Poor water quality is a major contributing factor to disease in developing countries. Silver-coated ceramic pot filters (CPFs) are a relatively common form of household water treatment system (HWTS) representing an effective and sustainable technology for poor communities. Water production seems to be the major limiting factor of the CPF's lifetime and sustainability since low flow rates do not produce an adequate daily volume of treated water. This paper describes a long-term study of CPF flow rates under controlled conditions using three different water sources. The relationship between water characteristics and flow rate was assessed with the intent of identifying the principal parameters that impact CPF water production. The study concluded that turbidity seems to be the principal indicator in determining CPF lifetime in terms of quantity of treated water. There is no evidence that biological activity also contributes to premature failure of CPFs and the data did not indicate that chemical precipitation is responsible for the filter clogging. Manufacturers commonly conduct initial flow rate tests using clear water as a measure of quality assurance. However, the relationship between initial flow rate and average flow rate during the lifetime of the CPF should be further studied.

INTRODUCTION

The United Nations World Water Assessment Programme (WWAP) (2015) demonstrates how water resources and services are essential to achieving global sustainability and states that water is at the core of sustainable development. The World Health Organization (WHO) and the United Nations Children's Fund (UNICEF) (2014) indicates that although the drinking water target of the Millennium Development Goal (MDG) was globally met in 2010, 748 million people still lacked access to an improved drinking water source in 2012. Most of these people are poor with over 90% living in rural areas and almost a quarter relying on untreated surface water. According to Kallman et al. (2011), the WHO's definition of “improved” water source is not based on water quality; therefore people that rely on this water source may face contamination and/or recontamination problems during water collection, transport and storage. UNICEF (2008) indicates that many of the improved sources do not provide safe water due to microbiological contamination and that water quality interventions have a greater impact when applied at the household level. A household water treatment system (HWTS) is effective if it produces sufficient safe drinking water for a family in the long-term. Hunter (2009) states that the most effective HWTSs in the long-term are ceramic filters, and Oyanedel-Craver & Smith (2008) conclude that silver-coated ceramic pot filters (CPFs) represent an effective and sustainable technology for poor communities.

CPFs are locally manufactured porous clay pots placed in a water container equipped with a lid and a spigot. The filter is made of a mixture of clay, a burn-out material (usually sawdust), and water. The mixture is pressed into pot-shaped molds, air-cured, fired in a kiln, and finally coated with a suspension of silver nanoparticles. The microbiological removal efficacy of CPFs has been documented by Lantagne (2001), van Halem et al. (2007), Oyanedel-Craver & Smith (2008) and others. According to van Halem et al. (2009), the decrease in water production during operation is due to filter clogging. The Ceramics Manufacturing Working Group (CMWG) (2011) and CPF manufacturers recommend scrubbing the filter when the flow rate reaches an unsatisfactorily low level. Lantagne (2001) and van Halem et al. (2007) show that the flow rate decrease is significant when surface water is treated and state that scrubbing the filter partially and temporarily restores the flow rate. According to van Halem et al. (2009), this cleaning method does not prevent long-term clogging which makes flow rate the limiting factor of the CPF's lifetime.

Flow rate is defined by CMWG (2011) as the volume of water treated by a saturated and full CPF during the first hour after production. A flow-rate test is one of the most common indirect quality tests, and it can be used as an indicator of production consistency, pathogen removal efficacy, and water production capacity. A survey described by CMWG (2011) indicated that flow rates accepted by factories range from 1.0 to 3.0 litres per hour (Lh−1) minimum to 2.0 to 5.0 Lh−1 maximum. According to Lantagne et al. (2010), the production process for a CPF is considered reliable when flow rates are maintained between 1 and 2 Lh−1. The minimum acceptable flow rate is established by CMWG (2011) as 1 Lh−1 based on consumer needs while the maximum is calculated based on the capacity of the filtering unit: 2.5 Lh−1 for 7.2-L capacity and 3.5 Lh−1 for 10-L capacity.

