ArticlePDF Available

Performance Evaluation of Gravity-Fed Water Treatment Systems in Rural Honduras: Verifying Robust Reduction of Turbidity and Escherichia coli during Wet and Dry Weather

Authors:
  • St. Joseph’s College of Maine

Abstract and Figures

This is the first study to document the reduction of turbidity and Escherichia coli throughout the processes of full-scale gravity-fed drinking water plants (GFWTPs) and their downstream distribution systems in rural Honduras. The GFWTPs, which in these cases were designed by AguaClara, use standard treatment processes: coagulation, sedimentation, filtration, and chlorination. During the dry season, we measured E. coli, turbidity, and chlorine residual at five GFWTPs with < 1,000 connections and at three alternative piped-water systems in neighboring communities. Samples were evaluated from the raw water, settled water, filtered water, post-chlorination in the distribution tank, and at a distant-piped household connection. During the dry season, the treated water and household connections serviced by the GFWTPs met World Health Organization (WHO) recommendations for E. coli (< 1 most probable number [MPN]/100 mL). Alternative plants with the same water sources had comparable or higher E. coli and turbidity measurements posttreatment. We examined performance robustness of two GFWTPs during the transition into the rainy season. The turbidity of the filtered water met WHO recommendations (< 1 nephelometric turbidity units). Escherichia coli was not detected in treated water, indicating that the two GFWTPs can consistently remove particulates and E. coli from source waters containing varying levels of turbidity. During two sampling events during the rainy season, E. coli was detected at the household connection of a GFWTP system with intermittent service and a substandard chlorine residual (geometric mean = 1.0 MPN/100 mL). Strategies to avoid contamination or inactivate E. coli in the distribution system are needed to ensure safe drinking water at the points of delivery, especially for systems with intermittent service.
Content may be subject to copyright.
Am. J. Trop. Med. Hyg., 99(4), 2018, pp. 881888
doi:10.4269/ajtmh.17-0577
Copyright © 2018 by The American Society of Tropical Medicine and Hygiene
Performance Evaluation of Gravity-Fed Water Treatment Systems in Rural Honduras:
Verifying Robust Reduction of Turbidity and Escherichia coli during Wet and Dry Weather
Yolanda M. Brooks,
1
* Erika A. Tenorio-Moncada,
2
Nisarg Gohil,
1
Yuqi Yu,
1
Mynor R. Estrada-Mendez,
2
Geovany Bardales,
2
and Ruth E. Richardson
1
1
School of Civil and Environmental Engineering, Cornell University, Ithaca, New York;
2
Department of Environment and Development, Panamerican
Agriculture University, Zamorano, Yeguare Valley, Municipality of San Antonio de Oriente, Francisco Moraz´an, Honduras
Abstract. This is the rst study to document the reduction of turbidity and Escherichia coli throughout the processes of
full-scale gravity-fed drinking water plants (GFWTPs) and their downstream distribution systems in rural Honduras. The
GFWTPs, which in these cases were designed by AguaClara, use standard treatment processes: coagulation, sedi-
mentation, ltration, and chlorination. During the dry season, we measured E. coli, turbidity, and chlorine residual at ve
GFWTPs with < 1,000 connections and at three alternative piped-water systems in neighboring communities. Samples
were evaluated from the raw water, settled water, ltered water, post-chlorination in the distribution tank, and at a distant-
piped household connection. During the dry season, the treated water and household connections serviced by the
GFWTPs met World Health Organization (WHO) recommendations for E. coli (< 1 most probable number [MPN]/100 mL).
Alternative plants with the same water sources had comparable or higher E. coli and turbidity measurements post-
treatment. We examined the performance robustness of two GFWTPs during the transition into the rainy season. The
turbidity of the ltered water met WHO recommendations (< 1 nephelometric turbidity units). Escherichia coli was not
detected in treated water, indicating that the two GFWTPs can consistently remove particulates and E. coli from source
waters containing varying levels of turbidity. During two sampling events during the rainy season, E. coli was detected at
the household connection of a GFWTP system with intermittent service and a substandard chlorine residual (geometric
mean = 1.0 MPN/100 mL). Strategies to avoid contamination or inactivate E. coli in the distribution system are needed to
ensure safe drinking water at the points of delivery, especially for systems with intermittent service.
INTRODUCTION
Access to safe drinking water is essential for basic human
health. Since 2000, the United Nations Millennium Develop-
ment Goals and Sustainable Management Goals have
spearheaded efforts to increase access of safe drinking
water. In 2015, 75% of the global population, 4.2 billion
people, had a piped water supply that was on premises.
1
However, instances of fecal contamination in piped systems
in low- and medium-income countries have been documented.
24
Consumption of polluted water is a major contributor to the
prevalence of diarrheal diseases.
5
Treatment of water supplies is necessary to remove partic-
ulates, including microorganisms of public concern, in piped
water supplies.
6,7
In high-resource settings, water treatment is
commonly accomplished through energy-intensive treatments.
Processes typically present in these plants include in the order
of operation: coagulation, occulation, sedimentation, ltration,
and disinfection. Energy-intensive treatment units are not fea-
sible in low-resource settings where electricity is expensive,
inconsistent, or simply not available. In theory and practice,
gravity can replace electrical requirements for metering, mixing,
and moving water through the drinking water treatment units.
However, there is little research evaluating the performance
of centralized gravity-powered water treatment systems to
remove particulates and microorganisms (e.g., fecal indicators
such as Escherichia coli) within the treatment units.
In 2010, 86% of the Honduran population had piped water
on premises.
8
Piped supplies in Honduras have been shown
to contain waterborne pathogens such as Giardia spp. and
Cryptosporidium spp.
9
There were an estimated 4.6 diarrheal
incidences per child-year in Honduran children < 5 years old
due in part to contaminated drinking water.
10
In an observational
cohort of hospital and clinic visits reported to the Honduran
Secretary of Health during 20002004, there was an estimated
1:123 risk of diarrheal deaths in children < 5 years old.
11
AguaClara is a collaboration between Cornell University,
Ithaca, NY, and Agua Para el Pueblo, Tegucigalpa, Honduras, a
Honduran nongovernmental organization (NGO). AguaClara
designs centralized water treatment plants for small towns in
rural Honduras. The plants include full-scale water treatment
processes such as occulation, oc blankets, plate settlers,
stacked rapid sand ltration, and chlorination. The treatment
processes and their downstream distribution systems are
powered by gravity and provide treated water to household
connections without relying on electricity. As of 2017, GFWTPs
designed by AguaClara provide drinking water to 65,000 people
in Honduras and India. Laboratory investigations of the treatment
processes in the AguaClara plants have led to the invention of
high-rate sedimentation tanks with oc blankets and ability to
self-clean without moving parts
12
as well as stacked rapid sand
lters that can be backwashed without pumps.
13
We are not aware of any studies that have investigated the
reduction of turbidity and fecal indicator bacteria of GFWTPs
during normal operation. Such investigations can provide a
benchmark for improving the understanding of the capabilities
and limitations of centralized, gravity-powered water treat-
ment plants. In addition, documentation of the performance of
GFWTPs will help accelerate adoption of these plants in small
communities within high-income countries, which often
struggle to afford highly mechanized treatment plants. In this
study, we investigated microbial, chemical, and physical in-
dicators of water quality along the treatment train and in
household connections of ve GFWTPs in rural Honduras. The
rst objective of our study was to evaluate and compare the
reduction of culturable E. coli and turbidity, and concentrations
* Address correspondence to Yolanda M. Brooks, Department of Food
Science, Cornell University, Stocking Hall Rm 204, Ithaca NY 14850.
