Homes with lead service lines (LSLs) in the City of Brandon, Manitoba, Canada, were found to exceed the provincial standard of 10 μg/L for lead in drinking water. Solids identified by X-ray diffraction of LSL scale were Pb5O8 and PbO2, indicating that lead(II) solids in the LSL scale have been oxidized to lead(IV) solids by free chlorine residuals. Natural organic matter (NOM) can reduce PbO2 within a few hours, and Brandon treated water has high levels of NOM at approximately 5–7.6 mg/L as total organic carbon (TOC). As water stagnates in the LSL during periods of no water use the free chlorine residual is depleted, permitting PbO2 to oxidize NOM and be reduced to more soluble lead(II) species, resulting in an increase in dissolved lead concentrations. Although it is generally believed that aquatic humic substances (AHS) are primarily responsible for the reductant capacity of NOM, removal of AHS from the treated water resulted in a 6% decrease in lead release from PbO2, while removal of 50% of total NOM resulted in a 75% decrease in lead release. AHS and TOC were not found to play a significant role in the reduction of PbO2 in this water.

Lead service lines (LSLs) were commonly installed in Canadian drinking water distribution systems prior to their ban in 1975 (Health Canada 2009). However, lead is toxic to humans and can leach into drinking water when in contact with lead pipes or plumbing materials (ATSDR 2007; Health Canada 2013). The Province of Manitoba, Canada, has established a standard for lead in drinking water of 10 μg/L (Province of Manitoba 2007), which is in line with standards set by Canadian Guidelines for Drinking Water Quality (Health Canada 2014) and the World Health Organization (WHO Guidelines for Drinking Water Quality 2011).

Lead released from lead pipes forms complexes with ligands, the predominant ligand most often being carbonate. Lead(II) carbonates can precipitate and form a scale coating the interior of LSLs. In systems maintaining free chlorine residuals throughout the distribution system for secondary disinfection, lead(II) carbonates in the scale can be oxidized to lead(IV) oxide (PbO2) (Schock et al. 1996; Lytle & Schock 2005).

The equilibrium solubility of PbO2 has not been experimentally measured at the pH and alkalinity relevant to drinking water systems. It has however been shown to be stable in the presence of free chlorine (Lytle & Schock 2005). When free chlorine levels are depleted, PbO2 in a solution of deionized (DI) water will be reduced back to divalent lead species, of a higher solubility, over a period of weeks to months (Lytle & Schock 2005). Natural organic matter (NOM) can act as an electron donor, and can decrease the time required for the reduction reaction of PbO2 to divalent lead species to hours (Dryer & Korshin 2007; Lin & Valentine 2008; Shi & Stone 2009a), even in the presence of the competing effects of free chlorine (Lin & Valentine 2009).

NOM and AHS

Due to the complexity of NOM and the poor understanding of its composition and behavior, studies often break NOM down into operationally-defined classifications. One operationally-defined class of NOM is aquatic humic substances (AHS), determined by the method adopted by the International Humic Substances Society (IHSS), or by the Standard Methods XAD Method 5510C. AHS, operationally-defined by these methods, are a highly reactive complex mixture of humic-like substances, displaying predominantly hydrophobic and acidic properties (Thurman 1981).

Understanding the properties of the matter responsible for reduction of PbO2 is important for the treatment plant to target this matter's removal. For example, treatment strategies for removal of total NOM include membrane filtration and changing the source water, while strategies for the targeted removal of the AHS fraction include advanced oxidation processes and optimizing coagulation efficiency. AHS are believed to be responsible for the majority of oxidation-reduction reactions of organic matter in natural water (Thurman 1985). Although it is not known exactly which components are responsible for the redox capacity of NOM and AHS, hydroquinones are one component of AHS that is thought to have particularly high reductant reactivity (Cory & McKnight 2005). Some studies have shown that the accumulation of lead and its toxicity can be influenced in the presence of AHS (Grosell & Gerdes 2006).

