Final Report

Point-of-Entry (POE) Cistern Water Purification Units (CPU) Development

Project Year: Apr. 1, 2009 ~ Feb. 28, 2010

Submitted by:
Sangchul Hwang, PhD
Associate Professor

Department of Civil Engineering and Surveying

University of Puerto Rico at Mayaguez

Submitted to:

US Virgin Islands Water Resources Research Institute

University of the Virgin Islands

February 28, 2010

Table of Contents
List of Tables

Listof Figures

List of Photos

Problem Statement

Objectives of Research

Materials

Lab-scale POE CPU

Setup and Operation

Backwash

Analysis

Results and Discussion

Physiochemical Water Quality

Filtration and Disinfection

Backwash

12

Biological Water Quality

13

CTValues on Microorganisms Removal

14

Student Training

15

Seminar at the USVIs

16

Future Studies

16

Results Dissemination

16

Acknowledgments

16

References

17

List of Tables
Table 1. Characteristics of rainwater
Table 2. Filtration media used for the sand filtration unit

5
7

Table 3. Backwash duration and intensity

12

Table 4. Physiochemical water quality parameters of backwash

12

List of Figures
Figure 1. Schematics of POE CPU

Figure 2. Trend of pH and conductivity during the lab-scale POE CPU run

10

Figure 3. Trend of turbidity and residual chlorine during the lab-scale POE CPU run

11

Figure 4. Results of fecal coliforms analysis

13
Figure 5. Effect of CT values on log removal of microorganisms in the influent amended with chicken
manure

14

Figure 6. Effect of CTvalues on log removal of microorganisms in the effluent from the sand filter

15

List of Photos
Photo 1. A view of the installed lab-scale POE CPU

Photo 2. Sand bed expansion during the backwash (left: before the backwash; right: after the backwash).
12

Problem Statement
Nearly 1.8 million people die every year due to diarrhoeal diseases (including cholera) in
developing countries and 90% of which are children under five years. Approximately 88% of these
diseases are attributed to unsafe water supply, inadequate sanitation, and poor hygiene practices. In

fact, about 1.1 billion people worldwide do not have access to drinking water from improved sources,
where as 2.6 billion people lack basic sanitation. With an improved water supply the diarrhea morbidity
rate is reduced by 6% and with an improved sanitation by 32%. In the Caribbean, 77% of the urban and
37% of the rural population don't have access to an improved drinking water, whereas 20% and 44% of

urban and rural population, respectively, have access to an improved sanitation (WHO/UNICEF, 2006).
Desalination supplies about 80% of the water used in the United States Virgin Islands (USVI).
Despite this fact, greater than 50% of the residences in the USVI use rainwater cistern (Brin et al., 2003).
Annual rainfall averages 40 inches in USVI and is very seasonal so that desalinated water is often

transported to refill the cisterns. From that point, there is a potential for improved water supplies to get
exposed to secondary contamination from pumps, pipes and the trucks, resulting in causing infectious
diseases and other enteric illness (Sobsey et al., 2008).

In addition, there might be contamination of rainwater before collected to the cisterns due to
urban pollution, bird and reptile waste materials and particulate matter deposited (Chang et al., 2004;
Evans et al., 2006; Simmons et al., 2001). Also, if the cisterns are old and dirty due to poor maintenance,

there is a potential of contamination of the stored cistern water. Another case of cistern water
contamination can be due to aged cistern structure, thereby infiltration of contaminants from the
surrounding contaminated soil and groundwater to the cisterns (Isquith and Winters, 1987).

Objectives of Research
The technology to be used for the provision of water services to communities and areas in
specific characteristics is always challenging. When water purification technology similar to what is used
in large urban areas is chosen, it will lead to high investments, which cannot be afforded by the majority
of small remote communities. It is important to choose technologies that can easily be operated and
maintained by local service providers or by community people.
This research aims to develop and evaluate a POE CPU which can be implemented in

communities of the USVIs. To meet this end, low-cost, simple, POE sand filtration and disinfection unit

was tested in a lab-scale. POE CPU to be developed in this proposed study will benefit many
communities which rely on rainwater as their drinking water source (e.g., USVIs). POE CPU could be used
by residents voluntarily with the expectation that the units can provide a protective measure if
contaminations are present due to dirty rainwater or mismanaged cisterns. The POE CPU can also be
used in a reactive mode, where the residents could take action in response to suspected or confirmed

cistern water contaminations. In other aspects, the results of the proposed study not only strengthen
the Pi's research thrust on small drinking water system evaluation and development, but also
significantly contribute to capacity building for USVIs through seminars and technology transfers in the
future.