E-mail: yb86@cornell.edu
881
of chlorine residual (storage tank and distribution system only)
along the treatment trains during dry weather of ve GFWTP
systems and three nearby drinking water systemstwo with
chlorination-only and one multistage ltration plant. The sec-
ond objective was to evaluate the concentrations of culturable
E. coli, turbidity, and free chlorine in the treatment processes
and delivered water of two of the GFWTPs during the transition
from the dry season to the rainy season, when the water
quality of the source water can degrade drastically.
METHODS
Site descriptions. During January 2016, we visited ve
GFWTPs designed by AguaClara in ve communities in
southern and western Honduras (labeled AC-1 through 5;
Figure 1). Their construction was overseen by a Honduran
NGO, Agua Para el Pueblo. Honduras has regulatory stan-
dards regarding the quality of treated water.
14
The water
supplies of all of the plants are surface waters such as springs
or rivers that are delivered to the plants via gravity conduction
lines. The Swiss Development Cooperation funded the con-
struction of all of the plants except AC-3, which was funded via
private donations. Information regarding the population ser-
viced, year built, and production volumes of all plants are listed
in Figure 1. AC-1 through 5 had the following processes: grit
removal, occulation, oc blankets, lamellar sedimentation,
stacked rapid sand ltration (except AC-3), and chlorination
and storage in a distribution tank before entering the distri-
bution system that delivered the treated water to household
premises (Table 1 and Supplemental Information S1). The
GFWTPs are operated, maintained, and governed by a water
board comprising elected citizens within the distribution
system and do not have Water Safety Plans modeled after
World Health Organization (WHO) recommendations. At each
GFWTP, turbidity measurements are routinely taken at least
two times per day in the raw water, settled water, and, where
applicable, ltered water. Each treatment plant records the
calculations of coagulant used and target chlorine dosage
applied to the ltered water.
We visited drinking water treatment systems (B-6 through 8)
with varying levels of treatment that had the same raw water
supply as one of the GFWTPs (Figure 1). These were a multi-
stage ltration plant (B-6) and two chlorination-only plants
(B-7 and B-8). We were unable to determine the funding
mechanism for the construction of B-7 and B-8. Construction
of B-6 was funded by the Swiss Development Cooperation.
B-6 has three successive ltration steps (two gravel steps
using dynamic and roughing ltration followed by a slow sand
lter) and equipment for chlorine gas disinfection. B-6 was
not chlorinating the nished water during our visit. The al-
ternative treatment plants are governed by a local board and
do not have Water Safety Plans modeled after WHO
recommendations.
Description of sampling events and sampling points. In
base ow conditions during January 2016, we visited ve
GFWTPs (AC-1 through 5) and three alternative treatment
plants (B-6 through 8). At each GFWTP, we sampled from ve
points within the AC GFWTPs and their downstream distribu-
tion systems: raw water after grit removal, settled water, ltered
water (exceptAC-3), distribution tank,and at one of the furthest
household connections from the distribution tank. At B-6, we
collected water samples from the raw water and from each of
the three successive ltration steps. At the chlorination-only
treatments (B-7 and B-8), we collected samples from the raw
water and at one of the furthest household connections from
the distribution tank. For samples from the tap, water was run
for at least 5 minutes before the sample was taken. Within each
sampling point, we collected duplicate samples of water.
During MayJuly 2016, we visited the two GFWTPs, AC-4
and AC-5, during six sampling events: three sampling events
occurred in the dry season during May 2016 (dry weather) and
three sampling events occurred during the onset of the rainy
season during JuneJuly 2016 (wet weather) < 24 hours after a
rain event that occurred at or in the vicinity of the sources of the
raw water. At each GFWTP, we took samples from ve sam-
pling points previously listed in the GFWTPs. During the
MayJuly 2016 sampling campaign, water was collected from
the same household connection in each town. Within each
sampling point, water samples were collected in triplicate.
Measurements of physical and chemical qualities and
concentrations of E. coli in water samples. During the
January 2016 and MayJuly 2016 sampling campaigns, for
FIGURE 1. Map of the locations of the gravity-fed drinking water treatment plants (AC-1 through 5) and the alternative treatment systems (B-6
through 8) that were evaluated in this study. Table 1 contains a description of each plant.
882 BROOKS AND OTHERS
chemical parameters we collected 50 mL from the sampling
points listed previously. During January 2016, we measured
turbidity using a MicroTPI Field Portable Turbidimeter (HF
Scientic, Fort Myers, FL) with a detection limit of 0.01 NTU.
During January 2016, the concentrations of free chlorine
(mg/L-Cl
2
) from the distribution tank and at the household
connection were determined using a free chlorine color
disc test (Hach, Inc., Loveland, CO). During MayJuly 2016,
turbidity of the treatment processes contained in the water
treatment plant (raw water after grit removal, sedimented
water, and ltered water) were measured by the plant operator
with a MicroTPI Field Portable Turbidimeter (HF Scientic),
whereas free chlorine and turbidity at the distribution tank and
at the household connection were measured with a digital
colorimeter DR-890 (Hach, Inc.; detection limits: 0.01 mg/L
and 1 NTU, respectively). The lower limits of detection of the
free chlorine measurements during the January 2016 were
0.02 mg/L. Values below the detection limits were reported as
less than the detection limit and evaluated at one-half of the
detection limit in subsequent analyses and graphs. During the
MayJuly 2016 sampling event, the turbidity measurements of
the raw water were measured by the plant operators within
24 hours of the sampling event during dry weather in AC-4 and
AC-5. Recent (within 30 days) turbidity measurements are
available online for the raw water, sedimented water, and l-
tered water of the GFWTPs at http://aguaclara.github.io/
settings.html).
Replicate 100-mL samples (duplicates in January 2016 and
triplicates in MayJuly 2016) were analyzed for concentrations
of E. coli in water samples using the manufacturersinstruc-
tions of the compartment bag test (CBT; AguaGenX, LLC,
Chapel Hill, NC).
15
Specically, samples of water were asep-
tically collected using 100-mL Whirl-Pak
®
Thio Bags
®
(Nasco,
Fort Atkinson, WI) and stored in a cooler on ice for £6 hours.
During the January 2016 sampling campaign, the CBT assays
were incubated at ambient temperatures, 2530°C for 4048
hours (time varied based on incubation temperatures as per
the manufacturers instructions), whereas the CBT assays
performed during the MayJuly 2016 sampling campaign
were incubated at 35°C for 24 hours.
16,17
The quantication
range of each CBT assay is 1100 MPN E. coli/100 mL. During
the MayJuly 2016 sampling campaign, one replicate of the
raw water was diluted 10× with bottled water and another was
diluted 100× to ensure at least one replicate was within the
quantication range. Negative controls were bottled water
and were processed every day of water collection. All negative
controls were negative for the presence of E. coli. For all
samples, we reported the geometric mean concentrations of
E. coli calculated from the replicates of each sample. During
the January 2016 and the MayJuly 2016 sampling cam-
paigns, there were distinct lower limits of detection of the
concentrations of E. coli in a water sample: 0.5 and 0.3 MPN/
100 mL, respectively, because of the distinct number of rep-
licates per sample (two versus three, respectively). Assays
that did not have detectable E. coli were reported as less than
the lower detection limit and evaluated at one-half of the de-
tection limit in subsequent analyses and graphs. Replicates of
each assay that were above the upper limit of detection, 100
MPN/100 mL, were reported as > 100 MPN/100 mL in sub-
sequent analyses and graphs.
