-
0
/ 12-
L
10-
S
t
a
n
d 8-
i
n
g
6-
4-
2-
0-
08/18/81
01/15/82 06/14/92 11/11/82
SflMPLE OflTE
LEGEND: TYPE * * * SITE 313
Figure 2
04/10/83 09/07/83
SITE 315
46
-------
SITE 314
1.3H
1.2-
1.1-
1.0-
0.9-
0.8-
n 0.7-
9
C
u
0.6-1
0.5-
0.4-
0.3-
0.2-
0.1-
0. 0.-
08/18/81 01/15/82 06/14/82 11/11/82
SflMPLE OflTE
LEGEND: TYPE * * * RUNNING
Figure 3
04/10/83 09/07/83
STRNDING
47
-------
concentration values were assumed to be due to loose corrosion products
being transported through the plumbing. Contamination of drinking water by
zinc corrosion products in West Germany and referred to as "Zinkgeriesel"
has been reported by Werner (16). Serious corrosion damage of galvanized
steel has occurred in new buildings in which the corrosion products were
observed. Older buildings did not experience this type of damage.
Irregular zinc coating, faulty weld seam, faulty connections, residues from
construction, brass fittings, and temperature were noted as factors
responsible for "Zinkgeriesel."
NATURE AND SOURCE OF METAL CONTENT
Due to the "spikes" and some extremely high metal concentrations
observed at some test sites, four sites were selected to determine the
nature of metal contamination. A filtering procedure was implemented to
periodically determine soluble metal concentrations in addition to the
total metal concentrations of samples from each site. Tables 7 and 8
illustrate how the metal concentrations are influenced by sampling
sequence, filtration, and plumbing materials. The brass sampling valve
appears to be responsible for the zinc and lead concentrations in the
standing samples, although the galvanized steel test loop has a high
background zinc concentration. The galvanized steel pipe is the major
contributor of zinc which is more than 85 percent soluble at site 315.
Soluble zinc concentrations were observed to range from 1.0 to 2.6
milligrams per liter (mg/1) Zn.
Lead concentrations appear to be due to particulate lead, rather than
soluble lead, with the sources being the brass valves and lead-tin solder
of the copper plumbing. Standing copper concentrations are 17 micrograms
per liter ( g/1) in the galvanized plumbing with higher concentrations of
57 to 370 g/1 observed in the copper plumbing. The copper tube is the
major source of copper at this site, while the contribution from brass
valves is not significant.
The major source of iron, manganese, and calcium concentrations
observed in Tables 7 and 8 is the water supply. Manganese and calcium are
present as a soluble species, whereas iron is present as both particulate
and soluble species. Sodium hexametaphosphate (5 mg/1) is being applied to
this water supply to control iron and manganese and may be influencing the
solubility equilibriums of metals at this site.
Another observation from these tables is that calcium, manganese, and
iron concentrations of standing samples are indicative of the water quality
at these sites 8 hours prior to the time the running samples were
collected. The practice of collecting standing samples prior to collecting
running samples infers that both samples have the same general chemical
characteristics, other than metal concentrations, which is often untrue.
48
-------
Table 7. Metal Concentrations from Copper Plumbing (Site 314)*
Sampling Procedure Total metal concentrations (yg/L)
Date
2/24
4/28
6/28
Flow condition
Standing,
Flowing,
Standing,
Flowing,
Standing,
Flowing,
1st
2nd
3rd
10
10
1st
2nd
3rd
10
10
1st
2nd
3rd
10
10
125
125
125
tnin.
min.
125
125
125
min.
min.
125
125
125
min.
min.
Filtered (0.4 uM) Zn
mL
mL
mL
mL
mL
mL
mL
mL
mL
No
No
Yes
No
Yes
No
No
Yes
No
Yes
No
No
Yes
No
Yes
217
102
74
50
50
180
160
80
30
60
280
230
150
60
30
Fe
72
36
16
1480
830
430
210
170
560
390
2000
1680
380
890
270
Cu
254
97
70
13
12
280
92
63
3
4
370
150
57
7
5
Pb
277
37
5
1
1
86
35
15
2
1
85
15
7
32
7
Mn
85
82
80
224
211
85
84
81
65
80
220
230
180
30
40
Ca
73
72
72
118
118
74
74
73
47
58
106
107
106
35
34
* Note: Galvanized Plumbing Precedes Copper Plumbing
-------
Table 8. Metal Concentrations for Site 315 (Galvanized Steel)
Sampling Procedure
Date
2/24
4/28
6/28
Flow Conditions
standing,
standing,
standing,
flowing, )
flowing, )
standing,
standing,
standing,
flowing, :
flowing, :
standing,
standing,
standing,
flowing, :
flowing, :
1st
2nd
3rd
>10
>1 0
1st
2nd
3rd
>1 0
>10
1st
2nd
3rd
>10
>1 0
125
125
125
min
min
125
125
125
rain
min
125
125
125
min
min
mL
mL
mL
ml.
mL
mL
mL
mL
mL
Filtered
(0.4 pM)
no
no
yes
no
yes
no
no
yes
no
yes
no
no
yes
no
yes
Total Metal Concentrations (pg/L)
Zn Fe Cu Pb Mn Ca
1220 1360 11
1210 1060 17
1040 240 5
60 1470 4
60 590 4
1350 2010
1230 380
1170 230
50 540
30 370
2960 2250
2880 1480
2640 180
90 1440
50 540
7
8
3
3
2
8
8
6
7
6
90 99
61 100
93
73
72
7 93 71
4 229 116
2 203 115
194
19
8
2
2
98
84
80
70
80
.76
74
76
58
63
18 240 107
8 240 106
2 210 107
3 60 38
2 80 45
50
-------
IMPACT OF DESIGN AND INSTALLATION PROCEDURES ON WATER QUALITY
The selection of the proper piping materials is the most important
single factor in designing a potable water system. The experiences of
Campbell (6), Obrecht and Myers (17), Cohen (10), Lane and Neff (18), and
others provides some guidelines for this selection; however, local
experience should weigh heavily on the selection process. Although the
proper material may be selected, poor pipe quality can be responsible for
rapid failure. Galvanized steel pipe in recent years has had frequent
occurrences of defective or inadequate zinc coatings. The quality of
copper tube has been consistent, although carbon films formed in the
manufacturing process have been reported to cause pitting. Piping should
be inspected prior to installation. Perhaps tighter piping specifications
and quality control procedures are required for materials exposed to
drinking water.
Poor plumbing practices or "shoddy" workmanship can also result in
corrosion of plumbing materials. It is not uncommon to observe dissimilar
metals joined together, excessive threading of pipe ends, excessive use of
flux and solder, inadequate reaming of pipe ends, or other manifestations
of poor plumbing practice. Each of these factors can contribute to copper,
zinc, or lead content of potable water, although the quantities contributed
are not known.
A contributing factor to the corrosion of materials is velocity, which
can be controlled by design. Copper tube is more susceptible to the
erosion-corrosion attack by water than is galvanized steel. This type of
corrosion is seen more often in reclrculating water systems which have
oversized pumps but has occurred in undersized copper tube in potable water
systems. Unreamed tube ends may also cause flow disturbances that erode
otherwise protective films in the vicinity of the fittings. It is not
clear whether less adherent corrosion products or soluble copper is being
generated by this corrosion process. Water velocity can influence
corrosion rates and increase the transport of zinc corrosion products in
galvanized steel plumbing.
Grounding of electrical circuits to water pipes in houses has been
reported (41) to cause copper corrosion with resulting blue-water and
copper concentrations of 6.5 mg/1. Corrosion by stray DC currents is well
documented; however, stray AC current corrosion has rarely been
documented. AC current corrosion has been reported (42) to increase with
decreasing frequency; and at 5-Hz AC currents, corrosion is 1 percent of
the equivalent DC current. Reliable information on the subject is not
available; further investigation is warranted.
IMPACT OF WATER TREATMENT PRACTICES
A brief discussion is warranted on the water supply operations and
maintenance program that can increase the corrosivity of the water toward
51
-------
copper or galvanized steel. Chemical treatment programs for the purpose of
corrosion control, we will assume, are reducing the corrosion.
CONTROL OF WATER SUPPLY QUALITY
The water quality of public water supplies has been found to vary
drastically over short time spans during the ISWS-EPA corrosion study.
These fluctuations in water chemistry have been due to equipment failures,
operation procedures, disinfection procedures, treatment change, and
seasonal variations. Public water supplies are dynamic systems where
chemical equilibriums control the solubility of metals and corrosion
mechanisms. Any change in temperature, pressure, or ionic concentration,
in an otherwise stable system, will influence corrosion. One example is a
small community in Central Illinois which employs five wells for a source
of water. Well operation follows a schedule which pumps one well for one
day, another well the next day, and three wells the following day. The
quality of the water from these wells is significantly different, as noted
by sulfate, chloride, and hardness concentration ranges of 77 to 590 mg/1,
51 to 404 mg/1, and 265 to 364 mg/1, respectively. Without elaborating
further, water plant operators in small communities need information and
assistance to improve their awareness of operating practices which may
upset a stable system (if one exists).
CHLORINATION IMPACT
Another significant change observed for water supplies in Illinois
during the last 10 years has been a decision by many water suppliers to
abandon lime-softening and by other water suppliers to change to free
residual chlorine disinfection methods. Economic reasons have probably
been responsible for the change from lime-softening to clarification-only
treatment, while the Illinois EPA has encouraged free residual chlorination
methods. We have observed 2 to 4 mg/1 free residual chlorine in some
supplies in order to maintain minimum residuals in the extremities of the
system. Application of sodium silicate and increased pH reduced the
corrosion that had been caused by high, free chlorine residuals in a copper
system of a state mental health facility. At another facility, pitting
failures of copper heat exchangers were reduced by lowering free chlorine
residuals from 2 mg/1 to 1 mg/1. Increased solubilization of copper by
chlorine, which is pH dependent, has been cited in the literature (43).
POLYPHOSPHATE TREATMENT
The usage of polyphosphates to sequester iron has found widespread
acceptance among small water suppliers, since this alternative is more
economical than other iron removal processes and has the added benefit of
removing tuberculation from cast iron piping. It is this latter ability to
solubilize corrosion products which concerns the author.
Early reports (44,45) on the effect of polyphosphates on lead pipe
concluded that polyphosphates inhibited lead solubility below pH = 7.0 but
52
-------
slightly increased lead solubility above pH = 8.8. One author (45)
concluded: "Since the amount of lead taken up by waters of high pH value
is relatively small, the increase due to the addition of metaphosphate does
not seem to indicate any serious danger to public health, except possibly
in softened waters which employ metaphosphate for recarbonation."
In this paper, the reported standing lead concentrations from lead
pipe exposed at pH = 8.8 increased from 0.33 mg/1 to 0.79 mg/1 Pb as the
metaphosphate dosage was increased from 0 mg/1 to 10 mg/1. To this author,
those lead concentrations appear significant, as are some lead values
observed in the ISWS-EPA study.
Preliminary mean metal concentrations are reported in Table 9 (copper
systems) and Table 10 (galvanized steel systems). Water supplies D and E
contained 1.0 and 4.0 mg/1 polyphosphate (as P04), while water supply A
temporarily contained 0.5 mg/1 polyphosphate. These three water supplies
have been exceeding the lead MCL value seriously at site 314, causing
concern that the polyphosphate may be increasing the solubility of lead
from lead-tin solder or brass fittings. Continued research on the effects
of polyphosphates on plumbing materials is required due to the potential
health hazard of these effects.
SUMMARY
Many studies have shown that copper tubing, galvanized steel pipe, and
brass fittings have increased the lead, zinc, iron, and copper
concentrations in drinking water. Water quality and water treatment will
influence the degree of impact these plumbing materials will have. The
MCLs for these metals have been exceeded most often in standing samples
from household taps. Brass sampling valves may have made a significant
contribution to the metal concentraions reported in many studies. Newly
installed copper or galvanized steel plumbing may require several months to
attain stable, minimum metal values in the drinking water. Stability may
never be achieved in some water supplies due to fluctuating water
chemistry. Disinfection methods and polyphosphate usage may also increase
the solubility of metals from copper or galvanized plumbing.
Continued research is required to identify the sources of these metals
and to control their solubility in drinking water. Basic information is
needed on the nature of protective corrosion films, the influence of
complexing agents on the film, and the water chemistry necessary to form
the films.
REFERENCES
1. EPA. 1976. National Interim Primary Drinking Water Regulations.
EPA-570/9-76-003. U.S. Environmental Protection Agency, Office of
Water Supply, Washington, D.C.
53
-------
Table 9. Mean Metal Concentrations (yg/L) Observed in Copper Plumbing
(approx. 45 samples per site)
Water Plumbing
Site Supply Age CuR CuS FeR FeS PbR PbS ZnR ZnS
301
303
305
307
308
310
312
311
316
A
A
B
B
C
C
D
E
F
new
old
new
old
new
new
new
new
new
37
7
25
8
489
555
258
125
15
267
25
188
37
1327
1277
675
419
287
89
53
60
50
144
164
177
1107
51
215
57
142
53
287
134
759
1408
93
14
1
1
1
4
5
30
455
14
106
3
17
2
42
10
170
2125
183
166
26
24
20
1116
1678
24
207
25
403
85
84
22
2358
4331
202
348
379
318 F new 10 95 56 58 4 70 66 597
R — running samples S — standing samples
54
-------
Table 10. Mean Metal Concentration (yg/L)
Observed Galvanized Steel Plumbing
(approx. 45 samples per site)
Water Plumbing
Age
old
old
new
new
new
old
new
new
new
R — running samples
Site
302
304
306
309
311
313
315
317
319
Supply
A
A
B
C
C
D
E
F
F
CuR
8
3
9
538
737
27
10
4
5
CuS
11
38
89
766
1111
79
39
51
16
FeR
67
54
78
701
244
150
1025
135
70
FeS
109
100
145
882
677
150
3009
150
320
PbR
1
1
1
13
6
4
5
3
3
PbS
6
3
12
17
15
8
213
16
80
ZnR
75
57
90
2214
3341
156
365
251
286
ZnS
134
122
411
5918
9845
418
4359
4376
6685
S — standing samples
55
-------
2. EPA. 1980. National Interim Primary Drinking Water Regulations,
Amendments. Federal Register 45(168):57332.
3. EPA. 1979. National Secondary Drinking Water Regulations. Federal
Register 44(140);42195.
4. NACE. 1983. Prevention and Control of Water-Caused Problems in
Building Potable Water Systems (TPC-7). NACE Publications 52186.
National Association of Corrosion Engineers.
5. Campbell, H.S. 1971. Corrosion, water composition, and water
treatment. Proceedings, Soc. Water Treat. & Exam. 20:11.
6. Campbell, H.S. 1954. The influence of the composition of supply
waters, and especially of traces of natural inhibitor, or pitting
corrosion of copper water pipes. Proceedings, Soc. Water Treat, and
Exam. 8:100.
7. Lucey, V.F. 1967. Mechanism of pitting corrosion of copper in supply
waters. Brit. Corrosion Jour. 2:175.
8. Cruse, H., and R.D. Pomeroy. 1974. Corrosion of copper pipes. Jour.
AWWA 8:479.
9. Obrecht, M.F., and M. Pourbaix. 1969. Corrosion of metals in potable
water systems. Proceedings, 3rd International Congress on Metallic
Corrosion, Vol. IV, Moscow.
10. Cohen, A. 1978. Copper in potable water systems. Heating, Piping,
and Air-Conditioning 5:81,
11. Kenworthy, L. 1943. The problem of copper and galvanized iron in the
same water system. J. Inst. Metals 69:67.
12. Britton, S.C. 1936. The resistance of galvanized iron to corrosion
by domestic water supplies. J. Soc. Chem. Industry 1:19.
13. Kenworthy, L., and M.D. Smith. 1944. Corrosion of galvanized
coatings and zinc by waters containing free carbon dioxide. J. Inst.
Metals 70:463.
14. Cox, G.L. 1931. Effect of temperature on the corrosion of zinc.
Ind. Engr. Chem. 23:902.
15. Slunder, C.J., and W.K. Boyd. 1971. Zinc: its corrosion
resistance. Zinc Institute Publication.
16. Werner, G. 1984. Galvanized Pipes in House Installations,
Zweckverband Landeswasserversorgung, West Germany.
56
-------
17. Obrecht, M.F., and J.R. Myers. 1973. Potable water systems in
buildings: deposit and corrosion problems. Heating, Piping, and
Air-Conditioning, May.
18. Lane, R.W., and C.H. Neff. 1969. Materials selection for piping in
chemically treated water systems. Materials Protection 8(2):27.
19. Gilbert, P.T. 1948. The corrosion of zinc and zinc-coated steel in
hot waters. Sheet Metal Industries, October.
20. Werner, G., E. Wurster, and H. Southeimer. 1973. Corrosion tests of
galvanized steel pipe by the Ground Water Supply Administration.
GWF-Wasser/Abwasser 114.
21. Bachle, A., E. Dessiner, H. Weiss, and I. Wagner. 1981. The
corrosion of galvanized and unalloyed steel pipes in drinking water of
different hardness and neutral salt content. Werkstoffe und Korrosion
32.
22. Geld, I., and McCaul, C. 1975. Corrosion in potable water. Jour.
AWWA 10:549.
23. Turner, M.E.D. 1961. The influence of water composition on the
dezincification of duplex brass fittings. Proc. Soc. Wtr. Treat.
Exam. 10.
24. Wormwell, F., and T.J. Nurse. 1952. The corrosion of mild steel and
brass in chlorinated water. J. Appl. Chem. 2, Dec.
25. Larson, T.E., R.M. King and L. Henley. 1956. Corrosion of brass by
Chloramine. Jour. AWWA 48, Jan.
26. Langelier, W.F. 1936. The analytical control of anti-corrosion water
treatment. Jour. AWWA 28:1500.
27. Ryznar, J.W. 1944. A new index for determining amount of calcium
carbonate scale formed by water. Jour. AWWA 36:472.
28. Larson, T.E., and R.V. Skold. 1958. Laboratory studies relating
mineral quality of water to corrosion of steel and cast iron.
Corrosion 14(6):285.
29. Dye, J.F. 1952. Calculations of the effect of temperature on pH,
free carbon dioxide, and the three forms of alkalinity. Jour. AWWA
44(4):356.
30. McCauley, R.F., and M.O. Abdullah. 1958. Carbonate deposit for pipe
protection. Jour. AWWA 50:1419.
31. Rossum, J.R., and D.T. Merrill. 1983. An evaluation of the calcium
carbonate saturation indexes. Jour. AWWA, Feb.
57
-------
32. Singley, J.E. 1981. The Search for a Corrosion Index. Jour. AWWA
73(11):179.
33. Karalekas, P.C., C.R. Ryan, C.D. Larson, and F.B. Taylor. 1978.
Alternative methods for controlling the corrosion of lead pipe. New
Eng. Water Works Assoc. 92(2) :159.
34. O'Brien, J.E. 1976. Lead in Boston water: its cause and
prevention. New Engl. Water Works Assoc. 90 (2): 172.
35. Zoeteman, B.C.J., and B.J.A. Haring. 1978. Introduction of Chemical
Compounds into Drinking Water During Distribution.. Ryksinstituut Voor
Drinkwatervoorziening Mededeling 78-6.
36. Stegavils, K. 1975. An investigation of heavy metal contamination of
drinking water in the City of Trondheim, Norway. Bull. Environm.
Contam. and Toxicol. 14(1) :57.
37. Haring, B. J.A. , and B.C.J. Zoeteman. 1980. Corrosiveness of drinking
water and cardiovascular disease mortality. Bull. Environm. Contam.
Toxicol. 25:658.
38. Sharrett, A.R., A. P. Carter, R.M. Orheim, and M. Feinlieb. 1982.
Daily intake of lead, cadmium, copper, and zinc from drinking water:
The Seattle study of trace metal exposure. Environmental Research
28:456.
39. Sharrett, A.R., R.M. Orheim, A. P. Carter, J.E. Hyde and M. Feinleib.
1982. Components of variation in lead, cadmium, copper, and zinc
concentration in home drinking water: The Seattle study of trace
metal exposure. Environmental Research 28:476.
40. Samuels, E.R., and J.C. Meranger. 1984. Preliminary studies on the
leaching of some trace metals from kitchen faucets. Water Res.
41. Guerrera, A. A. 1980. Grounding of electric circuits to water
sources: one utility's experience. Jour. AWWA 2:82.
42. Shreir, L.L. 1963. Corrosion. Vol. 2, Corrosion Control. John
Wiley and Sons.
43. Atlas, D., J. Coombs, and O.T. Zajicek. 1982. The corrosion of
copper by chlorinated drinking waters. Water Res. 16:693.
44. Hatch, G.B. 1942. Inhibition of lead corrosion with sodium
hexametaphosphate. Jour. AWWA 33:85.
45. Moore, E.W., and E.E. Smith. 1943. Effect of sodium
hexametaphosphate on the solution of lead. Jour. AWWA 34(9):1415.
58
-------
SUMMARY OF IMPACT
OF METALLIC SOLDERS ON WATER QUALITY
by: Norman E. Murrell
H2M Consulting Engineers
125 Baylis Road, Suite 140
Melville, New York 11747
BACKGROUND
An incident of lead poisoning of a dentist's son in Smithtown, New
York, prompted the consumer's request for testing of his home water
supply. Water as a source of lead was discounted at first because the
Water District had a 15-year record of lead-free distribution samples.
Nevertheless, first-draw (zero to one minute) sampling showed lead in
excess of the Drinking Water Standard of 50 micrograms per liter (wg/1).
After three minutes of flow, sampling showed water well under the standard
limit.
Lead solder was suspected as the source of the lead. Every home in
the 8-year-old subdivision was tested. The two occupied homes with high
first-draw lead levels both had recent plumbing additions using lead
solder. A new home under construction had an even higher first-draw lead
value of 7,100/«g/l.
Consulting engineers H2M/Holzmacher, McLendon & Murrell, P.C.,
conducted this testing and, at the same time, initiated an American Water
Works Association (AWWA) literature search for studies of lead solder.
Prior studies in the United States, Canada, and Europe showed high lead
values in first-draw water after periods of non-use. The lead values were
higher in water systems with low pH and soft water. Further studies were
then conducted by H2M in Nassau and Suffolk counties, New York, to
investigate the occurrence of lead leaching from new and old lead solder
joints.
PRELIMINARY INVESTIGATIONS
In the Smithtown Water District, at the site of the original
complaint, a time series test was run after a 4-hour period of nonuse with
a first-draw value of 300/fg/l. It took about 48 seconds for the lead
59
-------
value to drop below the drinking water standard. In the home under
construction, the first-draw lead value was 25,000 g/1 and, after 80
seconds of flushing, the lead value still exceeded the drinking water
standard. In a Melville office building with 23-month-old plumbing, first
draw after a 71.5-hour shutdown (long weekend) was 120 g/1. The highest
level (200 g/1 at 1.5 minutes) was calculated to coincide with the
location of lead-soldered joints within the plumbing system.
Continuing the investigation, more than 200 first-draw tests were
conducted by H2M in Nassau and Suffolk counties — a 1,200-square-mile area
dependent on naturally low pH and generally soft groundwater for its water
supply. There were many results above the drinking water standard. County
health departments, the 13 towns in the bi-county area, and water suppliers
were informed of these findings. Independent testing by the Suffolk County
Water Authority, Suffolk County Health Department, Nassau County Health
Department, and Town of Hempstead Water Department have confirmed H2M's
findings (see Table 1).
In January 1983, a meeting was held with U.S. Environmental Protection
Agency (EPA) personnel in Cincinnati, Ohio, to discuss the investigations
held to date and to suggest further research focusing on the variables of
age of plumbing and the pH and hardness of the water supply. The South
Huntington Water District agreed to participate and a proposal for a
cooperative study was submitted to EPA in March 1983. The "Lead Solder
Aging Study" was approved by EPA to commence October 1, 1983.
EPA/SOUTH HUNTINGTON WATER DISTRICT COOPERATIVE STUDY
As of May 1984, 90 sites have been identified and selected for this
study, the South Huntington Water District portion of the low pH sampling
has been completed, and much raw data has been gathered but not yet fully
analyzed. Ten homes were selected in each of nine age groups from zero to
20+ years old. These sites would be sampled three times at different pH
levels: less than 6.2, 7.2, and 8.2. The homes were randomly selected to
obtain a geographic distribution in the 19.4-square-mile service area of
the District. The type of existing solder was verified in each home
through scrapings of an exposed solder joint and testing by atomic
absorption spectrophotometer. Of 95 homes so tested, only one had less than
0.50 percent lead in the solder; 67.3 percent of the homes had lead content
ranging from 55 to 65 percent in the solder.
First-draw samples were tested for copper and cadmium in addition to
lead. An additional sample was drawn for determination of various water
quality parameters including pH. Copper values above the 1 milligram per
liter (mg/1) Secondary Maximum Contaminant level (up to 7.77 mg/1 at 6.9 pH
and 0 amps) were found. These sites were checked for stray electric
currents in the water service pipe to determine whether or not the high
copper values are caused through grounding of electric systems to the water
plumbing system.
60.
-------
TABLE 1
NASSAU COUNTY HEALTH DEPARTMENT TESTS
LOCATION
Locust Valley
Port Washington
Manorhaven
Woodbury
North Port Washington
North Hills
North Hills
LEAD
(yg/D
17,000
4,400
3,500
2,900
930
750
530
pH
7.0
6.8
6.8
7.3
7.0
6.7
6.7
HARDNESS
(mg/1)
41
49
49
64
49
23
23
% LEAD
SOLDER
60.3
61.7
47.9
58.4
50.3
56.2
60.0
AGE
( YEARS )
0
1
0
0
0
0
0
TOWN OF HEMPSTEAD WATER DEPARTMENT TESTS
WATER DISTRICT
East Meadow
East Meadow
Uniondale
East Meadow
East Meadow
Roosevelt Field
Roosevelt Field
East Meadow
DTI TT T^TKir1
tSU XljJL/lINvj
AGE
8 months
8 months
3 years
8 months
8 months
6 months
6 months
3 months
TTMTT nPT?r>T3P
ij.ri.nj Ddn\Jt\£i
FIRST DRAW
8 hours
24 hours
8-10 hours
8-10 hours
24 hours
12 hours
12 hours
24-48 hours
LEAD
0 MIN
2010
1500
1049
810
760
569
338
186
(yg/D
1 MIN
7
240
5
5
380
385
5
174
61
-------
During low pH sampling of 64 first-draw tests for cadmium, only one
sample at 42.5x*g/l was above the drinking water standard. Since the
second sample at 10 seconds contained only 3x^g/l cadmium, we assume the
presence of cadmium was caused by the faucet.
FIRST-DRAW LEAD FINDINGS
Of the first-draw lead samples taken, 61.84 percent have been above
the drinking water standard. The results ranged from a low of 2.0/«*g/l in
a 1968 home at pH = 5.9, to values ranging from 500 to l,200/«g/l (pH
ranges from 5.6 to 6.9) at 10 homes constructed between 1981 and 1983.
However, the highest lead value (l,300>t«g/l) was obtained at pH = 5.9 in a
1968 home.
With only the initial findings at a low pH, there appears to be a
correlation between the age of the lead solder and its ability to leach
lead into the drinking water. Ignoring the first zero-second flush and
allowing the first 300 milliliters (ml) of flush for the faucet's leaching
of lead, the second sample at 10 seconds was compared for 75 homes. For
homes constructed in 1980, 1981, 1982, and 1983, the second 125-ml sample
(taken in March and April, 1984) indicated that 23 homes, or 70 percent,
exceeded the 50- g/1 drinking water standard, and 10 homes were less than
this maximum contaminant level. For homes constructed between 1955 and
1979, only seven, or 17 percent, exceeded the drinking water standard at a
low pH with 35, or 83 percent, less than this maximum contaminant level.
pH AND HARDNESS
Additional investigations have been undertaken to determine locations
outside Long Island with water conditions that may be conducive to lead
leaching. Research throughout the world indicates that leaching of lead is
most acute in areas with low pH and soft water. This condition occurs not
only on Long Island, but in many locations in New York State and throughout
the nation. In New York State, 2,802 samples from 784 community water
systems have been gathered as part of the Water Resources Investigation of
the Chemical Quality of Water from Community Systems conducted by the U.S.
Geological Survey and the New York State Department of Health. Review of
this data shows many locations in New York City and throughout the state
with pH and hardness values which may be conducive to lead leaching.
HEALTH EFFECTS OF LEAD
Adverse effects of lead on infants have been well documented,
including anemia, colic, encephalopathy, nephropathy, and neuropathy.
Reduced I.Q. has been associated with high lead levels. It has been
recommended in at least two research studies of lead in drinking water that
the safe value of lead for infants should be reduced from 50x«g/l to
25x*g/l. First-draw testing is important since young children might
62
-------
normally take a drink of water in the middle of the night without proper
flushing. Infants might also get that lead-saturated first draw of water
in their liquid concentrated formula.
LEAD SOLDER
The random sample of homes selected for the South Huntington study
showed 67.3 percent of test sites have solder with a lead content between
55 and 65 percent. This tends to confirm that, during the past 20 years,
the predominant solder used in home plumbing in this area was lead solder.
Tin/Antimony and Tin/Silver Solder
Alternatives to lead solder include tin/antimony solder and tin/silver
solder. Recent wholesale costs per pound for these three solders in New
York State have been: tin/lead, $4.50; tin/antimony, $8.00; tin/silver,
$20.00. An estimate of one pound of solder per new home construction is
probably high.
The melting range of tin/silver and tin/antimony solder is slightly
higher than that for most tin/lead solder, 473°F for 95/5 tin/silver and
464°F for 95/5 tin/antimony, as opposed to 460°F for 40/60 tin/lead
solder. Tin/antimony solder has been tested for leaching in a prior U.S.
