Recent trends in off-flavor problems were investigated by a nationwide questionnaire survey targeting 239 drinking water treatment plants (DWTPs) maintained by 79 regional water suppliers (RWS) in Japan. The results indicated that 43 of 79 (54%) RWS and 89 of 239 (36%) DWTPs were affected by off-flavor problems in a recent 2-year period (October 2010–September 2012). Off-flavor incidents occurred in all regions in Japan. Most DWTPs affected by off-flavor incidents take in water from closed water areas, such as dams and lakes, where phototrophic microorganisms, such as cyanobacteria and algae, can grow easily. The most common off-flavor was an earthy–musty odor, while a fishy smell was also reported in some cases. The common producers of geosmin were Anabaena spp., while those responsible for 2-methylisoborneol were Phormidium spp. and Oscillatoria spp. The off-flavor events frequently occurred during summer and early fall months (June–September). The water temperature and pH of raw water for the production of drinking water when off-flavor problems occurred were higher than those during normal periods. To eliminate off-flavor compounds, most DWTPs (about 58%) adopt powdered activated carbon treatment, which considerably increases operational costs.

INTRODUCTION

The production of off-flavor compounds by organisms in water is a problem for water utilities worldwide (Watson 2003; Giglio et al. 2011). In Japan, off-flavors are a serious problem in water supply systems because surface water is commonly used as raw water for drinking water (Kishida et al. 2013). Photosynthetic microorganisms, such as algae and cyanobacteria, which produce off-flavor compounds, often occur in water sources (Matsumoto & Tsuchiya 1988; Yano et al. 1988; Oikawa & Ishibashi 2004). In 1990, more than 20 million people, accounting for about 18% of the total Japanese population, were affected by off-flavors in drinking water in Japan. Although the number decreased markedly with the adoption of advanced treatment processes, more than two million people still experience off-flavors in drinking water in Japan, and the number has increased slightly in recent years (Ministry of Health Labour and Welfare Japan 2013). The Ministry of Health, Labour and Welfare, Japan, decided to include geosmin and 2-methylisoborneol (2-MIB) in the drinking water quality standards, which are the most important criteria, in 2003. Current standard values of geosmin and 2-MIB are 10 ng/L. The threshold odor number (TON) has also been adopted as a management goal, which is a second-class criterion of the Japanese water quality management system (Kishida et al. 2013). Even when the concentration of off-flavor compounds in finished water exceeds the standard values, water supply is normally continued. However, drinking water that does not contain off-flavor compounds is sometimes delivered from other drinking water treatment plants (DWTPs) to consumers by water tank trucks. In rare cases, water charges are reduced during the periods of concentration excess.

To investigate recent trends in organism-related off-flavor problems, we performed a nationwide questionnaire survey targeting 142 DWTPs maintained by 24 regional water suppliers (RWS) in Japan (Kishida et al. 2013). The results indicated that off-flavor incidents had occurred in all regions in Japan. Cyanobacteria were the most common organisms responsible for producing off-flavor compounds. However, the numbers of target RWS and question items were limited, and detailed information on off-flavor incidents was not available. Hence, in the present study, we performed a detailed questionnaire survey about organism-related off-flavor problems targeting a wider range of water utilities in Japan.

MATERIALS AND METHODS

The questionnaire survey was conducted in 2012. The target period of the questionnaire ranged from October 2010 to September 2012. A total of 239 DWTPs maintained by 79 RWS from all regions of Japan were targeted for the questionnaire survey. As shown in Table 1, at least 25 DWTPs were targeted from each region. The DWTPs with very small and large water treatment capacities were included, as shown in Figure 1. Total amounts of distributed water of target DWTPs accounted for about 41% of the total amount of distributed water in Japan. Most target DWTPs take in surface water.

Table 1

Summary of target DWTPs for questionnaire survey

Region Japanese region Area (km2Population (in thousands) No. of target RWS No. of target DWTPs 
North Hokkaido–Tohoku 150,347 14,842 14 45 
East Kanto 32,424 42,607 13 40 
Central Chubu 66,795 21,715 25 
West Kinki 33,112 22,755 17 48 
Southwest Chugoku–Shikoku 50,720 11,539 14 42 
South Kyushu–Okinawa 44,467 14,597 12 39 
Total  377,865 128,055 79 239 
Region Japanese region Area (km2Population (in thousands) No. of target RWS No. of target DWTPs 
North Hokkaido–Tohoku 150,347 14,842 14 45 
East Kanto 32,424 42,607 13 40 
Central Chubu 66,795 21,715 25 
West Kinki 33,112 22,755 17 48 
Southwest Chugoku–Shikoku 50,720 11,539 14 42 
South Kyushu–Okinawa 44,467 14,597 12 39 
Total  377,865 128,055 79 239 
Figure 1

Amounts of distributed water of target DWTPs.

