C.R. Hayes,     Swansea University, UK

10.1 Introduction

Lead metal has been used since Roman times as a building material, particularly for conveying water. It has also been used widely as a roofing material. Lead metal is malleable and corrodes very slowly, making it an ideal material for these applications. Unfortunately, lead is toxic, the full extent of which has been increasingly realised in the modern era since the 1970s, as outlined in Section 10.3 below. Earlier accounts of the impact of lead on human health speculate the fall of the Roman Empire being influenced by lead poisoning (Dutrizac et al., 1982; Hodge, 1981; Nriagu, 1983) whereas Troesken (2006) gives an historical account of lead poisoning from lead pipes that spans the nineteenth and early twentieth centuries.

Health problems from lead roofing and guttering will be limited to occupational exposure through skin absorption and are unlikely to be much of a problem today if simple precautions are observed, particularly the use of gloves. Routine health and safety risk assessments should identify any workers at risk and lead exposure can easily be checked in blood samples.

The major issue relates to the historic use of lead pipes to supply drinking water. Lead pipes were commonly used to connect dwellings to the water main in the street, the latter being constructed of cast iron and more recently of asbestos-cement and various plastics. Lead pipe connections (also known as service lines) had the advantage of being able to withstand localised ground movements better, such as the expansion and shrinkage of clay soils, resulting in fewer bursts and less leakage.

This chapter describes the extent of lead piping in Europe, the associated potential extent of health problems and options for control.

10.2 Use of lead in buildings and drinking water contamination

It is common for a lead connection pipe to come under joint ownership. The first part, owned by the water company or local municipality, runs from the water main to the boundary of the property. The second part (in continuity) runs from the boundary of the property to the dwelling (house or apartment building) and is normally owned by the property holder. This split ownership has legal and financial implications when lead connection pipe removal is contemplated. As lead connection pipes are buried in the ground, their presence or absence is difficult to determine. Lead piping was also used in household plumbing and will often be more conspicuous except when buried in walls.

The use of lead connection pipes and lead pipes within dwellings began with the industrial revolution of the early 1800 s when industrialisation prompted urbanisation and mass housing. Despite concerns about the health hazards of lead pipes being expressed since the nineteenth century (Troesken, 2006), the installation of lead pipes continued until the mid-1980s in both Europe and North America.

The greater the length and diameter of the lead piping, the greater the risk, which depends on the surface area of the lead piping in contact with the water. Lead pipe lengths commonly vary from just a few metres to several hundred metres, depending on the connection distance and size of dwelling. The most common lead pipe internal diameter is half-inch or 12 mm and is typical for connections to individual dwellings that house a single family. For larger buildings such as apartment blocks, internal diameters are often 18 or 25 mm and sometimes 37 mm.

It is estimated (Hayes and Skubala, 2009) that up to 25% of houses could be supplied with drinking water through a lead pipe in Europe, on the basis of the early estimates shown in Table 10.1. The percentage of homes with lead pipes is known to be as high as 75% in some older cities and towns in Europe. In the US and Canada, the percentage of homes with lead service lines is estimated at between 2 and 3% (IWA, 2010a).

The amount of lead that builds up in the drinking water depends on its corrosivity and how long the water is in contact with the lead pipe; concentrations can vary from < 1 to several hundred μg/l. The basic mechanism is one of dissolution of the lead corrosion film inside the lead pipe. The corrosion film is typically dominated by lead carbonate unless phosphate-based corrosion inhibitors are in use. Particulate lead can also form if loose iron corrosion products (rust from old cast iron water mains) come into contact with the lead pipe corrosion film. A further mechanism for lead dissolution is organic chelation by natural organic matter present in the drinking water, particularly humic and fulvic acids (organic colour from peat bogs).

