Rapid monochloramine decay has been observed in the product water of three River Murray water treatment plants (WTPs). Previous investigations identified that rapid monochloramine decay was microbiological in nature and observed in samples taken after media filtration but was absent in filtered water samples from a fourth WTP of similar design. The filters at the WTP not exhibiting rapid decay are backwashed with filtered non-disinfected water whereas the other WTPs backwash with treated chloraminated water. It was therefore hypothesised that backwashing filters with chloraminated water was the cause of the rapid monochloramine decay. A pilot-scale study was conducted to investigate the impact of backwashing with chloraminated water on the occurrence of microbiologically accelerated monochloramine decay. Additional samples were analysed to assess the impact of chloraminated backwash water on N-Nitrosodimethylamine (NDMA) formation and biological degradation of taste and odour compounds 2-methyl isoborneol (MIB) and geosmin in the filter media. Backwashing with chloraminated filtered water was concluded to be the cause of the observed rapid monochloramine decay, with rapid decay observed within 8 weeks for the filters backwashing with chloramines. Additionally, backwashing with chloraminated filtered water was observed to increase NDMA formation and impair the biological degradation performance of MIB and geosmin.

INTRODUCTION

The River Murray is relied upon as the primary source of drinking water for the majority of consumers in regional South Australia. Above-ground transportation, often over several hundred kilometres, means high water temperatures and long residence times are commonly experienced. Chloramination is favoured over chlorination for these systems to better comply with Australian Drinking Water Guidelines (NHMRC & NRMMC 2011) for disinfection by-products, however loss of chloramine residual due to nitrification and microbiologically accelerated decay is also possible.

Monochloramine decay is the result of chemical and biological processes (Sathasivan et al. 2005). Chemical factors impacting the rate of monochloramine decay include pH, temperature, dissolved organic carbon (DOC), nitrite and chlorine to ammonia ratios (Sathasivan et al. 2005). Biological factors impacting the rate of decay include the presence of ammonia oxidising archaea, ammonia oxidising bacteria, nitrite oxidising bacteria and biological activity in general (Wolfe et al. 1988; Cunliffe 1991).

Rapid monochloramine decay has been observed in the distribution systems of the Swan Reach, Summit and Tailem Bend water treatment plants (WTPs) in South Australia (Cook et al. 2014). All three WTPs source water from the River Murray and are of conventional design (coagulation/flocculation/sedimentation/media filtration) with ultraviolet (UV) irradiation followed by chloramination (ammonia then chlorine) for disinfection. Interestingly, rapid monochloramine decay was not observed at the Morgan WTP, which also sources water from the River Murray and is of similar conventional design but does not practice UV disinfection and applies chlorine before ammonia for chloramination. A key operational difference between the four WTPs is that the filters at Morgan are backwashed with disinfectant-free water to enhance the biological degradation of taste and odour compounds instead of the chloraminated water used at the other WTPs.

An investigation by Cook et al. (2014) identified that the rapid decay was microbiological in nature using the method described by Sathasivan et al. (2005). Further investigations identified greater monochloramine decay rates in filtered water than settled water for all WTPs except Morgan. Laboratory testing at the time revealed that the order of ammonia addition had no significant impact on the decay rates unless a free chlorine Ct of 42 mg min/L or greater was maintained (Cook et al. 2014). It was subsequently hypothesised that backwashing the filters with chloraminated product water was the cause of the observed rapid chloramine decay (Cook et al. 2014). Similar observations have been made in the literature where rapid monochloramine decay was observed in the effluent of a granular activated carbon (GAC)/sand filter backwashed with chloramines (Wilczak et al. 2003) and nitrification was observed following the installation of GAC filters exposed to free ammonia at Ann Arbor (Skadsen 1993).

Four pilot filters were constructed to investigate the impact of backwashing media filters with chloraminated water on the occurrence of rapid monochloramine decay in filtered water. Additional testing was performed to determine if backwashing filters with chloraminated water impacted N-Nitrosodimethylamine (NDMA) formation in filtered water and the ability of the filters to biologically degrade the taste and odour compounds MIB and geosmin.