Statistical analyses. All analyses were performed in IBM
SPSS Statistics for Windows, Version 23.0 (IBM, Corp.,
Armonk, NY) and signicance was considered at α= 0.05.
Within the data gathered from the evaluation of the transition
of the rainy to dry season over six sampling events in AC-4 and
AC-5, MannWhitney tests analyzed the differences between
the concentrations of E. coli and turbidity measured at the raw
water during the dry and wet weather sampling events.
RESULTS
Turbidity reduction in gravity-fed water treatment sys-
tems and alternative systems. Turbidity was measured at
ve sampling locations within each of the GFWTPS (only four
sampling locations in AC-3) in the raw water, settled water,
ltered water, in the distribution tank, and at one of the furthest
household connections. All GFWTPs showed a decrease in
turbidity from raw water to treated water, although the extent
of the reduction varied. At AC-1, AC-2, AC-4, and AC-5, the
turbidity decreased from the raw water (1.6, 2.8, 7.4, and 4.5
NTU, respectively) to the ltered water (0.3, 1.3, 0.2, and 0.02
NTU, respectively). Turbidity levels at a distant household
connection were 0.7, 2.2, 0.9, and 0.6 NTU, respectively
(Figure 2A). There was no ltration process at AC-3, and the
turbidity decreased from the raw water (3.5 NTU) to the settled
water (2.2 NTU), followed by a slight increase measured at the
household connection (2.5 NTU; Figure 2A).
Turbidity was also measured in the alternative treatment
systems. In B-7 and B-8, the piped water systems only in-
cluded chlorination and were sampled in the raw water and at
one of the furthest household connections. In B-8, there was a
slight increase in the turbidity measured from the raw water
(7.8 NTU) to household connection (8.3 NTU; Figure 2B). In
B-7, the turbidity decreased from the raw water (11 NTU) to
the household connection (5.0 NTU; Figure 2B). B-6 was a
multistage ltration system where turbidity was measured at
the raw water and after each of the three ltration steps.
At B-6, the turbidity oscillated between the raw water and
TABLE 1
Description of key treatment processes and features of each plant
Plant Flocculation and sedimentation Filtration Chlorination Population served Production volume (L/s) Year built Location
AC-1 Y Y Y 6000 32 2014 San Nicolas, Santa Barbara
AC-2 Y Y Y 5000 20 2015 Jesus de Otoro, Intibuca
AC-3 Y N N* 2000 6 2009 Cuatro Comunidades, Francisco Morazan
AC-4 Y Y Y 4500 16 2016 Morocel´
ı, El Para´
ıso
AC-5 Y Y Y 3800 14 2016 San Mat´
ıas, El Para´
ıso
B-6 N YN* 1600 25 2000 Jesus de Otoro, Intibuca
B-7 N N Y 180 3 1982 Suyate, El Para´
ıso
B-8 N N Y 145 2 1995 San Marcos Abajo, El Para´
ıso
* Chlorination at AC-3 was inoperable on the sampling day.
B-6 had three successive ltration steps, two gravel steps and a nal rapid sand lter.
TURBIDITY AND E.COLI DECREASES IN GRAVITY-FED WATER PLANTS 883
after the three consecutive ltration steps (3.711.3 NTU;
Figure 2B).
In addition, the turbidity measured at the household con-
nection of AC-5 were > 10-fold less (0.7 NTU) than the house-
hold connection of the corresponding alternative treatment
plant (B-8; 8.3 NTU). Comparing the turbidity of the house-
hold connection in AC-4 and B-7, AC-4 had less turbid water
(2.2 versus 5.0 NTU).
Free chlorine measurements in the distribution tank
and at the household connections of GFWTPs and
alternative systems. Free chlorine was measured in the
distribution tank and at one of the furthest household connec-
tions of the ve gravity-fed water treatment systems. Free
chlorine was also measured at one of the furthest household
connection of the two chlorination-only systems, B-7 and
B-8. All GFWTPs except AC-3 had detectable residual chlorine
in the distribution tanks and at the household connections
(Table 2). In AC-2, residual chlorine increased from the distri-
bution tank to the household connection (Table 2). Chlorinaton
was ofine on the sampling date of AC-3. Because a single
parcel of water was not followed from the plant to the tap, it is
possible that variability in dosing caused slight differences in
chlorine between the distribution tank and household tap. The
free residual chlorine at the household connections of the al-
ternative systems (B-7 and B-8) are listed in Table 2 and were
comparable with the range of concentrations at the household
connections of the GFWTP systems.
Reduction of E. coli throughout the treatment processes
and distribution systems of the gravity-fed water treat-
ment systems and alternative treatment systems. In both
the alternative treatment systems and the GFWTP systems,
concentrations of E. coli were measured in the same sampling
points as the turbidity measurements. At each sampling point,
the concentrations of E. coli were measured in duplicate and
the geometric mean was reported. In all AC GFWTPs, there
was an overall reduction of the mean concentrations of E. coli
from raw water to the distribution tank (Figure 3A). At least a
10-fold decrease of the geometric mean concentrations of
E. coli between raw water and settled water were measured in
AC-2 through 5, where the concentrations of E. coli in the raw
and the settled water were both above the upper detection
limit for CBT (> 100 MPN/100 mL; Figure 3A). There was de-
tectable E. coli (³0.5 MPN/100 mL) in the ltered water before
chlorination in AC-4. In all GFWTPs, there was no detectable
E. coli (< 0.5 MPN/100 mL) in the distribution tank and at the
household connections serviced, indicating that AC1 through
5 had the following reduction efciencies: > 99.5, > 99.0,
> 97.6, > 97.8, and > 91.5%, respectively.
The geometric mean concentrations of E. coli measured in
the raw water of B-6 was 13.6 MPN/100 mL and oscillated
with successive ltration steps (between 5.4 and 48.3 MPN/
100 mL) with water leaving the nal ltration step with 8.0
MPN/100 mL (Figure 3B), representing 41.2% overall re-
duction. Escherichia coli was detected at the household
FIGURE 2. Comparison of turbidity measurements (NTU) from the treatment train processes and at a distant household connection from (A) from
ve gravity-fed water treatment plants (AC-1 through 5)
a
; and (B) from three alternative treatments (B6, B-7, and B-8 )
b,c
during the dry season. Each
data point represents one measurement. Display of the data using an arithmetic scale is displayed in Supporting Information Figure S2.
a
AC-3 did
not have a ltration process.
b
Chlorination was the only treatment for B-7 and B-8.
c
B-6 included three consecutive ltration processes. Chlorination
was not functioning during sampling and we did not collect a sample at a household connection or in the distribution tank.
TABLE 2
Free residual chlorine (mg/L) measured in the stored and chlorinated water and at a distant household connection of ve gravity-fed water treatment
plants (AC-1 through AC-5) and three alternative treatment plants (B-6 through B-8) during dry weather in January 2016
Mean free residual chlorine concentration, mg/L
Gravity-fed water treatment plants
Alternative treatment
plants
AC-1 AC-2 AC-3* AC-4 AC-5 B-6* B-7 B-8
Distribution tank 0.9 0.14 0.010.27 0.2 NDNDND
Household
connection
1.3 0.4 0.010.25 0.1 ND0.23 0.05
Measurements within the World Health Organ ization recommendations (0.52mg/Lfreechlorineinthedistributiontankand0.22 mg/L at the household) are represented in bold.
