An assessment of lead exposure potential from residential cutoff valves. (Features).

Introduction

Lead is now recognized as a significant public-health hazard in the United States. Medical and epidemiological studies over the past decade have shown that measurable and permanent neurological damage will likely result, especially in infants and young children, from even very low lead exposures (U.S. Environmental Protection Agency [U.S. EPA], 1995; Needleman & Bellinger, 1991; Wolfe, 2000). In a meta-analysis of every large published medical study of the association between IQ and blood lead concentration, Schwartz found that estimates of IQ score reductions ranged between I and 6 points for an increase of 10 [micro]g/deciliter ([micro]g/dL) blood level concentration (Schwartz, 1994). In addition to neurological and other health effects, exposure to lead has been found to cause pregnant women to deliver prematurely and to have low-birth-weight babies (Lippmann, 2000). Studies used by U.S. EPA estimate that children's blood lead levels are typically increased by approximately 0.16 [micro]g/dL for each microgram per day of lead ingested (Department of Health and Human Services, 1993). The purpose of this study was to determine the extent of lead leaching from new residential brass valves and to use these results to calculate and predict the approximate public lead exposure associated with these components.

Drinking-water contamination is clearly recognized as one of the important sources of lead exposure to the U.S. population, with estimates of exposure from this source typically ranging between 14 percent and 20 percent of the total average U.S. exposure (American Water Works Association, 1990; Bois, Benet, Tozer & Zeise, 1989). The sources of lead contamination in U.S. drinking water have been addressed steadily over the past 15 years. Lead was banned in solder joints, and plumbing components were restricted to less than 8 percent lead by mid-1988 (Safe Drinking Water Act Amendments, 1986). The Federal Lead and Copper Rule issued in 1991 required large public water suppliers and other water suppliers with corrosive waters to try to reduce the corrosivity of their water so that less lead would be dissolved from the leaded solder joints, leaded-brass faucet fixtures and other lead-containing components in household plumbing systems (U.S. EPA, 1991). It was estimated that the rule would reduce public lead expos ure from drinking water by about 50 percent when fully implemented by 2010 (U.S. EPA, 1991).

Until recently, nearly all household kitchen and bathroom faucets were constructed of 2 to 7 percent leaded brass, and studies have documented that these faucets almost always leach high amounts of lead into residential drinking water when new (Maas, Brown, Morgan, Patch, & Rainey, 1993) and in many cases even after many years of use (Patch, Maas, & Pope, 1998). In a 1995 litigation settlement, all major U.S. faucet manufacturers agreed to phase out the use of leaded-brass alloys in faucets by the end of 1999, and nearly all were in compliance by the end of 1997 (Consent Judgment and Settlement Agreements, 1995). Presently, virtually all residential brass water meters installed in the United States are made of 5 or 7 percent leaded brass. These meters have been shown to leach significant amounts of lead into drinking water both when new (Maas, Kawaguchi, Morgan, & Patch, 1997) and after many years when supplied by a more corrosive water composition (Maas, Patch, Morgan, & Pedler, 1998). A lawsuit filed under California Proposition 65 (Barclays California Code of Regulations, 1993) led to an August 1999 settlement under which the nine major U.S. manufacturers of these meters have agreed to switch to lead-free brass alloys by 2001 (J. Wheaton, director of Public Interest Law Group, Oakland, California, personal communication, July 14,1999). In late 1994, the manufacturers of submersible well pumps, which provide drinking water to over 40 million U.S. residents, switched to lead-free brass alloys, effectively eliminating this significant lead contamination source in new rural construction and replacement (Maas, Patch, Pope, & Thornton, 1998).

Through these various regulatory and litigatory initiatives, the exposure of the U.S. public to lead in drinking water has steadily decreased over the past decade, especially for the residents of new buildings. Decrease in exposure risk has been much less for residents of older buildings, especially if served by individual wells without corrosion reduction treatments. A recent longitudinal survey of tap water lead contamination risk in North Carolina found that the percentage of homes with significant contamination (first-draw lead concentrations of 15 [mu]g/L or greater) decreased from 15.9 percent in the 1989-1991 period to 9.2 percent in the period of 1994-1997 (Maas, Matthews, Morgan, & Patch, 1999). This is probably representative of national trends.