Although several theoretical and experimental studies have been conducted with the aim to understand flow-rate behavior of CPFs, its relationship to the quantity and quality of water is not well understood. Mathematical models have been developed with the aim of describing the hydraulic characteristics of CPFs (Plappally et al. 2009; Elmore et al. 2011; Schweitzer et al. 2013; Yakub et al. 2013). Lantagne et al. (2010) conducted a study comparing flow-rate behavior and coliform removal efficacy of several filter designs, and indicated the need for long-term testing. Van Halem et al. (2007, 2009) described a long-term study of CPFs and an investigation of three possible clogging mechanisms conducted through assessment of the effects of the rehabilitation of the filters. Van Halem et al. (2009) concluded that neither organic nor inorganic fouling were the principal causes of short-term clogging, but physical fouling by colloids was observed as a potential cause of failure.

This paper describes a long-term study of CPF flow rates under controlled conditions using three different water sources. The primary purpose of this study is to assess the relationship between water characteristics and flow rate with the intent of identifying the principal parameters that impact CPF lifetime.

METHODS

A long-term performance study was conducted in the groundwater hydrology laboratory of the Missouri University of Science and Technology to collect flow-rate and water-quality data using production CPFs. The experiments were conducted using nine silver-coated CPFs, manufactured and quality-control tested by a factory near Antigua, Guatemala. There, the flow rate of CPFs is tested using clear water, and filters outside the range of 1 to 2 Lh−1 are discarded. The subject filters were selected randomly from a stock of 100 filtering units and divided into three sets of three. Each set was used to establish three different systems with the same setup and conditions, but with differing water sources. The water sources were as follows:

  • Surface water (SW) collected in the Little Prairie Lake, a small fresh-water body near Rolla, Missouri, USA.

  • Challenge water (CW) created by mixing tap water (97%) and influent waste water (3%) from the Rolla Waste Water Treatment Plant.

  • Municipal tap water (TW) from Rolla, Missouri, USA.

Figure 1 depicts the constant head apparatus used to maintain constant flow with a relatively constant head through each CPF. Each system consisted of a 1,000-L tank used to periodically collect, store, and/or mix the source water. A timer-controlled pump moved the source water from the tank to a 100-L container that gravity-fed the three CPFs using float valves which maintained the water level at approximately 21 centimetres (cm) in each CPF. This kept each filter element almost completely full at a volume of 9 L. This maximized the flow rate over the time that the experiments were conducted. Once treated, the water from each CPF was collected in a calibrated container used to measure the total volume of treated water. Three constant head apparatuses were fabricated so that experiments could be conducted on all three water sources simultaneously.
Figure 1

Constant head apparatus.

Figure 1

Constant head apparatus.

Experiments were conducted for a total of 113 days. After the first 15 days of testing, the experiments were suspended for 30 days for logistical reasons, but subsequent analysis of the data did not indicate that the suspension period affected the experimental results. Therefore, all the data were subsequently considered in the study. According to van Halem et al. (2009) long-term clogging is not prevented by scrubbing the CPFs; therefore, the CPFs were not cleaned in order to avoid cross-contamination. However, at the end of the experiments, the CPFs were cleaned following the manufacturer's instructions and tested for 6 additional days to confirm that the lack of cleaning did not affect the study.

During the performance study the filter flow rate was measured daily using two different methods: (1) measuring the discharge in a graduated cylinder for a period of 1 hour (instantaneous flow rate); and (2) dividing the total volume of treated water by the number of hours (on average 23 hours) that had passed since the previous measurement, (daily average flow rate). The daily average flow rate was compared to the commonly used instantaneous flow rate in order to assess the flow-rate variability between measurements.

Influent and effluent water was tested 1 day after the source water tank was filled and the day before it was refilled (on average every 12 days) for the following parameters: turbidity (Hach 2100P Turbidimeter), hardness (Hach Hardness, Iron, and pH Test Kit HA-62), free chlorine (Hach Free and Total Chlorine Test Kit CN-70), temperature, pH, conductivity, and total dissolved solids (TDS) (YSI 556 Multiparameter System), and total and fecal coliforms (IDEXX Colilert Quantitray 2000). All the parameters were measured in duplicate and the data were averaged for analytical purposes. Influent water samples were taken from the ball valves installed upstream of the float valves and immediately analyzed. Effluent water samples for microbiological, chlorine, turbidity and hardness analyses were taken from the 5-gallon buckets and collected in disinfected containers used for the instantaneous flow-rate test. The maximum residence time of the sample in the receptacle was 1 hour. The rest of the tests were conducted using effluent water samples collected in the treated water containers with a maximum residence time of 24 hours.

