Laboratory plumbosolvency testing

Laboratory plumbosolvency testing to directly determine the corrosivity of the water supply to lead pipes (Section 10.1). 

The ortho-phosphate doses needed to achieve the UK criterion, that no more than 2% of random daytime samples exceed 10 pg/1, varied from 0.5 to 2.0 mg/1 (P) in the great majority of cases. Exceptionally, with poorly treated, highly coloured, low alkalinity waters, a dose of 2.8 mg/1 (P) was found to be necessary. Table 8.3 illustrates the dose requirements for low and high alkalinity waters, based on extensive laboratory plumbosolvency testing. Such doses of ortho-phosphate are considered to be entirely safe, given that ortho-phosphate concentrations are many thousands of times higher in many carbonated soft drinks. 

As M (the initial mass transfer rate which determines the initial slope of the dissolution curve) and E (the equilibrium concentration) reduce, the water is less plumbosolvent (less lead dissolves curves A to C) and these factors can be determined by stagnation sampling at appropriate reference houses or by laboratory plumbosolvency testing. Curves A1 and A2 differ in shape as a consequence of the relationship between the 30 minutes stagnation and equilibrium concentrations, which vary for individual waters (Hayes, 2008). The exponential curve and the assumption of plug flow are both approximations, but they enable the computational demands of the model to be greatly reduced. Extensive research (Hayes, 2002 and Van der Leer et al, 2002) has demonstrated that these approximations are adequate when compared to the more scientifically exact diffusion model and the three dimensional simulation of turbulent flow. 

From 1999 to 2002, 17.5% of random daytime samples had lead >10 pg/1 and 7.9% >25 pg/1. Laboratory plumbosolvency testing gave an average lead concentration after 30 minutes contact of around 50 pg/1 Zonal modelling indicated that an average ortho-phosphate dose in the range 0.6 to 0.9 mg/1 (P) should achieve >98% compliance with the fiiture lead standard of 10 pg/1. 

Laboratory plumbosolvency testing coupled with computational comphance modeUing to quickly determine the likely optimum dose, which should then be confirmed (and adjusted if necessary) by routine monitoring of in-situ lead pipes at consumers houses this approach can minimise the number of iterative changes to water treatment conditions and can save both time and money. 

The initial mass transfer rate (M in pg/m /s) and the equilibrium concentration for lead E in pg/1) can be derived from rapid laboratory plumbosolvency testing if Pb(ll) chemistry dominates the corrosion scale within the lead pipes. However, no such data was available. Instead, reference was made to sequential sampling exercises by both the Utility and the State Health Authority, and to LCR survey results, in an attempt to pick out high lead concentrations that might be reasonably representative of equUibrium conditions ( ). 

The orthophosphate dosing scenario A with M = 0.02 and = 30 is typical of many optimised systems in the UK and should be achievable with a relatively low orthophosphate dose. It is predicted that LCR compliance would be achieved, based on the first litre drawn after stagnation, but not if compliance was based on further sequential samples. This is also the case with B when M= 0.015 and E = 22.5. To achieve LCR compliance based on sequential sampUng would require M= 0.01 and =15 (scenario C), likely to be at the extreme range of orthophosphate s abihty to suppress lead solvency. Laboratory plumbosolvency testing will be necessary to confirm the orthophosphate dosing response of the water in City A (i.e. the doses required for scenarios A, B and C), and would help to confirm these comphance predictions (which would be subject to operational feasibility). 

Analysis of corrosion deposits from 6 lead pipes exhumed from City A indicated a dominance of Pb(II) compounds, which are more soluble than Pb(IV) compounds. This is consistent with the moderately high lead concentrations found in both LCR surveys and sequential sampling exercises. Despite pH elevation to 10.0 at the treatment works (which does not normally fall below 9.5 in the distribution network), 7 of the 9 surveys failed the LCR for lead over the period 2007 to 2011. The dominance of Pb(II) compounds also means that rapid laboratory plumbosolvency testing can be used in the quantification of M and E, and that the supply system should be amenable to orthophosphate treatment (albeit the transition characteristics of lead corrosion deposits under high pH conditions would have to be determined). 

With this in mind, the three orthophosphate dosing scenarios were modelled. The values of M (0.02) and E (30) are readily achievable, based on UK experience, and demonstrated a significant further reduction in lead concentrations that would (if achieved in practice) comply with the LCR. To determine the likely dose of orthophosphate, it would be a fairly simple matter to undertake laboratory plumbosolvency testing with results obtainable within a month. Any organic influences should be captured by such testing. 

However, if the values of M and E for City B were found to be higher, as might be determined by laboratory plumbosolvency testing, it would cast into doubt the assumptions about pipe lengths and diameters, which are key drivers in determining... 

There is further evidence of problems with lead in Europe. Extensive laboratory based plumbosolvency testing (see Section 10.2) in the UK using the method of Colling et al. (1987) indicates (Hayes, 2008) that most, and possibly aU, types of drinking waters in supply are likely to be sufficiently plumbosolvent so as to cause non-compliance with both the EU standards for lead (i.e. 10 and 25 pg/1) wherever lead pipes are present (if corrosion inhibitors are not dosed). This data is summarised in Table 5.2 for the treated water from 158 water treatment works in the UK (obtained over the period 1999 to 2004), prior to any dosing of corrosion inhibitor, at the test temperature of 25 °C, for three simplified categories. 

Step 2 Determine the plumbosolvency of the treated water input(s) to the water supply zone and a representative number of locations within the zone this will typically require five samples to be tested using the established laboratory procedure of Colling et al. (1987). Test results can be obtained within one month. 

Laboratory testing using sections of new lead pipe is reproducible and can determine the plumbosolvency of drinking water. 

Laboratory testing to confirm the current plumbosolvency of City B s water supplies and orthophosphate dose responses at different pH values could be considered. 

Presently, too little data has been provided to be able to calibrate the zonal compliance model to the extent possible for Cities A and B. However, a speculative exercise has been undertaken, based on the pipe-work characteristics of Cities A and B and assuming plumbosolvency factors ranging from Af= 0.03 to 0.05 and = 45 to 75. These plumbosolvency factors reflect (i) the lead concentrations observed in the sequential sampling surveys, and (ii) reference to extensive UK data from laboratory testing for low orthophosphate doses to moderately high alkalinity waters. 

To utilise these computational modelling tools further will require better calibration data. Better indicative results could quickly be obtained using assumptions about pipe-work circumstances if the level of plumbosolvency could be determined by laboratory testing. It would also be necessary to firm up on an estimate of the percentage occurrence of lead service lines in the City. This would enable a better assessment of the extent of plumbosolvency problems in City C to be determined, together with a better understanding of the effectiveness of orthophosphate dosing. 

Laboratory testing could quickly clarify the plumbosolvency characteristics of the City s water supplies. 

---