Authors: Professor Tom Casey, Aquavarra Research Limited; Hugh Kerr, Donegal County Council; and Pat Kearney, UCD School of Civil, Structural and Environmental Engineering Chlorine has been the drinking water disinfectant of choice in Ireland and worldwide for many years and as a result there is a well-established underlying process design methodology coupled with an operational focus on the delivery of process reliability. In this regard, the recent publication of an updated version of the EPA Water Treatment Manual on Disinfection (EPA, 2011), providing comprehensive design and operational disinfection guidelines, is a welcome source of up-to-date information for technical personnel engaged in water supply practice in Ireland. However, despite these facts there remains a significant degree of non-compliance with the current Irish Drinking Water Regulations (S.I. No. 278, 2007) associated with disinfection/chlorination. According to the EPA Report on the Provision and Quality of Drinking Water in Ireland (EPA, 2011) the number of Public Water Supplies reporting water quality parametric non-compliance to the EPA in 2009 and 2010 predominantly related to E.coli and total trihalomethanes (TTHMs) transgressions. The presence of E.coli in drinking water indicates a failure of the disinfection process while the presence of excessive THMs reflects the reaction between the added chlorine and natural organic matter (NOM) in the treated water. It is therefore opportune to review chlorination practice both in relation to its efficacy in providing primary disinfection and its role in the generation of THMs. [login type="readmore"] PRIMARY DISINFECTION The term primary disinfection, as used here, refers to the disinfection process as applied at a water treatment works (WTW), where the process design requirement is to ensure that the water leaving the process is hygienically safe and meets regulatory disinfection standards. The disinfection effectiveness of chlorine depends on the chlorine dose and chlorine contact time. Added chlorine reacts with various constituents in the water to which it is dosed resulting in a progressive reduction in the residual free chlorine (RFC) with contact time. The product of RFC concentration and contact time gives the process CT-value (mg.min/l) which is a key measure of the disinfection potential of the disinfection process and hence is a key process design variable. Hence, the RFC decay profile over time has a very significant influence on the CT value achieved in a given contact time. It is commonly observed that the decay rate of chlorine is rapid immediately after dosing and then experiences a decreasing rate of reduction with increasing contact time. Chlorine reacts rapidly with inorganic substances in a reduced valence state such as iron, manganese, ammonia and bromide. In general, the concentration of these inorganic reactants in Irish surface waters is very low. Chlorine also reacts with natural organic matter (NOM) at varying rates to form a range of organochloride compounds, some of which may pose a potential threat to the health of consumers (Hrudey & Charrois, 2012). Of these resulting disinfection by-products (DBPs), only the THMs are currently regulated by EU and Irish drinking water standards. NOM is the ill-defined complex matrix of organic material found in natural waters, derived mainly from living or decayed vegetation and microbial decomposition processes. While the overall NOM amount in water is conventionally quantified by its total organic carbon (TOC) content or dissolved organic carbon (DOC) content, these parameters reveal little about its detailed chemical composition or the reactivity of component constituents with chlorine. There is no universally accepted design chlorination CT value for drinking water production. The EPA Disinfection Manual (EPA, 2011) cites the original WHO recommendations for the use of chlorine as a disinfectant, which stipulated a minimum free chlorine concentration of 0.5 mg/l (C) after 30 minutes contact time (T) at a pH of less than 8, provided that the turbidity is less than 1 NTU, which it took to be equivalent to a CT of 30 x 0.5 = 15 mg.min/l. The same CT value could be achieved by having a longer contact time at a lower RFC, for example an RFC of 0.3 mg/l for a contact time of 50 minutes. While extending the contact time allows the RFC to be reduced, it should not be set below 0.2 mg/l. As protozoan pathogens such as Giardia and Cryptosporidium are effectively resistant to disinfection by chlorine, the primary disinfection targets of chlorination are pathogenic bacteria and viruses, the latter being significantly more resistant than the former. Table 1 contains virus inactivation recommendations given in Annex 1 of the WHO Guidelines for Drinking Water Quality, 4th Edition, 2011:

Table 1: Disinfection CT for 99% Virus Inactivation (WHO, 2011)