Figure 1

Amounts of distributed water of target DWTPs.

Questionnaire items are listed in Table 2. Although the items included detailed information, such as concentrations of off-flavor compounds in raw and finished water, not all of the RWS and DWTPs analyzed and/or recorded these data. Therefore, recovery ratios were not the same for each item. However, a relatively high recovery ratio (>78%) was obtained for all items in this study. The recovery ratios were calculated using the following equations.

Table 2

Questionnaire items and recovery ratios

Question items Recovery ratio (%) 
Presence or absence of off-flavor problems 100 
Types of water source of target DWTPs 100 
Time of year of off-flavor events 97 
Types of off-flavor 93 
Types of organisms associated with off-flavor events 91 
Concentration of off-flavor compounds in raw water 86 
Concentration of off-flavor compounds in finished water 78 
Water temperature of raw water when off-flavor events occurred 92 
pH of raw water when off-flavor events occurred 92 
Countermeasures against off-flavor problems 94 
Question items Recovery ratio (%) 
Presence or absence of off-flavor problems 100 
Types of water source of target DWTPs 100 
Time of year of off-flavor events 97 
Types of off-flavor 93 
Types of organisms associated with off-flavor events 91 
Concentration of off-flavor compounds in raw water 86 
Concentration of off-flavor compounds in finished water 78 
Water temperature of raw water when off-flavor events occurred 92 
pH of raw water when off-flavor events occurred 92 
Countermeasures against off-flavor problems 94 

(Recovery ratios for ‘presence or absence of off-flavor problems’ and ‘types of water source’) = (Number of DWTPs responding to each question item)/(Number of total target DWTPs)

(Recovery ratios for other items) = (Number of off-flavor events with information on each question item)/(Number of total off-flavor events*)

*If the type of off-flavor or time of year were different, the off-flavor events were counted as separate events even for the same DWTP.

The presence of off-flavor problems was counted only if operational conditions were changed and countermeasures were taken due to the occurrence of off-flavor compounds in raw water for drinking water treatment. In general, determining whether a countermeasure is taken is decided by the individual DWTP based on concentrations of off-flavor compounds in raw water. The DWTPs normally have individual management criteria for off-flavor compounds, such as geosmin, which are stricter than the national drinking water quality standards. Only off-flavor events that would be related to organisms were included in the responses from DWTPs, and other off-flavors, such as chemical and chlorinous odors, were not counted in this survey. The ‘organisms associated with off-flavor events’ were those present in source and/or raw water when off-flavor events occurred and were presumed by DWTPs to be responsible for producing off-flavor compounds. The organisms were identified by trained staff at 85% of the target RWS, while the remaining 15% of RWS outsourced or did not perform such identification. In general, identification was performed using a microscope according to the Japanese standard manuals for the examination of water (Japan Water Works Association 2008, 2010).

RESULTS AND DISCUSSION

Proportion of DWTPs affected by off-flavor problems

The results of the questionnaire survey indicated that 43 of 79 (54%) RWS and 89 of 239 (36%) DWTPs were affected by off-flavor problems in a recent 2-year period (October 2010–September 2012). Figure 2 shows the regional distribution of the proportion of DWTPs affected by off-flavor problems. Japan has a varied climate due to the shape of the country, which stretches over 2,500 km from north to south. Typical maximum and minimum water temperatures of surface water in the north region were 20 ± 5 °C and 5 ± 5 °C, respectively, while those in the south region were 30 ± 5 °C and 12 ± 5 °C, respectively, although these were dependent on altitude and geography (Ministry of Land, Infrastructure, Transport and Tourism Japan 2013). Nevertheless, off-flavor incidents occur in all regions, although the proportion differed between regions. In this study, it was not possible to determine the reasons for these regional differences, and the eutrophication status of each water source should be investigated in future studies.

Figure 2

Regional distribution of the proportion of RWS and DWTPs affected by off-flavor problems in Japan.

Figure 2

Regional distribution of the proportion of RWS and DWTPs affected by off-flavor problems in Japan.