Flow-rate and water-quality measurements from the three systems were then compared with the aim of assessing how the different water parameters impact flow-rate behavior. The data analysis was performed using the statistical software MINITAB 15.0.

RESULTS AND DISCUSSION

The statistical information that describes the flow-rate data of the three systems is summarized in Table 1. The average of the three CPFs associated with each set was used for each flow-rate measurement. It is important to note that the mean value of both SW and CW is below the lower limit value of the recommended flow-rate range (1 to 2 Lh−1) while the mean value of TW is almost at the upper limit. The higher standard deviation in TW reflects the higher variability with respect to the other systems.

Table 1

Statistical summary of flow-rate data

SW
CW
TW
VariableInst. flow rateDaily avg. flow rateInst. flow rateDaily avg. flow rateInst. flow rateDaily avg. flow rate
113 113 113 113 107 107 
Minimum (L/h) 0.40 0.38 0.35 0.33 1.40 1.25 
Median (L/h) 0.65 0.56 0.62 0.57 2.15 1.98 
Maximum (L/h) 1.76 1.67 1.76 1.82 3.18 2.96 
Mean (L/h) 0.74 0.68 0.72 0.68 2.12 2.01 
Std dev. (L/h) 0.35 0.32 0.32 0.30 0.36 0.37 
SW
CW
TW
VariableInst. flow rateDaily avg. flow rateInst. flow rateDaily avg. flow rateInst. flow rateDaily avg. flow rate
113 113 113 113 107 107 
Minimum (L/h) 0.40 0.38 0.35 0.33 1.40 1.25 
Median (L/h) 0.65 0.56 0.62 0.57 2.15 1.98 
Maximum (L/h) 1.76 1.67 1.76 1.82 3.18 2.96 
Mean (L/h) 0.74 0.68 0.72 0.68 2.12 2.01 
Std dev. (L/h) 0.35 0.32 0.32 0.30 0.36 0.37 

Instantaneous and daily average flow-rate data, as well as the differences between their paired measurements, were tested for normality and in all tests the null hypothesis that the data are normally distributed was rejected. Therefore, non-parametric test procedures for the statistical data analysis were employed. The Wilcoxon signed-rank test was used to assess flow-rate variability between measurements. This test allowed for the analysis of the difference between the paired observations resulting from the two different measurements (instantaneous and daily average flow-rate) and the determination of whether they came from the same population, as described in Helsel & Hirsch (1992). The test was conducted for all filters and in all cases the null hypothesis that the median difference between paired observations equals zero was rejected with p-values equal to 0.000. Therefore, the test shows that there is a statistically significant difference between the two methods and that flow-rate variability exists between measurements. The average percentage differences between instantaneous and daily average flow rate were 9.4% for SW, 7.9% for CW, and 5.5% for TW. In order to reduce the effect of this variability, the instantaneous flow-rate data were discarded and the daily average flow-rate measurements were used for further flow-rate analysis. Figure 2 shows the flow-rate behavior for the three systems by depicting the 10-day averaged flow rate versus time.
Figure 2

Flow-rate data from laboratory experiments.

Figure 2

Flow-rate data from laboratory experiments.

Inspection of the time series makes evident the similarity between SW and CW flow-rate behavior, while TW presents higher values. All three data sets showed a decreasing trend in flow rate with time. However, individual flow rates were not always lower than the immediately preceding flow rate.

The similarity between SW and CW and their difference with TW were tested using the non-parametric two-sample Wilcoxon rank sum test (Mann–Whitney test). The pair combinations of the three systems (SW-CW, SW-TW, and CW-TW) were tested to determine whether the two groups came from the same population. The null hypothesis was that the probability of having one group's flow rate greater than the other group was equal to 0.5 (same median). SW-CW yielded a p-value of 0.8951 (null hypothesis not rejected) while both SW-TW and CW-TW rejected the null hypothesis with a p-value of 0.0000. In addition, the difference in variance between the paired groups was tested using the non-parametric Levene's test, described by Ryan (2007), with the null hypothesis that the variances’ difference is equal to zero. Also in this case SW-CW did not reject the null hypothesis (p-value of 0.901) while both SW-TW and CW-TW rejected the null hypothesis with a p-value of 0.0000. Therefore, it can be stated that there is no statistically significant difference between SW and CW, while TW presented statistically different flow-rate behavior.