TREATED WATER CHLORINE DEMAND SURVEY

The Chlorine Demand parameter, as used in this presentation, is defined as the reduction in RFC with contact time due to its reaction with various constituents in the water. Hence, its value increases with contact time. Also, the term Treated Water, as used here, refers to surface water that has been WTW-processed to the pre-chlorination stage. For the purposes of this presentation, a survey of the immediate chlorine demand of treated waters from 13 WTWs was undertaken. The WTWs were selected to provide a broadly representative sample of Irish water supplies, both in terms of source type (all were surface water sources) and treatment technology. The survey was limited to one-off grab samples, taken during August 2012. Thus, while the resulting data set presents a useful overall estimate of the potential range and profile of chlorine demand of WTW-treated surface water, it provides no information on the temporal or seasonal variation in chlorine demand that would be experienced at any of the WTWs sampled. Two chlorine decay tests were carried out on each water sample, one at an applied dose of ca. 2 mg/l and a second test at a lower applied dose of 1.5 mg/l. Test sample volumes of 5 litre or 10 litre were used. The test vessels were incubated in an insulated water bath at about 15 ^oC. The free chlorine was measured by the DPD colourimetric method using Hach instrumentation. The RFC was generally measured at 15 minute intervals for the first hour of contact and at 30 minute intervals for the second hour of contact. The measured chlorine demand profile of a representative set of the tested waters is illustrated graphically in Fig 1a through 1f, inclusive. Figs 1a and 1b relate to low chlorine demand waters; Figs 1c and 1d relate to intermediate demand waters, while Figs 1e and 1f relate to high chlorine demand waters. The 15-min and 120-min measured chlorine demands for all samples tested are set out in Table 2, which also includes the water pH and test temperature. [caption id="attachment_2945" align="aligncenter" width="423"] Fig 1a: Chlorine demand of WTW BE treated water[/caption]   [caption id="attachment_2947" align="aligncenter" width="409"] Fig 1b: Chlorine demand of WTW DLM treated water[/caption]   [caption id="attachment_2955" align="aligncenter" width="411"] Fig 1c : Chlorine demand of WTW DMD treated water[/caption]   [caption id="attachment_2963" align="aligncenter" width="404"] Fig 1d:Chlorine demand of WTW GY treated water[/caption] [caption id="attachment_2959" align="aligncenter" width="408"] Fig 1e: Chlorine demand of WTW RD treated water[/caption] [caption id="attachment_2960" align="aligncenter" width="463"] Fig 1f: Chlorine demand of WTW SN treated water[/caption]

Table 2: Primary chlorine demand test data summary

C: indicates conventional treatment i.e. chemical coagulation/clarification/filtration. SSF: indicates slow sand filtration. N: indicates no pre-treatment The chlorine demand plots exhibit a common free chlorine reactivity profile in which there is an initial very high reaction rate that very rapidly decreases with contact time. The sample set covers a wide range of chlorine demand as reflected in the fact that the largest 120-minute demand measured is approximately three times the smallest measured value. In the case of high chlorine demand waters up to one third of the 120-minute chlorine demand is exerted in the first minute of contact time, while between 50% and 70% of the 120-minute chlorine demand is exerted in the first 15 minutes of contact time. Analysis of the measured chlorine demand data has shown that the demand time profile can be accurately represented by a power-type correlation of the following form:  D_t = k_Dtⁿ (1), where D_t is the chlorine demand (mg/l) at time t (min), k_D and n are empirical coefficients. The coefficient k_D corresponds to the fitted model 1-minute chlorine demand value. In all cases the chlorine demand was measured at two chlorine dose rates, the larger dose rate (ca. 2 mg/l) exceeding the lower dose rate (ca. 1.5 mg/l) by about 33%. However, while the 120-minute chlorine demand at the higher chlorine dose equalled or exceeded that for the lower dose in all samples tested, the increase in all cases was considerably less than 33%, indicating that the free chlorine reaction rate was not a first-order reaction in respect of chlorine. In the case of the low and medium demand waters, the increase in demand at the higher dose rate was found to be marginal. The slow-sand filtered waters exhibited the largest dose-related differential demand at about 20%. The treated water chlorine demand profiles reflected in the foregoing plots are not unique to Irish waters. The USEPA (1992) proposed division of the chlorine reaction process into three sequential time periods (1) zero-order reaction for contact time t = 0 to 5 min, second order for 5 min to 5 h, and (3) first-order for t > 5 h. A variety of mathematical models of the chlorine decay process are to be found in the research literature (Bocelli et al., 2003; Powell et al., 2000; Warton et al., 2000). However, their application as a predictive tool, under primary chlorination conditions, is problematical due to the catchment-specific and heterogeneous nature of the water constituents that consume chlorine. As the measured chlorine decay data presented in Figs 1(a) – 1(f) indicate, the reaction rate is partly chlorine concentration-dependent for some waters but not others. It would appear that the primary decay rate is very significantly influenced by the organic content of the water, which is conventionally measured as undifferentiated TOC. A correlation of chlorine demand and TOC for a range of Irish treated surface waters is presented in Fig 2. [caption id="attachment_2929" align="aligncenter" width="538"] Fig 2: Chlorine demand/TOC relationship[/caption]

