By Richard Maas, Steven Patch and Michael Green
Summary: Previous lab tests by found that some brands of new filtration units would leach lead into the filtered water from the leaded brass dispenser faucet. In this follow-up study to that article (Michael Green, “Filtration Study Finds Brass Components Leach Lead into Drinking Water,” WC&P, November 1998, pp. 84-86), 504 filtration systems in actual household use were tested and 12 percent were found to be leaching 1.25 micrograms per liter (µg/L) of lead or greater into an 8-ounce (oz.) water serving after an overnight dwell time.
The issue of lead leaching from leaded-brass (about 2 percent lead) dispenser faucets commonly associated with water filtration systems has received considerable attention over the past 18 months. In fact, the issue was ranked as the third biggest story affecting the POU/POE water treatment industry in 1998.4 A study conducted by the Environmental Quality Institute at University of North Carolina-Asheville released in June 19982 examined lead discharges from these faucets under plumbed laboratory conditions using an input water typical of California finished tap water composition. The study found that nearly all under-counter systems leach lead into drinking water and expose the typical user to more than the 0.5 µg/day, or parts per billion (ppb), of lead permitted under California Proposition 65 (Prop. 65) and two of the 16 brands tested were found to potentially expose the typical user to more than 15 µg/day of lead, which would classify the product as hazardous by the U.S. Consumer Products Safety Commission. These two of the models, the Omni OT2 and Franke Uniflow DW100, leached high amounts of lead (60-to-71 times the Prop. 65 level) apparently because—at least when new—the filtration media actually greatly increased the acidity and corrosivity of the input water even after the suggested pre-conditioning period.3
A case for change
Based on this study, the Center for Environmental Health (CEH) filed lawsuits under Prop. 65 against the companies with excessive lead discharges. Omni quickly changed its faucet to a non-brass, high impact plastic model and Franke pulled its Uniflow DW100 from the market and began using zero-lead faucets. Nearly all of the other major manufacturers reached quick settlements of the lawsuits by agreeing to provide either stainless steel, lead-free brass, plastic or plastic-lined dispenser faucets with their filtration units by early 1999, an agreement that apparently has been implemented.
To date, the chronology of events represents a classic case where a potential problem was uncovered and, through a combination of good industry citizenship, public awareness and litigation threat, the issue was quickly and effectively addressed to the benefit of all filtration system purchasers. There are probably only a few home water filtration units still being sold in the United States (either by very small foreign or U.S. manufacturers), which still contain leaded-brass dispenser faucets.
The important remaining issue concerns the degree to which leaded-brass filter faucets sold in the United States prior to 1999 continue to leach lead under actual residential usage conditions and for how long. To answer these questions as well as to produce practical information for concerned filtration system owners, we initiated a research study whereby any U.S. residential filtration system owner could have the water from their system analyzed for lead at the Environmental Quality Institute in return for their participation in the study. This included filling out a questionnaire regarding their filter type, filter brand, filter age, date of last cartridge replacement, water source type and daily filtered water consumption patterns. The first several hundred volunteers received free testing; a nominal fee to cover laboratory costs was requested for later participants. Volunteer participants were sent two laboratory-washed polyethylene sample bottles with detailed instructions for sample collection and return. One sample was a 30-milliliter (ml) first draw taken after an overnight standing time. Because most of the dispenser faucets have an internal volume of 10-to-20 ml, this sample represented the internal dwell slug plus about one-half to two flush volumes. Immediately following this first draw sample, water was run through the system at full flow for 10 seconds and a second 30-ml sample taken. This enabled any residual dissolved lead in the dispenser faucet to be completely flushed, and thus this second sample was representative of the typical lead concentration discharged from the filter under typical flowing water usage conditions.
First flush results
A total of 724 sample kits were sent out to volunteer participants, of which 504 were returned for inclusion in this study. Of these, about half were from California where the opportunity to participate in the study was most publicized. Of the 504 systems, 60 (11.9 percent) were found to contain more than 10 µg/L of lead in the first draw sample. A total of 123 (24.4 percent) contained at least 5 µg/L. The analytical results were for 30-ml samples and thus for a typical 8-oz.—approximately 240-ml—serving, the resultant concentration for a 30-ml sample with 10 µg/L would be about 1.25 µg/L added by the dispenser faucet assuming no additional lead after the first 30-ml. Although certainly no one purchasing a filtration system for the purpose of removing lead contamination from their tap water would want any lead added back into the water by the faucet, we used 5 µg/L as a threshold level of concern in our “results” letter to the volunteer participants. The highest 30-ml first draw observed was 82 µg/L (equal to about 10 µg/L in an 8-oz. serving) and the 99th percentile—the top 1 percent— concentration was 50.3 µg/L (~6.3 µg/L per 8-oz. serving). These results indicate that small to moderate lead contamination from the filter faucet was relatively common.
Because the study population of 504 was spread out over 40 brands of filtration systems, the study contains insufficient data to quantify the lead contamination risk from most of the brands. Table 1 shows the percentage of units producing first draw water with lead concentrations above 5 µg/L and 10 µg/L for those brands for which we received more than 25 samples.
