Simulation Pilot Study On The Effect Of Water Source Switch On Iron Release In Drinking Water Distribution Systems

In a northern China city, Danjiangkou water will be the main source water and thenfrequent source water switch between Danjiangkou reservoir water and local source waterswill be a routine water supply pattern when the middle route ofSouth-to-North-Water-Diversion Project is completed in2014. In the paper, the possibleeffects of coming source water switch to Danjiangkou water and frequent source water switchon the iron corrosion product release in drinking water distribution systems were studied.Unlined cast iron pipes from4different sites were harvested from existing distributionsystems of the northern city and transferred to the pilot study base in Danjiangkou city, and4pipe-loop simulating distribution systems were established. The test iron pipes werehistorically transporting different source waters, including local surface and ground waters.More than sixteen month-consecutive monitoring was conducted to investigate the waterquality changes after water sources switched, which was divided into seven stages.In the first stage, six months of consecutive monitoring was conducted to investigate thewater quality changes after source water switched to Danjiangkou water. Results showed thatthe Larson ratio of Danjiangkou water was around0.5, lower than the surface water of thenorthern city but higher than the ground water of the northern city. After source water switch,the iron pipe historically transported ground water underwent a long period (75days) of highiron release with obvious “red water” phenomenon, while the other three pipes historicallymainly transported surface waters had much lower iron release except during the initial phase(about30days) of source water switch. The total iron concentration and turbidity of effluentwater had significant linear positive relationship. Compared with the influent water qualityparameters, the pH, dissolved oxygen, alkalinity and hardness of effluent water decreased with the iron release processes in the simulating pipe systems. The different iron releaseresponses of4pipe systems were dependent on the characteristics of corrosion scales formedunder different water source conditions. Corrosion scales on pipes historically transportedground waters were relatively thin and no layered structure, while the corrosion scales onpipes historically transported surface waters were rather thick and mainly in tubercle formwith typical layered structure.In order to mimic more source water switch scenarios, based on more than6months ofconsecutive experiment conducted to investigate the effluent water quality changes afterDanjiangkou water source switched, the sulfate, chloride concentrations, the pH value,chloramines and dissolved oxygen (DO) of Danjiangkou water were adjusted to observe theiron release behaviors. Results showed that both increased sulfate and chloride concentrationshad slight effect on iron release in the cast iron pipes after the test pipes were stable forDanjiangkou water, and both iron concentration and turbidity in effluent met nationalStandards for Drinking Water Quality (GB5749-2006). However, when the hydraulicresidence time (HRT) increased from8h to24h and with chloride being100mg/L, both ironconcentration and turbidity rose significantly. In addition, the iron release corresponding tothe iron pipes historically transported ground water was above the ones historicallytransported surface water. When pH was between7.0and8.2, changing pH had a little effecton iron release, with both iron concentration and turbidity in effluent meeting the nationalstandards for drinking water quality of China. When HRT was constant, the total ironconcentration rose with chloramines increasing. What’s worse, iron released seriously frompipe wall when HRT lengthened. There were two sides to iron release with DO changing. Theiron release was controlled with DO being below a value, beyond which iron release waspromoted, but HRT still had a significant impact on iron release. Source water qualitycharacteristics, corrosion scale structural features and HRT are among the most importantfactors affecting iron release in drinking water distribution systems under multi-source waterswitch conditions.With SEM, XRF, XRD and XPS being used, results showed that external corrosionscales were smooth and dense, and inner corrosion scales were porous and loose. β-FeOOHand FeCO3that are instable existed in corrosion scales. The chemical constituents changedaccordingly after water sources switched. As a consequence of this, stable corrosion scalesdominated by Fe3O4and α-FeOOH were formed. XRD results showed that Fe3O4andα-FeOOH were above50%in the corrosion scales of stable drinking water distributionsystem.