One of the remaining potential sources of lead in residential drinking water is the leaded-brass cutoff valves that are used in residential plumbing systems. Although manufacturers of faucets, submersible well pumps, and now water meters have switched to truly lead-free brass alloys, cutoff valves continue to be manufactured out of leaded-brass alloys typically composed of 5 to 7 percent lead alloys. A high percentage of homes use brass cutoff valves. Therefore, it is important to be able to quantify the exposure to the public expected from a typical brass cutoff valve.

Methodology

In 1997, the 14 brands and models of residential brass valves used in this study were purchased from various retail plumbing supply outlets in California, A combination of six different outlets, ranging from plumbing supply outlets to large home supply centers to smaller neighborhood hardware stores, was included, It is likely that the 14 brands probably represent the majority of valve brands and models currently used in new California residential construction and plumbing renovation. All the available information associated with the type of valve and type and place of manufacture was recorded in the laboratory, and the features anticipated to be relevant to lead leaching were summarized as shown in Table 1.

The testing for lead leaching was conducted according to the protocol specified by NSF International for drinking-water end-use devices (NSF-61 Section 9), with some slight modifications as noted below. The details of this testing protocol are published elsewhere (NSF, 1997), but can be summarized as follows. After an initial 15-minute rinse by flowing tap water, each valve was tested in triplicate, rinsed three times with deionized water, then rinsed with three volumes of extractant water. The internal volume of each of the 42 valves was also determined at this time. Prior to actual testing, a portion of each internal surface, which in typical installation would be covered or otherwise separated from direct water contact by overlapping threaded fittings, was carefully coated with caulk. Thus, only part of the internal surface was left exposed to the extractant water. To test that the caulk did not leach lead, samples of water exposed only to the caulk were analyzed. All had nondetectable levels of lead. The extractant water was deionized water to which NSF-61-specific reagents were added to produce a pH of 8.0 ([+ or -]0.5), a chlorine residual of 2 mg/L, and a carbonate alkalinity of 50 mg/L as calcium carbonate ([CaCO.sub.3]). NSF-61 Section 9 specifies a carbonate alkalinity of 500 mg/L, but a 50 mg/L alkalinity is more typical of finished California and U.S. public water supplies. A "dump-and-fill" water-changing schedule was followed for 19 days. Each valve was filled with extractant water that was changed at two-hour intervals, five times throughout a 9-a.m.-to-5.-p.m. workday. Valves were left full of extractant water over the weekend. On days 3, 4, 5, 10, 11, 12, 17, 18, and 19, which correspond to Wednesday, Thursday, and Friday of each of the three weeks, samples were taken after internal dwell times of 16 hours, two hours, 30 minutes, and 10 minutes. The exposed internal volume of each valve was measured by filling the valve with water and emptying it into a graduated cylinder. The NSF-61 Section 9 p rotocol requires only 16-hour dwell samples; however, inclusion of the shorter dwell-time samples enables a more accurate estimation of actual human exposure to be made with only a very small decrease in resultant concentrations as a result of the extra two water changes (30 minutes and 10 minutes) on the nine sampling days.

The bottom of each valve was plugged with a rubber stopper coated with lead-free caulk, and the open top of each valve was protected from possible external lead contamination with plastic sheet wrap. A control vial of extractant water was mounted along with the experimental valves to document the absence of atmospheric lead inputs to the extraction waters. Samples were taken in acid-washed polyethylene bottles and were immediately acidified with Fisher[TM] trace-metal-grade concentrated nitric acid to a final acid concentration of 0.5 percent. Following a minimum 28-hour holding time, the authors analyzed samples for total lead by graphite furnace atomic absorption spectrophotometry using a TJA Video 12 ramped-program furnace. The lead detection limit was approximately 0.7 [micro]g/L, quantified by both independent standard calibrations and the method of standard additions. Raw sample lead concentrations were multiplied by the internal volume (in liters) of the corresponding valve to obtain the actual mass o f lead content in each sample. Predicted human exposure was calculated under the assumption that a person would consume four 500-mL water servings, one after each of the internal dwell times of 16 hours, two hours, 30 minutes, and 10 minutes, respectively.