EPA/Seattle study which determined that the maximum leaching of antimony
was 3.7>6«g/l after 27 hours of standing. The U.S. EPA/South Huntington
study found less than 4/c/g/l leaching of antimony after overnight non-use
at a low pH.
Legislative Impacts of the Research
In New York State, the lead content of solder could be controlled in
one of three ways. The first is the State Building Code, which now permits
each of 931 towns and 553 villages to ban lead solder if the County Health
Department determines it a health hazard. The second method would be
through the State Plumbing Code, which has considered a ban on lead solder
for 18 months, but has taken no action. The third avenue is through state
legislation. Such legislation has been introduced, but no bill has yet
been reported out of committee. Based on the research conducted to date,
H2M has recommended both state legislation and local initiatives to limit
lead content in solder to 0.50 percent. In monitoring for compliance, we
have found a few tin/antimony solders that contain 0.30 percent and 0.35
percent lead. The Nassau County Health Department even found one with 0.50
percent lead. Based on our local findings, in order to assure compliance
without undue hardship on plumbers, we recommend that state and federal
legislation set the maximum lead content in solder at 0.50 percent.
On Long Island, 10 towns serving a population of 2,334,329 have
already moved to amend their plumbing codes in order to ban lead solder,
and some have instituted monitoring programs to test solder for its lead
content in order to verify that plumbers are adhering to the new code.
63
-------
CONCLUSION
Leaching of metallic solders, particularly lead solder, appears to be
affected by many factors:
• pH of water.
• Plumbing workmanship.
• Hardness of water.
• Time since last use of water.
• Age of solder.
• Percentage of lead in solder.
Where these conditions are conducive to lead leaching, large numbers
of consumers are exposed to high-lead, first-draw wa.ter. If all evidence
points to a need to ban lead solder — why not now?
64
-------
IMPACT OF LEACHING BY PLASTIC PIPE, FITTINGS.
AND JOINING COMPOUNDS
by: Thomas Podoll
SRI International
333 Ravenswood Avenue
Menlo Park, California 94025
The California Department of Housing and Community Development (DHCD)
is proposing to expand the use of plastic plumbing pipe in residential
construction. The principal change with regard to water quality is to
allow polybutylene (PB) and chlorinated polyvinyl chloride (CPVC) pipe in
hot and cold potable water supply systems inside dwellings. SRI
International, under contract to the DHCD, prepared a document in March
1983 entitled "Environmental Review of Proposed Expanded Uses of Plastic
Plumbing Pipe" (1). This environmental review provides an overview of
existing information about the environmental impact of the new applications
of plastic pipe and identifies areas where better information is needed.
This presentation focuses on one part of this review, namely, the leaching
of CPVC and PB plumbing.
OVERVIEW
The impact of leached chemicals can depend on either the instantaneous
concentration of the chemicals in the water or the cumulative exposure to
varying concentrations over time. Thus, we would ideally like to know
concentrations of leachates as a detailed function of time after
installation of plumbing systems. Figure 1 shows a decreasing
concentration with time as the chemicals in the pipe are depleted or
immobilized. Different chemicals would show more or less rapid declines
over time after installation, depending on their properties and their
initial distribution in the pipe.
The general pattern, however, is only a large-scale picture; many
events disturb the smooth trend. Concentrations build up when water stands
in the pipe for a period of time ("dwell time"), for example, overnight or
during a vacation. When the system is flushed, concentrations drop to the
background levels of the incoming water supply. The variation over a day
or two might look something like that in Figure 2.
65
-------
8
I
INSTALLATION
REPLACEMENT
TIME
HA-4910-9
FIGURE 1 LONG-TERM PATTERN OF DECLINE IN CONCENTRATION
PARTIAL RINSING FOLLOWING
A WEEKEND AWAY
BATHROOM TAP FOLLOWING
EVENING BATHS
FIRST WATER DRAWN
NEXT MORNING
TYPICAL LONG-TERM
LEVEL OF RISE TIME
AFTER INSTALLATION
TIME
HA-4010-10
FIGURE 2 SHORT-TERM VARIATIONS IN CONCENTRATION
66
-------
LEACHING MODELS
If water stands in a pipe, the concentration of a specific chemical in
the water will be determined by the leaching rate, or flux, of the chemical
integrated over time. The leaching rate, in turn, will be determined by
the amount and distribution of the chemical in the pipe, the diffusion of
the chemical in the pipe and in the water, the equilibrium partitioning of
the chemical between the pipe and the water, and the diffusion of the
leachate out of the pipe into the air.
The differential equations that describe mass transfer in the pipe and
in the water are given by Pick's second law of diffusion. The solution of
these equations depends on the initial and boundary conditions for the
pipe/water system. Two relatively simple cases for pipe leaching appear to
be reasonable. In the first case, a surface film of leachate covers the
inner surface of the pipe and diffuses into an initially pure water phase.
In the second case, the leachate is initially distributed uniformly in the
pipe and diffuses into an initially pure water phase.
In Model 1 the leachate is concentrated at the pipe/water interface.
The leachate equilibrates with the water relatively rapidly, and the amount
of leachate in the water depends on the equilibrium partitioning of the
leachate between the surface film and the water (described by Hf in Figure
3) and on the number of equilibrium dwell periods that water has stood in
the pipe. Figure 3 simply shows that as H' decreases (where the leachate
becomes more soluble in water and less soluble in the pipe material), the
concentration in the dwell water decreases more rapidly with the number of
equilibrium dwell periods.
In Model 2 the leachate is initially uniformly distributed in the
pipe. Equilibrium is not attained during a dwell period, and the amount of
leachate in the water in a given dwell period will be diffusion controlled
and over a long elapsed time will diminish with a square-root-of-time
dependence. Figure 4 shows the concentration versus time profile for this
model, assuming daily sampling and the concentration in units of the
initial sample concentration.
Unfortunately, few data exist on pipe/water partitioning coefficients
or plastic pipe diffusion coefficients and, because these coefficients are
likely to be very sensitive to specific chemical/pipe interactions, they
are difficult to predict. Although these models cannot be used to predict
leaching from plastic pipes, they are useful for evaluating and
understanding experimental data. This is particularly important for
evaluating CPVC leaching data because of the complex composition of CPVC
and the use of solvent cement to join the piping.
67
-------
NUMBER Of DWELL «KIODS (nl
F,GURE 3 CONCENTRATOR LEGATE ,N WATER
ACCORDING TO MODEL 1
100 "<>
TIME <*V>)
300
280
FIGURE 4 CONCENTRATION VERSUS T.ME PROFILE FOR MODEL 1
A»umt dV.lv ampling .nd conc.ntrit.on in unit, of initnl
i«r\plt conctntrition.
68
-------
PLASTIC PIPE COMPOSITION
The composition of plastic pipes should be known to determine possible
leachates. Nonproprietary specifications supplied by manufacturers of
plastic pipe and the National Sanitation Foundation (2) are summarized in
Table 1. PB is composed primarily of the resin polymer and contains small
amounts of pigments (such as titanium dioxide, carbon black, or talc) and a
small amount of antioxidant (such as Irganox 1010). The composition of
CPVC is more complex and requires a solvent cement for joining pipes.
Potential health problems associated with toxic chemicals leached from
joined CPVC pipes are accordingly more difficult to assess than potential
water-quality problems with PB. Therefore, most of the remainder of this
presentation focuses on CPVC leaching.
QUALITY ASSURANCE
There are several important reasons for closely examining the quality
of plastic pipe leaching data:
1. The composition of pipe leachate water Is likely to be a complex
mixture (particularly for CPVC) of organic compounds in very
dilute concentrations.
2. Many of the potential leachates of plastic pipe are commonly
found in drinking water at measurable background concentrations.
3. Much of the work on plastic pipe leaching has been performed by
parties that had institutional interest in the question of pipe
water quality.
Excellent QA guidelines for plastic pipe leaching studies are detailed
in the Interim Report by the National Sanitation Foundation entitled
"Proposed Organohalide Leachate Testing Protocol for Plastic Piping."
CPVC LEACHING
The most important reports to date on CPVC leachability are those by
James M. Montgomery (3) and Boettner et al. (4).
James M. Montgomery, Consulting Engineers, Inc., under contract to the
California Department of Health Services, measured the water leachability
of PVC and CPVC pipes. The testing protocol consisted of static,
simulation, and kinetic tests. The static test was designed to estimate
the concentrations of chemicals in water expected during initial
occupancy. The kinetic test was designed to estimate the rate of leaching
so that the concentration of the chemical in the water could be predicted
as a function of time. The simulation test was designed to estimate
leachability during typical home use.
69
-------
TABLE 1
COMPONENTS OF PLASTIC PIPE SYSTEMS
PRODUCT
FUNCTION
PB
PERCENT
CPVC
PERCENT
Pipe and
fittings
Solvent
Cement
Primers
Resin
Impact modifiers
Stabilizers/
antioxidants
Lubricants
Pigments/fillers
UV stabilizers
Solvents
Resin/compound
Pigments
Solvents and
pigments
Poly(butene-l) >98
None
Irganox 1010 <0.5
None
Titanium dioxide <2.0
Carbon black <0.5
Talc <2.0
None
None
Chlorinated PVC >80
ABS >3
Methyl-methacrylate- <15
butadiene-styrene-
alpha-methyl styrene
Organotins <3.5
Oxidized polyethylene <1.5
wax
Titanium dioxide <5.0
Carbon black <0.05
Others <0.1
Tetrahydrofuran
Cyclohexanone 80-90
2-Butanone (methyl
ethyl ketone)
N,N-dimethyl formamide
CPVC or CPVC pipe <20
compound
Titanium dioxide <0.5
Carbon black
Others
Same as for solvent
cement
-------
The major limitations of this study were a lack of replicate
experiments and the consequent inability to quantify the statistical
validity of the data. In spite of these and other limitations, the
Montgomery report does give a semiquantitative picture of CPVC leaching.
Data compiled from the static and simulation tests indicate that the
following organic chemicals were consistently found in significant
concentrations in pipe water:
CEMENT SOLVENTS VOLATILE ORGANICS
MEK Dichloromethane
THF Carbon tetrachlorlde
Cyclohexanone Tetrachloroethene
DMF Trichloroethene
Toluene
The cement solvents were found in the parts-per-million range, but the
amounts found appear to diminish rapidly, as expected, with rinsing. The
volatile organics (except toluene) were typically found in the pipe water
in the 1- to 10-ppb range for systems that approximated normal initial use
of plumbing systems. Toluene was typically detected under 1 ppb.
Chloroform may leach from the pipe; however, the data are inconclusive.
Two of the tests found chloroform at levels no different from those in
controls (kinetic and simulation), but it was found at levels marginally
significantly different from those in controls in the static tests.
For the static tests, no correlation of concentration with dwell time
or elapsed time was observed for the low-molecular-
weight halogenated organics that were found in significant amounts. The
cement solvents, however, show a definite decrease in concentration during
the first few rinses of the cemented pipe. The absolute magnitude of these
concentrations is probably unrealistic, even for initial use of a plumbing
system, because of the low number of void volume rinses.
For the kinetic tests, no significant correlations between dwell time
and concentration could be deduced for the volatile organics. Conversely,
the concentrations of the cement solvents appeared to increase with
increasing dwell time up to a limiting equilibrium value. Moreover, these
concentrations fell significantly during the refill dwell kinetic
experiments.
The work of Boettner et al. (1981) was initiated to develop an
analytical method for detecting organotin stabilizers in water. In the
process, useful information on CPVC leachability was obtained.
This report confirmed the findings of the Montgomery report on the
leachability of cement solvents. High initial concentrations of these
solvents were found in the pipe water and these concentrations decreased
71
-------
with elapsed time. Boettner et al. also detected the presence of organotin
leachates in CPVC pipe water. The concentrations of these leachates were
initially in the 0.5- to 3-ppb range for 1-day dwell periods, but fell
rapidly to less than 0.1 ppb per day after 3 weeks elapsed time.
It is interesting to note that the concentrations of the organotin
leachates appeared to diminish in a biphasic manner with time; that is, the
concentrations diminished, then increased, then diminished again. This
biphasic behavior suggests two types of leaching in cement-joined pipes.
The initial leaching of organotins occurs from areas of the pipe where
solvent cement has not been applied. This leaching diminishes rapidly as
the surface concentration is depleted. In the areas where solvent cement
has been applied, there is no organotin present initially, but within a few
days, organotin species begin to diffuse through the cement/pipe surface
layer into the water, and the leachate concentration correspondingly rises.
These data suggest a possible explanation for a lack of correlation
between leachate concentrations of noncement solvents and time that was
described in the 1980 Montgomery report. If solvent cements were applied
in differing thicknesses to the joints of the test systems, it would be
expected that the leaching of pipe constituents through the cement to the
water would be retarded in varying degrees. Thus, while the leaching rate
in one part of the system may have been diminishing, the leaching rate in
another part of the system may have been increasing. For a plumbing system
with several joints, the variability of the leaching rate with elapsed time
over a period of several weeks may well have been sporadic, as observed.
SUMMARY OF PLASTIC PIPE LEACHING DATA
CPVC leaching data indicate that:
• Pipe cement solvents (MEK, THF, cyclohexanone, and DMF) are
initially leached at very high concentrations, but these
concentrations diminish rapidly with repeated flushing.
• Low-molecular-weight chlorinated organics (dichloromethane,
carbon tetrachloride, tetrachloroethene, and trichloroethene) are
leached in the low (1-10) ppb concentration range during initial
dwell periods. The lack of correlation of the concentrations of
these leachates with dwell time or elapsed time is probably
caused by the relative uncertainty of the measured values (which
were typically within a factor of 5 of control values) and by the
cumulative effect of leaching of these organics at different
rates from areas covered or not covered with solvent cement.
• Chloroform is a possible leachate. More work is required on
chloroform leaching and on the leaching of chemicals other than
the dominant solvents from the solvent cements.
72
-------
POLYBUTYLENE LEACHING
The leachabllity studies of PB were generally of lower quality than
those of CPVC. However, a study by Shell (5) indicated the presence of
Irganox 1010 derivatives in leachate waters. Irganox 1010 is an
antioxidant that is present in PB at less than 0.5%.
REFERENCES
1. SRI International. 1983. "Environmental Review of Proposed Expanded
Uses of Plastic Plumbing Pipe." Prepared for California Department of
Housing and Community Development, Menlo Park, California.
2. National Sanitation Foundation. 1983. "Proposed Organohalide
Leachate Testing Protocol for Plastic Piping," Interim Report.
3. James M. Montgomery, Consulting Engineers, Inc. 1980. "Solvent
Leaching from Potable Water Plastic Pipes," Final Report. Prepared
for the Hazard Alert System, California Department of Health
Services/Department of Industrial Relations, James M. Montgomery,
Consulting Engineers, Inc., Pasadena, California.
4. Boettner, E.A., G.L. Ball, Z. Hollingsworth, and R. Aguino. 1981.
"Organic and Organotin Compounds Leached from PVC and CPVC Pipe," U.S.
Environmental Protection Agency, Cincinnati, Ohio.
5. Shell Chemical Company. 1982. "Analysis of Water Leachates of
Polybutylene (PB) Pipe Supporting the Conclusion that the Use of PB
Pipe for Domestic Water Service Is Without Measurable Hazard," Shell
Chemical Company, Houston, Texas.
73
-------
NSF STANDARD AND CERTIFICATION PROGRAM
FOR PLASTICS PIPE: STANDARD 14
by: Nina I. McClelland
President and Chief Executive Officer
National Sanitation Foundation
3475 Plymouth Road
Ann Arbor, Michigan 48105
The National Sanitation Foundation (NSF) is a private, not-for-profit
corporation, chartered in 1944 under Michigan law. Its mission is to
develop and administer programs relating to public health and the
environment in areas of services, research, and education. Through its
headquarters and laboratories in Ann Arbor, Michigan, a wastewater
equipment testing facility in Chelsea, Michigan, and regional offices in or
near Los Angeles (Upland) and Sacramento (Davis), California; Ann Arbor,
Michigan; Philadelphia (Chalfont), Pennsylvania; Atlanta, Georgia; and
Dover, England, NSF's programs and services reach throughout the United
States and into 25 foreign countries.
For operational purposes, programs are grouped into Listing,
Certification, and Assessment Services. Listing Services programs are
those for which there is an NSF standard, authorization to display the
appropriate seal or logo, and a published product Listing. Certification
Services programs involve standards other than NSF's official regulations,
or specifications. They include authorization to display the certification
mark, and a published product Registry. All other programs and special
studies are provided through Assessment Services. A controlled use report
is provided for special studies; the identity mark is the company logo.
Each of these programs is available for domestic and international
application.
A highly qualified professional and technical staff — engineers,
chemists, microbiologists, and environmental scientists — is retained by
NSF in state-of-the-art facilities. Consulting faculty from the University
of Michigan are retained for toxicological, radiological, and other
expertise, as required. Recent capital acquisitions focus on laboratory
and information processing resources intended to optimize quality and
timeliness of all operations. A $3.25 million laboratory expansion program
(current) attests to the high performance and growth objectives to which
the Trustees and staff are fully committed.
74
-------
Plastics piping system components have been tested at NSF since 1955,
first in a comprehensive special study of leachate and toxicological
testing (1), then under the provisions of Standard 14, Plastic Piping
System Components and Related Materials. This Standard includes products
for potable water; drain, waste, and vent; and other plumbing systems
applications.
Ingredients used in products listed for potable water must comply with
rigid acceptance and qualification procedures prior to listing. These
requirements are summarized in Figure 1.
With an application for listing, a manufacturer provides complete,
detailed chemical identity of all ingredients in the formulation. This
information is reviewed by staff, and held under terms of confidential
disclosure. For ingredients generally regarded by the Food and Drug
Administration (FDA) as safe, or sanctioned previously for food contact,
e.g., listed in Title 21, Code of Federal Regulations (21CFR), no further
toxicological testing is usually required. For nonsanctioned ingredients,
Ames and animal feeding studies must be provided. The animal data must
include established no-effect, effect, and intermediate levels of the
proposed ingredient, using protocols and a testing laboratory accepted by
NSF in advance of the testing.
Chemical leachate testing is required for all proposed new
ingredients, using pipe and compound formulated to contain the ingredient
at two times the level proposed for maximum use. Chemical leachate
(extraction) testing for inorganic contaminants is performed consistent
with the testing protocol in Standard 14; i.e., exposure to "formulated
water" at pH 5.0 + 0.2 at 37 degrees (°C) for periods of 24, 24, and 72
hours. (Cold-water end-use applications are assumed; for hot-water
applications, exposures are 1 hr at 82° C, 1 hr at 82° C, and 72 hr at 37°
C.) Fresh water is added following each of the 24-hour exposures,
providing data for three separate extractions. The first exposure is
assumed to simulate worst- case end use, where the user would ingest the
first, very aggressive water drawn from a new installation; the third
exposure represents water which could be ingested as the first draw of
water left standing over a weekend. The protocol for organic chemical
leachate testing is similar except that pH 8.0 + 0.15 is used for the
exposure. ~
The chemical parameters monitored and their maximum permissible limits
(MPLs) are shown in Figure 2. (Currently, five volatile organic compounds
proposed by EPA for regulation are also being monitored to acquire a data
base. They are CC14; TCE; PCE; 1,2-dichloroethane; and 1,1,1-tri-
chloroethane.) For proposed new ingredients, the MPL for the first
exposure (MPL-1) is set at ten times the established, third-exposure MPL
(MPL-3). MPL-3 is equivalent to maximum contaminant levels (MCLs) in the
National Interim Primary Drinking Water Regulations (NIPDWR) for all
regulated chemicals included in Standard 14. By policy, all chemicals
regulated by NIPDWR are included in Standard 14 when pertinent to the
listed products.
75
-------
Acceptance1
(New on.
Qualification1
(New Vn.odu.ci on. Change.
in
Monitoring2
Disclosure statement with
complete chemical identity
Product and compound tested
at 2 X maximum recommended
use level
Chemical leachate testing'
1st exposure _<10 X MPL*
3rd exposure
-------
A special test protocol has been developed and validated for organics
leachate testing. Nineteen samples of polyvinylchloride (PVC) and
chlorinated polyvinylchloride (CPVC) were used for validation testing. To
date, none of the data suggests that the listed products contribute
significant levels of trihalomethanes to contacted water.
To confirm compatibility with other ingredients in a product recipe,
an accepted new ingredient must also be "qualified" by additional chemical
leachate testing. Products formulated by the ingredient user (pipe,
fittings, or appurtenance manufacturer) at the maximum proposed use level
are tested in accordance with the procedure described for acceptance
testing.
Following acceptance of an ingredient and qualification of product,
listing is authorized. The appropriate logo is then displayed on the
product; i.e., NSF-pw, -we, for potable water/well casing applications;
NSF-dwv, -tubular, for drain, waste, and vent/continuous waste systems;
NSF-cw, for corrosive wastes; NSF-sewer, for sewer main applications.
Standard 14, adopted in 1965, is revised and updated on a continuing
basis. In 1978, a requirement for residual vinyl chloride monomer (RVCM)
was added. By modeling and leachate testing, it was established that
levels to 10 parts per million (ppm) RVCM in the wall of pipe or fittings
would not leach detectable levels of the monomer to water exposed to
product (where "detection" is 2 ppb). Results of RVCM monitoring
experience show that none of the 460 samples tested in 1982 or 472 samples
tested in 1983 exceeded the established MPL of 10 ppm.
From the outset, standards development at NSF has included
representatives at all levels of government, the affected industry, and
users of the subject products or services. Typically, the request for a
standard may originate with any of the three sectors. Interest and
commitment are established in an exploratory meeting. A small task
committee is appointed to draft the document. The draft is reviewed and
revised or accepted by the Joint Committee (JC), where each sector —
regulatory, industry, and user — has voting representation. The Joint
Committee proposes the standard to the Council of Public Health
Consultants, a group of 36 public health professionals with no industry
representation. The Council recommends a standard to the Board of
Trustees, where final adoption occurs.
All NSF standards have a requirement for periodic review at intervals
not to exceed five years. The formal, ongoing review and revision or
reaffirmation process is accomplished through established Joint Committee
policies. All voting is in compliance with procedures described in the
Office of Management and Budget (OMB) Circular A119 (2).
Three annual inspections are required for all plastics production
facilities included in the listing. These are minimums and apply to
foreign and domestic sites. Nonconformance and other problems result in
77
-------
additional inspections. At the plant, products in production and inventory
are checked to verify consistency with previous design and testing
records. Use of accepted ingredients or compounds and in-plant quality
assurance are verified; and selected evaluations may be performed (e.g.,
dimensioning products). Seven hundred and eleven (711) inspections are
projected for 1984. Violations identified either in testing or inspection
result in various actions, ranging from corrective measures to delisting.
There are three general categories of regulation — official,
self-regulation, and third-party. Third-party programs have established
credibility world-wide. They effectively and appropriately place cost
burdens with the private sector. Their objectivity and long-term success
are a matter of record and a source of pride! We look forward to expanded
opportunities with plastics plumbing systems components and related
products through voluntary consensus standards and third-party
certification programs like NSF Standard 14.
REFERENCES
1. Tiedeman, W. and N.A. Milone. 1955. A Study of Plastic Pipe for
Potable Water Supplies, NSF, June 1955.
2. Office of Management and Budget. 1982. OMB Circular A119, Federal
Participation in the Development and Use of Voluntary Standards, Vol.
47, November 1, 1982.
78
-------
EVALUATION OF THE PERMEATION OF ORGANIC
SOLVENTS THROUGH GASKETED JOINTED AND
UNJOINTED POLY(VINYL CHLORIDE), ASBESTOS
CEMENT, AND DUCTILE IRON WATER PIPES
by: James P. Pfau
Group Leader, Coating Science
Polymer Science and Technology Section
Battelle Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
and
Donald Goodman
Manager, Polymer R&D
Tenneco Polymers, Inc.
R.D. 6, Box 177
Flemington, New Jersey 08822
SUMMARY*
Phase one of the Vinyl Institute's research program, "Evaluation of
the Permeation of Organic Solvents through Gasketed Jointed and Unjointed
Poly(vinyl chloride), Asbestos Cement, and Ductile Iron Water Pipes," has
been completed. Phase one involved the exposure of commercially available
4-inch water pipe joints to three neat organic solvent environments
(toluene, hexane, and 1,1,1-trichloroethane) for a period of 6 weeks. Each
pipe specimen contained demineralized water under a pressure of 40 pounds
per square inch gauge (psig) and throughout the 6-week exposure period,
samples of water were removed from each pipe and analyzed for traces of the
appropriate organic liquid.
The objective of this program was to study when and how permeation of
selected solvents occurs through the jointed and unjointed pipes and to
develop a hypothesis for the mechanism of permeation. The tests results
indicate that the mechanism for fastest breakthrough (permeation) is
through the gasketing material. The unjointed pipe showed slower
permeation or no permeation compared to the jointed pipe. When permeation
*Copies of original work are available on request to: Dr. Roy T.
Gottesman, Executive Director, Vinyl Institute, 355 Lexington Avenue, New
York, New York 10017.
79
-------
did occur with the jointed pipe, it happened much earlier than with the
straight unjointed pipe. Another mechanism for breakthrough is based on
the composition of the pipe. The moreporous asbestos/cement unjointed pipe
showed early permeation with toluene and 1,1,1-trichloroethane, and the
unjointed PVC pipe swelled to permeation with toluene after day 37 of the
42-day test.
The initial work was performed under "worst case model" conditions.
The breakthrough of neat solvents in the jointed test specimens indicates
that the joints of all three materials are susceptible to permeation. The
second phase of the program, expsoure to aqueous solutions of the same
organic solvents, appears to be warranted in order to determine permeation
under more realistic conditions. Consideration should be given in
characterizing the rubber compounds used in the gaskets. Differences in
the gaskets used with each type of pipe may explain the apparent
differences in permeation, particularly with regard to hexane.
CONCLUSIONS
1. Permeation of neat solvents is related to type of solvent, type
of pipe, whether the pipe is jointed or unjointed, and time of
exposure.
2. For all the organic solvent/pipe combinations investigated, the
jointed pipes showed organic solvent breakthrough (permeation).
The sole exception was the ductile iron/hexane, although the
jointed ductile iron pipe demonstrated permeation with toluene
and 1,1,1-trichloroethane.
3. The unjointed pipe showed slower permeation or no permeation
compared to the jointed pipe. Also, when permeation did occur
with the jointed pipe it happened earlier than with the straight
unjointed pipe.
4. The permeation data indicate that the mechanism for fastest
breakthrough is through the gasketing material. The gasketing
material used to seal the jointed pipes is more susceptible to
organic solvent (toluene, hexane, and 1,1,1-trichloroethane)
permeation than is the pipe material.
5. Another mechanism for permeation is based on the composition of
the pipe. The more porous asbestos/cement unjointed pipe showed
early permeation by toluene and 1,1,1-trichloroethane. The
unjointed PVC slowly softened to point of permeation with toluene
between sampling days 37 and 41, and the unjointed iron pipe did
not show permeation.
6. The order of aggressiveness (i.e., attack on the pipes and
gaskets) for the organic solvents from greatest to least is
toluene > 1,1,1-trichloroethane > hexane.
80
-------
7. Toluene permeated all jointed pipe sections. Toluene did not
permeate the ductile iron unjointed pipe, and the PVC unjointed
pipe showed breakthrough after day 37 of the 42-day test.
8. 1,1,1-Trichloroethane permeated all jointed pipe sections. It
did not permeate the PVC straight pipe or the iron straight
pipe. The asbestos straight pipe showed permeation at about 2
weeks.
9. Hexane permeated PVC jointed pipe and asbestos/cement jointed
pipe. Iron jointed pipe and all straight pipes showed no
permeation. The presence of a dark oily residue in the jointed
PVC pipe substantiates the conclusion that the gasketing material
has been attacked. The absence of permeation with jointed iron
pipe suggests that this gasket is not solvated by hexane, whereas
it was attacked by toluene and 1,1,1-trichlorethane in parallel
tests.
10. Entry of organic solvent into the demineralized water against 40
psig internal pressure is corroborating evidence that the
mechanism is permeation through gasketing material or pipe and
not hydraulic flow. Constant internal pressure does not exclude
contamination by permeation.
81
-------
TREATMENT OR WATER QUALITY ADJUSTMENT TO ATTAIN MCLs
IN METALLIC POTABLE WATER PLUMBING SYSTEMS
by: Michael R. Schock
Associate Chemist
Aquatic Chemistry Section
Illinois State Water Survey
6050 E. Springfield
Box 5050, Station A
Champaign, Illinois 61820
INTRODUCTION
Virtually all of the common household or building plumbing materials
— copper tubing, galvanized steel, lead pipe, brass fittings, and tin/lead
solder — will oxidize and dissolve to some extent in potable waters. The
byproducts of these corrosion reactions can be of concern because of
aesthetic or toxicological reasons; hence, the development (1,2) of maximum
contaminant levels (MCL) or of secondary (2) maximum contaminant levels
(SMCL) for lead, cadmium, copper, zinc, and iron. Additionally, corrosion
is economically costly, because of decreased service lifetimes of piping
materials, loss of water during distribution, and decreases of hydraulic
efficiency.
Traditionally, corrosion control measures taken by water utilities
have been aimed at preventing tuberculation and perforation of cast-iron
distribution mains, eliminating "red water" complaints, and protecting
dissolution of cement-mortar-lined or asbestos-cement pipe. Treatment
processes designed to alleviate some or all of these distribution system
problems may or may not effectively limit the corrosion of the different
types of metallic materials used in the service lines and interior systems
of buildings and dwellings.
The purpose of this paper is to examine the impact of several common
treatment approaches on the formation of corrosion byproducts from common
building plumbing materials. The treatment approaches covered in this
paper will be: calcium carbonate saturation, pH adjustment, pH plus
carbonate adjustment, orthophosphate addition, polyphosphate addition, and
silicate addition. First, however, a background will be developed to aid
in the understanding of the feasibility of different corrosion control
approaches.