METHODS

Four pilot filter columns (9.6 cm diameter) were erected at the Tailem Bend WTP (Figure 1). Filter media was sourced from the filters at the Tailem Bend (sand: 0.45–0.55 effective size (ES), uniformity coefficient (UC) <1.45; anthracite: 1.1–1.2 ES, UC < 1.45) and Morgan (sand: 0.45–0.55 ES, UC < 1.5; anthracite: 1.0–1.1 ES, UC < 1.5) WTPs with layer depths of 300 mm for sand and 750 mm for anthracite added to each pilot filter (Table 1). Settled water from the Tailem Bend WTP was continuously fed to the filters via a common header tank with a distribution weir. Limited hydraulic head was available from the header tank resulting in a maximum filtration rate of 3 m/h for clean media. Effluent from filters 1 and 2 was stored for use as the disinfectant-free backwash water (Table 1). Chloraminated product water from the WTP was sourced as the backwash water for filters 3 and 4. Backwashing was performed on a weekly to fortnightly basis. Samples for monochloramine decay tests and water quality analysis (turbidity, UV254 absorbance, true colour, conductivity and DOC) were collected at the end of the filter cycles immediately prior to backwashing.
Table 1

Pilot filter column media and backwashing arrangement

Filter column1234
Filter media Morgan WTP Tailem Bend WTP Morgan WTP Tailem Bend WTP 
Backwash source Filtered water Filtered water Filtered chloraminated water Filtered chloraminated water 
Filter column1234
Filter media Morgan WTP Tailem Bend WTP Morgan WTP Tailem Bend WTP 
Backwash source Filtered water Filtered water Filtered chloraminated water Filtered chloraminated water 
Figure 1

Pilot filter schematic.

Figure 1

Pilot filter schematic.

Samples were chloraminated, ammonia addition followed by chlorine, to achieve a monochloramine concentration of 4.5 mg/L (4.5:1 Cl2:NH3 ratio) at room temperature (22 ± 2 °C) and pH of 8.4 (±0.1). Free chlorine stock water (2,000 to 5,000 mg/L as Cl2), prepared by the addition of gaseous chlorine to ultrapure water, and analytical-grade liquid ammonia (920 mg/L as N) were dosed to achieve the target concentration. Free chlorine, monochloramine and total chloramine concentrations were measured by the N,N-diethyl-p-phenylenediamine (DPD)–ferrous ammonium sulphate titrimetric method procedure described in Standard Methods for the Examination of Water and Wastewater (1998). Monochloramine decay was monitored in 0.2 μm membrane filtered (to partition bacteria) and unfiltered samples to discriminate between the microbiological and chemical components of decay by calculation of a microbiological decay factor (Fm) as outlined by Sathasivan et al. (2005). MIB and geosmin concentrations were measured by solid phase microextraction and gas chromatography–mass spectrometry (GC/MS) using the method outlined by Graham & Hayes (1998). NDMA samples were analysed at the Australian Water Quality Centre using GC/MS/MS.

True colour, UV254 absorbance, pH, DOC and conductivity were monitored throughout the trial to identify changes in water quality. True colour and UV254 absorbance were measured by using a Thermoscientific Evolution 60 spectrophotometer. Turbidity was measured by a HACH 2100AN turbidimeter and DOC by a Sievers 900 laboratory total organic carbon analyser (Standard Methods 1998). Samples for UV254 absorbance and DOC were prepared for analysis by filtering through pre-rinsed 0.45 μm mixed cellulose ester (MCE) membrane filters (Standard Methods 1998) and through 0.2 μm MCE membranes for true colour. Conductivity and pH were measured by WTW portable probes.

RESULTS AND DISCUSSION

The pilot filters commenced operation in December 2014 and the first chloramine decay test was performed 8 weeks later in February 2015 (Figure 2(a) and 2(b)). No significant component of microbial decay was observed for filters 1 and 2 (Figure 2(a)). Rapid monochloramine decay was observed in the outlet waters from filters 3 and 4 (chloraminated backwash) (Figure 2(b)) with microbial decay (Fm = 1.0 and 1.8, respectively) impacting on monochloramine demand. These results are similar to those observed by Cook et al. (2014). The onset of microbiologically accelerated decay in filter 3, which sourced media from Morgan WTP where rapid decay is not observed, indicates that the filter media may be conditioned to promote microbiological decay within 8 weeks given appropriate conditions.
Figure 2

Monochloramine decays for pilot filter effluent (including 0.2 μm membrane filtered samples) after 8 weeks of operation: (a) filters 1 and 2, and (b) filters 3 and 4.

Figure 2

Monochloramine decays for pilot filter effluent (including 0.2 μm membrane filtered samples) after 8 weeks of operation: (a) filters 1 and 2, and (b) filters 3 and 4.