While there is the expected positive correlation between chlorine demand and TOC, the dependency varies from one catchment to another. This is not an unexpected finding considering that TOC is an umbrella parameter, embracing a range of unidentified organic constituents of natural origin and reflecting the organic profile of the source catchment. Thus, there is no universal model that can be applied to quantify the primary chlorine decay rate of treated surface waters. Hence, the importance of its experimental measurement to disinfection process design.

CHLORINE DISINFECTION PROCESS DESIGN The objective in drinking water chlorine disinfection process design is to ensure that the entire outflow from the chlorine contact tank has been in contact with free chlorine at a concentration and for a sufficient time period to ensure compliance with the selected design CT-value. The CT-value is calculated as the product of RFC and contact time, integrated over the period of contact, taking into account the decreasing RFC concentration with time. This product is illustrated graphically in Fig 3 as the hatched area bounded by the chlorine dose line and the chlorine demand curve. The CT value is calculated by integration as follows: CT = ∫₀^tc[C₀ – k_Dtⁿ]dt = C₀t_c – (k_Dt_c(1+n))/((1+n))    (2). For a given contact time t_c, the CT value may be usefully considered as being made up of two parts (a) the product of the outflow RFC by the contact time (upper hatched are in Fig 3), and (b) the contribution of the reacted free chlorine at contact time t_c (lower hatched are in Fig 3). The latter, designated CT^R for discussion purposes, is given by: CT^R  = k_Dt_c¹⁺ⁿ (1 – 1/(1+n))   (3).  [caption id="attachment_2935" align="aligncenter" width="530"] Fig 3: CT computation using chlorine demand curve[/caption]

The selection of a process CT value is a primary design consideration. The original WHO recommendation of an RFC of 0.5 mg/l after 30 minute contact has long been used as an authoritative design guideline. It is therefore of interest to examine the CT-value range it would provide for the set of treated surface waters in the survey outlined above. Inserting the calculated chlorine demand coefficients k_D and n for a chlorine dose of 2 mg/l in the relevant equations, as outlined above, produced the following value ranges:

Chlorine dose range required to generate a 30-min. RFC of 0.5 mg/l: 0.9 – 1.8 mg/l Corresponding CT-value range (30 x 0.5 + CT^R): 18 – 23 mg.min/l Thus, the calculated required chlorine dose for the water with the highest chlorine demand was twice that required for the water with the lowest chlorine demand. However, the CT-value delivered by the higher dose exceeded that delivered by the lower dose by only 28%. Thus, at a short contact time of 30 minutes, the CT^R contribution of the higher chlorine dose is diminished by the initial high chlorine demand rate. The required chlorine residual (C_t) and corresponding chlorine dose (C₀) required to deliver the selected process design CT value are calculated as follows:

C_t = CT/t – k_Dt (1 – 1/(1+n))     (4)

C₀ = Ct + k_Dtⁿ      (5)