82
-------
SOLUBILITY AND MASS-TRANSPORT PROCESSES
Unlike most distribution system mains, household and building systems
frequently have large sections of piping in which the water can be stagnant
for hours. During these long periods of standing, corrosion reactions can
continue to add byproducts to the water until a state of solubility
equilibrium is nearly or actually attained. Thus, the metal levels
observed in the water are initially governed by the mass transport of
oxidizing or solubilizing agents to the pipe surface, and of the corrosion
products away from the surface. The upper-limiting case, from a dissolved
byproduct standpoint, is that ultimately provided by the solubility
equilibria of the surface pipe material in the water. Physical factors,
such as nonadherence of the corrosion product film or the dislodging of
scales by normal turbulent flow, can add particulate metal species to the
water, increasing the observed "total" metal concentration. Also,
corrosion reactions can continue, even if the solubility equilibrium is
attained.
Clearly, there are numerous tradeoffs in corrosion control programs.
For instance, the minimization of metal levels in order to attain MCLs may
not result in the formation of an effective barrier to continuing metal
attack, shortening the lifetime of the plumbing material. It may merely
convert the metal into an adherent corrosion product while surface attack
and/or pitting continues. Alternatively, an absence of bulky corrosion
products and tubercles on iron lines brought about by treatment chemicals,
such as polyphosphates, might result in increased levels of soluble iron
through complexation.
In general, the strategy of corrosion control through water quality
adjustment involves interference with at least one link of the chain of
interconnected anodic and cathodic corrosion cell reactions (4).
Ordinarily, this consists of the formation of a "barrier" of a "passivating
film" that both limits the transport of the metallic species into solution
(mostly through solubility limitation), and limits the diffusion of the
oxidizing agents (normally dissolved oxygen or chlorine species) to the
pipe surface. These complicated couplings of redox and solubility
reactions, plus the differences in relative rates of the reactions, prevent
simple relationships between corrosion rates determined from pipe coupon or
insert data and solubility levels from being developed.
An example of the role of mass transport in governing the metal level
in nonequilibrium systems has been elegantly developed by Kuch and Wagner
(5) for lead. The approach is generally suitable for other types of pipe,
as well. These authors show that the level of lead in a water undergoing
turbulent flow can be predicted by knowing the lengths and diameters of the
lead pipes, the observed (or computed) concentration of lead at solubility
equilibrium, the volume flow rate of the water, the water temperature, and
the diffusion coefficient of the dissolved species. For water under
stagnating conditions, the parameters that need to be known to compute the
83
-------
lead level after a given stagnation time are: the diameter of the pipe,
the diffusion coefficient (approximated by that for Pb2+), the maximum
equilibrium lead concentration, and the lead concentration during flushing
(computed from the previous model). In hard waters, the mass transfer
coefficient may also need to be included.
FORMATION OF PASSIVATING FILMS
Because the in situ formation of tight, adherent films that are
impervious to ionic or oxidant migration is virtually impossible, a
discussion of the solids that can precipitate in pipes and produce a limit
for metal solubility is in order. Only common building plumbing materials
— lead, copper, new galvanized steel, and brass — will be considered here.
Table 1 gives a list of solid corrosion products that have been found
that are likely to be significant in controlling the solubility of the
different pipe or coating metals. Each metal behaves uniquely, and these
solids are largely responsible for determining whether or not a corrosion
control technique will work. Other deposits have been reported, but
systematic and comprehensive surveys in many water qualities at different
temperatures have not been done. Solids containing sulfate and chloride
are often common in saline waters, waters with a high ionic strength, or in
the presence of pits in copper tubing (6). Other deposits may also locally
form, but they may not act as solubility controls.
Detailed research has not been performed to identify the films of
corrosion products or inhibitor precipitates formed on tin/lead solder or
brass fittings under different water quality conditions. Until that time,
we must assume formation of the same products as would be present on pipes
of similar material.
The relationship of passivating film formation to water quality
adjustment and inhibitor dosage will be covered for galvanized steel
(zinc), copper, and lead in the next section. To assist in predicting
orthophosphate inhibitor effectiveness and the effect of pH and alkalinity
(through the inorganic carbonate content), solubility calculations were
performed in a similar manner to those reported previously for lead (7-10)
and zinc (10-11). A new computer program was created for calculation of
copper (II) solubility. The equilibrium constants used are reported in
Tables 2 and 3. Considerable uncertainty exists for much of the
thermodynamic data, and all computations were restricted to 25°C because
temperature effects on the equilibrium constants are largely unexplored.
An ionic strength of 0.0005 moles per liter (mol/L) was assumed as a
general approximation. The pH interval of 7 to 11 was considered, as well
as total inorganic carbonate (TIC) concentrations of from 5 to 500
milligrams calcium carbonate per liter (mg CaC03/L) (0.05-5.0 mmole
C/L). Only orthophosphate, hydroxide, and carbonate solids with a
documented history of occurrence in plumbing systems were included; hence,
84
-------
Table 1 . Scale Minerals That May Be Important in
Limiting the Solubility of the Indicated Metals and Metallic
Materials in Most Potable Water Systems
Galvanized Steel Pipe
Basic Zinc Carbonate (hydrozincite, Zn^OHjgtCCb^)
Zinc Hydroxide, Zn(OH)2, a-, &-, 5-, and e- forms
=-Hopeite, ZngCPCty^'^HgO
Zinc Oxide (Franklinite) , ZnO or ZnFe20ij
Hemimorphite, ZnjjSi2C>7(OH)2'H20 (hot water systems)
Copper Pipe
Cuprite, Ci^O
Tenorite, CuO
Basic Copper Carbonate (Malachite),
Brochantite,
Atacamite,
Lead Pipe
Basic Lead Carbonate (hydrocerussite),
Lead Carbonate (cerussite),
Hydroxypyromorph ite,
Plattnerite, Pb02
85
-------
Table 2. Reactions and Log Equilibrium Constants (6) at 25°C
and I - 0 mol/L Species Considered in Solubility Diagram Construction
M - Pb, Cu or Zn. All constants are for I - 0 and 25°C
log B
M2+
M2+
M2+
M2+
2M2+
2M2+
3M2+
4M2+
6M2+
M2+
M2+
M2+
M2+
M2+
+ H20
+ 2H20
+ 3H20
+ 4H20
+ H20
+ 2H20
+ 4H20
+ 4H20
+ 8H20
+ C032~ + H20
+ C032-
+ 2C032-
+ P0i,3- + H*
+ P0ji3~ + 2H+
0 MOH+ + H+
e M(OH)2° + 2H+
0 M(OH)3- + 3H+
0 M(OH)n2- + 4H+
0 M2OH3+ + H+
e M2(OH)22* + 2H+
0 M3(OH)j,2^ + 4H+
e MjjCOHjj,1** + 4H
0 M6(OH)8J4+ + 8H+
«> MHC03*
O MC03°
e M(co3)22-
0 MHPOn0
O MHoPOn*
Pb
-7.23
-16.93
-28.10
-39.7
-6.36
—
-23.88
-20.88
-43.61
—
7.1
10.3
15.45
21.1
Zn
-8.96
-16.9
-28.4
-41.2
-9.0
—
—
—
—
11.73
5.2
7.5
15.66
21-16
Cu
-7.93
-16.23
-26.9
-39.6
—
-11.21
-22.05
—
—
12.41
6.8
9.8
16.36
21.33
86
-------
Table 3. Equilibrium Constants at 25°C and I - 0 mol/L for the
Solids Considered in the Exploratory Solubility Calculations
Pb(OH)2(s) + 2H+ e Pb2+ + 2H20
PbC03(s) O Pb2+ + C032~
Pb3(C03)2(OH)2(s) + 2H+ O 3Pb2+ + 2C032- + 2H20
Pb5(POij)3OH(s) + H+ e 5Pb2+ + 3POn3~ + H20
CuO(s) + 2H+ e Cu2+ + H20
Cu(OH)2(s) + 2H+ O Cu2+ + 2H20
Cu2(OH)2C03(s) + 2H+ O 2Cu2+ + C032~ + 2H20
CuC03(s) O Cu2+ * C032-
Cu3(POj4)2«2H20(s) O 3Cu2+ + 2POn3~ -t- 2H20
e-Zn(OH)2(s) + 2H+ O Zn2+ + 2H20
ZnC03(s) O Zn2+ + C032~
Zn5(OH)6(C03)2(s) + 6H+ e 5Zn2+ + 2C032~ + 6H20
+ l|H20
87
-------
the absence of solids such as Pb3(P04>2 and Pb3(OH)5Cl. An exception was
CU3(PC>4)2 * 2H20, which was the only copper orthophosphate for which data
was available.
The results of some of the computations are presented in Figures 1
through 5, which will help show trends in solubility behavior of three of
the trace metals of interest. Even though computations went up to 2 mg
P04/L, the diagrams are shown only for an orthophosphate concentration of
0.5 mg P04/L, which is a common dosage. For copper (II), no
precipitation of orthophosphate was predicted. These diagrams show
"contour lines" for the metal solubility, as both pH and TIC are varied. ^
three-dimensional version of Figure 1 for lead (II) is shown in Figure 6,
for comparison. The units of "log mg/L" were chosen to produce smooth
contour line intervals and good detail in the diagrams. Reference will be
made to these figures in the following section.
TREATMENT APPROACHES
CALCIUM CARBONATE SATURATION
This is the approach traditionally followed by utilities with hard and
moderately hard water supplies. Also, many utilities who use lime
softening for clarification also use this strategy. Monitoring the calcium
carbonate saturation state was also included as one of the requirements of
the U.S. EPA "Corrosivity" regulation (1). Many expositions on the
rationale for CaC03 saturation are available, but the primary intent can
be summarized by saying that the objective has been to prevent
tuberculation and "red water" in iron distribution mains by covering the
pipe with a thin and adherent layer of calcium carbonate. However,
research by Stumm in the 1950s and 1960s (12, 13), and more recently by
primarily German scientists (14, 15), indicates that the effective
corrosion control mechanism involves the formation of a dense siderite
(FeCC>3) scale layer, and is also related to the buffer intensity of the
water and the Fe(II)/Fe(III) ratios in the pipe deposits.
Numerous indices are available for the estimation of the calcium
carbonate saturation state, and they have been thoroughly reviewed
elsewhere (16-18). The concept of using a calcium carbonate film to "seal"
the surface of a copper, lead, or galvanized pipe in household or building
systems in order to prevent metal leaching has been widely assumed to have
been validated. However, virtually no research has been published to
define the water quality conditions (i.e., pH, temperature, calcium, and
carbonate concentrations) or the time necessary to form the film, or to
examine the metal leaching behavior while film formation is in progress.
For zinc and copper, Figures 3 and 5 indicate that concentrations below the
SMCLs should remain easily attainable at pH values above 7.5 for most
conceivable TIC levels, in the absence of complete surface coverage by the
calcium carbonate deposit. Obviously, the thickness and homogeneity of the
88
-------
PH 9
100
150 200 250 300 350
mg CaC03/L Inorganic 003
400
450
500
Figure 1. Contour diagram of lead (II) solubility in the
system Lead (II)-Water-Carbonate at 25°C and an ionic
strength of 0.005 mol/L. The MCL is -1.30 (0.05 mg
Pb/L). The points marked "L" and "H" indicate local low
and high points. The contour interval is 0.05 log units,
89
-------
pH 9
50 100 150 200 250 300 350
mg CaC03/L Inorganic C03
400 450 500
Figure 2. Contour diagram of lead solubility in the presence
of 0.5 mg P04/L at I = 0.005 mol/L and 25°C. The MCL =
-130 on this scale, which is: Log (mg Pb/L) x 100. The
contour interval is 0.05 log units.
90
-------
pH 9
i 11 i i 111 i ii i i i it 11 ii M 11 111 n i i i i i it i M i i it i i 11 i i M i i i i 111 i i i i i i i i i M i i i i i 1111ii i iiit 11(1111111
5 50 100 150 200 250 300 350 400 450 500
mg CaCC>3/L Inorganic
Figure 3. Contour diagram of zinc solubility in the system
Zinc-Water-Carbonate at 25°C and I = 0.005 mol/L. The
SMCL is 0.699 (5 mg Zn/L), and the contour interval is
0.05 log units.
91
-------
PH 9
50
100
150 200 250 300 350
mg CaCO3/L Inorganic CO3
400
450 500
Figure 4. Contour diagram of zinc solubility in the presence
of 0.5 mg P04/L at I = 0.005 mol/L and 25°C. All points
lie below the 5-mg Zn/L SMCL (0.699 in diagram units).
The contour interval is 0.05 log units. Lines were
removed in the lower left to avoid crowding.
92
-------
pH 9
JT\ L>-r| IJ-n i I 1->TI i J-I-*-<-TTJ J.IJIII-IIII-- • _ -fc-fc
5 50 100 150 200 250 300 350 400 450 500
mg CaCO3/L Inorganic C03
Figure 5. Contour diagram of copper (II) solubility in the
system Copper (II)-Water-Carbonate at 25'C and I - 0.005
mol/L. All points lie below the SMCL of 1 mg/L (0.00 in
diagram units). The contour interval is 0.1 log units.
93
-------
-1.4
600
Figure 6. Three-dimensional solubility diagram for lead (II)
corresponding to the conditions of Figure 1.
94
-------
galvanized surface layer is of concern, because even at modest zinc
solubilities the layer could be perforated if not "sealed" in time by the
calcium carbonate coating.
In contrast to zinc and copper, lead (and soldered joints in copper
pipe containing lead) could cause a serious problem (Figure 1). Assuming a
pH range of 7.5 to 8.5 and a TIC concentration of approximately 200 to 400
mg CaC03/L (carbonate alkalinity of approximately 100 to 200 mg
CaCC>3/L), the dissolved lead concentration could easily attain 0.16 to
0.25 mg/L upon overnight standing. Flowing samples could readily approach
the MCL of 0.05 mg/L, given a sufficient length of pipe. Of additional
concern is the fact that fresh leaded solder would generally be exposed
immediately after house or building construction, or copper pipe repair.
Leaded solder has also been noted for connections of plated brass household
faucet assemblies (19). In some communities, lead is still either allowed
or specified (20) for new service lines of small diameters (approximately
0.75 to 2 inches).
There are three special practical considerations in implementing a
control program based on calcium carbonate saturation that should be
noted. While the Langelier Index has been included as a parameter to be
monitored in the EPA drinking water regulations (1), numerous researchers
have pointed out that two waters of the same Langelier Index would not
necessarily have the same potential mass of calcium carbonate to deposit
(13, 18, 21). This value, the "calcium carbonate precipitation potential"
(CCPP), is really the most enlightening index, and deserves to replace the
Langelier Index as a control parameter.
The second consideration is that the identity of the calcium carbonate
phase that precipitates is frequently the more soluble aragonite form,
rather than calcite, the latter being the assumed phase in almost all
calculations. Exactly what pipe, temperature, and water quality conditions
govern this phenomenon have not been investigated to any great degree. To
further complicate matters, a third form, vaterite, has been observed in
some hot water systems (22). Log solubility constants vary from -7.91 for
vaterite to -8.48 for calcite at 25°C and an ionic strength of zero. The
solubility of each different form of CaC03 also varies in its response to
temperature changes.
The third consideration is the fact that the most widely used values
for the solubility constant of calcite at different temperatures have been
shown to be in considerable error by studies over approximately the last
decade (10, 23). Revised values based on the latest research have been
presented (10, 23), and the widely used method of "total dissolved solids"
correction originally developed in 1942 (24) has been revised (25).
PH ADJUSTMENT
Figures 1, 3, and 5 show that for a simple system consisting of the
metal, water, and carbonate, the solubilities of lead (II), copper (II),
95
-------
and zinc can all be greatly reduced by increasing pH into the range of 9 to
10. In this pH range, in cold or room-temperature waters, the solubilities
are controlled by hydrocerussite (Pb3(C03)2(OH)2), tenorite (CuO),
and probably hydrozincite (Zn5(C03)2(OH)6). Lead is particularly
sensitive to both pH and TIC, while pH is the dominant factor for copper
(II) and zinc at TIC levels above approximately 20 mg CaC03/L.
While hydrocerussite and tenorite form tight and adherent scale
layers, the films formed on zinc may not be as stable. Bachle et al.
suggest that the corrosion products in the cover layer have no
transport-inhibiting effect (26). Several studies have indicated
conversions of an initially precipitated zinc oxide or hydroxide to a basic
carbonate, and have indicated dehydration and phase changes in hot water
systems (27-29). Changes in the morphology of the corrosion product of
zinc with pH and TIC warrant continued investigation, in order to attempt
to prevent removal of the thin galvanized layer, even while dissolved zinc
remains below the SMCL level of 5 mg Zn/L. Essentially, no data is
available for the dependence of complex formation and solubility constants
of important lead (II), zinc, or copper (II) solids on temperature.
PH AND CARBONATE ADJUSTMENT
This treatment approach has been extremely successful in the case of
lead, both in laboratory (8, 9) and field studies (30). Many waters
contain sufficient carbonate to help limit lead (II) solubility, if only
their pH is increased. A low alkalinity does not necessarily mean a low
carbonate content, particularly when the pH is below 7. Because lead (II)
forms strong carbonate complexes, only waters of very low carbonate
concentrations are amenable to this treatment. Waters with high
alkalinities and neutral to basic pH values may have to have carbonate
removed to enable the attainment of the MCL.
Copper (II) also forms strong carbonate complexes, and its solubility
is increased by carbonate supplementation in the pH range of pH 7 to 10.5.
Copper (II) solubility remains below about 0.01 mg/L above pH 8.5, so
maintenance of the SMCL should not often be a problem. Of considerable
concern, however, is the possibility of rapid pitting (generally called
"Type I" pitting) and perforation of the pipe wall. No consensus has been
reached on the causes and mechanisms of pit propagation. Mattsson and
Fredericksson (6) examined numerous corrosion products from systems with
pitting, and concluded that in waters high In sulfate and chloride, basic
copper (II) chloride and sulfate salts formed, rather than just tenorite
and malachite. They suggested the ratio of bicarbonate to sulfate, or
chloride, or both, could be important, but the effect might be
pH-dependent. In this case, carbonate supplementation in addition to pH
adjustment should prove to be beneficial.
Except for the pH range of 9 to 10.8 at high TIC concentrations, and
the extremely low TIC region ( 10 mg CaC03/L) in the pH range of 8 to
10.5, zinc solubility is relatively Insensitive to TIC. Thus, the benefit
96
-------
of carbonate supplementation might be analogous to the concept of calcium
carbonate saturation control, wherein an additional mass of carbonate might
tend to form a more effective film. There is no published experimental
verification of this, however, and an improvement in corrosion resistance
with continually increasing TIC should not be presumed. Here, there may be
a difference in behavior as the underlying steel alloy layer becomes
exposed, and the corrosion reactions of iron take over.
ORTHOPHOSPHATE ADDITION
In 1970, Murray (31) developed a corrosion inhibitor chemical from a
mixture of zinc sulfate, monosodium orthophosphate, and sulfamic acid
(31). He stated that the mechanism of corrosion inhibition was the
formation of a protective coating of Zn3(P04)2(s) because "... zinc
phosphate is insoluble in water at all concentrations. ..." That is
obviously false, because zinc orthophosphate solid does have a measurable
aqueous solubility constant (Table 3), but the chemical combination has
become a useful corrosion inhibitor.
Lead (II) forms at least one orthophosphate solid of low solubility in
realistic drinking water conditions, which can serve as the basis for
corrosion control (9, 17, 30, 32, 33). Figure 2 shows that the predicted
equilibrium lead (II) solubility can now fall below the MCL at low
carbonate concentrations, with a dosage of only 0.5 mg P04/L.
Experiments with both lead and galvanized pipes at U.S. EPA have
demonstrated that supersaturation with zinc orthophosphate is not necessary
for lead solubility control because the lead forms a passivating lead
orthophosphate film (30). Investigations by the Water Research Centre in
England (32) have indicated the solid to be hydroxypyromorphite,
Pb3(P04)20H. Successful field applications in England have been
reported by Sheiham and Jackson (32).
The passivating action of the orthophosphate depends on the pH, the
TIC concentration, the orthophosphate concentration, and temperature. No
data is available on the effect of the latter. On new lead surfaces, a
very thin, smooth film is formed. Taylor (34) has suggested that the
orthophosphate might not be as effective if the pipe surface is already
covered with corrosion products and deposits (34).
If the solubility constant for Cu3(P04)2 * 2H20 is accurate,
and if it is a realistic phase that would precipitate in drinking water
systems, then orthophosphate addition would be of little benefit for copper
level control. Solubility calculations were performed for orthophosphate
levels up to 2 mg P04/L, and no precipitation of a copper (II)
orthophosphate solid was predicted. If either or both of the solubility
constants used for tenorite and malachite are too low, then this conclusion
would have to be reevaluated. A zinc orthophosphate coating on the copper
could only be deposited if its solubility product would be exceeded. High
concentrations of zinc and orthophosphate might be required, depending upon
the pH.
97
-------
Figure 4 shows the solubility diagram for zinc when a dosage of 0.5 mg
P04/L is added, and Z^CPO^ • 4H20 ( -hopeite) is the
probable solid orthophosphate precipitate. At pH values under 7.5 and at
low TIC concentrations, some solubility suppression is likely. Increasing
the orthophosphate concentration would extend the stability domain of
-hopeite. Zinc leaching from galvanized pipe might be slowed or limited by
the mass-action principle, if zinc ion were added to the water along with
the orthophosphate inhibitor. More research in this area is definitely
needed.
Four advantages to orthophosphate addition plus pH adjustment are
apparent. First, for at least lead and zinc, the same solubility can be
attained at a lower pH than using pH plus carbonate alone. This assists
disinfection and slows organic halogen formation in chlorinated supplies.
Second, orthophosphate may be beneficial to the iron-zinc alloy substrate
beneath the galvanized layer, yielding superior protection where the
surface layer is finally penetrated. This area needs significant
investigation. Third, the same treatment strategy might also benefit
asbestos cement, cast iron, or cement mortar-lined distribution pipe
lines. A fourth advantage is that for waters with high carbonate contents,
a significant reduction in lead solubility would be possible without
softening or decarbonation.
POLYPHOSPHATE ADDITION
Polyphosphates are strong chelating agents for calcium, magnesium,
iron, and other trace metals. While they have useful properties for
reducing scaling, cleaning tuberculation from distribution mains, and
reducing "red water" complaints through sequestration of Fe2+ ion, they
have not been proven useful in reducing the solubility of zinc, copper, or
lead. Even though they may be somewhat effective relative to nontreatment,
particularly in acidic waters, alternative treatment strategies will
virtually always prove superior in reducing metal levels. Any joint use of
pblyphosphate chemicals with another treatment (such as to prevent scaling
at high pH, or with sodium silicate) must be accompanied by thorough
monitoring of trace metal concentrations. This is necessary because the
actual polyphosphate ligands have not been identified, and their complex
formation constants have not been determined, making even semiquantitative
estimates of effect through solution modeling calculations very unreliable.
SILICATE ADDITION
Silicate addition has been demonstrated to be an effective procedure
for corrosion inhibition of galvanized steel pipes, particularly at high
temperatures. Studies by Glasser and Lachowski of the structures of
silicate solutions (35) concluded that there is essentially no difference
between those prepared by the dissolution of anhydrous sodium metasilicate
or by salts with sodium orthosilicate groups. Therefore, either type of
additive should provide similar results for corrosion inhibition.
98
-------
Early studies with galvanized and brass drinking water pipes indicated
that the silicate additive (8 to 12 mg Si02/L) helped form amorphous
films binding corrosion products of the metal, and additionally enmeshing
iron oxide and organic particles present in the water supply (36). Thus,
an initial corrosion product needed to be present in order for the silicate
to work effectively. Similar conclusions may be drawn about the silicate
action on limiting the dissolution of zinc electrodes, zinc oxide, and zinc
dust through the formation of adsorption compounds that control zinc ion
diffusion (37). Silicate treatment may, therefore, bridge the gap between
control of zinc levels in solution by the precipitation of basic zinc
carbonate, zinc hydroxide, or zinc oxide, and sustaining the life of the
zinc coating by the formation of a uniform and adherent passivating film.
An indication of this potential has been given by studies done at the
Illinois State Water Survey with simulated domestic hot water systems (38,
39). In waters of low carbonate content, a film of zinc pyrosilicate was
apparently formed at a pH of 8 and a dosage of 8 mg SiC>2/L at 60°C
(140°F). However, in waters with higher carbonate contents, an adherent
scale of basic zinc carbonate mixed with silicate was found (38). Further
experiments were conducted to better define the optimum pH and silicate
dosage in a water of approximately 100 mg CaC03/L total hardness, 35 to
55 mg CaC03/L total alkalinity, and 8 mg S102/L natural silicate (39).
An initial high dosage of silicate (28 mg Si02/L total) was found to be
useful to provide more immediate corrosion inhibition and less bulky scale
formation, but the optimum dosage for this water quality was about 18 mg
SiC>2/L at a pH of approximately 8.5. Corrosion inhibition in these
experiments was monitored by the use of standardized ASTM pipe coupon
inserts for weight-loss determinations (40). Silicate requirements tended
to be less as the pH was increased by caustic soda additions. The authors
concluded that in the hot water systems, the optimum silica concentration
and pH may be different for waters of higher alkalinity, where more basic
zinc carbonate corrosion product would be formed, whereas primarily
hemimorphite (Zn4Si207(OH)2 * 1^0) was found in this study.
Little or no calcium carbonate deposition was detected. They also
indicated that other water quality factors, such as chloride, sulfate,
calcium, and magnesium, could change the dosage requirements and could
influence properties of the scale.
In a study of two fairly soft waters in Seattle, sodium silicate
addition at a pH of 8.0 considerably reduced the corrosion rates of black
steel pipes compared with other treatments investigated (41). After two or
three months of exposure of galvanized steel pipe coupons, the silicate
treatment showed slightly inferior results to that attained by lime plus
soda ash. Obviously, the interplay of pH, silicate dosage, and initial
corrosion product formed on the pipe is of critical importance at low
temperatures as well as in hot water systems. Further investigation is
necessary before treatment guidelines can be given with much confidence.
Silicate addition has not been found to be particularly effective for
copper pipe, although few studies have actually been conducted. No
99
-------
apparent reaction was found (38) between copper metal and a soft water
containing 8 mg Si02/L at pH 8.4 at 60°C (140°F). The "control"
experiment in this study was at pH 6.7, making a direct comparison of the
effect of pH adjustment alone to the different inhibitors tested very
difficult. Also, in waters tested having a high hardness, calcium
carbonate precipitation was suspected, making resolution of the role of the
sodium silicate inhibitor difficult. The scales seemed to offer good
corrosion resistance, however.
In the Seattle study (41), silicate addition proved to be advantageous
only in keeping even the initial corrosion rates of the copper specimens
low. After three months of exposure, alternate treatments such as lime
plus soda ash or lime plus zinc orthophosphate were equivalent or superior.
Some experimental tests have been conducted with lead pipe (30, 32),
and observations of the inhibitory effect of naturally occurring silica
have been summarized (30). After approximately 8 to 9 months of exposure
to 20 mg Si02/L, lead levels were slowly reduced in a soft,
low-alkalinity water at pH 8.2. In systems with both new galvanized and
lead pipes and 10 mg Si02/L concentration, lower lead levels were
observed with continued exposure. These levels were at approximately the
0.05-mg Pb/L MCL, but presumably the pH and TIC content of the water would
be important variables to consider before extrapolating results. These
experiments tend to support the idea of the slow formation of a film on new
pipe, but which might become effective over the long term as a "cementing"
agent for other corrosion products. Corrosion films on lead pipe always
tend to be very thin and uniform, unlike those on galvanized pipe. Sheiham
and Jackson (32) conducted some experiments with low alkalinity moorland
waters and concluded that 10 mg Si02/L had little effect on lead
dissolution.
CONCLUSIONS
Case studies and solubility calculations for zinc, copper (II), and
lead (II) in drinking water show that several control strategies have been
proven to be adequate for the attainment of current MCLs and SMCLs. The
most difficult metal to control is lead, and for some water chemistries
flushing of the lines before taking water for human consumption may be
required if optimization of treatment for lead corrosion control cannot be
accomplished. Lowering of the lead MCL to 0.025 or 0.01 mg/L would cause
considerable difficulty in light of current knowledge about available
inhibitors.
In order to develop efficient and accurate treatment techniques,
considerable effort must be expended in the future to understand the mode
of action of phosphate and silicate inhibitors, as well as the chemistry of
lead, copper, and zinc in response to changes of pH, TIC, and temperature.
This research is vital both to ensure the toxicological purity of the
water, and to extend the life of plumbing materials of diverse composition
in the same water distribution system.
100
-------
REFERENCES
1. EPA. 1976. National Interim Primary Drinking Water Regulations.
EPA-570/9-76-003. U.S. Environmental Protection Agency Office of
Water Supply.
2. EPA. 1980. Interim Primary Drinking Water Regulations; Amendments.
Federal Register 45(168):57332 (August 27, 1980).
3. EPA. 1979. National Secondary Drinking Water Regulations. EPA
EPA-570/9-76000. U.S. Environmental
Protection Agency Office of Drinking Water.
4. Ryder, R.A. In preparation. Corrosion Inhibitors. In: Internal
Corrosion of Drinking Water Systems. AWWARF.
5. Kuch, A., and I. Wagner. 1983. A mass transfer model to describe
lead concentrations in drinking water. Water Res. 17(10):1301.
6. Mattsson, E., and Fredericksson. 1968. Pitting corrosion in copper
tubes — cause of corrosion and counter-measures. Br. Corros. J.
3(9):246 (Sept. 1968).