Results and Discussion

This study examined the lead leaching into drinking water from three types of valves commonly used in residential plumbing systems: angle stops, ball valves, and gate valves. Angle stops are cutoff valves with an ovalshaped handle generally located just below kitchen and bathroom sinks. They generally also serve to reduce the pipe flow diameter from one-half to three-eighths of an inch. Because they are usually located within 4 feet of the faucet, their internal dwell water will nearly always become part of any water serving. (For typical one-half-inch residential plumbing, a 500-mL draw is equivalent to the internal volume of 11.5 feet of plumbing.)

Ball valves are cutoff valves installed near various junctures of the plumbing systems and are used to cut off water to specific sections of the household, usually by shifting the internal brass open-ended ball by 90 degrees. Often they are located within 11.5 feet of the kitchen or bathroom faucet, in which case their internal dwell water would become part of a 500-mL faucet draw. If they are located more than 11.5 feet away, however, their internal dwell slug would be transported approximately this distance toward the faucet with some longitudinal dispersion. This translocated internal dwell slug could then be dispensed with the next faucet usage or could be expelled by other subsequent water uses such as showering and toilet flushes.

Gate valves have a circular handle that is turned clockwise several revolutions to close a "gate" within the valve. They generally have larger internal volumes and can be used to shut off water to the whole house or, like ball valves, to portions of the building. Average ingestion potential of their internal dwell water is probably slightly less than for ball valves because of their somewhat greater average distance from a kitchen or bathroom sink.

Mean concentrations for the three types of valves averaged over the three replicates and the three sampling days of each week are plotted against the week of sampling in figures 1 through 3. Within each type of valve, brands that had the highest initial concentrations tended to drop the most during the course of the study.

The lead-leaching results for the valves during the third week of sampling are summarized in Table 2. Under the ingestion assumptions noted previously, the five brands of angle stops all produce a lower exposure than the stringent 0.5-[micro]g/day limit specified by California Proposition 65 as a lower threshold for any observable effect to the most sensitive segments of the population. These low exposures resulting from the angle stops are attributable to the low-to-moderate leachate concentrations (8.6 [micro]g/L to 57.6 [micro]g/L for 16-hour dwells) combined with the very small internal volume of these devices, typically on the order of 2 mL.

From Table 2, it can be seen that seven of the 14 models resulted in calculated daily exposures above 0.5 micro]g/day. This result occurred with all five gate valve models (which typically have greater internal volumes) and two of the four ball valve models. The Nibco gate valve, by combining high lead leachate concentrations with a larger internal volume (about 7 mL), produced the highest calculated exposure (9.73 [micro]g/day), which is more than half of the U.S. Consumer Products Safety Commission daily exposure threshold of 15 [micro]g/day for classifying a consumer product as hazardous. Ironically, Nibco is also one of the very few companies currently producing some models of completely lead-free residential valves. The four valves (three ball and one gate) that had chrome plating on the inside of the body tended to produce lower lead concentrations, but these levels were still moderately elevated, suggesting that a substantial portion of the lead dissolved may originate from the internal ball or gate it self, neither of which is chrome-plated.

It is interesting to note that disproportionately high percentages of the 16-hour dwell time lead concentrations are achieved after short dwell times. In fact, on average, 62.7 percent of the lead leached out over 16 hours was leached in the first two hours, and 32.4 percent occurred in just the first 10 minutes. This observation is consistent with other studies on lead-leaching dynamics and kinetics in brass kitchen faucets (Maas, Brown, et al., 1993) and brass water meters (Maas, Kawaguchi, et al., 1997), and suggests that, if a household has lead contamination, the contamination must be considered to be present even when the household plumbing system is being used frequently.