Effluent from pilot filters 1 and 2 both maintained mean 7-day monochloramine demands of 0.89 mg/L (standard deviations (SD) of 0.23 mg/L and 0.16 mg/L, respectively, n = 4) over the pilot trial compared to 2.13 mg/L (SD = 0.18 mg/L, n = 4) and 2.73 mg/L (SD = 0.80 mg/L, n = 4) for filters 3 and 4, respectively. The absence of rapid monochloramine decay in filtered waters from filters 1 and 2 for the entire duration of the trial is consistent with the hypothesis that backwashing with chloraminated water is the cause of rapid monochloramine decay (Figure 3). Variations in monochloramine demand between filters 3 and 4 indicate that other unmeasured factors may also influence the severity of monochloramine demand. Differences in DOC, pH, conductivity, UV254 absorbance and true colour between the filtered waters were not significant and no apparent trends in these parameters were observed during the investigation (Table 2).
Table 2

Summary of water quality data for each filter (±1 SD, n = 7)

 pHConductivity (μS/cm)DOC (mg/L)UV254 Abs. (/cm)True colour (HU)
Feed water 7.1 ± 0.3 337 ± 63 3.2 ± 0.8 0.047 ± 0.01 3 ± 1 
Filter 1 product 7.4 ± 0.2 351 ± 70 3.1 ± 1.0 0.043 ± 0.01 3 ± 1 
Filter 2 product 7.6 ± 0.3 354 ± 69 3.3 ± 0.9 0.044 ± 0.01 4 ± 1 
Filter 3 product 7.5 ± 0.3 353 ± 68 3.4 ± 1.0 0.046 ± 0.02 3 ± 2 
Filter 4 product 7.5 ± 0.3 354 ± 64 3.2 ± 0.8 0.048 ± 0.01 3 ± 1 
 pHConductivity (μS/cm)DOC (mg/L)UV254 Abs. (/cm)True colour (HU)
Feed water 7.1 ± 0.3 337 ± 63 3.2 ± 0.8 0.047 ± 0.01 3 ± 1 
Filter 1 product 7.4 ± 0.2 351 ± 70 3.1 ± 1.0 0.043 ± 0.01 3 ± 1 
Filter 2 product 7.6 ± 0.3 354 ± 69 3.3 ± 0.9 0.044 ± 0.01 4 ± 1 
Filter 3 product 7.5 ± 0.3 353 ± 68 3.4 ± 1.0 0.046 ± 0.02 3 ± 2 
Filter 4 product 7.5 ± 0.3 354 ± 64 3.2 ± 0.8 0.048 ± 0.01 3 ± 1 
Figure 3

Monochloramine decays for pilot plant filters.

Figure 3

Monochloramine decays for pilot plant filters.

The frequency of backwashing of filters was identified as a potential operational factor that may impact the occurrence of accelerated chloramine decay (Cook et al. 2014). Monochloramine demand in the outlet waters of each filter was monitored over the course of a filter cycle (at 1 hr, 3 days and 8 days) to identify the short-term impacts of backwashing with chloraminated water on monochloramine residual stability (Figure 4). The microbial component of monochloramine decay in filter 4 was greatest immediately following the backwash (Fm = 4.27) and decreased as the filter cycle progressed (Fm = 1.02 and 0.80 at 3 days and 8 days, respectively) (Figure 4(b)). A decrease in monochloramine stability in the 0.2 μm membrane filtered sample in the filter 4 effluent immediately following the backwash may indicate the possible presence of soluble microbial products (Bal Krishna et al. 2012). Monochloramine stability in the filter 2 product was not impacted by the position of the filter cycle (Figure 4(a)).
Figure 4

Impact of filter cycle position on monochloramine decay for (a) non-chloraminated and (b) chloraminated backwash waters.

Figure 4

Impact of filter cycle position on monochloramine decay for (a) non-chloraminated and (b) chloraminated backwash waters.

The presence of chloramines in the backwash water is expected to impact the microbiological community within the filter media. It is hypothesised that the addition of chloramines to the filter provides a competitive advantage to the microorganisms responsible for the rapid monochloramine decay, by the disinfection of competing microorganisms, with the advantage diminishing over the course of the filter cycle. The observed decrease in monochloramine stability immediately following backwashing with chloramines suggests that shorter filter cycle times resulting from greater backwashing frequency, such as those observed during the high DOC and turbidity challenge event in 2010/2011 (Cook et al. 2014), contribute to the occurrence of microbiologically accelerated monochloramine decay events.

Previous investigation of factors impacting NDMA formation identified increased formation potential following media filtration at WTPs that backwash with chloraminated water, however the cause of the increased formation was not established (Morran et al. 2009). A one-off test was performed using the pilot filters to identify if backwashing with chloraminated water impacted NDMA formation. Filtered water samples were collected from each pilot filter at the end of a 7-day filter cycle and chloraminated as per previous decay tests. Samples were analysed for NDMA 7 days after chloramination. NDMA concentrations were greater in outlet waters from filters 3 and 4 (65 and 40 ng/L, respectively) than filters 1 and 2 (20 and 26 ng/L, respectively) and were consistent with the findings of Morran et al. (2009) indicating that the presence of chloramines in backwash water increases NDMA formation. More detailed investigations are required to better understand this relationship and the mechanism for the increased formation.