7. Schock, M.R. 1980. Computer modeling of solid solubilities as a
guide to treatment techniques. Seminar on Corrosion Control in Water
Distribution Systems. U.S. EPA, Cincinnati, May 20-22.
8. Schock, M.R. 1980. Response of lead solubility to dissolved
carbonate in drinking water. Jour. AWWA 72(12);695. Errata 73(3):36
News (1981).
9. Schock, M.R., and M.C. Gardels. 1983. Plumbosolvency reduction by
high pH and low carbonate; solubility relationships. Jour. AWWA
75(2):87.
10. Schock, M.R. and C.H. Neff. 1982. Chemical aspects of internal
corrosion; theory, prediction, and monitoring. Proc. AWWA Water
Quality Technology Conference X, Nashville, Tennessee, Dec. 5-8.
11. Schock, M.R., and R.W. Buelow. 1981. The behavior of asbestos-cement
pipe under various water quality conditions: Part 2, theoretical
considerations. Jour. AWWA 73(11);609.
12. Stumm, W. 1956. Calcium carbonate deposition at iron surfaces.
Jour. AWWA 48:300 (1956).
13. Stumm, W. 1960. Investigations of corrosive behavior of waters.
Trans. ASCE - San. Engrg. Div. 86:27.
101
-------
14. Sontheimer, H., W. Kolle, and V.L. Snoeyink. 1981. The slderite
model for the formation of corrosion-resistant scales. Jour. AWWA
73(11):572.
15. Kuch, A., and H. Sontheimer. 1984. Corrosion in Iron and Steel
Pipes, or the "Protective-Layer-Story." DVGW Forschungsstelle am
Engler-Bunte-Institut, Postfach 6380, 7500 Karlsruhe 1, West Germany.
Unpublished.
16. Singley, J.E. 1981. The search for a corrosion index. Jour. AWWA
73(11) :579.
17. Gardels, M.C., and M.R. Schock. 1981. Corrosion indices: invalid or
invaluable? Proc. AWWA Water Quality Technology Conference IX,
Seattle, Washington. EPA 600/D-82-108.
18. Rossum, J.R., and D.T. Merrill. 1983. An evaluation of the calcium
carbonate saturation indexes. Jour. AWWA 75 (2): 95.
19. Samuels, E.R., and J.C. Meranger. 1984. Preliminary studies on the
leaching of some trace metals from kitchen faucets. Water Res.
18(1): 75.
20. Chapters of the Municipal Code of Chicago Relating to Plumbing, with
Amendments to October 19, 1978. Index Publishing Corp.
21. Loewenthal, R.E., and G.v.R. Marais. 1976. Carbonate Chemistry of
Aquatic Systems: Theory and Applications. Ann Arbor Science.
22. Carlson, W.D. 1983. The polymorphs of CaC03 and the
aragonite-calcite transformation. In: R.J. Reeder (ed.),
Carbonates: Mineralogy and Chemistry, Reviews in Mineralogy, Vol. 11,
Mineralogical Society of America.
23. Plummer, L.N., and E. Busenberg. 1982. The solubilities of calcite,
aragonite and vaterite in Co2~H20 solutions between 0 and 90°C,
and an evaluation of the aqueous model for the system
Geochim. Cosmochim. Acta 46:1011.
24. Larson, T.E., and A.M. Buswell. 1942. Calcium carbonate saturation
index and alkalinity interpretations. Jour. AWWA 34:1667.
25. Schock, M.R. 1983 (in press). Temperature and Ionic Strength
Corrections to the Langelier Index (Revisited). Submitted manuscript.
26. Bachle, A., et al. 1981. The corrosion of galvanized and unalloyed
steel pipes in drinking water of different hardness and neutral salt
content. (In German.) Werkstoffe und Korrosion 32:435.
27. Anderson, E.A., and M.L. Fuller. 1939. Corrosion of zinc. Metals
and Alloys (Sept. 1939).
102
-------
28. Gilbert, P.T. 1952. The nature of zinc corrosion products. Jour.
Electrochem. Soc. 99(1):16.
29. ILZRO. 1967. Surface Characteristics of Zinc in Aqueous Media.
Final Report. ILZRO Project ZE-20 (May 1967).International Lead
Zinc Research Organization, Inc.
30. Schock, M.R. 1984 (in preparation). The corrosion and solubility of
lead in drinking water. In: Internal Corrosion of Drinking Water
Systems, AWWARF.
31. Murray, W.B. 1970. A corrosion inhibition process for domestic
waters. Jour. AWWA 62(10):659.
32. Sheiham, I., and P.J. Jackson. 1981. The scientific basis for
control of lead in drinking water by water treatment. Jour. Inst.
Wtr. Treatment Engr. and Scientists 35(6):491.
33. Gregory, R., and P.J. Jackson. 1983. Reducing Lead in Drinking
Water. Water Research Centre Report 219-S. Stevenage Laboratory,
Elder Way, Stevenage, Herts SGI 1TH, England (May-June 1983).
34. Taylor, F.B. 1980. Metals — Corrosion or Dissolution? Proc. of the
Seminar on Corrosion Control in Drinking Water Systems, NEWWA (Mar.
24-25, 1980).
35. Glasser, L.S.D., and E.E. Lachowski. 1980. Silicate Species in
Solution. Part 1, Experimental Observations. J.C.S. Dalton Trans.;
393-398.
36. Lehrman, L., and H.L. Shuldener. 1951. The role of sodium silicate
in inhibiting corrosion by film formation on water piping. Jour. AWWA
43(3):175.
37. Glasser, L.S.D., et al. 1978. The reaction of zinc oxide and zinc
dust with sodium silicate solution. J. Appl. Chem. Biotechnol. 28:799.
38. Lane, R.W., et al. 1973. Silicate treatment inhibits corrosion of
galvanized steel and copper alloys. Mat. Perform. 12(4):32.
39. Lane, R.W., et al. 1977. The effect of pH on the silicate treatment
of hot water in galvanized piping. Jour. AWWA 69(8):457.
40. ASTM. 1976. Standard Methods of Test for Corrosivity of Water in the
Absence of Heat Transfer (Weight Loss Methods). D2688-70, ASTM
Standards, Part 31;141.
41. Ryder, R.A. 1978. Methods of Evaluating Corrosion. Proc. AWWA Water
Quality Technology Conference VI, Louisville, Kentucky, Dec. 3-6
(1978).
103
-------
DRINKING WATER REGULATIONS AND THEIR IMPACT
ON MATERIALS USED IN WATER SYSTEMS
by: James R. Boydston
Manager, Drinking Water Program
Oregon State Health Division
P.O. Box 231
Portland, Oregon 97207
Regulations and plumbing code requirements have a very definite
influence on the materials used in water systems and house plumbing. Such
regulations are necessary to provide for protection of the public health
where the combination of water quality and piping materials may result in
hazardous conditions. One of the problems of greatest current concern is
the presence of heavy metals in drinking water resulting from lead or other
piping materials being used to convey waters that are highly corrosive.
A number of regulations have been developed at both federal and state
levels in an attempt to prevent health hazards. The U.S. Environmental
Protection Agency (EPA) is currently in the process of adopting national
revised primary drinking water regulations (1). The current interim
regulations specify maximum contaminant levels for microbiological,
inorganic, and organic chemicals and radionuclides (2). The maximum
contaminant level (MCL) means "the maximum permissible level of a
contaminant in water which is delivered to the free flowing outlet of the
ultimate user of a public water system." These MCLs are currently being
reviewed and in a number of instances may be significantly lowered.
In addition to the MCLs, the interim regulations also include
requirements to identify the presence of specific materials in distribution
systems and to monitor for corrosivity. This requirement stems from
increasing concern about corrosion byproducts and particularly the much
higher levels of metals observed in corrosive water when that water stands
for some time in house plumbing systems. The interim regulations require
that all community water systems sample their water twice during a one-year
period and report the Langelier index as an indication of the corrosivity
of the water. This testing is required in order to enable the primary
enforcement agency to determine which water supply systems should initiate
corrosion control measures. The approach being considered for the revised
primary drinking water regulations is to set specific monitoring
requirements for corrosion byproducts such as lead and cadmium. The
systems known to have corrosive water will be required to take sufficient
samples to insure that MCLs for the corrosion byproducts will not be
104
-------
exceeded. Under this interim regulation the states were encouraged to
implement corrosion control measures at systems distributing corrosive
waters. Much has been written about the shortcomings of the Langelier
index as a measure of corrosivity but, since no better test has been
developed, this remains the index of choice.
EPA has also developed Water Supply Guidance Number 73 on sampling
techniques to be used for compliance monitoring for metals in drinking
water (3). These sampling techniques are not federally enforceable but are
intended to provide a standardized method for testing for corrosion
byproducts. This guidance provides for collection of three samples: from
a house faucet to represent the water that has been standing overnight in
the house plumbing, from the service line to the house, and from the
distribution system. Rather than using this sampling protocol to determine
MCL violations, it is to be used more as an indication of the potential for
exposure to higher levels of corrosion byproducts than would be identified
from the usual primary inorganics testing.
In addition to federal and state drinking water regulations, plumbing
code requirements also influence materials used in house plumbing systems.
The plumbing codes of most states are based upon some national model such
as the Uniform Plumbing Code. The Oregon Code with which I am familiar is
based upon the Uniform Plumbing Code of the International Association of
Plumbing and Mechanical Officials with local variations to meet Oregon
requirements (4). One interesting feature of the Oregon Plumbing Code is a
provision that the Administrator of the Health Division may prohibit the
use of any pipe, solder, fillers, or brazing materials which do not conform
to product acceptability criteria adopted by the Health Division. The
Oregon Code specifies that the approved materials used in house plumbing
may include brass, copper, cast iron, galvanized malleable iron, galvanized
wrought iron, galvanized steel, polybutylene (PB), chlorinated polyvinyl
chloride (CPVC), or other approved material. Polyethylene (PE) and
polyvinyl chloride (PVC) water pipe may be used only for cold water systems
outside of buildings. Lead pipe is not permitted in potable water systems.
Oregon law was changed in 1979 to give the Health Division more direct
influence over plumbing regulations. This was due in part to problems in
the 1970s with galvanized pipe imported from overseas that was not
manufactured to American standards and resulted in serious taste and odor
problems, rapid corrosion, and some potential for gastrointestinal upset.
The State Legislature amended the law relating to regulation of plumbing to
require that all potable water pipe sold in Oregon be clearly marked with
the identification of the manufacturer and plant of origin. This permits
easier tracking of inferior pipe.
Recent regulatory changes throughout the country have taken a number
of directions. This may be due in part to a lack of federal standards for
corrosion testing and MCLs for corrosion byproducts. A number of cities
have become concerned about the potential health effects of corrosion and
corrosion byproducts and have funded engineering and medical studies to
105
-------
evaluate the impact of corrosion. The problem with lead in drinking water
is certainly not new in this country. In an 1845 report to the City of
Boston (5), it was concluded that: "Considering the deadly nature of lead
poison, and the fact that so many natural waters dissolve this metal, it is
certainly the cause of safety to avoid, as far as possible, the use of lead
pipe for carrying water which is to be used for drinking." The advice was
not heeded, however. In 1920 another Boston report (6) stated: "If water
supplying such works be found to attack lead, the best plan is to replace
them gradually, meanwhile protecting consumers by proper treatment of the
water supplied through existing services." Boston's approach to resolving
the problem was to provide treatment to reduce the corrosivity of water.
Tests run on water from consumers' taps in Boston before and after
implementation of treatment showed a significant drop in lead concentration
from above 0.05 mg/L to consistently below that level. This is
accomplished by pH adjustment to approximately 8.5 using sodium hydroxide.
Much lead pipe plumbing remains in use in the Boston system.
In other parts of the country, notably Seattle and Portland, some lead
service connections remain, but a significant source of lead appears to be
from lead-based solders used to join copper pipe and fittings.
The State of Delaware is the only state to date that has elected to
ban the use of lead-based solder for potable water systems. The Delaware
plumbing code requires that soldered joints in copper tubing and fittings
shall be made with solder conforming to ASTM B32, provided that the lead
content shall not exceed 0.20 percent for potable water systems. This in
effect will require that tin-antimony or tin-silver solders be substituted
for the usual 50/50 tin-lead solder.
The City of Seattle has used two soft surface-water sources for many
years. Complaints about rusty water or blue stains and resulting taste
problems led the City of Seattle to sponsor an extensive engineering study
of corrosion and corrosion effects in the Seattle water system (7). The
study confirmed that both sources of water were corrosive and exposed the
Seattle customers to corrosion byproducts such as lead in addition to the
esthetic effects. Seattle's approach to the problem was two pronged — the
adoption of a number of amendments to the Uniform Plumbing Code and the
provision for treatment of water to neutralize corrosivity. Amendments to
the Plumbing Code included the banning of lead solder, the requirement for
an approved dielectric fitting when connecting piping of dissimilar
materials, and the approval to use PB or CPVC as hot or cold water piping
in buildings not exceeding four stories in height. In addition the City
now prohibits the use of house plumbing systems for electrical ground.
The treatment applied in Seattle consists of the addition of lime and
sodium bicarbonate to increase the pH and alkalinity of both sources.
Incidentally, Seattle's engineers estimated that there would be a net
annual savings to the consumer, looking at a benefit/cost ratio of 5.4:1
when comparing the total annual cost of the treatment versus the projected
net annual consumer plumbing maintenance cost savings.
106
-------
Portland also had an engineering study performed which confirmed that
the untreated water from Bull Run River is also highly aggressive (8).
Although the Portland system still contains roughly 10 percent lead
services or pigtails, the engineers felt the major health risk was from
lead solder in newer copper-plumbed systems. A supplemental study was done
on some non-copper-plumbed houses that were known to have lead pigtails,
and it was concluded that the lead spike that occurred in water standing
within the pigtails would be dispersed below MCL levels before reaching the
kitchen tap. It was concluded then that Portland should continue to
replace lead pigtails as they were found and that banning lead solder in
new copper plumbing would be sufficient to protect the public health.
Portland's first approach to the problem was to consider banning lead
solder by ordinance in the City of Portland, but later requested that the
State Health Division ban lead solder statewide since Portland serves water
to a number of water districts outside the city limits. A city ordinance
would have no effect within these districts. The Health Division was
concerned that, although banning lead solder would solve a major part of
the problem with new construction, there would still be concern about the
health risk exposure from existing plumbing systems. The state held a
number of informational hearings to receive comments on a proposed lead
solder ban and other options. Much comment from the plumbing industry
centered on the difficulty of using the alternative solders due to their
narrower working temperature range and the consequences of greater expense
for new construction. Many comments reflected a concern that such a
proposal would be unenforceable and was not necessary in many areas of the
state where hard groundwater sources are used. Some testimony was received
from communities that already provided full treatment, including pH
correction, that the lead solder ban was unnecessary.
The State of Oregon is still considering a number of alternatives,
including a statewide ban on lead solder with certain regions of the state
eligible for an exemption from the ban, and a monitoring program to improve
identification of communities that have corrosive water followed by a
requirement that those communities prepare plans for corrosion control.
Oregon will shortly draft a proposed rule incorporating the alternatives
and will schedule a series of formal hearings to receive testimony prior to
adoption or modification of the rule.
It is obvious that regulations and plumbing code requirements take a
number of directions as concerned officials search for methods to safeguard
against potential health risks. There are two major difficulties with the
problem of corrosion byproducts. First, in the case of lead and perhaps
other heavy metals, the effects are accumulative and chronic exposure to
low levels may result in health impacts that are difficult or impossible to
measure. Studies done by Dr. Needleman of the Harvard Medical School (9)
confirm that children exposed to low lead dose levels performed less well
than children not so exposed, even though neither group showed any other
symptoms of lead poisoning. A second difficulty with the problem of lead
in water is the lack of adequate methods for measuring corrosivity.
107
-------
Under these circumstances regulatory officials have an obligation to
take a conservative approach and require all reasonable methods to minimize
or to prevent any uptake of lead through drinking water. Since there are
reasonable treatment or control methods available, communities should
voluntarily be providing corrosion control. Such treatment also will have
esthetic benefits that will more than offset treatment or control costs.
Although there has been a great amount of research done on corrosion,
the water supply industry still has no simple reliable test to measure it.
Coupon testing has been suggested as an accurate measurement of corrosion,
but time required for the testing makes this impractical. I would
challenge the research community to develop a standard method for measuring
corrosivity by exposing a volume of water to a specimen of lead, leaving
the water standing for a selected time, then measuring the lead uptake in
the water. It appears a reproducible laboratory test could be developed
that would approximate water standing in a plumbing system. This type of
test is urgently needed to help provide answers to questions being raised
by a concerned public.
REFERENCES
1. Federal Register Vol. 48, No. 194, October 5, 1983. National Revised
Primary Drinking Water Regulations, Advance Notice of Proposed
Rulemaking.
2. U.S. EPA. 1976. National Interim Primary Drinking Water
Regulations. EPA-570/9-76-003. U.S. Environmental Protection Agency.
3. U.S. EPA. 1982. Guidance for Monitoring and Sampling Techniques to
Determine Corrosion Products, Including Lead in Water Supply
Distribution System. Water Supply Guidance Number 73. U.S.
Environmental Protection Agency.
4. International Association of Plumbing and Mechanical Officials.
1979. Uniform Plumbing Code.
5. Report of Commissioners Appointed by Authority of the City Council to
Examine the Sources From Which a Supply of Pure Water May Be Obtained
for the City of Boston. J.H. Eastburn, City Printer. Boston,
Massachusetts, 1845.
6. R.S. Weston. 1920. Lead poisoning by water and its prevention.
JNEWWA 34:239.
7. Kennedy Engineers. 1978. Internal Corrosion Study, Phases 1, 2, 3,
Seattle Water Department.
8. James M. Montgomery, Consulting Engineers. 1982. Internal Corrosion
Mitigation Study, Final Report, Portland Bureau of Water Works.
108
-------
9. Needleman, H.L., C. Gunnoe, A. Leviton, R. Reed,
H. Peresie, C. Maher, and P. Barrett. 1979. Deficits in psychologic
and classroom performance of children with elevated dentine lead
levels. New England Journal of Medicine 300 (13):689.
109
-------
EUROPEAN DEVELOPMENTS IN USE
OF PLUMBING MATERIALS
by: Ivo Wagner
Head, Corrosion and Material
Testing Section
Engler-Bunte Institute
University of Karlsruhe
Karlsruhe, Germany
INTRODUCTION
The plumbing materials used in Europe during the last century were
lead, copper, and galvanized steel. Recently plastic materials, especially
cross-linked polyethylene, came on the market for residential
drinking-water installations. Depending on the distribution method,
individual countries made greater or lesser use of different plumbing
developments.
Changes in the quality and use of plumbing materials were due to
changes in price or technical quality (mainly due to corrosion), and health
considerations. With these changes, use of lead plumbing in Europe
decreased nearly to nothing, and the installation of cross-linked
polyethylene pipes in a double-tube system with metallic joints became
commonplace.
Although the cost of plumbing systems in new-house installations,
mainly in Germany, increased over the last 30 years (with the ratio of
galvanized steel to copper pipes changing from 6:1 to 3:2), the use of lead
was ruled out because of health problems. Polyethylene pipe was developed
mainly to correct local corrosion problems caused by metallic pipe. More
detailed information on the different materials, their quality standards,
and their characteristics are given below.
LEAD
Lead plumbing and lead distribution systems were used in Europe to a
great extent when central water-distribution systems were being developed,
in the later decades of the 19th century. In spite of warnings to be found
in the literature since the end of the 18th century, lead was the preferred
plumbing material up until the late 1930s.
110
-------
The regulations covering lead varied by country. For example, to
reduce lead uptake by water, Prussia recommended that lead had to be free
from other metals. The Netherlands passed a regulation requiring a layer
of tin inside lead pipes and, when that was discovered to be hazardous,
amended the regulation to require the tin content to be 99.5 percent.
In other countries, mainly in some German states, lead was banned
completely, beginning with Wuerttemberg in 1878 (1). Lead was finally
ruled out in Germany in 1972 with the introduction of the new German
Standard DIN 2000. (A general survey of this development is given in Table
1.)
Similar developments took place throughout Europe at a pace dependent
on the national regulations and possibilities. In the United Kingdom, for
example, a recommendation against the use of new lead pipes was given in
the early 1970s. The last new lead plumbing systems were installed in
1976. Table 2 shows the development of recommendations in Great Britain.
These very restrictive measures concerning lead plumbing were
supported, mainly with respect to the old distribution systems, by a lot of
research into water quality and levels of lead in drinking water. Numerous
studies and surveys have been performed and published in the last 10 years
(3,4,5). The main conclusion reached was that the level of lead in both
soft and hard waters nearly everywhere exceeded the Maximum Acceptable
Concentration (MAC) values, often to a high extent. The solution to this
problem was either replacement of lead by other materials or treatment of
water to reduce lead solubility (6).
The questions to be answered in this case were addressed by studies
performed in the United States (7) and the United Kingdom (8) on the
stability or solubility of lead compounds. Additional data were gathered
(9) by monitoring and calculating lead levels as a function of standing
time; the maximum solubility of lead in water was reached after long
s tanding time.
With respect to the arguments of European toxicologists (10, 11) that
even low lead concentrations have to be seen as a cumulative poison, the
longtime uptake should be evaluated. With this goal, further research was
performed on consumption patterns of households in the United Kingdom and
Germany (12, 13). Lead levels were compared depending on maximum
concentration, kind of pipe, pipe length and diameter, flow velocity, and
standing time. These data allow for the calculation of general and
individual lead intake by a minimum of monitoring and analysis.
COPPER
The use of copper pipes was generally unrestricted due to health
considerations. Nevertheless, blue or green discoloration of water due to
stagnation (standing time) — mainly in acid waters — demonstrates that
copper pipe can cause heavy metal uptake in water after a long stagnation
111
-------
TABLE 1
LEAD; STANDARDS AND RECOMMENDATIONS
IN FEDERAL REPUBLIC OF GERMANY
YEAR(S)
REGULATION
1878-1972
1972-
present
1980/81-
present
Ban on lead pipes for distribution and house
installation in different German states,
beginning with Wuerttemberg (1878)
Total ban for new installations in lead
(DIN 2000)
Study on Drinking Water and Lead
(DVGW-Research Institute, Karlsruhe)
Results;
• Connecting pipes to be replaced depending on
pipe length and water quality.
• In house installations, MAC (40 ppb) is
regularly exceeded.
• Flushing and orthophosphate dosage may be
used as immediate measure.
• House installations in lead must be replaced
as quickly as possible.
112
-------
TABLE 2
LEAD: RECOMMENDATIONS AND MEASURES
IN GREAT BRITAIN
YEAR(S)
REGULATION
1976-
present
1976-78
from 1980
from 1982
Lead plumbing not further in use for new
plumbing systems
Countrywide survey on "Lead in Drinking Water"
(DOE)
Drinking water treatment:
• by pH increase in soft water
• by o-phosphate dosage in hard water
Survey on consumption patterns in households
(WRC)
General exception in Great Britain: No
pressurized systems in house installations; roof
tanks, only kitchen tap under pressure
113
-------
time. Water treatment by pH adjustment was therefore recommended in order
to reach drinking-water standards. As a result of corrosion problems, the
development of German Standards on copper pipe qualities was recommended.
Contrary to British and North European recommendations, the German
experience with copper pipe allowed for a certain level of carbon or
lubricant in the pipes. Since copper was used normally only in soft waters
or in hot-water installations, this allowance worked well. However, with
an increasing amount of copper-plumbed cold-water systems in regions with
hard groundwater, Type I pitting began to occur more frequently and finally
accounted for more than 70 percent of all corrosion cases (14). The
standards on lubricant, mainly carbon, on the inner pipe surfaces were
changed and the permissible levels decreased sharply. As a consequence of
that change, more hot (hard) soldering took place; after a while, pitting
in cold-water pipes increased once again. Thus the permissible carbon and
lubricant concentrations have been reduced again and a new surface
treatment on soft copper pipes has been introduced. These new pipes, which
have been in use for about 3 years, seem to meet the necessary
specifications; only a few cases of copper pipe failures have been reported.
The development of the German Standards governing copper pipe is given
in Table 3.
Although the use of copper pipe has presented no health problems, the
use of copper pipe in house installations is a concern due to the soft
soldering used to join the pipe. Most solders were tin/lead or
tin/antimony alloys or mixtures with differing lead content.
In the previously cited survey (2) on lead in the U.K., nearly 3,000
houses were monitored; with a certain astonishment, analysts registered
concentrations of lead in standing waters of copper-plumbed systems that
were higher than in lead-plumbed systems. Detailed research at the Water
Research Centre (15) in the U.K. and at KIWA (16) in the Netherlands has
shown that soft solders with lead content allow a high quantity of that
lead to migrate.
Similar research in Germany (17) studied the migration of lead out of
soft solders and of cadmium out of hard solders. The results were not
considered to be a problem of heavy metal uptake by water but, interpreting
the data correctly, it must be accepted that under actual pH conditions,
lead and cadmium uptake could become a severe health problem like that
found in U.K. and the Netherlands.
However, standards on solders have changed now in the U.K., the
Netherlands, and Germany. Soft solders now have to be free of lead; this
requirement is most important to the British plumbing systems, which were
normally installed with soft solder. The development of standards in
Europe is illustrated by the change in German Standards for soft and hard
solders. Table 4 shows the changes in the allowed solders, comparing the
standards for 1972 and 1983. Note that the change to solders that are free
of cadmium, antimony, and lead is complete.
114
-------
TABLE 3
DEVELOPMENT OF STANDARDS FOR COPPER PIPE
IN DRINKING WATER INSTALLATIONS1
1972
1981
1983
Cu-hard
Lubricant
(10 mg/dm2)
Cu-hard
Lubricant
(4 mg/dm2)
Cu-hard
Lubricant
1 mg/dm2 (diam. 2")
2 mg/dm2 (diam. 2")
Cu-soft
Carbon
2 mg/dm2
Cu-soft
Carbon
0.8 mg/dm2
Cu-soft
Carbon
0.2 mg/dm2
(no carbon films
allowed)
^Derived from DVGW-Arbeitsblatt GW 392, Standard of German
Gas and Water Association.
115
-------
TABLE 4
DEVELOPMENT OF STANDARDS FOR SOLDERS ON COPPER PIPES
IN GAS AND DRINKING WATER INSTALLATIONS1
SOLDER 1972 1983
Hard solders Ag 40 Cd Ag 34 Sn
Ag 30 Cd Ag 44
Ag 2 P Ag 45 Sn
Ag 2 P
Cu 6 P
Soft solders Sn Ag 5 Sn Ag 5
Sn Sb 5 Sn Cu 3
Sn 50 Pb
^-Derived from DVGW-Arbeitsblatt GW 2, Standard of German
Gas and Water Association.
TABLE 5
GERMAN STANDARD FOR MILD STEEL PIPES
TO BE GALVANIZED1
1961 1972 1978
Clean and complete Surface must be free Mechanical
welding seam from scales and shells tests included
bending stress
Welding seam has to
be cut; maximum height
is limited to 0.3 d +
0.05 d (d = thickness of
pipe wall)
^-Derived from Standard DIN 2440.
116
-------
GALVANIZED STEEL
Galvanized steel as a plumbing material is used hardly at all in the
U.K. and only to a small extent in northern Europe. In contrast, central
European countries commonly use galvanized steel as a plumbing material, to
the extent that it accounts for 60-80 percent of all piping material.
When standards for galvanized steel were set, especially in Germany,
the primary concern was to avoid corrosion due to deterioration of the
material. Standards therefore had to meet noncorrosion requirements. The
German Standard developed as shown in Table 5 above; the 1961 standard for
the steel pipes was upgraded in 1972 to recommend cutting the welding seam,
and in 1978 mechanical strength tests on the welded pipes were added to the
standard. Experience had shown that the welded seams in galvanized steel
pipes were the primary source of material failure due to pitting (18).
An additional standard governed the quality of the zinc layer in the
galvanized steel pipes. Again, protection against corrosion (of the zinc
layer) was the primary concern: "Zinkgeriesel," a type of selective
corrosion of the zinc layer, leads to the presence of sandy, zinc corrosion
products in house installations, often combined with heavy red water
problems. Research over a period of 15 years proved that most of these
problems were caused by the bad quality of the zinc layer of airblown pipes
(19).
Table 6 shows the continuing development of the standard governing
galvanized steel pipes. The health aspect was addressed in 1972 by
reducing the cadmium level in zinc to a maximum of 0.1 percent and fixing
the lead content of the zinc layer at a maximum of 1 percent. Additional
recommendations on the quality of the zinc layer were introduced, and
steam-blown pipes with smooth, tight zinc surfaces became standard. The
most recent (1978) change to the standard occurred after research on the
cadmium and lead uptake of water by corrosion of the zinc layer showed that
cadmium and lead levels often were higher than drinking-water standards
allowed (20), even after standing times of only a few hours. The amount of
cadmium allowed in the zinc layer was therefore reduced to a maximum of
0.01 percent. The lead content of the layer, made necessary by the
galvanizing process (the galvanizing tub requires a lead surface to avoid
crystallization of metallic zinc), was reduced to a maximum of 0.8
percent. Lower concentrations should be reached. Data show that the mean
level of 0.5 percent lead in the zinc layer of galvanized steel is not
generally reached.
PLASTIC MATERIALS
The installation of plumbing systems made completely of plastic piping
is only now becoming common in Europe. Many such Installations have been
temporary, made by introducing cross-linked polyethylene as a plumbing
material. Previous attempts, mainly with polyvinyl chloride (PVC) piping
117
-------
TABLE 6
GERMAN STANDARD FOR GALVANIZING
MILD STEEL PIPES1
1963
1972
1978
Zinc must be a
minimum of 98.5%
pure
Coating must cover
whole surface and
be free from
local accumula-
tions of zinc
Coating must have
minimum quantity
of 400 g/m2
Cadmium content in
zinc limited to 0.1%
Lead content is
limited in zinc
layer to 1.0%
Inside surface must
be free of impur-
ities and produced
by steam blowing
Zinc to be used must
be free of cadmium
Cd content of the zinc
is maximum of 0.01%
Lead content is
limited to the
technological
necessity: less than
or equal to 0.8% Pb
Production must be
supervised by inde-
pendent material
testing institute.