Particularly with corrosive waters, elevated lead in drinking water can occur, even with newer plumbing systems when no lead pipes are present, arising from brass or leaded solder. Brass alloys contain lead to improve machining characteristics and brasses with up to 8% lead have been permitted in the US since 1986, although recent legislation in the US will in the future limit the lead content of brass to 0.25%. Leaching of lead from brass to drinking water has been demonstrated (Kimbrough, 2007) and Triantafyllidou and Edwards (2011) point out that 77% of the homes in the USA (built pre-1986) could have brass plumbing components in which the lead content exceeds 8%, putting 81 million homes at risk in the USA from ‘legacy leaded brass’.

Solders containing high percentages of lead were in common use for jointing copper pipes but were banned in the mid-1980s throughout Europe and North America. Galvanic corrosion of the lead in the solder can result in very high concentrations (> 100 μg/l) in drinking water (Gregory, 1990), although the extent to which such effects diminish over time is largely unquantified. Water quality changes that would affect galvanic corrosion reactions, such as changes in disinfection or oxidants, lowering of pH and significant increases in the chloride content of the drinking water relative to sulphate, have been observed (Edwards and Triantafyllidou, 2007) to aggravate lead release from ‘legacy leaded solder’.

10.3 Toxicity of lead to humans

Lead is toxic to humans and lead poisoning is exhibited by a wide range of clinical conditions (Canfield et al., 2003; Gump et al., 2008; Jusko et al., 2008; Labat et al., 2006; Lanphear et al., 2000, 2002; Menke et al., ; Khalil et al., 2009; Pocock et al, 1994; Rabin, 2008; Tararbit et al., 2009; Wilhelm and Dieter, 2003). Adverse health effects include interference with haemoglobin biosynthesis, interference with calcium and vitamin D metabolism, gastrointestinal irritation, dullness, restlessness, irritability, poor attention span, headaches, muscle tremor, abdominal cramps, kidney damage, hallucination, loss of memory, encephalopathy, hearing impairment, gonad dysfunction, and violent behaviour. Lead can accumulate in bone and fatty tissue, with subsequent release, particularly during the later stages of pregnancy. In the US for several years, the trigger for corrective action has been lead in blood above 10 μg/dl (CDC, 1991). Most attention has been directed towards the retardation of child development, especially reductions in IQ. The World Health Organization in its booklet on Childhood Lead Poisoning (WHO, 2010) has drawn attention to the following:

The Joint FAO/WHO Expert Committee on Food Additives re-evaluated lead in June 2010 and withdrew the provisional tolerable weekly intake guideline value for lead on the grounds that it was inadequate to protect against IQ loss. This guideline value had been used as a basis for determining WHO’s guideline value for lead in drinking water of 10 μg/l. The new WHO guidelines (WHO, 2011) retain the guideline value for lead in drinking water of 10 μg/l, but as a provisional value due to current achievability.

Environmental exposure to lead includes paint, dust, petrol in those few countries that have yet to stop using lead additives, air pollution from burning waste containing lead and lead battery recycling, in addition to lead from drinking water (WHO, 2010). Circumstances will vary very much on a local basis; however, lead from drinking water will pose risks if lead pipes are still in service (Hayes and Hoekstra, 2010) and from lead leaching from brass and solder in some cases.

Lead in drinking water has been correlated with lead in blood in numerous studies (IWA, 2010a) in general terms. The relationship between lead in drinking water and lead in blood will vary for individuals as a function of the amounts ingested, age and body weight. On the basis of the epidemiological studies reported by Quinn and Sherlock (1990), a general curvilinear relationship applies where an average water lead concentration of around 20 μg/l equates to a blood lead level of 10 to 15 μg/dl (as illustrated in Fig. 10.1).

10.4 Assessing the risk associated with lead in drinking water

The extent of problems with lead in drinking water has been underestimated or even overlooked completely as a consequence of poor sampling or not even taking samples at all (IWA, 2010a), undermining regulatory efforts to control lead in drinking water in both Europe and North America. On the basis that lead contamination arises predominantly from domestic pipework systems (pipes and fittings), samples taken upstream from a distribution network are meaningless. The most appropriate sampling point must be where water is drawn for drinking or cooking purposes. However, samples taken from the consumer’s point of use (i.e. at the tap outlet, normally in the kitchen) after flushing the domestic pipework system are equally meaningless.