Conclusions

These study results indicate that, on average, a U.S. resident may be exposed to about 2.66 [micro]g/day of lead from each new valve in the household plumbing system. Little is currently known about how the lead leaching from these plumbing devices changes over time. Similar studies conducted with water meters (Consent Judgment and Settlement Agreements, 1995; Patch et al., 1998) and kitchen faucets extended over longer time periods indicate that lead levels will typically decrease to about 25 percent to 30 percent of the concentration means observed over the first 19 days and then will remain relatively constant for a long period thereafter. Both the amount of lead leaching from plumbing fixtures and the longevity of the leaching depend greatly on the corrosivity of the water supplying the residence (Maas, Patch, Morgan, & Pedler, 1998). The water used in these experiments was of approximately average U.S. corrosivity. Water supplies in the top 10 percent of the U.S. corrosivity range would be predicted to c ause several times greater lead leaching from brass plumbing components for an indefinite period. Exposure of the U.S. public to lead from angle stops, ball valves, and gate valves constitutes a relatively small source, but one that could easily be eliminated with non-leaded-brass alloys. Medical studies show that any additional lead exposure will result in corresponding increases in neurological damage, especially in infants, and U.S. EPA has set a maximum contaminant level goal of zero in view of the evidence on the health effects of lead. Manufacturers of kitchen faucets and submersible well pumps have already switched to completely lead-free brass alloys.

Lawsuits under California's Proposition 65 are currently in progress in California and hopefully will result in a conversion of residential cutoff valves to lead-free brass alloys. For all current residential buildings, however, residents, particularly those in households with pregnant women and children less than six years of age, should consider having their drinking water tested for lead.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

[FIGURE 3 0MITTED]

TABLE 1

Summary of Valve Characteristics

Brand        Valve Type    Average      Primary Interior     Country
                         Volume (mL)  Surface Construction  of Origin

B&K              BV         2.73              MCP           Italy

B&K              GV         7.97              SC            Indonesia

Brass Craft      AS         1.80              M             U.S.

Eastman          AS         2.03              N             UNK

Jameco           AS         1.83              M             UNK

Matco            BV         2.37              MCP           Taiwan

Nibco            BV         3.00              M/SC          U.S.

Nibco            GV         6.93              SC            U.S.

Orbit            GV         7.57              Sc            China

Pegler           BV         6.03              MCP           Italy

Pegler           GV         8.97              M             England

Roberts          AS         2.20              M             UNK

TAHO             GV         3.23              MCP           Taiwan

Wolverine        AS         2.10              M             U.S.

AS = angle stop.

BV = ball valve.

GV = gate valve.

M = machined.

MCP = machined with chrome plating.

SC = sand-casted.

UNK = unknown.
TABLE 2

Mean Lead Leachate Concentration([micro]g/L) and Calculated Lead
Ingestion ([micro]g/day) by Valve Brand, Third Week of Sampling *

Valve        Valve  16-Hour Dwell  2-Hour Dwell   30-Min. Dwell
Brand        Type   Concentration  Concentration  Concentration
                    ([micro]g/L)   ([micro]g/L)   ([micro]g/L)

B&K           BV         71.5           45.6           41.1

B&K           GV        161.2           85.7           75.4

Brass Craft   AS         16.7            8.3            9.5

Eastman       AS          8.6            3.6            2.0

Jameco        AS         18.6           11.2           10.9

Matco         BV        186.0          205.6           77.3

Nibco         BV        181.9          139.7          114.8

Nibco         GV        455.4          302.8          374.8

Orbit         GV        127.0           73.3           55.2

Pegler        BV         25.0           15.7           11.1

Pegler        GV        151.6          112.0           67.7

Roberts       AS         39.8           22.8           18.6

TAHO          BV         67.0           58.5           37.1

Wolverine     AS         57.6           33.9           27.4

Mean           -        112.0           79.9           65.9

Valve        10-Min. Dwell  Calculated Daily
Brand        Concentration      Exposure
             ([micro]g/L)     ([micro]g)

B&K               15.5            0.47

B&K               37.2            2.87

Brass Craft        4.1            <0.1

Eastman            1.0            <0.1

Jameco             5.2            <0.1

Matco             57.8            1.25

Nibco             57.6            1.48

Nibco            271.2            9.73

Orbit             27.0            2.14

Pegler             5.6            0.34

Pegler            39.5            3.33

Roberts           12.0            0.21

TAHO              18.1            0.58

Wolverine         15.1            0.28

Mean              40.5            1.63

* Samples are averaged over the three replicates and the three days of
weekly sampling.