Types of water source related to off-flavors

Figure 3 shows the proportion of types of water source of all target DWTPs and the DWTPs affected by off-flavor problems. As expected, all DWTPs affected by off-flavor incidents take in surface water. In particular, most DWTPs affected by off-flavor incidents take in water from closed water areas, such as dams and lakes, where phototrophic microorganisms, such as cyanobacteria and algae, can grow easily.

Figure 3

Proportion of types of water source of all target DWTPs (left) and DWTPs affected by off-flavor problems (right) on the basis of the number of DWTPs. (For DWTPs that take in water from several types of water source, the main type of water source was counted.)

Figure 3

Proportion of types of water source of all target DWTPs (left) and DWTPs affected by off-flavor problems (right) on the basis of the number of DWTPs. (For DWTPs that take in water from several types of water source, the main type of water source was counted.)

Types of off-flavor, concentrations of off-flavor compounds and organisms associated with off-flavor events

Table 3 shows types of off-flavor, concentrations of off-flavor compounds, and organisms associated with off-flavor events. As expected, an earthy–musty odor was the most common off-flavor encountered. There were more DWTPs with problems associated with geosmin than with 2-MIB in Japan. The maximum concentrations of geosmin and 2-MIB in raw water in some DWTPs were >500 ng/L. Although the concentrations in finished water were much lower than those in raw water due to the countermeasures implemented, the maximum concentrations in some DWTPs were higher than the Japanese drinking water quality standard values (10 ng/L). Cyanobacteria were the most common earthy–musty odor producers, although actinomycetes were also reported in some cases. Anabaena spp. were the most common cyanobacteria producing geosmin. It has been reported that the concentrations of geosmin and Anabaena spp. in water sources are significantly correlated (Uwins et al. 2007). Phormidium spp. and Oscillatoria spp. were common cyanobacteria responsible for production of 2-MIB. Microcystis spp. have not been reported to produce geosmin, and Anabaena spp. have not been shown to produce 2-MIB (Jüttner & Watson 2007). However, they were reported as responsible organisms in this questionnaire survey, although the number of responses was small. This indicates that the organisms associated with off-flavor events estimated by DWTPs were not always correct. This is because various organisms associated with earthy–musty odor are present in water sources, and identification requires a high level of skill. Indeed, organisms could not be identified in many cases, as shown in Table 3.

Table 3

Types of off-flavor, concentrations of off-flavor compounds and organisms associated with off-flavor events

Types of off-flavor Concentration of off-flavor compoundsa Types of organisms associated with off-flavor events 
  Raw water Finished water   Genus 
Name Ratiob (%) Mean Max. Mean Max. Name Ratiob (%) Name Ratiob (%) 
Earthy–musty (geosmin) 39 30 520 3.0 13 Cyanobacteria 57 Anabaena 61 
Oscillatoria 16 
Phormidium 
Microcystis 
Unidentified 12 
Actinomycetes – – 
Unidentified 37 – – 
Earthy–musty (2-MIB) 29 79 1,400 3.2 14 Cyanobacteria 68 Phormidium 54 
Oscillatoria 42 
Anabaena 
Actinomycetes 18 – – 
Unidentified 14 – – 
Fishy smell 26 130 600 1.3 Chrysophyta 58 Uroglena 97 
Dinobryon 
Fish eggs 29 – – 
Cyanobacteria Oscillatoria 100 
Unidentified 12 – – 
Others 18 40 < 1 Cyanobacteria 50 Microcystis 38 
Anabaena 38 
Oscillatoria 13 
Phormidium 13 
Diatom 17 Nitzschia 33 
Unidentified 67 
Chlorophyta 11 Unidentified 100 
Amoeba 11 Asterocaelum 100 
Unidentified 11 – – 
Types of off-flavor Concentration of off-flavor compoundsa Types of organisms associated with off-flavor events 
  Raw water Finished water   Genus 
Name Ratiob (%) Mean Max. Mean Max. Name Ratiob (%) Name Ratiob (%) 
Earthy–musty (geosmin) 39 30 520 3.0 13 Cyanobacteria 57 Anabaena 61 
Oscillatoria 16 
Phormidium 
Microcystis 
Unidentified 12 
Actinomycetes – – 
Unidentified 37 – – 
Earthy–musty (2-MIB) 29 79 1,400 3.2 14 Cyanobacteria 68 Phormidium 54 
Oscillatoria 42 
Anabaena 
Actinomycetes 18 – – 
Unidentified 14 – – 
Fishy smell 26 130 600 1.3 Chrysophyta 58 Uroglena 97 
Dinobryon 
Fish eggs 29 – – 
Cyanobacteria Oscillatoria 100 
Unidentified 12 – – 
Others 18 40 < 1 Cyanobacteria 50 Microcystis 38 
Anabaena 38 
Oscillatoria 13 
Phormidium 13 
Diatom 17 Nitzschia 33 
Unidentified 67 
Chlorophyta 11 Unidentified 100 
Amoeba 11 Asterocaelum 100 
Unidentified 11 – – 

aUnits: Geosmin and 2-MIB: ng/L; other off-flavor: TON. The means and maximum concentrations were calculated based on the maximum concentration of each off-flavor event.

bRatio was calculated on the basis of the number of off-flavor events. If multiple answers were obtained, all answers were counted.