The water analyses were conducted before and after filtration to characterize treatment efficacy. The average and standard deviation are summarized in Table 2. Measured concentrations at or below detection limits were assumed to be one-half of the detection limit for the purpose of calculating average and standard deviation. If the calculated average value was below detection limits the standard deviation was not calculated. The Missouri Department of Natural Resources (MoDNR) (2015) states that no microbiological contaminants were detected in the calendar year of 2014 in the municipal tap water from Rolla, Missouri. Therefore, the microbiological analyses for TW were limited to a monthly presence-absence test to check for cross-contamination in the laboratory, which resulted negative for total and fecal coliforms. These results are not included in Table 2. The relatively high variability of coliform counts in both raw SW and CW shown in Table 2 is attributed to die-off of the microorganisms. However, coliform samples were collected from the influent and effluent at the same time so that treatment efficacy calculations were not impacted by the die-off phenomenon.

Table 2

Water-quality testing data

Raw SW
Treated SW
Water propertyAverageStd dev.AverageStd dev.
Turbidity (NTU) 3.26 2.15 0.33 0.09 
Hardness (gpg CaCO30.5 0.4 
Conductivity (mS/cm) 0.125 0.009 0.133 0.006 
TDS (g/L) 0.082 0.006 0.086 0.004 
pH 6.96 0.16 6.80 0.10 
Temperature (°C) 22.27 0.93 20.67 0.39 
Free chlorine (mg/L) ND(0.02) N/A ND(0.02) N/A 
DO (mg/L) 6.26 0.59 6.24 0.33 
Total coliform (MPN/100 mL) 457.2 723.6 2.1 5.1 
Fecal coliform (MPN/100 mL) 0.4 0.8 0.0 0.0 
 Raw CWTreated CW
Turbidity (NTU) 3.53 4.12 0.37 0.09 
Hardness (gpg CaCO316 0.8 17 1.1 
Conductivity (mS/cm) 0.513 0.030 0.501 0.028 
TDS (g/L) 0.333 0.019 0.325 0.019 
pH 6.80 0.23 7.12 0.17 
Temperature (°C) 21.98 0.35 20.64 0.36 
Free chlorine (mg/L) ND(0.02) N/A ND(0.02) N/A 
DO (mg/L) 4.26 1.74 6.07 0.18 
Total coliform (MPN/100 mL) 1,089.7 1,068.2 43.4 64.0 
Fecal coliform (MPN/100 mL) 315.8 532.1 0.6 1.4 
Raw TWTreated TW
Turbidity (NTU) 0.66 0.43 0.23 0.07 
Hardness (gpg CaCO316 1.4 17 1.0 
Conductivity (mS/cm) 0.502 0.016 0.485 0.052 
TDS (g/L) 0.326 0.011 0.323 0.012 
pH 6.81 0.40 7.02 0.09 
Temperature (°C) 21.96 0.37 20.85 0.28 
Free chlorine (mg/L) 0.28 0.18 ND(0.02) N/A 
DO (mg/L) 5.82 1.00 6.11 0.20 
Raw SW
Treated SW
Water propertyAverageStd dev.AverageStd dev.
Turbidity (NTU) 3.26 2.15 0.33 0.09 
Hardness (gpg CaCO30.5 0.4 
Conductivity (mS/cm) 0.125 0.009 0.133 0.006 
TDS (g/L) 0.082 0.006 0.086 0.004 
pH 6.96 0.16 6.80 0.10 
Temperature (°C) 22.27 0.93 20.67 0.39 
Free chlorine (mg/L) ND(0.02) N/A ND(0.02) N/A 
DO (mg/L) 6.26 0.59 6.24 0.33 
Total coliform (MPN/100 mL) 457.2 723.6 2.1 5.1 
Fecal coliform (MPN/100 mL) 0.4 0.8 0.0 0.0 
 Raw CWTreated CW
Turbidity (NTU) 3.53 4.12 0.37 0.09 
Hardness (gpg CaCO316 0.8 17 1.1 
Conductivity (mS/cm) 0.513 0.030 0.501 0.028 
TDS (g/L) 0.333 0.019 0.325 0.019 
pH 6.80 0.23 7.12 0.17 
Temperature (°C) 21.98 0.35 20.64 0.36 
Free chlorine (mg/L) ND(0.02) N/A ND(0.02) N/A 
DO (mg/L) 4.26 1.74 6.07 0.18 
Total coliform (MPN/100 mL) 1,089.7 1,068.2 43.4 64.0 
Fecal coliform (MPN/100 mL) 315.8 532.1 0.6 1.4 
Raw TWTreated TW
Turbidity (NTU) 0.66 0.43 0.23 0.07 
Hardness (gpg CaCO316 1.4 17 1.0 
Conductivity (mS/cm) 0.502 0.016 0.485 0.052 
TDS (g/L) 0.326 0.011 0.323 0.012 
pH 6.81 0.40 7.02 0.09 
Temperature (°C) 21.96 0.37 20.85 0.28 
Free chlorine (mg/L) 0.28 0.18 ND(0.02) N/A 
DO (mg/L) 5.82 1.00 6.11 0.20 