* Chlorination was ofine at AC-3 and B-6 during the sampling event.
The lower detection limit was 0.02 mg/L. Measurements below the detection limit were reported at one-half of the lower detection limit.
ND = no data collected. We were unable to collect samples from the distribution tank in B-6, B-7, and B-8, and at a distant household connection for B-6.
884 BROOKS AND OTHERS
connection of B-8 (geometric mean = 1.5 MPN/100 mL)
representing 89.0% reduction, whereas E. coli was not de-
tected at the household connection in B-7 (> 96.3% re-
duction of E. coli).
Transition to the rainy season: reduction of turbidity
throughout the processes and distribution systems of
two GFWTPs. During six sampling events spanning MayJuly
2016, turbidity was measured at ve sampling points in AC-4
and AC-5: raw water, settled water, ltered water, in the dis-
tribution tank, and at one of the furthest household connec-
tions. In AC-4 and AC-5, the mean turbidities of the raw water
in wet weather (and dry weather) were 77.3 (1.7) and 45.3 NTU
(1.3 NTU), respectively (Figure 4AB). The turbidities of the raw
water during wet weather in AC-4 and AC-5 were not signi-
cantly larger than during dry weather (P= 0.10). In both plants,
the mean turbidity consistently decreased from the raw water
to the household connections during dry and wet weather
conditions (Figure 4AB).
Transition to the rainy season: free residual chlorine
measurements in the distribution tank and at the house-
hold connections of two gravity-fed water systems. During
six sampling events spanning MayJuly 2016, free chlorine
was measured in the distribution tank and at one of the furthest
household connections of the AC-4 and AC-5 systems. The
mean target dosages of chlorine applied to ltered water were
calculated by the plant operators and their ranges are listed in
Table 3. Independent of weather conditions, the mean con-
centrations of free chlorine in the distribution tank and at
the household connection were larger in AC-4 than in AC-5.
During dry and wet weather in both plants, the mean con-
centrations of free chlorine in the distribution tank were larger
than at the household connection (Table 3).
Transition to the rainy season: reduction of E. coli
throughout the treatment processes and in the distribu-
tion systems of two gravity-fed water treatment systems.
During each of the six sampling events, concentrations of
FIGURE 3. Comparison of the geometric mean concentrations of E. coli (MPN/100 ml) from the treatment train processes and at a distant
household connection (A)fromve gravity-fed water treatment plants (AC-1 through 5)
a
;and(B) from three alternative treatments (B6, B-7, and B-8)
b,c
during the dry season. Data points that are outside the quantication range (0.5-100 MPN/100 ml) are represented at one half the lower detection
limit with the label BD (below detection) or at the upper detection limit. Each data point represents the geometric mean of two replicates.
a
AC-3 did
not have a ltration process.
b
Chlorination was the only treatment for B-7 and B-8.
c
B-6 included three consecutive ltration processes. Chlorination
was not functioning during sampling and we did not collect a sample at a household connection.
FIGURE 4. The mean turbidity measured in the treatment processes and downstream distribution systems of two gravity-fed water treatment
plants, AC-4 (A) and AC-5 (B) during three dry weather and three wet weather sampling events between MayJuly 2016. Data points where all
measurements were below the lower detection limit (0.01 NTU for raw water, settled water and ltered water; and 1 NTU for distribution plant and at
the household connection) were represented at one-half of the detection limit with the title BD (below detection). The error bars represent one
standard error.
TURBIDITY AND E.COLI DECREASES IN GRAVITY-FED WATER PLANTS 885
E. coli were measured in each of the sampling points alongside
turbidity. The geometric mean concentrations of E. coli in the
raw water of AC-4 (and AC-5) during dry and wet weather were
16.5 and 718 MPN/100 mL, respectively (7.4 and 1,231 MPN/
100 mL, respectively; Figure 5AB). The concentrations of
E. coli in the raw water in AC-4 and AC-5 were not signicantly
larger during wet weather than dry weather (P= 0.10). During
wet weather, the largest decrease of the geometric mean
concentrations of E. coli in the plants in AC-4 and AC-5 oc-
curred between raw water and settled water (718 to 20.0
MPN/100 mL and 1,231 to 11.6 MPN/100 mL, respectively;
Figure 5AB). After subsequent treatment processes (sedi-
mentation and ltration) in AC-4 and AC-5, there were similar
concentrations of E. coli during wet and dry weather. In wet
weather in AC-4 and AC-5, reduction efciencies between
raw water and the distribution tank were consistently > 96%.
During dry and wet weather in AC-4 (and dry weather in AC-5),
E. coli was not detected (< 0.3 MPN/100 mL) at either the
distribution tank or the household connection (Figure 5AB).
However, E. coli was detected at the household connection
during two wet weather sampling events in AC-5 and had a
geometric mean concentration of 1.0 MPN/100 mL across
the three wet weather sampling events.
DISCUSSION
To our knowledge, this is the rst peer-reviewed study to
evaluate the reduction of E. coli and turbidity during the
treatment trains of GFWTPs. During dry weather in January
2016, we visited eight treatment plants, ve GFWTPs (AC-1
through 5), and three alternative plants (B-6 through 8), in rural
Honduras. Within the GFWTPs, there was an overall reduction
of turbidity from the raw water to the household connection.
Reduction throughout the treatment train was dependent
on the plant, with AC-3 and A-4 having the smallest reduction
in turbidity from raw water to the household connection
(Figure 2A). The absence of a ltration unit could explain the
lower reduction efciency in AC-3. In AC-4, the muted re-
duction of turbidity may be due to insufcient coagulant added
to the occulation tanks or a decreased residence time in the
sedimentation tanks. Compared with their companion
GFWTP systems, the alternative systems had higher mea-
surements and lower free residual chlorine at the points of
delivery (except B-7 which had comparable chlorine levels
with AC-4 near 0.2 mg/L). Also, E. coli was detected at a point
of delivery in B-8. The results from a singular sampling event
during dry weather indicated that the drinking water at one of
the furthest household connections from the gravity-fed
treatment systems was of higher quality than the evaluated
alternative systems. Because of variables such as the age of
treatment plant and its downstream distribution system, hy-
draulic conditions, variation in water quality conditions, and
size of the distribution systems, we are unable to conclude
that improved water quality of the gravity-fed water treatment
systems was due solely to the design of the GFWTP systems.
We also evaluated the performance of two GFWTPs, AC4
and AC-5, during a multi-week sampling campaign during the
transition from dry to wet seasons during MayJuly 2016. The
FIGURE 5. Geometric mean concentrations of E.coli in the treatment processes and downstream distribution systems of two gravity-fed water
treatment plants, AC-4 (A) and AC-5 (B) during three dry weather and three wet weather sampling events between MayJuly 2016. Data points where
all measurements were below the lower detection limit, 0.3 MPN/100 mL, were represented at one-half of the detection limit with the title BD(below
detection). Each data point represents the geometric mean of three sampling dates. The error bars represent one standard error.