AHS is also the fraction of NOM that has been found to contribute the most to the formation of trihalomethanes (THMs), a regulated disinfection byproduct (Singer 1999). The standard for THMs in the Province of Manitoba is 100 μg/L (Province of Manitoba 2007). Water systems with total organic carbon (TOC) concentrations considered high in their treated water (generally >3 mg/L), are at a higher risk of producing THMs after disinfection with free chlorine. Therefore, these water systems are motivated to remove the AHS fraction from their water, which could prove beneficial for reducing concentrations of lead at the tap as well. However, to date, studies have yielded conflicting results as to whether characteristic properties associated with humic-like substances (i.e. indicated by specific ultraviolet absorbance (SUVA) at 254 nm) are found to increase or decrease reduction of PbO2 (Dryer & Korshin 2007; Shi & Stone 2009a). This study elucidates the role of the operationally-defined AHS class of NOM in the reduction of PbO2. This study also examines how PbO2 reduction is specifically affected by reduction in total NOM, and how it is affected by TOC removal.

Corrosion control strategies

Common corrosion control strategies for water distribution systems include LSL replacement, pH and alkalinity adjustment, and phosphate addition (Schock et al. 1996). Complete LSL replacement should be the ultimate goal of water systems; however, this method is costly and takes time. A strategy is often needed as an interim solution while complete LSL replacement is underway. pH and alkalinity control would not have a significant impact on the solubility of scale composed of PbO2 because PbO2 is already sparingly soluble at the pH of drinking water (Lytle & Schock 2005). Some cities may also be motivated to avoid phosphate addition for corrosion control as phosphates contribute to eutrophication.

Water source

This study focused on the City of Brandon, Manitoba, Canada. The City of Brandon is located in south-western Manitoba, approximately 200 km west of Winnipeg, the province's capital. Approximately 3,600 homes in the City of Brandon (population: 46,000) may still have LSLs (City of Brandon 2014). The source water of the City of Brandon Water Supply System (BWSS) is the Assiniboine River, though groundwater is blended during periods when the river's turbidity, organics or hardness are high, typically in the spring and summer months. BWSS uses conventional coagulation-flocculation treatment processes with lime softening and chlorination. Phosphates are not currently added for corrosion control, and have not been in the past. The system disinfects with UV and free chlorine, and free chlorine residual is maintained throughout the distribution system. NOM is very high in the source water, ranging from approximately 11–15 mg/L as TOC, with 5–7.5 mg/L TOC remaining in the treated water (Table 1). A flow diagram of the Brandon water treatment plant can be found in Figure 1, and a map of the distribution system can be found in Figure 2. Raw and treated water quality parameters are summarized in Table 1.
Table 1

Typical drinking water quality parameters in raw water, treated water, and distribution system in the City of Brandon

ParameterRaw waterTreated water (at the treatment plant)Distribution system water
pHa 7.9–8.3 7.5–7.8 7.6–7.9b 
Alkalinity (mg/L as CaCO3)a 236–302 50–81 – 
NOM (mg/L as TOC)c 11.2–15.2 5–7.6 – 
Free chlorine residual (mg/L as Cl2)a – 2.3–2.9 0.3–0.9b 
Total THMs (μg/L)d – – 40–151 
ParameterRaw waterTreated water (at the treatment plant)Distribution system water
pHa 7.9–8.3 7.5–7.8 7.6–7.9b 
Alkalinity (mg/L as CaCO3)a 236–302 50–81 – 
NOM (mg/L as TOC)c 11.2–15.2 5–7.6 – 
Free chlorine residual (mg/L as Cl2)a – 2.3–2.9 0.3–0.9b 
Total THMs (μg/L)d – – 40–151 

apH, alkalinity and free chlorine in raw and treated water analyzed daily from 2007 to 2012 by the plant.

bValues of pH and free chlorine in the distribution system are based on three samples collected from the distribution system in May 2014.

cNOM values are based on 12 samples collected throughout the year between 2006 and 2013 by the City of Brandon and the Manitoba Office of Drinking Water.

dTHMs were analyzed quarterly by the City of Brandon in 2013 (City of Brandon 2013).

Figure 1

Flow diagram of the Brandon water treatment plant.

