The potential of powdered activated carbon (PAC) adsorption for odor removal in low turbidity drinking water was investigated. The batch experiments were conducted under various conditions including PAC species, dosage, contact time and dosing point. The effects of pre-chlorination and PAC dosage on turbidity were also studied. Results showed that adsorption was quite vulnerable to initial threshold odor number (TON), and higher influent TON required a larger dosage. Both PAC species (derived from coal and wood) presented excellent adsorption capacity for odorants. The adsorption process versus time had three steps and the adsorption kinetics were well fitted by the second order model. Pre-chlorination had an adverse effect on both raw water TON and odor removal. PAC adsorption was enhanced by dosing during coagulation and could, in turn, save coagulant dosage.

INTRODUCTION

It is widely acknowledged that several organic compounds at ng/L to mg/L concentrations can cause odor problems in drinking water (Chen et al. 1997). Among the group of earthy-musty odorants, geosmin (GSM), 2-methylisoborneol (MIB), 2-isopropyl-3-methoxy-pyrazine (IPMP), 2-isobutyl-3-methoxy-pyrazine (IBMP) and 2,3,6-trichloroanisole (TCA) are five major components, within which MIB and GSM are of particular importance (Chen et al. 1997; Ho et al. 2007). The treatment and control of odor compounds has come to the attention of researchers in the field of drinking water supply, since odorants have long been related to drinking water quality and safety (Burlingame et al. 2004). The removal of most odor compounds from drinking water is a difficult task due to their extremely low threshold odor number (TON), usually in the range of ng/L (Antonopoulou et al. 2014). Therefore, it is necessary to seek an effective method of removal.

Powdered activated carbon (PAC) is often used in treatment plants because it is relatively inexpensive and flexible (Wnorowski 1992; Cook et al. 2001). Numerous studies have researched the adsorption of odorants by PAC and verified its availability and effectiveness (Wnorowski 1992; Bansal et al. 2005; Srinivasan & Sorial 2009; Srinivasan & Sorial 2011). Chen et al. (1997) evaluated the odor removal process by activated carbon derived from four different materials. Cook et al. (2001) investigated MIB and geosmin removal in four raw waters for dosage prediction. Effects of chlorine oxidation were studied by Jung et al. (2004). Experiments were also conducted to reveal the effects of contact time (Gholizadeh et al. 2013) or dosing point (Boonanuntanasarn et al. 2014).

However, systematic investigations into odor removal by PAC in low turbidity drinking water under various conditions including PAC species, dosage, contact time, dosing point, chlorine oxidation and coagulation are rarely reported (Turk et al. 1972; Wnorowski 1992; Srinivasan & Sorial 2011; Zhang et al. 2015). We conducted batch experiments to look into PAC adsorption of odorants in low turbidity drinking water, aiming to add some insights into its applications.

MATERIAL AND METHODS

Water quality and PAC preparation

The water targeted in this study is from a drinking water source of a city in the south of China. The characteristics of the water are given in Table 1.

Table 1

Raw water quality in this study

ParameterUnitValue
pH a 6.97 
Turbidity NTU 23.9 
CODMn mg/L 6.88 
DO mg/L 1.37 
Ammonia mg/L 4.97 
Nitrate mg/L 2.21 
Nitrite mg/L 0.448 
Chloride mg/L 66.1 
TON a 300 
Algae 104/L 337 
ParameterUnitValue
pH a 6.97 
Turbidity NTU 23.9 
CODMn mg/L 6.88 
DO mg/L 1.37 
Ammonia mg/L 4.97 
Nitrate mg/L 2.21 
Nitrite mg/L 0.448 
Chloride mg/L 66.1 
TON a 300 
Algae 104/L 337 

ano unit. Experimental methods.

Two kinds of PAC (coal carbon and wood charcoal) were used in this study, and were purchased from Shanxi Activated Carbon Corporation (Shanxi Province, China). Only particles which passed through a 320 mesh sieve were used for experiments. Both types of PAC had similar iodine and methylene blue values (906 and 182 mg/g for coal carbon, 910 and 185 mg/g for wood charcoal, respectively, data were from the corporation).