A ‘fishy’ smell accounted for 26% of total off-flavor events. Although the maximum TON in raw water in some DWTPs was >100, those in finished water always met the management goals of the Japanese drinking water quality management system (3 TON). Interestingly, the off-flavors that were caused by Chrysophyta and fish eggs were categorized into the same ‘fishy’ smell, although the off-flavor compounds would be different. It has been reported that Chrysophyta, such as Uroglena spp., have the potential for producing a ‘cucumber’ or ‘fishy’ smell (Yano et al. 1988; Watson et al. 2008). The large freshwater fish Hypophthalmichthys molitrix is known to spawn in the upper basin of the intake points of DWTPs and the eggs cause a severe off-flavor problem (Japan Water Works Association 2000). Other off-flavors such as a ‘seaweed’-like smell were also reported in this questionnaire survey although the number of such events was small.

Time of year of off-flavor events and water temperature and pH of raw water when off-flavor events occurred

Figure 4 shows the time of year of off-flavor events. As expected, the events occurred frequently during the summer and early fall months (June–September). In temperate regions such as Japan, thermal stratification, which is a favorable condition for cyanobacterial growth, generally occurs seasonally (summer–fall) in water sources (Newcombe et al. 2010). This is one of the reasons why off-flavor events occurred during the summer and early fall months. In addition, water temperature is high in these seasons, thus promoting the growth of such organisms. As shown in Figure 5, the water temperature of raw water when off-flavor problems occurred was higher than that during normal periods. The average water temperature when off-flavor problems occurred was approximately 19 °C. It has been reported that maximum growth rates are achieved at temperatures above 25 °C for most cyanobacteria, which are the most common off-flavor producers. It has been suggested that temperature optima for cyanobacteria are higher than those for Chlorophyta and diatoms (Robarts & Zohary 1987). Meanwhile, off-flavor events were reported even if the water temperature of raw water was lower than 10 °C, as shown in Figure 4. This was mainly because Chrysophyta (Uroglena spp.) can grow and produce off-flavor compounds under low-temperature conditions. The average water temperature when organisms associated with off-flavor events were Uroglena spp. was 13 °C.

Figure 4

Time of year of off-flavor events. (If off-flavor events stretched over a period of multiple months, all months were counted.)

Figure 4

Time of year of off-flavor events. (If off-flavor events stretched over a period of multiple months, all months were counted.)

Figure 5

Distribution of water temperature (left) and pH (right) in raw water. *Distribution data were obtained from a large database of Japanese drinking water quality (Japan Water Works Association 2013).

Figure 5

Distribution of water temperature (left) and pH (right) in raw water. *Distribution data were obtained from a large database of Japanese drinking water quality (Japan Water Works Association 2013).

The pH of raw water when off-flavor problems occurred was higher than that during normal periods, indicating that off-flavor problems often occur when water sources are eutrophicated. It is well known that photosynthesis by cyanobacteria and algae in eutrophicated dams and lakes increases pH. The average pH when off-flavor problems occurred was approximately 7.6.

Countermeasures against off-flavors

Figure 6 shows the proportions of countermeasures against off-flavors in the questionnaire survey. To eliminate off-flavor compounds, most DWTPs (approximately 58%) adopt powdered activated carbon treatment, which considerably increases operational costs as in other countries (Cook et al. 2001; Giglio et al. 2011). All of the DWTPs determine the quantity of powdered activated carbon to add based on empirically obtained individual correlation equations or tables for the concentration of off-flavor compounds in raw water and addition quantity. A few DWTPs adopt other countermeasures, such as changes in point of chlorine addition and enhancement of coagulation.

Figure 6

Proportion of countermeasures against off-flavors in Japan on the basis of the number of off-flavor events. (If multiple answers were obtained, all answers were counted.)

Figure 6

Proportion of countermeasures against off-flavors in Japan on the basis of the number of off-flavor events. (If multiple answers were obtained, all answers were counted.)