Figure 1

Flow diagram of the Brandon water treatment plant.

Close modal
Figure 2

Map of the Brandon water distribution system. 1: Location of the Brandon water treatment plant. 2 and 3: Locations of the system's two chlorine booster stations. The dashed lines outline the regions where LSLs are located.

Figure 2

Map of the Brandon water distribution system. 1: Location of the Brandon water treatment plant. 2 and 3: Locations of the system's two chlorine booster stations. The dashed lines outline the regions where LSLs are located.

Close modal

Objectives

The first goal of this study was to investigate the occurrence of lead in drinking water in the City of Brandon, Manitoba, Canada. The second goal was to identify the mechanisms of lead release from the LSL scale into the water. The third goal was to investigate the effectiveness of removing total NOM, TOC and AHS from the BWSS treated water, in the presence of the competing effects of free chlorine, as a strategy for controlling lead release from the reduction of PbO2 to the more soluble lead(II) species.

Lead concentrations in the Brandon water distribution system

Twenty Water single-unit dwellings with LSLs (according to the City's records) were selected in the City of Brandon, located within the regions identified in Figure 2. Water samples were collected from each house following a minimum stagnation period of 6 hours in order to capture maximum lead concentrations. A 6-hour stagnation period would occur frequently while residents were sleeping or at work. Due to the logistical complications of collecting samples from numerous residences following a period of 6 hours of stagnation, residents of the selected houses were contacted and given explicit instructions to collect the first four consecutive liters of water from the kitchen's cold water tap, following a minimum 6-hour period of stagnation, and to collect a fifth liter after allowing the water to run continuously for 5 minutes, in order to flush the lines and obtain water from the water main. Based on the typical distance between houses and the curb in Canada, and the typical diameter of service lines of 5/8-inch (1.6 cm) to ¾-inch (1.9 cm) (Health Canada 2009), this sample should be representative of water drawn from the water main, entering in contact with the LSL only in passing.

The sample bottles were dropped off with written instructions. The residents then collected the samples following the instructions, and the bottles were picked up within 2 days. The samples were analyzed for total lead by ALS environmental laboratories by inductively coupled plasma–mass spectrometry (ICP-MS). Samples were acid digested by USEPA Method 200.2, Revision 2.8, and samples were analyzed by USEPA Method 200.8, Revision 5.4. Sampling and analysis was completed between July and October 2012.

X-ray powder diffraction analysis of LSL scale

The City of Brandon excavated an LSL from their water distribution system in the summer of 2013 (Figure 3(a)). The interior scale was allowed to dry, and samples were harvested and stored in a desiccator. The scale was analyzed by X-ray powder diffraction (XRD) using a Siemens D5000 powder diffractometer. Scans were carried out over the range of 6–70 degrees 2θ, with 0.02-degree steps. Peak patterns in the resulting diffractogram were processed using MDI Jade software to identify the crystalline species.
Figure 3

(a) Image of LSL cut lengthwise with scale visible. (b) Magnified image of a piece of scale with four layers visible.

Figure 3

(a) Image of LSL cut lengthwise with scale visible. (b) Magnified image of a piece of scale with four layers visible.

Close modal

Free chlorine in the distribution system

Three houses with LSLs were selected for monitoring free chlorine in the distribution system, in flushed samples and samples following a stagnation period of a minimum of 8 hours. In each of the three houses, residents were instructed not to use the water during the night. In the morning, measurements of free chlorine were taken immediately from a sample of water that had been stagnating overnight; and subsequently, from a sample of water collected following a 10-minute flush, a flush time exceeding that required to flush the LSL in order to assure that this sample was drawn from the water main.