The proposed research was proposed with a budget of $19,750 during the period of April 1,
2009 to Feb 28, 2010. It encompassed lab-scale research, student training, results dissemination, and a
seminar.

Materials
Three different sizes of sand were purchased from the Standard Sand & Silica, Co.. The sands
were manufactured specifically for the compliance with AWWA Standard B100 and are listed with NSF
Standard 61 as an approved filtration sand supplier and for the use of filtration systems. Commercial

gravels in sizes of 0.25~0.75" was purchased for construction of the gravel filter. The lab-scale filtration
unit was constructed with glass columns (7" in diameter and 12" long). For disinfection, 0.1% sodium
hypochlorite (NaOCI) solution was applied to filtered effluent.

Rainwater was collected periodically and stored in a refrigerator at 4C prior to use. It was
analyzed for the key water quality parameters (Table 1).
Table 1. Characteristics of rainwater.

Villarreal and Dixon

Appan (1999)

This Study

4.1(0.4)

5.8-6.2

(2005)
PH

5.2-7.9

BOD (mg/L)

7-24

COD(mg/L)

44-120

TOC(mg/L)

6-13

Turbidity (NTU)

10-56

77

4-6(5.7)

Nitrate (NO3-, mg/L)

5-7

Phosphate (PO4-,mg/L)

0.025

TS(mg/L)

60-379

SS(mg/L)

3-281

9.1(8-9)

Conductivity (uS/cm)
Total colifroms

0.7-1.3

26-28

92.0 (97.1)

(#/ioomL)
Fecal coliforms

(#/ioomL)

6.7(8.9)

2-10

Lab-scale POE CPU

Setup and Operation

According to the schematics of the POE CPU proposed (Figure 1), a lab-scale filtration and
disinfection unit was constructed as shown in Photo 1. For the sand filtration column, three different
sizes of sand were used, as shown in Table 2. Distribution and selection of sands were decided in

accordance to another small drinking water system project that the PI is working on at a field (Hwang et
al., 2009).

^XD\

from the cistern

->

to the house

Legend:
MM

\^J

pump

lead gravel filtration

3 way stop valve

lag sand filtration

sampling port
**

Treatment flow

>

Bypass flow

disinfection

""> System wash flow

Figure 1. Schematics of POE CPU.

Photo 1. A view of the installed lab-scale POE CPU.

Table 2. Filtration media used for the sand filtration unit.

Media

Size

Total Weight (kg)

Global No. SS 30/65

1.87

Sand

Effective size = 0.18 mm

Global No. SS 20/30

3.75

Sand
Effective size = 0.55 mm

Global No. SS 6/20

1.87

Sand
Effective size = 1.10 mm

For the filtration, rainwater was pumped at 0.09 gal/min which is corresponding to 0.50

gal/min/ft2 as shown below:

0.09 ^

SFR = , ^J7"7! = 0.50

0.182 ft2
mm ft2
For disinfection, a commercial liquid NaOCI solution (5.25%) was diluted to 0.1%. The solution

was pumped to the effluent from the sand filtration unit at a flowrate of 0.03154 mL/min. The target
disinfectant concentration was set at 0.5~1.0 mg/L in the effluent.
Backwash

Backwash of the sand filtration unit was conducted at a flow rate of 0.275 gal/min. The stored
disinfected filtration effluent was used for the backwash. The flowrate was corresponding to 1.51

gal/min/ft2 as shown below:

0.275 qal/min

qal

BWR = .,.. ' =1.51 "

0.182 ft2
mm ft2

Analysis
Samples were collected for analyses of physiochemical and biological characteristics of water.
For physiochemical water quality parameters, measured were total residual chlorine concentrations, pH,

specific conductivity, particle counts and turbidity. Total residual chlorine (TRC) concentrations were
monitored with HACH calorimetric method. The value of pH was monitored with Oakton pH Meter 300
series. Specific conductivity was analyzed with Orion Specific Conductivity Meter Model 162. Turbidity
was measured with LaMotte 2020 Turbidimeter. Particle counts were measured with a particle counter

(9703 Liquid Particle Counting System, Pacific Scientific, Co.).