The more recent (2011) WHO recommendations, as set out in Table 1, suggest a CT value of 35 mg.min/l for a 99% virus inactivation under adverse pH and temperature conditions. Adopting a design CT value of 35 mg.min/l and a 60 minute contact time, the calculated required chlorine dose and corresponding effluent RFC values for the tested treated waters are plotted in Fig 4. The required dose varied from 0.94 mg/l for the water with the lowest chlorine demand to 1.85 mg/l for the water with the highest chlorine demand. The outflow RFC, which is the more important parameter from a process control viewpoint, varied from 0.28 mg/l for the highest chlorine demand water to 0.46 mg/l for the lowest chlorine demand water. It is also noteworthy that the CT^R fraction of the total CT varied from 20% in the lowest chlorine demand water to 52% in the highest chlorine demand water. [caption id="attachment_2932" align="aligncenter" width="532"] Fig 4: Chlorine dose and outflow RFC required for CT of 35 mg.min/l[/caption] CONTACT TANK DESIGN  The ideal contact tank has plug-flow characteristics i.e. all elements of the flowing water mass have the same residence time, providing equal exposure to chlorine and a common CT, which can be calculated as indicated in the previous section. In practice, ideal plug-flow conditions are not achievable and instead of a common residence time the flow-through characteristics are quantified by a residence time distribution about the nominal or average residence time as defined by V/Q, where V is the tank volume and Q is the flow rate. The key design objective is the contraction of the residence time spread by the elimination of short-circuiting. The extent of short-circuiting is mainly influenced by the inlet and outlet arrangements and the internal geometry of the contact tank. Selection of retention time is an important process design decision as it determines the tank volume. While it would be feasible to achieve the WHO disinfection criterion of an RFC of 0.5 mg/l at a retention time of 0.5h, a somewhat longer retention time, say 1-2h, provides a more stable design basis. The longer retention time ensures that the immediate rapid chlorine demand is satisfied within the contact tank and ensures that the RFC in the tank outflow will not be dissipated very rapidly. In practice, the mixing characteristics of contact tanks fall between being fully mixed and plug flow, resulting in a residence time distribution about the average value. This can be taken into account in contact tank sizing by applying a correction factor to the mean residence time. The residence time distribution can be determined by tracer testing (EPA Manual, 2011). US EPA contact tank design guidance (USEPA, 1999) for disinfection proposes a residence time correction factor based on the t₁₀ value which is defined as the time for 10% of the injected tracer to be discharged from a contact tank. The recommended contact time correction factor F_m is taken as the ratio of the t₁₀ value to the nominal average residence time RT = V/Q. Guidance design F_m values are given in Table 2. In some water supply developments treated water is pumped from the WTW to a remote elevated service reservoir. Where the rising main has an adequate residence time, it has the potential to provide the ideal contact tank. However, the use of a rising main in this way is only feasible where there are no branch supplies off the rising main or local supplies in the environs of the WTW. The contact tank features that are designed to reduce short-circuiting include:

• Flow path length large relative to flow cross-sectional dimensions, further enhanced by internal baffle wall arrangement.
• Inflow kinetic energy dissipated over full water depth by use of a vertical standpipe with multiple orifices
• Perforated baffles used to generate flow over full cross-section of tank
• Outflow collected over water depth by vertical standpipe with multiple orifices and discharged over a fixed weir that defines the contact water surface elevation and hence defines the contact tank volume.

Table 2: Guidance values for t₁₀/RT for indicated baffling (USEPA, 1999; EPA Manual, 2011)