Investigation of PbO2 reduction by NOM

The effect of NOM, TOC and AHS on the reduction of PbO2 was investigated in the laboratory. PbO2 was synthesized by adding Pb(NO3)2 ( ≥ 99%, certified ACS, Fisher Scientific, New Jersey, USA) to a solution of NaOCl (reagent-grade, Sigma Aldrich, St Louis, MO, USA) in a final volume of 500 mL to a molar ratio of Pb2+/NaOCl of 0.5. This solution was stirred for 24 hours. The resulting dark brown particles were then filtered through 0.45 μm mixed cellulose ester filters (Fisher Scientific, New Jersey, USA), and the filter papers with the captured particles were rinsed with 0.1 N HNO3 (Fisher Scientific, New Jersey, USA) to remove residual Pb(II) species, and then rinsed thoroughly with DI Milli-Q water (Millipore, Bedford, MA, USA) to remove any remaining acid. This acid rinse has been shown not to significantly impact lead dissolution, and is only an issue at low pH in the presence of strong acids (Lin & Valentine 2008). The particles remaining were rinsed off the filter paper, suspended in DI water and freeze-dried using a Labconco FreeZone freeze drier. The freeze-dried particles were analyzed by XRD using a Siemens D5000 powder diffractometer to be sure that they were pure PbO2.

Samples of treated water were collected from the Brandon water treatment plant prior to chlorination on July 14, 2014. Historic general water quality parameters are outlined in Table 1. Four jars with varying fractions of NOM and corresponding TOC concentrations were prepared using 1-L amber glass jars, with the following conditions: The first jar contained de-ionized water and no NOM; this was confirmed as it had a recorded TOC concentration of <0.5 mg/L. The second jar contained Brandon treated water collected prior to chlorination, with a TOC concentration of 7.1 mg/L. The third jar contained the Brandon treated water used in jar 2, diluted by nearly half with DI water (50% NOM) which yielded a TOC concentration of 3.9 mg/L. The fourth jar contained the Brandon treated water with the operationally-defined AHS fraction removed which yielded a TOC concentration of 4.9 mg/L. The AHS fraction of NOM was removed using the XAD method 5510C from Standard Methods for the Examination of Water and Wastewater. The concentration of TOC remaining after removal of AHS was 4.7 mg/L. TOC was measured using a Teledyne Tekmar Fusion Total Organic Carbon Analyzer. Each jar was adjusted to the approximate average pH, alkalinity and free chlorine conditions of water in the Brandon distribution system (as shown in Table 1). pH was adjusted to 7.8 using concentrated HNO3 or NaOH (Sigma Aldrich, St Louis, MO, USA); alkalinity was adjusted to 74 ± 2 mg/L as CaCO3, using NaHCO3 (Fisher Scientific, New Jersey, USA); and free chlorine was adjusted to 1.0 mg/L as Cl2 at the beginning of the experiment in order to replicate the residual chlorine concentration in the distribution system at houses with LSL, and was allowed to deplete, as occurs during water stagnation in LSLs. Free chlorine was adjusted using NaOCl. All jars were run in triplicate, with the exception of the jar containing DI water, which was run in duplicate.

PbO2 (23.0 ± 0.4 mg) was added to each of four 1-L amber glass jars at time = 0. Jars were shaken on a shaker plate at 100 rpm throughout the 21-day experiment. Samples were taken at 12, 24, 48 and 72 hours and 7, 14 and 21 days in order to capture a profile of lead release over time. Samples were filtered through a 0.2 μm nylon syringe filter (25 mm diameter) (Fisher Scientific, New Jersey, USA) in order to remove the PbO2 particles. It was verified that PbO2 particles could not pass through a 0.2 μm filter by measuring the lead concentration in the filtrate. This ensures only soluble lead was analyzed. It was also verified and measured that NOM colloids could pass through this pore size. If released dissolved lead complexed with or adsorbed to NOM colloids they would pass through the filter and be detected in the filtrate. The samples were acidified to 2% HNO3 and dissolved lead released by reduction of the PbO2 particles throughout the experiment was analyzed by a Perkin-Elmer 800 Graphite Furnace Atomic Absorption spectrometer (GFAA) with a Perkin-Elmer Lead Lumina Hollow Cathode Lamp. Lead was measured at 283.3 nm with a 0.7 nm slit width. All glassware used in the experiments was acid-rinsed.