The most appropriate sampling methods for use at the consumer’s tap are as follows:

An advantage of methods 1 and 2 is that they are amenable logistically to the use of trained sampling personnel, as opposed to having to rely on consumers (method 3).

In Europe the current standard for lead in drinking water is 25 μg/l, dropping to 10 μg/l in December 2013, the same as the current WHO guideline value. In all cases, these standards relate to the weekly average concentration ingested. At the zonal level, RDT sampling provides an adequate basis for assessing compliance (IWA, 2010a). In the US, the lead standard is 15 μg/l based on the 90th percentile of survey samples (first draw after at least 6 hours’ standing) taken from dwellings where lead service lines are present and/or leaded solder was used after 1983. Recent developments in risk assessment at the zonal level (Hayes, 2010) indicate that many water supply systems in Europe can be expected to have significant levels of non-compliance with the WHO guideline value of 10 μg/l (and EU standard from December 2013), with up to 50% dwellings affected. This extent of non-compliance is borne out by numerous case studies (IWA, 2010a) and confirms the need for widespread corrective action. Computational modelling (Van der Leer et al., 2002; Hayes, 2010) can predict the extent of non-compliance with the WHO guideline value, based on the plumbosolvency of the water supply and the system’s pipework characteristics, and can predict a zonal failure profile of the extent and severity of non-compliance. This is illustrated in Fig. 10.1 for a city in which 70% of houses have a lead pipe and for high plumbosolvency water supplies. The figure shows an assumed general relationship between water lead and blood lead (from Quinn and Sherlock, 1990), and the percentage of houses where blood lead concentrations are predicted to be 10, 17 and 20 μg/dl. In this example, 23% of houses are associated with a risk condition equating to an IQ loss of 6 points and 1% of houses are associated with a risk condition equating to an IQ loss of between 8.5 and 11 points (with 5% of houses having an intermediate risk level).

The borderline nature of the current EU standard of 25 μg/l, in terms of risk reduction, is apparent and even the future EU standard of 10 μg/l (and current WHO guideline value) can be seen not to afford total risk reduction as defined by IQ loss. For the time being, the WHO guideline value of 10 μg/l should be taken as the basis for determining risk, even though it is now a provisional guideline value (WHO, 2011). In recognition of the severity of health effects in children, there is a clear need to establish the extent of compliance with the WHO guideline value, in every water supply system operated by a water company or municipality, consistent with recommendations made to the European Commission (Hoekstra et al., 2008) concerning revisions to EC Directive 98/83/EC. Risks from lead in drinking water also need to be assessed in the very numerous small and very small water supplies that are often privately owned (IWA, 2010b).

Assessment tools that can help to establish the extent of risk in a water supply system are as follows (IWA, 2010a):

Health concerns relating to lead in drinking water at individual properties can be assessed by sequential stagnation sampling (to get an indication of how much lead is present), pipework inspection and blood lead surveillance (particularly in children).

10.5 Lead pipe replacement and fittings containing lead

There are two common myths which need to be dispelled:

Partial lead pipe removal by water companies does not solve the problem when, as is commonly the case, the home owners’ lead service pipes and/or internal lead piping remain (DWI, 2010). Indeed, it can exacerbate the situation, at least in the short term (Renner, 2010). Where concerted efforts have been made to remove all lead pipes, home owners have mostly failed to cooperate (as in The Hague) because of the costs and inconvenience involved. Total lead pipe removal is of course the ambition, but because this is both expensive and disruptive, it will remain unrealistic until such time as legislation requires owners to make their homes lead pipe-free. One way would be to require a home to be certified lead pipe-free at the time of sale or letting. The estimated cost for total lead pipe removal at a home (IWA, 2010a) is between €1000 and €4000, equating to many tens of billions of euros at the European scale.