Odor removal was carried out by batch tests, 1 L of raw water with a certain amount of PAC was agitated and settled (agitation procedure as described in Table 2) for measuring TON. TON, turbidity and color were measured according to Standard Methods (APHA 2005) at a temperature of 60 °C. All experiments were conducted in triplicate and the average values were used in the figures and tables.

Table 2

Agitation procedures used in this study

StepTimeaSpeedbDosage
30 60 PAC 
300 Coagulant 
200 c 
100 c 
60 c 
20 c 
StepTimeaSpeedbDosage
30 60 PAC 
300 Coagulant 
200 c 
100 c 
60 c 
20 c 

aunit: min.

bunit: r/min.

cno chemicals added.

RESULTS AND DISCUSSION

Effect of PAC species

Figure 1 shows odor removal performance by two kinds of PAC, namely coal carbon and wood charcoal. Both coal carbon and wood charcoal exhibited a good capacity to remove odor at an initial TON of 300 (over 83.33% for both). Even though the removal efficiency of coal carbon exceeded that of wood charcoal at a PAC dosage of 20–30 mg/L, there was only a maximum discrepancy of 1.45%. In view of its odor removal performance and economic status, we chose coal carbon for further studies.
Figure 1

Effect of PAC species on odor removal.

Figure 1

Effect of PAC species on odor removal.

Choi & Chung (2015) reported that coal carbon has a smaller pore radius than wood charcoal, thus enhancing its adsorptive force with organic compounds. It could be concluded that coal carbon is more effective for adsorption of organics with lower molecules (Bansal et al. 2005), which is in accordance with the present concept that major groups of odorants have a molar mass of about 200 (Antonopoulou et al. 2014).

Effect of PAC dosage

It can be seen from Figure 2 that a good linear relationship between effluent TON and dosage ranging from 10 to 40 mg/L was achieved (the correlation coefficient (R2) was over 0.90), and effluent TON decreased with the addition of PAC dosage. In addition, the average removal rate (76–87%) increased with the increasing dosage (10–40 mg/L). The removal rate increased slightly with the initial TON as there was only a maximum of 5% difference among samples with an initial TON between 100–300. PAC dosage was a key factor in odor removal, as an average increase of 20% in the removal rate was achieved with increasing the dosage from 10 to 40 mg/L.
Figure 2

Effect of PAC dosage: TON (—), fitting line (…), initial TON 100 (▪), 150 (●), 200 (▴), 250 (▾), 300 (◆), average removal rate (×).

Figure 2

Effect of PAC dosage: TON (—), fitting line (…), initial TON 100 (▪), 150 (●), 200 (▴), 250 (▾), 300 (◆), average removal rate (×).

PAC dosage is a vital factor in the application of PAC for odor removal, for both overdosing and under-dosing could lead to either an unacceptable cost or the ineffective use of adsorbents (Cook et al. 2001). The increase of the odor removal rate was due to a more adsorbent surface area and more adsorption sites available (Gallego et al. 2013). Considering the odor removal rate and the residual TON, 40 mg/L was considered to be the optimum concentration of PAC dosage for other experiments in this study.

Effect of contact time

The plot of adsorption performance versus contact time with an initial TON of 400 and PAC dosage of 40 mg/L is illustrated in Figure 3. The odor removal process can be divided into three steps, i.e., rapid adsorption in the first 5 min, during which about 80% of TON was eliminated, slower removal in the following 25 min (over 90% of odor compounds was removed during the first 30 min), and a stable period from 30–150 min. It is apparent that adsorption reached an equilibrium at about 120 min. In addition, the adsorption process was well fitted by the second-order model (R2 = 0.9985). The rate constant and equilibrium adsorption capacity were 0.00287 and 373.34, respectively.
Figure 3

Effect of contact time: initial TON 400, PAC dosage 40 mg/L.

Figure 3

Effect of contact time: initial TON 400, PAC dosage 40 mg/L.

Bansal et al. (2005) and Han et al. (2011) described the three steps of adsorption. Initially odor compounds rapidly reach the boundary layer by mass transfer, then they slowly diffuse from the boundary layer film onto the adsorption surface when most of the external sites available have been occupied, and finally they diffuse into the porous structure of the PAC. According to Savasari et al. (2015), the modelling results showed that the TON removal process was concentration dependent and that the rate determining step involved chemical adsorption. The conclusions were also in accordance with related research investigating the kinetics of odor removal by adsorption (Gholizadeh et al. 2013; Watanabe et al. 2014; Xin et al. 2014).