Lead concentrations in the Brandon water distribution system

Total lead concentrations were found to exceed the provincial standard of 10 μg/L at the tap in 16 of the 20 houses with an LSL sampled in Brandon, following a 6-hour period of stagnation (Figure 4). Eleven of the 20 houses exceeded 50 μg/L and two houses exceeded 100 μg/L. The highest concentration seen was 280 μg/L. Fifteen of these houses had lead concentrations exceeding the standard in the first litre of the first-draw sampling; 15 in the third liter; and 14 in the fourth liter. As a general rule, based on the typical diameter of LSLs and typical distance from the curb to the house in Canada, lead occurring in the first litre can be attributed to leaching from leaded materials from premise plumbing, while lead in the third and fourth litres should originate from the LSL (Health Canada 2009). This indicates that lead is leaching from both the premise plumbing and the LSLs. Lead concentrations in the samples collected after the water had been flushed for 5 minutes and were expected to decrease to below the standard, as this water was drawn from the water main and had not been stagnating in the LSL (Health Canada 2009). However, lead concentrations continued to exceed the standard after the 5-minute flush in 14 of the 20 houses. It is possible that lead in this sample may be due to particulate lead; however, an attempt at size fractionation of water samples from a small number of houses with LSLs yielded unreliable results. This conclusion was based on the fact that particulate lead measured was higher than total lead. This is likely due to adsorption of lead onto the sides of the containers during a holding period of approximately 4 months before analysis. This research is ongoing; the data on particulate lead concentrations will be provided in subsequent publications.
Figure 4

Lead concentrations at the tap of homes with LSLs in Brandon. N = 20. Min-max, 25th and 75th percentiles, and median shown. Standard is shown as the dashed line.

Figure 4

Lead concentrations at the tap of homes with LSLs in Brandon. N = 20. Min-max, 25th and 75th percentiles, and median shown. Standard is shown as the dashed line.

Close modal

XRD analysis of LSL scale

Four layers were observed in the LSL scale (Figure 3(b)). It was not possible to reliably separate the individual layers due to their amalgamation in the LSL. XRD analysis was performed on a powdered sample of all four layers of the scale. As a result, the intensity peaks were not well-defined; however, crystalline solids identified in the scale were PbO2 and Pb5O8 (Figure 5) (Mosseri et al. 1990). PbO2 was confirmed to be of high purity as no other major or minor peaks were present in the XRD of the synthesized PbO2. Kim & Herrera (2010) reported the occurrence of Pb3O4 in a Canadian drinking water system and suggested that the species was an intermediate between oxidation of cerussite (lead carbonate mineral PbCO3) to Pb3O4 to PbO2. The reaction in the case of the BWSS is proposed to be as follows:
formula
1
formula
2
Figure 5

XRD diffractogram of a scale sample from a LSL excavated from the Brandon distribution system, with reference diffractograms below. Parenthesis indicated intensities next to their corresponding peaks.

Figure 5

XRD diffractogram of a scale sample from a LSL excavated from the Brandon distribution system, with reference diffractograms below. Parenthesis indicated intensities next to their corresponding peaks.

Close modal

The identification of PbO2 and Pb5O8 in the LSL scale shows that the scale is being oxidized by the free chlorine residual maintained throughout the distribution system. This also indicates that lead release during stagnation periods is governed by redox reactions, rather than dissolution as would be expected in lead carbonate scales, as PbO2 is sparingly soluble at the pH range of drinking water (Lytle & Schock 2005). During the 6-hour stagnation period in the Brandon water distribution system, the free chlorine residual is depleted by the high concentrations of NOM in the treated water, promoting the formation of more soluble divalent lead solids (primarily lead carbonates). This effect is compounded by NOM reducing PbO2, as reported often in the literature (Dryer & Korshin 2007; Lin & Valentine 2008, 2009; Shi & Stone 2009a).

Free chlorine in the distribution system

Free chlorine was found to decrease in water samples following an 8-hour stagnation period compared to flushed samples drawn directly from three homes in the distribution system. Flushed samples from the tap were found to have free chlorine concentrations in the range of 0.82–0.96 mg/L. After 8 hours of stagnation this range dropped to 0.3–0.72 mg/L. This confirms that free chlorine concentration drops during periods of 8 hours of water when stagnating in the LSLs (such as would commonly occur when residents are at work and asleep), permitting the reduction of PbO2.