Pipes must be signed
by durable continuous
printing.
^Derived from Standard DIN 2444.
118
-------
in cold-water installations, aroused no interest except in industrial or
technical water supplies where special corrosive waters (e.g., fully
softened or desalinated) are distributed and most plumbing is placed on the
wall.
Nevertheless, numerous nonmetallic materials are used in
drinking-water distribution and plumbing installations. In addition to the
hard-PVC pipes and polyethylene pipes in distribution networks, many
coatings of reservoirs, fittings and armatures, gaskets, and other products
in contact with drinking-water plumbing systems are made from plastics (or,
more generally, nonmetallic materials).
Recommendations were developed in different European countries in
order to guarantee that the quality of nonmetallic materials would have no
undesirable health effects. Initial recommendations and test methods
covering materials in contact with drinking water were developed (in the
Netherlands and the Federal Republic of Germany) to regulate specifically
the leaching of lead from lead-stabilized, hard-PVC tubes in contact with
drinking water.
Later recommendations covered other questions, for example
organoleptics, migration of total organic carbon (TOC), chlorine and oxygen
demand, and microbiological aspects. The emphasis differs from country to
country; while in Great Britain the recommendations are based mainly on
microbiological aspects (21), in Germany the recommendations and methods
have a more chemical and analytical direction (22).
Table 7 shows the general points of view of the German regulations
(the KTW regulations). In addition to the general points covered in Table
7, the parameters vary depending not on the specific material but on the
range of use. This regulation was based on the premise that the
surface-area/volume ratio and the contact time negatively influence water
quality. Table 8 gives the limits of the basic requirements depending on
the range of use. If special requirements concerning specific compounds
are found to be necessary depending on the particular use, these additional
parameters will be limited with respect to the range of use of the product.
Microbiological tests are likely to be added to KTW recommendations;
currently only a DVGW-Arbeitsblatt (23) covers this matter especially to
describe requirements on lining materials of drinking-water reservoirs.
REFERENCES
1. Anon. 1878. Amtsblatt d. koenigl. Wuertt. Min. d. Innern.
Stuttgart, No. 7, April 29, 1878: 101-106.
2. Anon. 1977. Lead in drinking water: A survey in Great Britain
1975-76. Poll. Paper No. 12, HMSO, London.
119
-------
TABLE 7
KTW EMPFEHLUNGEN (RECOMMENDATIONS)
FOR PLASTIC AND OTHER NONMETALLIC MATERIALS
TO BE USED IN DRINKING WATER DISTRIBUTION SYSTEMS1
1. Material compounds (recipe) must be in accordance with
positive lists, including limited amounts of compounds
(each compound to be approved by toxicological working
group)
General limits:
• on taste and odor, color, turbidity and test water.
• on migration of total organic carbon in test water.
• on chlorine demand in chlorinated test water.
Exposure time: 3x3 days, every 3 days change of test water
Test water quality: Desalinated water, for Cl2 dem. with
0.6-0.7 ppm free chlorine
Surface/volume ratio: Migration test - 1:1 cm^/cm^
demand test (depends on range
of use)
3. Additional requirements on special compounds to be
migrated depending on material composition (for example,
aromatic amines from epoxy resins, phenols from phenolic
resins, lead from Pb-stabilized PVC pipes, etc.).
^Derived from recommendations given by the German
Federal Health Bureau, Berlin.
120
-------
TABLE 8
MAXIMUM VALUES TO BE ALLOWED
IN KTW RECOMMENDATIONS DEPENDING
ON RANGE OF USE
RANGE OF USE
ORGANOLEPTICS
(TASTE,
ODOR, COLOR)
TOC
CHLORINE
DEMAND
A: Distribution
pipes
Not affected
S/V ratio 1:1
2.5
2.0
B: Reservoirs
Not affected
S/V ratio 1:4
10
C: Fittings,
Armatures
Not affected
S/V ratio 1:6
15
12
Dl: Gaskets with a
lot of surface
in contact
with water
D2: Gaskets with
little surface
in contact
with water
Not affected
S/V ratio 1:25
Not affected
S/V ratio 1:50
cmVcm^
63
75*
125
150*
*In 1985, these values will change to 50 and 100.
121
-------
3. Gregory, R. 1981. WRC Case Studies, 2nd Seminar on Lead in Drinking
Water, Paper 16 (March 3-4, 1981). Water Research Centre, Medmenham,
Great Britain.
4. Elzenga, C.H.J., and A. Graveland. 1981. Studies in the Netherlands,
2nd Seminar on Lead in Drinking Water, Paper 11 (March 3-4, 1981).
Water Research Centre, Medmenham, Great Britain.
5. Wagner, I., and A. Kuch. n.d. Trinkwasser und Blei,
Veroeffentlichungen des Bereichs und des Lehrstuhls fuer
Wasserchemie. Karlsruhe, Heft 18, ZfGW-Verlag, Frankfurt/Main.
6. Sheiham, I., and P.J. Jackson. 1981. J. Inst. Wat. Engr. Sci. 35:
491-515.
7. Schock, M.R. 1980. Journal AWWA 72: 695-704.
8. Jackson, P.J., and I. Sheiham. 1980. Calculation of lead solubility
in water. TR 152, Water Research Centre, Medmenham, Great Britain.
9. Kuch, A., and I. Wagner. 1983. Water Research 17: 1303-1307.
10. Ohnesorge, F.K. 1980. Gas/Wasser 121: 515-522.
11. Department of Environment (DoE). 1982. The Glasgow duplicate diet
study 1979-80. Poll. Rept. No. 11, HMSO, London.
12. Wagner, I. n.d. Verbrauchergewohnheiten - Stagnationszeiten.
Auswertung von Basisdaten zum Wasserverbrauch aus unterschiedlichen
Staedten der Bundesrepublik Deutschland. Unpublished.
13. Bailey, R. n.d. Influence of consumption patterns in households on
lead levels in drinking water. Water Research Centre, Medmenham,
Great Britain.
14. von Franque, 0. 1972. Werkstoffe & Korrosion 23: 241-246.
15. Oliphant, R.J. 1982. Special Subject 18, IWSA-Congress, Zurich: 5-9.
16. Anon. 1982. Invloed van loden dienst- en binnenleidingen op het
loodgehalte van drinkwater in Nederland. Report from VEWIN, KIWA, and
RID. April 1982.
17. Winkler, B., and 0. von Franque. 1981. Metall 35: 222-227.
18. Friehe, W. 1973. Helzung Lueftung Haustechnik 24: 51-54.
19. Werner, G., E. Wurster, and H. Sontheimer. 1973. Gas/Wasser 114:
105-117.
122
-------
20. Meyer, E. 1978. La Tribune de CEBEDEAU 31: 431-441.
21. British Standards Institution (BSI). 1982. DD 82.
22. KTW-Empfehlungen des Bundesgesundheitsamtes Berlin,
Bundesgesundheitsblatt Jarhgang 1977, 1. and 2. Mitt. ff.
23. DVGW-Arbeitsblatt W 270, Vermehrung von Mikroorganismen auf
Materialien fuer den Trinkwasserbereich. 1983. ZfGW-Verlag,
Frankfurt/Main.
123
-------
REPORT: PANEL NO. 1
Chairman:
Recorder:
Panelists:
EPA Resource
Person:
JOINING ALTERNATIVES FOR COPPER PIPE
Rhodes Trussell
James M. Montgomery
Consulting Engineers, Inc.
Pasadena, California
Peter Karalekas
EPA, Region I
Boston, Massachusetts
A.C. Kireta
Copper Development Association, Inc.
Merrimack, New Hampshire
Richard Ballentine
Silver Institute
Cincinnati, Ohio
Vince Doyle
Plumbing and Air Conditioning
Contractors of Arizona
Phoenix, Arizona
John Courchene
Seattle Water Department
Seattle, Washington
Jerome F. Smith
Lead Industries Association, Inc.
New York, New York
Bill Hampshire
Tin Research Institute, Inc.
Columbus, Ohio
Allen Hammer
Virginia Bureau of Water Supply Engineering
Richmond, Virginia
Marvin Gardels
Water Supply Research
Cincinnati, Ohio
124
-------
CHARGES TO PANEL
The panel was charged with four major items for deliberation. They
were:
1. to assess extent of water quality impacts from lead/tin solder;
2. to determine if there are acceptable alternatives to lead/tin
solder;
3. to determine if there are mitigating steps which may be taken to
reduce contamination;
4. to consider other remedial actions that could be taken by
government agencies, water utilities, and plumbing and various
other private groups.
To accomplish its work the panel was balanced in composition to
include a water quality consultant as chairman, four representatives of
solder and plumbing material suppliers (copper, silver, lead and tin), an
expert in plumbing, a municipal water quality engineer, a state water
supply engineer, and an EPA resource person who was expert in corrosion
research. In addition there were 18 attendees representing a range of
industrial, water utility, academic, and public interest groups and state
and federal government agencies.
DISCUSSION
After a review of the panel charge by Chairman Trussell, the seven
panelists spoke briefly in turn on various subjects related to the panel
charges. A.C. Kireta of the Copper Development Association (CDA) felt that
any problem was confined to soft acidic waters. In a CDA study of 142
buildings in Connecticut ranging in age from zero to 120 years old, only
two had lead concentrations greater than the maximum contaminant level
(MCL) for lead. Mr. Kireta recommended that all new plumbing systems be
flushed before being placed in service and questioned present sampling
methods for lead in tap water. Among the alternatives to lead/tin solder
is tin/silver, which Kireta felt was a good alternative. Brazing of pipe
was not desirable because of the many fittings now being manufactured with
plastic internal parts which would not withstand high brazing
temperatures. Tin/antimony is also an acceptable alternative; however, its
use on potable water lines would be difficult to enforce because of the
problem of switching by plumbers who would have lead/tin on hand for waste
piping. Kireta felt that water treatment to reduce corrosion was the major
solution to solving any existing water quality problems.
Richard Ballentine, representing the Silver Institute, pointed out
that 95-5 tin/antimony was not a good wetting solder whereas tin/silver was
very good and had a higher strength. He estimated that the cost for
plumbing a house with tin/silver solder was only $5 to $10 more than
125
-------
tin/lead because of reduced labor costs and the need for less solder due to
shorter overlap in couplings.
Vince Doyle, of the Plumbing and Air Conditioning Contractors of
Arizona, felt that the solution to the problem with lead/tin solder was to
upgrade training procedures for plumbers. He showed examples of poorly
soldered joints with excessive flux and solder on the interior of the pipe
as well as perfectly soldered joints with only a very thin line of solder
exposed to water. Mr. Doyle felt there would be no problem if all joints
were soldered efficiently. He also felt that tin/antimony was more
difficult to work with.
John Courchene, of the Seattle Water Department, discussed the
corrosion problems that had occurred in Seattle. Studies had shown that 60
percent of water samples had exceeded the MCL for iron and 13 percent for
lead. Among the problems associated with corrosion were high plumbing
repair bills ($7.5 million/year), staining, and health. The solution
developed for Seattle involved raising the pH of water from 6.8 to 8.3,
encouraging the use of noncorrosive materials such as plastic, and banning
the use of solder containing lead. These measures have resulted in the
reduction of lead, copper, cadmium, zinc, and iron in Seattle water.
Jerome Smith, Lead Industries Association, conceded that high lead
concentrations could be found in the first sample drawn from a faucet in
the morning. However, he questioned the meaning of the sample and urged
that epidemiological studies be done to determine if lead/tin solder had
any effect on blood lead levels. Among the solutions recommended by Smith
were treating water to reduce its corrosivity, chemical flushing of all
newly plumbed water lines, and the use of preforms in the joints of copper
tubing to prevent solder from flowing into the pipe itself which would
result in its retention in the joint.
Bill Hampshire, Tin Research Institute, pointed out that lead from
soldered joints in tin cans and from gasoline had been reduced
significantly. In studies by the Tin Research Institute there are a number
of solders that can replace tin/lead. Among the more promising are a
tin/copper solder which is weaker than lead/tin but still strong enough for
its intended use; other solders under development are stronger.
Allen Hammer, Virginia State Health Department, said that some
communities still had lead service lines and that lead had been found there
in excess of 10 times the MCL. He described a situation occurring in June
1983 in a house where 18 of 31 samples were greater than the MCL. The
piping was removed and replaced using tin/antimony solder. Further
sampling found no samples containing lead greater than the MCL. In another
house, the highest lead concentration found was 6.1 mg/1 before flushing
the water. Hammer felt there was no predictability of lead results. Among
the actions taken by Virginia were a news release recommending flushing of
water lines before consumption, appointment of a task force on building
code changes, and further study of the problem. In preliminary results of
126
-------
this new study 11 of 65 samples had lead greater than the MCL, with ranges
from 0.001 to 0.4 mg/1 lead in waters that ranged from pH 6.0 to 9.1. In
another study by a private lab only 1 of 105 samples contained lead greater
than the MCL.
In the discussion period following the presentations by panel members,
a number of questions and issues were raised. Opinions differed, from one
individual who felt there was no problem to others who advocated a total
ban on solder containing lead.
A total ban would deal with any future problems, but a number of
individuals pointed out the necessity of dealing with the current problem.
It was suggested that water utilities be required to treat their water to
reduce its corrosivity toward plumbing. The observation was made that
there is no standardized technique for sampling for lead. Regarding the
issue of the occurrence of the problem being confined to soft water areas,
Schock pointed out that in studies in Illinois, the problem of lead was
identified in communities with very hard, high-pH water, indicating a more
widespread problem. Regarding the ability of manufacturers to meet the
demand if lead/tin solder were banned, one individual felt that the
marketplace would meet any demand that developed.
FINDINGS AND CONCLUSIONS
1. Lead is present in some waters due to corrosion of solder. The
problem has occurred in both hard and soft water areas.
2. New installations, which also may contain particles of lead solder and
flux, and supplies with soft, acidic water, may be the worst-case
situations.
3. Alternative joining materials such as tin/silver or tin/antimony
solder are available should lead/tin solder be displaced.
4. Water treatment methods are available to reduce the problem in certain
areas. Also, properly soldered joints, chemical flushing, and
preforms may be useful in reducing lead in drinking water.
5. Proper plumbing practice achieved through communication with plumbing
groups and training can have a role in reducing the problem.
RECOMMENDATIONS
Although a number of recommendations were put forth by various panel
members and the audience, it was generally felt a consensus was reached on
the following areas.
127
-------
1. More study is needed on the presence of lead from corrosion of solder
in a variety of water qualities.
2. A standardized sample collection method should be developed to be used
in these studies and in monitoring of public water supply systems.
3. Additional studies are needed on treatment techniques to reduce
corrosion of lead from solder since simple pH adjustment may not be
appropriate for all waters.
128
-------
REPORT; PANEL NO. 2
METAL PIPE AND FITTING ALTERNATIVES
FOR PLUMBING
Chairman: J. Edward Slngley
James M. Montgomery
Consulting Engineers, Inc.
Gainesville, Florida
Recorder: Frank A. Bell, Jr.
Office of Drinking Water
Washington, D.C.
Panelists: James R. Boates
Boston Plumbing Company
Natick, Massachusetts
Dodd S. Carr
International Lead Zinc
Research Organization, Inc.
New York, New York
Anthony Colucci
Colucci and Associates
Morgan Hill, California
Larry Galowin
National Bureau of Standards
Washington, D.C.
Joe Glicker
Bureau of Water Works
Portland, Oregon
Stanley Mruk
Plastic Pipe Institute
New York, New York
Robert A. Wilson
Copper Development Association
Purcellville, Virginia
129
-------
CHARGES TO PANEL
Charges to the panel were to assess the extent of water quality
impacts from metal plumbing pipe and fittings and to consider mitigating
steps such as water quality control, use of alternative grounding
electrodes, limitation or control of some materials, and possible standards
and certification programs. Possible training and educational materials
development or teaching programs were also to be considered.
To accomplish its charges the panel was balanced in composition and
included a water quality/corrosion expert as chairman with discussants
including industrial representatives in the areas of copper, galvanized
steel, and plastic materials and metal fittings; a plumbing contractor; a
municipal water quality engineer; and a government researcher in plumbing
systems. In addition there were 15 attendees from a range of industrial,
consulting, and municipal and state government sources.
DISCUSSION
Metal pipe and fitting alternatives were outlined by the discussants;
key points were as follows:
• Copper pipe is the dominant household plumbing material with
principally esthetic effects from copper leaching and staining.
• Galvanized steel pipe has had a level rate of use in the United
States and is commonly used in many communities. Leaching effects
are principally esthetic with elevated zinc levels, although traces
of lead and cadmium may be present.
• Lead pipe and fittings are seldom used except in Chicago where the
plumbing code requires that service lines be lead pipe. (See "Lead
Product Utilization Survey of Public Water Supply Distribution
Systems throughout the United States," April 1984, which is
included as Appendix A to this panel report.) Lead
levels are elevated in water standing in lead pipes, more so in
low-pH, low-alkalinity waters but also in a variety of other waters.
• Metal fittings may also serve as a source of leached contamination,
for example lead, where the subject material is contained in the
fitting.
Water quality control was urged as the main solution to control
corrosion in all types of metal pipe by the metal industry
representatives. In general they urged that local utilities provide
corrosion control treatment and that decisions regarding piping materials
be made at the local level on a case-by-case basis.
13Q
-------
On the other hand the water utility representative said that plumbing
materials, where identified as the source of contaminants, should be
controlled by plumbing code modification. He also discussed the
utilization of EPA's Drinking Water Advisory #73 in pursuit of appropriate
corrosion control sampling and inquired what national results had been
found from the 1980 corrosion monitoring regulations. While not a part of
this panel discussion, attention was frequently turned to the subject of
lead/tin solder. An undercurrent of concern regarding recent and ongoing
moves to ban lead/tin solder was evident through the whole panel session.
A lack of communication was cited by the plumbing representative, who
was unaware of the water quality impacts of plumbing materials. He
emphatically stated that plumbers wanted to safely and efficiently transmit
drinking water to the consumer in the same quality received from the public
utility, but he was unaware of some of the problems that had been mentioned.
Other needs for research and development were highlighted in the areas
of:
• standardized sampling and analytical procedures for the study of
leached contamination.
• improved means of studying corrosion scientifically and efficiently.
• the problem of corrosion from grounding and stray electric current
and the effective use of dielectric or separating fittings between
dissimilar materials.
FINDINGS AND CONCLUSIONS
While this was a stimulating panel session, the divergent and
conflicting views and time limitations prevented a full coverage of all
subjects. For the subjects covered, the chief conclusion appeared to be
that lead pipe contributes some lead contamination even in waters
considered noncorrosive; its use in new plumbing installations has been
abandoned in nearly all jurisdictions; there is no logical requirement for
its use to continue. The majority conclusion was that its future use
should be discouraged or banned.
Panel consensus was evident on the general subject of communication.
It was agreed that more formal communication on plumbing materials and
water quality needs to be provided, particularly to the plumbing and code
development and enforcement communities.
Regarding other materials or problems, there was insufficient
information or time to cover the assigned subjects. The questions at hand
were important and should be covered in subsequent meetings in more depth.
131
-------
APPENDIX TO PANEL SESSION 2.
LEAD PRODUCT USE SURVEY OF PUBLIC WATER SUPPLY DISTRIBUTION SYSTEMS
THROUGHOUT THE UNITED STATES
INTRODUCTION AND SUMMARY
A survey was conducted by each of the 10 EPA Regional Water Supply
Branches to assess the extent of lead utilization in water distribution
systems throughout the United States. A total of 153 public water systems,
selected from 41 states, including the District of Columbia and Puerto
Rico, were examined.
In a previous survey that was conducted almost 60 years ago ("Action
of Water on Service Pipes" by Wellington Donaldson, Journal AWWA, 11:649
(1924), 48 percent of the 539 cities that were evaluated had utilized lead
service pipes. The results of the present survey confirm that lead
products still exist in many water distribution systems and that the
utilization of such products in the past was widespread.
With regard to lead in drinking water, many studies show that lead is
seldom found in measurable concentrations in water sources. Lead, however,
is found frequently in corrosive tap water. Corrosive water may be reduced
by a number of measures, such as adjusting the pH, etc.
The problem of corrosion of lead has become such an important issue
that the EPA has proposed interim regulations requiring the determination
of the presence of specific materials in water distribution systems and
monitoring for characteristics of corrosivity of the water. Furthermore,
recent studies have indicated lead-tin solder can impart excessive lead
concentrations to drinking water.
ACKNOWLEDGEMENT
This survey was prepared and reviewed by David Chin and Peter
Karalekas of the U.S. EPA Region I Water Supply Branch staff. Special
thanks go to Terry Regan for performing the tedious task of typing this
survey. Region I would like to commend the efforts of Water Supply Branch
members from the other nine regions who contributed the data and
information for this survey and to thank Frank Bell for his encouragement
and assistance.
METHODOLOGY
The survey consisted of conducting telephone interviews or sending out
the questionnaires to state counterparts or to the particular water
systems. Most of the systems that were surveyed represented those that
132
-------
served large populations (91 out of the 153 systems served populations
exceeding 100,000 people).
The main problem with this survey was the reliability of the responses
that were obtained. A number of questions that were asked involved making
rough estimates of the use of lead products in water distribution systems.
In some of the systems that we evaluated, recordkeeping just did not exist
in the past. Some water system operators or superintendents were more
experienced and knowledgeable than others, which in turn resulted in more
reliable estimates. Also, some superintendents may just be more
cooperative than others, and would be willing to divulge more information
and make an effort to do some research. It is difficult to take into
account a number of these subjective factors.
In order to obtain uniform response from each of the ten regions, the
following questionnaire was developed and suggested to be utilized in this
survey:
• What is the total population served by the water system?
• What is the total number of services in the water system?
• What service line materials are being used for new installations?
• What part of the service line does the water system own?
• Were lead or lead-lined service pipes ever installed in the
distribution system? If so, how many still remain in the
distribution system and where are they located?
• Were lead gooseneck service connections ever used? If so, how many
still remain in the distribution system and where are they located?
• Were lead sweatjoints ever used? If so, how many still remain in
the distribution system and where are they located? Are they still
being used?
• What other service line materials have been used in the past?
• Does the water system have an ongoing replacement program for lead
materials utilized in the distribution system?
• Has the system had any problems with high lead concentrations in
the distribution system?
• What is the pH of the water in the distribution system? Has the pH
been adjusted higher to prevent corrosion problems?
133
-------
OBSERVATIONS
We primarily focused our attention on the information obtained on the
water systems' utilization of lead services, lead goosenecks and lead
sweatjoints. The following table summarizes our findings:
# of Systems That Had # of Systems That Had
Type of Utilized Lead Services Used Lead Goosenecks
Response (Out of 153 Responses) (Out of 152 Responses)
Affirmative 112 (73.2%) 94 (61.8%)
Unknown 5 (3.3%) 15 (9,9%)
None Known 36 (23.5%) 43 (28.3%)
Not Asked or
No Response — 1
Approximately 73 percent of the systems we surveyed indicated that
lead service lines had been installed in the past. In fact, the City of
Chicago still installs lead services for those locations utilizing a size
pipe of under and including 2". Lead services are still part of the
accepted plumbing code in Chicago.
Lead goosenecks have also been widely used. Nearly 62 percent of the
water systems had utilized lead goosenecks. Approximately half of the
water systems questioned reported the use of lead sweatjoint in their
systems. It is assumed that, because of the universal use of lead-tin
solder for joining copper pipes, lead in solder would be found in all of
the systems surveyed.
As for the responses regarding the ownership of the service line, they
varied widely from the water supplier owning the service line from the main
to the meter, property line, or curb, to the consumer owning the entire
service line. The significance of these responses is that for many water
systems, water suppliers have no responsibility over the utilization of
lead products in the home plumbing area.
Few water suppliers have active ongoing replacement programs for any
lead materials that have been installed in the distribution system. It
appears that water suppliers will only replace a lead service line or a
lead gooseneck, for example, when a leak exists.
Many system reported pH values of less than the optimum for
controlling lead corrosion. Yet, nearly all of the responses we received
on the question of whether there is a problem with high lead concentrations
in the distribution system were negative. Once-per-year monitoring for
lead required by the NIPDWR would certainly account for many of the
negative responses.
We have summarized, alphabetically by state, the information that was
submitted from all of the ten EPA Regions in the following table.
134
-------
CO
Ul
Name of Utility
and/or
Public Water Supply
Location
Birmingham,
Alabama
Montgomery/
Alabama
Fayetteville,
Arkansas
Fort 9nith,
Arkansas
Hot Springs,
Arkansas
Pine Bluff,
Arkansas
El Centro,
California
San Diego,
California
Arvada,
Colorado
Aurora ,
Colorado
Denver,
Colorado
Estimated
I of people
served
650,000
175,000
45,000
71,629
43,000
60,000
26,402
929,000
86,409
187,000
1,000,000
Total
# of
services
150,000
54,232
11,765
29,000
17,700
20,077
6,100
208,946
24,500
45,000
225,000
Service line
material
used for old
installations
galvanized
iron, PVC
galvanized
iron, lead
galvanized
iron
copper,
galvinized
iron
PVC, copper,
galvanized
copper, PVC
copper, lead
lead used
until 1940
copper,
galvanized
polyethylene
copper
Service line
material
used for new
installations
copper
copper
copper, PVC
copper, PVC,
polyethylene
copper
PVC
copper
plastic,
copper
PVC
copper
copper
Use of lead
services
% or t
still left
none
known
# unknown,
replaced if
found
none known,
replaced if
found
t unknown,
general
area known
Yes, none
remain
unknown
300
10,000
Yes, #
unknown
almost all
have been
replaced
tes, #
unknown
Use of lead
goosenecks
% or t
still left
none
known
none
known
none
known
Yes,
# unknown
Yes,
< 200
unknown
unknown
10,000
Yes, 1
unknown
Yes,
12
Yes, *
unknown
Use of lead
sweatjoints
% or #
still left
none
known
Yes, all
have been
replaced
Yes,
3-4%
Yes,
t unknown
none
known
unknown
unknown
10,000
t unknown,
still being
used
none
known
Yes, t
unknown
pH in the
distribution
system and if
adjusted
8.5
lime added
8.9
lime added
7.5
no adjustment
7.5
no adjustment
8.2
no adjustment
7.1
no adjustment
7.4 - 7.6
7.6 - 8.2
7.4
adjusted
7.5
no adjustment
7.2 - 7.8
adjustment at
one plant
-------
Name of Utility
or
Public Water Supply
Bridgeport
Hydraulic Co. ,
Bridgeport, CT
Hartford HOC,
Hartford, CT
New Britain
Water Department
New Britain, CT
New Haven
Water Department
New Haven, CT
Stamford
Water Company
Stamford, CT
Waterbury
Water Department
Waterbury, CT
Dover
Water Department
Dover, Delaware
Artesian Water Co.
Newark, Delaware
Newark
Water Department
Newark, Delaware
Wilmington
Water Department
Wilmington, DE
Wilmington, DE
Suburban Water Co.
of Claymont, DE
District of
Columbia
Estimated
1 of people
served
336,185
391,000
99,302
371,259
84,000
124,000
27,500
140,000
30,000
140,000
75,000
756,500
Total
* of
services
90,000
87,000
15,802
90,000
19,714
25,000
7,428
36,449
6,800
34,252
23,000
120,000
Service line
material
used for old
installations
brass ,
galvanized
iron
N/A
N/A
brass, lead,
galvinized
iron
brass, lead,
galvinized
iron
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Service line
material
used for new
installations
PVC, copper
copper,
brass
copper
copper, FVC
ductile iron
copper
copper,
ductile iron
copper
copper, PVC
copper
copper
polyethylene
copper
Use of lead
services
% or #
still left
1%
none
known
Yes, #
unknown
most have
been
removed
24
none
known
none
known
none
known
•none
known
<5%
<1%
60,000
Use of lead
goosenecks
% or #
still left
2*
Yes, #
unknown
1,500
30,000
532
Yes
10,000
50%
4%
none
known
none
known
<1%
none
known
Use of lead
sweatjoints
% or *
still left
still being
used for
services
Yes
1500
still in
use
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
pH in the
distribution
system and if
adjusted
6.8 - 7.5
7.01
8.2 - 8.3
7.2
6.7 - 7.1
7.2
8.0 - 8.2
7.3 - 7.7
6.7 - 7.2
7.0 - 7.7
7.1 - 7.9
7.6 - 8.3
-------
Name of Utility
or
Public Water Supply
Orlando, Florida
Tallahassee,
Florida
Atlanta, Georgia
Savannah, Georgia
Boise, Idaho
Cour d'Arlene,
Idaho
Idaho Falls, Idaho
Lewiston, Idaho
Twin Falls, Idaho
Chicago, Illinois
(city proper only)
Decatur, Illinois
Gary - Hobart
Water Corp.