The TW turbidity values were significantly lower than the turbidity values measured for the raw water from the other two sources. SW and CW showed very similar levels in both influent and effluent water. According to Sawyer et al. (2003) turbidity may be caused by a wide variety of suspended materials that range in size from colloidal to coarse and include both organic and inorganic substances. Turbidity removal was significant in all the systems: in the TW effluent it was lower than the regulatory standard limit set by the US Environmental Protection Agency (EPA) (1999) while in SW and CW it was just slightly higher. These results are consistent with the flow-rate behavior of the three systems analyzed above demonstrating that turbidity is an indicator that suspended particles do impact CPF flow rate.

The microbiological analysis showed a higher coliform concentration, especially fecal, in CW than in SW, which reflects a higher biological activity in CW. This is confirmed by the low concentration of dissolved oxygen (DO) found in CW. According to EPA (2012) wastewater from sewage treatment plants, such as that used to create CW, often contains organic materials that are decomposed by microorganisms. This causes an increase in the biochemical oxygen demand (BOD) and a decrease in the concentration of dissolved oxygen. Natural organic matter (NOM) is described by Crittenden et al. (2012) as a complex matrix of organic chemicals originating from a water body due to biological activity, including secretion from the metabolic activity of microorganisms and algae. It was hypothesized that the higher biological activity in CW would result in a higher potential for biological fouling and a corresponding decrease in the flow rate. However, since there is no significant difference between the measured SW and CW flow rates, that hypothesis does not appear valid.

Percentage differences between influent and effluent water in hardness, conductivity, and TDS were not significant in any of the systems. Similar levels were reported in CW and TW, while SW presented lower values, as expected. According to Sawyer et al. (2003) hardness is caused by multivalent cations capable of reacting with anion and precipitate. These results are not consistent with the flow-rate behavior, demonstrating that inorganic fouling is not responsible for the change in flow rate.

No free chlorine was detected in the influent water of SW and CW while TW showed an average concentration of 0.28 parts per million (ppm). Free chlorine concentration in effluent water was below detection limits in all the systems. According to EPA (2013), the maximum residual disinfectant level for chlorine is 4 ppm, 14 times higher than the maximum detected concentration. Therefore, the measured free chlorine concentrations were considered too low to impact the behavior of the CPFs. Temperature and pH measurements were consistent between the influent and effluent for all three water sources. In general the results described above agree with the performance study and rehabilitation experiments described in van Halem et al. (2007, 2009).