TABLE 3
The range of target chlorine concentrations and mean concentrations of free chlorine (mg/L) measured in the distribution tank and at a household
connection of two gravity-fed drinking water treatment systems (AC-4 and AC-5) sampled in wet and dry weather during MayJuly 2016
Mean free residual chlorine concentration, mg/L ± std. error
AC-4 AC-5
Dry weather Dry weather Dry weather Wet weather
Target chlorine concentration (range
measured across all sampling dates)
1.011.04 0.71.0
Distribution tank 0.43 ± 0.08 0.51 ± 0.08 0.23 ± 0.15 0.51 ± 0.08
Household connection 0.36 ± 0.12 0.27 ± 0.09 0.02 ± 0.02 0.27 ± 0.09
The lower detection limit was 0.02 mg/L.
886 BROOKS AND OTHERS
wet weather sampling took place during the rst major rain
event of the 2016 rainy season when overland storm runoff
washes ground-deposited feces and particles into the surface
water supplies (rst-usheffect). During wet weather, the
ranges of turbidity in the raw water of AC-4 and AC-5 were
10.04271.40 NTU and 0.75314.70 NTU, respectively. Al-
though turbidity and E. coli levels increased in the source wa-
ters after the onset of the rainy season (Figures 4ABand
5AB), the water in the distribution tanks consistently met the
WHO recommendations for E. coli (< 1 MPN/100 mL; Table 4),
indicating that the GFWTPs robustly removed fecal pollution in
the raw water. The microbial quality of the water was consis-
tently maintained at the point of delivery in AC-4 (Table 4,
Figure 5A). Despite the high quality of produced water even
during the rainy season, E. coli was detected at a household
connection in AC-5 during two sampling events (geometric
mean concentrations of 2.1 and 3.2 MPN/100 mL during each
sampling event; Table 4, Figure 5B). The contamination found
during two sampling events suggest that at the time of sam-
pling, there were fecal intrusion(s) in the distribution system and
insufcient residual chlorine to inactivate microorganisms of
public concern in the water of the distribution system.
It is worth noting that the distribution system in AC-5 pro-
vided intermittent service during MayJuly 2016. Intermittent
service may have contributed to E. coli contamination at the
point of delivery during two wet weather sampling events.
Comparably, a systemic review determined that improved
water supplies had signicantly larger loads of fecal contam-
ination during rainy seasons compared with dry seasons.
18
In
addition, there were more occurrences of E. coli contamina-
tion and signicantly larger concentrations of E. coli in distri-
bution systems receiving intermittent supplies shortly after a
rain event than during dry weather in Hubli and Dharwad,
Karnataka, India.
19
These results reinforce the need for routine
monitoring of E. coli at points of delivery and increasing
chlorine residual of intermittent water supplies to inactivate
fecal intrusion within the distributions systems.
There are various pathways that could be responsible for
the E. coli contamination observed in the distribution system in
AC-5. A global review of intermittent water supplies de-
termined that corrosion and depressurizing/repressurizing
cycles put excess strain on the pipes and suggest that large
volumes of intrusions may enter the pipes during extended
periods of low pressure.
20
In addition, backush and low
pressure were identied as mechanisms that introduced fecal
contamination in distribution systems in Hubli and Dharwad
Karnataka, India.
21
A study of an intermittent water system
in Arraij ´an, Panama, suggested that possible fecal intrusions
were attenuated by longer supply durations, increased
pressure in the distribution system, and a higher chlorine
residiual.
22
Therefore, increasing the service duration, water
pressure, and/or chlorine residual in the distribution system of
AC-5 could improve the quality of the intermittent water
supply at the points of delivery and decrease associated
health risks.
Overall, the largest reduction in turbidity and concentrations
of E. coli in two GFWTPs during wet weather occurred be-
tween raw water and the settled water (Figures 4AB and
5AB), indicating that occulation, oc blankets, and sedi-
mentation are responsible for most of the reduction of
particulate matter, including suspended microorganisms.
Throughout the sampling events, the turbidity measured after
ltration in AC-4 and AC-5 (except one wet weather sampling
event in AC-4) were < 1 NTU, which met the WHO recom-
mendations before disinfection.
23
Compared with the large
water system in Arraij ´an, Panama, servicing 263,000 people,
overall the points of delivery of the two rural GFWTPs in
Honduras had comparable turbidity and concentrations of
residual chlorine (AC-4 only in dry weather).
22
Our study indicates that the GFWTPs designed by Agua-
Clara and built/operated by local labor produce high-quality
drinking water and their treatment processes are robust
against the increases in turbidity and fecal pollution in raw
water that accompany rain events. However, intrusions in the
distribution system can erode the accomplishments of the
treatment processes. The data generated in our study provide
a strong recommendation to initiate routine monitoring of the
gravity-fed water treatment systems in rural Honduras. Mon-
itoring schemes should include operational monitoring of
E. coli, free residual chlorine, and turbidity throughout the
treatment processes and the distribution system. Although
turbidity and chlorine residual tests are already carried out
locally at the GFWTPs, conventional efforts to monitor E. coli
concentrations require high costs including transportation to a
centralized laboratory and equipment and labor to perform the
method. The data collected in our study demonstrated that
routinely measuring E. coli is possible in low-resource settings
using CBT and the data generated by the CBT agree with
TABLE 4
Comparison of the performance of two gravity-fed drinking water treatment systems (AC-4 and AC-5) during wet and dry weather to the WHO
recommendations for turbidity, Escherichia coli, and residual chlorine of treated drinking water
Parameter
Mean in the gravity-fed water treatment systems accross
six sampling dates (range)
WHO recommendationsAC-4 AC-5
Turbidity (NTU) Before disinfection 1.17 (< 0.016.00) 0.5 (< 0.013.00) < 1 with full treatment*
Distribution tank 2.17 (< 0.0111.00) 1.17 (< 0.015.00) NA
At the household connection 0.7 (< 0.012) < 0.01 NA
E. coli concentrations (MPN/100 mL) Distribution tank < 0.3 < 0.3 < 1
At the household connection < 0.3 0.4(< 0.33.12) < 1
Residual chlorine (mg/L) Distribution tank 0.47 (0.340.6) 0.29 (0.030.52) 0.5
At the household connection 0.30 (0.100.57) 0.08 (< 0.020.19) 0.22
Source This study This study WHO (2010)
WHO = World Health Organization.
* Or < 5 NTU before disinfection in systems with limited or no treatment available.
Not applicable.
Geometric mean.
§ After 30 minutes of contact.
TURBIDITY AND E.COLI DECREASES IN GRAVITY-FED WATER PLANTS 887
conventional E. coli enumeration methods such as Colilert and
membrane ltration.
15,17,24
Training plant operators or water
board members to routinely monitor water safety will allow the
identication of fecal intrusions, ensuring residual disinfection
throughout the distribution system, and conrming the micro-
bial and physicochemical quality of the water at the points of
delivery. This training will also help move resource-limited
communities closer to consistently meeting the UNsSus-
tainable Management Goal 6.1 of universal access to safe
and affordable drinking water.
1
Received July 19, 2017. Accepted for publication May 22, 2018.
Published online August 8, 2018.
Note: Supplemental information appears at www.ajtmh.org.
Acknowledgments: We wouldlike to thank Walker Grimshaw, Jonathan
Christensen, Aminta Nuñez, Jac ´obo Nuñez, and Antonio Elvir at Agua
Para el Pueblo for logistical support; AnaReuss and Evan Greenberg for
help with sample collection; Monroe Weber-Shirk for help organizing
communication; Lourdes Espinal for support in the laboratory.
Financial support: This research was funded by Cornell Universitys
David R. Atkinson Center for a Sustainable Future (ACSF).