CONCLUSIONS

The results of the questionnaire survey indicated that 43 of 79 (54%) RWS and 89 of 239 (36%) DWTPs were affected by off-flavor problems in a recent 2-year period (October 2010–September 2012). Off-flavor incidents occurred in all regions in Japan. Most DWTPs affected by off-flavor incidents take in water from closed water areas, such as dams and lakes where phototrophic microorganisms, such as cyanobacteria and algae, can grow easily. The most common off-flavor was an earthy–musty odor, while a fishy smell was also reported in some cases. The common producers of geosmin were Anabaena spp., while those responsible for 2-MIB were Phormidium spp. and Oscillatoria spp. The off-flavor events occurred frequently during the summer and early fall months (June–September). The water temperature and pH of raw water for the production of drinking water when off-flavor problems occurred were higher than those during normal periods. The average water temperature and pH when off-flavor problems occurred were approximately 19 °C and 7.6 °C, respectively. To eliminate off-flavor compounds, most DWTPs (about 58%) adopt powdered activated carbon treatment.

ACKNOWLEDGEMENTS

This study was partially supported by a Health Labour Sciences Research Grant (H24-Kenki-Ippan-003) from the Ministry of Health, Labour and Welfare, Japan, and JSPS KAKENHI Grant Number 24760440. We thank Mr Kitamura, Dr Oikawa and Dr Fujise for helping to make the questionnaire survey sheet. We are grateful to Japan Water Works Association for helping to gather the survey data. We thank the workers at the DWTPs for providing data for the survey.

REFERENCES

REFERENCES
Giglio
S.
Chou
W. K.
Ikeda
H.
Cane
D. E.
Monis
P. T.
2011
Biosynthesis of 2-methylisoborneol in cyanobacteria
.
Environ. Sci. Technol.
45
(
3
),
992
998
.
Japan Water Works Association
2000
Animals
. In:
Organisms of Water Supplies in Japan – Photographs and Descriptions
.
Japan Water Works Association
,
Tokyo, Japan
, pp.
103
168
.
Japan Water Works Association
2013
Database of water quality of aqueduct (in Japanese)
.
Available from: www.jwwa.or.jp/mizu/list.html (accessed 28 September 2013)
.
Jüttner
F.
Watson
S. B.
2007
Biochemical and ecological control of geosmin and 2-methylisoborneol in source waters
.
Appl. Environ. Microbiol.
73
(
14
),
4395
4406
.
Kishida
N.
Konno
Y.
Nemoto
K.
Amitani
T.
Maki
A.
Fujimoto
N.
Akiba
M.
2013
Recent trends in microorganism-related off-flavor problems in drinking water treatment systems in Japan
.
Water Sci. Technol. Water Suppl.
13
(
5
),
1228
1235
.
Matsumoto
A.
Tsuchiya
Y.
1988
Earthy-musty odor-producing cyanophytes isolated from five water areas in Tokyo
.
Water Sci. Technol.
20
(
8/9
),
179
183
.
Ministry of Health, Labour and Welfare Japan
2013
Trends in off-flavours in water supply systems (in Japanese)
. .
Ministry of Land, Infrastructure, Transport and Tourism Japan
2013
Water information system (in Japanese)
.
Available from: www1.river.go.jp/ (accessed 7 September 2013)
.
Newcombe
G.
House
J.
Ho
L.
Baker
P.
Burch
M.
2010
Management Strategies for Cyanobacteria (Blue-Green Algae): A Guide for Water Utilities
.
Water Quality Research Australia
.
Oikawa
E.
Ishibashi
Y.
2004
Species specificity of musty odor producing Phormidium tenue in Lake Kamafusa
.
Water Sci. Technol.
49
(
9
),
41
46
.
Organisms of water supplies in Japan – photographs and descriptions
2008
revised edn
,
Japan Water Works Association
,
Tokyo, Japan
(in Japanese)
.
Standard Method for the Examination of Water
2010
Japan Water Works Association
,
Tokyo, Japan
(in Japanese)
.
Watson
S. B.
Ridal
J.
Boyer
G. L.
2008
Taste and odour and cyanobacterial toxins: impairment, prediction, and management in the Great Lakes
.
Can. J. Fisheries Aqua. Sci.
65
(
8
),
1779
1796
.
Yano
H.
Nakahara
M.
Ito
H.
1988
Water blooms of Uroglena americana and the identification of odorous compounds
.
Water Sci. Technol.
20
(
8–9
),
75
80
.