In order to compare the flow-rate trends of the systems, flow rate was plotted versus the total volume of treated water in Figure 3. All three data sets showed that early time flow-rates increase, and after a stabilization period all three flow-rates begin decreasing with an almost linear trend. The flow-rate growth during the first phase is consistent with other flow-rate observations documented by Lantagne et al. (2010), Hubbel & Elmore (2012), Hubbel et al. (2015), and others. According to Lantagne et al. (2010), this initial increase could be due to the washing of combustible material trapped in the CPF during the production process. The total volume of treated water was 1,759 L in SW, 1,758 L in CW and 4,961 L in TW. All the filters passed the quality test at the production location, and thus should have an initial clear-water flow rate in the range of 1 to 2 Lh−1. This is true for all the systems, but the graph shows a change in flow-rate behavior after the first approximately 400 L have been treated. The vertical line in Figure 3 depicts the separation between this first phase when the flow rate increases, and the second phase when the flow rate stabilizes prior to gradually declining.
Figure 3

CPF flow rate trend analysis.

Figure 3

CPF flow rate trend analysis.

The equations of the linear trend of the second-phase data were calculated to allow for comparison of the slopes. SW and CW showed a similar slope significantly steeper than TW.

SW and CW failed, in terms of water production, after the first 400 L since the flow rate dropped below the lower limit of 1 Lh−1 during the second phase. The TW flow rate was above the manufacturer's specification of 2 Lh−1 during the first 3,000 L, but below the upper limit of 3.15 Lh−1 published by CMWG (2011). Using the CMWG criterion, the TW result can be considered valid and the lifetime can be estimated. The lifetime in terms of water production capacity was calculated adding the volume of water treated during the first phase (400 L) and the second phase. The latter was calculated using the trend line slope with acceptable flow rates inside the range of 1–3.15 Lh−1. The result for TW was a total volume of treated water equal to 8,000 L before it reached the expiration flow rate. Similar estimates could not be calculated for CW and SW because their flow rates were too low.

CONCLUSIONS

Considering the results of both the flow-rate experiments and water quality analysis, it can be concluded that turbidity seems to be the principal indicator in characterizing CPF lifetime in terms of quantity of treated water. Indeed, the results of the study show how an increase in water turbidity impacts both the average flow rate and the rate at which it decreases. Estimates of CPF lifetime with water having different turbidity levels showed that the total volume of treated water before failure can range between 400 L and 8,000 L. However, further research is required to identify the relationship between turbidity and flow rate, and to characterize the suspended particles responsible for clogging.

There is no evidence that biological activity also contributes to premature failure of CPFs and the data did not indicate that chemical precipitation is responsible for the filter clogging.

Flow rate is a powerful indicator of CPF performance since it readily provides information about water production capacity and performance of the filter in terms of removal efficacy, but its behavior has to be better understood. This study indicates that the results from the initial flow-rate tests that are commonly conducted by manufacturers as quality tests using clear water could be unrepresentative of the average flow rate during the lifetime of the CPF.

ACKNOWLEDGEMENTS

The authors are grateful to the management and staff of the CPF factory in Antigua, Guatemala, which provided the CPFs for this project. The Missouri S&T Geological Engineering program provided financial support for the importation of the CPFs.