Because of the practical and financial constraints associated with total lead pipe removal, the most robust solution in the short to medium term for larger systems will be centralised corrective water treatment. Flushing pipework prior to drinking water use has been advocated, but recent research (Hayes and Croft, 2011) suggests that flushing would be required prior to every flow event, resulting in unacceptable increases in water consumption. The application of point-of-use filters that can remove lead from water might also be considered, particularly for very small supply systems or as an interim measure at individual dwellings.

Brass fittings containing > 8% lead were used until the mid-1980s. Thereafter, lead-free brass with up to 8% lead has been permitted. As brass fittings have been in common use for many years, it follows that a very large number of homes that have a domestic pipework system will have brass fittings in service that have an appreciable lead content. It is difficult to envisage the removal of these brass fittings except under specific circumstances where lead leaching problems are considered significant. The use of leaded solder was also permitted up to the mid-1980s to joint copper pipes. Again, a very large number of homes will still have pipework in service that was jointed by leaded solder and it is clearly impracticable to replace these pipework systems, except under specific circumstances. The implication is that all water supply systems need a corrosion control strategy to minimise these potential legacy effects. Concurrently, all countries need to move towards the use of brass with < 0.25% lead (as in the US) and prohibit the use of leaded solders, using effective regulatory controls.

10.6 Corrective water treatment

Laboratory testing (Hayes, 2008) has indicated that all types of drinking water are sufficiently plumbosolvent to be capable of dissolving lead to concentrations that exceed the WHO Guideline Value of 10 μg/l when in contact with a lead pipe for 30 minutes, unless specifically treated to reduce plumbosolvency. To reduce the plumbosolvency of drinking water it is necessary to either increase pH (often to above 9.0) or add a corrosion inhibitor (orthophosphate works best) or both. Plumbosolvency reduction by treatment is widely practised in the UK, the Netherlands and the USA (and increasingly in Canada) but not elsewhere to the same extent. This means that most drinking water outside the UK, the Netherlands and North America continues to be sufficiently plumbosolvent to cause problems (as defined by the WHO Guideline Value).

Again, there are several common myths which need to be dispelled:

The relationship between lead solubility, pH and alkalinity (Fig. 10.2) can of course be exploited and pH elevation can be successful in overcoming problems with lead in drinking water in some areas. However, particularly in areas with a high number of lead pipes, pH elevation alone may not be successful, a good example being The Hague (IWA, 2010a) where after corrective water treatment (softening and pH elevation) and the removal of most of the water company’s lead pipes, 23% of random samples continued to exceed the WHO Guideline Value. In such cases, and in cases where natural organic matter is present in the drinking water, a corrosion inhibitor must be used if high levels of compliance with the WHO Guideline Value are to be achieved.

The most successful corrosion inhibitor is orthophosphate (IWA, 2010a; Cardew, 2009). In the UK, 95% of the public water supplies are dosed with orthophosphate and after optimisation 99% of random samples in England and Wales complied in 2010 with the WHO’s Guideline Value. In Scotland and Northern Ireland the level of compliance was slightly lower but improving. The UK experience clearly demonstrates that effective plumbosolvency control can be achieved by dosing orthophosphate and that childhood lead poisoning from drinking water can be considerably abated during the time it will take to remove all the lead pipes. There is no evidence that orthophosphate dosing has any health or adverse environmental consequences. It is apparent that orthophosphate dosing also reduces cuprosolvency (Comber et al., 2011) and the leaching of nickel from chrome-nickel plated components (Hayes, 2011).

The unit cost of orthophosphate is about 0.2 p/m³ and NPV analysis confirms that this is a considerably lower cost option than lead pipe replacement (IWA, 2010a).

10.7 Recommendations

10.8 Sources of further information and advice

The Best Practice Guides published by the International Water Association (IWA, 2010a, 2010b) will provide the reader with much more detail than that given in this chapter. These are available via www.iwap.co.uk.

10.9 References