Authorsaddresses: Yolanda M. Brooks, Nisarg Gohil, Yuqi Yu, and
Ruth E. Richardson, Cornell University, Civil and Environmental Engi-
neering, Ithaca, NY, E-mails: yb86@cornell.edu, nisarg@genotekbio.
com, yuqiyu802@gmail.com, and rer26@cornell.edu. Erika A. Tenorio-
Moncada, Mynor R. Estrada-Mendez, and Geovany Bardales, Escuela
Agricola Panamericana Zamorano, Water Resources, Municipality of
San Antonio de Oriente, Francisco Moraz´an, Honduras, Central Amer-
ica, E-mails: etenorio@zamorano.edu, mynoresmen@gmail.com
and gbardales2018@gmail.com.
This is an open-access article distributed under the terms of the
Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the
original author and source are credited.
REFERENCES
1. WHO/UNICEF Joint Monitoring Programme on Water Supply and
Sanitation, 2017. Safely Managed Drinking Water. Geneva,
Switzerland: World Health Organization. Available at: https://
www.wssinfo.org/leadmin/user_upload/resources/JMP-SMDW-
TR-1-March-2017.pdf. Accessed March 7, 2017.
2. Bain R, CronkR, Wright J, Yang H, Slaymaker T, Bartram J, 2014.
Fecal contamination of drinking-water in low- and middle-
income countries: a systematic review and meta-analysis.
PLoS Med 11: e1001644.
3. Heitzinger K, Rocha CA, Quick RE, MontanoSM, Tilley DH, Mock
CN, Carrasco AJ, Cabrera RM, Hawes SE, 2015. Improved
but not necessarily safe: an assessment of fecal contamination
of household drinking water in rural Peru. Am J Trop Med Hyg
93: 501508.
4. Baum R, Kayser G, Stauber C, Sobsey M, 2014. Assessing the
microbial quality of improved drinking water sources: results
from the Dominican Republic. Am J Trop Med Hyg 90: 121123.
5. Ashbolt NJ, 2004. Microbial contamination of drinking water and
disease outcomes in developing regions. Toxicology 198:
229238.
6. Betancourt WQ, Rose JB, 2004. Drinking water treatment
processes for removal of Cryptosporidium and Giardia.Vet
Parasitol 126: 219234.
7. Geldreich EE, 1996. Microbial Quality of Water Supply in Distri-
bution Systems. Boca Raton, FL: CRC Press LLC.
8. WHO/UNICEF Joint Monitoring Programme on Water Supply and
Sanitation, 2013. Progress on Sanitation and Drinking Water
2013 Update. Geneva, Switzerland: World Health Organization,
140. Available at: http://apps.who.int/iris/bitstream/10665/
81245/1/9789241505390_eng.pdf?ua=1. Accessed October
1, 2004.
9. Solo-Gabriele HM, LeRoy Ager A Jr., Fitzgerald Lindo J, Dub ´on
JM, Neumeister SM, Baum MK, Palmer CJ, 1998. Occurrence
of Cryptosporidium oocysts and Giardia cysts in water supplies
of San Pedro Sula, Honduras. Rev Panam Salud Publica 4:
398400.
10. Fischer Walker CL, Perin J, Aryee MJ, Boschi-Pinto C, Black RE,
2012. Diarrhea incidence in low- and middle-income countries
in 1990 and 2010: a systematic review. BMC Public Health 12:
220.
11. Sol ´orzano Gir ´onJO, Molina IB, Turcios-Ruiz RM, Quiroz Mejia CE,
Amendola LM, de Oliveira LH, Andrus JK, Stupp PW, Bresee
JS, Glass RI, 2006. Burden of diarrhea among children in
Honduras, 20002004: estimates of the role of rotavirus. Rev
Panam Salud Publica 20: 377384.
12. Garland C, Weber-Shirk M, Lion LW, 2016. Inuence of variable
inlet jet velocity on failure modes of a oc blanket in a water
treatment process train. Environ Eng Sci 33: 7987.
13. Adelman MJ, Weber-Shirk ML, Will JC, Cordero AN, Maher WJ,
Lion LW, 2013. Novel uidic control system for stacked rapid
sand lters. J Environ Eng 139: 939946.
14. Rep´ublica de Honduras Ministerio de Salud, 1995. Norma Tecnica
Nacional Para la Calidad Del Agua Potable. Available at: http://
www.ersaps.hn/documentos/normativa/Norma Tecnica cal-
idad agua.pdf. Accessed February 26, 2018.
15. Stauber C, Miller C, Cantrell B, Kroell K, 2014. Evaluation of the
compartment bag test for the detection of Escherichia coli in
water. J Microbiol Methods 99: 6670.
16. AquaGenX, LLC, 2017. Compartment Bag Test (CBT) for E. coli In-
structions for Use: Drinking Water. Raleigh, NC: AquaGenX, LLC.
Available at: https://www.aquagenx.com/wp-content/uploads/
2013/09/-v6-CBT-Ecoli-Instructions-_DrinkingWater_20170920.
pdf. Accessed March 20, 2018.
17. WangA, McMahan L, Rutstein S, Stauber C, Reyes J, Sobsey MD,
2017. Household microbial water quality testing in a peruvian
demographic and health survey: evaluation of the compartment
bag test for Escherichia coli.Am J Trop Med Hyg 96: 970975.
18. KostylaC, Bain R, Cronk R, Bartram J, 2015. Seasonal variation of
fecal contamination in drinking water sources in developing
countries: a systematic review. Sci Total Environ 514: 333343.
19. Kumpel E, Nelson KL, 2013. Comparing microbial water quality in
an intermittent and continuous piped water supply. Water Res
47: 51765188.
20. Kumpel E, Nelson KL, 2016. Intermittent water supply: preva-
lence, practice, and microbial water quality. Environ Sci Tech-
nol 50: 542553.
21. Kumpel E, Nelson KL, 2014. Mechanisms affecting water quality
in an intermittent piped water supply. Environ Sci Technol 48:
27662775.
22. Erickson JJ, Smith CD, Goodridge A, Nelson KL, 2017. Water
quality effects of intermittent water supply in Arraij ´an, Panama.
Water Res 114: 338350.
23. World Health Organization, 2011. Guidelines for Drinking-Water Qual-
ity, 4th edition. Geneva, Switzerland: WHO Press. Availabl e at: http:/ /
whqlibdoc.who.int/publications/2011/9789241548151_eng.pdf.
Accessed January 25, 2017.
24. Brooks YM, Collins SM, Mbullo P, Boateng GO, Young SL,
Richardson RE, 2017. Evaluating human sensory perceptions
and the compartment bag test assays as proxies for the pres-
ence and concentration of Escherichia coli in drinking water
in western Kenya. Am J Trop Med Hyg 97: 10051008.
888 BROOKS AND OTHERS
... Community water management stands at the confluence of community initiatives and environmental management. While a lack of social capital can hinder community water and sanitation initiatives , higher social capital has been associated with group cooperation (Tantoh and McKay, 2020) and improved outcomes (Bowles and Gintis, 2002;Kähkönen, 1999, 2002;Narayan, 1995;Watson et al., 1997). 1 Social capital has also been linked to the diffusion of water-related behaviors . ...
... The plants are built in water systems that, before the plants, provided untreated piped water to household taps. Researchers at Cornell University create hydraulic engineering designs with the criteria to treat water to international standards of quality for turbidity, free chlorine residual concentration, and microbial quality (Brooks et al., 2018). They function by gravity, without electricity, and without mechanical components. ...