AS = angle stop.

BV = ball valve.

GV = gate valve.

Acknowledgement: This research was supported by a grant from the California Public Health Trust.

References

American Water Works Association. (1990). Water quality and treatment (4th ed.). New York, NY: McGraw Hill Publishing.

Consent Judgment and Settlement Agreements of Aug. 1, 1995 (People v. American Standard et al.), San Francisco, California Superior Court #948017.

Barclays California Code of Regulations. (1993). Register #93, Title 22, P. 179.

Bois, F.Y., Benet, L.Z., Tozer, T.N., & Zeise, L. (1989). Application of clearance concepts to the assessment of exposure to lead in drinking water. American Journal of Public Health, 79(7), 827-831.

Department of Health and Human Services, Food and Drug Administration, Part II. 58 Federal Register 117 (June 1, 1993) (codified at 21 C.F.R. ss 109 and ss 189).

Lippmann, M. (2000). Environmental toxicants: Human exposures and their health effects. New York: John Wiley and Sons.

Maas, R.P., Brown, G.M., Morgan, D.M., Patch, S.C., & Rainey, E.R. (1993). The rate and extent of lead leaching from faucet fixtures and hot water dispensers under laboratory conditions (Technical Report #93-009). Asheville, NC: University of North Carolina at Asheville Environmental Quality Institute.

Maas, R.P., Kawaguchi, H., Morgan, D.M., & Patch, S.C. (1997). Lead leaching from brass water meters under pressurized flow conditions. American Water Works Annual Conference Proceedings, Atlanta, Georgia, June 15-19,1997, C, 589-602.

Maas, R.P., Matthews, G., Morgan, D.M., & Patch, S.C. (1999). Lead contamination of North Carolina residential tapwater. Proceedings of the Mountain Water Research Conference, Asheville, North Carolina, April 17, p. 21.

Maas, R.P., Patch, S.C., Morgan, D.M., & Pedler, R.C. (1998). Lead leaching from in-service leaded-brass water meters. Technical Journal of the North Carolina Section of the American Water Works Association, 1, 270-279.

Maas, R.P., Patch, S.C., Pope, J.P., & Thornton, L.W. (1998). Lead-leaching characteristics of submersible residential water pumps. Journal of Environmental Health, 60(6), 8-13.

NSF International. (1997). Drinking water system components health effects (ANSI/NSF-61-1997b). Ann Arbor, MI: Author.

Needleman, H., & Bellinger, D. (1991). The health effects of low level exposure to lead. Annual Review of Public Health, 12, 111-140.

Patch, S.C., Maas, R.P., & Pope, J.P. (1998). Lead leaching from faucet fixtures under residential conditions. Journal of Environmental Health, 61(3), 18-21.

Safe Drinking Water Act Amendments. (1986). Public Law 99-339.

Schwartz, J. (1994). Low-level lead exposure and children's IQ: A meta-analysis and search for a threshold. Environmental Research, 65(1), 42-45.

U.S. Environmental Protection Agency, Drinking water regulations: Maximum contaminant level goals and national privacy drinking water regulations for lead and copper. 53 Federal Register 110 (1991).

U.S. Environmental Protection Agency (February 1995). Lead poisoning and your children (EPA 800-B-92-0002). Washington, DC: Environmental Health Center, National Lead Information Center, A Division of the National Safety Council. http://www.nsc.org/ehc/nlic/ledepa.htm (8 Dec. 1999).

Wolfe, C.A. (2000). Pediatric alert: Assess for lead. RN, 63(8), 26-30.

Corresponding author: Richard P. Maas, University of North Carolina--Asheville Environmental Quality Institute, One University Heights, Asheville, NC 28804.

E-mail: <[e-mail]>.