*Refer EPA manual, 2011 for additional details

INFLUENCE OF TEMPERATURE ON CHLORINE DEMAND As the data in Table 1 indicate, the CT requirement for disinfection is strongly influenced by temperature, as reflected in the fact that the recommended 99% virus inactivation value at 5 ^oC is approximately 1.5 times that at 10 ^oC. As a consequence, the influence of temperature on primary chlorine demand is an important consideration in disinfection process design, since for a given chlorine dose and contact time, the demand rate affects the resultant CT value. A frequently used approximate rule, enunciated by van’t Hoff (Fair et al., 1968), is that aqueous reaction rates typically double for a rise in temperature of 10^oC, corresponding to about 8.4% per degree rise in temperature. Research findings (Powell et al., 2000) have shown a 1.8- to 3.2-fold increase in chlorine decay rate in distribution pipework for a temperature rise from 10 ^oC to 20 ^oC. However, in a limited series of chlorine decay tests carried out as part of this study, a temperature rise of 10 ^oC (from 5 to15 ^oC) was found to result in only about a 10% rise in the 120-minute primary chlorine demand value. The significant difference between these findings may indicate that primary chlorine demand is less influenced by temperature than is the ongoing subsequent distribution system chlorine demand. However, this indication should be regarded as tentative and would need to be verified by further testing. Thus, while it is clearly the case that a given chlorine dose would deliver a greater CT value at 5 ^oC than it would at 15 ^oC, the difference is most reliably obtained by chlorine demand measurements at both temperatures. There is anecdotal evidence that public water supplies using surface water sources experience a seasonal variation in chlorine usage, with increased demand in the summer and early autumn periods. While this increase is very probably temperature-related it may also be caused by a seasonal NOM variation both in composition and concentration. ONGOING CHLORINE DEMAND The primary chlorine demand data reported above is limited to a contact duration of 120 minutes. There is obviously a continued growth in chlorine demand beyond this terminal point. While the empirical model that has been used to quantify the primary chlorine demand is not necessarily valid outside the contact time period for which it has been verified, the slope of the demand graph at its terminal point measures the ongoing rate of increase in chlorine demand at this point. Analysis of the graph slopes shows the rate of increase varied within the range 0.06 - 0.24 mg/l per hour, the lower end of the range relating to the low chlorine demand waters and the upper end of the range relating to the high chlorine demand waters. This points to the fact that waters with a high primary chlorine demand are also likely to have a much greater ongoing chlorine demand than waters with a low primary chlorine demand. ACKNOWLEDGEMENTS The authors wish to acknowledge the support of water analysis facilities provided by the UCD School of Civil, Structural and Environmental Engineering and Donegal County Council in carrying out the water analyses reported in the paper. The collaboration of the various Local Authorities that provided test water samples is also gratefully acknowledged.   REFERENCES Bellar et al. (1974)  The Occurrence of Organohalides In Finished Drinking Waters, J. AWWA, 66:12:703. Boccelli, D., Tryby, M. E., Uber, J. G., and Summers, R. S. (2003):  A reactive species model for chlorine decay and THM formation under rechlorination conditions, Water Research, 37, 2654-2666. Brezonik, P. L. (1994) Chemical Kinetics and Process Dynamics in Aquatic Systems, CRC Press, Boca Raton. Chowdhury, S. and Champagne, P. (2008):  An investigation on parameters for modelling THM formation, Global NEST Jour., Vol. 10, No. 1, pp80-91. Chua, K. H. (1996):  THM Formation in Drinking Water, Ph.D Thesis, Department of Civil Engineering, UCD. Clark, R. M. (1998):  Chlorine demand and TTHM kinetics:  a second order model. J. Environ Eng ASCE, 124(1), pp16-24. Collins, M. R. and Eighmy, T.T. (1988):  Modification to the slow sand filtration process for improved trihalomethane removal; Sect. 4.4, pp281-304 in Slow Sand Filtration: Recent developments in water treatment technology, Ed. N.J.D. Graham, Ellis Horwood. Engerholm, B. A. and AMY G. L. (1983):  A predictive model for chloroform formation from humic acid, JAWWA, 75(8), 418-423. EPA(Irl) (2011)  Water Treatment Manual: Disinfection, Environmental Protection Agency, Johnstown Castle, Co. Wexford. EPA(Irl) (2011)  The Provision and Quality of Drinking Water in Ireland, A Report for the Year 2010, Environmental Protection Agency, Johnstown Castle, Co. Wexford. Fair, G. M., Geyer, J. C. and Okun, D. A. (1968)  Water and Wastewater Engineering, Vol. 2, John Wiley & Sons, Inc. Hrudey, S. E., Charrois, J. W. A. (2012):  Disinfection By-products and Human Health, IWA Publishing, London. Hutton, P. L. and Chung, F. I. (1994):  Bromide distribution factors in THM formation. Jour. Water Resource, Planning and Management ASCE, Vol. 120, No. 1. Kim, J., Chung, Y., Shin., D., Kim, M., Lee, Y., Lim, Y. and Lee, D. (2002):  Chlorination by-products in surface water treatment process, Desalination, 151, 1-9. Oliver, B. G. and Lawrence, S. (1979):   Haloforms in drinking water: A study of precursors and precursor removal, JAWWA, 71(3),161-163. Powell, J. C., Hallam, N. B., West, J. R., Forster, C. F. and Simms, J. (2000): Factors which control bulk chlorine decay rates, Wat. Res. Vol. 34, No. 1 pp. 117-126. Rook, J. J. (1976)  Haloform in Drinking Water, J. AWWA 68:3:168 Snoeyink, V. L. and Jenkins, D. (1980)  Water Chemistry, John Wiley & Sons, Inc. USEPA (1992) Water Treatment Simulation Program User’s Manual Version 1.21, EPA No. 811B92001. USEPA (1999)  Enhanced Coagulation and Enhanced Precipitative Softening Guidance Manual, EPA 815-R-99-012 USEPA (1999)  EPA Guidance Manual, Alternative Disinfectants and Oxidants, EPA 815-R-99-0.14. Warton, B., Heitz, A., Joll, C. and Kagi, R. (2006):  A new method for calculation of the chlorine demand of natural and treated waters, Water Research, 40, 2877-2884. WHO (1992)  Guidelines for Drinking Water Quality: Vol. 2: Health Criteria and Other Supporting Information, Geneva. WHO (2011)  Guidelines for Drinking Water Quality, 4^th. Edition, Annex 1: Water treatment and pathogen control: Process efficiency in achieving safe drinking water, Geneva.