The purpose of this study was to identify the effect that water quality, pipe material, pipe size, flow conditions and the use of corrosion inhibitors would have on the rate of free chlorine and chloramine decay in distribution systems. Empirical models were developed to predict the disinfectant residual concentration with time based on the parameters that affected it. Different water treatment processes were used to treat groundwater and surface water to obtain 7 types of finished waters with a wide range of water quality characteristics. The groundwater was treated either by conventional treatment by aeration (G1) or softening (G2) or high pressure reverse osmosis (RO) and the surface water was treated either by enhanced coagulation, ozonation and GAC filtration (CSF-O3-GAC or S1) or an integrated membrane system (CSF-NF or S2). The remaining two water types were obtained by treating a blend of G1, S1 and RO by softening (S2) and nanofiltration (G4). A pilot distribution systems (PDS) consisting of eighteen (18) lines was built using old pipes obtained from existing distribution system. The pipe materials used were polyvinyl chloride (PVC), lined cast iron (LCI), unlined cast iron (UCI) and galvanized steel (G). During the first stage of the study, the 7 types of water were blended and fed to the PDS to study the effect of feed water quality changes on PDS effluent water quality, and specifically disinfectant residual. Both free chlorine and chloramines were used as disinfectant and the PDSs were operated at hydraulic retention times (HRT) of 2 and 5 days. The PDSs were periodically tested for free and combined chlorine, organic content, temperature, pH, turbidity and color. The data obtained were used to develop separate models for free chlorine and chloramines. The best fit model was a first-order kinetic model with respect to initial disinfectant concentration that is dependent on the pipe material, pipe diameter and the organic content and temperature of the water. Turbidity, color and pH were found to be not significant for the range of values observed. The models contain two decay constants, the first constant (KB) accounts for the decay due to reaction in the bulk liquid and is affected by the organics and temperature while the second constant, KW, represents the reactions at the pipe wall and is affected by the temperature of the water and the pipe material and diameter. The rate of free chlorine and chloramine decay was found to be highly affected by the pipe material, the decay was faster in unlined metallic pipes (UCI and G) and slower in the synthetic (PVC) and lined pipes (LCI). The models showed that the rate of disinfectant residual loss increases with the increase of temperature or the organics in the water irrespective of pipe material. During the second part of the study, corrosion control inhibitors were added to a blend of S1, G1 and RO that fed all the hybrid PDSs. The inhibitors used were: orthophosphate, blended ortho-polyphosphate, zinc orthophosphate and sodium silicate. Three PDSs were used for each inhibitor type, for a total of 12 PDSs, to study the effect of low, medium and high dose on water quality. Two PDSs were used as control, fed with the blend without any inhibitor addition. The control PDSs were used to observe the effect of pH control on water quality and compare to the inhibitor use. One of the control PDSs (called PDS 13) had the pH adjusted to be equal to the saturation pH in relation to calcium carbonate precipitation (pHs) while the pH of the other control PDS (PDS 14) was adjusted to be 0.3 pH units above the pHs. The disinfectant used for this part of the study was chloramine and the flow rates were set to obtain a HRT of 2 days. The chloramine demand was the same for PDS 14 and all the PDSs receiving inhibitors. PDS 13 had a chloramine demand greater than any other PDS. The lowest chloramine demand was observed in PDS 12, which received silicate inhibitor at a dose of 12 mg/L, and presented the highest pH. The elevation of pH of the water seems to reduce the rate of decay of chloramines while the use of corrosion inhibitors did not have any effect. on the rate of chloramine decay. The PDS were monitored for chloramine residual, temperature, pH, phosphate, reactive silica, and organic content. Empirical models were developed for the dissipation of chloramine in the pilot distribution systems as a function of time, pipe material, pipe diameter and water quality. Terms accounting for the effect of pH and the type and dose of corrosion inhibitor were included in the model. The use of phosphate-based or silica-based corrosion inhibitors was found to have no effect on the rate of chloramine dissipation in any of the pipe materials. Only the increase of pH was found to decrease the rate of chloramine decay. The model to best describe the decay of chloramine in the pilot distribution systems was a first-order kinetic model containing separate rate constants for the bulk reactions, pH effect and the pipe wall reactions. The rate of chloramine decay was dependent on the material and diameter of the pipe, and the temperature, pH and organic content of the water. The rate of chloramine decay was low for PVC and LCI, and more elevated in UCI and G pipes. Small diameter pipes and higher temperatures increase the rate of chlorine decay irrespective of pipe material. Additional experiments were conducted to evaluate the effect of flow velocity on chloramine decay in a pilot distribution system (PDS) for different pipe materials and water qualities. The experiments were done using the single material lines and the flow velocity of the water was varied to obtain Reynolds' numbers from 50 to 8000. A subset of experiments included the addition of blended orthophosphate corrosion inhibitor (BOP) at a dose of 1.0 mg/L as P to evaluate the effect of the inhibitor on chloramine decay. The effect of Reynolds' number on the overall chloramine decay rate (K) and the wall decay rate constant (W) was assessed for PVC, LCI, UCI, and G pipes. PVC and LCI showed no change on the rate of chloramine decay at any flow velocity. UCI and G pipes showed a rapid increase on the wall decay rate under laminar conditions (Re ≤ 500) followed by a more gradual increase under fully turbulent flow conditions (Re ≥ 2000). The use of the BOP inhibitor did not have an effect on the rate of chloramine decay for any of the pipe materials studied. Linear correlations were developed to adjust the rate of chloramine decay at the pipe wall for UCI and G depending on the Reynolds' number.
Identifer | oai:union.ndltd.org:ucf.edu/oai:stars.library.ucf.edu:etd-4066 |
Date | 01 January 2007 |
Creators | Arevalo, Jorge Miguel |
Publisher | STARS |
Source Sets | University of Central Florida |
Language | English |
Detected Language | English |
Type | text |
Format | application/pdf |
Source | Electronic Theses and Dissertations |
Page generated in 0.0027 seconds