Gary, Indiana
Estimated
# of people
served
300,000
100,000
650,000
150,000
105,000
21,000
41,000
17,000
26,000
3,000,000
94,000
100,000
(Gary Only)
Total
# of
services
76,100
42,000
127,636
45,000
38,482
9,000
15,600
4,900
9,489
400,000
31,000
30,000
Service line
material
used for old
installations
galvanized,
copper
galvanized,
PVC, lead
galvanized,
PVC
cast iron,
steel
code accepted
galvinized
iron, copper,
kalmain
code accepted
code accepted
galvanized
ductile iron,
lead
lead,
galvinized
iron
lead until
1950,
copper
Service line
material
used for new
installations
PVC, copper
PVC, copper
copper
polyethylene
or
polybutylene
code accepted
polyethylene,
PVC
code accepted
copper
plastic,
code accepted
lead for 2" &
under;ductile
iron for >2"
copper
copper
ttee of lead
services
% or I
still left
Yes, all
have been
replaced
# unknown,
some being
replaced
f unknown
Yes, #
unknown
100
known
none
known
none
known
# unknown,
very early
use
none
known
all pipe 2"
and under
remain
<100
replaced
when found
25,000
replaced
when found
Use of lead
goosenecks
% or #
still left
none
known
# unknown,
some being
replaced
none
known
Yes, t
unknown
100
known
200
10,000
not
known
Yes, #
unknown
For all
pipe 2" and
under
<100
replaced
when found
25,000
replaced
when found
Use of lead
sweatjoints
% or #
still left
none
known
none
known
none
known
none
known
Yes
4,350
none
found
20,000
# unknown,
very early
use
none
?
?
?
pH in the
distribution
system and if
adjusted
7.8
no adjustment
8.1
adjusted with
caustic soda
7.3
phosphate and
lime added
7.0 - 7.3
no adjustment
7.1 - 7.7
7.0 - 7.2
7.2 - 7.3
7.4 - 7.6
adjusted
7.9
adjusted
8.1 - 8.5
adjusted
9.5
7.6
-------
00
Name of Utility
or
Public Water Supply
Des Moines, Iowa
Water Works
Dubuque , Iowa
Water Works
Sioux City, Iowa
Utilities Dept.
Kansas City, Kansas
Board of Public
Utilities
Johnson County,
Kansas
Water District #1
Topeka, Kansas
Water Department
Wichita, Kansas
Dept. of Water and
Water Poll. Control
Lexington, Kentucky
Louisville,
Kentucky
Portland, Maine
Water Distript
Annapolis, Maryland
DPW
Baltimore, Maryland
Estimated
* of people
served
253,078
62,321
82,000
160,000
205,000
140,000
280,000
290,000
670,000 -
720,000
141,240
95,000
1,100,000
Total
t of
services
68,301
20,390
25,000
56,000
65,972
45,000
115,000
70,000
202,000
38,000
75,000
368,000
Service line
material
used for old
installations
galvanized,
lead
lead used
prior to 1928
N/A
galvanized
galvanized
and cast iron
galvanized
wrought iron
lead prior
to early
1940 's
galvanized,
lead, copper
PVC
galvanized
galvanized
iron
N/A
N/A
Service line
material
used for new
installations
copper,
ductile iron
copper,
ductile iron
copper, PVC,
ductile and
cast iron
copper, PVC
copper
copper,
ductile iron,
brass
PVC, copper
PVC
copper
copper,
ductile iron
N/A
N/A
Use of lead
services
% or #
still left
Yes, #
unknown
5,000
12,000
a few,
t unknown
none known
135
<15,000
older parts
of city
# unknown,
replacement
program
Ifes, #
unknown
none known
none known
<0.1%
Use of lead
goosenecks
% or #
still left
Yes, #
unknown
not known
Yes, f
unknown
several
hundred
none known
none known
100
# unknown,
replacement
program
none known
200
none known
<0.1%
Use of lead
sweat joints
% or *
still left
*es, #
unknown
tes, #
unknown
2,000
several
hundred
none known
none known
none known
none known
none known
N/A
N/A
N/A
pH in the
distribution
system and if
adjusted
9.3
9.5 - 9.6
7.2 - 7.5
7.8 - 8.1
9.0
9.2 - 9.5
8.2
7.2
phosphate and
lime added
8.2,
lime added
6.6 - 6.7
8.7
7.8
-------
Name of Utility
or
Public Water Supply
Cumberland,
Maryland
Hagerstown,
Maryland
Washington Suburban
Sanitary Comnission
Maryland
Boston, MA
Water and Sewer
Comnission
Brockton, MA
Water Department
Cambridge, MA
Water Department
Fall River, MA
Water Department
Lawrence, MA
Water Department
Lowell, MA
Water Department
Lynn, MA
Water Department
New Bedford, MA
Water Department
Newton, MA
Water Department
Estimated
I of people
served
40,000
70,000
1,705,000
600,000
89,040
100,361
100,000
62,000
94,000
80,000
100,169
91,066
Total
# of
services
17,200
16,892
N/A
90,000
21,455
13,250
15,500
12,000
18,000
17,800
22,000
25,612
Service line
material
used for old
installations
N/A
N/A
N/A
cast iron,
Lead
N/A
N/A
N/A
N/A
lead , copper ,
tin alloy
copper, lead,
cement-lined,
galvinized
N/A
N/A
Service line
material
used for new
installations
N/A
N/A
N/A
copper ,
ductile iron,
cement-lined
copper
copper
PVC, copper
copper
copper
copper
copper
copper, FVC,
galvanized,
cement-lined
Use of lead
services
% or #
still left
none known
none known
none known
50%
none known
<10%
10%
1500
<25%
4,500
20%
3,500
30%
6,600
1,000
Use of lead
goosenecks
% or #
still left
none known
none known
none known
50%
<10%
2,000
<10%
10
none known
<25%
4,500
40%
7,000
30%
6,600
4,000 -
5,000
Use of lead
sweatjoints
% or *
still left
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
pH in the
distribution
system and if
adjusted
7.4
7.5
7.7
8.8
9.6 - 9.8
8.6 - 8.8
8.5
8.0
7.2 - 7.3
6.5 - 7.5
7.0
8.8
-------
Name of Utility
or
Public Water Supply
Quincy, MA
Mater Department
Scmerville, MA
Water Department
Springfield, MA
Water Department
Worcester, MA
Water Department
Kalamazoo,
Michigan
Rochester ,
Minnesota
Jackson ,
Mississippi
Meridian,
Mississippi
Cape Girardeau,
Missouri
Mo. Utilities Co.
Columbia, Missouri,
Water and Light
Department
Florissant,
Missouri
Water Department
Independence ,
Missouri
Mo. Water Company
Estimated
1 of people
served
91,200
80,600
241,500
172,342
DO, 000
60,000
210,000
49,000
35,000
62,000
70,000
120,000
Total
* of
services
21 ,000
14,000
43,381
37,000
35,000
17,500
66,000
16,000
12,332
17,000
15,000
40,000
Service line
material
used for old
installations
N/A
N/A
galvanized
iron
N/A
lead
lead, copper,
galvanized
PVC,
galvanized
galvanized
galvanized
plastic
N/A
galvanized
Service line
material
used for new
installations
copper
copper
copper
copper,
ductile iron
copper
copper
copper, brass
copper
polyethylene ,
copper
copper
copper , brass ,
ductile and
cast iron
copper,
ductile iron
Use of lead
services
% or #
still left
<200
<20%
100-200
replaced/yr
none known
unknown
5%
30%
Yes, all
have been
replaced
Yes, 1
unknown
Yes, none
remain
# unknown,
old section
of town
50
very few
exist
Use of lead
goosenecks
% or #
still left
Yes, I
unknown
Yes, #
unknown
unknown
unknown
5%
30%
Yes, all
have been
replaced
none known
unknown
# unknown,
old section
of town
none known
me
2,000
Use of lead
sweatjoints
% or *
still left
N/A
N/A
N/A
N/A
?
?
none known
none known
none known
none known
not in use
since 1944,
jy still exis
none known
pH in the
distribution
system and if
adjusted
8.8
8.8
6.8
6.5 - 6.6
7.0
no adjustment
7.4
no adjustment
8.9 - 9.0
lime added
8.5
lime added
8.3
8.5
9.6 - 9.8
5t
9.8
-------
Name of Utility
or
Public Water Supply
Jefferson City,
Missouri, Capital
City Water Conpany
Joplin, Missouri
Water Works Co.
Kirkwood, Missouri
Water Department
Lee's Summit ,
Missouri
Water Conpany
Ray town, Missouri
Water Conpany
St. Joseph,
Missouri
Water Conpany
St. Louis, Missouri
Water Division
St. Louis County,
Missouri
Water Conpany
Sedalia, Missouri
Water Department
Grand Island,
Nebraska
Utilities Dept.
Lincoln, Nebraska
Water Department
Omaha, Nebraska
Utilities District
Estimated
1 of people
served
33,000
55,000
27,987
38,000
15,000
91,518
455,000
1,000,000
22,000
33,944
174,560
431,000
Total
# of
services
9,300
17,000
9,761
10,500
6,500
30,213
118,500
260,910
8,833
10,739
52,488
126,234
Service line
material
used for old
installations
galvanized,
cast iron
copper, cast
iron, tubeloy
galvanized
iron
galvanized
galvanized
iron tubeloy,
plastic, A-C,
lead
galvanized
cast iron,
galvanized,
A-C
galvanized
cast and
ductile iron
galvanized,
cast iron
galvanized
iron, tube
alloy
Service line
material
used for new
installations
copper
polyethylene
tubing
copper
copper
copper
PVC, copper
copper,
ductile iron
copper, PVC,
ductile iron
copper,
ductile and
cast iron
copper
copper,
ductile iron
copper,
brass ,
ductile iron
Use of lead
services
% or *
still left
Yes, #
unknown
1,500
Yes, #
unknown
30-40 %
none known
8779
unknown as
to number &
location
28,000
Yes, #
unknown
2,000
innumerable
25,000 -
30,000
Use of lead
goosenecks
% or #
still left
Yes, #
unknown
none known
Yes, #
unknown
unknown
\fery few,
if any
8779
Yes, #
unknown
Yes, #
unknown
Yes, t
unknown
none known
innumerable
25,000 -
30,000
Use of lead
sweatjoints
% or #
still left
Yes, #
unknown
none known
none known
Yes, *
unknown
Vary few,
if any
17,558
still being
used
Yes, #
unknown
Yes, #
unknown
none known
innumerable
wiped- lead
joints
pH in the
distribution
system and if
adjusted
9.7 - 9.9
7.5
10.0
7.8 - 8.3
9.9
7.6
8.4 - 10.2
9.6 - 9.8
8.0 - 8.4
7.3
7.5
9.0
-------
Name of Utility
or
Public Water Supply
Manchester,
New Hampshire
Water Works
Albuquerque,
New Mexico
Sante Fe,
New Mexico
Elizabeth town.
New Jersey
Water Conpany
Hackensack ,
New Jersey
Water Company
Monmouth, NJ
Consolidated
Water Conpany
Newark, New Jersey
Water Department
Passaic Valley,
New Jersey
Water Commission
Buffalo, New York
Division of Water
Erie, County,
New York
Water Authority
Jamaica, New York
Water Supply Co.
New York City,
New York
Aquaduct System
Estimated
# of people
served
100,000
342,000
50,000
500,000
800,000
249,000
380,000
287,316
358,000
317,000
650,000
7,071,000
Total
# of
services
22,000
104,400
14 ,000
150,080
180,709
70,121
58,000
63,000
90,000
99,148
118,000
820,000
Service line
material
used for old
installations
galvanized
iron
galvanized
galvanized
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Service line
material
used for new
installations
copper
polystyrene ,
copper,
cast iron
cast or duct-
ile iron,
copper <2"
PVC for <2",
copper,
ductile iron
copper for
<2",
ductile iron
copper, PVC,
ductile iron
copper,
brass ,
ductile iron
copper ,
ductile iron
copper ,
ductile iron,
PVC
copper,
ductile iron,
PVC
copper,
ductile iron
copper, brass
ductile and
cast iron
Use of lead
services
% or t
still left
none known
1,000
locations
known
none known
<5%
very few
<5%
150
replaced/yr
2,972
about 75%
about 25%
# unknown
# unknown
# unknown,
many
# unknown,
many
Use of lead
goosenecks
% or #
still left
250
none known
1,000
locations
known
<1%
very few
25,000
100
replaced/yr
unknown
none known
unknown
sane
sane
# unknown,
many
# unknown,
many
Use of lead
sweatjoints
% or #
still left
none known
none known
several
miles of
old lines
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
pH in the
distribution
system and if
adjusted
6.6 - 7.8
8.1
No adjustment
7.8
adjusted
6.8 - 7.6
7.8
7.5 - 7.8
7.5 - 7.9
7.1 - 7.3
7.5 - 7.8
7.8
6.0 - 6.9
6.3 - 7.4
-------
Name of Utility
or
Public Water Supply
Suffolk County,
New York
Water Authority
Charlotte,
North Carolina
Raleigh,
North Carolina
Bismarck,
North Dakota
Fargo,
North Dakota
Grand Forks,
North Dakota
Lawton, Oklahoma
Oklahoma City,
Oklahoma
Tulsa, Oklahoma
Corvallis, Oregon
Eugene, Oregon
Hillsboro, Oregon
Estimated
# of people
served
850,000
350,000
165,000
35,000
61,281
53,060
100,000
460,000
420,000
43,000
103,000
35,000
Total
t of
services
246,898
106,342
47,000
8,881
25,159
8,927
40,000
165,000
140,000
12,000
34,777
10,000
Service line
material
used for old
installations
N/A
galvanized,
black iron,
PVC
galvanized
iron
cast iron
iron, lead,
copper
black iron
galvanized
N/A
galvanized,
copper,
PVC
galvanized
polyethylene ,
copper, "
polybutylene
galvanized,
copper,
polybutylene
Service line
material
used for new
installations
copper,
ductile iron
copper
copper,
brass
copper,
plastic
copper
copper
copper
copper
copper, plas-
tic, brass,
cast iron
3/4" meter
uses r'copper
or plastic
Copper
code accepted
Use of lead
services
% or #
still left
<1%
Yes, all
have been
replaced
none known
none known
Yes
2,500
Yes
2,000
Yes, 250
remain in
older areas
Very few
remain in
older areas
Yes, very
few remain
none known
Yes, #
unknown
none known
Use of lead
goosenecks
% or #
still left
none known
Yes, all
have been
replaced
Yes, all
have been
replaced
none known
# unknown,
used a long
time ago
none known
Yes, 150
remain in
older areas
Yes, in old
sections of
town
Yes, very
few remain
Yes, all
known
removed
Yes, #
unknown
none known
Use of lead
sweatjoints
% or *
still left
N/A
none known
none known
none known
Yes,
2,500
Yes,
4,000
none known
Yes, in old
sections of
town
none known
none known
Yes
none known
pH in the
distribution
system and if
adjusted
6.7-7.2
9.3
lime added
7.2
NaOH & phos-
phate added
9.3
9.0 - 9.5
lime added
8.6 - 9.2
lime added
7.6
10.5
7.5
N/A
7.2 - 7.3
adjusted
6.9
adjusted
-------
Name of Utility
or
Public Water Supply
Portland, Oregon
Salem, Oregon
Erie, Pennsylvania
Bureau of Water
Philadelphia,
Pennsylvania
Water Department
Philadelphia, PA
Suburban
Water Company
Pittsburgh,
Pennsylvania
Water Department
Western Pennsylvania
Water Company of
Pittsburgh, PA
Metropolitano A
San Juan,
Puerto Rico
Enrique Ortega
La Plata,
Puerto Rico
Ponce Urbano,
Puerto Rico
Aqua Dilla,
Puerto Rico
Caquas ,
Puerto Rico
Estimated
# of people
served
650,000
120,000
205,000
1,685,000
902,000
424,000
i
500,000
711,999
363,936
241,540
114,497
109,236
Total
# of
services
120,000
45,000
52,000
522,000
287,000
89,000
126,512
178,000
90,984
60,385
28,624
27,309
Service line
material
used for old
installations
galvanized
code accepted
copper,
galvanized
iron
copper,
galvanized &
black iron
N/A
copper
galvanized,
cast and
wrought iron
N/A
N/A
N/A
N/A
N/A
Service line
material
used for new
installations
copper
is used up
to the meter
code accepted
Type k
copper
copper
copper
copper
copper and
ductile iron
copper, PVC,
ductile iron
copper, PVC,
ductile iron
copper, PVC,
ductile iron
copper, PVC,
ductile iron
copper, PVC,
ductile iron
Use of lead
services
% or #
still left
Yes, #
unknown
unknown
none known
1
Yes, #
unknown
<200
30%
Yes
5,318
none known
none known
none known
none known
none known
Use of lead
goosenecks
% or #
still left
Yes,
10,000
none found
2% remain
from 1920's
service area
Yes, #
unknown r
none known
none known
Yes, #
unknown
none known
none known
none known
none known
none known
Use of lead
sweatjoints
% or #
still left
Yes
20,000
some, f un-
known, still
in use
used until
early
1960 's
leadite
joints used
rost replacer
N/A
Yes, 5%
remain
Yes, #
unknown
N/A
N/A
N/A
N/A
N/A
pH in the
distribution
system and if
adjusted
6.8
No adjustment
low 6's
No adjustment
7.3 - 7.6
7.0 - 8.5
]
7.0 - 7.6
7.6 - 7.8
7.1
7.4 - 8.2
7.4 - 8.0
7.4 - 8.0
7.8 - 8.2
7.2 - 7.6
-------
Name of Utility
or
Public Water Supply
Pawtucket ,
Rhode Island
Water Department
Providence,
Rhode Island
Water Department
Warwick,
Rhode Island
Water Department
Charleston ,
South Carolina
Columbia,
South Carolina
Huron, South Dakota
Water Department
Pierre,
South Dakota
Sioux Falls,
South Dakota
Memphis, Tennessee
Nashville,
Tennessee
Logan, Utah
Odgen, Utah
Estimated
1 of people
served
108,750
283,816
78,500
350,000
231,700
13,000
11,973
81,343
800,000
350,000
28,000
64,000
Total
# of
services
21,000
69,121
26,000
60,000
70,440
3,500-
4,000
3,800
25,000
221,709
105,000
56,000
20,000
Service line
material
used for old
installations
brass ,
galvanized &
cast iron
N/A
N/A
galvanized
steel
galvanized
lead and
galvanized
ductile and
cast iron
galvanized
PVC,
galvanized
cast iron,
galvanized,
PVC
galvanized
and ductile
iron
N/A
Service line
material
used for new
installations
copper
copper and
ductile iron
copper
copper
Type k
copper
copper
PVC
copper
copper
copper
copper with
brass/bronze
fittings
copper
Use of lead
services
% or #
still left
lead-1 ined
10%
2,100
25%
100
Yes, #
unknown
none known
Yes, #
& location
unknown
Yes #
unknown
little used
Yes, #
& location
unknown
Yes, sane
remain in
older areas
# unknown,
in old part;
of the city
Yes, a
little used
none known
Use of lead
goosenecks
% or #
still left
none known
unknown
none known
Yes, #
unknown
none known
possibly, #
& location
unknown
Possibly, #
& location
unknown
Yes, #
& location
unknown
none known
3 none known
Yes, a lot
used
Yes, many
years ago
1 unknown
Use of lead
sweatjoints
% or #
still left
N/A
N/A
N/A
Yes, #
unknown
none known
possibly, #
& location
unknown
Possibly, #
& location
unknown
Yes, #
& location
unknown
none known
Yes, *
unknown
Yes, some
used with
goosenecks
none to
curb; many
inside bldgs
pH in the
distribution
system and if
adjusted
7.2 - 7.3
10.0
10.0
8.0 - 8.3
6.9
adjusted
8.8 - 9.2
adjusted
7.5 - 8.2
no adjustment
7.0 - 8.0
no adjustment
7.2
phosphate
added
7.8 - 8.0
caustic soda
added
N/A
7.1 - 7.2
s
-------
Name of Utility
or
Public Water Supply
Provo, Utah
Arlington County,
Virginia
Fairfax County,
Virginia
Water Authority
Henrico County,
Virginia
Newport News,
Virginia
Norfolk, Virginia
Richmond, Virginia
Bellingham,
Washington
Everett, Washington
Seattle, Washington
Spokane, Washington
Tacoma, Washington
Estimated
t of people
served
75,000
165,000
650,000
170,000
333,000
270,000
225,000
50,000
300,000
1,115,000
180,000
214,000
Ttotal
# of
services
15,000
34,000
134,000
46,200
90,000
61,700
62,500
17 ,000
90,000
168,371
62,000
68,362
Service line
material
used for old
installations
galvanized
N/A
N/A
N/A
N/A
N/A
N/A
galvanized
galvanized
and copper
galvanized,
copper, plas-
tic, kalmain
copper,
galvanized
galvanized,
copper
Service line
material
used for new
installations
copper
copper
copper
plastic,
copper
copper
PVC
copper
copper or
plastic
anything
Code allows
Code Accepted
Type F copper
galvanized,
plastic,
galvanized
copper
Use of lead
services
% or #
still left
Yes, a few
remain
200
none known
none known
<2%
3%
25,000
not known
believed
most has
been removec
Yes, #
unknown
If used all
have been
replaced
none known
Use of lead
goosenecks
% or *
still left
unknown
3,000
none known
1%
<2%
N/A
none known
Yes
probably
1
unknown
If used all
have been
replaced
none known
Use of lead
sweat joints
% or *
still left
none known
N/A
N/A
N/A
N/A
N/A
N/A
Yes
none known
probably
may have
Yes,
a lot still
in use
pH in the
distribution
system and if
adjusted
8.4 - 8.5
8.0
7.0+
7.3 - 7.9
7.1
7.2
7.2 - 7.6
6.8 - 6.9
adjusted
7.4
Adjusted
8.00
soda ash and
lime added
7.8
none
6.8 - 7.2
-------
REPORT: PANEL NO. 3
PLASTIC PIPE AND FITTINGS
Chairman:
Recorder:
Panelists:
EPA Resource
Persons
Gary Englund
Minnesota Division of Environmental Health
Minneapolis, Minnesota
Harry Von Huben
EPA, Region V
Chicago, Illinois
Alan J. Olson
B.F. Goodrich Co.
Cleveland, Ohio
David Spath
California State Department of Health
Berkeley, California
Tom Adams
Adams, Broadwell and Russell
San Mateo, California
Nina McClelland
National Sanitation Foundation
Ann Arbor, Michigan
Ray Lee
American Water Works Service Co., Inc.
Hadden Heights, New Jersey
Daniel W. Hurley
Associated Lead, Inc.
Philadelphia, Pennsylvania
Ray Jones
EPA, Region X
Seattle, Washington
Koge Suto
EPA, Region III
Philadelphia, Pennsylvania
147
-------
CHARGES TO PANEL
The panel was charged with assessing the extent of impact to drinking
water quality due to leaching from plastic pipe, permeation of plastic
pipe, and leaching from joining compounds and solvents; further charges
were to assess possible responses to the leaching and permeation problems.
Other approaches to the issue of public health and the use of plastic pipe
were also considered.
To accomplish its work the panel was balanced in composition to
include a state water supply engineer as chairman; representatives of
various industrial (plastics and metals), plumbing, and third-party
standards interests; water utility and State health engineers; and two
knowledgeable EPA technologists as resource persons. In addition there
were 23 attendees representing a range of water utility, industrial,
research, and state and federal governmental interests.
Charge No. It Assess the extent of impact to drinking water quality due to
leaching from various types of plastic pipe (PVC, CPVC, PB,
etc.).
Discussion
(Discussion on this item was strictly limited to leaching from the
pipe itself, with the problems of joining materials considered as a
separate issue.)
The argument was made by the plastic pipe interests that plastic pipe
has been used and tested for over 20 years with no indication of adverse
health effects due to leaching. The point was also made that plastic pipe
is regularly tested by the National Sanitation Foundation (NSF) with no
adverse health risk being noted.
The counter was made that the report being prepared by SRI
International may indicate significant leaching problems. The point was
also made that NSF works only from established health limits. The example
was given that if chloroform at 99 g/1 was found in a sample of pipe, it
would be "approved" by NSF — yet, most authorities would agree that the
level should be lower if possible. It was also noted that NSF carefully
looks at the analysis of the pipe material, but does not necessarily try to
determine other materials that may be present due to the manufacture of the
pipe.
Although everyone, more or less, had to concede that additional
adverse health effects due to pipe material leaching could be found in the
future, the plastic pipe interests were reluctant to endorse a general
statement that more research should be done on possible leaching problems.
148
-------
Panel Opinion
Majority opinion: The majority did not feel there is a problem of
adverse water quality due to leaching from plastic pipe, based on known
health effects information.
Minority opinion: The minority felt that there is a potential problem
and more information and research is needed. They particularly felt there
is a greater potential problem with some types of plastic pipe than with
others.
Charge No. 2; Assess the extent of impact to drinking water quality due to
permeation of plastic pipe.
Discussion
The point was made by plastic industry interests that a great amount
of plastic pipe is in use and, percentage-wise, the instances of detection
of permeation are extremely small. It was also stated that most
authorities will agree that it is not good practice to install any pipe in
contaminated soil. Neither of these points were contested.
But, the counter point was made that most pipe is installed when the
soil is "clean," with the expectation that it will remain clean for many
years, and then, in some instances, the soil becomes contaminated later,
creating a problem if the pipe is permeable.
Plastic industry interests stated that they do not deny there is a
permeation problem. They said they want to "get a handle on the problem"
and "want to be the first to make recommendations on specific areas where
plastic pipe should not be used." They also stated that they are looking
at protection measures that possibly could be incorporated into manufacture
or added during the installation of the pipe.
The point was made that a number of jurisdictions at a local level had
already taken steps to control locations where permeable pipe may be used.
Panel Opinion
It was unanimously agreed that there is a demonstrated susceptibility
of plastic pipe to permeation by organic chemicals. However, it was
recognized that the susceptibility is somewhat variable for different types
of plastic.
149
-------
Charge No. 3: Assess the extent of impact to drinking water quality due to
leaching from joining compounds, solvents and other
materials.
Discussion
There was general agreement that there is some degree of leaching of
compounds and solvents used in pipe joining. There appears to be little
information at present on the degree and duration of the problem.
There were opposing views on whether the leaching poses a potential
public health risk. One faction stated that there is no data demonstrating
a health risk, that the amount of leached material is, over all, very
small, and that the problem only persists for a few years after the pipe is
installed. Others felt that the exposure is significant enough to
seriously investigate the potential health danger.
There was also discussion on the potential of greater-than-necessary
leaching of joining compounds due to what is perceived as widespread use of
incorrect or "multi-purpose" joining compounds. According to plastic
industry representatives, multi-purpose compounds should be used only when
pipes of different types are joined; otherwise, only the material specified
for the particular type of pipe should be used.
Panel Opinion
Majority opinion: The majority agreed there is a problem of leaching
from joining compounds and more information is needed on the health risks
that this creates.
Minority opinion: The minority felt there is some leaching that takes
place from joining compounds, but that it is not of public health
significance.
Charge No. 4; Assess possible responses to leaching problems from piping,
joining compounds, and other materials:
• Limit the use of some materials.
• Improve installation techniques.
• Strengthen third-party standards and
monitoring programs.
• No change.
Discussion
It was generally agreed that, although there is concern, we are
nowhere near considering prohibiting the use of any materials for general
150
-------
use. It was also mentioned that any "limiting" of use in the future might
be in regard to ingredients, rather than particular products.
The fact that California and various local jurisdictions have already
imposed some regulation on use demonstrates that there is a definite need
to strengthen third-party standards and model codes.
Panel Opinion
There was agreement that there is a definite need to strengthen
third-party agreements and monitoring to review state, local, and model
codes and to continue to review and improve installation techniques. There
is a continuing need, as an industry, to review the compounds and solvents
used in the manufacture and installation of the products. There is also a
definite need for improved information, communications, and instruction
from the manufacturers, regulators and trade groups.
Charge No. 5: Assess possible responses to the problems of permeation and
water quality:
• Limit the use of plastic pipe.
• Develop new non-permeable pipe materials.
• No change.
Discussion
It was generally agreed that there is a need to notify users and the
public that plastic pipe should not be used in contaminated soil.
There was considerable interest in the California regulation
prohibiting use of plastic pipe under conditions considered to pose a
possible adverse health effect. The regulation is not well known or well
enforced, and it was written several years before permeation problems were
understood, but it is of interest as a start in recognizing and addressing
the problem.
It was agreed that there is a great need for the plastic pipe
industry, trade groups, the plumbing industry, and others to bring about an
awareness of the impacts of possible permeation.
The plastic industry representatives stated that they are working on
development of non-permeable pipe materials. There are also new
technologies being developed for improved installation (including soil
barriers), preventing migration, etc., that might also be applicable in
preventing permeation of plastic pipe.
It was unanimously agreed that "no change" in addressing the problems
of permeation was not appropriate.
151
-------
Panel Opinion
It was agreed that:
• There are areas where there is a need to limit the use of plastic
pipe because of existing or potential ground or environmental
contamination.
• There is a need for the industry or other bodies to inform
potential users of plastic pipe of the limitations of the products.
• There is a need for further investigation into the development of
non-permeable pipe or other means of preventing pipe permeation.
Charge No. 6; Detail any other approaches to coping with problems relating
to public health and use of plastic pipe.
Discussion
It was again discussed that there is a definite need to positively
identify the various types of pipe material and joining compounds or
solvents, so that only "compatible" materials will be used. Color coding
or other means could be used. This would make it easier for the user to
determine what products are appropriate and it would make it much easier
for the field inspector to assure that correct materials are being used.
It was stated that NSF is doing some testing for leaching and
permeation, and indications are that this program will be expedited in view
of the remarks and discussions at this meeting.
There is a great need for the manufacturers and trade associations to
"get the word out" to the users and installers on any problems, special
conditions of use, etc.
It was suggested that possibly a clause could be incorporated into the
NSF or other certification, stating that use of incorrect joining compounds
would nullify the certification.
Panel Opinion
It was agreed that:
• There is a need for better communications between the industry, the
trades, and consumers and installers on potential problems and/or
locations where plastic pipe should not be used.
• There appears to be a need for better identification (color coding,
labeling) of plastic pipe and joining compounds/solvents to better
assure the use of the proper joining material for each type of pipe.