Investigation of PbO2 reduction by NOM and AHS

The treatment plant is motivated to remove the AHS fraction in order to meet THM requirements (City of Brandon 2013). The AHS fraction has also been reported to be particularly redox-reactive (Cory & McKnight 2005). For these reasons, the effect of the removal of AHS on dissolved lead release from PbO2 reduction was investigated. Dilution of total NOM non-preferentially was investigated as well.

Results of the jar tests show that after 21 days, removal of the AHS fraction resulted in a 6% reduction of lead release compared to the jar with 100% of NOM, while dilution (1:1) of 50% of the total NOM resulted in a 75% reduction in lead release (Figure 6). This highlights that even after AHS extraction and a 33% removal of TOC, there was only a 6% reduction in lead release. It follows from these numbers that the removal of AHS and TOC are not associated with a significant amount of lead release, but other properties of NOM are. Some of the reasons for this are discussed below.
Figure 6

Concentration of lead (<0.2 μm) released from PbO2 reduction in μg/L by varying concentrations and compositions of DOC and NOM over 21 days. pH is 7.8, alkalinity is 74 mg/L as CaCO3, free chlorine is 1.0 mg/L as Cl2 initially, and is allowed to deplete.

Figure 6

Concentration of lead (<0.2 μm) released from PbO2 reduction in μg/L by varying concentrations and compositions of DOC and NOM over 21 days. pH is 7.8, alkalinity is 74 mg/L as CaCO3, free chlorine is 1.0 mg/L as Cl2 initially, and is allowed to deplete.

Close modal

Although it was unexpected that AHS did not play a significant role in the release of lead, these findings are supported by other studies (Dryer & Korshin 2007; Shi & Stone 2009a). One study found that coagulation of raw water with aluminum sulphate and ferric chloride, which decreased the SUVA at 254 nm (considered to be representative of humic-like substances), resulted in increased lead release from PbO2 relative to the untreated water (Shi & Stone 2009a). This was explained by the presence of a reductive fraction of NOM and a fraction of NOM inhibiting reduction. Furthermore, while another study found that lead release from PbO2 reduction was associated with changes in absorbance at the wavelengths associated with humic-like substances, the authors found that these changes in absorbance due to oxidation of this matter were not sufficient to explain the extent of PbO2 reduction, and suggested that another fraction of NOM was responsible for PbO2 reduction in addition to the humic-like substances (Dryer & Korshin 2007). It has been shown that various inorganic species such as Fe(II) and Mn(II) can have a significant impact on lead released (Shi & Stone 2009b), and the speciation changes depending on oxidation-reduction potential (ORP) of the water. The concentration of these metals were not monitored in this study, but historical concentrations of total Fe and Mn were found to be <0.1 and 0.001 mg/L, respectively, in Brandon city waters. This may be one explanation as the removal of AHS and TOC did not coincide with the decrease in lead release, but the dilution of 50% NOM did. In addition, it has also been found that the presence of monochloroamine may cause lead release via reductive dissolution (Zhang & Lin 2013). These are things that can be further investigated; and knowledge of these effects may be useful in determining a treatment method that will be effective in targeting the removal of the portion of NOM responsible for reducing PbO2.

Characterization of DOC in the Brandon water source

Characterization of the DOC in the source water of the BWSS, the Assiniboine River, was performed in May 2011, with water collected from the Portage la Prairie water treatment plant, approximately 100 km east of the BWSS (Sadrnourmohamadi et al. 2013). The raw water was separated into six fractions: hydrophobic acid (HPOA), neutral and base, and hydrophilic acid, neutral and base. The largest component of the raw Assiniboine River water was the hydrophilic neutral (HPIN) fraction, at 42.2% of the raw water, followed by the HPOA fraction (the fraction containing AHS), at 39.9% (Sadrnourmohamadi et al. 2013). Furthermore, the study found that coagulation significantly decreased the HPOA fraction, and following coagulation, the vast majority of dissolved organic carbon was comprised of the HPIN fraction (Sadrnourmohamadi et al. 2013). It is likely that the dominant fraction remaining after coagulation of the Brandon water is the HPIN fraction, which has been reported to consist of polysaccharides (Leenheer 1981), and can be attributed to microbial degradation. It is possible that the fraction of NOM responsible for its reductant capacity is the HPIN fraction. Further studies are required to investigate the effect of HPIN, and other operational classifications of NOM, on the reduction of PbO2, in order to inform water treatment plant engineers on which processes could have a beneficial effect on concentrations of lead at the tap.