Because of the relatively small number of samples of each brand, these results should probably be viewed as only very approximate for each brand. In most cases, random inclusion or exclusion of one or two high lead concentration units changes the percentages greatly. It’s interesting to note that all of the systems, including the Franke and Omni systems, which previously tested much higher than the others when new under laboratory conditions, tended to be remarkably close to the overall means. It would appear water that the water corrosivity increasing properties of the Franke and Omni models, which were observed when new, dissipate over time in actual residential usage. Conversely, several models that showed relatively low lead leaching in the laboratory studies, (e.g., Ametek) demonstrated at least average risk in actual usage. Considering that dispenser faucets on most of the models are the same 2 percent leaded-brass alloy, it makes sense that long term lead leaching would be controlled by the corrosivity of the water to which filter was plumbed. Thus, we believe that if the system is supplied by a corrosive water, it will tend to leach more lead regardless of the filter brand. In the laboratory tests, all of the units were connected to the same relatively non-corrosive water.
To test this theory further, we measured the pH (the single most important variable affecting water corrosivity) of the water samples and found that, for water below pH 7.0 (i.e., more corrosive), 29.7 percent of the samples exceeded 10 µg/L, while for waters with a pH above 7.5 (i.e., less corrosive), only 5.7 percent of the samples had lead concentrations above 10 µg/L. Further evidence of water corrosivity being the major factor controlling lead contamination risk is provided from the observation that 33 percent of the 140 reverse osmosis (RO) filters (known to often increase water corrosivity) had first draw lead levels above 5 µg/L, while only 24 percent of the non-RO-type units exceeded this level. Of course, the analysis is only very approximate since other factors (i.e., hardness, alkalinity, temperature, sulfate, etc.) can also affect lead corrosivity.
One serendipitous observation from our study is that overall the water filtration systems studied do a remarkably good job of removing lead. Although we had no way of knowing the incoming lead concentrations to the filters, only three of the 504 systems exhibited a 10-second purge lead concentration of above 5 µg/L (highest = 15.2 µg/L), and all three of these indicated on their questionnaire that it had been more than one year since they had changed their filter cartridge.
Finally, it should be noted that the amount of lead leached from the leaded-brass dispenser faucets was not found to be significantly associated with age of the system. Many of the systems observed to be leaching higher levels of lead had been in continual residential service for more than five years. Apparently, if a system is plumbed to a corrosive water that promotes the dissolution of lead from the faucet, it will continue to do so indefinitely. This observation is consistent with our laboratory results, which have demonstrated that for three models subjected to extending testing (five water flushings per day), lead concentrations remained essentially constant from Day 20 to Day 282 when the experiments were terminated.
Until late 1998 or early 1999, most residential water filtration systems were equipped with a leaded-brass dispenser faucet composed of about 2 percent lead. In a series of Prop. 65 settlement agreements with the Center for Environmental Health, nearly all major manufacturers agreed to modify these systems to employ stainless steel, plastic, lead-free brass, or plastic-lined dispenser faucets. The current study of 504 actual residential systems found that 24 percent produced first draw lead concentrations above 5 µg/L and 11.9 percent produced first draw concentrations of about 10 µg/L (equivalent to 1.25 µg/L in an 8-oz. serving). The risk of lead contamination from lead-containing dispenser faucets appears to be primarily related to the corrosivity of the incoming household water supply rather than to the manufacturer of the filtration system or the age of the system. We recommend that the millions of residential owners of filtration systems purchased before 1999 either have the filter faucet replaced with a stainless steel, lead-free brass, plastic or plastic-lined version or have their filtered water tested for lead. We’re continuing to accept volunteer participants for our on-going study and will report additional findings in the future.
- Maas, R.P., S.C. Patch, M.S. LaGoy and D.M. Morgan, Lead Exposure From Brass
Fixtures Connected to Commercially Available Water Filters, UNC Asheville Environmental Quality Institute, Asheville, N.C., Technical Report #98-053, 26 p., 1998.
- “Newsreel: Lead Leaching Claim Slams Faucet-Filter Unit Manufacturers,” WC&P, August 1998, p. 12.
- Green, M., “Filtration Study Finds Brass Components Leach Lead into Drinking Water,” WC&P, November 1998, pp. 84-86.
- “The Biggest Stories Affecting the POU/POE Water Treatment Industry in 1998,” WC&P, December 1998, pp. 34-37.
About the authors
Richard P. Maas, Ph.D., and Steven C. Patch, Ph.D., co-direct the Environmental Quality Institute at the University of North Carolina-Asheville. The EQI has long been recognized as the leading U.S. university research institute for the study of lead in drinking water. The institute has conducted the research for several previous California Proposition 65 litigations including faucet fixtures, submersible well pumps and water meters. Their lead research has been funded frequently by the USEPA, and the State of California, and they have authored over 40 publications on lead contamination of tap water. Maas and Patch can be reached at EQI, UNC Asheville, Asheville, NC 28804, or e-mail: [email protected] or [email protected]
Michael Green is executive director and founder of the Center for Environmental Health. He has a master’s degree in natural resources and a master’s degree in public policy, both from the University of Michigan. The mission of CEH is to protect the public from environmental health hazards and toxic exposures by directly influencing corporate behavior in the public interest. If you would like to participate in ongoing research related to the above article, call (510) 594-9864. Green can be reached at CEH, 528 61st St., Suite A, Oakland, CA 94609, or e-mail: [email protected]