152
-------
• "Universal" solvents are supposed to be used only for joining
different types of pipe, but are apparently being indiscriminately
and generally used due to incomplete information furnished by the
suppliers, or because the trades feel they are the easiest to use.
Charge No. 7; What are the indicated desirable or recommended remedial
actions by governmental, manufacturing, trade, code, or
testing groups?
Discussion
There was unanimous agreement that there has not been adequate
communications from any of the groups regarding the plastic pipe problems
that have been discussed at this meeting.
It was also discussed that there is a real need for additional testing
and research, particularly in regard to "real-world" conditions of
installation as they may differ from laboratory-controlled conditions.
There was also mention of a need for consistency in the information
put out by the various interested groups. It was also stated that it is
often difficult for someone with problems to find a source of information.
A "hotline" number for plastic pipe would be very helpful in this regard.
The matter was also discussed, without resolution, that there is a
need for each of the interested groups to be doing something specific
toward investigating the problems that have been discussed.
Panel Opinion
It was agreed that there is a need to improve communications between
all of the interested groups to get consistency in the information being
generated.
OTHER PANEL OBSERVATIONS AND RECOMMENDATIONS
A representative of the cast-iron pipe industry stated that gaskets
are available for cast-iron pipe that are used when the pipe is to convey
organic compounds, so the same material should be usable to prevent
contamination of water in the pipe if the soil is contaminated. It seems
that this is an important point if permeation of regular gaskets is
considered to be a problem, and should be made readily available to the
trade and public.
Although it was not specifically discussed by the panel, it should be
recognized that the "problems" of plastic pipe actually sort out into three
different "jurisdictional" areas:
153
-------
1. The water mains and water services, generally to the service
stop, are usually of a material approved by and installed in a
manner specified by the water utility.
2. Some jurisdictions allow the installation of other material past
curb stop, so the material could be different. The problems of
permeation through pipe is primarily a problem with these
jurisdictions.
3. The third problem the panel felt may be significant is from
joining material or solvents. This is primarily a problem of
interior piping, which is under the jurisdiction and inspection
(if any) by the local building inspector.
Somehow, it will be necessary to communicate with each of these
jurisdictions regarding the problems that will affect them.
154
-------
REPORT: PANEL NO. 4
REGULATORY AND COMPLIANCE ASPECTS
Chairman:
Recorder:
Panelists:
EPA Resource
Persons:
Raymond Jarema
Connecticut Department of Health Services
Hartford, Connecticut
Charles D. Larson
EPA, Region I
Boston, Massachusetts
James Sargent
Wisconsin Bureau of Plumbing
Sun Prairie, Wisconsin
Richard B. Howell, III
Delaware Bureau of Environmental Health
Dover, Delaware
Donald Dickerson
Dickerson Associates
Van Nuys, California
Charles Buescher, Jr.
Continental Water Co.
St. Louis, Missouri
Hugh F. Hanson
Consumers/Pirnie Utility Services Co.
White Plains, New York
Peter Lassovszky and Arthur Perler
Office of Drinking Water
Washington, DC
CHARGES TO PANEL
The charges to the panel were:
1. Examine the means by which materials/contamination problems can
be controlled through Maximum Contaminant Level (MCL) and
155
-------
monitoring requirements (either existing or as potentially modified) of the
Primary Drinking Water Regulations, as regards:
• Lead pipe and fittings.
• Copper pipe, joints, and fittings.
• Galvanized pipe and fittings.
• Plastic pipe, joining compounds, and fittings.
2. Examine the means by which materials/contamination problems can
be controlled through:
• State and local building or plumbing codes.
• Regional or national model codes.
3. Examine monitoring aspects of code compliance. (For example, how
would compliance with a code/solder requirement be monitored?)
4. Examine other mechanisms for regulatory or compliance change.
5. Assess any indicated desirable or recommended remedial actions by:
• Federal and state drinking water agencies relating to
changes in drinking water regulations, monitoring, and
implementation.
• Regional or national model code groups.
• Local and state code groups.
To accomplish its work the panel was balanced in composition to
include a state water supply engineer as chairman, a state code official,
two engineers knowledgeable in water utility operations, a former chairman
of a national model code development committee, a state water supply
engineer involved in changing his state's plumbing code, and two EPA
resource persons who are knowledgeable in corrosion and additive aspects.
In addition there were 20 attendees representing a range of water utility,
industry, research and state and federal governmental interests.
DISCUSSION
Panel members as well as most of the attendees of this panel session
felt the approach to solving regulatory and compliance problems must be
multifaceted. Limiting the toxic elements in materials used for plumbing
as well as treating water to reduce corrosiveness are both essential.
Because of the short time allotted for the panel to come up with
156
-------
recommendations, the best approach appeared to be to define
responsibilities and assign those responsibilities to the most appropriate
organizational entities.
The U.S. Environmental Protection Agency (EPA) was assigned the
responsibility of defining the problem, assessing or compiling existing
data on a regional or national basis, developing primarily MCLs, and
monitoring requirements to regulate corrosion byproducts. It was also
thought that EPA should be responsible for preparing source documents for
the states. It was recommended that the EPA conduct a national sampling
survey similar to the National Organics Reconnaissance Survey (NORS) or
National Organics Monitoring Survey (NOMS) surveys. Before undertaking the
survey there should be a standardization of sampling techniques and a study
designed which would incorporate uniformity in sampling. It was also
considered very critical that educational materials for states and local
government authorities should be developed by EPA. Water Supply Guidance
No. 73 (WSG 73) was passed out to those attending the panel discussion.
WSG 73 specifically deals with monitoring and sampling techniques to be
used for evaluating the effects of corrosion. There was concern voiced
that WSG 73 was not and had not been distributed widely enough and it
should certainly be made available to utilities. Also WSG 73 may serve as
the basis of a standardized sampling program. A copy of WSG 73 is included
as Appendix A to this panel report.
The state's responsibility was defined as being the enforcement agent
over the water purveyor. It was also felt that a model code should be
developed at the state level. However, there was some debate over whether
the model code should be at the state or at the federal level. There was
some consensus among the group that communication was necessary to improve
state agency interaction and coordination. This interaction includes water
purveyors, health departments, and building and housing departments at the
state level.
The water purveyor's responsibility was defined as the provision of
safe drinking water to its consumers. Water safety is assured through a
regular monitoring program; therefore, it is the water purveyor who will be
collecting the data in a uniform manner. It is the water purveyor who
implements treatment to reduce corrosivity. It is also the water
purveyor's responsibility to transmit educational material on an ongoing
basis to the consumer, thereby helping to educate the public.
Manufacturers of materials, including materials used to make pipe and
joints, should reassess standards on an ongoing basis. Third-party review,
for example by the National Sanitation Foundation, was discussed as being a
fairly good mechanism for a central standard-setting process.
At the local government level, the consensus was that the
responsibility for enforcement of plumbing as well as building code
requirements should be administered at this level. Because the inspection
and approval process takes place at this level, enforcement should also be
administered at this level.
157
-------
Consumers also have responsibilities and they should use the
educational materials distributed to them. The consumer-water purveyor
relationship is an essential one that must have credibility and trust
incorporated into it.
A building owner's responsibility is to comply with existing codes,
standards, and regulations that pertain to his facilities, including
building, plumbing, and health codes. Building owners have an ultimate
responsibility to their clients in meeting compliance requirements by
ensuring that public safety is addressed.
The responsibility for research was not designated, although the
general consensus was that additional epidemiological and health effects
research is needed.
In summary, communication and public education must be immediately
developed and implemented. Corrosive water should be made less corrosive,
and exposure to toxic compounds in plumbing materials should be reduced as
much as possible. EPA should define the corrosion problem and develop
primary MCL and monitoring requirements to regulate hazardous corrosion
byproducts. EPA should develop a standardized sampling procedure to help
identify distribution system corrosion problems. A national survey of
corrosion byproducts using the standardized sampling procedure is
recommended.
APPENDIX TO PANEL SESSION 4
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
DATE 'AU6 4 1982
SUBJECT: Water Supply Guidance Number
FROM vi-ctor J- Kimm, Director
Offi^r- of Drinking Water
T0 Water Supply Representatives, Region I-X
Holders of Water Supply Guidance Series
Attached is Water Supply Guidance Number 73 for Monitoring and
Sampling Techniques to Determine Corrosion Products, including
lead, in water supply distribution systems.
158
-------
Several Regional Offices have requested clarification and
guidance on sampling techniques to be used for compliance
monitoring for metals in drinking water, and on the applica-
bility of MCLs to corrosivity and corrosion products in water.
Neither the National Interim Primary Drinking Water Regulations
(NIPDI-7R), nor the Manual for Interim Certification of Labora-
tories address the issue of sampling techniques for metals.
By definition, an MCL is the maximum permissible level of a
contaminant in the water delivered to the consumer. Conven-
tional sampling techniques, in which lines are flushed in an
effort to provide a sample representative of the water pro-
vided by the supplier of water, is usually considered to be
appropriate in terms of the MCL definition. The NIPDWR con-
tained minimum requirements for the MCL, monitoring and re-
porting, with the expectation that States and other authorities
would adapt them to meet local needs. '•In the cases of sub-
stances that are primarily introduced into drinking water during
transit, it is probable that more extensive site specific
monitoring and sampling methods should be chosen by the local
or State authority to obtain a better indication of actual
human exposure from drinking water.
Paragraph 141.42 of the NIPDWR requires that community water
suppliers identify specific materials of construction that are
present in their distribution systems. However, additional
monitoring requirements were not specified to determine the
presence of the corrosion products that are contributed from
these materials to the drinking water. Corrosivity of the
water is to be determined on the basis of the Langelier Index,
and in certain cases, the Aggressive Index. When there is a
potential for an MCL to be exceeded, additional actions are
obviously warranted.
Although the regulations are not explicit regarding sampling
techniques for corrosion products in water, the States ob-
viously have the authority to require specified monitoring in
instances where it is felt that the water quality indices are
inadequate to determine the corrosive tendency of the water or
where regulated contaminants are potentially present. In such
instances, the attached guidance produced by the Criteria and
Standards Division can provide a practical tool to determine
the representative range of specific contaminant levels to
which the consumer might be exposed as a result of the inter-
action of the water and the materials of construction in the
distribution system. The sampling techniques described are
not federally enforceable, but are intended only as a guidance.
This approach was adopted from the methods that have been success-
fully used by Region I in dealing with lead contamination resulting
from aggressive water in contact with lead pipe and services.
159
-------
Water Supply Guidance Number 73
Guidance for Monitoring and Sampling Techniques to Determine
Corrosion Products, Including Lead, in Water Supply Distribution
Systems
Introduction
The presence of contaminants in the water at the consumer's tap
is a function of both treated water quality and distribution
system factors that can lead to localized contaminant problems.
Because the occurrence of corrosion related contaminants
in the water distribution system are often localized, random
sampling techniques are not always successful in detecting
their presence. Thus, the minimum monitoring requirements as
prescribed by the NIPDWR in 1975 may not be appropriate in some
situations to realistically determine the extent of consumer
exposure to contaminants resulting from corrosive water.
The 1980 amendments to the NIPDWR's established a one year corro-
sivity monitoring program which was intended to cause systems to
identify circumstances where corrosion related contamination may
occur, and to encourage appropriate corrective actions. Corro-
sivity measurements required under these regulations are deter-
mined indirectly using the indices (see 40 CFR 141.42(c)) which
are based on calcium carbonate saturation.
The indices specified in the 1980 amendments generally measure
the tendency of the specific water to form a protective calcium
carbonate (CaCO~) layer. The index value is dependent on the
pH, alkalinity, hardness, total dissolved solids (TDS) and the
temperature of the water. Each of these parameters may inde-
pendently affect the corrosive tendencies of the water so that
waters may be corrosive even though the index indicates non-
corrosive conditions.
Therefore, it is generally agreed that the indices may only be
applicable to determine the corrosive tendencies of the water
within narrow pH ranges, and then only if sufficient amounts
of calcium and alkalinity are present in the water. These
indices do not take into account the use of corrosion inhibiting
chemicals such as hexametaphosphate and bimetalic phosphates.
Additionally, calcium carbonate stability, as well as other
parameters, may change in the distribution system because of
retention in distribution reservoirs and rechlorination.
The sampling techniques suggested herein provide a method to
monitor and verify the effective corrosivity of the water and
the effectiveness of treatment to control corrosion for those
160
-------
instances where the water quality indices are not applicable.
For additional details involving the applicability of corrosion
indices and monitoring, refer to the Statement of Basis and
Purpose for the August 27, 1980, Amendments to the National
Interim Drinking Water Regulations.
Purpose
This Guidance provides a suggested protocol for water suppliers
who desire to determine if corrosion products are present in
those situations where the existing regulations may not provide
sufficient specificity. Specifically, the sampling techniques
herein can help to determine the range of specific contaminant
concentrations which a consumer might experience as a result of
the interaction of the water and the materials comprising the
water delivery system. In addition, the Guidance provides
criteria to determine when additional actions might be appro-
priate. The Agency realizes that, due to large variations in
water quality and distribution materials in any water supply
system, implementation of sampling protocol should be done on
a case-by-case basis. The protocol below is a suggested model
that States and utilities may use in its entirety or in part
as determined by specific situations.
Guidance for the Selection of Appropriate Sample Sites
Sample sites for monitoring should be established at locations
where the potential for consumer exposure to a specific con-
taminant would be greatest. For instance, monitoring for lead
should be done at sample sites in the distribution system where
the presence of lead pipes or service has been verified. The
rationale for problem specific site selection as opposed to random
sampling, is that corrosive by-products (i.e., lead, zinc, copper
and others) are not normally found in distribution waters in high
concentrations except where they are being dissolved from piping
materials due to corrosion. Table 1 provides information on the
type of contaminants associated with various delivery system con-
struction materials. Table 2 contains a suggested ratio of sampling
sites to population served. The design of an appropriate sampling
regimen also needs to reflect local conditions including hydraulic
configuration, water aggressiveness, sampling costs, analytical
costs and other factors. In instances where the presence of a
material in the distribution is known to be localized, the ratio
of sample sites to population served should be adjusted to reflect
the portion of the distribution system affected, rather than the
whole water supply system. In addition, sections of the distribu-
tion system that are served by multiple sources of different water
quality should be treated as a separate entity for sampling purposes
to account for variations in water quality in the affected portions
served by the various sources.
161
-------
Table 1
Recommendations for Sampling Parameters
The following parameters are suggested for analysis for
different types of pipe material:
Pipe Material Parameter of Interest
Ductile and Cast Iron (Unlined) iron
Copper Copper, Lead from Solder
Joints
Lead Lead
Galvanized Steel Zinc, Cadmium, Lead, Iron
Copper Alloys Copper, Zinc, Lead
Note: Asbestos fibers from asbestos cement pipe was not included
as a suggested parameter. Monitoring asbestos cement pipe
deterioration is best done in the distribution mains rather
than by the sampling program suggested in this guidance.
Table 2
Recommended Number of Sampling Sites to be Monitored
on a Quarterly Basis for the First Year
Population Served Number of Sampling Sites
<100 1
101 - 1000 2
1001 - 5000 5
5001 - 10,000 7
>10,001 10
Note: It is expected that localized variables affecting
corrosion are more numerous in larger distribution systems
that in smaller ones. For this reason, proportionately
more sample sites are recommended for larger water systems
in order to better account for these variables.
162
-------
The water purveyor should use discretion in determining the
need for monitoring for specific contaminants. Where previous
records and experience has indicated that the water is not
corrosive to a specific material or that certain piping materials
have not been used, the purveyor should modify his sampling and
analytical program to reflect that information.
Once the number and type of samples to be analyzed are estab-
lished, the water supplier can rank the location of points to
be sampled using the following suggested order of priority:
1. Locations where corrosion problems are suspected due to
consumer complaints or other indications,
2. Buildings with interior plumbing containing specific
materials of interest such as lead, galvanized steel,
copper, etc.,
3. Service lines containing specific materials of interest
such as lead, galvanized steel, copper, etc.,
4. Recently constructed buildings served by the utility, and
5. Several randomly selected control sites throughout the
distribution system.
For example, in lead monitoring, a water supplier would sample
for lead at locations meeting priority "1" first. If not enough
buildings are available with appropriate interior plumbing, he
would determine the remaining locations using priority "2", "3",
"4" and/or "5". Sample sites to monitor other contaminants such as
iron, copper, zinc and cadmium should be established in a
similar manner. The suggested order of priority was established
on the basis of maximum potential exposure to the consumer.
Priorities "1", "2" and "3" are self explanatory. Priority "4"
was established on the basis of field data which indicates that
corrosion rates are much higher in the newly constructed plumbing
than older ones.
There are a number of options available for the utilities to
perform a survey to locate and identify specific sample sites.
Consultation of records and archives may be one of the best
approaches for some larger and older systems to determine the
general location and presence of contamination sources. This
approach will reduce manpower and time requirements in the
field to pinpoint the exact location of the source of contamina-
tion. Often, the local or county health department might have
already conducted a survey involving health-related contaminants
in the water as a result of distribution system corrosion.
163
-------
An alternative approach is to have utility personnel, such as
meter readers or meter repairmen develop information as a
routine part of their normal jobs. It is usually possible to
determine the type of material used for service lines by checking
the connections at the meter. Another method which could be
utilized is to mail questionnaires. However, it should be noted
that such an approach usually produces poor results, since only
a small portion of the customers may reply, and the information
provided may not be reliable. The surveys could be conducted
by initiating a direct house-to-house search for the identity,
type and source of materials in the distribution system. This
latter approach would only be attractive for small utilities
having few connections.
Guidance for Collection of Samples for Analysis*
To determine the representative range of contaminant levels to
which the consumer is exposed during the course of daily activi-
ties, the samples should be collected early in the morning. The
samples should be drawn from interior faucets to get water that
has been in intimate and sustained contact with the interior
plumbing, the service line and water main as follows:
Sample 1 - The sample reflecting interior plumbing condition is
collected immediately upon opening the faucet. This represents
water that usually has been standing in the fixture and interior
plumbing for a 6 to 8 hour period.
Sample 2 - The sample reflecting service line condition is
collected after the water has been running and the sample
collector feels the water temperature change from warm to cold.
Since water would be expected to warm slightly after standing
in interior plumbing, this cold water represents water that
had been standing just outside the foundation of the house in
contact with the underground service line for the same 6 to 8
hour period.
Sample 3 - The sample representative of the water main is
collected after allowing the water to run for at least three
minutes after the initial temperature change was noted. This
water would have had a minimum contact time with the service
line and interior plumbing and probably represents the ambient
distribution water quality. The length and size of pipes in
the building should be considered to determine if the 3 minute
contact criterion is applicable.
*Karalekas, P. C., Jr.f Ryan, C. R., Larson, C.f Taylor, F.,
"Alternative Methods for Controlling the Corrosion of the
Lead Pipe." Paper presented at the Annual Conference of the
New England Water Works Association, Boston, Massachusetts, 1977,
164
-------
This method of sample collection provides an indirect profile
of consumer exposure to corrosion related contaminants. Water
that has been stationary in the interior plumbing overnight has
had maximum contact time with the source of contamination and
represents water that may be drunk by the consumer early in
the morning. The second sample represents the water that was
standing in the service line and may represent water used for
preparation of breakfast beverages such as coffee or tea. The
third sample, taken after running the water, represents water
that is commonly drawn during the same day when water is more
frequently used. It is obvious that no single sampling protocol
can represent every scenario of exposure. The suggested proto-
col is an attempt to estimate typical exposure potential. Utili-
ties and States are encouraged to use other protocols if a specific
exposure situation can be sampled with more confidence. Further
study is required before a more accurate universal protocol can
be articulated.
Recommendations for Sampling Frequency and Interpretation of
Results"
To account for seasonal variations in water quality affecting
corrosion, the sampling should be repeated at least quarterly
for a period of one year. This sampling frequency may be reduced
in instances where the water obtained is solely from the
ground and it has been established that the water quality is
constant through the year. In reviewing these quarterly samples
and in making determinations about corrective action, the water
supplier should consider any first draw sample or any average
value (average of 3 samples taken at one location) that are
greater than the ambient or background levels for the particular
contaminant being evaluated. Neither the first draw sample
nor the average values of the 3 samples taken at each location
would constitute an MCL violation per se, but they would give
good indications of the possibility that an MCL is being exceeded
for some portion of the time and that a part of the population
is being exposed to elevated levels of the corrosion product.
It may be that a reassessment of the MCL compliance monitoring
regimen should be undertaken to assure that the monitoring program
being used is indicative of the quality of water being consumed
in those parts of the system that are being affected by corrosive
waters.
If the above sampling results show that the highest average
concentration from any one set of the four seasonal samples
at each sampling point is below MCL values, reducing the fre-
quency of sampling at each site to one sample per year is justi-
fied, but this sample should be obtained during the seasonal
period that revealed the greatest corrosion induced contamina-
tion. If the above sampling survey results show that the
average concentration of contaminants from any one set of the
165
-------
four seasonal samples at any sampling point exceeds the MCL
(or other level of concern for unregulated contaminants) re-
medial action should be contemplated. Such actions might
include treatment additions or modifications to reduce the
corrosivity of the water, the use of corrosion inhibiting
chemicals, or changes to piping materials to alleviate the
health risk associated with isolated local problem areas.
Quarterly sampling should continue until the average concen-
tration of contaminants is known to be consistently below the
MCL (or other level of concern for unregulated contaminants.)
A maximum contaminant level is defined to be, "...The maximum
permissible level of contaminant in water which is delivered to
the free flowing outlet of the ultimate user of a public water
system...contaminants added to the water under circumstances
controlled by the user, except those resulting from corrosion
of piping and plumbing caused by water quality,are excluded
from this definition." MCLs have been established for lead and
cadmium in the National Interim Primary Drinking Water Regu-
lations, and for copper, zinc and iron in the National Secondary
Drinking Water Regulations. These MCLs define the level of
contaminant at which corrective actions are appropriate and
required. However, the wording of the Safe Drinking Uater Act
and the legislative history that accompanied the Act direct
that any hunan exposure to potentially harmful contaminants
should be minimized to the extent feasible. Concentrations
of corrosion by-products such as lead and cadmium can reach
levels in excess of the MCL in very short periods of time (40
minutes or less in some studies).* Therefore, even though the
average values obtained using the previously discussed sampling
protocol or samples taken in accordance with the minimum require-
ments set out in the National Interim Primary Drinking Water
Regulations do not technically exceed the MCL, any elevated
contaminant levels indicate the potential for excessive exposures
and prudent public health policy would dictate corrective or
remedial actions.
*Peacock, Stuart J., "Factors Influencing Household Water Lead:
A British National Survey". Archives of Environmental Health,
Vol. 35, No. 1, Jan-Feb 1980, pp. 45-51.
Briton, A., "Factors Influencing Plumbosoluency in Scotland",
Journal of the Institution of Water Engineers and Scientists,
Vol. 35, July 1981, pp. 350-364.
166
-------
REPORT: PANEL SESSIONS
SUMMARY OF CONCLUSIONS AND RECOMMENDATIONS
by: Frank Bell
Office of Drinking Water
U.S. EPA
401 M Street, S.W.
Washington, D.C. 20460
BACKGROUND
The panel sessions of the seminar were divided into four subject areas:
1. Joining Alternatives for Copper Pipe
2. Metal Pipe and Fitting Alternatives for Plumbing
3. Plastic Pipe and Fittings
4. Regulatory and Compliance Aspects.
Charges to each panel followed the general framework of assessing the
current state of knowledge, identifying problems, assessing alternative
solutions, and recommending future steps to cope with the identified
problems.
KEY CONCLUSIONS AND RECOMMENDATIONS
The following conclusions and recommendations represent a sense of the
meeting in the important areas of concern, especially emanating from the
panel discussions.
LEAD/TIN SOLDER
Lead is present in drinking waters, both hard and soft, due to
corrosion of lead/tin solder. New installations, which also may contain
particles of lead solder and flux, and supplies with soft, acidic water may
be the worst-case situations.
167
-------
Several state and local jurisdictions have acted and others are
considering acting to ban lead/tin solder from potable water plumbing.
There were strong differing opinions and no consensus could be reached
regarding the desirability of banning the future use of lead/tin solder.
Alternative joining materials such as tin/silver or tin/antimony solder are
available should lead/tin solder be replaced.
Water treatment methods are available to reduce the problem in certain
areas. Since poor plumbing practice appears to be involved in several
incidents of excess potable water lead levels, improvements in plumbing
practice through communication and training can have a role in reducing the
problem. In addition the use of chemical flushing and preforms may be
useful in reducing lead contamination levels.
Recommendations;
a. While the occurrence of high lead levels attributable to lead/tin
solder was accepted, it was recommended that more study and
analysis of data were required to assess the_ extent and severity of
lead contamination in a variety of water qualities.
b. Additional investigative and research work were also recommended to
develop a standardized sampling method and to refine water quality
treatment and control measures.
LEAD PIPE AND FITTINGS
The use of lead pipe for potable water has been abandoned by most
jurisdictions throughout the United States with the exception of the City
of Chicago, where the plumbing code requires that service lines of 2 inches
or less in diameter be lead pipe. Lead levels are elevated in water
standing in lead
pipes, more so in low-pH, low-alkalinity waters but also in a variety of
other waters. Since there are economic alternative materials, including
copper, galvanized steel, and plastic (currently used in many cities as
shown in an EPA survey of April 1984), there is no logical requirement for
the continued use of lead pipe and fittings.
Recommendation: The majority recommendation was that the future use of
lead pipe and fittings should be discouraged or banned.
TREATMENT OR WATER QUALITY ADJUSTMENT
A basic paper was presented on techniques of water treatment to
minimize corrosion. In the panels on joining alternatives for copper pipe
and on metal pipe and fittings, a repeated theme emphasized the problems of
controlling leaching from existing systems. Corrosive waters will increase
the rate of leaching from lead pipes, a residual of which exists in many
communities, and from copper pipe joined by lead/tin solder. Excess
168
-------
contaminant leaching has been associated with a variety of water
qualities. Since simple pH adjustment may not be appropriate for all
waters, more study and research are needed on treatment techniques to
reduce corrosion.
Recommendations;
a. All public water utilities should review their water quality and
potential corrosion problems and should take steps to make their
water quality less corrosive where elevated contaminant levels
occur.
b. Additional study and research should be conducted to define the
occurrence of corrosion in a spectrum of water qualities and to
develop appropriate anticorrosion treatment methods.
MONITORING FOR CORROSION BYPRODUCTS
Several panels pointed out questions with respect to corrosion
sampling techniques and the need for developing a standardized approach.
When EPA's Water Supply Guidance 73 (WSG 73) was passed out at one panel
meeting, concern was expressed that it had not been distributed widely
enough and was virtually unknown in many communities. EPA also promulgated
corrosion monitoring regulations in 1980 and their results remain to be
determined.
EPA was assigned responsibility for defining the corrosion problem and
for developing primary Maximum Contaminant Levels (MCLs) and monitoring
requirements to regulate corrosion byproducts. The state's responsibility
was defined as being the enforcement agent over the water purveyor. The
water purveyor's responsibilities were defined to include implementing
corrosion monitoring in a uniform manner, adjusting water quality to
minimize corrosion, and rendering a safe quality water to consumers.
Recommendations; EPA should define the corrosion problem and develop
primary MCL and monitoring requirements to regulate corrosion byproducts.
Development of a standardized sampling procedure that would help identify
distribution system corrosion problems and a national survey of corrosion
byproducts utilizing the standardized sampling procedure are recommended.
PLASTIC PIPE AND FITTINGS/LEACHING
The leaching of toxic substances from plastic pipe to drinking water
has been the subject of long-term standards and monitoring under the
National Sanitation Foundation (NSF) third-party program. Few problems and
no adverse health risk have been noted in over 20 years of surveillance for
covered products. However, not all byproducts of plastic manufacture may
have been covered in the NSF program: a minority felt the need for further
research, particularly with respect to some materials more than others.
169
-------
Contamination of drinking water from the use of solvents and joining
compounds in the field assemblage of some plastic pipes represents a
significant concern. There was also discussion on the potential of greater
than necessary leaching from the misuse of incorrect or "multipurpose"
compounds. The multipurpose compound would be used when two pipes of
different materials are joined; otherwise, only the material specified for
a particular type of pipe would be used. Apparently a lack of materials
identification and incomplete use instructions from suppliers may be
responsible for much of the misuse of multipurpose joining compounds.
Recommendations; There was majority agreement that there is a problem of
leaching from joining compounds and that more information is needed on the
health risks that this creates. Specific recommendations were to
strengthen third-party agreements and monitoring; to review state, local,
and model codes; and to review and improve installation techniques. The
industry was urged to review the compounds and solvents used in the
manufacture and installation of plastic products.
PLASTIC PIPE AND FITTINGS/PERMEATION
Research has shown that plastic pipe and fittings are permeable to
organic chemicals such as gasoline and trichloroethylene and that drinking
water contamination has occurred in several instances where such chemicals
have come into contact with plastic pipe. Research has also shown that
asbestos/cement pipe and the gaskets used for joining ductile iron pipe are
permeable to some organic chemicals.
While the extent of contamination from the permeation of plastic pipe
appears to be limited, there was general agreement that it does occur and
that plastic pipe should not be used in areas subject to organic chemical
contamination. California does have a health department regulation
prohibiting the use of plastic pipe under certain conditions considered to
pose a possible adverse health effect. Plastic industry representatives
stated that they are working on development of nonpermeable pipe
materials. There are also new techniques being developed for improved
installation, such as soil barriers, that might be helpful in preventing
the permeation of plastic pipe.
Recommendations; There were general recommendations that the use of
plastic pipe be limited in areas of existing or potential contamination;
that industry or other bodies inform potential users regarding limitations
of products relating to permeation, and that further investigation should
be made into the development of nonpermeable pipe or other means of
preventing pipe permeation.