Effect of dosing point

The difference in odor removal performance by adding PAC before and during coagulation is presented in Figure 4. It is apparent that better removal efficiency was obtained when PAC was dosed during coagulation (when visible flocs occurred) rather than before coagulation. A higher removal rate of 2.22–4.92% existed between the two kinds of dosing points. With the increase in PAC dosage, the results got even closer.
Figure 4

Effect of dosing point: initial TON 150.

Figure 4

Effect of dosing point: initial TON 150.

This is mainly because odor compounds in raw water can be removed by both coagulation and PAC adsorption (Alfonsin et al. 2015). Adsorption sites, especially macropores and mesopores were occupied by these compounds thus restricting the migration of odor compounds within the porous structure (Bansal et al. 2005; Zhang et al. 2015). This could be avoided by dosing PAC at mid-coagulation so that the major amount of these organics were removed by coagulation (Bazafkan et al. 2015).

Effect of chlorine addition

Compared with raw water, pre-chlorination led to the increase of effluent TON (Figure 5(a)). Although as much as about 5 mg/L chlorine could counteract the side effect of TON increase, the odor species turned from earthy odor (0–1 mg/L), fishy odor (2–3 mg/L) to medicinal odor (above 4 mg/L). Effluent TON reached its peak at a chlorine concentration of 2 mg/L and decreased afterwards, which verified the odor removal capacity of pre-chlorination. Overall chlorine addition in the present study brought about a larger effluent TON and raised the loading rate of following processes. Pre-chlorination should be avoided regarding its adverse effect on further treatment.
Figure 5

Effect of chlorine addition: (a) effect of chlorine addition concentration on raw water TON; (b) effect of chlorine addition on odor removal by PAC, without chlorine addition (●), with 5 mg/L chlorine addition (▪); (c) effect of chlorine oxidization time on odor removal by PAC, chlorine concentration 5 mg/L, initial TON 200, PAC dosage 40 mg/L.

Figure 5

Effect of chlorine addition: (a) effect of chlorine addition concentration on raw water TON; (b) effect of chlorine addition on odor removal by PAC, without chlorine addition (●), with 5 mg/L chlorine addition (▪); (c) effect of chlorine oxidization time on odor removal by PAC, chlorine concentration 5 mg/L, initial TON 200, PAC dosage 40 mg/L.

In addition, chlorine addition along with PAC weakened the adsorption capacity of odor compounds in contrast with PAC only (Figure 5(b)). However, this trend declined while the chlorine concentration increased. There was almost no significant discrepancy when chlorine addition was above 4 mg/L. Effect of pre-chlorination time before PAC addition was also studied (Figure 5(c)). It is obvious that better removal efficiencies were obtained at larger time lags.

Trihalomethanes and haloacetic acid as the main byproducts of pre-chlorination have adverse effects on odor removal (Jung et al. 2004). Many researchers have found that natural organic matter (Wnorowski 1992; Srinivasan & Sorial 2011; Zhang et al. 2015) could lead to competitive adsorption and result in the failure of effective removal of odorants in aqueous environments. The non-odor compounds occupied adsorption sites in the PAC, hence requiring a greater dosage to obtain the same removal efficiencies (Turk et al. 1972; Watanabe et al. 2014). Therefore, eliminating interference of other organics is vital for removal effectiveness as well as reducing cost (Srinivasan & Sorial 2011).

Effect of PAC addition on turbidity

The effect of PAC addition on effluent turbidity reduction by polyaluminium chloride (PACI) coagulation is shown in Figure 6. It can be seen from Figure 6 that coagulation can be strengthened by PAC addition, as effluent turbidity decreased from 0.97 NTU (no PAC) to 0.79 NTU (40 mg/L PAC). However, even with PAC adsorption and PACI coagulation, the raw low turbidity water had a poor turbidity removal rate (maximum 18.56% at PAC dosage of 40 mg/L).
Figure 6

Effect of PAC dosage on water turbidity: PACl contact time of 30 min followed by 40 mg/L PAC.

Figure 6

Effect of PAC dosage on water turbidity: PACl contact time of 30 min followed by 40 mg/L PAC.