REFERENCES

REFERENCES
Ceramics Manufacturing Working Group
.
2011
Best Practice Recommendations for Local Manufacturing of Ceramic Pot Filters for Household Water Treatment
.
CDC
,
Atlanta, GA
,
USA
.
Crittenden
J. C.
Trussell
R. R.
Hand
D. W.
Howe
K. J.
Tchobanoglous
G.
2012
MWH's Water Treatment: Principles and Design
.
John Wiley & Sons
,
Hoboken, NJ
,
USA
.
Elmore
A. C.
Fahrenholtz
W. G.
Glauber
L. G.
Sperber
A. N.
2011
Calculation of ceramic pot filter hydraulic conductivity using falling-head data
.
Water Science and Technology: Water Supply
11
(
3
),
358
363
.
Helsel
D. R.
Hirsch
R. M.
1992
Statistical Methods in Water Resources
.
Elsevier Inc.
,
New York, NY
,
USA
.
Hubbel
L. A.
Elmore
A. C.
2012
Quantification of the lifetime of ceramic pot filters
. In:
World Environmental and Water Resources Congress 2012: Crossing Boundaries
,
Loucks
E. D.
(ed.).
ASCE, Reston, VA, USA
, pp.
900
910
.
Hubbel
L.
Elmore
A. C.
Reidmeyer
M. R.
2015
Comparison of a native clay soil and an engineered clay used in experimental ceramic pot filter fabrication
.
Water Science and Technology: Water Supply
15
(
3
),
569
577
.
Kallman
E. N.
Oyanedel-Craver
V. A.
Smith
J. A.
2011
Ceramic filters impregnated with silver nanoparticles for point-of-use water treatment in rural Guatemala
.
Journal of Environmental Engineering
137
(
6
),
407
415
.
Lantagne
D.
Klarman
M.
Mayer
A.
Preston
K.
Napotnik
J.
Jellison
K.
2010
Effect of production variables on microbiological removal in locally-produced ceramic filters for household water treatment
.
International Journal of Environmental Health Research
20
(
3
),
171
187
.
Lantagne
D. S.
2001
Investigation of the Potters for Peace Colloidal Silver Impregnated Ceramic Filter Report 1: Intrinsic Effectiveness
.
Alethia Environmental
,
Allston, MA
,
USA
.
Missouri Department of Natural Resources
.
2015
Rolla 2014 Annual Water Quality Report
.
http://dnr.mo.gov/ccr/MO3010700.pdf (accessed 22 June 2015)
.
Oyanedel-Craver
V. A.
Smith
J. A.
2008
Sustainable colloidal-silver-impregnated ceramic filter for point-of-use water treatment
.
Environmental Science and Technology
42
(
3
),
927
933
.
Plappally
A. K.
Yakub
I.
Brown
L. C.
Soboyejo
W. O.
Soboyejo
A. B. O.
2009
Theoretical and experimental investigation of water flow through porous ceramic clay composite water filter
.
Fluid Dynamics & Materials Processing
5
(
4
),
373
398
.
Ryan
T. P.
2007
Modern Engineering Statistics
.
John Wiley & Sons
,
Hoboken, NJ
,
USA
.
Sawyer
C. N.
McCarty
P. L.
Parkin
G. F.
2003
Chemistry for Environmental Engineering and Science
.
McGraw-Hill Education
,
New York, NY
,
USA
.
Schweitzer
R. W.
Cunningham
J. A.
Mihelcic
J. R.
2013
Hydraulic modeling of clay ceramic water filters for point-of-use water treatment
.
Environmental Science and Technology
47
(
1
),
429
435
.
United Nations Children's Fund
.
2008
UNICEF Handbook on Water Quality
.
UNICEF
,
New York, NY
,
USA
.
United Nations World Water Assessment Programme
.
2015
The United Nations World Water Development Report 2015: Water for a Sustainable World
.
UNESCO
,
Paris
,
France
.
US Environmental Protection Agency
1999
Guidance Manual for Compliance with the Interim Enhanced Surface Water Treatment Rule: Turbidity Provisions
. .
US Environmental Protection Agency
2012
Water: Monitoring & Assessment
. .
US Environmental Protection Agency
2013
Water: Basic Information about Regulated Drinking Water Contaminants
. .
Van Halem
D.
Heijman
S. G. J.
Soppe
A. I. A.
van Dijk
J. C.
Amy
G. L.
2007
Ceramic silver-impregnated pot filters for household drinking water treatment in developing countries: material characterization and performance study
.
Water Science and Technology: Water Supply
7
(
5–6
),
9
17
.
Van Halem
D.
van der Laan
H.
Heijman
S. G. J.
van Dijk
J. C.
Amy
G. L.
2009
Assessing the sustainability of the silver-impregnated ceramic pot filter for low-cost household drinking water treatment
.
Physics and Chemistry of the Earth, Parts A/B/C
34
(
1–2
),
36
42
.
World Health Organization & United Nations Children's Fund
.
2014
Progress on Sanitation and Drinking-Water – 2014 Update
.
Joint Monitoring Programme for Water Supply and Sanitation
,
WHO
,
Geneva
,
Switzerland
.
Yakub
I.
Plappally
A.
Leftwich
M.
Malatesta
K.
Friedman
K. C.
Obwoya
S.
Nyongesa
F.
Maiga
A. H.
Soboyejo
A. B. O.
Logothetis
S.
Soboyejo
W.
2013
Porosity, flow, and filtration characteristics of frustum-shaped ceramic water filters
.
Journal of Environmental Engineering
139
(
7
),
986
994
.