... One option is to cancel the project. Although cancellation could be supported based on the critical role of social capital for sustainability (Bowles and Gintis, 2002;Kelly et al., 2017) and to avoid the sunk cost fallacy, it would be fraught in practice. Canceling a project motivated by humanitarian concern, that involves a public good and in which stakeholders have invested financially and reputationally can be challenged on ethical grounds and resisted by stakeholders. ...
Article
Since social capital permeated the conceptualization of “community” in the early 1990s, evidence has mounted as to its significance for water and sanitation projects. But understanding how to apply knowledge of social capital towards sustainable development has lagged far behind. To address this gap, we combined insights from household surveys, semi-structured interviews, performance evaluations, and extensive field observations into comparative case narratives of six community-managed drinking water treatment plants built through an NGO initative in Honduras. The aim was to not only measure social capital but to use those measurements as a framework to understand how social capital affected each project. Our findings suggest that social capital amplified or hindered plant performance through three key mediating mechanisms: Resource mobilization; resilience to change and conflict; and participation and communication. Plant performance, in turn, influenced community social capital by affecting the relationship between the NGO and community and community residents' trust in the project. These mechanisms constitute a feedback loop by which relative levels of social capital and water system operation are mutually reinforcing. Drawing on our results, we propose four accessible proxies and four initial strategies practitioners can use to operationalize social capital scholarship and identify and disrupt negative patterns during the project cycle. We encourage additional translational research to leverage the ever-growing knowledge of social capital into improved outcomes in practice.
... The AguaClara treatment system incorporated coagulation, flocculation, and settling tanks to account for the variable turbidity of incoming source water. 20 Several other evaluations reviewed here coupled chlorinators with other forms of treatment such as gravity-driven membrane filtration systems 21 or natural (gravel and sand) filtration upstream of chlorination. 22,23 The capital cost of commercially available passive chlorinators ranged from $3 to $662, with an average cost of $140 (values reported in 2021 USD, adjusted for inflation). ...
... 21,38,39 For studies evaluating existing chlorinators installed prior to the evaluation, the study period was often not more than a few weeks to a few months. 20,28,29 The results from our NGO survey suggest that longer-term evaluations could be conducted for many chlorinators that have already been deployed and are currently in use in resource-constrained communities. ...
Article
Full-text available
The world is not on track to meet Sustainable Development Goal 6.1 to provide universal access to safely managed drinking water by 2030. Removal of priority microbial contaminants by disinfection is one aspect of ensuring water is safely managed. Passive chlorination (also called in-line chlorination) represents one approach to disinfecting drinking water before or at the point of collection (POC), without requiring daily user input or electricity. In contrast to manual household chlorination methods typically implemented at the point of use (POU), passive chlorinators can reduce the user burden for chlorine dosing and enable treatment at scales ranging from communities to small municipalities. In this review, we synthesized evidence from 27 evaluations of passive chlorinators (in 19 articles, 3 NGO reports, and 5 theses) conducted across 16 countries in communities, schools, health care facilities, and refugee camps. Of the 27 passive chlorinators we identified, the majority (22/27) were solid tablet or granular chlorine dosers, and the remaining devices were liquid chlorine dosers. We identified the following research priorities to address existing barriers to scaled deployment of passive chlorinators: (i) strengthening local chlorine supply chains through decentralized liquid chlorine production, (ii) validating context-specific business models and financial sustainability, (iii) leveraging remote monitoring and sensing tools to monitor real-time chlorine levels and potential system failures, and (iv) designing handpump-compatible passive chlorinators to serve the many communities reliant on handpumps as a primary drinking water source. We also propose a set of reporting indicators for future studies to facilitate standardized evaluations of the technical performance and financial sustainability of passive chlorinators. In addition, we discuss the limitations of chlorine-based disinfection and recognize the importance of addressing chemical contamination in drinking water supplies. Passive chlorinators deployed and managed at-scale have the potential to elevate the quality of existing accessible and available water services to meet "safely managed" requirements.
... Recognizing that conventional drinking water treatment utilized in larger urban contexts may not be appropriate for rural or low-income communities, more suitable alternatives have been and continue to be developed and tested (Skinner, 2001;Montgomery and Elimelech, 2007;Smyth, 2011;Pickering et al., 2015Pickering et al., , 2019Amin et al., 2016;Brooks et al., 2018) providing more options to communities, government agencies, and other organizations to better serve a wider range of contexts (Hunter et al., 2010a;González Rivas et al., 2014;Mihelcic et al., 2017). Technologies are usually classified based on the scale or level at which the water is treated, including centralized or community-level, point-of-entry, and point-of-use (POU) systems, each presenting advantages and disadvantages in terms of treatment efficiency as well as the social dimensions to ensure proper and sustained use, maintenance, and monitoring. ...
... As such, this system can enhance both safety and sustainability of rural water systems by easing operation and reducing potential hurdles to disinfection (Arnold and Colford, 2007;Kearns et al., 2015). In addition, the GAC filters provide protection against other organic contaminants potentially present in the water, including disinfection by-products (DBPs) and pesticides, among others (Hopman et al., 1994;Lantagne, 2001;Schwarzenbach et al., 2006;Kim and Kang, 2008;Jusoh et al., 2014;Kearns et al., 2014;Benstöm et al., 2017). ...
Article
Full-text available
Intermittent drinking water supply is common in low- and middle-income countries throughout the world and can cause water quality to degrade in the distribution system. In this study, we characterized water quality in one study zone with continuous supply and three zones with intermittent supply in the drinking water distribution network in Arraiján, Panama. Low or zero pressures occurred in all zones, and negative pressures occurred in the continuous and two of the intermittent zones. Despite hydraulic conditions that created risks for backflow and contaminant intrusion, only four of 423 (0.9%) grab samples collected at random times were positive for total coliform bacteria and only one was positive for E. coli. Only nine of 496 (1.8%) samples had turbidity >1.0 NTU and all samples had ≥0.2 mg/L free chlorine residual. In contrast, water quality was often degraded during the first-flush period (when supply first returned after an outage). Still, routine and first-flush water quality under intermittent supply was much better in Arraiján than that reported in a previous study conducted in India. Better water quality in Arraiján could be due to better water quality leaving the treatment plant, shorter supply outages, higher supply pressures, a more consistent and higher chlorine residual, and fewer contaminant sources near pipes. The results illustrate that intermittent supply and its effects on water quality can vary greatly between and within distribution networks. The study also demonstrated that monitoring techniques designed specifically for intermittent supply, such as continuous pressure monitoring and sampling the first flush, can detect water quality threats and degradation that would not likely be detected with conventional monitoring.
Article
Full-text available
Intermittent water supplies (IWS), in which water is provided through pipes for only limited durations, serve at least 300 million people around the world. However, providing water intermittently can compromise water quality in the distribution system. In IWS systems, the pipes do not supply water for periods of time, supply periods are shortened, and pipes experience regular flow re-starting and draining. These unique behaviors affect distribution system water quality in ways that are different than during normal operations in continuous water supplies (CWS). A better understanding of the influence of IWS on mechanisms causing contamination can help lead to incremental steps that protect water quality and minimize health risks. This review examines the status and nature of IWS practices throughout the world, the evidence of the effect of IWS on water quality, and how the typical contexts in which IWS systems often exist - low-income countries with under-resourced utilities and inadequate sanitation infrastructure - can exacerbate mechanisms causing contamination. We then highlight knowledge gaps for further research to improve our understanding of water quality in IWS.