For biological water quality parameters, measured were total coliforms (TC), fecal conforms (FC)
and total heterotrophic bacteria (THB) via a membrane filtration technique. Microbial analysis was done
within 24 hours of sampling. TC was quantified with m ENDO Broth. All colonies that are red and have
the characteristic metallic sheen are counted as TC after incubation for 24 hrs at 350.5 C. For the FC,
m FC broth was used. Tryptic soy broth was used for THB analysis. After incubation for 24 hours at

44.50.2 C, all colonies developed in blue were counted for the FC. THB analysis was done with 372hour incubation at 35C.

Results and Discussion

Physiochemical Water Quality

Filtration and Disinfection

Rainwater was pumped at a loading rate of 0.50 gal/min/ft2 to the lab-scale POE CPU for 10 mins.
Samples were taken after 10 mins. Then, the POE CPU was run for 1 hour and samples taken after 1
hour and analyzed. It should be noted that different rainwater characteristics were observed in the
cistern.

Water temperatures were measured to be 20 1C throughout the experiment. As shown in

Figure 2 and Figure 3, the values of pH were maintained at a neutral level (~7) in the system.
Conductivity in the final disinfected effluent was increased by ~2 times, but the value was still low
enough at ~53 u5/cm. Reduction in turbidity was observed despite the initial increase during the first

trial. Residual chlorine concentration was achieved in the range of 0.2~0.7 mg/L which was similar to the
targeted concentration (0.5~1.0 mg/L).

after 10 mins
after 1 hr

Rainwater

After Gravel Filter

After Sand Filter

After Disinfection Tank

80

70

after 10 mins

after 1 hr
60

E 50

40

8 30

20

10

Rainwater

After Gravel Filter

After Sand Filter

After Disinfection Tank

Figure 2. Trend of pH and conductivity during the lab-scale POE CPU run.

2.5

after 10 mins
after 1 hr

I 1.5

0.5

Rain water

After Gravel Filter

After Sand Filter

After Disinfection Tank

0.8

0.7

-Free Chlorine
-Total Chlorine
0.6

0.5
no

E.
2

0.4

0.3

0.2

0.1

after 10 mins

after 1 hr

Figure 3. Trend of turbidity and residual chlorine during the lab-scale POE CPU run.

11

Backwash

During the backwash, 2.5 inches of bed expansion was achieved and the backwash effluent
quality was analyzed (Table 3, Table 4, and Photo 2).
Table 3. Backwash duration and intensity.
Backwash Process

Duration (min)

Flow rate(gal/min)

Bed Expansion (inches)

Sand Column

0.275

2.5

Photo 2. Sand bed expansion during the backwash (left: before the backwash; right: after the backwash).

Table 4. Physiochemical water quality parameters of backwash.

Parameters

Temperature (oC)
pH

Backwash Influent*

20.0

Backwash Effluent
From Gavel Filter

From Sand Filter

23.4

21.2

7.1

7.0

7.0

Conductivity (uS/cm)

73.3

60.2

69.6

Turbidity (NTU)
Free chlorine (mg/L)
Total Chlorine (mg/L)

0.6

53.1

4.9

0.06

0.05

0.29

0.10

0.08

0.54

*: disinfected effluent from the sand filtration unit.

Biological Water Quality

As shown in Figure 4, FCwere detected at a concentration of 10 CFU/100 mL in the rainwater

collected for the 1st trial (i.e., after 10 mins). However, no FC was found in the rainwater collected and

used for the 2nd trial (i.e., after 1 hr). Differences in FC trend were due probably to the dissimilar
properties of the influent rainwater with respect to FC concentrations. Unusually high FCs were
detected from the gravel filter-backwashed water. However, no FCwas found in the sand filter
backwashed water.