The media filters at the Morgan WTP plant are biologically active and capable of degrading MIB and geosmin, often to below the reporting limit of 4 ng/L (McDowall 2008), leading to significant savings in the dosing of PAC during taste and odour challenge events. The Tailem Bend WTP, despite being of similar design and sourcing water from the same river system, is entirely reliant on PAC for control of taste and odour compounds. The impact of chloraminated backwashes on the capability of filters to degrade MIB and geosmin was assessed by continuously spiking the filters with MIB and geosmin over 25 hours. The ability of the Morgan WTP media to degrade MIB and geosmin was retained, and potentially enhanced due to the high empty bed contact time of the pilot plant (McDowall 2008), when backwashed with undisinfected water (filter 1) but was severely impacted when backwashed with chloraminated water (filter 3) with the MIB degradation capacity lost completely (Figure 5). The ability of the Tailem Bend WTP media to degrade MIB and geosmin improved when the filters were no longer exposed to chloramines; however, the degree of removal remained limited and was below reported thresholds for physical removal of 30% and 50% for MIB and geosmin, respectively (Elhadi et al. 2004). This result demonstrates that whilst the presence of residual disinfectant in backwash water may inhibit biological degradation of taste and odour compounds, its absence alone is not sufficient to develop this capability.
Figure 5

Mean removal of MIB and geosmin after 22–25 hours of exposure to MIB and geosmin.

Figure 5

Mean removal of MIB and geosmin after 22–25 hours of exposure to MIB and geosmin.

CONCLUSIONS

The backwashing of media filters with chloraminated water was found to be an essential contributor to the occurrence of rapid monochloramine decay. Additionally, the severity of the monochloramine demand decreased with time since backwashing. Further testing found that backwashing with chloraminated water can lead to elevated NDMA formation in filter effluent and significantly impairs the capability of the filters to biologically degrade MIB and geosmin.

ACKNOWLEDGEMENTS

The authors thank Miriam Nedic for her assistance in completion of laboratory decay tests and Corey Legg for assisting in the operation and maintenance of the filtration pilot plant.

REFERENCES

REFERENCES
Cook
D.
Morran
J.
Mobius
W.
Drikas
M.
2014
Impact of water quality change in the River Murray on monochloramine decay
.
Water: Journal of the Australian Water Association
41
(
2
),
110
116
.
Cunliffe
D. A.
1991
Bacterial nitrification in chloraminated water supplies
.
Applied and Environmental Microbiology
57
,
3399
3402
.
Elhadi
S. L. N.
Huck
P. M.
Slawson
R. M.
2004
Determination of system losses of geosmin and MIB in bench-scale filtration apparatus
.
Water Quality Research Journal Canada
2
(
39
),
207
212
.
Graham
D.
Hayes
K. P.
1998
Application of solid phase microextraction for analysis of off-tasting flavours in water
. In:
Proceedings of the WaterTECH Conference
,
Brisbane, Australia
.
McDowall
B.
2008
Removal of Geosmin and 2-methylisoborneol from Drinking Water through Biologically Active Sand Filters
.
PhD Thesis
,
School of Chemical Engineering, the University of Adelaide
,
Adelaide
,
Australia
.
Morran
J.
Slyman
N.
Newcombe
G.
2009
Factors affecting NDMA formation in treated drinking water
.
Water
36
(
3
),
57
61
.
NHMRC & NRMMC
2011
Australian Drinking Water Guidelines Paper 6 National Water Quality Management Strategy
.
National Health and Medical Research Council/National Resource Management Ministerial Council/Commonwealth of Australia
,
Canberra
,
Australia
.
Sathasivan
A.
Fisher
I.
Kastl
G.
2005
Simple method for quantifying microbiologically assisted monochloramine decay in drinking water
.
Environmental Science and Technology
39
(
4
),
5407
5413
.
Skadsen
J.
1993
Nitrification in a distribution system
.
Journal of American Water Works Association
85
(
7
),
95
103
.
Standard Methods for the Examination of Water and Wastewater
1998
20th edn.
American Public Health Association/American Water Works Association/Water Environment Federation
,
Washington, DC
,
USA
.
Wilczak
A.
Hoover
L. L.
Lai
H.
2003
Effects of treatment changes on chloramine demand and decay
.
Journal of American Water Works Association
95
(
7
),
94
106
.
Wolfe
R. L.
Means
E. G.
Davis
M. K.
Barrett
S. E.
1988
Biological nitrification in covered reservoirs containing chloraminated water
.
Journal of American Water Works Association
80
(
9
),
109
114
.