Article
Full-text available
Background: Access to safe drinking-water is a fundamental requirement for good health and is also a human right. Global access to safe drinking-water is monitored by WHO and UNICEF using as an indicator "use of an improved source," which does not account for water quality measurements. Our objectives were to determine whether water from "improved" sources is less likely to contain fecal contamination than "unimproved" sources and to assess the extent to which contamination varies by source type and setting. Methods and findings: Studies in Chinese, English, French, Portuguese, and Spanish were identified from online databases, including PubMed and Web of Science, and grey literature. Studies in low- and middle-income countries published between 1990 and August 2013 that assessed drinking-water for the presence of Escherichia coli or thermotolerant coliforms (TTC) were included provided they associated results with a particular source type. In total 319 studies were included, reporting on 96,737 water samples. The odds of contamination within a given study were considerably lower for "improved" sources than "unimproved" sources (odds ratio [OR] = 0.15 [0.10-0.21], I2 = 80.3% [72.9-85.6]). However over a quarter of samples from improved sources contained fecal contamination in 38% of 191 studies. Water sources in low-income countries (OR = 2.37 [1.52-3.71]; p<0.001) and rural areas (OR = 2.37 [1.47-3.81] p<0.001) were more likely to be contaminated. Studies rarely reported stored water quality or sanitary risks and few achieved robust random selection. Safety may be overestimated due to infrequent water sampling and deterioration in quality prior to consumption. Conclusion: Access to an "improved source" provides a measure of sanitary protection but does not ensure water is free of fecal contamination nor is it consistent between source types or settings. International estimates therefore greatly overstate use of safe drinking-water and do not fully reflect disparities in access. An enhanced monitoring strategy would combine indicators of sanitary protection with measures of water quality.
Article
We compared the concentrations of Escherichia coli quantified with Colilert(™) and the compartment bag test (CBT) in source water and household stored drinking water (SDW) in 35 households in western Kenya. We also investigated the associations of the perceptions of organoleptic properties and overall quality with ≥ 1 MPN/100 mL E. coli in SDW. Participants who rated the taste or smell of their SDW "< 5" on a 1 = "poor" to 5 = "excellent" Likert scale were 8.71 or 7.04 times more likely, respectively, to have ≥ 1 MPN/100 mL E. coli. Organoleptic properties are innate, albeit imperfect, indicators of fecal pollution in water. Within their shared quantification range, concentrations of E. coli enumerated with Colilert and CBT were similar and had a significant correlation coefficient, 0.896 (95% confidence interval = 0.691-1.101). The methods had moderate agreement within the World Health Organization's health risk levels (Cohen's Kappa coefficient = 0.640). In low-resource settings, CBT provides comparable assessments of E. coli concentrations to Colilert.
Article
The Joint Monitoring Program relies on household surveys to classify access to improved water sources instead of measuring microbiological quality. The aim of this research was to pilot a novel test for Escherichia coli quantification of household drinking water in the 2011 Demographic and Health Survey (DHS) in Peru. In the Compartment Bag Test (CBT), a 100-mL water sample is supplemented with chromogenic medium to support the growth of E. coli, poured into a bag with compartments, and incubated. A color change indicates E. coli growth, and the concentration of E. coli/100 mL is estimated as a most probable number. Triplicate water samples from 704 households were collected; one sample was analyzed in the field using the CBT, another replicate sample using the CBT was analyzed by reference laboratories, and one sample using membrane filtration (MF) was analyzed by reference laboratories. There were no statistically significant differences in E. coli concentrations between the field and laboratory CBT results, or when compared with MF results. These results suggest that the CBT for E. coli is an effective method to quantify fecal bacteria in household drinking water. The CBT can be incorporated into DHS and other national household surveys as a direct measure of drinking water safety based on microbial quality to better document access to safe drinking water. © 2017 by The American Society of Tropical Medicine and Hygiene.
Article
Improving designs of facilities that employ floc blankets in a water treatment process train to ensure stable performance is desirable, as is an understanding of suitable operating conditions to maintain a functional floc blanket in the system. By considering sequential processes when choosing design parameters, the whole system can be optimized to produce high quality effluent at low cost. A lab scale water treatment system with flocculator, floc blanket, and lamellar sedimentation was used to evaluate the effect of energy dissipation rates (EDR) in the inlet jet to the floc blanket on performance of the system as a whole. Results show that presence of a floc blanket provided an additional factor of 8 decrease in settled water suspended solids concentration at an upflow velocity 1.2 mm/s. Inlet jet EDR did not impact system performance until ∼300 mW/kg after which settled water turbidity increased. At the lower end of inlet jet EDR tested, the jet was unable to resuspend settled flocs. Given that plant performance was acceptable at higher inlet jet EDR, smaller inlet jets with a higher velocity could be used to ensure resuspension of flocs for continuous hydraulic cleaning.
Article
The indicator used to measure progress toward the Millennium Development Goal (MDG) for water is access to an improved water supply. However, improved supplies are frequently fecally contaminated in developing countries. We examined factors associated with Escherichia coli contamination of improved water supplies in rural Pisco province, Peru. A random sample of 207 households with at least one child less than 5 years old was surveyed, and water samples from the source and storage container were tested for E. coli contamination. Although over 90% of households used an improved water source, 47% of source and 43% of stored water samples were contaminated with E. coli. Pouring or using a spigot to obtain water from the storage container instead of dipping a hand or object was associated with decreased risk of contamination of stored water (adjusted prevalence ratio [aPR] = 0.58, 95% confidence interval [CI] = 0.42, 0.80). Container cleanliness (aPR = 0.67, 95% CI = 0.45, 1.00) and correct handwashing technique (aPR = 0.62, 95% CI = 0.42, 0.90) were also associated with decreased contamination risk. These findings highlighted the limitations of improved water supplies as an indicator of safe water access. To ensure water safety in the home, household water treatment and improved hygiene, water handling, and storage practices should be promoted. © The American Society of Tropical Medicine and Hygiene.
Article
Infrastructure for water treatment faces numerous challenges around the world, including the high failure rate of digital, electronic, pneumatic, and mechanical control systems due to their large number of components and their dependency on proprietary parts for repair. The development of more efficient, reliable, easily repaired water treatment controls that rely on simple fluidics rather than on complex systems has the potential to significantly improve the reliability of drinking water treatment plants, particularly for cities and towns in developing countries. A stacked rapid sand filter (SRSF) has been proposed as a more robust and sustainable alternative to conventional rapid sand filters because each filter can backwash at the same flow rate used for filtration without requiring pumps or storage tanks. While the concept of this filter has been demonstrated in previous studies, this paper presents a novel control system for the SRSF based on fluidics that eliminates the need for mechanized controls. The water level in the filter is regulated by a siphon pipe, which conveys flow during backwash and which contains an air trap to block flow during filtration. The state of the siphon pipe and the ensuing state of the filter are controlled by one small-diameter air valve. This fluidic control system was tested in pilot-scale experiments, which demonstrated its ability to set the mode of operation of the filter and served as the basis for the derivation of design equations. In addition, the first full-size SRSF was built at a municipal water plant in Honduras using this fluidic control system, which provided a full-scale demonstration of its effectiveness. This simple and robust control system shows promise as part of a sustainable rapid sand filtration process. (C) 2013 American Society of Civil Engineers.