18

16

14

after 10 mins

12

after lhr
10

e^

v<^

>

##

or

.<e

if

4T

*r

*F

*?

JB*

?F

Figure 4. Results of fecal coliforms analysis.

THB was also analyzed for both 1st and 2nd trials. However, the CFU was too many to count for al
the samples taken and analyzed, even with 100-time dilution of the samples. Additional run focusing on
bacterial water quality is warranted to conduct.

13

CT Values on Microorganisms Removal

Microorganism concentrations (CFU/lOOmL) were artificially increased to assess the effect of CT
values on microbial reduction. CT values are the product of disinfectant concentration (C) and contact
time (T). Leachate of chicken manure was added to the rain water for this purpose. The initial

concentrations of total coliforms and THB were measured to be 823 CFU/100 mIL and 42xl03 CFU/100
mL, respectively.

Results from the disinfection of the influent water are shown in Figure 5. When the influent was
disinfected with a chlorine concentration at 6 mg/L, smaller than 1-log removal was achieved for both

total coliforms and THB despite 60-min contact time. Also, very low concentrations (<0.2 mg/L) of free
chlorine was quantified after 20-min contact time.

6
u

"o"
u
euo
o

ba
c
ro

- 4 1

y=0.0168x

BO
V-

O
O

R2 = 0.9802

3 5
01
QJ

Total Coliforms

Total Heterotrophic Bacteria

ro

>
o

E
oj

-C-Free Chlorine

cc
bO
o

Linear (Total Coliforms)

1

Power (Free Chlorine)

-Q_
10

20

30

40

50

60

70

Time (min)
Figure 5. Effect of CTvalues on log removal of microorganisms in the influent amended with chicken manure.

However, when the effluent of the sand filter was disinfected with the same initial

concentration of chlorine, 1-log removal and 1.5-log removal of both microorganisms was achieved with
a contact time of 30 mins and 60 mins, respectively (Figure 6). Residual free chlorine concentration was

14

maintained at 3 mg/L. It should be noted that the sand filtration reduced the concentrations of total

coliforms from 823 to 400 CFU/100 mL and THB from 42xl03 to 7.4xl03 CFU/100 mL

iy= 0.0291xi
:R2- 0 9196'

y=0.0259x
,''

R2 = 0.9939

_i

QO

4
a)

/+

1.0

c
v.

v..

.c

u
QJ
OJ

Total Coliforms

A Total Heterotrophic Bacteria

-O-Free Chlorine
-

Linear (Total Coliforms)

Linear (Total Heterotrophic Bacteria)

0.0
10

20

30

40

50

60

70

Time (min)
Figure 6. Effect of CTvalues on log removal of microorganisms in the effluent from the sand filter.

Student Training
The current project started administratively in April 2009. Prior to the kick-off, the project had
focused on the training of two graduate students (Daniel Concepcion and Juan Falcon) with literature

reviews with respect to filtration, disinfection and rainwater collections. They also helped to construct
the lab-scale POE CPU. Later, two undergraduate students worked on the project (Laura Rodriguez and
Maria Sevillano).

15

Seminar at the USVIs

A seminar was given by the PI (Dr. Hwang) to the science students in the Division of Science and

Mathematics at the University of the Virgin Islands at St. Thomas on February 19, 2010. The title of the
seminar was "Reclamation and Reuse of Rainwater". The seminar addressed both conventional and

advanced technologies for rainwater reclamation and subsequent reuse. Biochemical quality of
rainwater collected and stored was emphasized. Fundamentals of engineering design and operating
parameters of sand filtration, chlorination, and ultra violet/peroxide were explained and applications of
those fundamentals to real-case scenarios were demonstrated.