Szlachta & Adamski (2009) and Terauchi et al. (1995) showed that PAC enhanced coagulation and reduced the required dose while increasing the process efficiency. The reason for this could be due to the very active surface of PAC and its high capacity to adsorb organic matter (Bazafkan et al. 2015). In addition, PAC in low turbidity water also increases the number of particles which can contribute to the nucleus for coagulation (Szlachta & Adamski 2009).

CONCLUSIONS

Various factors affecting odor removal from low turbidity drinking water were investigated by batch tests. The main results were as follows:

  1. Excellent and similar removal rates were obtained by both PACs derived from coal and wood. In view of the greater removal efficiency and lower cost, coal carbon was more suitable for application.

  2. PAC adsorption was vulnerable to an initial TON ranging from 100 to 300. PAC dosage was a vital factor and the optimal dosage was 40 mg/L for raw water with a TON of 100–300.

  3. Equilibrium could be achieved as rapidly as 2 h with an initial TON of 400 and PAC dosage of 40 mg/L. In addition, the adsorption process was well fitted by the second order kinetics model.

  4. Pre-chlorination had an adverse effect on PAC treatment. The longer time lag of chlorine oxidization before PAC dosage was more beneficial for odor removal. Dosing PAC during coagulation instead of before coagulation was more favorable for odor removal.

  5. The coagulation process by PACI was aided with a PAC dosage from 10 to 40 mg/L.

ACKNOWLEDGEMENTS

This research was financially supported by the National Natural Science Foundation of China (NSFC) (No. 51378400), and the Natural Science Foundation of Hubei Province, China (No. 2013CFB289).

REFERENCES

REFERENCES
Alfonsin
C.
Lebrero
R.
Estrada
J. M.
Munoz
R.
Kraakman
N. J.
Feijoo
G.
Moreira
M. T.
2015
Selection of odour removal technologies in wastewater treatment plants: a guideline based on Life Cycle Assessment
.
J. Environ. Manage.
149
,
77
84
.
Antonopoulou
M.
Evgenidou
E.
Lambropoulou
D.
Konstantinou
I.
2014
A review on advanced oxidation processes for the removal of taste and odor compounds from aqueous media
.
Water Res.
53
,
215
234
.
APHA
2005
Standard Methods for the Examination of Water and Wastewater
.
American Public Health Association (APHA)
,
Washington, DC
,
USA
.
Bansal
R. C.
Goyal
M.
Bansal
R. C.
Goyal
M.
2005
Activated Carbon Adsorption. CRC Press, UK
.
Bazafkan
M.
Takdastan
A.
Eslami
A.
2015
Survey of poly aluminum chloride with powdered activated carbon as coagulant aid in TOC removal from Karoon River in Iran
.
Exchange (MIEX resin)
5
,
7
.
Burlingame
G. A.
Suffet
I. H.
Khiari
D.
Bruchet
A. L.
2004
Development of an odor wheel classification scheme for wastewater
.
Water Sci. Technol.
49
(
9
),
201
209
.
Jung
S. W.
Baek
K. H.
Yu
M. J.
2004
Treatment of taste and odor material by oxidation and adsorption
.
Water Sci. Technol.
49
(
9
),
289
295
.
Terauchi
N.
Ohtani
T.
Yamanaka
K.
Tsuji
T.
Sudou
T.
Ito
K.
1995
Studies on a biological filter for musty odor removal in drinking water treatment processes
.
Water Sci. Technol.
31
(
11
),
229
235
.
Turk
A.
Haring
R. C.
Okey
R. W.
1972
Odor control technology
.
Environ. Sci. Technol.
6
(
7
),
602
607
.
Wnorowski
A.
1992
Tastes and odours in the aquatic environment: a review
.
Water, S.A.
18
(
3
),
203
214
.
Xin
X. D.
Wang
M. Q.
Ge
X.
Zhao
Q. H.
Sun
S. H.
Jia
R. B.
2014
Highly efficient removal of geosmin and 2-methylisoborneol by carboxylated multi-walled carbon nanotubes
.
Monatsh. Chem.
145
(
5
),
747
754
.
Zhang
Y.
Zhao
X.
Zhang
X.
Peng
S.
2015
A review of different drinking water treatments for natural organic matter removal
.
Water Sci. Tech.-W. Sup.
15
(
3
),
442
.