While previous studies have shown contradictory results on whether humic-like substances are primarily responsible for the reduction of PbO2, this study finds that operationally-defined AHS is not responsible for the reductant properties of NOM in the treated water studied. The targeted removal of AHS at the treatment plant may be beneficial to reduce THMs, but will not help in reducing lead concentrations at the tap. However, reduction of total NOM appears to be quite effective in reducing concentrations of lead released from reduction of PbO2. Inorganic and microbial water constituents also have the ability to play a significant role in reduction of PbO2.

This study was funded by the City of Brandon. We thank Patrick Pulak and Alexia Stangherlin of the City of Brandon, and Brad McIntosh, manager of the Brandon Water Supply System, for providing information about the system and for providing the portion of a lead service line. The monitoring portion of this study was the initiative of and was funded by the Manitoba Office of Drinking Water. We thank Dr Housseini Coulibaly and Kim Philip for their help. We also thank the University of Manitoba for their aid in funding this project. The funding sources were not involved in study design.

Agency for Toxic Substances and Disease Registry (ATSDR)
2007
Toxicological Profile for Lead
. .
City of Brandon
2013
Public Water Supply Annual Report
.
(accessed 3 January 2015)
.
City of Brandon
2014
Lead Water Services Information
.
(accessed 13 August 2014)
.
Health Canada
2009
Guidance on Controlling Corrosion in Drinking Water Distribution Systems
. .
Health Canada
2013
Final Human Health State of the Science Report on Lead
. .
Health Canada
2014
Guidelines for Drinking Water Quality–Summary Table. Water and Air Quality Bureau, Healthy Environments and Consumer Safety Branch, Health Canada, Ottawa, Ontario
.
Leenheer
J.
1981
Comprehensive approach to preparative isolation and fractionation of dissolved organic carbon from natural waters and wastewaters
.
Environ. Sci. Technol.
35
,
4805
4816
.
Lytle
D.
Schock
M.
2005
Formation of Pb(IV) oxides in a chlorinated water
.
J. AWWA
97
(
11
),
102
114
.
Province of Manitoba
2007
The Drinking Water Safety Act: Drinking Water Quality Standards Regulation
. .
Schock
M.
Wagner
I.
Oliphant
R.
1996
Corrosion and solubility of lead in drinking water, Chapter 4
. In:
Internal Corrosion of Water Distribution Systems
, 2nd edn.
American Water Works Association Research Foundation
,
Denver, CO
.
Thurman
E. M.
Malcolm
R. L.
1981
Preparative isolation of aquatic humic substances
.
Environ. Sci. Technol.
15
(
4
),
463
466
.
Thurman
E.
1985
Organic Geochemistry of Natural Waters
.
Martinus Nijhoff/Dr. W. Junk Publishers
,
Dordrecht
,
The Netherlands
.
USEPA (United States Environmental Protection Agency) Method 200.2, Revision 2.8
.
Sample Preparation Procedure for Spectrochemical Determination of Total Recoverable Elements
.
Office of Research and Development
,
Cincinnati, OH
.
USEPA (United States Environmental Protection Agency) Method 200.8, Revision 5.4
.
Determination of Trace Elements in Waters and Wastes by Inductively Coupled Plasma – Mass Spectrometry
.
Office of Research and Development
,
Cincinnati, OH
.
World Health Organization
2011
Guidelines for Drinking Water Quality
, 4th edn.
Water Sanitation Health
,
Geneva
.