Future Studies
The goal of the first-year project was to develop and evaluate the performance of a POE CPU

with respect to physiochemical and bacteriological water quality. To stretch the first-year project' goal,
the second-year project is proposed to execute, focusing on the disinfection byproducts (DBPs)

production potential in the final effluent from the POE CPU. Pathogens, such as Giardia, are often found
in source water, and can cause gastrointestinal illness and other health risks (e.g., diarrhea, cramps,
vomiting). To kill (or inactivate) these pathogenic microorganisms, water needs to be disinfected.
However, disinfectants like chlorine can react with naturally-occurring materials (NOM) in the water to

form DBPs. These byproducts include trihalomethanes (THMs), haloacetic acids, chlorite and bromate.
The second-year research will also focus on how these DBPs production can be minimized in the
system, while ensuring microbial water quality. These byproducts may lead to increased health risks as

potential carcinogens and the cause of reproductive and development defects in laboratory animals and
human, if consumed in excess of EPA's standard over many years (Hamidin et al., 2008; Wang et al.,
2007). The second-year work will encompass a series of lab-scale research on potential of DBPs
occurrence in the effluent of the POE CPU, engineering approach to minimize DBPs production in the

system, construction of a pilot-scale POE CPU, and a technical seminar by the PI to the students and
professionals in the USVIs

Results Dissemination
Results obtained from the current research were presented at a regional professional
conference as follows:

Hwang S., Concepcion D., Falcon J. "Small Filtration and Disinfection Unit for a Point-of-Entry Cistern
Water Purification", 2009 CWWA Conference, October 4-10, St. Thomas, USVI.

Acknowledgments
This research was made possible with financial support by the United States Geological Survey
104B Program through the US Virgin Islands Water Resources Research Institute. We appreciate Mrs.
Elsa Arce at the Puerto Rico Water Resources and Environmental Research Institute for her sincere

administrative supports.
16

References

Appan, A. (1999), "A dual-mode system for harnessing roofwater for non-potable uses", Urban Water, 1,
317-321.

Brin, G., Bossi, L, Buscemi, P., Chesiek, H., Gangemi, A., Milu, M., Rose, D., Stevens, C "Water resources

and wetlands for Haiti and the Virgin Islands," MITfinal report June 2003;
(ceemeng.mit.edu/finalproposal.doc: accessed in Nov. 2008)

Chang, M., McBroom, M.W., Beasley, R.S. (2004), "Roofing as a source of nonpoint water pollution",
Journal of Environmental Management, 73, 307-315.
Evans, C.A., Coombes, P.J., Dunstan, R.H. (2006), "Wind, rain and bacteria: The effect of weather on the
microbial composition of roof-harvested rainwater", Water Research, 40, 37-44.
Hamidin, N., Yu, Q.J., Connell, D.W. (2008), "Human health risk assessment of chlorinated disinfection

by-products in drinking water using a probabilistic approach", Water Research, 42 (13), 32633274.

Hwang, S., Concepcion, D., Fonseca, A., Falcon, J., Sinha, R. "Evaluation of Small Drum Filtration and

Disinfection Units for Small, Rural Communities", Proceedings of 2009 American Water Works
Association Annual Convention, San Diego, CA, June 14-18, 2009.

Isquith, I.R., Winters, H. (1987), "The microbial analysis of domestic cistern water in the U.S. Virgin
Islands", Desalination, 66, 415-421.

Simmons, G., Hope, V., Lewis, G.,Whitmore, J., Gao, W. (2001), "Contamination of potable roof-collected

rainwater in Auckland, New Zealand", Water Research, 35(6), 1518-1524.

Sobsey, M., Stauber, C, Casanova, L, Brown, J., Elliott, M. (2008). "Point of Use Household Drinking
Water Filtration: A Practical, Effective Solution for Providing Sustained Access to Safe Drinking
Water in the Developing World", Environ. Sci. technol., 42(12), 4261-4267.

Villarreal, E.L, Dixon, A. (2005), "Analysis of a rainwater collection system for domestic water supply in
Ringdansen, Norrkoping, Sweden", Building and Environment, 40,1174-1184.
Wang, W., Ye, B., Yang, L, Li, Y., Wang, Y. (2007). "Risk assessment on disinfection by-products of
drinking water of different water sources and disinfection processes", Environment
International, 33(2), 219-225.

WHO/UNICEF. Meeting the MGD drinking water and sanitation target: the urban and rural challenge of
the decade. Geneva: The World Health Organization and the United Nations Children's Fund.
2006.

17