INFORMATION TRANSFER AND COMMUNICATION
From its inception, this seminar represented a first meeting of many
varied interests in which government, industry, plumbing experts, etc.,
took an active part in considering drinking water corrosion contamination
170
-------
problems and possible solutions. A common observation'was that
insufficient information transfer and dialogue had occurred in the past,
particularly between the governmental water supply interests and the
industrial and plumbing sectors.
Further, it was felt that insufficient time was available at this
seminar and that the subjects covered were too broad to allow adequate time
for meaningful discussion and resolution of issues.
Recommendations; There was a general recommendation for more study and
deliberation in the substantive areas and for subsequent meetings planned
with more focussed discussions and more time allowed. A further
recommendation was made for more formal communication such as articles,
papers, joint small discussions, etc., on plumbing materials and water
quality, so that problems and solutions can be properly communicated among
all the varied interests.
171
-------
SUMMATION - PLUMBING MATERIALS AND
DRINKING WATER QUALITY
by: James M. Symons
Department of Civil Engineering
University of Houston
Central Campus
4800 Calhoun
Houston, Texas 77004
The Workshop was opened by Cotruvo, who introduced the subject and
gave the themes of the program: to permit the airing of views and to
provide directions for the future. Cotruvo suggested that the Workshop
focus on three questions:
1. Do problems exist?
2. What technical solutions are available?
3. What are the roles of the various organizations?
I. DO PROBLEMS EXIST?
A. Materials
1. Copper
Anderson reported that copper piping has several good
features: it is popular, functions well, has a good program
of quality assurance, is based on the eddy current method,
and is available during manufacture; also, the industry has
developed techniques that remove the drawing chemicals. He
noted, however, that copper pipe does give some problems;
for instance, pitting can cause leakage, and corrosion will
occur if utilities do not treat their water properly. This
is particularly true in small utilities, where treatment is
less sophisticated.
2. Plastic
Plastics were discussed by Mruk and Podoll, who noted that
the favorable features of this material are its immunity to
corrosion, the smooth surface of the piping material, the
172
-------
lack of need for soldering joints, and the flexibility of
the pipe. Mruk pointed out some unfavorable factors,
however, such as problems with jointing materials, the
additives in the plastic pipe, its susceptibility to high
temperatures, its combustibility, and the reduction in
strength as the duration of loading increases. Podoll
studied chlorinated polyvinyl chloride (CPVC) pipe and found
that a few volatile organic compounds leached out at the 1
to 10-microgram/liter (ug/L) concentration. He also noted
that some jointing compounds were found during the leaching
test initially, but then were rinsed out. Organo-tin
initially leached at a concentration of 0.5 to 3 ug/L, but
the concentration fell to 0.1 ug/L after 3 weeks.
3. Lead
None of the speakers or the panelists had many favorable
comments to make about lead pipe, but Karalekas reported on
severe lead problems in the New England area. He pointed
out that even 5.5 percent of running samples exceed the
regulated lead concentration. Karalekas discussed his
sampling protocol of a standing sample, a short-term running
sample, and a long-term running sample; several other
speakers and panelists mentioned the importance of
standardizing these sampling procedures as they do affect
the overall metal concentration in the resulting water
sample.
4. Galvanized Pipe
Neff pointed out that new pipe dissolves more zinc into the
water than pipe that has been in place for a period of
approximately 2 years. Neff also pointed out that
chrome-plated brass valves, such as are found in the
household, produce metal concentrations in water; he
suggested that studies on piping systems should take into
account the contribution of metals from the valves and
faucets.
5. Solders
Murrell presented data from a study on Long Island, where the
authors found that 67.3 percent of the piping systems had
solders containing from 55 to 65 percent lead content. They
found that this solder leaches metals on standing and that in
new homes, very high lead levels were initially found. They
also pointed out that slugs of water with high concentrations
of lead occurred periodically as water is drawn through a
plumbing system. As water that has been in contact with a
soldered joint comes out of the tap it will contain leached
metals.
173
-------
B. Penetration Problems
Pfau and Goodwin reported on the penetration of several solvents
through jointed and unjointed ductile iron, polyvinyl chloride
(PVC), and asbestos/cement (A/C) pipe. When the ground was
saturated with solvents only the solid ductile iron pipe resisted
penetration. All joints permitted the penetration of solvent
into the water and solid A/C pipe passed the solvent quite
easily. Solid PVC pipe also passed the solvent, but only after
32 days. Goodwin reported similar data when the pipe was
challenged with approximately 300 mg/L of solvent.
C. Importance of Water Chemistry and Environment
Anderson noted that copper pipe would corrode if water velocities
are greater than 5 ft/sec, the water has a high concentration of
carbon dioxide, and the water is soft. Karalekas reported the
impact of soft water and low pH on lead solution. Neff indicated
that with galvanized piping, chlorine residual had an influence
on metal loss as did low pH, and commented that polyphosphates
may aggravate the metal loss situation. Murrell noted that low
pH and soft water caused increased problems from soldered joints.
II. WHAT TECHNICAL SOLUTIONS ARE AVAILABLE?
A. Water Treatment
Schock reviewed the impact of pH, carbonate concentrations,
orthophosphate, polyphosphate, and silicate on lead, copper, and
zinc. Schock pointed out that the solubility of lead actually
goes up as the carbonate concentration goes up, a find that might
not be intuitively obvious to the unsuspecting. He pointed out
that many people use the Langelier Index to calculate calcium
carbonate deposition, but that several factors in the Langelier
Index are based on rather old data; therefore, the actual pH of
saturation is frequently a few tenths of a pH unit different than
calculated by the traditional Langelier Index. He also pointed
out that polyphosphates are quite different from orthophosphate
and that in some cases polyphosphates may actually cause higher
metal content in water. Finally, he noted that silicates often
produce a favorable effect on galvanized pipe. Karalekas showed
several case histories in which the benefits of increasing pH had
a dramatic effect on the lead content in New England waters. The
importance of proper water treatment was also discussed in the
panels.
174
-------
B. Regulations
Several alternative techniques (other than water treatment) for
controlling corrosion byproducts in water were suggested by
various speakers. For instance, Murrell recommended that lead
solder be banned and alternatives such as silver or tin-antimony
be used to avoid the dissolution of metal from the solder.
Boydston discussed altering plumbing codes as a way of
controlling the resultant water quality in household distribution
systems. He also pointed out that, in the absence of Federal
regulations, state and local ordinances could be passed to
control the use of certain plumbing materials. Wagner thoroughly
reviewed European practice, pointing out that many countries in
Europe had regulations restricting the use of materials that
would have a deleterious effect on water, for instance lead pipes
and lead solder. He also pointed out that plastic pipe material
was controlled by statute in many European countries. Another
approach to control would be Federal regulation by the U.S. EPA.
Currently many metals are regulated, either as primary or
secondary drinking water regulations and the corrosivity of water
should be measured. Vigorous corrective action through
enforcement after finding corrosive water would help reduce the
insult to the consumer from dissolution of metal. The importance
of education as a method of improving the current situation was
pointed out by the panels.
C. Third-Party Review
McClelland reported on the extensive review, inspection, and
testing protocol of the National Sanitation Foundation (NSF) for
plastic pipe. By investigating the manufacturing products that
went into pipe and the manufactured product itself, the NSF is
able to ensure that pipe containing their seal meets rather
rigorous standards. In any manufacturing product quality control
is important. Mruk, for example, told of subsequent problems
that resulted from a plastic pipe manufactured without a
necessary additive.
No matter which approach, or combination of approaches, is used
for controlling the deterioration of water quality during passage
through distribution piping and plumbing systems, inspection and
enforcement is essential to make this system work. For instance,
during the Workshop examples were given of plumbers who switch
from more desirable to less desirable solders when inspectors are
not looking; of water utilities that will not treat water to the
degree recommended by the consulting engineer; of pipe that may
be sold in an area where it is unsuitable (problems occurred some
years ago with asbestos/cement pipe in soft water areas; and of
installation practices not properly followed, for example, the
kinking of plastic pipe in an effort to have it bend beyond its
traditional flexibility.
175
-------
III. WHAT ARE THE ROLES OF THE VARIOUS ORGANIZATIONS?
All three parties represented at this Workshop have an important
interest in this subject. The manufacturers want their products to
enjoy a good reputation, both with the public and with the
regulators, so they can sell their product. The regulators want to
be able to offer users alternative piping materials and they want to
protect the health of the public. The independent service groups
such as the NSF provide an essential link between the other two
parties.
The Workshop was successful in generating a dialogue among the three
parties mentioned above. Only through cooperative effort, such as
the one demonstrated in this Workshop, can the ultimate goal of safe
water for the consumers be realized.
176
-------
APPENDIX A
LIST OF PARTICIPANTS
SEMINAR ON PLUMBING MATERIALS AND DRINKING WATER QUALITY
Moderators
John Courchene
Seattle Water Dept.
1509 South Spokane Street
Seattle, WA 98144
206-625-4305
James Manwarlng
American Water Works Association Research Foundation
6666 West Quincy Avenue
Denver, CO 80235
303-794-7711
James J. Westrick
Chief, Water Supply Technology Branch
Technical Support Division
Office of Drinking Water
U.S. EPA
26 West St. Clalr Street
Cincinnati, OH 45268
513-684-7908
Speakers
Paul Anderson
Vice President, Building Construction Markets
Copper Development Association, Inc.
Greenwich Office Park 2
Box 1840
Greenwich, CT 06836
203-625-8223
Frank Bell
Office of Drinking Water
U.S. EPA
401 M Street, S.W.
Washington, DC 20460
202-382-3034
177
-------
James Boydston
Manager, Drinking Water Program
State Health Division
P.O. Box 231
Portland, OR 97207
503-229-6302
Dr. Joseph Cotruvo
Office of Drinking Water
U.S. EPA
401 M Street, S.W.
Washington, DC 20460
202-382-7575
Donald Goodman
Manager, Polymer R&D
Tenneco Polymers, Inc.
R.D. 6, Box 177
Flemington, NJ 08822
201-782-4011
Peter C. Karalekas, Jr.
Water Supply Branch
U.S. EPA, Region I
J.F. Kennedy Federal Building
Boston, MA 02203
617-223-3973
Francis T. Mayo
Director, MERL
U.S. EPA
26 West St. Glair Street
Cincinnati, OH 45268
513-684-7951
Dr. Nina I. McClelland
National Sanitation Foundation
3475 Plymouth Road
P.O. Box 1468
Ann Arbor, MI 48106
313-769-8010
Stanley A. Mruk
Plastics Pipe Institute
c/o SPI
355 Lexington Avenue, 6th Floor
New York, NY 10017
212-573-9400 X420
178
-------
Norman E. Murrell
H2M Consulting Engineers
125 Baylis Road
Suite 140
Melville, NY 11747
516-752-9060
Chester H. Neff
Illinois State Water Survey
605 E. Springfield Avenue
Box 5050, Station A
Champaign, IL 61820
217-333-4954
James P. Pfau
Group Leader, Coating Science
Polymer Science and Technology Section
Battelle Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Dr. R. Thomas Podoll
SRI International
333 Ravenswood Avenue
Menlo Park, CA 94025
415-859-5834
Michael R. Schock
Illinois State Water Survey
605 E. Springfield
Box 5050, Station A
Champaign, IL 61820
217-333-9322
Dr. James M. Symons
Dept. of Civil Engineering
University of Houston
Central Campus
4800 Calhoun
Houston, TX 77004
713-749-1703
Ivo Wagner
Lehrstuhl fur Wasserchemie der
Universitat-Willstatter-Allee 5
University of Karlsruhe
Postfach Nr. 6380
Karlsruhe, Federal Republic of Germany
011-49-721-34893
179
-------
PANEL 1. JOINING ALTERNATIVES FOR COPPER PIPE
Chairman
Dr. Rhodes Trussell
James M. Montgomery,
Consulting Engineers
555 East Walnut Street
Pasadena, CA 91101
213-796-9141
Recorder
Peter Karalekas
Water Supply Branch
U.S. EPA, Region I
J.F. Kennedy Federal Building
Boston, MA 02203
617-223-3973
Panelists
Richard Ballentine
(The Silver Institute)
J.W. Harris Co.
10930 Deerfield Road
Cincinnati, OH 45242
513-891-2000
John Courchene
Seattle Water Dept.
1509 South Spokane Street
Seattle, WA 98144
206-625-4305
Vince Doyle
Executive Director, Plumbing and Air
Conditioning Contractors of Arizona
1616 East Maryland Avenue
Phoenix, AZ 85016
602-277-2634
Allen Hammer
Bureau of Water Supply
Engineering
Virginia State Health Dept.
109 Governor Street, 9th Floor
Richmond, VA 23219
804-986-1776
Bill Hampshire
Tin Research Institute, Inc.
1353 Perry Street
Columbus, OH 43201
614-424-6200
A.C. Kireta
Copper Development
Association, Inc.
P.O. Box 515
Merrimack, NH 03054
603-429-0376
Jerome F. Smith
Lead Industries
Association, Inc.
292 Madison Avenue
New York, NY 10017
212-578-4750
EPA Resource Personnel
Marvin Gardels
Drinking Water Research
MERL
U.S. EPA
26 West St. Glair Street
Cincinnati, OH 45268
513-684-7236
180
-------
PANEL 2. METAL PIPE AND FITTING ALTERNATIVES FOR PLUMBING
Chairman
Dr. Edward Singley
Environmental Science and
Engineering, Inc.
P.O. Box ESE
Gainesville, FL 32602
904-332-3318
Recorder
Frank Bell
Office of Drinking Water
(WH-550)
U.S. EPA
401 M Street, S.W.
Washington, DC 20460
202-382-3037
Panelists
James R. Boates
Boston Plumbing Co.
15 Ridge Avenue
Natick, MA 01760
617-542-3777
Dodd S. Carr
Manager, Chemistry and Patents
International Lead Zinc Research
Organization, Inc.
292 Madison Avenue
New York, NY 10017
212-532-2373
Dr. Anthony Colucci
Colucci and Associates
17705 I Hale Avenue
Suite 2
Morgan Hill, CA 95037
408-778-1100
Larry Galowin
Center for Building Technology
National Bureau of Standards
Washington, DC 20234
202-921-3293
Joe Glicker
Bureau of Water Works
City of Portland
1120 S.W. 5th Avenue
Portland, OR 97204
503-796-7471
Stan Mruk
Plastic Pipe Institute
c/o SPI
355 Lexington Avenue
6th Floor
New York, NY 10017
212-572-9400 X420
Robert A. Wilson
Copper Development
Association
P.O. Box 839
Purcellville, VA 22132
703-338-5300
181
-------
PANEL 3. PLASTIC PIPE AND FITTINGS
Chairman
Gary Englund
Water Supply and General Engineering
Minnesota Division of
Environmental Health
715 Delaware Street, S.E.
Minneapolis, MN 55440
612-296-5330
Recorder
Harry Von Huben
Water Supply Branch
U.S. EPA, Region V
230 S. Dearborn Street
Chicago, IL 60604
312-353-2151 „
Panelists
Tom Adams
Adams, Broadwell and Russell
400 South El Camino Real
San Mateo, CA 94402
415-342-1660
Daniel W. Hurley
Associated Lead, Inc.
2545 Aramingo Avenue
Philadelphia, PA 19125
215-427-3000
Ramon G. Lee
American Water Works
Service Co., Inc.
500 Grove Street
Haddon Heights, NJ 08035
609-546-0700
Nina McClelland
National Sanitation
Foundation
3475 Plymouth Road
Ann Arbor, MI 48106
313-769-8010
Alan J. Olson
Manager, Industry Affairs
B.F. Goodrich Co.
Chemical Group
6100 Oak Tree Boulevard
Cleveland, OH 44131
216-447-6132
David Spath
California State Department
of Health
2151 Berkeley Way
Berkeley, CA 94704
415-540-2172
EPA Resource Personnel
Roy Jones
Water Division
U.S. EPA, Region X
1200 6th Avenue
Seattle, WA 98101
206-442-1900
Koge Suto
Water Supply Branch
U.S. EPA, Region III
6th and Walnut Streets
Philadelphia, PA 19106
215-597-2786
182
-------
PANEL 4. REGULATORY AND COMPLIANCE ASPECTS
Chairman
Raymond Jarema
Chief, Water Supplies Section
Preventable Diseases Division
Connecticut Department of Health
Services
Hartford, CT 06106
203-566-1253
Recorder
Charles D. Larson
Chief, Technical Assistance
Section
Water Supply Branch
U.S. EPA, Region I
J.F. Kennedy Federal Building
Room 2113
Boston, MA 02203
617-223-6486
Panelists
Charles Buescher, Jr.
Chairman of the Board
St. Louis County Water Co.
535 North New Ballas Road
St. Louis, MO 63141
314-997-3404
Donald Dickerson
Donald Dickerson Associates
6901 Hayvenhurst Avenue
Van Nuys, CA 91406
213-989-0505
Hugh F. Hanson
President, Consumers/Pirnie Utility
Services Company
2 Corporate Park Drive
White Plains, NY 10602
914-694-2182
Richard B. Howell, III
Office of Sanitary
Engineering
Bureau of Environmental
Health
Dover, DE 19901
302-736-4731
James Sargent
Director, Bureau of Plumbing
Wisconsin Dept. of Industry,
Labor and Human Relations
1852 Wisconsin Avenue
Sun Prairie, WI 53590
608-266-8984
183
-------
EPA Resource Personnel
Peter Lassovszky
Science and Technology Branch
Office of Drinking Water (WH-550)
U.S. EPA
401 M Street, S.W.
Washington, DC 20460
202-382-3030
Arthur H. Perler
Office of Drinking Water
U.S. EPA
401 M Street, S.W.
Washington, DC 20460
202-382-3040
184
-------
Attendees
Barry Back
Circuit Rider
Kentucky Rural Water Assoc.
P.O. Box 1424
Bowling Green, KY 42102-1424
502-843-2291
Charles R. Bell
Manager, Field Operations
Division of Public Water
Supply
Illinois Environmental
Protection Agency
Springfield, IL 62706
217-785-0561
Gerald Boland
Supervisor, Inspection Office
Water Works Engineering
4747 Spring Grove Avenue
Cincinnati, OH 45232
513-352-4653
Vincent Bower
Vice President, Standard
Products Div.
Mueller Brass Co.
1925 Lapeer Avenue
Port Huron, MI 48060
313-987-4000
Herb Brass
TSD
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7936
Gary Cahall
Salesman
Cerro Copper Products Company
7557 South 78th Avenue
Bridgeview, IL 60458
312-594-7600
Keith Carswell
DWRD
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7228
Michael J. Cassady
Researcher
Battelle Columbus Labs
505 W. King Avenue
Columbus, OH 43201
614-424-5568
Robert Clark
DWRD
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7228
A. Cohen
Manager, Standards/Safety
Engineering
Copper Development Assoc.
Inc.
Greenwich Office Park 2,
Box 1840
Greenwich, CT 06836
203-625-8232
James S. Crall
Superintendent of
Laboratories
Orlando Utilities Commission
P.O. Box 3193
Orlando, FL 32802
305-423-9179
James Cummings
Chief, Plumbing Division
State of Michigan
7150 Harris Drive
Lansing, MI 48909
517-322-1804
185
-------
George A. Cypher
Technical Director -
Chemistry
International Copper
Research Association, Inc.
708 Third Avenue
New York, NY 10017
212-697-9355
Thomas H. Danison
Researcher
Battelle Columbus Labs
505 W. King Avenue
Columbus, OH 43201
614-424-5599
Bill Davis
Environmental Engineer
Region VI, Water Supply
Branch
U.S. EPA
1201 Elm Street
Dallas, TX 75270
214-767-1762
Jack DeMarco
Supervisor, Water Research
Cincinnati Water Works
5651 Kellogg Avenue
Cincinnati, OH 45228
513-231-7825
Richard Eilers
DWRD
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7809
Bernard J. Enright
Field Representative
American Iron & Steel
Institute
1000 16th Street, N.W.
Washington, DC 20036
Samuel F. Etris
Senior Consultant
The Silver Institute
Suite 1138
1001 Connecticut Avenue
Washington, DC 10036
202-331-1227
Kim Fox
DWRD
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7236
Thomas M. Franklin
Sanitary Engineer
Penn. Dept. of Environmental
Resources
P.O. Box 2357
Harrisburg, PA 17120
717-783-3795
Ronald A. Frano
Executive Director
Plastics Pipe Institute
355 Lexington Avenue
New York, NY 10017
212-573-9418
Ed Frindt
DWRD
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7236
Bud Gardiner
Executive Vice President
Gardiner Metal Co.
4820 S. Campbell Avenue
Chicago, IL 60632
312-847-0100
Harold L. Garrison
Water Quality Supervisor
Kentucky American Water Co.
P.O. Box 7500
2300 Richmond Road
Lexington, KY 40502
606-269-7859
186
-------
Edwin E. Geldreich
Chief Microbiologist
Drinking Water Research Div.
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7232
Betsy Gibson
Administrative Assistant
P.J. Higgins & Assoc.
P.O. Box 97
Damascus, MD 20872
301-963-0493
Dillard Griffin
Water Quality Superintendent
Kentucky American Water Co.
P.O. Box 7500
2300 Richmond Road
Lexington, KY 40502
606-269-7859
Richard C. Haenke
Product Manager/P&H Div.
Mueller Brass Co.
1925 Lapeer Avenue
Port Huron, MI 48060
313-987-4000
David G. Hansen
Environmental Engineer
Dept. of Natural Resources -
Horicon Hdqtrs.
1210 N. Palmatory Street
Horicon, WI 53032
414-485-4434
William S. Harris
Vice President Administration
American Water Works Service
Co., Inc.
525 Grove Street
Haddon Heights, NJ 08035
609-547-3211
David Hartman
Water Research Chemist
Cincinnati Water Works
5651 Kellogg Avenue
Cincinnati, OH 45228
513-231-7825
Steve Hathaway
DWRD
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7453
Robert M. Henson
Technical Services Manager
J.W. Harris Co., Inc.
10930 Deerfield Road
Cincinnati, OH 45242
513-891-2000
David W. Heumann
Asst. Water Quality Engineer
Department of Water & Power
City of Los Angeles
P.O. Box 111, Rm. A-18
Los Angeles, CA 90051
213-481-7752
Pat Hilgard
Toxicologist
Office of Toxic Substances
U.S. EPA (TS796)
401 M Street, S.W.
Washington, DC 20460
202-382-3491
James Ho
EMSL
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7321
Marvin E. Holmgren
Vice President Manufacturing
Elkhart Products Corporation
1255 Oak Street
Elkhart, IN 46514
219-264-3181
187
-------
Kenneth M. Howard
Environmental Program
Coordinator
Commonwealth of Kentucky
18 Reilly Road
Fort Boone Plaza
Frankfort, KY 40601
502-564-3410 X406
Richard Huddleston
Region V
U.S. EPA
230 S. Dearborn
Chicago, IL 60402
Walter Jakubowski
Chief, Parasitology and
Immunology
HERL
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7385
Dean J. Jarog
Supervisor of Research
Everpure Inc.
660 N. Blackhawk Drive
Westmont, IL
312-654-4000 X319
James Jenkins
Project and Product Engineer
Manager
Lee Brass Co.
P.O. Box 1229
Anniston, AL 36202
205-831-2501
Taj M. Khan
Environmental Engineer
U.S. EPA, Region II
26 Federal Plaza
Room 824
New York, NY 10278
212-264-1800
Edward C. Kispert
Superintendent, Water
Quality & Research
Cincinnati Water Works
5651 Kellogg Avenue
Cincinnati, OH 45228
513-231-7825
Michael P. Kovach
Regional Sanitary Engineer
Water Supply Division
Michigan Dept. of Public
Health
3500 N. Logan
Lansing, MI 48909
517-373-1376
S. Bala Krishnan
OEET/WMD, R&D
U.S. EPA
401 M Street, S.W.
Washington, DC 20460
202-382-2617
Larry B. Landsness
Environmental Engineer
Dept. of Natural Resources
P.O. Box 7921
Madison, WI 53707
608-267-7647
Russell W. Lane
Water Treatment Consultant
1207 Devonshire Drive
Champaign, IL 61821
217-352-3528
Gary Larimore
Executive Director
Kentucky Rural Water
Association
P.O. Box 1424
Bowling Green, KY 42102-1424
502-843-2291
Richard Lauch
MERL
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7467
188
-------
Raymond D. Letterman
Professor, Department of
Civil Engineering
Syracuse University
Syracuse, NY 13210
315-423-2311
Edwin Lippy
HERL
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7226
Gary Logsdon
DWRD
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7345
Tim Lohner
Environmental Scientist
Div. of Public Water
Water Quality Section
Ohio Environmental
Protection Agency
361 E. Broad Street
Columbus, OH 43215
614-466-8307
Ron Long
Regional Manager
Cerro Copper Products Company
7557 South 78th Street
Bridgeview, IL 60458
312-594-7600
Jack Longacre
Technical Services Engineer
NIBCO, Inc.
500 Simpson Avenue
Elkhart, IN 46515
219-295-3421
0. Thomas Love, Jr.
Chief, Organic Control -
Exploratory Activity
Drinking Water Research
Division
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7281
Benjamin Lykins, Jr.
DWRD
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7460
W.S. Lyman
Senior Vice President
Technical & Market Services
Copper Development Assoc.
Inc.
Greenwich Office Park 2
Box 1840
Greenwich, CT 06836
203-625-8230
Patricia A. Mack
Environmental Surveillance
Lab. Supervisor
Div. of Environmental Quality
Commonwealth of No. Islands
Government
P.O. Box 217 CHRB
Saipan, CM 96950
6984/6114
Casper Martin
Director, Research &
Development
NIBCO, Inc.
500 Simpson Avenue
Elkhart, IN 46515
219-295-3417
Lee McCabe
TMD Science Advisor
TMD
2314 Kenlee Drive
Cincinnati, OH 45230
513-231-5965
189
-------
John F. McKenna, Jr.
Attorney
John F. McKenna Jr.
Professional Corp.
909 East Green Street
Pasadena, CA 91106
213-796-0001
Frank Messinco
General Water Foreman
Passaic Valley Water
Commission
1525 Main Avenue
Clifton, NJ 07015
201-792-3900
J. Robert Metz
Senior Engineer
Ductal Iron Pipe
Research Association
28 W. 118 Hillview Drive
Naperville, IL 60565
312-355-7003
David J. Miller
Assistant Western Regional
Sales Manager
United States Pipe & Foundry
Co.
P.O. Box 247
Westmont, IL 60559
312-920-0050
Robert Miller
HERL
U.S. EPA
26 West St. Glair Street
Cincinnati, OH 45268
513-684-7454
James Millette
HERL
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7462
Richard Miltner
DWRD
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7403
R.M. Mitchell
Technologist, Product Safety
& Compliance
Oil & Chemical, HS&E
Shell Oil Company
P.O. Box 4320
Houston, TX 77210
713-241-2268
Marvin H. Nobis
District Sales Manager
U.S. Pipe & Foundry Co.
5550 Este Avenue
Cincinnati, OH 45232
513-242-1200
Frederick W. Pontius
Water Quality Engineer
American Water Works
Association
6666 West Quincy Avenue
Denver, CO 80235
303-794-7711
George T. Rochford
Staff Representative
American Iron & Steel
Institute
1000 16th Street, N.W.
Washington, DC 20036
202-452-7288
Meredith N. Scheck
Manager, Government and
Regulatory Affairs
The Vinyl Institute
(Division of Society of
Plastics Industry, Inc.)
355 Lexington Avenue
6th Floor
New York, NY 10017
212-573-9553
190
-------
John Schuster
Cerro Copper Products Company
P.O. Box 681
East St. Louis, IL 62202
618-337-6000 X262
Craig Secover
Director of Product
Engineering
Delta Faucet Company
55 E. lllth Street
Indianapolis, IN 46280
317-848-0604
Gayle Smalley
Supervisor of Water Quality
North Marin County Water
District
P.O. Box 146
Novato, CA 94947
415-897-4133
Thomas J. Sorg
Research Engineer
DWRD
U.S. EPA
Cincinnati, OH 45268
513-684-7370
H.A. Sosnin
American Welding Society -
American Society of
Mechanical Engineers
234 Wyncote Road
Jenkintown, PA 19046
215-884-3408
Alan Stevens
DWRD
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7342
Ralph E. Stewart
Chief, Residential
Sanitation Section
New York State Department of
Health
Room 342, Tower Bldg.
Empire State Plaza
Albany, NY 12237
518-474-3766
Nancy S. Ulmer
DWRD
U.S. EPA
26 West St. Clair Street
Cincinnati, OH 45268
513-684-7583
Rosalind Volpe
Lead Industries Assoc.
242 Madison Avenue
New York, NY 10017
Robert P. Walker
Executive Director
Uni-Bell PVC Pipe Association
2655 Villa Creek Drive
No. 155
Dallas, TX 75234
214-243-3902
Robert Warner
Sierra Club
129 E. Altadena Drive
Altadena, CA 91001
213-798-5379
Vincent Williams
Associate Chemist
Louisville Water Co.
3018 Frankfort Avenue
Louisville, KY 40206
502-582-2431
Melvin Wilson
Supervisor, Maintenance
Division
Cincinnati Water Works
4747 Spring Grove Avenue
Cincinnati, OH 45232
513-352-4626
191
-------
Robert C. Witman
Vice President & Technical
Director
Carstab Corporation
2000 West Street
Cincinnati, OH 45232
513-733-2142
R. Scott Yoo
Water Quality Supervisor
East Bay Municipal Utility
District
P.O. Box 24055
Oakland, CA
415-891-0608
U.S. Envl
Rogion -!;,
250 S. O.-'.
Agency;
'L-oi'ii £>t..-oet, F.oora 1670
o0604
192
US GOVERNMENTm,NTlNBOFFICE ,985- 559-1H/107B8
------- |